Groundwater Studies UNESCO

krus-couv-avec-dos.
qxd 27/11/2003 11:35 Page 1
G ro u n d w a t e r
Groundwater studies
studies
An international guide
for
hydrogeological investigations
Edited by
V. S. Kovalevsky,
G. P. Kruseman,
K. R. Rushton
IHP-VI, SERIES ON GROUNDWATER NO. 3

kruseman-00.qxd 27/11/2003 11:19 Page 1
Groundwater
studies
An international guide
for
hydrogeological investigations
IHP-VI, SERIES ON GROUNDWATER NO. 3

The designations employed and the presentation

of material throughout the publication do not
imply the expression of any opinion whatsoever
on the part of UNESCO concerning the legal
status of any country, territory, city or area or of
its authorities, or the delineation of its frontiers
or boundaries.
Published in 2004 by the United Nations

Educational, Scientific and Cultural Organization
7, place de Fontenoy, 75352 Paris 07 SP
Layout and typesetting by Marina Rubio 93200 Saint-Denis

Printed by UNESCO
ISBN 92-9220-005-4
UNESCO 2004
Preface
Understanding of groundwater has developed significantly since 1972 when the first part of the original
volume of Groundwater Studies was published by UNESCO. Yet for someone who is just commencing the
study of groundwater, there is still a need for a text which will help in starting their work. For those with
a greater experience in hydrogeological investigations, there is a need to increase awareness both of
more recent work and information about techniques which are outside their previous experience. This
new volume is intended to meet both of these needs and therefore it has the subtitle An International
Guide for Hydrogeological Investigations.
This document was prepared in the framework of the Fourth Phase of the International
Programme as Project M-1-3 that was supervised and directed by Habib Zebidi, Water Science Specialist,
Division of Water Sciences, UNESCO and after his retirement in 1999 by his successor Ms Alice Aureli.
Dr Habib Zebidi has drawn contributors and members for an editing committee from different
countries. They brought to this volume their own distinctive perspectives, for all are acknowledged
experts with extensive experience of practical groundwater issues. The members of the editing
committee have provided a certain unity of style end presentation and also they have tried to keep the
size of this document in hand. Each chapter is intended to provide sufficient information to comprehend
the fundamentals of the topic; in addition reference is made to publications where further information
can be obtained for more detailed study. The contributing authors are listed below per chapter and
details are given in Appendix A-3.
Authors
Chapter 1
Ken R. Rushton (United Kingdom) and Gideon P. Kruseman (The Netherlands)
Chapter 2
Jasper Griffioen (The Netherlands)
Chapter 3
Henny A. J. van Lanen (The Netherlands)
Chapter 4
Frank D. I. Hodgson (South Africa)
Chapter 5
Robert Becht (The Netherlands)
Chapter 6
Robert Sporry (The Netherlands)
Chapter 7
Aldo da C. Rebouas (Brasil)
Chapter 8
Jasper Griffioen (The Netherlands) and Gideon P. Kruseman (The Netherlands)
Chapter 9
P. K. Aggarwal (Austria), K. Froehlich (Austria), K. M. Kulkarni (India) and
E. Garcia-Agudo (Austria)
Chapter 10
Jean Margat (France)
Chapter 11
Ken R. Rushton (United Kingdom) and Vlademir S. Kovalevsky (Russia)
Chapter 12
Jean Margat (France)
Chapter 13
Vlademir S. Kovalevsky (Russia) and J. Vrba (Czech Republic)
Chapter 14
Philip E. LaMoreaux (United States of America) and Yuan Daoxian (China)
Chapter 15
S. D. Limaye (India)
Chapter 16
Emilio Custodio (Spain)
Acknowledgements
The contributing authors and the editing committee like to thank the UNESCO Division of Water Science
for their patience, advice and support during the long preparation of this document.
The editors are very grateful to Dr H. Speelman, Director of TNO Netherlands Institute of Applied
Geosciences for the the processing of the figures, for the funding of the correction of the text on
the proper use of the English language and for making funds available to assist G. P. Kruseman in his
editing work.
Contents
Chapters
1 Occurrence of groundwater, regime and dynamics 19

1.1 Hydrological cycle 19
1.2 Groundwater flow 21
1.2.1 Groundwater flow processes and continuity 21
1.2.2 Groundwater head 21
1.2.3 Darcys Law, hydraulic conductivity and transmissivity 23
1.2.4 Inflows and outflows 23
1.2.5 Storage coefficients and time dependency 24
1.3 Groundwater composition 24
1.3.1 Physical and chemical properties 24
1.3.2 Risk of groundwater pollution 25
1.4 Groundwater assessment and exploration 29
1.4.1 Aquifer and groundwater systems 29
1.4.2 Data collection 32
1.5 Groundwater exploitation and management 36
1.5.1 Monitoring conditions within the aquifer system 36
1.5.2 Water balances and simulation models 38
1.5.3 Potential consequences of changed groundwater conditions 39
1.5.4 Cost of investigations and groundwater development 40
1.5.4 Management of groundwater 40
1.6 References and additional reading 41
2 Groundwater quality 43
2.1 Introduction 43
2.1.1 Definition and scope of chemical hydrogeology 43
2.1.2 General approach 43
2.2 Basic principles of groundwater chemistry 46
2.2.1 Precipitation/dissolution reactions 46
2.2.2 Redox reactions 46
2.2.3 Sorption reactions 47
2.2.4 Aqueous complexing 48
2.2.5 Gas transfer 48
2.2.6 Ion filtration and osmosis 49
2.2.7 Radioactive decay 49
2.3 Acquisition of chemical data 49
2.3.1 Sampling procedures 49
2.3.2 Groundwater analysis 50
2.4 Evaluation of groundwater chemical data 51
2.4.1 General procedure 51

5
2.4.2 Characterisation of groundwater quality 51

2.4.3 Compiling groundwater quality data 55
2.5 Process interpretation and modelling 55
2.5.1 Modelling 55
2.5.2 Coupling of hydrochemical reactions 59
2.5.3 Redox zoning 59
2.6 Groundwater supply and health 61
3 Groundwater networks and observation methods 67

3.1 Introduction 67
3.2 Place of monitoring networks in groundwater management 68
3.2.1 Background monitoring networks 68
3.2.2 Specific monitoring networks 68
3.3 Hydrological variables 70
3.4 Network design 71
3.4.1 Network density 72
3.4.2 Sampling frequency 78
3.4.3 Simultaneous design of network density andsampling frequency 82
3.5 Groundwater information system 82
3.6 Observation methods 85
3.6.1 Observation wells and piezometers 85
3.6.2 Methods for monitoring groundwater quantity 86
3.6.3 Methods for monitoring groundwater quality 89
3.7 Concluding remarks 92
4 Processing and presentation of data 95

4.1 Scope
4.2 Data types and presentation possibilities 95
4.2.1 Point data 96
4.2.2 Spatial information 96
4.2.3 Presentation possibilities 96
4.3 Graphic processing and presentation of data 96
4.3.1 Borehole information 97
4.3.2 Hydraulic properties of the aquifer 98
4.3.3 Time-dependent data 98
4.3.4 Specialised hydrochemical diagrams 99
4.3.5 Other graphics 106
4.4 Statistical processing and presentation of data 108
4.4.1 Basic statistics 108
4.4.2 Time series 110
4.5 Spatial information systems 113
4.5.1 Introduction 113
4.5.2 Entry of spatial information 114
4.5.3 Conversion of point data 114
4.5.4 Hydrogeological maps and the UNESCO code 116
4.6 Conclusions 118

6
5 Remote sensing techniques for groundwater prospection 121

5.1 Introduction 121
5.1.1 General 121
5.1.2 Techniques and approaches 121
5.2 Principles of remote sensing and GIS 123
5.2.1 Remote sensing 123
5.2.2 Physical fundamentals 123
5.2.3 Aerospace imaging systems 124
5.2.4 Frequency domains 126
5.2.5 Imaging satellites 126
5.2.6 Digital image processing 126
5.3 Geographic information systems (GIS) 128
5.4 Required resources 130
6 Geophysical techniques in groundwater investigations 133

6.2 Magnetic methods 133
6.2.1 The source of the earths magnetic field 133
6.2.2 Data processing and interpretation 136
6.2.3 A case-study of aeromagnetic surveying
for aquifer exploration 140
6.3 Electromagnetic (EM) methods 140
6.3.1 Principles and survey techniques 141
6.4 Resistivity method 150
6.4.1 Resistivity of rocks 150
6.4.2 Resistivity measurements 151
6.5 Induction polarisation 156
6.5.1 Principles and procedures of IP 156
6.6 Seismic method 160
6.6.1 Principles and instrumentation 160
6.6.2 Reflection and refraction of seismic waves 161
6.6.3 Recent developments 166
6.7 Gravity surveying 167
6.7.1 Principles and field procedures 167
6.7.2 Interpretation of gravity anomalies 169
6.7.3 Applications to groundwater exploration 169
6.8 Ground penetrating radar 172
6.9 Geophysical borehole logging 174
6.9.1 Principles and instrumentation 174
6.9.2 Logging physical parameters 176
7 Well drilling and design methods 185

7.2 The development of the well drilling techniques 185
7.3. Water quality protection for wells 186
7.4 Standards for water and monitoring wells 187

7
7.5 Well drilling techniques 187

7.5.1 Cable tool method and variations 187
7.5.2 Direct circulation rotary drilling and variations 191
7.5.3 Reverse circulation rotary drills 192
7.5.4 Air drilling systems 194
7.6 Auger hole drilling 196
7.6.1 Hand augers 196
7.6.2 Solid-stem augers 197
7.6.3 Hollow-stem augers 198
7.7. Well design and construction 199
7.7.1 General 199
7.7.2 Particulars of design 200
7.7.3 Main components of a water well design 201
7.8 Well design in deep confined aquifer systems 210
7.8.1 Hydrogeological conditions 210
7.8.2 Typical well drilling procedures 212
8 Determining hydrodynamic and contaminant transfer

parameters of groundwater flow 217
8.2 Parameters of flow in homogeneous aquifers 217
8.3 Determination of hydraulic characteristics by pumping tests 217
8.3.1 The pumping test 217
8.3.2 The well 219
8.3.3 The piezometers 220
8.3.4 The measurements 221
8.4 Analysis of pumping test data on homogeneous
and isotropic aquifers 223
8.4.1 Data analysis 223
8.4.2 The well flow formula for confined aquifers 223
8.4.3 Theiss curve-fitting method 225
8.4.4 Jacobs straight line method 225
8.4.5 Recovery analysis 227
8.4.6 Well flow formula for other conditions 228
8.5 Determination of hydraulic characteristics by other methods 229
8.5.1 Single-well tests 229
8.5.2 Flowing well tests 229
8.5.3 Slug tests 229
8.5.4 Tidal movements 229
8.5.5 Lithology and grain size analysis 231
8.5.6 Laboratory tests 231
8.5.7 Water balance and groundwater flow simulation model 231
8.6 Determination of hydraulic characteristics of aquifers
in fractured rocks 232
8.6.1 Double-porosity model 232
8.6.2 Single vertical fractures 233
8.6.3 Single vertical dikes 233
8.7 Determination of hydraulic characteristics of volcanic rocks 235
8.8 Determination of hydraulic characteristics of limestones 235
8.9 Diffusion, dispersion and macrodispersion 236

8
9 Nuclear techniques in groundwater investigations 239

9.2 Environmental isotopes 239
9.2.1 Stable isotopes 240
9.2.2 Radioactive isotopes 243
9.2.3 Applications in groundwater studies 247
9.2.4 General remarks on environmental isotopes 254
9.3 Artificial isotopes 255
9.3.1 Radioactive tracers 255
9.3.2 Techniques 256
9.3.3 Applications 257
9.3.4 Practical considerations 261
10 Hydrogeological mapping 271

10.2 The role and place of hydrogeological mapping 271
10.2.1 Maps among other methods of storing and representing
information 272
10.2.2 Role of mapping in the development of hydrogeological
insight into an area 273
10.3 What can and must hydrogeological maps depict? 274
10.3.1 Inventory 274
10.3.2 Data of field surveys and subsurface investigations 276
10.3.3 Shapes and dimensions 276
10.3.4 Mapping programme 277
10.3.5 Hydrogeological map sensu stricto 277
10.3.6 Presentation of surface water 278
10.3.7 Specific purpose maps 278
10.4 Classification of hydrogeological maps 278
10.4.1 Programme and purposes 279
10.4.2 Scale 280
10.4.3 Scientific maps and practical maps 280
10.4.4 Terminology 281
10.5 Hydrogeological map making 282
10.5.1 The language of maps; properties and constraints 282
10.5.2 Key 283
10.6 Producing hydrogeological maps 285
10.6.1 Preliminary choice 285
10.6.2 Programming 285
10.6.3 Explanatory note 286
11 Assessment of groundwater resources and groundwater

regime forecasting 289
11.2 Identifying and quantifying groundwater resources 290
11.2.1 Formulation 290
11.2.2 Groundwater resource components 296
11.2.3 Examples of identification and formulation of aquifer
flow mechanisms 299

9
11.2.4 Important questions when identifying resources 301

11.2.5 Selection of the co-ordinate system 302
11.2.6 Preliminary flow balances 302
11.3 Use of models for quantifying groundwater resources 303
11.3.1 Fundamentals of groundwater modelling 303
11.3.2 Various types of models 304
11.3.3 Important issues in developing models 304
11.3.4 Description of models used for case studies 307
11.4 Using forecasting to identify safe yields 308
11.4.1 Forecasting and prediction 308
11.4.2 Forecasting 308
11.4.3 The use of models for predictions 313
11.4.4 Methodologies in the use of models for predictive purposes 314
11.4.5 Examples of predictions 314
11.5 Concluding remarks 316
12 Groundwater management 319

12.1.1 Scope 319
12.1.2 Groundwater resource management preliminary remarks 319
12.2 Physical conditions of groundwater management 321
12.2.1 Flow management and storage management 321
12.2.2 Overall exploitation strategies of an aquifer 321
12.2.3 Types of aquifer systems and management conditions 325
12.3. Socio-economic conditions management actors and objectives 325
12.3.1 Management actors 325
12.3.2 Management levels
12.3.3 Management objectives 358
12.4 Management constraints and criteria 329
12.4.1 Internal constraints 329
12.4.2 External constraints 330
12.5 Management decision methods and aids 330
12.5.1 Modelling and predictive management 330
12.5.2 Forecasting unit production costs 331
12.5.3 Forecasting external costs 331
12.5.4 Optimisation methods 331
12.5.5 Management control 332
12.6 Management instruments 333
12.6.1 Regulation 333
12.6.2 Financial incentives 334
12.7 The future of groundwater management 334
12.7.1 Towards groundwater management implementation 334
12.7.2 A more integrated management 335
12.7.3 A more ambitious form of management 335
13 (A) The influence of changes in hydrogeological conditions

on the environment and (B) Groundwater quality protection 339
13A.1 Introduction 339
13A.2 The environmental impact of groundwater withdrawal 340

10
13A.2.1 Effects on the relation between groundwater and surface water 340
13A.2.2 Land subsidence 340
13A.2.3 The influence on karstification 341
13A.2.4 Effects on plants and animal life 342
13A.2.5 Influence on seismicity 342
13A.3 The impact of man-induced groundwater level rise on the environment 343
13A.3.1 General 343
13A.3.2 Effects of waterlogging 343
13A.3.3 Effects on agricultural lands 344
13A.3.4 Effects of surface water reservoirs 344
13A.3.5 Effects on karstification 345
13A.3.6 Biological effects 345
13A.3.7 Effects on seismicity 345
13A.3.8 Effects in urban areas 345
13A.4 References and additional reading 346
13B.1 Introduction 347
13B.2 Groundwater quality protection strategy 347
13B.3 Groundwater quality protection policy 348
13B.4 Groundwater protection management 350
13B.4.1 General protection of groundwater 350
13B.4.2 Comprehensive groundwater protection 350
13B.4.3 Delineation of groundwater protection zones 351
13B.5 Groundwater pollution control 352
13B.5.1 Influence of natural processes and human impacts 352
13B.5.2 Point pollution control of groundwater 355
13B.5.3 Non-point pollution of groundwater 356
13B.5.4 Non-point pollution control 356
13B.5.5 Impact of groundwater on human health 358
13B.6 References and additional reading 360
14 Hydrogeology of carbonate rocks 363

14.2 Factors determining the occurrence of groundwater
in carbonate rock 365
14.2.1 Structure 365
14.2.2 Fracture systems 366
14.2.3 Joints 366
14.3 Hydrogeological features of carbonate rocks 368
14.3.1 Porosity 368
14.3.2 Permeability 369
14.3.3 Groundwater flow 369
14.4 Examples of groundwater flow systems in carbonate rocks 370
14.4.1 Karst hydrological systems 370
14.4.2 Characteristics of karst hydrological systems 372
14.5 Hydrochemical character of carbonate rock aquifer 377
14.5.1 Important hydrochemical features of carbonate rocks 378
14.5.2 Environmental aspects and recommended references 378
15 Hydrogeology of hard rocks 383


11
15.2 Occurrence of groundwater 384

15.3 Groundwater development 385
15.4 Types of wells 386
15.5 Drinking water supply 389
15.6 Exploration 391
15.7 Recharge augmentation 392
15.8 Sustainability and pumpage control 393
16 Hydrogeology of volcanic rocks 395

16.2 Volcanic rocks and formations 395
16.3 Hydrogeological properties of volcanic formations 399
16.4 Groundwater flow in volcanic formations 407
16.5 Hydrogeochemistry and mass transport in volcanic formations 711
16.6 Groundwater quality issues in volcanic formations 416
16.7 Groundwater exploitation in volcanic formations 418
16.8 Groundwater balance in volcanic formations 420
16.9 Groundwater monitoring in volcanic formations 421
16.10 Geothermal effects in volcanic formations 423
Appendices
Appendix A-1 Conversion factors for physical data 427
Appendix A-2 Conversion factors for hydrochemical data 428
Appendix A-3 List of authors 429
Tables
Table 1.1 Overview of the types of (bio)geochemical reactions that control the fate
of contaminants in groundwater 28
Table 2.1 Schematic flow chart for a chemical hydrogeological project 44
Table 2.2 Nomenclature for water 52
Table 2.3 Normal range of chemical composition of groundwater, seawater and rainwater
away from the coast 52
Table 2.4 Sources of major constituents 53
Table 2.5 Sources of minor constituents 54
Table 2.6 Sources of trace constituents 54
Table 2.7 Concentration ratios and their meaning in terms of process control 56
Table 2.8 Redox classification of groundwater 60
Table 2.9 Guidelines for remediation of groundwater withdrawal systems 62
Table 2.10 WHO guideline values for species in drinking water (1993) 65
Table 3.1 Summary of instruments commonly used to measure groundwater heads 86
Table 3.2 Accuracy of various methods to measure groundwater heads 88
Table 4.1 Correlations between chemical constituents in water from a mining environment 110
Table 6.1 Magnetic susceptibilities of minerals, ores, and rocks 134
Table 6.2 Resistivities of some rock types and fluids 150
Table 6.3 Elastic wave velocities in some media 161
Table 6.4 Density values of some common rocks 168
Table 7.1 Advantages and disadvantages of cable tool method 189
Table 7.2 Major types of drilling fluids used in the water well industry 192

12
Table 7.3 Advantages and disadvantages of the direct circulation rotary drilling method 193
Table 7.4 Advantages and disadvantages of the reverse circulation rotary method 193
Table 7.5 Advantages and disadvantages of the air rotary drilling systems 196
Table 7.6 Major advantages and disadvantages of the auger techniques 198
Table 7.7 Recommended diameters for pump chamber and permanent
surface casing 204
Table 7.8 Maximum discharge rates for optimum diameter of riser pipe 206
Table 7.9 Open areas for some representative screens 209
Table 8.1 Preferred distance between pumping well and piezometer 220
Table 8.2 Frequency of water level measurements in the pumped well 222
Table 8.3 Frequency of water level measurements in piezometers 222
Table 8.4 Range of variation of the specific yield 231
Table 9.1 Environmental stable isotopes used in groundwater studies 240
Table 9.2 Cosmogenic and anthropogenic radioisotopes used in hydrology 243
Table 9.3 Radioisotope tracers commonly used in groundwater investigations 256
Table 10.1 Mappable subjects 275
Table 10.2 Clustering of mappable objects 276
Table 10.3 Users and information requirements 279
Table 10.4 Classification of hydrogeological maps 280
Table 10.5 System for classifying hydrogeological maps 281
Table 11.1 Approximate water balance for Miliolite Limestone aquifer 303
Table 11.2 Different types of groundwater models 305
Table 12.1 Advantages and disadvantages of groundwater as a source of supply 320
Table 12.2 Types of aquifer system and management conditions 326
Table 12.3 Classification of the actors in groundwater development,
explotation and management 327
Table 13B.1 Classes of groundwater in the United States of America 349
Table 13B.2 NO3 content in groundwater 352
Table 13B.3 Sources of groundwater pollution 354
Table 14.1 Classification of voids and porosity 368
Table 14.2 Important characteristics of carbonate rocks 373
Table 16.1 Common porosity values of volcanic formations 403
Table 16.2 Common ranges of permeability for water 404
Figures
Fig. 1.1a The hydrological cycle 19
Fig. 1.1b Flow chart of the hydrological cycle 20
Fig. 1.2 Significance of groundwater head 22
Fig. 1.3 Description of storage effects 25
Fig. 1.4 Aquifer types 30
Fig. 1.5 Porosity systems 31
Fig. 1.6 Fractured rock models 31
Fig. 1.7 Aquifer boundary conditions 34
Fig. 1.8 Unreliability of open boreholes 37
Fig. 1.9 Unreliability of open boreholes 38
Fig. 2.1 Factors controlling groundwater composition 45
Fig. 2.2 Illustration of the principles of forward and inverse modelling 57
Fig. 2.3 PH/redox potential diagram with redox clines for the major redox
couples in groundwater 61
Fig. 2.4 Exposure route in groundwater of dissolved species harmful to man 64
Fig. 3.1 General layout of the groundwater monitoring procedure 69
Fig. 3.2 Adaptation of the first version of a specific monitoring network 70

13
Fig. 3.3 Hydrological variables to be monitored 72

Fig. 3.4 Measured and simulated hydrological variables 73
Fig. 3.5 Procedure for the determination of network density and sampling frequency 74
Fig. 3.6 Experimental (dots) and theoretical semi-variogram (solid line) 75
Fig. 3.7 Network density graphs 76
Fig. 3.8 Network density derived from allowable standard deviation
of the interpolation error (SDIE) 77
Fig. 3.9 Step trends in the groundwater heads 78
Fig. 3.10 Nitrate concentrations in the groundwater outflow of the Noor catchment 79
Fig. 3.11 Periodic fluctuations and a linear trend in the groundwater heads in Gujarat, Western
India 80
Fig. 3.12 Difference between groundwater heads measured with different sampling
frequencies in a observation well in the Hupselse Beek catchment (The Netherlands) 81
Fig. 3.13 Outline of a groundwater information system 83
Fig. 3.14 Comparison of some measured ion concentrations with conventional analytical
techniques and with a set of integrated ion-selective electrodes HYDRION-10 92
Fig. 4.1 Composite hydrogeological borehole log 98
Fig. 4.2 Time dependent plot of water levels, electrical conductivity and rainfall 99
Fig. 4.3 Plotting procedures for the Piper diagram 100
Fig. 4.4 Plotting procedures for the Expanded Durov diagram 101
Fig. 4.5 Piper plot of groundwater chemistries from a mining environment 102
Fig. 4.6 Durov plot plot of groundwater chemistries from a mining environment 103
Fig. 4.7 SAR plot of groundwater chemistries from a mining environment 104
Fig. 4.8 Expanded Durov plot of groundwater chemistries from a mining environment 105
Fig. 4.9 Stiff plots of water chemistries 106
Fig. 4.10 Pie diagrams of water chemistries 107
Fig. 4.11 Chernoff faces of water chemistries 107
Fig. 4.12 Stacked bar charts of water chemistries 108
Fig. 4.13 Box and whisker plot of groundwater chemistries from a mining environment 109
Fig. 4.14 Box and whisker plot of groundwater chemistries from a mining environment 109
Fig. 4.15 Histogram of groundwater pH levels from a mining environment 110
Fig. 4.16 Bivariate histogram of groundwater pH and sulphate from a mining environment 111
Fig. 4.17 Fifth order polynomial regression interpolation of groundwater levels 111
Fig. 4.18 Multiple linear regression showing actual and simulated water levels 112
Fig. 4.19 Factor analysis of groundwater chemistries from a mining environment 113
Fig. 4.20 Smoothed water-level contours, using a 5,000 m search radius 115
Fig. 4.21 Smoothed water-level contours, using a 500 m search radius 115
Fig. 4.22 Three-dimensional presentation of groundwater levels 117
Fig. 4.23 Three-dimensional presentation of groundwater contours at metre intervals 117
Fig. 4.24 Grid presentation for evaluation of averaged groundwater levels 118
Fig. 5.1 Electromagnetic spectrum 124
Fig. 5.2 Interaction between EM radiation and matter 124
Fig. 5.3 Path and interactions between sun and sensor 125
Fig. 5.4 Satellite characteristics 127
Fig. 5.5 From scanning to display 128
Fig. 5.6 Raster and vector data structures 129
Fig. 6.1 Quantitative interpretation of the magnetic anomaly over a normal
fault in a basalt sheet 137
Fig. 6.2 Contour map of magnetic anomalies 138
Fig. 6.3 (a) Shaded relief aeromagnetic image of a granite in the Yilgarn Block
of Western Australia, (b) Interpretation of the fracture 139
Fig. 6.4 Magnetic anomalies of the Palla Road area, eastern Botswana 141
Fig. 6.5 Interpretation of Figure 6.4 142

14
Fig. 6.6 Principle of the Slingram EM technique 143

Fig. 6.7 Principle of time domain EM 145
Fig. 6.8 Example of a frequency domain EM profile 147
Fig. 6.9 Example of the interpretation of airborne VLF data 148
Fig. 6.10 Results of a single frequency horizontal loop EM survey 149
Fig. 6.11a Principle of the resistivity method, using two current electrodes
and two potential electrodes 151
Fig. 6.11b, Layout of different electrode arrays: Wenner, Schlumberger
c, d and Dipole-Dipole 152
Fig. 6.12 Example of a VES in the Wenner array 154
Fig. 6.13 Example of resistivity profiling 155
Fig. 6.14a Correlation of hydraulic transmissivity values with normalised
transverse resistance TN 156
Fig. 6.14b Hydraulic transmissivity contour map 157
Fig. 6.15 Example of a continuous depth sounding survey by electrode
array scanning 158
Fig. 6.16 Principle of measuring the chargeability in the time-domain IP technique 159
Fig. 6.17a,b In a layered medium three different waves can be observed 162
Fig. 6.18 Example of a seismograph recording with a 24-channel system 164
Fig. 6.19 Example of a refraction seismic profile across a valley 165
Fig. 6.20 Principle of the CDP technique in reflection seismics 166
Fig. 6.21 Example of a high resolution seismic reflection section after full
processing of the data, as an application in groundwater surveying 167
Fig. 6.22 Example of the combined effect on a gravity anomaly by a shallow
and a deep structure 170
Fig. 6.23 Gravimetry used for a structural study of a sedimentary basin 171
Fig. 6.24 Careful surveying allows the detection of faulted offset of the
basement-sediment interface 173
Fig. 6.25 The sounding mode for GPR 173
Fig. 6.26 Example of a GPR section of push moraine deposits 174
Fig. 6.27 Basic components of geophysical borehole logging 175
Fig. 6.28 Electrode configurations for single point resistance logging, normal
resistivity logging and latero logging 177
Fig. 6.29 Example of a borehole log as used in groundwater surveys 178
Fig. 7.1 Typical truck-mounted cable tool equipment for drilling wells 188
Fig, 7.2 Major components of a rotary drilling rig 190
Fig. 7.3 Diagrams of a direct (a) and a reverse (b) rotary circulation system 191
Fig. 7.4 Basic components of an air rotary drilling system 194
Fig. 7.5 Guide for the use of bit types in air drilling system 195
Fig. 7.6 Diagram of a hand auger 197
Fig. 7.7 Diagram of a solid-stem auger 197
Fig. 7.8 Diagram of a hollow-stem auger 199
Fig. 7.9 Examples of well construction 202
Fig. 7.10 Diagram showing the main parts of a deep water well design 203
Fig. 7.11 Fence diagram of the Paran sedimentary basin 211
Fig. 7.12 Typical water well designs to tap the confined Botucatu aquifer 213
Fig. 8.1 Drawdown in a pumped confined aquifer 218
Fig. 8.2 Well penetration 219
Fig. 8.3 Cluster of piezometers in a heterogeneous aquifer intercalated
with aquitards 221
Fig. 8.4 Log-log and semi-log curves of drawdown versus time 224
Fig. 8.5 Analysis of data from pumping test with Theiss curve-fitting method 226
Fig. 8.6 Analysis of piezometer data with Jacob straight line method 227

15
Fig. 8.7 Time drawdown and residual drawdown 228

Fig. 8.8 Geohydrological conditions to which the formula of Bosch may be applied 230
Fig. 8.9 Porosity systems 232
Fig. 8.10 Semi-log time-drawdown plot for an observation well in a fractured
rock formation of the double-porosity type 233
Fig. 8.11 A well that intersects a single plane fracture of finite length and
infinite hydraulic conductivity 234
Fig. 8.12 Composite dike/aquifer system 235
Fig. 8.13 The principal processes of hydrodynamic dispersion and
macrodispersion 237
Fig. 9.1 The annual mean 18O of precipitation as a function of the mean
annual air temperature at surface 242
Fig. 9.2 Long-term tritium concentration in precipitation 244
Fig. 9.3 Carbon-14 distribution in atmospheric CO2 246
Fig. 9.4 Tracer probe for dilution and direction logging 257
Fig. 9.5 Schematic diagram showing the tracer concentration changes
with time in a borehole dilution experiment 259
Fig. 9.6 Arrangement of a set of detectors for determining vertical flow in a borehole 261
Fig. 9.7 Diagram of water flow direction in a borehole0 262
Fig. 9.8 Unconfined porous aquifers. Relation of hydraulic conductivities
determined by tracer dilution techniques and by pumping tests 263
Fig. 10.1 Spatial graphics 272
Fig. 10.2 Relation between the tools and the modes of expression 273
Fig. 10.3 Flow chart of the production of a hydrogeological map 286
Fig. 11.1a, Example of weathered-fractured aquifer in India: (a) pumped
b, and c and observation wells, (b) general form of drawdowns
291for pumping test, (c) numerical model to represent aquifer system 291
Fig. 11.1d Example of weathered-fractured aquifer in India,
and e (d) detailed comparison between field and modelled results,
(e) main flow mechanisms 292
Fig. 11.2a Example of alluvial aquifer: (a) simplified diagram of aquifer system,
and b (b) history of exploitation of the aquifer system 293
Fig. 11.2c Example of alluvial aquifer: (c) response of shallow and deep observation wells,
and d (d) change of flow mechanisms from before to after exploitation 294
Fig. 11.2e Simplified model of the aquifer system 295
Fig. 11.2f Comparison of field and modelled groundwater head fluctuations 296
Fig. 11.3 Example of Miliolite limestone aquifer 297
Fig. 11.4 River/aquifer interaction 300
Fig. 11.5 Examples of forecasting 310
Fig. 12.1 The roles of flow and storage in groundwater management 321
Fig. 12.2 Exploitation under a dynamic balance strategy 322
Fig. 12.3 Exploitation under a prolonged imbalance strategy 323
Fig. 12.4 Mining of stored groundwater storage, under a depletion strategy 324
Fig. 13B.1 Impact of nitrate pollution on groundwater in the Czech Republic 357
Fig. 13B.2 Changes in hydrogeological profiles in a shallow aquifer 358
Fig. 14.1 Location map of the major outcrops of carbonate rocks 364
Fig. 14.2 Vertical section through limestone outcrops showing diminution of fissure
widths with depth 367
Fig. 14.3 Porosity types in soluble carbonate rocks 369
Fig. 14.4 The relationship between stratigraphy, folding pattern, relief
and karst hydrological system 371
Fig. 14.5 The hydrological, hydrochemical and isotopic response to a storm 371

16
Fig. 14.6 Components of water flow and monitoring facilities in Yaji karst hydrological system,
Guili, China 376
Fig. 14.7 Hydrograph of spring S31 in Yaji karst system, Guilin, China 377
Fig. 14.8 Correlation between precipitation and discharge of karst spring 375
Fig. 15.1 Dug wells with horizontal and vertical bores 388
Fig. 15.2 Effect of natural drainage on a village well 390
Fig. 16.1 Structure of volcanic formations affecting groundwater 399
Fig. 16.2 Simplified cross section through island volcanic formations 400
Fig. 16.3 Schematic internal structure of a volcanic massif conditioning
the permeability distribution 402
Fig. 16.4 Two modes of interpreting groundwater in the volvanic island of Oahu 406
Fig. 16.5 Cross section along the Fluvi river valley, Catalonia, Spain 407
Fig. 16.6 Schematic, highly idealised cross sections through volcanic formations 408
Fig. 16.7 Schematic cross section through volcanic islands of the high type 410
Fig. 16.8 Experimental water gallery in the Famara massif 412
Fig. 16.9 Relationship between rock and water composition 414
Fig. 16.10 Water isotopes applied to volcanic areas 416
17
1 O c c u r r e n c e o f g ro u n d w a t e r ,
regime and dynamics
1.1 Hydrological cycle
Most water on our planet occurs as saline water in the oceans and deep underground or is
contained in polar ice caps and the permanent ice cover of the high mountain ranges. So, only
30 million km3 of fresh water, that is only 2 percent of all water, plays an active part in the
hydrological cycle and in the maintenance of all life on the continents.
The hydrological cycle (Figs 1.1a and 1.1b) depicts how part of the ocean water evap-
orates, the water vapour turns into fresh water precipitation (rain, hail, snow) on the earths
surface (seas, land), then flows over the land surface (glaciers, runoff, streams) and partly
infiltrates into the soil (soil water) to be used by the vegetation (evapotranspiration), or to
recharge the groundwater bodies. Subsequently, most groundwater returns either by being
pumped or by natural outflow, to surface water bodies which subsequently discharge back into
the sea.
Figure1.1a The hydrological cycle (After De Wiest,1965)

19
Groundwater studies
Figure1.1b Flow chart of the hydrological cycle (After Freeze and Cherry, 1979)
Water on land masses is always in motion, either moving quickly (vapour transport,
precipitation, surface flow) or slowly (groundwater flow, glaciers). The slowness of groundwater
flow means that most fresh water is in the form of groundwater. Consequently, groundwater is
the main storage reservoir of fresh water, while surface water can be considered as the surplus
precipitation that could not infiltrate, or that has been rejected as overflow from the ground-
water reservoir (springs and other outflows).
When water enters the soil (infiltration) it becomes soil water; soil water may not com-
pletely fill the pores between the soil particles. Thus the zone through which the water moves is
unsaturated (unsaturated or vadose zone). The flow through the unsaturated zone is essentially
vertical. At the top of the vadose zone the vertical flow may be downward under the influence
of gravity, or it may be upward (capillary flow) resulting from the evapotranspiration processes.
To a depth of 5 m below the ground all unsaturated flow is downward (deep-rooting trees and
shrubs, e.g. salt cedar, may still abstract water from the downward flow).
When vertical flow is impeded by an impervious layer, the pores above this layer fill
completely and become saturated. The voids filled by the groundwater may have different
origins and may occupy a smaller or larger part of the gross volume. The porosity is the per-
centage of a gross volume of soil or rock that is filled by air or water; the soil or rock containing
such pores is called a porous medium. When the pores have their origin in the genesis of the
rock, the rock is said to have primary porosity (sedimentary deposits, weathered hard rocks).
When their origin is from events that occurred during a later stage (jointing, faulting,

20
Occurrence of groundwater, regime and dynamics
dissolution) the porosity is said to be secondary. Most sedimentary rocks with primary porosity
also contain some joints and fracture that add secondary porosity to the total porosity of the
porous medium. Similarly dense rocks (quartzites and granites) may have some primary
porosity which is subordinate to the secondary porosity. Occasionally the effects of both forms
of porosity can be recognised in the behaviour of a single groundwater body; this phenomenon
is called dual porosity.
1.2 Groundwater flow
1.2.1 Groundwater flow processes and continuity
Groundwater is an important source of water; it may provide the base flows for rivers, or act as
an underground reservoir from which water can be pumped as a location into which water can
be drained. Consequently it is the flow of groundwater which must be examined.
Usually, groundwater travels very slowly; one hundred metres per year is a typical
average horizontal velocity and one metre per year is a typical vertical velocity. When these
velocities are multiplied by the cross-sectional areas through which the flows occur, the quan-
tities of water involved in groundwater flow are often substantial. Consequently, the essential
feature of an aquifer system is the balance between the inflows, outflows and quantity of water
stored.
Unlike a surface reservoir, the upper surface of the groundwater (the water table or
phreatic surface) is not horizontal; a sloping water table results from the resistance to flow
caused by the hydraulic conductivity. Due to the slow movement of groundwater, care is
necessary when positioning any man-made outflows, such as pumped boreholes, to ensure
that they collect water efficiently from the aquifer system. The water balance of the aquifer
system is the key to the identification of the aquifer resources and the consequences of changes
in exploitation. The water balance is based on the principle of the continuity of flow.
1.2.2 Groundwater head
Although the flow of groundwater is an important process, the actual groundwater flows cannot
be measured directly. Consequently, an alternative method of identifying groundwater
conditions is required and this is provided by the groundwater head (or groundwater potential).
The groundwater head at a location in an aquifer is the height to which water will rise in a
piezometer (or observation well). So that the conditions at a specific location in an aquifer can be
identified, the open section of the piezometer which monitors the conditions should extend for no
more than a metre. Figure 1.2(a) shows a typical piezometer; the groundwater head, h, equals the
sum of the pressure head p / g, plus the datum head, z. Groundwater flows from a higher to a
lower groundwater head. Typical examples of the use of the groundwater head to identify the
direction of the groundwater flow are shown in Fig. 1.2. In Fig. 1.2(b) a confined aquifer is rep-
resented in which there are two piezometers to identify the direction of flow. In the left-hand
diagram the flow is from left to right because the lower groundwater head is in the piezometer
to the right, that is in the direction of the dip of the strata. For the right- hand diagram the
groundwater gradient is to the left, since the water level in the left-hand piezometer is lower;
consequently the direction of the flow is up-dip in the aquifer. This flow could be caused by the
presence of a pumped well or a spring to the left of the section.
The diagram in Fig. 1.2(c) refers to an unconfined aquifer in which the water table has a
significant slope. There are three piezometers (which can be considered as two pairs) provided
to identify the horizontal and vertical components of the flow. Piezometer (ii) penetrates just below

21
Groundwater studies
Figure 1.2 Significance of groundwater head; (a) definition, (b) identifying flow directions in one dimension,
(c) identifying flow directions in two dimensions

22
the water table, hence it provides information about the water table elevation; the other two
piezometers are within the aquifer and are only open at the bottom of the piezometer. Therefore
they provide information about the groundwater head at the bottom of the piezometer. Since the
open sections of piezometers (i) and (ii) are at the same horizontal elevation, they provide
information about the horizontal flow component (or velocity component); the horizontal flow is
clearly from left to right. Piezometer (iii) is positioned directly below piezometer (ii) and pro-
vides information about the vertical flow component on that section. Since the groundwater head
in piezometer (iii) is below that of piezometer (ii) there is a vertically downward component of
flow. The horizontal and vertical flow components can be combined vectorially to give the
magnitude and direction of the flow.
1.2.3 Darcys Law, hydraulic conductivity and transmissivity
The magnitude of the flow can be calculated from Darcys Law; the Darcy velocity, v, can be
determined from
v=K.i (1.1)
where
K is the hydraulic conductivity or (permeability) [L/T],
i is the hydraulic gradient between two piezometers,
the minus sign signifies that flow is in the direction of falling groundwater head.
This equation can be used to estimate the flows for the two examples of Figure 1.2(b). The
same equation can be used to calculate the horizontal and vertical flow components for the
example of Fig 1.2(c).
The Darcy velocity calculated from Darcys Law is an artificial velocity, since it assumes
that flow occurs through the whole cross section. In practice the flow occurs only through
certain pores and fissures; consequently, an approximation to the actual velocity can be obtained
by dividing the Darcy velocity by the effective porosity.
In regional groundwater studies, an important quantity is the horizontal flow through the
aquifer system, Qh; this flow can be calculated from the equation
Qh = T . i (1.2)
where T, the transmissivity is the sum of the permeabilities over the saturated depth. For
example, the transmissivity in the x direction,
Tx = kx dz (1.3)
sat depth
1.2.4 Inflows and outflows
A study of the groundwater flow within an aquifer requires information about inflows and
outflows. The term recharge is used for the inflow to an aquifer system arising from precipitation,
return flow from irrigation and flows from various surface water bodies such as rivers, canals
and lakes. The magnitude of the recharge is likely to change significantly with time. Two books
published by the International Association of Hydrogeologists provide extensive information
about recharge; Groundwater Recharge (Lerner et al., 1990) reviews the methods of estimating

23
Groundwater studies
recharge in a range of climates whereas in his book Simmers (1997) focuses on semi-arid and
arid areas. There can also be inflow from other aquifers.
Outflows from the aquifer system can be divided into natural outflows and man-made
outflows. Natural outflows occur when water leaves the aquifer at springs or into rivers. Other
natural outflows include low-lying areas which act as a sink to groundwater systems; this form
of outflow may be associated with areas of evapotranspiration especially from deep-rooting
vegetation. These low-lying areas often form wetlands which have a high ecological value. One
further natural outflow occurs when water flows into other aquifers.
There are also man-made outflows. Pumped wells and boreholes are the main means of
withdrawing water from an aquifer; different designs of wells and boreholes are required for
different types of aquifer and different discharge rates. Since the velocities in the vicinity of the
pumped borehole are far higher than the natural groundwater velocities, there is a risk of
deterioration of the aquifer in the vicinity of the well or borehole and a deterioration of the bore-
hole structure. Horizontal wells or adits provide alternative means of collecting water from an
aquifer; this approach is especially suitable for shallow aquifers or for aquifers with thin lenses
of good quality water.
1.2.5 Storage coefficients and time dependency
Decreases in the volume of the water stored in an aquifer release water to flow through the
aquifer, especially during periods with little recharge. There are two types of storage coefficients
(see Figure 1.3):
storage coefficient of a confined aquifer SA [dimensionless]; this is the amount of water
released from a column of unit cross-sectional area of a confined aquifer for a unit
decline of the piezometric surface.
specific yield of an unconfined aquifer SY [dimensionless], this is the amount of water
released from a column of unit cross-sectional area of an unconfined aquifer for a unit
decline of the water table (phreatic surface)
The storage properties of an aquifer allow continuing exploitation of the aquifer during
periods of poor recharge. Consequently it is necessary to consider the time-variant behaviour of
an aquifer. In periods of high recharge, inflows to the aquifer system in excess of the outflows
may be stored in the aquifer although the resultant rise in the groundwater heads may lead to
increased outflows to springs or rivers. During periods with little or no recharge, water is
withdrawn from storage. Due to the important time-variant response of aquifers, it is essential to
obtain all data on a time-variant basis and it is advisable to study aquifer conditions over a
number of years before reaching conclusions about the aquifers behaviour
1.3 Groundwater composition
1.3.1 Physical and chemical properties
A water molecule consists of two atoms of hydrogen (H) and one atom of oxygen (O), so it has
the chemical formula H2O. At sea level its freezing point is 0C and its boiling point is 100C.
Water is a good solvent and natural water always contains some chemicals in solution. The total
amount of dissolved solids in a water sample (TDS) is expressed in mg/l and water is classified
according to the TDS as fresh, brackish, saline or brine. The limits in this classification vary from
country to country and even from study to study (see also Chapter 2).
The major cations in groundwater are usually sodium (Na+), potassium (K+), calcium
(Ca ), magnesium (Mg++) and the major anions are chloride (Cl), bicarbonate (HCO3+),
++

24
Figure 1.3 Description of storage effects; (a) specific yield, (b) confined storage
sulphate (SO4 ) and nitrate (NO3). Of the solutes that occur in minor amounts the following
are mentioned because of their influence on the water use: iron (Fe; taste, staining), boron (B;
toxicity to plants), fluoride (F; health risk), aluminium (Al; health risk), nitrate (NO3; health
risk). Very small amounts of other ions, usually called trace elements, are often present in
natural water. Furthermore, small amounts of the isotopes of hydrogen, such as deuterium 2H,
tritium 3H and oxygen 18O occur in all natural waters. The use of the analysis of the isotope
content of these environmental isotopes is discussed in Chapter 9.
Groundwater that comes from deep aquifers (> 2000 m) or from aquifers in contact with
subterranean (volcanic) heat sources may have high temperatures and may be used as a source
of geothermal energy; this topic will not be discussed in this publication.
Water with a particular chemical composition may be exploited as mineral water for
bottling and for medical use in health resorts; this is another topic not covered in this
publication.
1.3.2 Risk of groundwater pollution
Groundwater pollution from human activities has become a major topic of groundwater
research and large amounts of money are currently being invested in the prevention of ground-
water pollution and in the rehabilitation of polluted groundwater bodies. The contaminants that

25
Groundwater studies
may pollute groundwater are grouped according to their physico-chemical characteristics in

order to characterise their fate in the groundwater environment:
metals;
oxy-anions;
dissolved organics;
non-aqueous phase liquids (NAPLs);
colloids and radionuclides;
bacteria and viruses.
(i) Metals
Dissolved metals usually occur as cations in groundwater, but important exceptions exist such
as chromium (Cr) and uranium (U), which may also occur as oxy-anions. The mobility of
metal cations often increases with decreasing pH, for a combination of two reasons. Firstly,
most minerals that are formed by metals are less soluble at increasing pH: carbonates, oxy-
hydroxides, sulphides. Secondly, the sorption capacity of solid phases for cations increases with
increasing pH.
(ii) Oxy-anions
Oxy-anions have less singular characteristics in groundwater than metals. The sorption capacity
for anions generally increases with decreasing pH. However, the importance of sorption strongly
varies for individual anions and oxy-anions and is still an active topic of research. However, it
has become clear that sorption is relevant for oxy-anions that can be considered as weak bases,
like phosphate, arsenate and chromate. Sulphate, being a strong base, is weakly adsorbed, and
adsorption of nitrate, which is also a strong base, is negligible.
Oxy-anions that behave as weak bases are most mobile under weakly acid conditions (pH
approximately 5 to 6). At higher pH the mobility is limited by solubility of minerals and at lower
pH it is limited by sorption.
Several oxy-anions are not stable within the entire range of redox conditions that can be
found in groundwater which makes them susceptible to redox processes. With decreasing redox
potential , the major anions, nitrate and sulphate may be converted to N2 and H2 S, respectively.
The former is unreactive in groundwater, the latter may form sulphides with Fe or other heavy
metals. Arsene occurs in three redox states in groundwater: arsenate, arsenite and arsenic. Their
mobility is distinctly different, since arsenic binds to sulphides. This limits As concentrations in
groundwater in strongly reduced environments. The redox state for chromium also varies from
Cr(VI) as Cr2O 7 2 to Cr(III) as Cr3+. The latter is also much more susceptible to sorption and
precipitation than the former.
(iii) Dissolved organics
Dissolved organics are relevant in the groundwater environment in two different ways. First,
dissolved organic matter may be a reductant of the groundwater system , i.e. its decomposition
brings about anaerobic conditions in the groundwater. Second, organic molecules may be
undesirable in groundwater because of their toxicity. The first condition is normally indicated by
the concentration of Dissolved Organic Carbon (DOC) and refers to organic matter as a major
contributor to the overall groundwater composition. The latter refers to individual organic
species at low concentrations and these species are referred to as micro-organics. Examples are
pesticides, polycyclic aromatic hydrocarbons and chlorinated aliphatic hydrocarbons.
The transport of micro-organics is controlled by aqueous solubility, sorption and

26
degradation. As soon as the organic carbon content of the aquifer exceeds 0.1% solid organic
matter is the major sorbing compound. Degradation of micro-organics may be controlled
biotically or abiotically. Often, biotic degradation is faster than abiotic. The degradation rate
strongly depends on the redox type of the groundwater/aquifer system.
(iv) Non-aqueous phase liquids
Following spills on the surface, non-aqueous phase liquids (NAPLs) may occur as immiscible
fluids in the subsurface. The flow of these liquids is hydrodynamically not geochemically
controlled. Their behaviour depends on their density; fluids like petrol, diesel, etc. are lighter
than water and form floating layers. On the other hand, several solvent fluids like trichloro-
ethene are heavier and form sinking layers. Non-aqueous phase liquids are important as a
source of dissolved organic matter in groundwater. The DOC may change the redox status of the
aquifer system and the soluble compounds of oil derivatives may cause deterioration in the
groundwater quality. Benzene, toluene, ethylbenzene and xylene, for example, are compounds in
oil that show relatively high mobility in groundwater.
(v) Colloids and radionuclides
Contamination associated with nuclear activities deserves special attention. Radionuclides

behave chemically in an identical manner to their non-radioactive isotopes, but physically they
may show some distinct features. Radionuclides may be attached to colloids.
Colloidal particles can be transported faster than the average linear groundwater velocity.
The reason is that a colloid has to be considered as a distinct particle having a specific radius.
This makes it impossible to enter pores having smaller radius and also impossible to move along
the edge of pores. Effectively, the flow rate is higher than for water molecules themselves or
dissolved species. The charge characteristics of the colloid compared to that of the solid matrix
complicate the description of colloidal transport. Colloids may be repelled from the edge when
they have a similar charge and may be attracted if they have opposing charges.
Contaminants that are adsorbed to colloids can thus be transported at unexpectedly high
flow velocities. One of the most relevant examples is radioactive caesium that is adsorbed
by colloidal Fe-hydroxide particles. Radioactive contamination via colloid-facilitated transport
usually has a local nature, since the aquifer acts as a filter for colloidal particles. Aquifers having
large pores such as gravel deposits, fractured zones or karstified zones, are more susceptible to
contamination caused by colloids that carry contaminants than fine-grained aquifers.
(vi) Bacteria and viruses
Bacteria and viruses may cause diseases if the contaminated water is used for drinking. These
bacteria are referred to as pathogenic bacteria. The contamination is often related to sewage
or waste water. Short-circuit flow from the surface to well screens is a well-known cause of
bacterial contamination, which can often be attributed to poor well construction.
Bacteria can be considered as living colloidal particles. Viruses are particles that have a
smaller radius than bacteria: in the order of 0.01 m versus 1 m, for bacteria. Like colloids, the
mobility of bacteria is primarily controlled by filtration. The movement is enhanced by large
pores, i.e. coarse matrices having small sorption capacities. Survival of pathogenic bacteria is
encouraged by high moisture contents, low temperature and neutral pH.
Filtration is less important for viruses due to their small size. The dominant factor
affecting their movement is adsorption. Survival of viruses is favoured by high moisture
contents and low temperatures. The mobility of bacteria and viruses is largely determined by the

27
Groundwater studies
same factors. Greatest movements occur in coarse aquifers and infiltration areas with thin
unsaturated zones. Fractured and karstified rock have the highest potential risk. Contamination
by bacteria or viruses is often local, for similar reasons to colloidal-facilitated radionuclide
transport.
Both pathogenic bacteria and viruses decay in groundwater, but this needs time. The
decay has been described as:
Nt = N0 e kt (1.4)
where:
N refers to the number of species,
k is the degradation constant, and
t is time.
Observed values for the degradation constant range from 0.001 to 0.06 hr 1 for the
groundwater environment.
Table 1.1 presents an overview of the types of geochemical and biogeochemical reactions
that control the fate of contaminants in groundwater.
Table 1.1 Overview of the types of (bio)geochemical reactions that control the fate of contaminants
in groundwater

28
1.4 Groundwater assessment and exploration
1.4.1 Aquifer and groundwater systems
Any groundwater assessment study requires knowledge of the nature of the aquifer system and
groundwater conditions.
(i) Aquifer system
The aquifer system comprises:

the geometry (extent and thickness) of the aquifer or aquifer system and possible
interlayered aquitards,
the boundary conditions: head controlled, flow controlled or no-flow boundaries
the aquifer type(s): confined, semi-confined (leaky), unconfined (phreatic), or perched
unconfined (Figure 1.4)
the hydraulic parameters are derived from the properties of the aquifer material:
- porosity n (dimensionless),
- intrinsic permeability k (dimension L2) is a function of the grain-size diameter,
rounding and sorting,
- compressibility of the rock matrix, a, that ranges from 10 6 to 10 8 for clay to 10 8 to
10 10 m2/N or Pa1 for gravel.
For fresh water these fundamental parameters combine into the commonly used terms:
k g
hydraulic conductivity: K = [L1T 1] (1.5)

where:
and are the density and viscosity of the water, and
g is acceleration of gravity
Tx = kx dz [L2 T 1]

transmissivity (1.3)
sat depth
The transmissivity for vertical flow through an aquitard is:
Tv = Kv/M [T 1]; (1.6a)

where
Kv and M are the vertical hydraulic conductivity and thickness of the aquitard,
respectively.
The reciprocal of this value is known as the hydraulic resistance of the aquitard:
c = M/K v [T] (1.6b)
Fractures in a rock formation (Figure 1.5) strongly influence the fluid flow in that formation.
Consequently, conventional well flow equations developed primarily for homogeneous aquifers
do not adequately describe the flow in fractured rocks. An exception occurs in hard rocks of
very low permeability if the fractures are numerous enough and are evenly distributed

29
Groundwater studies
Figure 1.4 Aquifer types: (a) confined aquifer, (b) unconfined aquifer, (c and d) leaky aquifer,
(e) perched aquifer, (f) composite aquifer system
throughout the rock; then the fluid flow will only occur through the fractures and will be similar
to that in an unconsolidated homogeneous aquifer.
A complicating factor is the fracture pattern, which is seldom known precisely. This means
that, based on the geological data, a fracture model must be assumed (Fig. 1.6). Although many
theoretical models have been developed in recent years, few of the associated well functions
have been tabulated. Therefore the discussion will be restricted to fracture models for which the
tables have been published (Kruseman et al., 1990).
The double-porosity concept regards fractured rocks as consisting of matrix blocks with a
primary porosity and low hydraulic conductivity, separated by fractures with a low storage

30
Figure 1.5 Porosity systems: (a, b) primary single porosity, (c, d) secondary single porosity,
(e, f) double porosity
Figure 1.6 Fractured rock models: a) a naturally fractured rock formation, b) an idealised three-dimensional,
orthogonal fracture system, c) idealised horizontal fracture system

31
Groundwater studies
capacity but a high hydraulic conductivity. This concept assumes that no variation of head
occurs in the matrix blocks, i.e. the inter-porosity flow is in pseudo-steady state. The flow
through the fractures in the vicinity of a pumped borehole will be radial and non-steady,
Figure 8.1. Curve-fitting methods and straight-line methods have been developed to analyse
pumping tests in double-porosity aquifers (Chapter 8).
(ii) Groundwater system
The groundwater system comprises many components:

the quantity of groundwater stored in the aquifer system and its quality,
water table (phreatic) levels and their fluctuations over time indicating changes in the
amount of water stored in the aquifer,
groundwater head (piezometric level) fluctuations of confined or semi-confined
aquifers indicating the changes over time of the hydraulic pressure in the aquifer,
recharge and discharge sources and the time-dependent rates of discharge and
recharge from each source (hydraulic stress),
groundwater budget, being a comparison between the sum of all recharge and other
inflow components and the sum of all discharge components plus the change in
storage over a specific period of time (e.g. six months, one year, etc.),
chemical composition.
(iii) Groundwater models
As an understanding of the groundwater system develops, mathematical groundwater flow

simulation models can be used to check the consistency of the data. Mathematical models range
from the Theis equation for radial flow to a well to complex three-dimensional flow and con-
taminant transport models. Adjustment of the parameters and variables within physically
realistic limits allows the refinement (calibration) of the model so that the model reproduces the
historical changes in groundwater heads and groundwater flows.
In the same way a mass transport model can be developed to integrate the hydrochemical
data.
As the precise parameter values are often scarce and are obtained indirectly (e.g. by
pumping tests, see Chapter 8) and recharge values may be difficult to estimate, adjustments may
be necessary before the model is properly calibrated. This implies that sufficiently long time
series data (groundwater heads, recharge and discharge rates) are required to develop a ground-
water flow simulation model. Without historical data a model may have no real similarity with
the actual physical problem.
When a calibrated model has been prepared, it is assumed that because it can reproduce
the past behaviour, it can also predict the future behaviour under changing hydraulic stress
conditions. For further details on models and modelling the reader is referred to Anderson and
Woessner (1992).
1.4.2 Data collection
A successful groundwater investigation depends on field data (see also Chapters 3 and 4). How-
ever, to minimise additional fieldwork each groundwater study should start with the collection
and analysis of the available information and documentation. Based on the results of this
analysis, it will be decided whether this information is sufficient to carry out the assessment or
whether additional information must be collected in the field. At that time, a preliminary report
is prepared that contains the available information and a plan of additional field studies. The

32
following sections consider the kind of information that is required and how and from where it
can be obtained.
(i) The geometry of the aquifer
The geometry of the aquifer or aquifer system and possible interlayered aquitards, is determined
by the extent and thickness of the aquifer. These geometrical parameters are derived from the
study of the geology of the area, borehole drilling data (for details see Chapter 7), geophysical
well logs and geophysical surface studies, e.g. geo-electrical surveys and seismic surveys (for
details, see Chapter 6).
(ii) Water table and piezometric level
The water table and piezometric level data are obtained from dedicated observation wells and
piezometers that are regularly monitored and from incidental measurements.
(iii) The aquifer type(s)
Figure 1.4 shows the main different aquifer types: confined, semi-confined (leaky), unconfined
(phreatic), perched (i.e. two independent unconfined aquifers, one above the other); in many real
life the aquifer system is often more complex (see Chapter 11).
A confined aquifer is completely filled with water and bounded above and below by an
impervious layer. The water level in a piezometer that taps the aquifer rises to a level above
the top of the aquifer (the piezometric level). If the pressure in the aquifer is such that the
piezometric level lies above the land surface, the piezometer may become a free-flowing
(artesian) well.
An unconfined aquifer is bounded below by an impervious layer, but is not restricted by a
confining layer above it. Its upper boundary is the water table that is free to rise and fall. Water
in a well which just penetrates an unconfined aquifer is at atmospheric pressure and does not
rise above the water table.
A semi-confined, or leaky, aquifer is an aquifer whose upper and lower boundaries
are aquitards, or one boundary is an aquitard and the other is an aquiclude. Water is free to
move up or down through the aquitards. If a leaky aquifer is in hydrological equilibrium, the
water level in a well tapping it may coincide with the water table. The water level in the well
may also stand above or below the water table, depending on the recharge and discharge
conditions.
(iv) The hydraulic parameters
Methods to determine the hydraulic parameters mentioned in section 1.4.1 are discussed in
Chapter 8.
(v) Boundary conditions
Figure 1.7 shows the three boundary types. If the boundary of an aquifer consist of an
impervious barrier, there will be no flow across that boundary; i.e. it is a no-flow boundary. If
the boundary is pervious, flow may occur if there is a head difference between the groundwater
on either side of the boundary. The amount of flow is determined by this head difference and
the transmissivity at the boundary; i.e. it is a head-controlled boundary. If the flow across the
boundary is not determined by a head difference, the boundary is said to be flow-controlled, e.g.

33
Groundwater studies
the inflow from a karst aquifer into an alluvial aquifer. The boundary conditions are determined
from an analysis of the hydrogeological context.
Figure 1.7 Aquifer boundary conditions; 1) flow-controlled boundary, 2) external zero-flow boundary,
3) internal zero-flow boundary, 4) and 5) internal head-controlled boundaries, 6) external
head-controlled boundary, 7) free surface boundary (After Boonstra and De Ridder, 1981)
(vi) Groundwater storage
The volume of groundwater stored in an aquifer is calculated as the product of the thickness of
the saturated part of the aquifer and its effective porosity; this volume is much larger than the
exploitable amount of groundwater, because:
part of the water is retained in the pores, i.e. the difference between the total porosity
and the specific yield,
the yield of a pumped well in an unconfined aquifer may decrease with the dimin-
ishing thickness of the saturated part of the aquifer.
The yield of a confined aquifer is determined by the confined storage coefficient, which is
several orders of magnitude smaller than the porosity. In deep aquifers, the pumping lift may
become uneconomically large before the pumping water level falls to the top of the aquifer.

34
(vii) Groundwater budget
The groundwater budget is given by:
groundwater recharge groundwater discharge = change in groundwater storage
The recharge components include:

percolation of precipitation,
percolation of irrigation water,
percolation from other surface water bodies, e.g. rivers, lakes,
lateral subsurface inflow,
vertical subsurface inflow through an underlying or overlying aquitard, artificial
recharge.
Discharge components include:

direct evapotranspiration from the water table,
spring flow and exfiltration,
lateral subsurface outflow,
vertical subsurface outflow,
discharge by pumping or other man-made devices.
(viii) Groundwater quality
Sampling of groundwater for water quality analysis not only comprises the sampling technique
itself but also the set-up of the sampling well, the materials used, the type of field measurements
and conservation techniques prior to analysis in the laboratory (see also Chapters 2 and 3). The
form of the sampling depends on the purpose of the water quality determination, e.g. regional
differentiation between aquifers, identification of recharge sources, study of local groundwater
pollution, monitoring of the supply of water for domestic use, etc.
A critical issue is the correct choice of the materials for the sampling equipment. Teflon
materials are preferred for the completion of the sampling well because Teflon is the most
chemically inert plastic and is therefore superior to other materials such as polypropylene and
polyethylene or rigid PVC. When the content of metals has to be determined, steel and iron
particularly the latter should be avoided.
The manner of collecting the groundwater sample may influence the outcome of the
analysis and is most critical for volatile compounds, in particular for volatile organic chemicals.
In order to collect a sample that is not influenced by a long presence in the well the latter should
be flushed repeatedly before a sample is collected. A practical rule of thumb is flushing for three
times the well volume. All materials that come into contact with the groundwater during
sampling should be cleaned between two samplings to avoid cross-contamination. Special care
should be taken to avoid contact with oil or grease from the engines or other parts of the
sampling equipment.
Sampling for the analysis of possible microbiological contamination requires sterile
equipment and sampling bottles, as well as procedures that prevent contamination of the sample
during its handling.

35
Groundwater studies
1.5 Groundwater exploitation and management
1.5.1 Monitoring conditions within the aquifer system
When considering groundwater exploitation and management it is essential to have a versatile

system of monitoring conditions within the aquifer system; a detailed presentation on aquifer
monitoring can be found in Chapter 3. This brief rsum concentrates on three important aspects
of groundwater monitoring.
(i) Monitoring groundwater heads
As explained in Section 1.2, groundwater heads provide invaluable information about the flow
conditions in the aquifer system. Consequently, the changing conditions in an aquifer can be
identified by the monitoring of groundwater heads preferably in purpose drilled observation
boreholes and piezometers. The use of boreholes which are open to the aquifer over a significant
proportion of their depth is likely to lead to erroneous results, especially if the borehole is close
to a pumped borehole. Figure 1.8 compares the response of an open borehole and the same bore-
hole when a number of individual piezometers were installed in the borehole; these piezometers
were influenced by the pumping from a nearby supply borehole. From the difference between
the continuous line which represents the individual piezometers and the broken line which
refers to the open borehole, it is clear that the single curve for the open borehole is misleading
(Rushton and Howard 1982). In fact the open borehole acts as a large vertical fissure transferring
water from top to bottom of the aquifer; this so disturbs the aquifer flows that the result from
the open borehole is certain to be misleading.
The frequency and duration of groundwater head measurements must reflect the changes
in aquifer conditions. Close to a pumped borehole, frequent readings are required, whereas at
greater distances less frequent readings are acceptable. However it is essential to take sufficient
readings to identify the peaks and troughs of groundwater heads; the minimum frequency of
readings should be every two months.
(ii) Monitoring groundwater flows
The direct measurement of flows in an aquifer is rarely possible, although valuable information
can be gained from the use of tracers. Estimates of inflows such as recharge can be derived from
tracers and from lysimeters; these estimates suffer from the limitation that they are point
readings. Integrated values of groundwater outflows can be gained from estimates of the change
in river flow between two locations; monitoring of spring flows can also provide information
about local outflows. One component of groundwater flow which can be measured accurately is
the quantity abstracted from boreholes.
(iii) Monitoring groundwater composition
The chemical composition of groundwater provides crucial information about the historical
aquifer conditions and current changes within the aquifer system. Because groundwater under
natural conditions moves so slowly, the overall distribution of the chemical components often
provides information about the long term conditions in the aquifer system. However for parts of
the aquifer in the vicinity of sources of inflow or outflow, changes in the chemical composition
can give clear information about the current groundwater flow processes. In determining the
chemical composition within the aquifer or at outlets such as springs, rivers or pumped

36
Figure1.8 Unreliability of open boreholes; hydraulic head readings of an open borehole compared
to individual piezometers close to a pumped borehole
boreholes, great care must be taken to ensure that the sample is not contaminated or allowed to
change in composition before testing.
The determination of the groundwater quality within an aquifer is often attempted using
open observation boreholes. Again, serious errors can arise due to the mixing of groundwaters
in open boreholes. Price and Williams (1994) compared sulphate concentrations obtained from
depth samples in an open borehole, both with porewater samples obtained when the borehole
was drilled and also with samples obtained from sections of the aquifer which were isolated by
packers. Figure 1.9 shows that the depth samples in an open borehole do not reflect accurately
the conditions within the aquifer. Also, samples obtained from pumping between packers fail to
reflect the water quality at depth, since the higher sulphate water from the upper part of the
aquifer flows into the lower part of the aquifer, causing contamination.

37
Groundwater studies
Figure 1.9 Unreliability of open boreholes: chemical sampling in an open borehole compared to cores
and packer testing
1.5.2 Water balances and simulation models
Before any changes are made to the exploitation of an aquifer, or management rules are intro-
duced, it is necessary to understand the flow processes within the aquifer system. The first step

38
is usually to develop a water balance; subsequently a simulation model may be devised. Water
balances should be prepared for regular periods, such as every six months, or every year. The
initial attempt at a water balance is unlikely to give components which sum to zero; however,
any lack of balance should provide insights into the components which require further
investigation. There are two main purposes in developing a simulation model:
To confirm whether there is a valid understanding of the overall aquifer system; this is
achieved by bringing together all the parameter values and estimates of quantities
such as the recharge and river/aquifer interaction to see whether historical responses
can be reproduced. If there is difficulty in reproducing certain historical responses, the
simulation model can be used to explore alternative parameter values and may lead to
the need for further fieldwork. Until this stage of the model development is completed,
the simulation model should not be used for making predictions.
Once the validity of the simulation model has been confirmed, it can be used for pre-
dictions related to master planning of a large system or examining the impact of a
limited number of boreholes. In the predictive simulations it is possible to examine the
consequences of various exploitation scenarios; alternatively the objective may be to
see how deteriorating conditions within the aquifer can be corrected.
1.5.3 Potential consequences of changed groundwater conditions
Groundwater exploitation, the construction of irrigation schemes and urban development can all
lead to major changes in groundwater conditions.
Groundwater development leading to heavy exploitation can cause significant falls in the
water table and a reduction in the yields of boreholes due to the extra height over which
the water has to be raised; if the aquifer is over-exploited it may become unusable. Another
consequence of heavy exploitation and the removal of large volumes of water can be settlement
of the ground (Downing, 1994).
Rising water tables can occur due to excessive irrigation, losses from canals or return
flow from irrigated fields; the result of the rising water table may be water-logging and
salinisation. In urban areas, the natural drainage by surface water courses and through
the aquifer system may be unable to take away excess water; unless sufficient
additional drainage is provided ,water tables may rise to the ground surface. Riyadh in
Saudi Arabia provides a typical example where it was thought that the aquifer would
provide natural drainage but the vertical permeability of underlying strata is too small
to take away the recharge which results from losses from the water supply system and
the irrigation of gardens; the result is that the water table is close to the ground surface
in many areas of the city (Rushton and Al-Othman, 1994).
In a number of major cities the original source of water for domestic and commercial
purposes was groundwater; due to heavy exploitation in the past there have been
significant falls in the water table elevation. Because of the unreliability of the ground-
water (both in terms of quantity and quality) water is imported from another
catchment. Consequently, the rate at which water is pumped from the aquifer has
decreased, resulting in a stabilisation and subsequent rise in the water table. This water
table rise is enhanced by leakage from water mains and sewers and may lead to
flooding of basements and underground services and a reduction in the load-bearing
capacity of foundations (Johnson, 1994; Wilkinson and Brassington, 1991).
Changing abstraction from an aquifer can influence the flows to springs, rivers, and
other areas of seepage; this can lead to a severe deterioration in the environmental
conditions. The maintenance of wetlands has become an important issue in a number
of countries. An increase in the volume of water pumped from an aquifer will affect

39
Groundwater studies
surface water features and will therefore have an environmental impact; however, it is
sometimes possible to relocate abstraction boreholes and modify abstraction patterns
so that the environmental impact can be minimised.
A further consequence of increased exploitation of an aquifer is that poor quality water
can be drawn in from other parts of the aquifer system. There are numerous field
examples of saline water being drawn into coastal aquifers or poor quality old water
being drawn up from depth (Wheater and Kirby, 1998).
1.5.4 Cost of investigations and groundwater development
A groundwater investigation can be costly since it may involve:

drilling trial boreholes,
monitoring groundwater heads, flows and chemical characteristics for a number of
years,
carrying out a thorough analysis of all the data and
developing a simulation model.
Nevertheless, investigations which are carefully planned, implemented and interpreted
are usually cost effective, since they can lead to a suitable abstraction regime minimising the risk
of providing too few or too many boreholes and can identify any undesirable effects on the
environment.
When a groundwater scheme has been implemented it is important to continue with
monitoring to ascertain whether the consequences, as identified in the initial investigation, are
observed in the field. If there are unforeseen results following the implementation, any
necessary modifications can be incorporated.
1.5.5 Management of groundwater 1
The management of groundwater resources is a difficult task because it requires the co-operation
of all the users. Frequently there is competition for the good quality groundwater and there is
insufficient water to meet all the demands. This brings a serious risk of over-exploiting and
damaging the aquifer, but the consequences of over-exploitation may take many years to become
unambiguously apparent. A fall in groundwater heads may be due to a period of reduced
recharge or a redistribution of abstraction sites and may not be a sign that abstraction exceeds
recharge. Careful monitoring of the water table elevation is the key to the identification of
unsustainable abstraction from the aquifer (Rushton, 1994).
Suitable control rules often involving legislation are needed for the management of an
aquifer. One approach is a system of licences to exploit the groundwater, but an aquifer may
be over-licensed historically. Furthermore, when considering a licence application it may be
difficult to be certain about the consequences of the increased abstraction; time-limited licences
are a sensible way of determining the effect of increased abstraction. In countries where there is
no enforceable legislation about groundwater abstraction, over-abstraction is likely to occur.
Studies have been made of the optimal use of aquifers, but they tend to have little rel-
evance in real-life situations. It is difficult to reduce or stop exploitation, since large investments
are required for drilling boreholes, with serious financial losses if the boreholes cannot be used.
If the borehole capacity exceeds the inflows into the aquifer system, one approach is to limit the
drawdown in all the pumped boreholes; this is similar to defining a minimum water level in a
surface reservoir.
1. Management of groundwater is considered in detail in Chapters 12 and 13.

40
1.6 References and additional reading
ANDERSON, M. P.; WOESSNER, W. W., 1992. Applied Groundwater Modelling, Academic Press, 381 pp.
BOONSTRA, J.; DE RIDDER, N. A., 1981. Numerical Modelling of Groundwater Basins; A User-oriented
Manual. International Institute of Land Reclamation and Improvement, Publication No. 29,
Wageningen, 226 pp.
DE WIEST, R. J. M., 1965. Geohydrology. John Wiley and Sons, New York, 366 pp.
DOWNING, R. A., 1994. Falling Groundwater Levels, a Cost-Benefit Analysis. In: Proceedings, ICE
International Conference on Groundwater Problems in Urban Areas, Thomas Telford, London,
pp. 21336.
FREEZE, R. A.; CHERRY, J. A., 1979. Groundwater. Prentice-Hall, Englewood Cliffs, NJ, 604 pp.
JOHNSON, S. T., 1994. Rising Groundwater Levels: Engineering and Environmental Implications.
In: Proceedings, ICE International Conference on Groundwater Problems in Urban Areas,
Thomas Telford, London, pp. 28598.
KRUSEMAN, G. P.; DE RIDDER, N. A.; VERWEIJ, J. M, 1990. Analyses and Evaluation of Pumping Test
Data (Second edition). International Institute of Land Reclamation and Improvement,
Publication No. 29, Wageningen, 371 pp.
LERNER, D. N.; ISSAR, A. S.; SIMMERS, I., 1990. Groundwater Recharge. A Guide to Understanding
and Estimating Natural Recharge. International Contributions to Hydrogeology, Verlag Heinz
Heise, Vol. 8, 345 pp.
PRICE, M.; WILLIAMS, A., 1993. The Influence of Unlined Boreholes on Groundwater Chemistry: A
Comparative Study Using Pore-Water Extraction and Packer Sampling. J. Inst. Water and
Eng. Man, Vol. 7, pp. 6519.
RUSHTON, K. R., 1994. Assessing Aquifer Exploitation Using Observation Boreholes. J. Inst. Water
and Env. Man., Vol. 8, pp. 15664.
RUSHTON, K. R.; HOWARD, K. W. F., 1982 The Unreliability of Open Observation Boreholes in
Unconfined Aquifer Pumping Tests. Ground Water, Vol. 20, pp. 54650.
RUSHTON, K. R.; AL-OTHMAN, A. A. R., 1994. Control of Rising Groundwater Levels in Riyadh,
Saudi Arabia. In: Proc. ICE International Conference on Groundwater Problems in Urban Areas,
Thomas Telford, London, pp. 299309.
SIMMERS, I., 1997. Recharge of Aquifers in (Semi-)arid Areas. International Contributions to Hydro-
geology, A. A. Balkema, Rotterdam, Vol. 19, 277 pp.
WHEATER, H.; KIRBY, C., 1998. Hydrology in a Changing Environment, Vol. II, John Wiley and Sons,
New York, 604 pp.
WILKINSON, W. B.; BRASSINGTON, F. C., 1991. Rising Groundwater Levels An International
Problem. In: Applied Groundwater Hydrology: a British Perspective. Oxford University Press,
pp. 3553.

41
2 G ro u n d w a t e r q u a l i t y
2.1 Introduction
2.1.1 Definition and scope of chemical hydrogeology
Chemical hydrogeology is an interdisciplinary science that relates the quality of groundwater to

various processes and reactions in aquifers and aquitards: the quality of groundwater in a
particular aquifer is described, evaluated and explained primarily by application of principles of
aquatic chemistry to a hydrogeological system. In other words chemical is the study of the
geological and hydrological controls on the groundwater chemistry. The objectives of chemical
hydrogeological investigations are to determine the sources, concentration, and fate of dissolved
constituents within the physical framework of flow and transport. This chapter emphasises the
application of chemical hydrogeology, including the techniques, and procedures required to
investigate groundwater quality. For general background information on chemical hydrogeology
see Back and Freeze (1983).
In the early phases of chemical hydrogeology, research focused on developing analytical
techniques to describe the solute concentrations and chemical characteristics of water. The
approach is generally referred to as hydrochemical facies mapping. The next phase was to
provide an explanation of these observations by identifying the chemical reactions and hydro-
geological controls that are associated with the solute concentrations in water and with the
chemistry of the water. The approach is called reaction modelling. Now increased exploitation of
groundwater resources along with contamination of some supplies has brought a need for
greater understanding of the concentration limits and the variations in the behaviour of the
dissolved substances. We are now entering a phase in which research is directed toward
predicting the solute concentrations and chemical character of water at a specified point in some
future time. The ultimate goal is multicomponent geochemical transport modelling.
2.1.2 General approach
A schematic flow chart in Table 2.1 shows the objectives of a typical quality of groundwater
investigation. The ultimate goal, rarely achieved at present, is to predict the quality of
groundwater changes in space and time and the rates of these changes.
As in any other project, in a groundwater quality project, the desired objectives must e
clearly formulated. For research, these objectives are the testing, verification or refuting of one or
more hypotheses; for applied studies they are identification of the problem, and of the
controlling parameters and an evaluation of alternative ways of solving or coping with the
problem. Hydrogeologists are frequently required to assess the quality of groundwater on a

43
Groundwater studies
Table 2.1 Schematic flow chart for a chemical hydrogeological project
Needed for sampling and monitoring
(1) Preliminary interpretation of the hydrogeological system, and

(2) Reliable techniques for chemical and isotopic analysis
In order to prepare hydrochemical maps and diagrams that will:
DESCRIBE IDENTIFY
The spatial and stratigraphic Probable chemical reactions

distribution of dissolved
constituents Significant parameters to monitor
Relation of chemical Mineralogical controls

constituents to hydrogeo-
logical parameters Source and flow of groundwater
Source of contaminants
TO EXPLAIN TO MONITOR
Source of constituents Chemical changes in space
Concentration and Chemical changes over time

distribution of constituents
Rate of chemical changes
Hydrochemical heterogeneity
PROVIDES ALL OF THE ABOVE
A better understanding of the functioning of the hydrogeological system
regional basis. This requires a sampling and monitoring programme, that is significantly dif-
ferent spatially and temporally from a site-specific study. When hydrogeological and geo-
chemical principles are applied appropriately, the result is efficient, meaningful and therefore
successful groundwater assessments.
To be able to identify the controlling hydrogeochemical reactions, one must know the
present and past geological, hydrological, biological, soil chemical, meteorological, and human
factors that affect the chemistry of water (Fig. 2.1). The input of chemical or biological species in

44
Groundwater quality
Figure 2.1 Factors controlling groundwater composition
the unsaturated zone of the recharge area is largely controlled by atmospheric deposition, land
use (including vegetation and soil cover), the frequency, amount and duration of precipitation
and irrigation, the mineralogy of the soil, the air temperature, the soil gas/air exchange rate, and
transport properties. The quality of infiltrating surface water is controlled by the composition of
the surface water together with the temperature, the composition of the surface water bottom
sediments and the residence time within these bottom sediments. Factors that affect chemistry of
groundwater in the saturated zone are chemical reaction rate, residence time within the
saturated zone, and mineralogy of the rock matrix. Here, the residence time and flow path are
determined by factors like aquifer thickness, permeability, porosity and amount of recharge. The
phenomena of mixing of water from different areas, aquifers or confining beds, from seawater
intrusion or trapped saline water, or contaminants impose a hydrological control on the
chemical character of groundwater. Superimposed on all these primarily natural factors are
anthropogenic effects leading to chemical and physical stresses on the hydrogeological system
that may be dominant in some areas.
To be able to formulate and carry out regional studies, the following scientific and
technical aspects associated with hydrogeology must be understood:
the basic concepts of organic and inorganic chemistry;
the use of environmental isotopes and geochemical modelling;
principles of advective/dispersive transport and the coupling with reactions
the consequences of surface water/groundwater interaction or soil moisture/ground-
water interaction on groundwater quality;

45
Groundwater studies
the effects of land use on water quality; and application of geochemical principles to
regional aquifer systems;
salt water intrusion; and specific contamination problems in karst aquifers.
A publication that thoughtfully demonstrates the application of these many concepts and
principles for regional assessment of water quality is Alley (1993).
2.2 Basic principles of hydrochemistry
2.2.1 Precipitation/dissolution reactions
(i) Dissolution of neutral salts

Dissolution of a neutral salt is a reversible reaction in which the acid/base equilibria of water are
not affected. An example is the dissolution of gypsum:
CaSO4 2H2O Ca2+ + SO42
A typical example of dissolution of neutral salts is the dissolution of evaporites. These salts are
very soluble and their dissolution strongly increases the salinity of the groundwater. When
saline groundwater exfiltrates in closed semi-arid or arid basins precipitation is likely.
(ii) Weathering reactions
In these reactions primary minerals become dissolved or altered and secondary minerals may be
produced. The secondary minerals are closer to chemical equilibrium with the earth surface con-
ditions than the primary ones. Weathering reactions are associated with shifts in the acid/base
equilibria. Consider for example the dissolution of Ca-carbonate in carbonic acid:
CaCO3 + H2CO3 Ca2+ +2HCO3
Carbonate reactions are reversible. Note that dissolved CO2 is hydrated in groundwater and
exchange with the gas phase may be limited or impossible. CO2 in groundwater is therefore
indicated as H2CO3. Another example of weathering is the incongruent dissolution of
Al-silicates. At neutral pH, Al remains in the solid and the process is represented in a general
way as:
Aluminium silicate + H+ Me m+ + Aluminium mineral + H4SiO4
where Me m+ refers to metals such as Na, K, Ca, and Mg. Silicon dissolves in pore water and/or
remains in the solid. Weathering reactions involving silicates can be considered as irreversible
under earth surface conditions.
2.2.2 Redox reactions
In these reactions ions electrons are transferred from one species to another. Redox half-reactions
are used to describe the change of an element from the reduced state to the oxidised state or vice
versa. Consider for example the half-reaction for the reduction of nitrate to nitrogen gas:
NO3 + 5e + 6H+ N2 + 3 H2O

46
Groundwater quality
All elements that exist in more than one valence state in nature are susceptible to this type
of reaction. In order to obtain a true redox reaction which occurs in aquatic systems two half
reactions have to be combined. Redox reactions also affect the acidity of the system, since during
the reaction protons are transferred. These reactions may co-occur with precipitation/dissolution
reactions, yielding redox-controlled dissolution/precipitation reactions. For example, pyrite
contains two reduced species, Fe(II) and S(I), which may be oxidised to Fe(III) and SO4,
respectively. Oxidation of pyrite by dissolved oxygen may, therefore, be complete or incomplete,
depending whether or not Fe(II) is oxidised:
FeS2 + 3 O2 + 3 H2O Fe(OH)3 + 2 SO42 + 4H+
or
FeS2 + 3 O2 + H2O Fe2+ + 2 SO42 + 2H+
Many redox reactions in groundwater systems are microbially mediated and kinetically control-
led. The bacteria use redox-sensitive compounds as a source of energy for their metabolism. The
thermodynamic feasibility of a redox reaction is determined by the redox status of the system.
That status is expressed in terms of the redox potential, the Eh. The combination of Eh and pH
determine whether or not a redox reaction is feasible.
A somewhat special type of redox reaction in the subsurface is the degradation of organic
matter. Organic matter is unstable and is oxidised to CO2 by the oxidants present, or is degraded
to CO2 and CH4. Organic matter is often simplified as CH2O. The oxidation of organic matter by
SO4 is represented by:
2CH2O + SO42 HCO3 + HS + H2CO3
Methanogenesis is represented by:
2CH2O + H2O H2CO3 + CH4
The degradation of organic matter is commonly a sequence of reactions that produces a variety
of carbon compounds. The degradation usually induces inorganic reactions, and in many
subsurface systems it is the primary control for the redox status. Degradable organic matter
occurs in the solid state as a primary material, or it enters the system in the dissolved state. In
the latter case it may be the result of natural or anthropogenic processes.
2.2.3 Sorption reactions
(ii) Cation exchange reactions
Cation-exchange reactions are the result of charge compensation of negatively charged clay
minerals and organic matter. The occupation of the exchange complex is determined by the
aqueous composition together with the affinity of the exchange complex for the cations present.
The charge compensation occurs by means of an increased concentration of cations near the
surface, possibly accompanied by a decreased concentration of anions. The cation-exchange
reaction has been formulated in several ways. A common thermodynamically justified way to
represent the exchange-reaction between Na and Ca, for example, is:
2NaX + Ca2+ CaX2 + 2Na+

47
Groundwater studies
where X represents the solid phase exchanger with charge 1. The cation-exchange capacity
(CEC) indicates the total amount of exchangeable cations on the porous medium. It depends on
the type of clay minerals and the amount of organic matter, and it is also pH-dependent. Cation-
exchange happens as soon as the native pore water composition is displaced by water of another
composition. This happens, for example, during seawater/fresh water displacement associated
with seawater intrusion, etc.
(ii) Surface complexation
Surface complexation reactions are reactions in which species are transferred from the solid
surface to the aqueous phase and vice versa. They typically occurs for oxides, humic and fulvic
acids. Sites at the surface of the solids are protonated or deprotonated, for example:
S OH S O + H+
where S refers to a surface site. The protonation/deprotonation is strongly pH-dependent.

Additionally, dissolved metals and oxyanions become sorbed at the surface sites. If the amount
of proton transferred differs during sorption, several types of reaction happen. Consider, for
example, sorption of copper and arsenate, which are strongly favoured by Fe-hydroxide, as
examples:
S OH + Cu2+ S OCu+ + H+
and
S OH + AsO43 + 3H+ S OH2AsO4 + H2O
Sorption of metals and oxyanions is strongly pH-dependent. Sorption of oxyanions is strongest

at acid to neutral pH, and sorption of metals is strongest at neutral to alkaline pH. This
behaviour is related to increasing deprotonation of the surface sites with increasing pH, which
gives rise to a gradual change in charge from positive to negative.
2.2.4 Aqueous complexing
Aqueous complexing does not involve solid minerals. However, it indirectly influences reactions
with the porous medium. It gives rise to an increased concentration of species, since the free
species react together to form complexes. The solubility of Ca-carbonate is increased by the
forming of the aqueous Ca-bicarbonate complex:
Ca2+ + HCO3 CaHCO3+
In groundwater systems aqueous complexing reactions are instantaneous.
2.2.5 Gas transfer
Some of the gases dissolved in groundwater are important reactants with the solid matrix.
Consider for example the role of CO2 in dissolving carbonates and that of O2 as an oxidant of
reduced compounds. Gases become dissolved in pore water proportionally to the partial gas
pressure in the associated gas phase:
CO2g CO2(aq)

48
Groundwater quality
where subscripts g and aq refer to the gaseous and aqueous phases. Gases remain dissolved as
long as the sum of partial gas pressures does not exceed the hydrostatic pressure. Groundwater
can often be considered as closed to gas transfer, i.e., gases that react are not replenished and gas
that is produced remains in the dissolved state during groundwater flow. Exceptions to this rule
are groundwater near the water table and groundwater in hydrothermal or volcanic systems.
2.2.6 Ion filtration and osmosis
These reactions occur in clays and other fine-grained sediments in which the unchanged water
molecule can pass freely through the semi-permeable membrane but the anions are repelled by
the negative charge on the clay layers and the cations remain with the anions to maintain
electrical balance. It is important in large, thick sedimentary basins, in areas with a long
residence time and at interfaces between seawater and fresh water in clayey deposits.
2.2.7 Radioactive decay
Radioactive decay and fractionation of isotopes produce radioactivity in the groundwater and
alter the isotopic composition of the water. This is discussed in Chapter 9.
2.3 Acquisition of chemical data
2.3.1 Sampling procedures
Ideally, the sampling programme for a geochemical groundwater investigation will collect the
minimum number of samples required to have adequate three-dimensional spatial and strati-
graphic coverage of the area being investigated. The fundamental task is to obtain samples that
are representative, diagnostic, and characteristic of the aquifer and to analyse them with
minimal change in composition. The volatile and reduced compounds are the most sensitive to
shortcomings in the sampling and conservation procedures. Poor procedures that can render
unreliable results include obtaining samples of water standing in a well casing, or from a well
having cross flow, the use of unclean or reactive bottles, the lack of an airtight seal to prevent
aeration or degassing, the lack of adequate preservation of non-stable constituents, contamin-
ation of the samples with the materials used in well construction (casing, cement, drilling mud,
or other additives), the lack of proper filtration of particulate matter from the sample, and long
storage time.
Before starting the groundwater sampling, the chemists in the laboratory who will
perform the analysis, should be consulted for advice on sample size and the preservation tech-
niques to be used in the field. The concise handbook on analytical techniques for water
published by APHA-AWWA-WPCF is useful. Volatile compounds must be collected in gas
bottles. Samples for analysis of dissolved cations must be collected in plastic bottles, filtered
through a 0.45 m filter, and acidified to pH 12 to prevent precipitation after sampling and
sorption to the bottle.
The alkalinity (bicarbonate concentration) together with the pH should preferably be
measured immediately in the field, together with temperature and electrical conductivity,
since these two change rapidly after sampling. Aeration of the groundwater sample gives rise
to entrance of O2 and associated oxidation of dissolved reduced species, and dissolved CO2
escapes when the partial pressure of dissolved CO2 is higher than atmospheric. When sampling
for reduced species, special precautions must be taken in order to preserve the actual

49
Groundwater studies
concentrations. When sampling for dissolved gases degassing must be prevented. Positive
displacement pumps are to be preferred above suction type of pumps.
2.3.2 Groundwater analysis
(i) Inorganic constituents
An analysis of groundwater may range from a value for a single component to a long list of
inorganic, organic and biological measurements. The selection of the constituents for analysis is
determined both by the overall objective of the investigation and the specific purpose of the
chemical and biological analysis. Because of the complexity of many diverse problems for which
analyses may be useful, there is no standard analysis. For example, the suite of inorganic con-
stituents for geochemical prospecting is appreciably different from the constituents used to
monitor a landfill or to determine the effects of mineral diagenesis. However, the conventional
point of view has been that a standard inorganic analysis consists of determination of the four
major cations (calcium, magnesium, sodium, and potassium) and the three major anions (bicar-
bonate, sulphate, chloride) together with pH, electrical conductivity and temperature. If there is
anthropogenic contamination, nitrate must also be determined. Aluminium becomes increas-
ingly important when the pH is 6 or lower. Dissolved Fe(II) and Mn(II) may occur as major
species in anaerobic groundwater. Ammonium, F, PO4, H2S and CH4 are of secondary impor-
tance; their analysis is often crucial to resolve the evolution of groundwater quality along flow
lines. A criterion for an accurate analysis is that the difference between the sum of cations and
the sum of anions is less than 5% of the sum of these two sums. Note that this criterion only
checks the major ions, and does not guarantee that the minor species have been properly
analysed.
(ii) Organic constituents
Due to increasing anthropogenic contamination with organic microspecies (pesticides, petroleum

hydrocarbons, chlorinated hydrocarbons, etc.) it is becoming increasingly important to obtain
samples of groundwater for determination of organic compounds. The first step towards
obtaining insight into the status of groundwater with respect to organic compounds determin-
ation of the total amount of dissolved organic carbon (DOC). Common compounds that make
up DOC include fatty acids, humic substances, carbohydrates, amino acids and uronic acids.
Organic material is frequently present in rainfall, groundwater, streams, rivers, lake and sea-
water. A properly carried out investigation will explain how and why the DOC varies over time
and space and what processes have controlled these variations. Principles of organic geo-
chemistry of water are increasingly applied for this purpose.
DOC is of little diagnostic value when dealing with organic microcontaminants. There are
special analytical procedures for volatile and non-volatile microspecies. In the first instance, we
should distinguish three major groups of interest:
chlorinated hydrocarbons that are associated with dry-cleaning facilities, metal-
working industries, chemical industry and military areas,
petroleum hydrocarbons (including the BTEX compounds; benzene, toluene,
ethylbenzene and the xylenes) which are related to petrol stations, petrochemical
industry, crude oil and natural gas exploitation sites, and military areas,
pesticides, herbicides, fungicides, which represent an extremely broad group of
compounds mainly related to agricultural activities, but also used to maintain
railways, parks, golf courses, etc.

50
Groundwater quality
(iii) Bacteria, viruses and microbial activity
Microbiota in the subsurface are studied either to find out whether pathogenic bacteria and
viruses are present in groundwater or to determine the role of microbial activity in predom-
inantly redox processes. If possible groundwater must be sampled under sterile conditions, to
avoid any contamination from the earth surface. Note that during drilling of wells contamin-
ation with allochtonous microorganisms may occur.
There are routine techniques t for analysing pathogenic microorganisms. Isolation pro-
cedures and plate counts are often the basis for identifying and counting the bacteria present.
The role of microorganisms in controlling quality of groundwater is increasingly being
demonstrated, especially their potential to degrade organic contaminants. An excellent book
that addresses this topic is Chapelle (1993). Microorganisms and microbial activity have been
studied in two different ways: the microbial activity is characterised by means of incubations
and temporal analysis of the compound of interest, or the microorganisms themselves are
cultivated, isolated, etc. We must realise that the latter analyses are not completely represen-
tative of the subsurface microbial community, since cultivation is selective and most bacterial
species have not been a identified. Instead of studying groundwater it may be worth studying
the microorganisms in a solid matrix, however, the spectrum of groundwater microorganisms is
not identical to that of the solid matrix and biased results will be obtained when characterising
groundwater.
2.4 Evaluation of groundwater chemical data
2.4.1 General procedure
Chemical groundwater data are interpreted to find out where the ions come from, how they
reach their concentration, what is their form and behaviour, where they are going, and how fast.
The major activities are:
determining the mineral equilibrium characteristics of water using speciation models;
displaying the spatial distribution of chemical concentrations in order to understand
the hydrogeological controls on the chemical variations within an aquifer or chemical
differences between aquifers;
identifying the controlling chemical reactions;
interpreting water analyses in terms of hydrological, geological, and anthropogenic
controls;
chemical reaction modelling to predict chemical changes in the aquifer during
exploitation.
Groundwater quality data may be used to independently validate the hypotheses put forward to
explain the physical functioning of the groundwater system. Hypotheses about residence times,
flow paths and rates can be checked using hydrogeochemical data in combination with data
about isotopes. Chemical, mineralogical and lithological data about the aquifer is of great help
during interpretation of groundwater quality data. This information must be collected during
drilling of the wells, piezometers, etc.
2.4.2 Characterisation of groundwater quality
A first evaluation of the groundwater quality is whether the water is fresh, brackish or salt.
Nowadays the amount of dissolved salts is most commonly expressed in one of the
following ways:

51
Groundwater studies
The electrical conductivity (EC) of water expressed in S/cm or in S/cm (10 6 S/cm) for
fresh water, or in mS/cm (10 3 S/cm) for salt water; where S stands for Siemens
(formerly called mho), the inverse of the resistance expressed in ohm ();
Salinity, where total dissolved solids is expressed as parts per thousand of unit weight
of water, for example, normal seawater has a salinity of about 35;
Total dissolved solids (TDS) which is the residue on evaporation at 105C or 180F.
A classification of water based on TDS is given in Table 2.2.
Table 2.2 Nomenclature for water
Classification TDS (mg/l)
Fresh water1 < l,000 to 2,000

Brackish water 1,000 to 20,000
Saline water 20,000 to 50,000
Brine >50,000
1. Limits vary from one country to another, depending on the salinity of available water.
The next step is to evaluate the individual constituents. Only about a dozen of the somewhat
more than one hundred chemical elements are normally analysed in water samples.
Over the years, hydrogeochemists have been guided in their selection of significant cons-
tituents by the relative concentration of a particular element and its physiological, agricultural,
industrial, or geochemical significance. In groundwater (Table 2.3) the concentrations are
Table 2.3 Normal range of chemical composition (in mg/l, except pH) of groundwater, seawater
and rainwater away from the coast
Groundwater Mean ocean water a Rainwater b
pH 5.5 9 8.2 4.1 5.6

Ca 10 200 400 0.1 2
Mg 0.1 100 1,350 0.05 0.2
Na 1 300 10,500 0.1 1
K 0.1 20 380 0.08 0.3
NH4 05 <0.5 0.1 2.3
Fe c 0 10 0.01 0 0.2
Mnc 02 0.002 0 0.02
SiO2 10 30 6.4
HCO3 80 400 142 0
SO4 10 100 2,700 0.4 8
Cl 1 150 19,000 0.25 2
NO3 0 50 <0.5 0.3 4
F 0.1 2 1.3 0.03 0.1
Br <5 65
B <2 4.6
a. Hem (1986); b. Rainwater away from the coast (Royal Netherlands Institute of Meteorology; Appelo and Postma, 1993);
c. The redox status is different for the water types.

52
Groundwater quality
intermediate between rainfall and ocean water. Rainwater is not pure water and in some coastal
areas or desert areas the chemical load of the atmospheric precipitation can be a significant
contribution to the chemical character of groundwater.
The chemistry of rainwater depends on such factors as the windward distance from the coast or
salty lakes and soils, wind intensity, the period of rainfall within the storm, distance from cities
and industrial centres.
The chemical composition of mean ocean water (Table 2.3) is typical of the open ocean not
concentrated by evaporation nor diluted by fresh water from river flow or groundwater dis-
charge. In coastal aquifers, seawater intrusion can dramatically alter the chemical composition of
the groundwater and result in chemical processes within the saline water and zone of
dispersion. A classification based on salinity of water from such environments (Table 2.2)
demonstrates the rather arbitrary limits for fresh, brackish, and saline water and brines.
In addition to rainfall and seawater, the sources of ions are from the dissolution of
minerals, sorption reactions, and degradation of organic material in the unsaturated and satu-
rated zones. Tables 2.4, 2.5 and 2.6 present a brief summary of the sources of various chemical
species in groundwater. Speciation calculations should be performed to determine the dis-
tribution of the total concentration of a compound among free species and aqueous complexes.
It is the chemically active concentration of the relevant aqueous species that is calculated.
Table 2.4 Sources of major constituents
Major constituents > 5 mg/l Source
Calcium, Ca primarily from carbonates, gypsum, feldspars
Magnesium, Mg feldspars, olivine, pyroxene, amphiboles, mica,

Mg-calcite
Sodium, Na feldspars, evaporites, cation exchange, seawater,

industrial waste
Potassium, K feldspar, fertiliser, K-evaporites, glauconite
Silicic acid, H4SiO4, SiO2 silicates
Ammonia, NH4 pollution, degradation of organic matter, reduced NO3,

cation-exchange
Sulphate, SO4 dissolution of gypsum and anhydrite, oxidation

of pyrite, seawater, windborne fertiliser salts
Chloride, Cl windborne rainwater, seawater and brines, evaporite

deposits, pollution
Nitrate, NO3 atmospheric deposition, decay of nitrogen-fixing

plants, oxidation of ammonia or organic nitrogen,
contamination
Carbonate, CO2, HCO3, CO3 soil and atmospheric CO2, carbonate rocks, oxidation
of organic material, volcanic gases
Oxygen, O2 soil gas and atmosphere

53
Groundwater studies
Table 2.5 Sources of minor constituents
Minor constituents, 0.01 to 10 mg/l Source
Aluminium, Al clays, feldspars, amphiboles, micas
Boron, BO3 tourmaline, evaporites, sewage, seawater, volcanic

emissions
Fluoride, F fluorite, some silicates, volcanic emissions
Sulphide, H2S pyrite, reduced sulphate, oil field gas, volcanic

emissions
Phosphate, PO4 apatite, fertiliser, sewage sludge, degradation of

organic matter
Iron, Fe oxides, sulphides, carbonates and clays
Manganese, Mn oxides and hydroxides
Strontium, Sr carbonates
Organic acids, mostly humic and fulvic organic matter decomposition
Argon, Ar air
Methane, CH4 organic matter degradation under intense reducing

conditions
Nitrogen, N2 air, nitrate reduction
Table 2.6 Sources of trace elements
Trace constituents, generally <0.1 mg/l

and commonly below detection level Source
Arsenic, As arsenic insecticides, pyrite oxidation
Bromide, Br evaporites, seawater, rainwater
Chromium, Cr contamination
Iodide, I marine vegetation, evaporites
Lithium, Li silicate weathering
Uranium, U uranium dispersed minerals, mill tailings
Vanadium, V not well known
Zinc, Zn, cadmium, Cd sulfides, industrial waste, PO4-fertiliser salts,

sewage sludge
Determination of the distribution of species and their active concentrations is an iterative

process, because a set of mass-balance equations has to be solved together with a set of thermo-
dynamic mass action equations. After the active concentrations have been calculated, the
saturation state of groundwater with respect to minerals is calculated.

54
Groundwater quality
Conventionally, the saturation index is provided, which is defined as the log value of the
ratio of the ion activity product to the equilibrium constant of the mineral. For example, the
saturation index with respect to calcite is calculated as:
SI cc = log {[Ca2+][CO32]} / K c c
where SI cc is the saturation index with respect to calcite, K cc is the thermodynamic solubility
product of calcite, and [Ca2+] and [CO32] is the active concentration of Ca2+ and CO32,
respectively. Unsaturation for a mineral (SI < 0) indicates that the mineral will dissolve if
present. Saturation (SI = 0) indicates that thermodynamic equilibrium exists and the mineral
equilibrium may control the concentration of the constituents present in that mineral. Super-
saturation (SI > 0) increases the possibility of precipitation; non-equilibrium is the result of slow
kinetics, non-ideal behaviour or inhibition of precipitation.
2.4.3 Compiling groundwater quality data
It is generally convenient to use one or more of the many classification schemes and geochemical
diagrams available, in order to identify and depict the significant chemical characteristics of
water in various parts of the groundwater regime or the chemical differences between aquifer
systems (Schoeller, 1962; Hem, 1989; Custodio and Llamas, 1976; Back and Freeze, 1983; see
also Chapter 4). Chemical information gained from such diagrams is commonly plotted on
hydrogeological maps and cross-sections to illustrate the areal and stratigraphic distribution of
various water types and the change over time in the chemical character of groundwater. In order
to prepare hydrochemical diagrams, it is frequently necessary to convert analyses from one set
of units to other units; these conversion factors are listed in the Appendix A-1, Conversion
factors.
Concentration ratios are also very helpful when interpreting groundwater analyses in
terms of hydrogeochemical processes. Table 2.7 lists species ratios that are of general use,
together with the information that these ratios may provide. Note that these rules are based on
many groundwater quality studies; however, there are extreme environments that do not behave
according to these general insights. Molar ratios or equivalent-concentration ratios are often
more convenient than weight-concentration ratios to identify hydrochemical ratios. To be able to
accept or reject hypotheses about the ongoing processes, the information derived from the ratios
needs to be combined.
2.5 Process interpretation and modelling
2.5.1 Modelling
The next step after the groundwater analysis has been interpreted in terms of the origin of
groundwater and chemical evolution, is to quantitative model the processes proposed. In
chemical reaction modelling, the data and calculations are used to determine:
which chemical reactions have occurred,
the extent to which the reactions have proceeded,
the hydrogeological conditions under which the reactions have occurred,
which changes may occur over time,
the speed of such changes.
Here the calculations of mineral equilibria determine the saturation state of the water and
identify which minerals will tend to dissolve or precipitate. In addition to the normal chemical

55
Groundwater studies
Table 2.7 Concentration ratios and their meaning in terms of process control
Ratio Application
SO4/Cl Seawater has a typical weight ratio of 0.14. When both constituents originate
from seawater smaller ratios suggest SO4-reduction. Higher ratios suggest
additional sources of SO4 and/or additional processes such as dissolution of
gypsum or oxidation of pyrite.
Na/Cl The seawater weight ratio is 0.55. Higher values imply Na-desorption during
freshening, if paleohydrological studies have revealed that marine water was
once present. Higher ratios can also be indicative of silicate weathering.
Mg/Cl The seawater weight ratio is 0.071. Deviations from this value together with
deviations for Na/Cl suggest that there has been cation-exchange for the
appropriate hydrogeological environment. Silicate weathering also gives rise to
high ratios.
Na/HCO3 If seawater/fresh water displacements can be excluded, high ratios indicate

substantial weathering of Na-feldspar or other Na-silicates.
Ca/HCO3 To identify Ca-carbonate dissolution (check with saturation index with respect
to calcite and other Ca-carbonates) or Ca-silicate weathering.
Ca/SO4 To identify gypsum dissolution.
Ca/Mg To identify dissolution or precipitation of dolomite, calcite, etc. (check with

saturation index with respect to dolomite, calcite, etc.). Deviations from the
seawater ratio occur in coastal regions due to cation-exchange, albeit combined
with carbonate reactions.
(Ca+Mg)/(Na+K) To identify aquifer lithology: importance of carbonate reactions versus silicate

weathering.
Br/Cl To differentiate between brines and seawater.
Fe/SO4 To identify oxidation of Fe-sulphide.
HCO3/sum anions To distinguish between weathering reactions and input of dissolved species
from the surface in the recharge area.
NO3/sum anions To identify the extent of anthropogenic contamination due to e.g. agricultural
activities.
(H+Al)/sum cations To indicate the extent of acidification (use equivalent concentrations).
F/Mg To identify biotite dissolution as the source of F.
constituents, isotopes have proved extremely useful in groundwater quality studies. Isotopes are
atoms of an element with the same number of protons but different numbers of neutrons in their
nucleus. The distribution of isotopic species in water provides additional information on sources
of groundwater, on flow paths and mixing, and on chemical reactions and sources of ions.
Isotopes measured in water include: 18O, 2H (or deuterium, D), 3H (or tritium, T), 3He, 13C, 14C,
34S, 15N, 37Cl and 234U. See Chapter 9 for a discussion of isotope hydrology. The ultimate
objective of a groundwater quality study, including isotope chemistry, is to provide chemical

input for the mass transport equation in order to predict concentrations of organic and inorganic

56
Groundwater quality
constituents at some point in space and time. Several hydrogeochemical software models are
freely available via internet, e.g., the U.S. Geological Survey offers a series of models at:
http://water.usgs.gov/software/geochemical.html.
Two types of approaches are distinguished in hydrogeochemical modelling and in other
types of hydrological modelling as well (Fig. 2.2): forward modelling and inverse modelling. In
forward modelling we define an initial state of the system studied and we impose a series of
processes on that system. The resulting state is calculated by mathematical quantification of the
processes imposed. The result is thus one unique state of the system. In inverse modelling we
consider two (or more) states of the system studied and from the difference between these two
(or more) statuses we quantify the extent of processes that are assumed to happen. The different
states of the system must of course be linked in some way, e.g., the second state of the system is
a groundwater that lies along a flow line downstream from the first state of the system.
Figure 2.2 Illustration of the principles of forward and inverse modelling
Both forward and inverse hydrogeochemical modelling require a proper understanding of

the ongoing hydrogeochemical processes. A conceptual model of the system and the likely
reactions in that system is thus necessary before any mathematical modelling can be performed.
The quality of a mathematical model is ultimately linked to the quality of the modellers model

57
Groundwater studies
concept: any ill-defined model concept will lead to a wrong mathematical model and con-
sequently to wrong model results. For groundwater systems we have to deal with:
thermodynamic constraints;
mineralogical constraints (including non-mineralogical reacting solids);
kinetic limitations;
hydrological transport limitations.
The mineralogical composition of the system studied determines the type of reactions that may
happen, and the thermodynamic constraints will determine the direction of the chemical pro-
cesses together with the maximum amount of mass transfer of chemical species from one phase
to another. Kinetic limitations will determine the actual extent to which the potentially occurring
reactions have happened within the time span that is available for the system studied and
whether or not a metastable thermodynamic state of the system is likely to be reached. The latter
is indicated by the Ostwalds kinetic step rule: a system follows the kinetically most favourable
path instead of the thermodynamically most feasible path. Kinetic limitations thus determine
the actual amount of mass transfer. For groundwater, the kinetic limitations have to be
compared with the travel times obtained from the hydrogeological study.
Forward hydrogeochemical modelling is the as prediction of an aquatic system from an
initial solution combined with a series of postulated reactions. Thermodynamic principles are
used in hydrogeochemical software models and forward hydrogeochemical modelling is thus
implicitly constrained by thermodynamic considerations. However, it is possible to violate
thermodynamic constraints if the modeller incorporates kinetics, e.g. by defining zero-order
reaction steps in the model. To what extent this may happen depends on the features of the
model used. Thermodynamic constraints can also be deliberately violated in order to consider
some non-equilibrium or irreversible reactions. Thermodynamic databases may be deliberately
changed into databases that need to be considered as non-equilibrium databases. One may, for
example, decouple the thermodynamic relationships for N-species (NH4, NO3, NO2 and N2),
because oxidation of ammonium to nitrate and reduction of nitrate to nitrogen gas are usually
kinetically controlled by irreversible reactions. Generally, redox reactions may have to be
modelled as irreversible thermodynamic processes instead as reversible thermodynamic
processes. Here, we should remember that kinetically controlled irreversible processes must
always be thermodynamically feasible, even those that are microbially controlled. Bacteria may
catalyse geochemical reactions but they cannot change the direction of geochemical reactions!
Inverse hydrogeochemical modelling uses existing data about aqueous systems and
calculates the net mass transfer for a series of reactions proposed. Mass balance constraints are
the only mathematical constraints that are explicitly included in inverse hydrogeochemical
modelling. Kinetic, thermodynamic and mineralogical considerations are implicitly included via
the modellers concept of which reactions are involved. The input data will at least be chemical
data about the aqueous compositions, but may include isotope data. The inverse problem is not
constrained by thermodynamic considerations: the models output is the net transfer of a master
species into or out of the aqueous phase. This net mass transfer can be the difference between
the dissolved amount of one mineral and the precipitated amount of another mineral. The
modeller is thus responsible for controlling the thermodynamic feasibility. One way to do this is
to perform a post hoc forward modelling for the initial composition, including thermodynamic
constraints with respect to mineral equilibria, etc. The outcome of an inverse modelling can be a
variety of possible reaction pathways. A major shortcoming of inverse models is that surface
reactions are not adequately included in the mathematical model: the extent to which sorption
reactions occur between two observation points is controlled by the total sorption capacity
between these points and the selectivity coefficients, whereas the extent to which precipitation/
dissolution can occur is independent of the amount of solid present as long as more is present
than can be dissolved (see Section 2.2). Forward hydrogeochemical transport modelling circum-

58
Groundwater quality
vents this problem and as soon as sorption reactions are involved is thus actually superior to
inverse hydrogeochemical modelling
2.5.2 Coupling of hydrochemical reactions
For groundwater chemistry it holds true that the behaviour of minor solutes is partly deter-
mined by the behaviour of major solutes. Two concepts deserve attention in how they interlink
different types and reactions, and couple the behaviour of minor species and that of major
species. They are the common-ion effect and sequential equilibrium control. The common-ion effect
refers to two reactions in which a similar species is involved. For example, carbonate equi-
librium and cation-exchange:
CaCO3 + H Ca2+ + HCO3

and
Ca X2 + 2 Na+ 2 NaX + Ca2+
Calcium is the common ion for these two reactions and the two reactions are coupled to each
other by means of the activity of Ca2+. Combining these two reactions results in salinisation of
an aquifer in which Ca-carbonate precipitates and freshening of the aquifer if Ca-carbonate
dissolves.
Sequential equilibrium control is comparable to the common-ion effect. Now, a major
species dictates the behaviour of a minor species in aqueous concentration by means of equi-
libria for two solid phases that have an identical co-ion. An example is the parallel behaviour of
the weak carbonate and phosphate acids when equilibrium is attained for siderite and vivianite:
FeCO3 + H Fe 2+ + HCO3
and
Fe3(PO4)2 . 2H2O + 4 H+ 3 Fe 2+ + 2 H2PO4 + 2 H2O
The thermodynamic dissociation constants can be combined after eliminating the activity of
Fe2+, which is the identical co-ion. Combination results in:
[HCO3]3 Ksi [H2O]

2
2=K +
[H2PO4] viv [H ]
The behaviour of the carbonate species thus determines the phosphate concentration for such a
system. The essential difference between the common-ion effect and sequential equilibrium
control is that for the latter, major processes dictate minor processes and, associatedly, major
solutes dictate minor or trace solutes, whereas for the former the processes balance each other.
2.5.3 Redox zoning
The redox status of an aquifer system deserves special attention for a combination of reasons:
the types of minerals present are related to the prevailing redox conditions,
degradation rates for microorganics such as pesticides often depend on the redox
status, because these rates depend on the oxidants involved,
aqueous speciation of redox-sensitive species such as arsene is related to the redox
status.
The major redox-sensitive elements in groundwater systems are oxygen, hydrogen, nitrogen,

59
Groundwater studies
manganese, iron, sulphur and carbon. The redox half-reactions for these elements in aqueous
systems are:
O2 + 4 H+ + 4e H2O
NO3 + 6 H+ + 5e N2 + 3 H2O
MnO2 + 4 H+ + 2e Mn2+ + 2 H2O
Fe(OH)3 + 3 H+ + e Fe2+ + 3 H2O
SO42 + 10 H+ + 8e H2S + 4 H2O
H2CO3 + 8 H+ + 8e CH4 + 3 H2O
2 H+ + 2e H2
in which e refers to an electron, which does not occur as free aqueous species in aqueous
systems. Note that all reduction reactions consume protons and, consequently, all oxidation
reactions produce protons. Also note that the oxidised forms of Fe and Mn occur in the solid
state, which reflects the insoluble nature of these oxides at neutral pH.
Infiltrating groundwater at the water table is usually oxic and redox potential decreases in
downstream direction due to the consumption of oxidants present by reductants present in the
subsurface. When the supply of reductant is not extremely high the oxidants are consumed in
order of decreasing energy yield. A characteristic sequence for the disappearance of dissolved
oxidants and the appearance of dissolved reductants therefore develops along a flow line.
Figure 2.3 presents the redox clines for the above redox half-reactions in a pH-pe diagram.
We can classify groundwater on its redox status based on the dissolved redox-sensitive species.
Table 2.8 present a classification scheme that is valid for pH 7. When using this scheme note that
deviations may occur at low pH when the Fe-redox cline and the N-redox cline lie close to each
other. The Fe-redox cline is also sensitive to the type of Fe-oxyhydroxide present. Crystalline Fe-
oxyhydroxides are much less soluble than amorphous ones. The redox cline shifts in response to
lower redox potential for crystalline oxyhydroxides. Finally, the redox clines for S and C lie close
to each other. Methanogenesis may happen before SO4 reduction or parallel to it, depending on
whether the reaction products are removed by secondary precipitation reactions (e.g. Fe-
sulphide precipitation). This lowers the active concentrations of the products and keeps the
forward redox reaction thermodynamically favourable.
Table 2.8. Redox classification of groundwater (x refers to a concentration above the detection limit,
which usually lies around 110 g/l)
Redox status O2 NO3 SO4 Mn Fe H2S CH4
Oxic x x x
Suboxic x x
Mn-anoxic1 x x x
Fe-anoxic x x x
SO4-reducing x x x x
Methanogenic x x x x
1. Due to kinetic limitations NO3 and Mn2+ are often jointly observed in the absence of both dissolved O2 and Fe.

60
Groundwater quality
Figure 2.3 PH/redox potential diagram with redox clines for the major redox couples in groundwater
2.6 Groundwater supply and health
Groundwater is abstracted from aquifers for several reasons: drinking water for man or live-
stock, process water in food-, beverage- and table luxuries industries, irrigation water for agri-
culture, process water in industries, cooling water and other low-quality uses. In the long term
technical problems may arise during the withdrawal of groundwater. The problems are a con-
sequence of specific, reactive species in groundwater together with improper installation of the
technical facilities. Some characteristics of malfunction are readily observable and guidelines for
their possible causes and standard remediation are given in Table 2.9.
Groundwater is used as drinking water in many areas in the world. However, some
groundwaters contain such high concentrations of species that their intake by man is undesir-
able. Toxic concentration levels present in groundwater not only interfere with human health
directly via consumption of drinking water, but also indirectly via consumption of food
(Fig. 2.4). Irrigation water is one of the factors which, together with geochemical and soil factors,
determine the trace element composition of crops. Here, a particular problem is that some
elements are more toxic to humans than to plants: in healthy plants these elements may
accumulate to levels that threaten human health.
Mercury, lead and cadmium are particularly toxic to humans (mammalian toxic), whereas
copper, nickel and cobalt are more toxic to plants (phytotoxic). The trace element composition of
leguminous crops, pulses and cruciferous crops fluctuates more widely than that of cereal
grains. The variability for a specific crop also varies for the different elements. No simple,

61
Groundwater studies
Table 2.9 Guidelines for remediation of groundwater withdrawal systems
Observation Possible causes Standard remedial practice
i. Rusty stains in High concentration Water aeration and pH

clothes and/or plumbing; of dissolved iron correction followed by
Water is red or brownish settling and filtration
when tap is opened or
after chlorination.
Black stains High concentration of Mn is more difficult to
dissolved manganese remove than Fe
ii. Hard, white or If soluble in HCl, caused Water softening, but may be
brown encrustations in by calcium carbonate or expensive. Acidification is
pipes and fittings iron carbonate effective but may cause
corrosion. Avoid the loss of
dissolved gases, rapid cross-
section reductions, excessive
well drawdown. Periodic
cleaning advisable. High water
velocity and constant tempera-
ture delays the carbonate
deposition. Removal of iron
may be effective
As above If not soluble in HCl, Difficult to prevent. Avoid

combination of iron carbo- evaporation, temperature
nate and silicates. High changes and sediments that
silica content (gypsum can accelerate the deposition
encrustations are rare)
iii. Loose sediments Particulate matter Improve the well or correct.

and foreign matter in the screen and tube. Well
pipes, elbow bends and completion or well develop-
tanks ment may have been inap-
propriate. Screen slots may
be too wide or the gravel pack
may not filter properly. Avoid
intermittent operation as much
as possible
Clay-like matter Clay Improve the well or eliminate

some portions of the screen.
Prevent the entry of water
that produces clay defloccul-
ation
Yellow to red flocculant Iron hydroxide Treat for iron as in part (i)
or black filamentous
organic matter
iv. Metallic corrosion Acidic water, high sulphide Correct the pH with lime
in pipes, fittings, and tanks content, high free CO2

62
Groundwater quality
Table 2.9 (contd.)
Observation Possible causes Standard remedial practice
v. Corrosion and Acidic water Correct the pH with lime

alteration of concrete,
cement and rocks High salinity Avoid evaporation and pro-
tect the surface from being
alternately wet and dry
vi. Water turbidity Clay Correct the well. See item (iii).
High silica Difficult to treat.
Gas bubbles Purge the pipes, avoid air
suction
vii. Water taste:
- Salty Excessive salinity Change water sources
- Rusty Excessive iron and/or
manganese See item (i)
- Astringent Excessive sulphate Change water sources
- Earthy High hardness Softening
- Tasteless Not fresh. Lack of dis-
solved oxygen Aeration
- Tingling High COs content Aeration and/or pH
correction.
- Disagreeable Hydrocarbons, chlorinated,
hydrocarbons, chlorophenols
which may appear after
chlorination Adsorption on charcoal
viii. Water odour, most

noticeable in low,
unventilated areas:
- Rotten egg Sulphides Aeration
- Petrol Hydrocarbons Adsorption on charcoal
ix. Water colour:

- Reddish Iron Treat for iron as in item (i).
Avoid air. Purge the pipes
- Blackish Manganese Same as above
- Yellow Hexavalent chromium in Difficult to treat. Reducing
toxic quantities (greater agents and filtration
than 0.5 mg/l)
- White Small air or gas bubbles Store the water in a quiescent
state
x. Warm water Deep well and/or volcanic

activity Cooling tower
xi. Retarded plant High salinity Change water source

growth, brown leaves Excess sodium Add lime or gypsum
Excess boron Change water source
Dissolved methane (rare) Aerate

63
Groundwater studies
straightforward relationships are known between the composition of irrigation water and the
trace element composition of crops.
The toxic concentrations in groundwater are the result of either natural processes or of
anthropogenic contamination. Groundwater with natural above-normal concentrations typically
occurs in areas with hydrothermal or volcanic activity and in areas with ore deposits.
The opposite of concentration excess is concentration deficiency: drinking water is an
important source of various elements for humans, several of which are essential for human
health. Essential ions for human health are Na, K, Ca, Mg, Cl, SO4, HCO3, NO3, F, PO4, Fe, Zn,
Cu, Mn, I, Se, Co, Cr and Mo and possibly also Si, Li, B, V, Ni, As, Sr, Nb and Sn.
Shortage of iodine, selenium and fluorine has been reported to cause deficiency-related
health problems. These problems arise in rural areas in developing countries, where the
inhabitants depend on local food and small-scale drinking water facilities. Iodine deficiency is
particularly associated with remote mountainous areas away from the sea.
Out of the two series of essential ions, Se, As, and Cr are also classified as character-
istically hazardous by the U.S. Environmental Protection Agency together with Ba, Cd, Pb, Hg
and Ag. It should be realised that the toxicity of redox-sensitive elements depends on the redox
status of that element. For example, Cr(VI)O42 is a human carcinogen, but Cr3+ is not. The
World Health Organisation (1993) has defined guideline values for drinking water. Here, a dis-
tinction is made between species that have an adverse effect on health and species that may give
rise to complaints from consumers.
Table 2.10 lists the guide values for inorganic constituents. Ideally, microorganic
constituents should not be present.
Figure 2.4 Exposure route in groundwater of dissolved species harmful to man

64
Groundwater quality
Table 2.10 WHO guideline values (in mg/l) for species in drinking water (1993)
Species Health significance Species Consumer complaints
Antimony (Sb) 0.005 P 1 Aluminium (Al) 0.2

Arsenic (As) 0.01 P Ammonium (NH4) 1.5
Barium (Ba) 0.7 Chloride (Cl) 250
Beryllium (Be) NAD 2 Iron (Fe) 0.3
Boron (B) 0.3 Sodium (Na) 200
Cadmium (Cd) 0.003 Sulphate (SO4) 250
Chromium (Cr) 0.05 P Sulphide (H2S) 0.05
Copper (Cu) 2P Zinc (Zn) 3
Fluoride (F) 1.5 TDS 1,000
Lead (Pb) 0.01
Manganese (Mn) 0.5 P
Mercury (Hg) - total 0.001
Molybdenum (Mo) 0.07
Nickel (Ni) 0.02
Nitrate (NO3) 50
Nitrite (NO2) 3P
Selenium (Se) 0.01
Uranium (U) NAD
1. P is provisional value.
2. NAD means no adequate data to permit recommendation of a health-based guideline value.
ALLEY, W. M., (ed.), 1993. Regional Ground-Water Quality. Van Nostrand Reinhold, New York,
634 pp.
APHA-AWWA-WPCF. Standard Methods for the Examination of Water and Waste Water. American
Public Health Association, Washington, D.C., (publication office).
APPELO, C. A. J.; POSTMA, D., 1993. Geochemistry, Groundwater and Pollution. A. A. Balkema,
Rotterdam, 536 pp.
APPLETON, J. D.; FUGE, R.; MCCALL, G. J. H., (eds.), 1996. Environmental Geochemistry and Health.
Geol. Soc. Spec. Publ. (113).
BACK, W.; BAEDECKER, M. J.; WOOD, W. W., 1993. Scales in Chemical Hydrogeology: A Historical
Perspective; In: W. M. Alley (ed.), RegionaI Ground-Water Quality. Van Nostrand Reinhold,
New York, pp. 11129.
BACK, W.; FREEZE, R. A., (eds.), 1983. Benchmark Papers in Chemical Hydrogeology. Hutchinson and
Ross, Stroudsburg, Pa, 416 pp.
CHAPELLE, F. H., 1993. Ground-Water Microbiology and Geochemistry. John Wiley and Sons, 424 pp.
CUSTODIO, E.; LLAMAS, M. R., 1976. Hidrologa subterrnea. Ediciones Omega, Barcelona, 2350 pp.
DEUTSCH, W. J., 1997. Groundwater Geochemistry. Fundamentals and Applications to Contamination.
Lewis Publ., Boca Raton, 221 pp.
DOMENICO, P. A.; SCHWARTZ, F. W., 1990. Physical and Chemical Hydrogeology. John Wiley and Sons,
824 pp.
DREVER, J. I., 1997. The Geochemistry of Natural Waters. Surface and Groundwater Environments.
Prentice Hall, Upper Saddle River, NJ, 436 pp.

65
Groundwater studies
FETTER, C. W., JR., 1994. Applied Hydrogeology. Prentice Hall, Upper Saddle River, NJ, third
edition, 691 pp.
, 1993. Contaminant Hydrogeology. Macmillan Publishing Co., New York, 458 pp.
FREEZE, R. A.; CHERRY, J. A., 1979. Groundwater. Prentice Hall, 604 pp.
HEM, J. D., 1985. Study and Interpretation of the Chemical Characteristics of Natural Water. U.S. Geol.
Surv. Water Supply Paper 2254, third edition, 263 pp.
LANGMUIR, D., 1997. Aqueous Environmental Geochemistry. Prentice Hall, Upper Saddle River, NJ,
600 pp.
LICHTNER, P. C., STEEFEL, C. I.; OELKERS, E. H., 1996. Reactive Transport in Porous Media. Reviews
in Mineralogy, Vol. 34. Miner. Soc. of America, Washington, D.C., 438 pp.
MATTHESS, G., 1982. The Properties of Groundwater (translated from the original German by John C.
Harvey), Wiley-Interscience, 406 pp.
Parkhurst, D. L.; Plummer, L. N., 1993. Geochemical Models. In: W. M. Alley (ed.), Regional
Ground-Water Quality. Van Nostrand Reinhold, New York, pp. 25594.
PLUMLEE, G. S.; LOGSDON, M. J., (1999). The Environmental Geochemistry of Mineral Deposits.
Part A: Processes, Techniques, and Health Issues. Reviews in Economic Geology, Vol. 6A. Soc.
of Econ. Geologists, Littleton, CO, 371 pp.
SCHOELLER, H., 1962. Les eaux souterraines. Mason et Cie, Paris, France.
THURMAN, E. M., 1985. Organic Geochemistry of Natural Water. W. Junk Publishers, 497 pp.
WORLD HEALTH ORGANISATION, 1993. Guidelines for Drinking-Water Quality. Volume 1. Recommen-
dations. WHO, Geneva, second edition, 188 pp.

66
3 G ro u n d w a t e r n e t w o r k s
and observation methods
3.1 Introduction
A groundwater monitoring network is an organised system for the continuous or frequent

measurement and observation of the actual, dynamic state of the underground environment,
often used for warning and control (definition adapted from UNESCO, 1992). Monitoring of
groundwater is essential for the characterisation of groundwater systems and a key activity
in defining sustainable environmental development of such systems. Hydrological processes are
highly variable in space and time making data collection over all scales difficult and expensive,
but groundwater systems can only be identified if ample hydrological data in space and time
have been collected.
To understand the prevailing system in a region both physical and chemical data need to
be systematically collected.
There should be no substantial use of groundwater resources (e.g. for drinking water,
irrigated agriculture, industry) unless its exploitation is based on a sound groundwater man-
agement plan, which is supported by a continuous effort of collecting hydrologic data from
observation networks. Then, possible detrimental effect of exploitation, such as the drying up of
springs or ecologically relevant wetlands, as well as the impact of human activity on the ground-
water resources (e.g. pollution or recharge reduction) can be recognised at an early stage of the
groundwater development.
Groundwater monitoring networks have been extensively developed in the humid-
temperate, mid-latitude, industrialised countries, although methods for the optimal design of
networks are still under development. Furthermore, more sophisticated observation instruments
are being developed to reduce monitoring costs or to collect more specific data. On the other
hand, in many developing countries data from groundwater monitoring networks are woefully
lacking, despite the great need for these data to manage the sometimes scarce groundwater
resources and the high water demand resulting from population increases. This is why UNESCO
promotes the elaboration of groundwater monitoring procedures, including the design and
analysis of networks and the description of observation methods, in the framework of the Inter-
national Hydrological Program (IHP-UNESCO, 1990).
The first operational procedures for hydrological data collection in networks were
proposed by Langbein (1954). Over the years many good reviews of groundwater monitoring
networks and observation methods have been published. Brown et al. (1972), Bachmat (1989)
and WMO (1994) give detailed guidelines on the design of groundwater monitoring networks
and the observation methods to be used. Other relevant information is provided by Moss (1982;
1986), Nielsen (1991) and OPEA (1993). It is not the intention of this chapter to cover in full all
the basic aspects that are well documented in the publications mentioned above. In this chapter
a summary and an update is presented, which is mainly based upon a recent UNESCO
document (Van Lanen, 1998).

67
Groundwater studies
3.2 Place of monitoring networks in groundwater management
Groundwater management should rely on hydrological data of the actual state of the ground-
water system. In this context two different types of groundwater monitoring activities should be
distinguished, namely background monitoring and specific monitoring.
3.2.1 Background monitoring networks
Background monitoring, or primary, networks are introduced because of the need to start
monitoring before significant human interference occurs. Background monitoring is applied to
large areas without significant human interference (low technology). Its objective is to identify
the actual state of the aquifer system. It establishes the initial conditions prior to significant
groundwater development and gives background data for future discussions on possible over-
exploitation.
Before any groundwater monitoring can start, the objectives to be derived from the
national land and water policy must be clear. In the upper block of Figure 3.1 the main steps of
the pre-monitoring investigations are presented.
Before a monitoring system can be designed , the groundwater flow system in the area
must be understood and described as far as existing data allow. These activities start with the
collection of time-independent data, e.g. about geological framework, hydraulic properties,
boundary conditions and hydrochemistry resulting in a definition of the prevailing groundwater
flow systems. It is likely that at the start of the background monitoring phase knowledge is
still insufficient for a full understanding of groundwater systems, because an objective of this
monitoring is to improve the knowledge about the prevailing groundwater systems, but similar
regions may have been investigated, which might allow transfer of knowledge to the region to
be monitored (Fig. 3.1).
After the system identification phase, a first version of the groundwater monitoring net-
work is designed (middle block of Figure 3.1). Systematic measurements of heads and chemical
composition of groundwater in existing and abandoned wells should start. The groundwater
monitoring should be supported by the collection of meteorological data, data on the vadose
zone, and spring flow and stream flow data in similar networks. After the first or second year of
systematic data collection, the initial version of the background monitoring network should be
thoroughly analysed. The earlier defined version of the groundwater system and associated
conceptual model should be refined. Statistics and geostatistics can help to improve the
sampling density and frequency. It is likely that the analysis based on the improved knowledge
of local hydrological phenomena of the region itself will lead to the first version of the
monitoring network being refined to provide better representation of the specific characteristics
of the groundwater systems in the area of interest (Fig. 3.1).
The refinement of the background monitoring network should be a continuous process of
analysing incoming data, refinement of the description of the groundwater system and
subsequent modification of the network. It is likely that the network will need to be modified
more than once, especially in regions with extreme meteorological conditions, such as the arid
and semi-arid or arctic areas. Background monitoring in these areas is a long-term effort.
3.2.2 Specific monitoring networks
Specific or secondary monitoring networks follow what happens in the underground envi-
ronment when it is substantially exploited for particular purposes or when ecologically relevant
wetlands, springs or streams are expected to be affected or polluted by human activities. Specific
monitoring characterises the transient state of the aquifer and acts as an early warning for over-

68
Groundwater networks and observation methods
Figure 3.1 General layout of the groundwater monitoring procedure (Van Lanen, 1998)
exploitation or pollution. The monitoring is restricted to those areas where effects are expected;
for example, because of significant abstraction and a more accessible aquifer. Potential problems
are the reduction of spring flow, falling groundwater heads, falling well yields, deterioration of

69
Groundwater studies
water quality (including sea water intrusion or upconing of brackish or salt water) and land
subsidence.
A specific monitoring network should be set up after it has been decided to develop
groundwater resources in a particular region or when contamination of the water resources is
expected. The specific network should be designed on the basis of a comprehensive analysis of
background monitoring data. The conceptual model of the groundwater flow system is replaced
by a numerical one (lower block of Figure 3.1). This groundwater simulation model must specify
the consequences of different abstraction scenarios or remediation measures in terms of ground-
water heads, groundwater flow lines, residence times, changes in recharge conditions (e.g.
induced recharge), chemical composition, and spring flow and streamflow. Subsequently, the
specific monitoring aspects can be formulated, e.g. the boundary of the affected area, type of
hydrological variables to be monitored, and sampling density and frequency. A specific ground-
water monitoring network is likely to require modification if the collected data show that the
response of the groundwater flow system differs from the simulated one. An example is
presented in Figure 3.2. Similar to background monitoring, specific monitoring needs continuous
efforts in terms of the collection of data, data analysis and refinement or redefinition of the
monitoring network.
The specific monitoring should be integrated into the background monitoring efforts
(Fig. 3.1). Eventually the ideal situation is to have a background monitoring network that covers
the whole country and specific monitoring networks in regions where groundwater resources
are significantly exploited or where pollution occurs (nested monitoring networks).
Figure 3.2 Adaptation of the first version of a specific monitoring network after differences were observed
between monitored and modelled effects (Lloyd, 1998)
3.3 Hydrological variables
The essence of a groundwater monitoring network is the measurement and observation of the
state variables of the groundwater body itself. The dynamic state of the groundwater system can
be monitored through observing:

70
depth to water table (x, y, t)1;

chemical composition of phreatic groundwater (x, y, z, t);
piezometric heads (x, y, t) of each aquifer, or
heads (x, y, z, t) of each aquifer unit if vertical differences occur;
chemical composition of deep groundwater (x, y, z, t) in each aquifer;
spring flow (x, y, t);
well yields (x, y, t); and
chemical composition of abstracted groundwater (x, y, t).
A first analysis of the chemical composition should preferably be carried out in the field.
Standard measurements that characterise the physical-chemical composition of groundwater,
e.g. temperature, dissolved oxygen (DO), redox potential (Eh), hydrogen concentration (pH) and
electrical conductivity (EC).
Furthermore, in a multiple aquifer system attention should also be paid to vertical
differences, both in aquifers and aquitards. Carrillo-Rivera (1998) showed that in Mexico City
subsidence caused by groundwater abstraction cannot be adequately understood unless the
vertical head differences in the upper aquifer are monitored.
Groundwater monitoring networks need to be complemented by other networks moni-
toring streamflow and hydrological variables for the determination of groundwater recharge
(Fig. 3.3). Unless this is done, the state of a groundwater system cannot be understood adequa-
tely. WMO (1994) systematically presents acquisition and analysis techniques for precipitation,
evapotranspiration and soil moisture to determine groundwater recharge. Lerner et al. (1990)
and Simmers (1997) give an extensive review of recharge estimating techniques and associated
collection of hydrological data for different types of recharge (e.g. precipitation recharge, river
recharge). A review of methods to measure streamflow is presented by Boiten (1993) and WMO
(1994). In Figure 3.4 the average results of integrated long-term monitoring and subsequent
modelling are presented for three different catchments in West and Central Europe. These results
were used to identify the prevailing groundwater systems.
The Hupsel catchment in the Netherlands has a small groundwater storage and a quick
response of the streamflow to excess precipitation. A pronounced summer drought occurs. The
Gulp basin consists of chalk with a large groundwater storage; the streamflow shows a
smoothed and delayed response on excess precipitation. The Cern Desn catchment, which is
located in the Jizera Mountains in the Czech Republic, has a small groundwater storage (weath-
ered granite overlying fractured granite) but a relatively constant discharge because of the
evenly distributed precipitation. The high peak in May-June is from snow melting.
The meteorological, vadose zone, groundwater and streamflow monitoring networks
should be integrated from the outset (Moss, 1986). Commonly, more than one organisation is
responsible for the acquisition of the data so coordination must be good. For example, locations,
monitoring frequency, accuracy, data processing, data transfer and publication, in the different
networks need to be harmonised.
3.4 Network design
Ideally the design of network density and sampling frequency would be based on an optimi-
sation of the cost of monitoring and of the accuracy of collected and derived data related to the
objectives of the network. Without a thorough understanding of the hydrogeological setting of a
region, there is little chance that a network would produce adequate information. At the start of
monitoring in a region, however, a classic problem is insufficient hydrogeological knowledge
1. x, y, z, t denotes position in space and time.

71
Groundwater studies
Figure 3.3 Hydrological variables to be monitored in a groundwater network and variables of other networks
which are complementary but necessary
and therefore an unknown spatial and temporal variability for each variable to be monitored,
although we know that groundwater heads and chemical components are spatially and
temporally correlated.
Occasionally, lack of prior hydrogeological knowledge on how to interpolate (in space and
time) between measurement points hampers even the beginning of monitoring. If hardly any
data are available, background monitoring should start anyway, by designing a network based
on the few existing data and on expert knowledge from similar regions. After some years of data
collection, statistical and geostatistical techniques can be applied to explore the spatio-temporal
structure of each hydrological variable in the region of interest (Fig. 3.5). Eventually, optimi-
sation theory and socio-economic analysis can be used in decision-making procedures to
propose optimal networks to the policy makers (WMO, 1994). The following sections on
network density and sampling frequency assume that sufficient data are available for statistical
analysis.
3.4.1 Network density
The effectiveness of a groundwater network in terms of network density is often defined as the
accuracy of the spatial interpolation, i.e. the standard deviation of the spatial interpolation error.

72
Figure 3.4 Measured and simulated hydrological variables. Average measured precipitation and streamflow,
and simulated real evapotranspiration and streamflow (BILAN; Kaprek and Krejcov, 1994)
for a Dutch, a Belgium-Dutch and a Czech catchment
73
Groundwater studies
Figure 3.5 Procedure for the determination of network density and sampling frequency
Therefore a spatial interpolation technique is required that not only estimates the groundwater
variables, but also provides the standard deviation of the estimation error (e.g. Zhou, 1992a).
Kriging is a well known and suitable technique for such a purpose (e.g. Marsily, 1986).
Kriging is a method for estimating the value of a regionalised random variable (e.g.
groundwater head or chemical component) at any point that has not been measured from a set
of measurements at different locations. The semi-variogram plays a key role in the kriging
procedure. It describes the spatial correlation structure of the regionalised variable, i.e. it shows
that observations closer to each other are likely to be more similar than observations further
away.
The semi-variogram is determined as follows (e.g. Zhou, 1992a):
Several groups (m) of distances between measurement locations with an average
distance d are defined. For example, d1 is the group where the distances between the
locations are small, whereas dm is the group with the largest distances;
For each distance group dk the possible pairs of measurements locations i and j are
identified (nk : all possible pairs);
For each distance group dk the variogram value (dk) is calculated using the following equation:
k
(dk) = 2 1n
k
(Z Z )
1
i j
where Zi and Zj the measured variables are at location i and j.

74
The semi-variogram is obtained by plotting the (dk) values against the distances of the various
groups. This semi-variogram is called the experimental one (dots in Fig. 3.6);
Figure 3.6 Experimental (dots) and theoretical semi-variogram (solid line)
Finally a theoretical semi-variogram is fitted through the experimental one (solid line in Fig. 3.6)
by using common models such as an exponential, Gaussian or spherical model with three
parameters, i.e. the nugget variance c0, the range parameter b and the sill A0. The nugget
variance describes the non-spatial variance, i.e. the variance of the measurement error, observer
error or micro-distance variability. The range parameter or correlation length and the sill
describe the spatial variation. The variation increases with increasing distance (Fig. 3.6) between
locations in space. The value of the range parameter is the variability that eventually occurs (A0)
when the distances between the observations increase.
In a heterogeneous region (e.g. geology, surface water system) it might be useful to find
out if a stratification of the region leads to better quality semi-variograms. Stratification implies
dividing the region into disjoint subareas (strata) with more homogeneous characteristics. To
model the spatial variability for each stratified area, a within-strata semi-variogram needs to
be determined. For a within-strata semi-variogram, only pairs of points belonging to the same
stratum contribute to the semi-variogram. If the hydrological variable is related to the
stratification, the within-strata semi-variogram will take lower values than the ordinary semi-
variogram, because the within-strata variability is less than the overall variability. However, if
this relation is not present, it will show higher variability, and hence stratification is useless.
If the semi-variogram is known, it is only a simple routine to compute the hydrological
variable on the nodes of a specified network, including the standard deviation of the inter-
polation error (SDIE), from the locations with measurements. In the subsequent network density
design the SDIE plays a key role. For example, the following procedure can be applied (Van
Bracht and Romijn, 1985):
divide the region into homogeneous subareas (e.g. geology, water table depth etc.) also
considering available observation wells per subarea. Regional hydrogeological knowl-
edge is a prerequisite for defining an adequate stratification;
determine within-strata semi-variogram for each Subaru for a particular date and
check persistence for other dates (e.g. Stein, 1998);

75
Groundwater studies
Calculate the network density graphs for each Subaru using the within-strata semi-
variogram (Fig.3.7). The graph specifies the relation between the SDIE and the number
of observation wells per km2 and is derived by calculating the hydrological variable on
the nodes of a specified network from a decreasing number of locations in the Subaru
for which measurements are available. The Subarus sensitivity to decreasing the
density of sampling points and the SDIE depend on the characteristics of the Subarus.
For instance, in a clay polder area with controlled water levels (subarea B) the SDIE
and the sensitivity are significantly lower than in a sandy area with natural drainage
and with a shallow impermeable base (subarea A).
For each subarea determine the allowable SDIE that can be related to the priority of the
subarea in terms of groundwater management. Van Bracht and Romijn (1985) derive
this priority from planned groundwater exploitation and the sensitivity of the subarea
to lowering of the water table. Some fictitious results of this step are illustrated in
Figure 3.8. First the priority of each area is converted in an allowable SDIE; the
allowable SDIE increases from 40 to 80 cm for subareas 5 to 2 (black bar in Fig. 3.8)
corresponding to a decreased priority. Then theoretically, the network density could be
derived from the network density graphs (Fig. 3.7). However, in some subareas the
allowable SDIE is even lower than the one belonging to the highest density in the
network density graph (subarea 1), whereas in others (subarea 4) the allowable SDIE is
still higher than the one belonging to the lowest density. Therefore a maximum and
minimum network density are introduced; in this example, one well per 2 and 10 km2,
respectively.
The SDIE belonging to these densities is derived from Figure 3.7 and plotted in
Figure 3.8 as the maximum and minimum (first and second bars). Now the network
density can be determined for each subarea. In subareas 2 and 4 the minimum density
can be used, i.e. one well per 10 km2 (1/10), whereas in subarea 1 the maximum density
Figure 3.7 Network density graphs; relationship between standard deviation of the interpolation error (SDIE)
and density of observation wells

76
Figure 3.8 Network density derived from allowable standard deviation of the interpolation error (SDIE),
minimum and maximum density for different subareas
is used (1/2), although the allowable SDIE is exceeded. In the subareas 3 and 5 the
network density is between the prescribed minimum and maximum density, namely
1/7 and 1/6, respectively.
Perform a sensitivity analysis (monitoring alternatives); i.e. specify other values for
allowable SDIE, and for minimum and maximum network density, and determine net-
work density for each subarea and total network density. Van Bracht and Romijn (1985)
show that in their example reducing the maximum allowable SDIE from 65 to 60 cm
entails increasing the number of observation wells by 33%. A 10 cm decrease from 65
to 55 cm implies a 77% increase in wells.
Compare cost and benefits of the different monitoring alternatives and eventually
select optimal network density.
Stein (1998) also gives an example of a redesign of a groundwater network using kriging.
He shows that in an area of 635 km2, where approximately 500 wells occur, the uncertainty can
even decrease if 50 existing wells are removed and 5 new ones are introduced at strategic
locations.

77
Groundwater studies
3.4.2 Sampling frequency
Besides exhibiting spatial variability, groundwater heads and chemical components show a
temporal variability, which introduces the question of how often a variable has to be monitored.
Temporal variability can be analysed with time series analysis procedures (e.g. Box and
Jenkins, 1976; Chatfield, 1989; WMO, 1994), which give information about trends, periodic fluc-
tuations and the mean. Detection of trends and periodic fluctuations is relevant for monitoring
the effects of groundwater abstraction, seasonal climatic effects, or the deterioration of the
groundwater quality. Sometimes, use of time series analysis is hampered by incomplete data
series, irregularities in sampling intervals, and shortness of the time series.
The analysis of groundwater time series is confronted with some special properties of the
groundwater system, e.g. the groundwater head or chemical composition at time t is dependent
on previous values at time t1 (autocorrelation), and non-stationarity due to trends and periodic
fluctuations. If these features are recognised the appropriate advanced time series analysis
techniques can be applied to evaluate the time series and subsequently to design the sampling
frequency (e.g. Zhou, 1992a; Zhou, 1992b). The following steps can be recognised:
Analysis of characteristics of groundwater time series. First possible trends in time
series of groundwater data need to be investigated. The probability of trend detection
depends on trend magnitude, correlation structure of the time series, the observation
period, and the required reliability of the statistical test. A plot of the data is likely to
show the presence of an obvious trend and if so, which type of trend prevails (step,
linear, or exponential trend).
Figure 3.9 shows an example of step trends of about 50 cm in groundwater heads of a
semi-confined aquifer, due to the seasonal groundwater abstraction. Step trends are
easiest to recognise by comparing the means of groundwater heads or other com-
Figure 3.9 Step trends in the groundwater heads because of a seasonal groundwater abstraction
in a semi-confined aquifer

78
ponents from the periods before and after a known or expected change in the
groundwater system. The detection of a step trend is simply a statistical test, where the
hypothesis that the means of the time series are equal or unequal either is accepted or
rejected. Students t distribution is often used. Nitrate concentrations in the ground-
water discharge of the Noor catchment (Belgium-The Netherlands) show a linear trend
(Fig. 3.10). Linear trends are commonly evaluated by using a classical linear regression
model. The detection of the linear trend is a statistical test against the slope of the
trend (Zhou, 1992a).
Figure 3.10 Nitrate concentrations in the groundwater outflow of the Noor catchment showing a linear trend
After trend analysis, the time series should be evaluated on its periodicity due to
seasonally varying rainfall, evapotranspiration or abstractions. A clear example of
periodic fluctuations and a linear trend in the piezometric heads of deep groundwater
in India due to abstraction for irrigation is presented in Figure 3.11. Spectral analysis is
applied to recognise periodic fluctuations. The periodic properties of groundwater
time series may be modelled by harmonic series. The linear least squares method is
applied to derive the constant and the harmonic coefficients (Zhou, 1992a).
Finally the stationary or stochastic component of the time series can be evaluated. This
component is determined by subtracting the trend and periodic components from the
original time series, which results in a groundwater time series of residuals. This time
series is usually stationary and autocorrelated. Zhou (1992a) proposes using the mean
of the residual time series to estimate the sampling frequency associated with the
stationary component of the time series.
Determination of the sampling frequency (Zhou, 1992a). The sampling frequency f
equals the maximum value of three different frequencies, if all three are relevant:
f = max ( fT , fP , f M )
fT : the sampling frequency for trend detection;

fP : the sampling frequency for periodicity detection;
fM : the sampling frequency associated with the mean of the stationary component.

79
Groundwater studies
Figure 3.11 Periodic fluctuations and a linear trend in the groundwater heads in Gujarat, Western India,
due to the over-exploitation of an alluvial aquifer (Rushton, 1998)
The relationship between accuracy of the estimate of the mean and the sampling
frequency (fM) is introduced by using the concept of standardised half width of the confidence
interval of the mean. Eventually a family of curves is obtained with the standard half width of
the mean versus the sampling frequency for various correlation coefficients, sampling periods,
and autoregressive model structures. These curves clearly show that increasing the sampling
frequency (fM) beyond a certain critical value results in only a marginal increase of accuracy. The
critical value is the key factor in determining fM. The critical value decreases with an increase of
the correlation coefficient. Therefore, for highly autocorrelated time-series, fM is lower than for
slightly autocorrelated series.
The analysis of periodic fluctuations using harmonic series reveals the highest significant
periodic fluctuation in the real time series fH. The sampling frequency to determine the trend fP
needs to be more than twice fH. For example, when the significant periodic fluctuations in
groundwater time series are monthly and annual, fP should be at least twice a month.
Furthermore the variances of the estimation of the harmonic coefficients can also be used as a
criterion for determining the sampling frequency fP (Zhou, 1992a).
Significant trends are explored by using Students t distribution. If these trends prevail in
the time series, Zhou (1992a) shows that for given standard trend magnitude and length of
observation period, the power of the trend detection (power=0 for no trend recognition, and
power=1 for complete recognition) will only be a function of the sampling frequency for a
particular confidence level. Obviously, for a given trend magnitude the power of the trend
detection increases with an increase of the sampling frequency. Furthermore the power of trend
detection decreases with a smaller trend magnitude. In this way a family of curves is obtained

80
with the power of the trend detection versus the sampling frequency for various trend
magnitudes, observation periods and confidence limits. The sampling frequency fT can be
derived by evaluating the shape of these curves.
The proposed sampling frequency can be verified by comparing the characteristics of the
original time series with the series obtained using the recommended frequency (Fig. 3.12). The
statistical characteristics (see step 1) of the original time series based on, for example, two-
weekly data should be nearly identical with the characteristics of the derived series based on the
recommended lower frequency (e.g. monthly data).
Figure 3.12 Difference between groundwater heads measured with different sampling frequencies
in a observation well in the Hupselse Beek catchment (The Netherlands)
81
Groundwater studies
Zhou (1992a) and Zhou and Li (1992) present comprehensive elaborations of the above-
mentioned stochastic procedure for the redesign of the sampling frequency in two cases:
groundwater level monitoring in the region around a Dutch well field for drinking water supply,
and the monitoring of shallow groundwater levels in the Central Yellow Plain in China.
The above-mentioned stochastic procedure for determination of the sampling frequency is
based on historical time series. Therefore the recommended frequency applies to a situation with
no interference, e.g. in the case of the redesign of the background or specific monitoring net-
work, after some years of monitoring without a significant change in the groundwater system
(Fig. 3.12). However, the sampling frequency should be adapted if human interference starts and
is expected to result in a trend and possibly other periodic fluctuations. The new sampling fre-
quency can be determined by a combined stochastic-deterministic approach. The stochastically
modelled time series are converted into a new time series using deterministic groundwater
models, such as MODFLOW (McDonald and Harbaugh, 1988).
Adaptation or introduction of a trend and periodic fluctuations can be based upon
the changes in groundwater heads simulated by these models. In this way the first sampling
frequency of the specific monitoring network can be determined (Fig. 3.12). After some years
the recorded time series can be used with the stochastic approach to refine the sampling
frequency.
3.4.3 Simultaneous design of network density and sampling frequency
Under some conditions, a trade-off between sampling frequency and network density occurs for
a particular hydrological variable (Van Geer and Van der Kloet, 1986).
Stein (1998) introduces the dynamic semi-variogram in this context. The parameters of
a hypothetical dynamic semi-variogram are allowed to change in time. The procedure to be
followed is essentially the same as already explained for the network density. With the dynamic
semi-variogram procedure, a value of a hydrological variable can be interpolated towards an
arbitrary, unmonitored site and time, based on monitored data. The specified prediction error
can serve as a criterion to find out whether the network and sampling density are already
adequate.
Van Geer and De Stroet (1990) and Zhou (1992a) propose applying a stochastic-
deterministic approach, i.e. a Kalman filtering model that consists of the Kalman filtering
algorithm and a deterministic groundwater model. In this way observation-based information
and physically-based information are incorporated in the network design. Usually the network
design is obtained by minimising the network density and sampling frequency under the
constraint of a prescribed threshold value of the standard deviation of the estimation error
(SDEE). A trial and error procedure is followed, which implies that for a given network
alternative a contour plot of the SDEE is computed with the Kalman model, and compared with
the threshold value. In areas where the SDEE is lower than the threshold, the network density
can be decreased, whereas the opposite occurs in areas where the SDEE is higher than the
threshold. Alternatively, a closer fit of the SDEE can be achieved by adjusting the sampling
frequency.
3.5 Groundwater information system
The collection of data from a groundwater monitoring network is useless, unless an information
system is available that organises the data flow from observation or measurement to dissem-
ination Although most steps are trivial, many mistakes are made in this context. The following
steps can be distinguished (Fig. 3.13; readers are also referred to WMO, 1994):

82
Figure 3.13 Outline of a groundwater information system
Recording. At specified times the value of a hydrological variable is recorded either in a

field notebook or is stored automatically in a data logger. Recording by a skilled
observer offers the possibility for an initial quality control. To maintaining good
quality observations it is essential that the observation station itself is regularly
inspected on possible malfunctioning of the observation devices.
Transmission. At specified times the observed data go to the data-processing centre.
Copies of field notebooks are mailed, or the observer calls the centre. Automatically
stored data are retrieved on site with a portable PC. In the field a first quality check
can be done, by plotting the observed data in a graph, and performing simple statistics

83
Groundwater studies
(e.g. determination of minimum and maximum). Fully automated stations can transmit
the recorded data instantaneously or at regular intervals to the centre. This relative
expensive option is attractive for remote areas, but skilled staff are needed. If the
groundwater systems involved are significantly developed, these fully automated
systems at far-off locations also can operate as an early warning system; this implies
that the system transmits a message if a specified level is exceeded (e.g. groundwater
head below a certain value or Cl concentration above a critical level). In general, the
recording and transmission is done by several observers. Therefore it should be
adequately organised and well supervised, i.e. clear, concise notebooks with obvious
instructions on coding, required accuracy and initial quality control, and a proper
manual for each observation instrument, data logger and PC used for data retrieval.
Central data storage. After receiving the data in various forms and storage media, the
processing can start. For instance, correction of data for zero-shift, detection of missing
values and replacement with an appropriate code, conversion of stage levels into
discharges and transfer of water tables or piezometric levels to metres above datum.
Quality control. A quality check must be carried out. Several methods are available, e.g.
visual control by plotting in a graph, detection of outliers, comparison with similar
time series, comparison with data of an allied hydrological variable (Delft Hydraulics,
1992). Information associated with the data element, such as the date and station code,
need to be checked as well. Data can be stored in databooks and quality control can be
done by hand, but usually it is more efficient to store the data in a computer system.
Then, quality control is more feasible because it can be done automatically.
Processing. After the data have been checked, descriptive statistics can be applied, such
as the calculation of totals over different periods, mean, median, minimum and
maximum and variation. Probability distributions can be computed, giving the
probability of occurrence of certain events or return periods. At this stage, missing data
also can be replaced by a predicted value by using statistical techniques, such as
regression.
Dissemination. People dealing with groundwater must have access to monitored data. If
the data are not stored on a computer system, data books should be published. If not
all basic data can be published, then at least some processed data are provided, and
basic data can be supplied on request. If the data are available in a computer system,
an on-line connection with the database is preferred. Such a connection makes it
possible to retrieve and to analyse only those data relevant for the user. Of course, the
database should be well organised, described and protected if people outside the data
processing centre have access to it.
In a groundwater information system not only dynamic data are stored. Actually it should
comprise two parts (Fig. 3.13), namely:
a part in which time-independent data are stored, and
a part with the dynamic data.
In the time-independent domain of the information system, for example, locations and
descriptions of drillings are stored, as well as permeabilities, thickness of aquifers, hydraulic
resistances, storativities. Observation locations and type of observation tools are also stored. It
also should contain a logbook with comments on each monitoring site, such as malfunctioning
or failure of the monitoring device, and checks with other, or more accurate tools.
Some groundwater information systems not only contain data, but also include analysis
techniques. An example is REGIS ( Kuipers, 1998). It comprises techniques to derive hydraulic
properties (permeabilities, hydraulic resistances) of the subsurface from pumping test data (see
also Chapter 8). Furthermore, different types of groundwater simulation models are linked to

84
the information system. Dynamic and time-independent data available in the database can be
tailored in different ways, so that they can serve as input data for the groundwater models.
3.6 Observation methods
3.6.1 Observation wells and piezometers
There have been extensive reviews of methods to monitor the groundwater variables (e.g.
WMO, 1994; Otto, 1996). Below only techniques that monitor dynamic variables of groundwater
in the field and the laboratory, such as elevation of water levels and water chemistry will be
discussed. In this section a summary is presented, and only some of the recent developments are
explained in more detail.
The data on groundwater heads and the chemical composition are obtained at observation
wells or piezometers. Observation wells and piezometers should be installed carefully, taking
account of rock type (consolidated, unconsolidated), and tested (WMO, 1994). An observation
well is a non-pumping well used to observe the height of the water table. It is generally of larger
diameter than a piezometer and may be screened throughout the thickness of the aquifer. The
observation well is a length of fully slotted tubing that is lowered into the bore hole and
backfilled with sand or soil around the side of the tube. Water can freely enter the tube along its
entire length. Comparisons between individual piezometers and an open well have shown that
the open well may act as a vertical conduit for water to flow, thereby distorting the groundwater
conditions in the vicinity of the measuring location.
A piezometer is a non-pumping well, generally small in diameter, that is used to measure
the groundwater head. It is an open-ended pipe, placed in a borehole that has been drilled to the
desired depth in the ground. The bottom tip of the piezometer is fitted with a perforated or
slotted screen (length about 1 m), to allow the inflow of water. The annular space around the
screen should be filled with a gravel pack. The remaining annular space around the pipe can be
filled with any material, except where the presence of aquitards requires a seal of bentonite clay
or cement grouting to prevent leakage along the pipe. The water levels measured in piezometers
represent the average hydraulic head at the screen of the piezometer. Rapid and accurate
measurements can best be made in small diameter wells. If their diameter is large, the volume of
water contained in the pipe may cause a time lag in water level changes.
Multi-piezometer installations are required for measuring the vertical hydraulic head
distribution in anisotropic aquifer systems. Well-known methods are:
Clusters of small-diameter piezometers that are placed in a single borehole at different
depths and insulated by impervious material. The bottom part of each piezometer is
screened. Leakage along the pipes from one piezometer to another has to be avoided.
Clusters of piezometers in different boreholes in one place. The piezometers are
installed at different depths. This is technically less demanding, but usually more
expensive.
Before an observation well or piezometer is used, it should be developed. This entails
injecting and abstracting water into and from the well, which induces groundwater flow alter-
nately from and to the well. This procedure removes clogging from the screens and fine
materials from the bottom of the well and the pack around the well. Subsequently a per-
formance test is needed. An adequately installed well or piezometer should quickly follow head
changes in the aquifer. A simple test is carried out by observing the fall of the water level in the
well after the recharge of a known volume of water injected in the well. If the decline is too slow,
the well should be further developed. Occasionally the tests of the wells or piezometers should
be repeated, e.g. every two or three years.

85
Groundwater studies
3.6.2 Methods for monitoring groundwater quantity
Methods to monitor groundwater heads are described below. Furthermore some information is
provided about the monitoring of groundwater well discharge. Methods to monitor ground-
water recharge and discharge are not dealt with, so interested readers should consult the
relevant publications (e.g. Lerner et al., 1990; Boiten, 1993; WMO, 1994; Simmers, 1997).
Both manual-operated and automated-recording instruments are available to measure
groundwater heads in observations wells or piezometers (Table 3.1).
Table 3.1 Summary of instruments commonly used to measure groundwater heads

(From: Driscoll, 1986; Nielsen, 1991; WMO, 1994; Otto, 1998)
Method Readout device Advantages Disadvantages Costs, skills
Manual
Wetted-tape Tape markings, Accurate if depth Several Low price and easy
or flexible sometimes with is not too large measurements to produce and to use
steel steel ruler needed to find
approximate depth
Dipper Tape markings, Accuracywithin Not-applicable in Low price and easy to

sometimes with 0.01 m, fast noisy environments produce and to use
steel ruler
Inertial Tape markings Accuracy within Calibration Moderately priced,

devices 0.01 m, fast and simple, easy to operate
to use in polluted
groundwater
Two- Tape markings Fast and simple, Calibration, regular Moderately to high
electrode accuracy decreases maintenance, priced dependent
devices with depth batteries on cable length, easy
to operate
Automatic
recording
Mechanical Drum chart Widely applied Float lag, High priced,

float or data logger mechanical failure, regular maintenance
recorder large well diameter and checking
system
Data logger Less components Temperature effect, High priced, regular
Pressure than float systems connection with the checking
transducer open air, calibration
Ultrasonic Data logger Less components Temperature and High priced, regular
sensors than float systems humidity effects; checking
for under-water
types effects of
pressure, solute
concentrations
and air bubbles

86
(i) Manual groundwater head monitoring
The most common manual method is to suspend a weighted plastic-coated tape or flexible steel
cable from the well head to a point below the water level. The water level is determined by the
difference between the length of the wound-out cable and the length of that part that has been
submerged. The length can either be read out directly from the markers, or by using a steel rule
and the nearest marker, if fewer markers are adjusted on the tape or cable. Sometimes chalk or
pastes that change colour are used to simplify the determination of the part submerged. Depths
of 50 to 100 m are measured with ease. At greater depth a thin steel cable or lightweight plastic-
coated tape is recommended. A skilled observer can achieve an accuracy of a few millimetres,
although usually the error increases with depth.
The dipper is a cylindrical probe with a hollow space at the end, which is connected to a
plastic-coated tape or flexible steel cable. If the dipper reaches the water table an audible signal
is produced. Usually the depth is determined after raising and lowering the probe a number of
times over a distance of a few centimetres. The depth of the water table equals the length of the
tape or cable, usually measured at the well head.
An inertial device is a portable instrument that is designed so that a weight attached to a
cable moves downwards at a constant velocity. If the weight reaches the water table, a braking
mechanism prevents further downward movement. A counter gives the depth of the water table
relative to the position of the portable instrument, usually the well head. With this equipment
depths greater than 100 m can be measured. Accuracy is high if the cable has negligible stretch.
An advantage is the high resistance, for example, to oil-polluted water and corrosive waters.
The two-electrode equipment consists of a portable reel with plastic-coated tape or cable
connected to a probe at the end of the cable. The probe with a length of about 10 to 20 cm has
two small adjacent electrodes. The circuit between the two exposed electrodes is closed when the
probe reaches the water table, and a visible and/or audible signal is produced by a lamp or
buzzer built-into the reel. The depth of the water table is read from the cable and is related to the
position of the well head. With this equipment large depths to over 300 m are measured. The
maximum depth is dependent on the length of the electrical cable, the design of the electrical
circuitry and the acceptable weight of the equipment. The accuracy is comparable to the inertial
device, although at great depth (in the order of 500 m) errors of 0.15 m are reported. However, if
differences in water level have to be measured at this great depth (e.g. pumping test) and the
cable is left suspended, an accuracy of millimetres is achieved.
Other techniques use other physical properties, such as the resistance or capacitance. The
electrochemical effect of two different metals is also used. Then no batteries are needed.
Measurable current flow can be produced in most groundwaters by immersing two electrodes
(e.g. magnesium and brass). There is even a single electrode version: a magnesium electrode in
the probe and a steel earth pin at the surface.
The above-mentioned instruments can also be used for free-flowing groundwater, as long
as the groundwater head and the piezometer are no more than approximately 1.5 to 1.8 m above
the soil surface. As an alternative the piezometer can be connected to a transparent tube just
above the soil surface. The head is read from the marked tube, or a steel rule is used. If the head
of the artesian aquifer is more than 1.8 m above the soil surface a pressure device needs to be
applied (see below).
(ii) Automatic recording of groundwater heads
Automatic recording of groundwater heads is sometimes necessary, i.e. to investigate fast

changes (e.g. pumping test, tidal effects, scientific research) or to monitor at remote places.
The mechanical float recorder is widely applied. This device is based on a float that is

87
Groundwater studies
linked to a counterweight by a cable that runs over a pulley. The device is above the well, and
the float and the counterweight are in the observation well if the diameter permits. In wells or
piezometers with small diameters a vertical pipe has to be installed in the ground for the
counterweight next to the well with the float. Float and float-line friction against the well casing
can be a problem in wells with deep water levels and should be avoided. A large-diameter float
is recommended because of its greater sensitivity to water level changes. The cable should be
long enough to account for the groundwater fluctuation between the dates that the site is
visited. The turn of the pulley is converted into a vertical movement of a pen on a drum chart
(analogue recorder), which is driven by a spring or an electrical clock. Usually the speed at
which the drum rotates can be adjusted, e.g. drum charts from one day to one month are
available. Instead of recording on a drum chart, the turn of the pulley also can be converted into
digital information (electronic recorder) that can be stored on a data logger. Then the water level
is not recorded continuously as with the drum chart, but it is stored only at prescribed intervals,
for example every minute, 15 minutes or hour. A proper data storage is necessary; the data
logger in the field should have a backup memory. Then the data can retrieved again, if
something goes wrong with the data exchange between the data logger, the portable PC in the
field and the central computer in the office (Fig. 3.13). Careful maintenance and checking of the
float recorder is a prerequisite to avoid malfunctioning. Every time the recorder is visited the
recorded level should be checked against manual observations. In reliability, versatility,
maintenance efforts and need for skilled staff, the float type method is considered to be a good
compromise compared to other automatic recording methods. Other methods based upon the
pressure transducer and ultrasonic sensors are being used increasingly (Table 3.1).
Table 3.2 Accuracy of various methods to measure groundwater heads (After Boiten et al., 1995)
Method Accuracy (m)
Floaters
analogue recorder
- diameter 0.10 m 0.0030.005
- diameter 0.20 m 0.0010.003
electronic recorder
- diameter > 0.08 m 0.0010.003
Pressure transducers
cheap types 0.0100.050 1
expensive types 0.0020.010 1
Ultrasonic sensors 0.0050.020
1. Depends on measuring range, etc.
In a pressure transducer the water pressure is converted to an electrical signal. The

transducers are designed so that the water pressure is linearly related to the electrical signal. A
calibration of the pressure transducer is recommended. The small pressure transducer is lowered
into the observation well or piezometer below the deepest possible water level and stays there.
The transducer is connected to the open air via a cable (atmospheric pressure is the reference)
and to a logger which periodically stores the water level reading. Temperature affects both the
electronic and the mechanical part of the transducer; at high or low temperatures the transducer

88
may not operate adequately. Nowadays there are probes that combine a pressure transducer and
a data logger. The 1520 cm long probes are lowered into the observation well and stay there for
one month, recording pressure data every 15 minutes. This device is a good alternative for
locations exposed to vandalism.
An ultrasonic sensor transmits ultrasonic pulses. These pulses are reflected at the water
surface and received again by the sensor. The time between transmission and receipt is linearly
related to the distance between the sensor and the water surface, and consequently to the water
level. Ultrasonic sensors can be placed either above or below the water surface. Below the water
surface the propagation velocity is lower than in the air. Ultrasonic sensors are sensitive to
temperature and humidity. Furthermore the performance of underwater types is affected by
variations in water pressure, solute concentration and occurrence of air bubbles. The ultrasonic
sensor is connected to a data logger.
An indication of the accuracy of the various methods is given in Table 3.2. Errors from the
conversion of readings related to the well head (local reference) to datum (e.g. m a.m.s.l.) are not
included.
(iii) Abstraction monitoring
Mechanical flow meters record the total water abstraction and the data is read from the meter at
periodic intervals. Electronic totalling flow meters are used if the data can be stored on a logger.
For proper groundwater management, each abstraction well within a well field should be
instrumented to monitor abstraction amounts and well performance.
3.6.3 Methods for monitoring groundwater quality
Methods for monitoring the chemical composition of groundwater are essential for the
identification of groundwater flow systems (background monitoring). Furthermore the moni-
toring is necessary to investigate the development of the chemical composition during the
exploitation of the groundwater resources (specific monitoring). This section will describe the
sampling of aquifers, reasons why the chemical composition of the sample may differ from the
composition of the aquifer itself, sampling devices and analysis techniques. For a more com-
prehensive description, readers are referred to WMO (1994) and Otto (1998).
(i) Sampling of aquifers
The objective of groundwater quality monitoring programmes is to obtain samples of water that
represent the groundwater within an aquifer, or within particular depth intervals in an aquifer.
For some purposes hydrochemical monitoring should be discrete in its location within the
aquifer system, which implies that additional hydrochemical monitoring boreholes have to be
constructed. Where possible installed observation wells or piezometers for head measurements
should be used, provided that the design is suitable for efficient sampling.
When sampling aquifers for monitoring groundwater quality a distinction should be
made between non-point and point sampling. Non-point samples obtained by pumping from
open boreholes or fully screened observation wells provide information on the overall changes
in groundwater quality; the samples might be a mixture of different groundwater types. Such a
sampling programme monitors broad changes in the chemical composition of groundwater in an
aquifer at relatively low cost. Non-point sampling, however, is inadequate for monitoring
groundwater quality changes at a local and three-dimensional scale. Therefore a different
sampling design to the non-point method is required to monitor site-specific groundwater
changes such as salt water intrusion or plumes from waste disposal sites. Important consider-

89
Groundwater studies
ations in the design include the need for close vertical interval point sampling and sampling
locations that take the groundwater flow patterns into account. Using piezometers for water
quality sampling, the screen length of the borehole casing or piezometer has to be considered.
There is little difference if hydraulic head is measured in a small or large volume of the aquifer,
because the vertical hydraulic gradient in the permeable unit is usually small. However, the
result can be dramatically different for water quality sampling.
The most reliable method for multilevel sampling can be achieved by piezometers in
different boreholes at various depths at one site. The second best method is to bundle piezo-
meters at different depths in one borehole and install samplers at discrete sampling points.
Sometimes it is possible to sample continuously in an uncased borehole or fully screened stand-
pipe by using a multi-layer sampling probe (MLS). The MLS collects undisturbed water samples
at small depth intervals (e.g. 3 cm). The MLS consists of a rod with tens of cylindrical cells on
top of each other with a small volume (e.g. 15 cm3). The MLS is lowered into the well to the
required depth. The cells are separated by rubber seals or inflated packers that prevent vertical
mixing of groundwater in the borehole. The sampling is based on the dialysis-cell method, using
the process of diffusion through a nylon membrane. For large ions the diffusion process reaches
an equilibrium after a maximum of 5 days. After the equilibrium stage has been reached the
MLS can be pulled out of the well, and the composition of the water in each cell can be analysed
(Krajenbrink et al., 1989; Ronen et al., 1986). The results obtained with the MLS are only reliable
if cross-migration of fluids in the well can be prevented.
(ii) Factors changing the chemical composition of groundwater samples
To obtain a representative sample of groundwater the well casing volume of a well has to be
purged by bailing or pumping at least three bore volumes. The adequacy of the purging can be
checked in the field by examining readily measured physical-chemical components (e.g.
temperature, pH, EC) of the purged groundwater that should approach a constant value. A
small diameter well would also mitigate purging problems for chemical sampling.
Three problems may seriously affect the chemical composition of a groundwater sample,
namely:
effects of well construction and contamination with drilling fluid. For example, the use
of cement rings in dug wells (higher pH) or bentonite as seal in a piezometer can affect
the water chemistry. Also, drilling fluid that has infiltrated into the surrounding
aquifer material is often difficult to remove.
sample deterioration. The chemistry of samples can change due to variations in
temperature and gas pressure. Cool and dark storage of samples is necessary. In situ
analysis, or prompt transportation to the laboratory can preserve data quality;
careless field and laboratory practices. Sample contamination caused by improper
bottle washing and filtering is a main concern. It is recommended to wash the bottle
with the groundwater to be analysed, prior to the collection of the sample. Organic
compounds can be absorbed by plastic containers, also the loss of volatile components
can be a problem. The quality of laboratory analyses should be checked by submitting
blanks and duplicates.
(iii) Sampling devices
The initial consideration in selecting a sampling device is whether the well will accommodate
the device. The smaller the diameter of the well, the more limited, complex and expensive the
available samplers become. Depending on the analytical requirements of a sampling
programme, the device should not affect the chemical and physical composition of the sample:

90
the material of the device should not sorb or leach contaminants;

pH, Eh of the sample should not alter;
no volatile should be released from the device;
the device should not introduce air or non-inert gas into the sample.
Also, the depth of sampling has to be considered. The deeper the sampling interval, the
more head a pump must overcome to deliver a sample to the surface. Ease of the sampling pro-
cedure, transportation, cleaning and maintenance of the device should be considered. Otto
(1998) lists the characteristics (e.g. well and device diameter, sampling depth, sample volume,
chemical alteration, relative costs) of some standard sampling devices, such as the cheap bailers
and the expensive submersible pumps.
Free-flowing or artesian wells are easy to sample by making a small hole in the obser-
vation well about 15 to 20 cm above the soil surface, which is normally closed to allow the
observer to measure the correct groundwater head. The water is allowed to flow for some time
to purge it, before taking samples.
(iv) Analysis techniques
Analysis techniques can be subdivided into the conventional analytical chemical techniques
predominantly used in the laboratory and sensors that measure one or more physical-chemical
components either in the laboratory or in the field. Excellent handbooks are available that
describe the former group (e.g. Clesceri et al., 1989; Velthorst, 1993). In this section only the
sensor-based approach that has produced encouraging results in recent years, is described.
For many years, sensors have been used to measure the physical-chemical properties (pH,
Eh, T, EC, turbidity) of groundwater in a well. There are many commercial sensors available.
Currently a few monitoring devices are on the market that can measure chemical components
(major anion and cations, dissolved oxygen) either in the laboratory or directly in the field.
These portable in-situ monitors are expensive, but when many samples have to be analysed the
costs per sample may be below those of the conventional analytical techniques. They require
repeated and careful calibration but are easy to operate. CSIRO 2 has developed an in-situ water
quality monitor that can measure a range of physical and chemical water parameters. Another
promising device, HYDRION-10 3, analyses samples with a set of integrated ion-selective or ion-
related electrodes (ISEs). It consists of a sensor unit and a measurement unit. The sensor unit is
made of stainless steel and contains the electrodes. This unit fits into an ordinary 500 ml glass
beaker. A water sample of 250 ml is sufficient for the analysis. The sensor unit can also be put
directly into the water, e.g. in a free-flowing well or in a surface water stream.
The measurement unit takes care of the operation and the data storage. The operational
features include the simultaneous measurement of EC, pH, K+, Ca2+, Mg2+, NH4+, Cl, NO3, and
HCO3; the specially-developed software automatically corrects for (mutual) effects affecting the
measurement of an individual component. No additives are needed, but filtering is necessary for
turbid water. About 1015 samples can be measured per hour. For about two years the outcome
of the HYDRION-10 has been compared with laboratory results using conventional analytical
procedures. Some results are given in Figure 3.14. Results correspond reasonably well, both for
weathering ions and for anthropogenic ions.
2. Dr G. Skying and Mr I. Jones, CSIRO Division of Water Resources, P.O. Box 1666, Canberra, Australia.
3. HYDRION, P.O. Box 522, 6700 AM Wageningen, The Netherlands.

91
Groundwater studies
Figure 3.14 Comparison of some measured ion concentrations with conventional analytical techniques
and a with set of integrated ion-selective electrodes HYDRION-10
3.7 Concluding remarks
Background monitoring networks should be installed if groundwater exploitation is expected

or planned. Monitoring should also be started when other future human interference can lead
to deterioration of the underground environment, ecologically relevant wetlands, springs or
streams due to changed groundwater conditions. Specific monitoring networks should be intro-
duced when a particular human interference starts.
Groundwater systems and the impact of changes can only be identified adequately if
both physical and chemical groundwater data are collected. Furthermore vertical
differences in groundwater components should not be neglected.
Groundwater monitoring networks should be supported by other networks collecting
data on groundwater recharge and stream flow.
Network density and sampling frequency are sometimes hard to define for back-
ground monitoring. Knowledge obtained in similar areas should be used. In sub-
sequent phases statistical and geostatistical techniques can be applied successfully.
Kriging has proved to be successful for the determination of network density, whereas
time-series analysis has been favourably applied to define sampling frequency. Oper-
ational procedures are available.
Groundwater information systems, which organise the flow of monitoring data from
observation to dissemination, are a prerequisite for an adequate monitoring network.
Several methods are available to monitor groundwater quantity data. Required
frequency (continuous or intermittent), accuracy and skills, and available budget
determine the best selection. Similar criteria apply to methods for quality monitoring.
Additionally, diameters of wells and devices, required sample volume, and changes in
chemical parameters should be considered.
Groundwater quality monitoring should consider the frequent occurrence of vertical
concentration gradients. For some purposes it is necessary to sample discrete parts of
the underground environment instead of taking a mixed sample over the entire depth
is a necessity. Special hydrochemical piezometers or sampling devices (e.g. multi-layer
sample probes) are needed.

92
Besides conventional analytical analysis techniques used in the laboratory, sets of inte-
grated ion-specific electrodes produce encouraging results. The latter analysis tech-
nique can also be applied in the field; this offers maximum flexibility and a minimum
of sample deterioration due to transport and storage.
BACHMAT, Y. 1989. Management of Groundwater Observation Orogrammes, World Meteorological

Organisation, Operational Hydrology Report No. 31, Geneva, 96 pp.
BOITEN, W. 1993. Flow-Measuring Structures. In: Sanderson et al. (eds.), Open Channel Measure-
ments. Special issue, Flow Measurements and Instruments, pp. 1724.
BOITEN, W.; DOMMERHOLT, A.; SOET, M., 1995. Handbook of Discharge Measurements in Open Water
Channels. Department of Water Resources Report No. 51, Agricultural University,
Wageningen, 158 pp. (In Dutch.)
BROWN, R. H.; KONOPLYANTSEV, A. A.; INESON, J.; KOVALEVSKY, V. S. (eds.), 1972. Groundwater
Studies; An International Guide for Research and Practice. Studies and Reports in Hydrology,
No. 4, UNESCO, Paris.
BOX, G. E. P.; JENKINS, G. M., 1976. Time Series Analysis: Forecasting and Control. Hilden Day, San
Francisco, California.
CARILLO-RIVERA, J. J., 1998. Monitoring Exploited Aquifers Resulted in Subsidence, Example:
Mexico City. In: H. A. J. van Lanen (ed.), Monitoring for Groundwater Management in
(Semi-)Arid Regions, Studies and Reports in Hydrology No. 57, UNESCO, Paris,
pp. 191208.
CHATFIELD, C., 1989. The Analysis of Time Series. Fourth edition, Chapman and Hall, London,
241 pp.
CLESCERI, L. S.; GREENBERG, A. E.; TRUSSELL, R. R. (eds.), 1989. Standard Methods for the Examination
of Water and Wastewater. Seventeenth Edition. American Public Health Association,
American Water Works Association and Water Pollution Control Federation.
DELFT HYDRAULICS, 1992. HYMOS: A Database Management and Processing System for Hydro-
Meteorological Data, Delft.
DE MARSILY, G., 1986. Quantitative Hydrogeology. Academic Press.
DRISCOLL, F. G. 1986. Groundwater and Wells. Johnson Filtration System Inc., St Paul, Minnesota,
1089 pp.
KAPREK, L.; KREJCOV, K., 1994. BILAN, Water Balance Model: Users Guide. T. M. G. Masaryk
Water Research Institute, Prague, 12 pp.
KRAJENBRINK, G. J.; BOUMANS, L. J. M.; MEINARDI, C. R., 1989. Hydro-Chemical Processes in the
Top Layer of Groundwater under Pasture-Land. Proceedings, Symposium on Nitrogen in
Organic Wastes, Applied to Soils. Aalborg, Denmark. Academic Press, London, pp. 31733.
KUIPERS, T. T., 1998. Geohydrological Information Management Systems. In: H. A. J. van Lanen
(ed.), Monitoring for Groundwater Management in (Semi-)Arid Regions. Studies and Reports
in Hydrology No. 57, UNESCO, Paris, pp. 10727.
LANGBEIN, W. B., 1954. Stream Gauging Networks. In: Proc. Assemble Gnrale de Rome, Rome,
IAHS Publication No. 38, pp. 293303.
LERNER, D. N.; ISSAR, A. S.; SIMMERS, I., 1990. Groundwater Recharge. A Guide to Understand and
Estimating Natural Recharge. IAH, Vol. 8, Verlag Heinz Heise GmbH, Hannover, 345 pp.
LLOYD, J. W., 1996. Specific Monitoring. In: H. A. J. van Lanen (ed.), Monitoring for Groundwater
Management in (Semi-)Arid Regions. Studies and Reports in Hydrology No. 57, UNESCO,
Paris, pp. 4764.

93
Groundwater studies
MCDONALD, M. G.; HARBAUGH, A. W., 1988. A Modular Three-Dimensional Finite-Difference

Groundwater Flow Model. United States Geological Survey.
MOSS, M. E., 1982. Concepts and Techniques in Hydrological Network Design. World Meteorological
Organisation, Operational Hydrology Report No. 19, Geneva, 30 pp.
MOSS, M. E. (ed.), 1986. Integrated Design of Hydrological Networks, IAHS Publication No. 158,
Wallingford, 415 pp.
NIELSEN, D. M., 1991. Practical Handbook of Ground-Water Monitoring. Lewis Publishers,
Chelsea, Michigan, 717 pp.
OEPA, 1993. Technical Guidance Manual for Hydrogeologic Investigations and Groundwater Monitoring
Programs. Ohio Environmental Protection Agency, Division of Drinking and Ground-
waters, Columbus, Ohio.
OTTO, C. J., 1998. Monitoring Tools and Type of Recording. In: H. A. J. van Lanen (ed.), Monitor-
ing for Groundwater Management in (Semi-)Arid Regions. Studies and Reports in Hydrology
No. 57, UNESCO, Paris, pp. 6590.
RONEN, D.; MARGARITZ, M.; LEVY, I., 1986. A Multi-Layer Sampler for the Study of Detailed Chemical
Profiles in Groundwater. Water Research 23, pp. 3115.
RUSHTON, K. R., 1998. Monitoring an Over-Exploited Aquifer in India. In: H. A. J. van Lanen,
(ed.), Monitoring for Groundwater Management in (Semi-)Arid Regions. Studies and Reports
in Hydrology, UNESCO, Paris, pp. 14150.
SIMMERS, I. (ed.), 1997. Recharge of Phreatic Aquifers in (Semi-)Arid Regions. IAH Intern. Contr.
Hydrogeol., A. A. Balkema, Rotterdam, 277 pp.
STEIN, A., 1998. Geostatistical Procedures for Analysing Spatial Variability and Optimising the
Collection of Monitoring Data. In: H. A. J. van Lanen (ed.), Monitoring for Groundwater
Management in (Semi-)Arid Regions. Studies and Reports in Hydrology No. 57, UNESCO,
Paris, pp. 91106.
UNESCO, 1990. International Hydrological Programme. Hydrology and Water Resources for a
Sustainable Development in a Changing Environment; detailed plan for the fourth phase of
the IHP (19901995). Paris, 54 pp.
UNESCO, 1992. International Glossary of Hydrology. Paris, 413 pp.
VAN BRACHT, M. J.; ROMIJN, E., 1985. Redesign of Groundwater Level Monitoring Networks by Appli-
cation of Kalman Filtering and Kriging Methods. Proceedings, Symposium on the Stochastic
Approach to Subsurface Flow, Montvillargenne.
VAN GEER, F. C.; DE STROET, C. B. M. , 1990. Kalman Filtering Packages for Program CISKA, a Com-
puter Model for the Interpolation of Heads with Kalman Filtering Algorithm. Report TNO
Institute of Applied Geoscience, Delft.
VAN GEER, F. C.; VAN DER KLOET, P. , 1986. Tradeoff Between Frequency and Density of a Ground-
water Measurement Network. In: M. E. Moss (ed.), 1986. Integrated Design of Hydrological
Networks, IAHS Publication No. 158, Wallingford, pp. 21329.
VAN LANEN, H. A. J., (ed.) 1996. Monitoring for Groundwater Management in (Semi-)Arid Regions.
Studies and Reports in Hydrology No. 57, UNESCO, Paris, 224 pp.
VELTHORST, E. J., 1993. Manual for Chemical Water Analysis. Department of Soil Science and
Geology, Agricultural University, Wageningen, 109 pp.
WMO, 1994. Guide to Hydrological Practices. Data Acquisition and Processing, Analysis, Forecasting
and other Applications. Fifth edition, WMO-No. 168, Geneva,735 pp.
ZHOU, Y., 1992a. Design and Analysis of Groundwater Monitoring Networks. IHE, Delft, 124 pp.
ZHOU, Y., 1992b. Sampling Frequency for Monitoring the Actual State of Groundwater Systems.
Research Report, IHE, Delft.
ZHOU, Y.; LI, B., 1992. Sampling Frequency for Monitoring Shallow Groundwater Levels in the Central
Yellow River Plain. Proceedings of International Workshop on Groundwater and Environ-
ment, Beijing.

94
4 P ro c e s s i n g a n d p r e s e n t a t i o n
of data
4.1 Scope
The processing and organised display of hydrogeological parameters are essential components
of hydrogeological interpretation and reporting. Processing and display of data are inseparable
and follow each other logically. They involve:
data collation,
data validation and storage,
retrieval of data for processing and creation of displays.
Typical use for hydrogeological displays may include:
identification of anomalies and outliers in data series,
depiction of trends and patterns,
comparison between several hydrogeological data sets,
interfacing hydrogeological data with other data sets,
spatial incorporation into Geographic Information Systems (GIS),
enhancement of reports.
These are but a few examples of a vast array of possible processing and display of hydro-
geological variables. An example of a previous publication that has partially dealt with the
subject, is that of Lloyd and Heathcote (1985).
Since the introduction of personal computers, significant developments in computer soft-
ware have come about. Within PC environments many displays that could previously only be
constructed with great difficulty can nowadays be created with relative ease. This does not rule
out the manual processing of data; indeed many of the displays discussed below can equally
well be drawn by hand. However, considering the massive amount of data that is generated
nowadays, preference should be given to the processing of data by computer,
4.2 Data types and presentation possibilities
For a meaningful discussion on the processing and display of hydrogeological variables, certain
data groupings are necessary. The groups discussed below have been selected to represent the
various environments in which hydrogeologists operate. There are two main groups: point data
and spatial information.

95
Groundwater studies
4.2.1 Point data
Point data are the data that are generated at a specific geographical point. In hydrogeological
terms, this data may assume many forms. Examples are:
geological data, such as borehole logs;
hydrogeological data, such as depths to water and yields;
construction data, such as casing, piezometer, borehole development and costs;
type of equipment, such as pumping equipment and recorders;
geophysical data, including surface and borehole measurements;
hydraulic properties of the aquifer(s), such as transmissivity and storage;
water abstraction and groundwater level records;
hydrochemistry;
other information, such as topography and surface drainage.
Even though many of these variables have spatial connotations, they refer back to the
groundwater abstraction point and are, for that reason, grouped under point data. Examples of
variables with spatial connotations in the above list include the geological log, which has a
vertical dimension, and surface geophysics, which has both horizontal and vertical dimensions.
4.2.2 Spatial information
The concept spatial information emphasises two important differences from point data. The
term spatial suggests that point data are processed or displayed in the geographical context,
such as on a map. The term information suggests that the point data have been processed,
usually to the extent that the original values for variables can no longer be recognised.
4.2.3 Presentation possibilities
Tools for the presentation of data and information can be grouped under three main headings.
These are:
graphs;
statistical tables and displays;
maps and cross sections.
A significant overlap is possible between these categories; for instance, maps may contain
graphs or statistical interpretations as insets. Alternatively, many of the commonly used
displays, such as specialised hydrochemical diagrams, have statistical undertones. Also, most of
the statistical displays that will be discussed below, fall in the category of graphs. Statistical
displays differ from ordinary graphs in that they usually contain processed information together
with raw data. As the discussion of the various displays evolves, the different categories of
displays that will be demonstrated in this chapter, will become clear.
4.3 Graphic processing and presentation of data
In order to present hydrogeological point data in as few concise diagrams as possible, certain
logical groupings are necessary. The following is suggested:
Group 1: Borehole information, such as geological, hydrogeological and construction data;
progressive borehole yield during drilling; packer testing results; borehole geophysics.
Group 2: Hydraulic properties of the aquifer(s), such as hydraulic conductivity, transmissivity,
storage and dispersion.

96
Processing and presentation of data
Group 3: Time-dependent data, such as groundwater abstraction; groundwater levels; water

quality.
Group 4: Specialised hydrochemical diagrams for detailed chemical interpretations.
Group 5: Other graphics, such as bar, stack, scatter and line graphs.
Most point data generated during hydrogeological investigations, can be accommodated
within these five groups.
Layouts for graphs to depict variables within each of these groups may vary from one
presentation to the next. There are, however, three basic rules to be considered when con-
structing displays. These are:
the displays must be clear, concise and legible,
they should contain only relevant data,
displays should be provided with the necessary headings and labels to ensure that
they will be meaningful entities on their own.
With these rules in mind, displays that may be used by the hydrogeologist will now be recom-
mended and discussed.
4.3.1 Borehole information
Vast amounts of valuable data are generated during the drilling of a borehole. Most of these are
of the one off type, in contrast to repeated measurements such as water levels for instance.
Essential one off variables that should be recorded during drilling are:
geological description of cores and cuttings,
depth to water and yield,
penetration rate during drilling,
measurements on water (electrical conductivity and pH for instance),
construction data (hole diameter, depth, casing properties).
Other data that are usually recorded after the hole has been completed, are:
borehole geophysics,
packer testing results.
The latter two variables are not truly of the once off type, because they may be measured
repeatedly, if so desired.
Each of the above variables should be recorded individually for the sake of establishing
complete records. Interpretation of the many data sets is, however, very difficult and time-
consuming. Ideally, all this information should be summarised onto a single sheet of paper in a
format that can easily be included in a report (Figure 4.1).
Figure 4.1 presents one such summary. In this diagram, the geological column has been
placed to the right, with abbreviated geological descriptions on the far right. This arrangement
allows other variables to be plotted to the left, leaving room for projections and interpretations.
This kind of presentation obviously has its limitations, the most important of which is the
limited number of vertical profiles that can be included on a single sheet of paper. In practice, it
has been found that up to seven vertical profiles can be combined on a normal-sized page.
Thereafter, the horizontal scales become so small that variations in values are no longer obvious.
If more than seven variables need to be plotted, they should be plotted onto two sheets,
repeating essential columns such as the geology on the second sheet.
The vertical scale is unlimited, since logs may be carried over to following pages. For the
sake of clarity, a single page that contains the complete log always should be included; certain
interesting portions of the log may then be presented, enlarged, on successive pages.
Many other examples of log summaries are found in practice. The exact layout of the logs
is not important, as long as the three basic rules for diagram construction, as indicated above,
are adhered to.

97
Groundwater studies
Figure 4.1 Composite hydrogeological borehole log
4.3.2 Hydraulic properties of the aquifer
Graphical presentation of hydraulic properties of an aquifer may be done on two levels.

The first level is plots of drawdown data for the calculation of the aquifer constants. Such
plots and the interpretation methods are discussed in Chapter 8 and have been well documented
in literature (Kruseman et al., 1991).
The second level is the presentation of hydraulic properties of aquifers in ways other than
time-dependent plots. This usually requires the data to be converted into spatial information.
This application will be discussed in detail in Section 4.5.
4.3.3 Time-dependent data
The hydrogeologist measures a variety of parameters that are time-dependent, e.g. water levels,
pumping rate and duration, rainfall and water chemistry. Plotting of this data does not normally
present a problem. Line or scatter plots and bar charts are used for this purpose. The data sets
may be visually compared by combining plots as in Figure 4.2. This type of presentation is
effective because of its clear and concise nature.
Processing and interpretation of time-dependent data are usually done through statistical
methods. Possibilities include variance within data sets, regression analysis, as well as corre-
lations between the different data sets. These procedures will be discussed under statistical
methods in Section 4.4.

98
Figure 4.2 Time dependent plot of water levels, electrical conductivity and rainfall
4.3.4 Specialised hydrochemical diagrams
Through the years, many special displays that meaningfully present hydrochemical data have
been devised. Of these, six displays stand out in terms of clarity and significance.
They are the Piper (Piper, 1944), Durov (Durov, 1948), Expanded Durov (Lloyd, 1965),
SAR (Wilcox, 1955; Bower et al., 1968), Schoeller (Schoeller, 1962) and Stiff (Stiff, 1951) diagrams.
All of these diagrams are so-called multivariate displays, simultaneously taking up to eight
variables into consideration, often projecting these variables to a single point on the diagrams.
Examples of the Piper and Expanded Durov diagrams, with their plotting procedures, are
included in Figures 4.3 and 4.4.
Advantages of using these diagrams include:
the plotting of numerous water analyses onto a single diagram,
the classification of waters according to their chemical characteristics,
the identification of trends.

99
Groundwater studies
Figure 4.3 Plotting procedures for the Piper diagram: Convert mg/l to meq/l by division
(Ca/20, Mg/12, Na/23, K/39, T.Alk./50, SO4/48,Cl/35.5, NO3/62. Add Na +K and Cl + NO3.
Calculate percentage of cations and anions. Plot cations by scaling off Ca, then Mg.
Plot anions by scaling off T.Alk. then SO4. Project cation and anion points to triangle

100
Figure 4.4 Plotting procedures for the Expanded Durov diagram: Convert mg/l to meq/l by division
(Ca/20, Mg/12, Na/23, K/39, T.Alk./50, SO4/48,Cl/35.5, NO3/62. Add Na +K and Cl + NO3.
Calculate percentage of cations and anions. Plot cations by scaling off Ca, then Mg.
Plot anions by scaling off T.Alk. then SO4. Project cation and anion points to square
As an example of the advantages of these diagrams, a data set from a coal-mining environment
has been selected and plotted in Figures 4.5 4.8. To demonstrate the kind of conclusions
that may be derived from these plots, the following information is provided:
Only about 30 per cent of the groundwater that has been sampled is unpolluted. The
latter is characterised by water of a calcium/magnesium bicarbonate composition.
The trend towards sodium enrichment, as depicted in the cation triangle, is due to
cation exchange, as groundwater flows from surrounding aquifers towards the mine.
The trend towards sulphate enrichment, as seen in the anion triangle, is due to pyrite
oxidation in the coal-mine. This shows the extent to which the groundwater regime has
already been polluted by mining.

101
Groundwater studies
Figure 4.5 Piper plot of groundwater chemistries from a mining environment
The Durov diagram is particularly handy because it also includes the electrical
conductivity of the water, thus giving a reflection of salt concentrations in the water. It
suggests that the increase in the salt load is associated with mining activities.
The calcium enrichment shown in the cation triangle of the trilinear diagrams
originates from the neutralisation of acid water from the mine by calcium carbonate in
the ground.
The natural groundwater and aquifer have significant buffering potential against
acidification, since only a few samples plot in the low pH range in the Durov diagram.
The Expanded Durov diagram categorises waters into nine classes. For this reason, this
diagram is for distinguishing between various groundwater populations, rather than
for studying trends, as was the case for the diagrams discussed above.
The SAR diagram suggests that sodium hazard to plants is low. Salinities are in the
medium to high range. Crops with some resistance to salinity and soils with a sandy-
loam character, may be irrigated successfully with the mine water.
Nitrate pollution, which is derived from fertiliser application by farmers, is present in
some borehole waters. This contamination cannot successfully be demonstrated by

102
Figure 4.6 Durov plot plot of groundwater chemistries from a mining environment
means of these diagrams, because of the dominance of other constituents. Other

diagrams such as bar, line or scatter will have to be used for this purpose.
The above example demonstrates the usefulness and limitations of these diagrams in
the interpretation of hydrochemical data. Clearly, many variations of the above may be
used to indicate values or demonstrate trends.
Through statistical methods, similar interpretations are possible. These methods will be
discussed in Section 4.4. Statistical evaluations of this kind are usually complex and only
understood by those familiar with statistical terminology. Reports for management should
therefore always include diagrams, which may be backed up by statistical evidence if considered
necessary.

103
Groundwater studies
Figure 4.7 SAR plot of groundwater chemistries from a mining environment

104
Figure 4.8 Expanded Durov plot of groundwater chemistries from a mining environment
105
Groundwater studies
4.3.5 Other graphics
Various possibilities exist to depict water chemistries for individual water samples. Presentations
which have been used are the line (including Schoeller), scatter, bar, vector, radial, star, pie and
polygon (including Stiff) diagrams (Lloyd and Heathcote, 1985) and Chernoff faces. Examples of
some of these plots are presented in Figures 4.94.12. It is important to note that in most dia-
grams the units are meq/l, thus reflecting the true reactive ratios of the constituents. However, in
the case of line, scatter and bar diagrams, the units may also be in mg/l. Results of chemical
analyses are usually reported by laboratories as mg/l.
Business graphics such as line, scatter and bar plots are often used to depict hydro-
geological data. These diagrams allow clear and concise presentations of a variety of aspects. By
changing the y-axis of the plot from a linear scale to a logarithmic scale, the range of values that
may be depicted becomes almost unlimited.
Figure 4.9 Stiff plots of water chemistries

106
Figure 4.10 Pie diagrams of water chemistries
Figure 4.11 Chernoff faces of water chemistries
107
Groundwater studies
Figure 4.12 Stacked bar charts of water chemistries
4.4 Statistical processing and presentation of data
Statistical interpretation of hydrogeological data has not received much attention in the past. The
main reason for this probably lies in the fact that groundwater flow and pollution transport are
governed by parabolic or elliptical differential equations, in contrast to surface water where
stochastic processes play a major role.
4.4.1 Basic statistics
The term Basis Statistics usually includes a variety of statistical interpretations, such as: mean,
mode, median, standard deviation, standard error, correlations, probabilities, t-test and analysis
of variance. An introduction to basic statistics can be found in handbooks such as Kachigan
(1986), and Runyon and Haber (1976). For more advanced discussions on the elementary theory
of statistics, the books of Hays (1988) and Kendall and Stuart (1979) may be consulted.
The results of many of these tests can be incorporated into a single diagram, referred to as
a box and whisker plot. A box and whisker plot usually depicts five statistical characteristics
for each variable (Figures 4.134.14). The central dots represent the mean, median or mode of
the variables. Around these points, boxes are constructed that show the upper and lower limits
of a statistic, such as the 25 and 75 percentiles or the standard errors of the means. The widths of
the boxes are arbitrary. Lastly, lines that extend as whiskers above and below boxes typically
reflect standard deviations or minima and maxima of variables. Box and whisker plots are
generally recognised for their usefulness in visual comparative statistics for frequently measured
variables.
Correlations amongst variables may be established by calculating correlation coefficients.
This application is particularly useful for studying water chemistries and the interdependence of
constituents within waters (Table 4.1).

108
Figure 4.13 Box and whisker plot of groundwater chemistries from a mining environment
Figure 4.14 Box and whisker plot of groundwater chemistries from a mining environment
109
Groundwater studies
Table 4.1 Correlations between chemical constituents in water from a mining environment (sample size 250)
Total
Ca Mg Na Cl SO4 alkaline
Ca 1.00 84* 0.15 0.07 0.97* 0.40

Mg 0.84* 1.00 0.21 0.12 0.85* 0.19
Na 0.15 0.21 1.00 0.90* 0.17 0.23
Cl 0.07 0.12 0.90* 1.00 0.07 0.13
SO4 0.97* 0.85* 0.17 0.07 1.00 0.50
Total alkaline 0.40 0.19 0.23 0.13 0.50 1.00
* Denotes a correlation better than 0.70.
Histograms may be used to display distributions of variables (Figure 4.15). Of particular

value is the bivariate histogram (Figure 4.16), in which two variables may be considered simul-
taneously. The interrelationship between sulphate concentrations and the pH of the water
becomes apparent in the latter presentation.
Figure 4.15 Histogram of groundwater pH levels from a mining environment
4.4.2 Time series
Many hydrogeological variables are time-dependent. This data may therefore be analysed by
time series methods, such as trend analysis and correlation methods.
Most commonly, polynomial regression methods are used to depict trends (for example,in
water-level fluctuations, for instance: Figure 4.17). A comparison between water-level responses
in different boreholes may be done by cross correlation, while auto correlation may be used to
investigate possible repetitive responses (Box and Jenkins, 1976).

110
Figure 4.16 Bivariate histogram of groundwater pH and sulphate from a mining environment
Figure 4.17 Fifth order polynomial regression interpolation of groundwater levels
111
Groundwater studies
Multivariate analysis can take on many forms. A commonly used predictive tool is
multiple linear regression.It enables a number of variables to be related to a single variable that
needs to be predicted (Hodgson, 1978). A typical example of how water levels may be predicted
and related to rainfall and abstraction is shown in Figure 4.18. The equation used for the
prediction of the water levels indicates a time lag of two months between the rainfall event and
maximum recharge to the aquifer. It should be stressed that prediction of this kind is only valid
within observed limits. If, for instance, a rainfall event occurs that exceeds all previous rainfall
intensities, predictions may no longer be valid.
Figure 4.18 Multiple linear regression showing actual and simulated water levels
Another form of multivariate analysis is that of grouping of variables that show relational
trends. Test statistics which may be used by the hydrogeologist include principal component
analysis, factor analysis and cluster analysis. All three of these evaluative procedures are based
on the principle of recognising patterns in data sets that contain many observations (Stevens,
1986). These techniques are excellent in, for instance, distinguishing between different popu-
lations. The results are similar to those obtained diagrammatically from the Piper and other
specialised hydrochemical diagrams.
Figure 4.19 is a factor analysis of the water quality data for the mining environment. The
close interrelationship between mine pollution variables (electrical conductivity, Ca, Mg, SO4
and Mn), versus nitrate pollution from farming and the natural pollution (Na, Cl and F) is
demonstrated by the groupings of the variables in the diagram.
Using advanced multivariate statistics has its pros and cons. An advantage of these
procedures is the wide range of tests that may be performed by informed statisticians The
disadvantage is that very few managers and planners understand the significance of these tests.
Many prefer graphs or maps above complex statistical arguments. This situation will, hopefully,
change, as more and more individuals become familiar with statistical procedures. Many current
statistical computer software packages are accompanied by well-documented manuals, making
the use of statistics easy and enjoyable.

112
Figure 4.19 Factor analysis of groundwater chemistries from a mining environment
4.5 Spatial information systems
4.5.1 Introduction
Spatial information systems have undergone significant development during the past five years
(Fulton, 1992; Haefner, 1992; Juracek, 1992). In its computerised form, these systems are referred
to as Geographic Information Systems (GIS). GIS in its simplest form constitutes two essential
components, namely a map drafting facility and a related data base. All elements, such as
points, lines, arcs and text that are drafted, are recorded as spatial entities within the data base.
All information in a GIS is structured. Structuring is done according to certain definable
conditions. A basic condition is usually that variables of the same type are stored together, to
allow ease of data retrieval. Each storage unit is usually referred to as a coverage. Hydrogeo-
logical data and information such as water levels, positions of boreholes and distribution
networks may therefore constitute separate coverages as part of a GIS data base.
One of the main advantages of standardising on GIS is that hydrogeological information
may be combined with other information sets. Typical information sets that are of use to the
hydrogeologist could be topography, soils, geology, meteorology, urban development, waste
disposal and mining and industrial development. The latter information is generally available
within other institutions where GIS has been in place for some time. It therefore does not need to
be entered by the hydrogeologist.
The advantage of having complementary information sets available at the start of a hydro-
geological investigation, is enormous. It allows, amongst other things, the identification of target
areas where hydrogeological investigations would be meaningful. As an example of the ver-
satility of such a system, the following constraints may for instance be imposed onto existing
information within the GIS, for the siting of a domestic waste facility:
it should not be closer than 1,000 m from a water-supply borehole;
it should not be within 500 m of urban or nature conservation areas;

113
Groundwater studies
it should not be on primary aquifers, chalk or dolomite;

the natural groundwater level should be at least 2 m below the surface;
the area should not have a surface gradient greater than 1:100;
it should be accessible from an existing highway;
sufficient clay should be available within a 2 km radius, for leachate control.
GIS is particularly well-suited for the above evaluation. It will, within seconds, identify all
the areas that conform to these requirements and show them on the computer screen. Detailed
hydrogeological investigations may then be concentrated within these areas, thus effecting a
considerable cost savings in terms of initial planning and site visits.
Many other scenarios may be investigated using GIS. The only constraints are:
the availability of information sets;
the quality of the information.
4.5.2 Entry of spatial information
Entry of information into GIS is complex and time-consuming. Point data may be imported
electronically from existing databases. Existing maps either have to be digitised or scanned.
Digitising involves the physical tracing of all lines, points and text into GIS coverages. Electronic
scanning of existing maps is initially much faster. This provides a raster image that may be used
as a coverage in the GIS. Raster images are not true GIS images and no intelligence can be
attached to them. They consist of millions of dots, with each dot representing a single scanned
pixel. Such images occupy vast storage space on computer disks. In view of these limitations, it
is almost always necessary to convert raster images to vector images.
Vector images may be obtained from raster images by activating suitable computer
software. This software is capable of recognising patterns and converting them into lines, text,
symbols and shadings. Vectorisation of raster images is a major task, mainly due to the relatively
poor printed quality of most maps. It may, for instance, take more than a month to vectorise a
single topographical map. Nevertheless, scanning and vectorising is the preferred technique for
inputing information into a GIS.
4.5.3 Conversion of point data
The conversion of point hydrogeological data into spatial information is the next logical step in
GIS application. This is most commonly done by contouring point values. Examples of variables
that may be contoured are transmissivities, storage coefficients, hydrochemical data and water
levels.
Various methodologies and numerous software packages are available for contouring of
data. Most of these methodologies are based on the transformation of irregularly spaced data to
regularly spaced values. This may be typically done through inverse distance or kriging
methodologies. Kriging provides the added advantage that confidence limits may be attached to
contours a facet that is often overlooked during contouring. A set of contours should therefore
ideally be accompanied by a second set of contours that show upper and lower limits for
contoured values.
Another aspect of contouring which is easily overlooked is that contouring should not be
done in areas where insufficient data are available. A check on the sufficiency of data may be
built into computerised programs, by limiting search distances. Areas where insufficient data are
available will thus be identified and no contours will be drawn there (Figures 4.204.21).
Contours should preferably be accompanied by a posting of the positions where data are
available, thus explaining the absence of contours in certain areas.
Three-dimensional contours are sometimes used to demonstrate specific aspects. It is

114
Figure 4.20 Smoothed water-level contours, using a 5,000 m search radius
Figure 4.21 Smoothed water-level contours, using a 500 m search radius
115
Groundwater studies
difficult for the viewer to orient himself with respect to these presentations. For this reason,
these contours should be accompanied by (i) two-dimensional contours, (ii) a vertical scale and
(iii) an indication of the drawing orientation. In addition, equi-potential contours may be
embedded onto the three-dimensional surface, thus enabling the quantification of variations in
the diagram (Figures 4.224.23).
A recent development in GIS is the transformation of spatial information into grid images.
In these applications, the grid images are very coarse, typically consisting only of a 40 x 30
matrix. The purpose of such transformations is to convert spatial information into averaged
grids (Figure 4.24) which in turn, may again be analysed by using standard GIS tools. A typical
application would be the transformation of water-level and topographic information into two
raster images of identical dimensions. Thereafter, the interrelationship between these two
variables may be studied. The application of grid procedure particularly lies in the comparison
of complex information sets.
The combination of GIS technology with hydrogeological tools such as specialised
interpretations and modelling, has been under development for a number of years. Examples are
the NWIS-II (USGS, 1991), REGIS (TNO, 1993; NITG-TNO, 2000) and GGIS (IGS, 1994) software.
All three software packages will have the capability of converting point data to spatial data,
integrating data sets with GIS and executing flow and mass transport models from within the
GIS environment.
4.5.4 Hydrogeological maps and the UNESCO code
Hydrogeological maps to depict groundwater related information (see also Chapter 10) have
been around for many years. Examples of such maps are plentiful (UNESCO, 1983). UNESCO
suggested standardising hydrogeological codes and maps in 1963; they were revised in 1983.
These publications have been well received and many maps have been produced using this code
or variations of it.
One of the many examples where the UNESCO code has been used is the Hydrogeo-
logical Map of England and Wales (United Kingdom, 1977). Many other hydrogeological maps
have been published of Europe and other parts of the world. The UNESCO code of 1983 lists
numerous hydrogeological maps world-wide.
The tendency for hydrogeological maps is to have a large central base map showing some
hydrogeological variables. Surrounding this map are other map inserts. In the case of the
Lesotho Map, for instance, (Ardino 1994), information on the following topics has been included
on the map:
hydrogeological legend;
summary of hydrogeology for Lesotho;
climatic information, including rainfall time series, contours and wind directions;
base flow and run-off information;
water quality information tabled and plotted onto Piper diagrams;
geological and hydrogeological sections;
groundwater contours and flow nets;
tables showing individual and interesting characteristics that relate to the geohydro-
logy at specific localities on the map.
Other features that may typically be considered are groundwater vulnerability in terms of
over-exploitation and pollution.
In view of the great number of hydrogeological maps that are being produced world-
wide, standardisation of colours and codes is recommended. It is recommended that the
UNESCO codes of 1963 and 1983 should be used as closely as possible.

116
Figure 4.22 Three-dimensional presentation of groundwater levels
Figure 4.23 Three-dimensional presentation of groundwater contours at metre intervals
117
Groundwater studies
Figure 4.24 Grid presentation for evaluation of averaged groundwater levels
4.6 Conclusions
In terms of hydrological data processing and presentation, the only limit is the imagination of
the individual dealing with these matters. In this Chapter, an attempt has been made to present
the techniques that are available to everybody. Many more sophisticated techniques are available
often requiring significant computing power. When searching for a specific processing and
presentation technique, the norm should always be:
the level of expertise of the individual;
the quality of the data;
the availability of equipment such as computing equipment;
the level of expertise of the individuals to whom the information is to be presented.
ARDUINO, G.; BONO, P.; DEL SETTE, P., 1994. Hydrogeological Map of Lesotho, Scale 1:300,000. Depar-
tment of Water Affairs, Maseru.
BOWER, C. A.; OGATA, G.; TUCKER, J. M., 1968. Sodium Hazard of Waters as Influenced by Leaching
Fraction and by Precipitation or Solution of Calcium Carbonate. Soil Science, Vol. 106,
pp. 2934.
BOX, G. E. P.; JENKINS, G. M., 1976. Time Series Analysis: Forecasting and Control. Holden-Day, San
Francisco.
DUROV, S. A., 1948. Natural Waters and Graphic Representation of their Composition. Dokl. Akad.
Nauk SSSR, Vol. 59, pp. 8790.

118
FULTON, J. L., 1992. Development of Spatial Data Guidelines and Standards: Spatial Data Set
Documentation to Support Hydrological Analysis in the U.S. Geological Survey. Geographic
Information Systems (GIS) and Mapping Practices and Standards, STP 1126, ASTM,
Philadelphia, PA, pp. 307.
HAEFNER, R. J., 1992. Use of a Geographic Information System to Evaluate Potential Sites for Public-
Water-Supply Wells on Long Island, New York. Open-file Report 91-182, USGS, Denver, 37 pp.
HAYS, W. L., 1988. Statistics (fourth edition). CDS College Publishing, New York.
HODGSON, F. D. I., 1978. The Use of Multiple Linear Regression in Simulating Groundwater
Responses. Ground Water, Vol. 16, pp. 24953.
IGS, 1994. GGIS A Geographic Groundwater Interpretation System under Arc/Info. Open-file Report
to the WRC, Institute for Groundwater Studies, Bloemfontein, South Africa, 120 pp.
JURACEK, K. E., 1992. Use of a Geographic Information System to Assist with Studies of the Availability
and Use of Water in Kansas. Open-file Report 92-142, USGS, Denver, 14 pp.
KACHIGAN. S. K., 1986. Statistical Analysis: An Interdisciplinary Introduction to Univariate and Multi-
variate Methods. Redius Press, New York.
KENDALL, M.; STUART, A., 1979. The Advanced Theory of Statistics (Vol. 2). Hafner, New York.
KRUSEMAN, G. P.; DE RIDDER, N. A., 1991. Analysis and Evaluation of Pumping Test Data (second
edition). International Institute for Land Reclamation and Improvement, Publ. 29,
Wageningen, 377 pp.
LLOYD, J. W.; HEATHCOTE, J. A., 1985. Natural Inorganic Hydrochemistry in Relation to Groundwater.
Clarendon Press, Oxford, 296 pp.
LLOYD, J. W., 1965. The Hydrochemistry of Aquifers of North-Eastern Jordan. Journ. Hydrol., Vol. 3,
pp. 319 30.
NITG-TNO, 2000. Regional Geohydrological Information System (REGIS). REGIS Demo CD.
NITG-TNO CD, Delft.
PIPER, A., 1944. A Graphic Procedure in the Geochemical Interpretation of Water Analysis. Trans.
Am. Geophys. Union, Vol. 25, pp. 91423.
RUNYON, R. P.; HARPER, A., 1976. Fundamentals of Behavioural Statistics. Addison-Wesley, Reading,
MA.
SCHOELLER, H., 1955. Gochimie des eaux souterraines, application aux eaux des gisements de ptrole.
Revue de LInstitut Franais du Ptrole et An. de Combustibles Liquides, 213 pp.
STEVENS, J., 1986. Applied Multivariate Statistics for the Social Sciences. Erlbaum, Hillsdale, NJ.
STIFF, H. A., 1951. The Interpretation of Chemical Water Analysis by Means of Patterns. J. Petr. Tech-
nology, Vol. 3, pp. 157.
TNO, 1993. REGIS Regional Hydrogeological Information System. Information Document, TNO
Institute for Applied Geoscience, Delft, 19 pp.
UNESCO, 1963. International Legend for Hydrogeological Maps. Unesco, Paris.
UNESCO, 1983. International Legend for Hydrogeological Maps. Revised Edition, Unesco, Paris.
United Kingdom, 1977. Hydrogeological Map of England and Wales, 1:625,000. Institute of Geological
Sciences, London.
USGS, 1991. System Requirements Specification for the U.S. Geological Surveys National Water Infor-
mation System II. Open-file Report 91-525, United States Geological Survey (USGS), Denver,
622 pp.
WILCOX, L.V., 1955. Classification and Use of Irrigation Xaters. U.S. Department of Agriculture.
Circular No. 969, 19 pp.

119
5 Remote sensing techniques

f o r g ro u n d w a t e r p ro s p e c t i o n
5.1 Introduction
5.1.1 General
The remote sensing (RS) techniques discussed in this chapter are based on the interaction of
electromagnetic radiation and matter. This interaction takes place at the surface of the earth (the
depth penetration of the traditional RS techniques is negligible).
The image of the surface can be interpreted using a variety of techniques. The visual inter-
pretation of a photographic hardcopy has long been a standard technique in earth and
vegetation sciences, but now that computers and appropriate software have become easily
affordable, many techniques that used to be restricted to specialised data processing centres
have an interpretation of an area should also involve extrapolation to the unseen deeper part
become available even to earth scientists with limited computer facilities. Personal computers
can now be used to process satellite images, scan aerial photographs, apply Geographical Infor-
mation Systems (GIS), and run sophisticated statistical packages and mathematical models.
5.1.2 Techniques and approaches
The many aspects of hydrology and hydrogeology that may be mapped include geological, soil
or geomorphological units, drainage pattern, slope, vegetation, water bodies, lineaments, geo-
logical structures, man-made structures, water points and villages. The speed and accuracy of
mapping can be increased if aerial photographs and/or photographic prints of satellite images
are used (Allum, 1966).
The availability, distribution and quality of groundwater as well as its natural flow
pattern is largely determined by the characteristics of the subsoil and the climatic conditions.
Since remote sensing techniques basically reflect properties of the surface, in order to better
understand the hydrogeology of the area, an interpretation of an area should also involve extra-
polation to the unseen underground. This will lead to a conceptual three-dimensional model of
the groundwater system. No standard recipes can be given for the development of such a model.
The interpreter has to integrate the features he can see on his image with all the other available
information on the area, such as thematic maps (geological, vegetation), meteorological data,
hydrological data. Fieldwork is imperative to check the validity of the interpretation (ground
truthing).
For simple groundwater exploration, aerial photographs and RS are used to map features
likely to be high-yielding zones, and to identify favourable structures or deposits for ground-
water accumulation.

121
Groundwater studies
The applications of remote sensing data in water resources studies (Hansmann et al., 1992;
Ling Hengzhang, 1985; Nieuwenhuis and Thunnissen, 1990; Schultz and Barrett, 1989; Waters et
al., 1990) can be divided into several approaches.
(i) Visual interpretation using computers with image processing
Modern image processing software facilitates interpretation. Using the traditional prints of an
image a trade-off had to be made between the overall quality of the image and the detectibility
of specific features. Todays high quality colour monitors enable the interpretation to be done
from the screen. Several digital enhancement techniques are usually available in the image
processing software, the analyst can use information from several bands, and a selected window
can be enlarged; this improves image interpretation. All these techniques can be fine-tuned in
order to enhance the detectibility of a specific feature, e.g. the drainage pattern, the lineaments,
the vegetation cover and the lithology. Once one feature has been mapped, another specific
process can be applied to another feature.
(ii) Modelling the interaction of EM radiation with matter
In water studies modern image processing software mainly focus on the energy balance at the
ground surface, the correlation between spectral response and vegetation characteristics and
meteorological applications (Baily, 1990; Belward, 1991; Lagouarde and Brunet, 1993).
The actual evapotranspiration and soil water content plays a very important role in the
energy balance at the ground surface. There are theoretical models that attempt to relate the
thermal infrared radiation with the soil moisture content and the rate of evaporation (Mauser,
1990). For example, the so called cold cloud temperature registered by meteorological satellites
can be correlated to rainfall. The spectral response of vegetation can be modelled and used to
estimate vegetation density, vegetation type and moisture conditions.
(iii) Classification, automatic interpretation and feature extraction
Much research is being done on the computerisation the interpretation of satellite images. Geo-
logists want geological maps, soil scientists pedological maps and hydrologists hydrological
maps all readily made by a computer. The processing is based on the fact that different
thematic units may have a different spectral response. There are simple or sophisticated rules,
usually statistically based, for deciding how to interpret the area. Often the algorithms must be
fine-tuned for the theme of interest. Pattern recognition techniques may also be used to decide
how to classify an area.
(iv) Integration of RS, GIS and modelling techniques
Modern RS/GIS software enables digital RS data to be merged with other data.
Digitised maps, e.g. road networks, or geological maps, can be combined with the images,
making the interpretation easier. An existing thematic map can be visualised together with the
information from a satellite image.
In most studies the conceptual models essential as the backbone of the investigations,
need to be quantified. It is not sufficient to know where water is stored; amounts, quality and
fluxes also need to be known. Here again the techniques combining RS, GIS and simple models
can be used to estimate fluxes. Mapping units can be based on geomorphology, geology, soils
and vegetation. Some of the components of the hydrological cycle, such as actual evapo-
transpiration and groundwater recharge potential, may be estimated for different mapping

122
Remote sensing techniques for groundwater prospection
units. Irrigated area, and the associated consumption of surface water and groundwater can be
estimated using RS techniques. Better interpolation of rainfall data is possible if the data are
combined with vegetation indices and digital terrain models.
Hydrological models may be directly programmed in the GIS, or be linked to the GIS,
facilitating the preparation of input files for the model and serving as a post processor to display
the results of the models calculations done by the model.
Flow systems can be schematised as cross-sections. Simple cross-section flow models can
be used to explain the observations, or force the investigator to review his model when the
results do not tally with the observations.
Airborne geophysical information may be displayed together with RS data, or a satellite
colour image may be wrapped around a Digital Terrain Model (DTM) in an perspective view.
It must be stressed that digital techniques offer many advantages but cannot replace a
sound knowledge of geomorphology, geology, hydrology, and soil and vegetation sciences.
On the following pages an introduction in remote sensing techniques and a selection of
case studies will be presented to demonstrate the different aspects of the use of RS, aerial
photography and GIS in hydrogeology.
5.2 Principles of remote sensing and GIS
5.2.1 Remote sensing
Remote sensing, in general terms is defined as collecting and interpreting information about a
target without physical contact with the object (Sabins, 1986). The observations are usually made
from aircraft or satellites. The definition would include the measurement of gravity, electrical
and magnetic fields, radioactivity and electromagnetic radiation, but these techniques are
considered part of airborne geophysics (see chapter 6).
The term remote sensing is commonly restricted to methods that employ electromagnetic
energy (light, heat, radar waves) as a means of measuring the characteristics of the earths
surface. The most commonly used RS products are aerial photographs and satellite images.
5.2.2 Physical fundamentals
(i) Electromagnetic radiation
Electromagnetic radiation is a wave phenomenon propagating with the speed of light, which is
3 . 10 8 m/sec.
For waves the velocity (c), wavelength (l) and frequency(f) are related by the formula
c = l . f.
The different classes of EM radiation, such as X rays, UV, visible light, infrared, radar and
radiowaves are based on the wavelength or frequency. Figure 5.1 gives an indication of the
nomenclature of different wavelength ranges.
(ii) Interactions with matter
Figure 5.2 schematically depicts the interaction mechanisms between electromagnetic radiation
and matter. When the radiation encounters matter it is called the incident radiation. The
following interactions can take place: reflection, scattering, transmission, absorbtion and
emission.
The interaction between EM radiation and matter is usually a combination of the basic

123
Groundwater studies
Figure 5.1 Electromagnetic spectrum
Figure 5.2 Interaction between EM radiation and matter
interaction phenomena described above. What is important is that all interactions vary as a
function of the wavelength. Some of these interactions are recorded by a remote sensing system
and the characteristics of the matter may be interpreted. By far the most important properties
recorded are the reflection characteristics.
5.2.3 Aerospace imaging systems
The radiation returning from the surface of the earth is recorded by a photographic or scanning
system. The two types of imaging systems are described below. A schematic representation of
the path of the radiation through the atmosphere is presented in Figure 5.3.

124
Figure 5.3 Path and interactions between sun and sensor
(i) Photographic systems
Black and white films record the average intensity of the reflected light, colour films record the
intensity of the three primary colours similar to the human eye, whereas infrared films are
sensitive to near-IR radiation. Aerial photographs are usually taken from an aircraft though
many photographs are have been taken during manned space missions. The aerial photographs
taken in such a way that there is approximately 60% overlap between adjacent photos. Exam-
ination of the shared part of two photographs (a so-called stereo pair) through a stereoscope
yields a three-dimensional impression of the landscape. These stereo pairs are used for a wide
range of applications, such as the production of topographical maps and the mapping of soil,
vegetation, land use, forestry, geology, geomorphology and hydrology features. (Verstappen,
1977; Meijerink and Van Wijngaarden, 1997; Sabins, 1986; Von Bandat, 1962).
(ii) Scanning systems
Scanning systems are more complicated than the lens-based photographic system. They divide
an image into an array of millions of small square areas, called pixels, (picture element). These
pixels are scanned and the incoming radiation for one or more wavelength intervals is recorded
for each individual pixel. The wavelength intervals are called bands. These systems can be
mounted in an aircraft, but most such data is collected by satellites.
Three types of resolution are distinguished. The spatial resolution is the size of the pixels,
e.g. 30 x 30 metres for the Landsat Thematic Mapper, (Landsat TM). The spectral resolution is
determined by the number of bands. The temporal resolution is determined by the frequency of
the satellite overpasses. The recorded radiation for every pixel and every band is recorded as an
8 bit digital number (1 byte), which is equivalent to a number between 1 and 256.

125
Groundwater studies
5.2.4 Frequency domains
The imaging system may record one or more frequency or wavelength intervals.
A black and white film will record all visible light. Colour film will record the three
primary colours. Sensors are designed to detect a specific wavelength or frequency interval. A
major division in domains is made between reflected and emitted radiation.
(i) Reflected radiation
Visible and near infra red radiation is reflected, absorbed and scattered by the earths surface.
The reflected light is recorded by the sensor or photographic film. The differences on the image
are mainly due to the difference in reflectivity of the Earths surface.
(ii) Emitted radiation
Matter emits electromagnetic radiation. The intensity and wavelength are functions of temper-
ature and properties of the radiating material. The wavelength of the radiation increases with
temperature. At high temperature the wavelength enters the domain of visible light (e.g. on
heating, a piece of iron changes in colour from deep red to yellow).
Thermal IR images may yield valuable hydrogeological information. In many cases the
discriminating power of thermal IR with respect to lithology is higher than that of visible or
near-infrared bands. Seepage and springs discharging into rivers, lakes or the sea may be
detected, due to the temperature differences. Fracture patterns may show up very distinctively,
as will be shown in one of the case studies. Thermal images can be used to estimate actual
evapotranspiration and its spatial distribution.
Nevertheless, thermal IR imagery is rarely used, because thermal IR images are most
useful if they are taken during night time, when solar heating is absent. Therefore, IR surveys
have to be flown at night and cannot be combined with, for example, a survey of normal aerial
photographs. Several satellites have thermal IR sensors. At present the Landsat TM with a pixel
size of 120 x 120 metres has the highest resolution. This resolution is not enough for detailed
investigations. Moreover, most of the TM images are acquired in the daytime, as during the
night only one of the seven bands is recorded.
5.2.5 Imaging satellites
Figure 5.4 shows the characteristics of common satellites.

A special feature of the French SPOT Satellite is its capability to look sideways (off-nadir).
This allows of an image of the same area to be produced from different orbital passes. This
property makes it possible to produce satellite stereo images that can be studied with a normal
stereoscope, as well as the high frequency imaging of an area. As the product has to be ordered
specifically, it is rather expensive.
5.2.6 Digital image processing
The raw data sent by the satellites needs processing to transform it into visual information,
either as a print or as an image displayed on the screen of a computer. The processing can be
done by the seller of the data, a specialised firm, or directly by the user on a PC or a work-
station.
The following data processing categories can be grouped:

126
Figure 5.4 Satellite characteristics
Corrections:
- correction for technical distortions;
- correction for atmospheric distortions;
- correction for geometrical distortions;
- image enhancements;
- contrast enhancements;
- filtering (edge and lineament enhancements);
- colour coding;
- transformations:
- principal component images;
- ratio images.
Classification:
decision rules are used to automatically classify (interpret) an image in different
thematic classes.These classes may be type of soil, type of geology, landuse,
vegetation, etc.
The image processing is schematised in Figure 5.5. The figure depicts four data arrays
representing four spectral bands. Every element of the arrays represents a pixel and is stored as
a digital number (DN). The higher the number, the higher the energy received (reflection) for
that wavelength. The two XYgraphs represent the spectral reflection curves or spectral sign-
atures for two pixels, A and B. They represent the reflectivity of a pixel as a function of the
wavelength (). This curve is representative of different land and vegetation types. A com-
bination of three bands can be represented by a three-dimensional vector. Clusters of points in
n-dimensional space representing n bands of a set of pixels are called the feature space. A band
can be represented as a grey-tone image by assigning black to the lowest digital number and
white to the highest; the intermediate values are represented by grey shades. Three bands can be

127
Groundwater studies
Figure 5.5 From scanning to display
represented by a colour image, by assigning three bands to the intensity of the three primary
colours: red, green and blue. For further reading on digital image processing and colour theory,
see one of the specialised books (e.g. Lillesand and Kiefer, 1987; Mather, 1987; Mulder, 1982 and
1988; Niblack, 1986; Sabins, 1986; Wilkinson, 1991).
5.3 Geographical information systems (GIS)
GIS has been a fast developing technology in spatial information processing. An overlap exists
with pure image processing and many software packages offer both image processing and GIS
capabilities. GIS can enter, store, retrieve, process and display spatial information in the form of
maps or images (satellite, aerial photographs) including a database which is linked to the
mapping units (Meijerink et al., 1987). This database is the equivalent of the legend in traditional
maps.
On a map three elementary entities can be distinguished: points, lines or segments and
surfaces or polygons. Points may represent villages, wells, met stations, lines may represent
roads, rivers, faults, whereas polygons may represent countries, districts, soil classes, geological
formations, or any other mappable units. Information on the points, lines, polygons are called
attributes and are stored in a database. One record in the database contains the information on a

128
geographical entity (point, line, polygon). The attributes stored in the database can be the
number of lanes, quality, type, etc. of a road, the discharge, water quality, etc. of a river, the
names, population, number of schools, etc. in towns and villages, the population of a district,
etc., soil data like permeability, thickness, salinity, etc.
The two main data structures to store spatial information are vector and raster data
structures.
In a vector-based system, points are represented by a coordinate pair (X,Y), a line is
represented as a series of coordinate pairs approximately defining the line and a
polygon is represented by one or more lines surrounding an area (Figure 5.6).
In a raster-based system, the area is overlain by a raster. A point is represented by
one cell, a line by a sequence of adjacent cells and a polygon by a cluster of cells
filling the polygon.
Figure 5.6 Raster and vector data structures
Data retrieval from non-spatial databases is only possible if selection criteria are applied
(conditions) to the database itself, e.g. retrieve all wells with a yield of more than 10 m3/s. A
GIS can also deal with geographical queries such as: select all wells which are at less than 3 km
from a village, or a combination with a logical condition such as: retrieve all wells closer than
3 km from a village and yielding of more than 10 m3/s.
Maps may be combined or modified using GIS. For example a detailed geological map
has an attribute database with the following three fields:
Era: Palaeozoic, Mesozoic, Tertiary;
System: Devonian, Carboniferous, Permian, Jurassic, Cretaceous, Palaeocene,
Miocene;
Lithology: granite_1, granite_2, granite_3, sandstone_1, sandstone_2, limestone.
One can combine all polygons from the same Era and the result will be a simple geo-
logical map with classes Palaeozoic, Mesozoic and Tertiary. Similarly, one can regroup the three
granites and two sandstones producing a simplified lithological map.

129
Groundwater studies
Another process a GIS can handle is the overlaying or combining of two maps. A simple
erosion index consists of a combination of slope, rainfall and soiltype. These parameters can be
mapped on three separate maps. A combination gives a new map of the relative erosion hazard.
Raster and vector data structures each have their advantages and disadvantages; which
approach is more suitable depends on the task to be done. In general, the vector approach
is more suitable for large database-oriented GIS projects, whereas the raster-based systems
are very suitable for geographical modelling and analysis. Some GIS packages have both
approaches integrated, and can convert from vector to raster and vice versa. The raster data
structure used in a GIS is also used to store satellite images. Many models (groundwater, surface
water, erosion, crop yield) use a raster as their main data structure, and Digital Elevation Models
(DTM) can also be represented in a raster data structure. Each cell of the raster has as attribute
the average elevation of that particular cell. Therefore the linking and combination of techno-
logies is straightforward. A cell of a raster based GIS representing a certain soil, may also
represent a pixel with a certain DN from a satellite image, or represent a cell in a groundwater
model for which the phreatic level can be calculated. The only condition is that the raster should
represent the same area on the ground and the raster dimensions and cell sizes must be equal in
all cases. All spatial data, can be integrated in a GIS be they thematic map units, remotely sensed
images, administrative boundaries, topography, demographical data, towns, roads, wells, rivers,
interpolated data from rainfall or groundwater levels, the simulated water levels, or river
discharges.
The GIS has the capability of processing the information, which may reveal certain
relationships, and ofvisualising different types of information simultaneously. So we can have
satellite images with a road map superimposed on a three-dimensional representation of the
landscape based on a DTM with a satellite colour composite image draped over it, or continuous
geophysical data like gravity or magnetic field combined with satellite images. For further
reading on GIS in general see Burrough and McDonnell (1998) or Aronoff (1989). For application
in hydrology see Meijerink et al. (1994).
5.4 Required resources
Preprocessing yields two main products:

a) The digital data on computer compatible storage medium like tape or CD-ROM.
b) A negative or a photographic colour print.
(i) Aerial photographs
Existing aerial photographs can often be obtained for the cost of printing (typically a few US$).
However, even for a relatively small area the complete coverage may consist of hundreds of
photographs. This is also due to the large overlap.
(ii) Satellite images
Satellite images are rather costly. In 1995 commercial prices ranged between US$4,000 and
US$ 4,500 for a full scene. It is also possible to order subscenes, ranging in price from a few
hundred to several thousands of dollars, depending on the size. However, many images dating
from before 1985 can be acquired for prices ranging from US$250 to US$ 500. It is likely that
prices will fall when the raw unprocessed data is available at low cost and commercial
companies can add value by processing. Information on products can also be obtained via
Internet.

130
The requirements for hardware are modest. Most software is available for Windows,
requiring a computer with a 486 or higher processor. Until recently, special equipment was
needed to read the computer compatible tapes containing the data. Since data can now be
ordered on a CD-ROM, only a CD-ROM player is required to transfer the image to the hard disk
of the computer.
Colour inkjet printers can produce images of acceptable quality; however, specialised and
expensive equipment is still needed for the high-quality photographic prints. With Microsoft
Windows evolving as a dominant operating system, the integration of maps and images in a
document will become routine.
ALLUM, J. A. E., 1966. Photogeology and Regional Mapping. Pergamon Press, Elmsford, N.Y.
ARONOFF, S., 1990. Geographic Information Systems: A Management Perspective. WLD Publications,
Ottawa.
BAILEY, J. O., 1990. The Potential Value of Remotely Sensed Data in the Assessment of Evapo-
transpiration and Evaporation. Remote Sensing Reviews, Vol. 4, pp. 249377.
BELWARD A. S., 1991. Spectral Characteristics of Vegetation, Soil and Water in the Visible, Near-
Infrared and Middle-Infrared Wavelengths. In: Belward and Valenzuela (eds.), Remote
Sensing Applications and GIS for Resources Management in Developing Countries (Eurocourses).
Kluwer Acad. Publ., Dordrecht, pp. 3154.
BURROUGH, P. A.; MCDONNELL, R. A., 1998. Principles of Geographical Information Systems. Oxford
University Press, Oxford.
COLWELL R. N., (ed.), 1983. Manual of Remote Sensing. Amer. Soc. of Photogrammetry, Virginia.
DURRET, H. J. (ed.), 1987. Color and the Computer. Boston Academic Press Inc., Boston, 299 pp.
ELACHI C., 1987. Introduction to the Physics and Techniques of Remote Sensing. John Wiley and Sons,
New York.
HANSMANN, B. C.; MEIJERINK, A. M. J.; KODITUWAKKU, K. A. W., 1992. An Inductive Approach for
Groundwater Exploration. ITC Journal 19921993, pp. 26976.
HENGZHANG, L., 1985. Applications of Remote Sensing Technology to Water Resources Exploi-
tation in Mountainous Areas in Southwest China. In: Remote Sensing Applications to
Highland Development. Proc. Regional Seminar, UNDP/ESCAP, 711 Nov. 1985, Kathmandu,
Nepal, pp. 805.
JACKSON R. D., 1983. Spectral Indices in N-Space. Remote Sensing of the Environment, No. 25,
pp. 295309.
LAGOUARDE J. P., 1991. Use of the NOAA AVHRR Data Combined with an Agrometeorological
Model of Evaporation Mapping. International Journal of Remote Sensing, Vol. 12,
pp. 185364.
LAGOUARDE, J. P.; BRUNET, Y., 1993. A Simple Model for Estimating the Daily Upward Longwave
Surface Radiation Flux from NOAA-AVHRR Data. International Journal of Remote Sensing,
No. 14, pp. 90725.
LILLESAND, T. M.; KIEFER, R. W., 1987. Remote Sensing and Image Interpretation. John Wiley and
Sons, New York.
MAUSER, W., 1990. Modelling the Spatial Variability of Soil Moisture and Evapotranspiration
with Remote Sensing. In: Proceedings, International Symposium Remote Sensing and Water
resources (IAH-BCRS), Enschede, The Netherlands, pp. 249260.
MEIJERINK A. M. J. ET AL., 1994. Introduction to the Use of Geographical Information Systems for
Practical Hydrology. ITC-UNESCO, Enschede.

131
Groundwater studies
MEIJERINK, A. M. J.; VALENZUELA, C. R.; STEWARD, A., 1987. The Integrated Land and Watershed
Management Information System (ILWIS). ITC Publication No. 7, 115 pp.
MULDER, N. J., 1982. Methodology of Colour Coding MSS and Other Data. Archives Int. Soc.
Photogrammetry and Remote Sensing, Helsinki Symposium, Commission III.
MULDER, N. J., 1988. Digital Image Processing, Computer Aided Classification and Mapping. In:
A. W. Kchler and I. S. Zonneveld (eds.), Vegetation Mapping. Kluwer Ac. Publ., Dordrecht.
NIBLACK, W., 1986. An Introduction to Digital Image Processing. Prentice Hall International, 215 pp.
NIEUWENHUIS, G. J. A.; THUNNISSEN, H. A. M., 1990. Application of Remote Sensing in Water
Management. In: Proceedings, International Symposium Remote Sensing and Water resources,
(IAH-BCRS), Enschede, The Netherlands, pp. 22948.
SABINS, F. F., 1986. Remote Sensing: Principles and Interpretation. W. H. Freeman & Co, New York.
SCHULTZ, G. A.; BARRETT, E.C., 1989. Advances on Remote Sensing for Hydrology and Water Resources
Management. UNESCO, Techn. Doc. IHP, Project 5.1, Paris.
VERSTAPPEN, H. TH., 1977. Remote Sensing in Geomorphology. Elsevier Scient. Publ. Co.
VON BANDAT, H. F., 1962. Aerogeology. Gulf Publishing Compagny, Houston, Texas.
WATERS, P.; GREENBAUM, D.; SMART, P. L.; OSMASTON, H., 1990. Applications of Remote Sensing to
Groundwater Hydrology. Remote Sensing Reviews, Vol. 4, pp. 22364.
WILKINSON, G. G., 1991. The Processing and Interpretation of Remotely-Sensed Satellite Imagery,
A Current View. In: Belward and Valenzuela (eds.), Remote Sensing Applications and GIS for
Resources Management in Developing Countries (Eurocourses). Kluwer Acad. Publ.,
Dordrecht, pp. 97110.

132
6 Geophysical techniques
i n g ro u n d w a t e r i n v e s t i g a t i o n s
6.1 Introduction
For many years geophysical exploration techniques have proved to be efficient tools in ground-
water exploration, not only in the direct detection of the presence of water, but also in the
estimation of aquifer size and properties and groundwater quality. As objectives in groundwater
exploration may be named: detection of the position and extent and volume of potential
aquifers, location of faults and fracture zones below undifferentiated surface covers, detection of
fresh/saline water interfaces, estimation of the depth of the water table, detection and tracing of
groundwater pollution, detection of aquitards and connections between different aquifers.
Through correlation with boreholes and pump tests an estimation can be made of the variation
of porosity, permeability and transmissivity through an aquifer.
This chapter presents an overview of methods which are at present applied to ground-
water exploration and it does not pretend to be complete. It should be noted that there has been
enormous progress in the development of applications since the early 1980s, so geophysical
methods other than the traditional ones have been included.
This chapter discusses the various methods in a sequence of large-scale surveys (regional
scale, airborne methods) to small-scale ones (local scale, ground-based methods). The rationale
behind this is that the surveying process aims at investigating large areas with fast, efficient and
relatively cheap methods, to select those parts which seem attractive or even favourable for
further, more detailed surveys. These will give more information about the target, but are at the
same time more costly. By surveying in a sequence of large scale to small scale an efficient and
cost effective result may be obtained. It should be realised here that the airborne and waterborne
techniques are actually variations on the original ground-based techniques.
6.2 Magnetic methods
From the point of view of mapping and exploration using the magnetic method, the whole earth
may first be considered in three parts: core, mantle and crust.
6.2.1 The source of the earths magnetic field
(i) Data collection
The magnetic field of the earth originates in the molten part of the earths core. This core field
varies smoothly and predictably over the surface of the earth and is defined conventionally by

133
Groundwater studies
the International Geomagnetic Reference Field (IGRF) which changes at any one locality slowly
and progressively with time. The field at any given point is usually defined in terms of its total
field strength (T), its inclination (I) and the deviation between true and magnetic north or
declination (D). In geophysical mapping, the scalar magnitude of the geomagnetic field is (with
few exceptions) the only element that has been measured; this frees the geophysicist from the
need to relate his observation to any reference direction and so makes rapid surveying, even
from a moving aircraft, relatively straightforward using suitable electronic magnetometers.
The earths mantle is too hot to be magnetic since the magnetic properties of rocks
generally disappear when they are heated above the Curie point temperature. This falls in the
range 550600C so, with typical geothermal gradients, this temperature is reached at depths
towards the base of the crust.
(ii) The crust, the source of magnetic anomalies
Most of the rocks within the earths shallow crust are cooler than the Curie point temperature and
may display magnetic properties, either remanent or induced. Remanent magnetisation is a per-
manent setting of the magnetic fabric of a rock that may persist for hundreds of millions of years
and may have been acquired when the rock was first formed or when it was re-heated during
metamorphism. Induced magnetisation, on the other hand, is a temporary effect, acquired whenever
a susceptible material is located in a magnetic field such as that of the earths core. It is very
difficult to say whether the magnetism present in rocks in situ is induced or remanent, except in
the case of outcropping rocks that can be sampled and subjected to laboratory measurements.
Both the induced and the remanent magnetisation of rock are carried only by the
magnetic minerals present. Magnetic properties are addressed as magnetic susceptibility: the
degree to which the mineral particles can be magnetised. Table 6.1 presents some susceptibility
values of minerals, ores and rocks.
Table 6.1 Magnetic susceptibilities of minerals, ores, and rocks
Mineral/ore Susceptibility (S.I.) Rock type Susceptibility (S.I.)
Pure magnetite 15 Sediments 0.00.001

Magnetite ore 0.0714 Metamorphic rocks 0.0001 0.001
Ilmenite ore 0.34 Granites 0.0001 0.01
Pyrrhotite 0.0010.1 Basic plutonics 0.0001 0.1
Hematite ore 0.0004 0.01 Volcanic rocks 0.00 0.1
Pyrite 0.00010.005 Banded iron formations Very high remanent
magnetisation
The combined effects of the induced and remanent magnetisation of rocks imposes local
variations anomalies on the earths main magnetic field. Departures of actual magnetic field
values from those predicted by the IGRF may be mapped out in a survey using ground or
airborne techniques to produce maps or images that reflect the distribution of magnetic pro-
perties of crustal rocks. Since most sedimentary rocks are virtually non-magnetic the anomalies
mapped out over sedimentary basins or over areas of surficial cover, dense vegetation or water
cover provide an important means of interpreting the depth and structure of underlying
igneous and metamorphic rocks. This has led to the widespread application of aeromagnetic
surveying to extend the mapping of the so-called magnetic basement into the large areas of the
world where the igneous and metamorphic rocks do not outcrop.

134
Geophysical techniques in groundwater investigations
(iii) Measuring the magnetic field and its anomalies; data collection
The magnetic field can be measured with sensors based on different principles; the oldest is the
fluxgate sensor. Modern instruments for ground operation use a sensor based on the proton
precession meter or the so-called (optically-pumped) caesium magnetometer. The latter is also
used in airborne operations, allowing fast sampling of the magnetic field.
The strength of a magnetic field is usually expressed in Tesla. The scalar magnitude of
the IGRF falls within the range 25,000 to 60,000 nT (nanoTesla) over almost all of the earths
surface. Anomalies due to igneous and metamorphic rocks measured at typical aircraft survey
altitudes (50 to 100 m above terrain) seldom exceed one or two thousand nT and mostly are of
the order of tens or hundreds of nT. Ground-based magnetometers record variations that are
perhaps ten times greater. Magnetometers are capable of measuring magnetic variations with a
sensitivity of at least 1 nT and often to small fractions of this quantity. Modern magnetometers
measure the absolute value of the so-called total magnetic field (i.e. core-field plus anomalous
component, added vectorially) to this accuracy in a measuring time of one second or less. So,
magnetic anomalies with amplitudes from less than 1 nT to several thousand nT can be detected
and mapped out conveniently and accurately in magnetic surveying, even though they often
represent less than 1% of the total magnetic field at any given location (Reeves, 1992).
The survey procedure is complicated by short-term variations in the geomagnetic field
(temporal variations) that have amplitudes of 50100 nT on an approximately daily cycle. They
can have much larger and unpredictable variations due to so called magnetic storms that have
their origin in events on the sun. Temporal variations can be eliminated from field observations
by recording them during the survey at a base station, or by measuring the magnetic gradient,
using two short-spaced sensors simultaneously.
The general procedure for data collection is to make observations at regular intervals
(stations) along a profile line. For mapping purposes, profile lines are surveyed at more or less
regular intervals (line spacing). For mapping purposes the line spacing must be such that no
significant anomalies will escape between adjacent lines. This is particularly important for
airborne operations, where closer lines lead to more flight kilometres and consequently to higher
survey costs. For most surveys, lines should not be separated by more than 45 times the
vertical distance from source to magnetometer. For the detection of shallow sources this leads to
a high line density in ground-based surveys. For airborne surveys a flight elevation of 100 m
(above terrain) would allow a line spacing of 400500 m, while at 50 m flight elevation a spacing
of 200250 m would be required. Increasing the flight height will reduce the survey cost, but the
consequence of that is a reduced resolution of smaller and deeper sources. Ultimately, the cost of
an airborne survey is determined by the desired resolution. The limit of the resolution is again
set by how low an aircraft can be flown with safety. Good sampling of a magnetic anomaly
requires at least 10, but preferably 20, observations over the length of the anomaly. An aircraft
flying at a height of 50 m at a velocity of 70 m/sec and fitted with sensors which can sample
10 times per second, will sample at an interval of 7 m on the ground. For good progress (i.e.
production) a ground survey would require an interval of 5 or 10 m, risking poor definition of
anomalies over shallow sources, particularly in areas with high levels of geological noise
(laterites, magnetic boulders).
When large areas need to be surveyed, an airborne operation will generally prove to be
most cost effective. For detailing of selective areas a follow-up by a ground based survey can
then be carried out, which reduces also the operational problems associated with less accessible
terrain.

135
Groundwater studies
6.2.2 Data processing and interpretation
When the local IGRF value is subtracted from the local magnetometer measurement to leave the
local anomaly, the value of the anomaly is a very close approximation of the component of the
anomalous field in the direction of the inducing field. Since the distribution of magnetic properties in
rocks can be quite complex, magnetic anomalies can be both positive and negative. Even a single
very simple geometrical source (e.g. a buried magnetic cube) will produce a magnetic anomaly
that has both positive and negative parts. Most of the negative part flanks the side of the body
nearest to the magnetic pole of the hemisphere, i.e. on the north side for the northern hemi-
sphere and on the south side for the southern hemisphere.
The magnetic anomalies over simple buried geometrical shapes can be predicted from the
physical and geometrical parameters of the sources by forward modelling calculations. Taken one
stage further, the model may be adjusted in shape and depth to produce an anomaly which
closely fits an observed anomaly. This process is called indirect interpretation and can be further
refined into a process of inversion whereby the physical and geometrical parameters of the body
are adjusted to generate a response which closely fits the observed magnetic anomaly. This is
particularly useful where an estimate of, for instance, the depth of a sedimentary layer or the
thickness of weathering needs to be estimated.
(iii) Estimating depths of magnetic sources
By forward modelling it can easily be demonstrated that, as the depth of a given body increases,
the shape of its anomaly changes. The most obvious change is the reduction in amplitude of the
anomaly. Depending on the extent of the source body in depth and strike-length, the amplitude
is found to decay by a power usually between one and three of the distance between source and
magnetometer. Doubling the distance between source and sensor will cause the anomaly to
reduce in amplitude by somewhere between one-half and one-eighth. However, the amplitude is
directly proportional to the magnetisation of the rock which can have a range that varies over
several orders of magnitude. It follows that the amplitude of an anomaly is of little use in
estimating the depth of its source, since the magnetisation of the source body is usually
unknown and even reasonable estimates could be in error by two or three orders of magnitude.
More helpful in estimating the depth of the source of an anomaly is the fact that the
wavelength of the anomaly increases approximately linearly with increasing depth of source (at
least for compact sources that are much smaller than their depth), regardless of the mag-
netisation. In the days before computer-based interpretation became routine Several graphical
methods to estimate the wavelength were used. Well defined anomalies may be interpreted by
inversion to produce quite reliable estimates of the depth and geometry of the source body,
given some reasonable geological constraints such as the likely shape of the source (dyke, fault,
horizontal plate, down-thrown block, etc.), as shown in Figure 6.1.
(iv) Maps and images
After removal of temporal variations the survey results are presented as contour maps (Fig-
ure 6.2) or images (Figures 6.3).
Most land areas of the world have by now been covered by aeromagnetic survey (though
if carried out many years ago these surveys are sometimes of inferior quality) and the ground-
water explorer should be aware of the potential value of surveys that already exist. Where
aeromagnetic data are available in digital format, display and enhancement using methods of
image processing has led to many methods of data presentation that can be tailored to the needs
of the user by seeking the clearest possible expression of the geological features of interest.

136
Figure 6.1 Quantitative interpretation of the magnetic anomaly over a normal fault in a basalt sheet
at position D in Figure 6.4. The basalt is 300 m thick north of the fault and only 100 m thick
to the south (Courtesy Dept. of Water Affairs, Government of Botswana)
Perhaps the most important development resulting from this has been as simple as making aero-
magnetic images attractive and more intuitively interpretable for those who are not necessarily
specialists in geophysics.
(v) Interpretation
Interpretation of aeromagnetic maps is usually both qualitative and quantitative. Probably the
most widespread purpose is to correlate the mapped magnetic expression with known geo-
logical mapping and to interpolate and extrapolate the magnetic response to areas with
obscured, unknown, geology. A full discussion of this process is beyond the scope of this
volume. However, some attention must be drawn to some of the deductive processes that are
particularly valuable in groundwater exploration.
Perhaps the most obvious application would be in typically hard rock areas. In such areas
the search is concentrated on localised areas of enhanced or secondary porosity that may be of
value. These could be areas with thick weathered bedrock or with buried channels. In view of
the uncertainty (approx. 20% of the source to sensor distance) in estimating the depth of source
rocks from magnetic data, the success of an airborne survey in this application would probably
be limited. This is particularly so in areas where intrusive granites appear which have such a
featureless magnetic signature that they can easily be confused with areas of sediment. Still,
magnetic ground surveys are useful for locating weathered bedrock and buried channels.
A second, more promising, example is in the tracing of faults and fractures through such
hard rock areas. Distribution of magnetic properties in rocks gives rise to a complex magnetic
anomaly pattern or signature which may be associated with a particular assemblage of rocks.

137
Groundwater studies
Figure 6.2 Contour map of magnetic anomalies for a survey flown at 100 m ground clearance,
east-west lines at 400 m spacing and a contour interval of 2 nT. Relatively magnetic
metasediments occupy the centre of the map, with a granite body to the east
and sediments to the west (Courtesy of the Australian Geological Survey Organisation)
This pattern may differ from the pattern over an adjacent rock type of different lithology, in
amplitude, in number of anomalies, in shape of the anomalies. etc. Criteria of this sort allow
many hidden boundaries to be mapped out between often only limited areas of exposure
suitable for conventional field mapping. Similarly, faults or joints can be revealed by the offset of
the magnetic pattern, as in Figure 6.3. More subtle expressions of such faults or joints can often
be recognised. In some cases, dykes are clearly seen to produce distinctive magnetic anomalies
when magnetite-rich mafic magma has been injected along fault planes.
The aeromagnetic interpretation maps resulting from interpretations of features of all
these sorts are often criss-crossed by interpreted dykes, faults and joints. Though not very inter-
esting for the geological mapping of the area, they may be crucial in the siting of successful
groundwater boreholes. A ground-breaking study by Astier and Paterson (1987) showed that, for
an area in Burkina Faso, West Africa, the yields of water boreholes in hard rock terrain decreased
markedly with distance from interpreted aeromagnetic fractures. Boreholes sited within 300 m of
a lineament were far more likely to give substantial yields than those sited further away. The
same study showed that, in the test area, far more lineaments were evident from aeromagnetic
data than from satellite and air photo interpretation. It can be concluded that in many areas the

138
Figure 6.3 (a) Shaded relief aeromagnetic image of a granite in the Yilgarn Block of Western Australia
with illumination from the east. (b) Interpretation of the fracture pattern revealed
from the magnetic data and the outline of the granite. Proterozoic dykes striking east-west cut
across the area near the north and the south margins of the figure (Courtesy of the Australian
Geological Survey Organisation)
139
Groundwater studies
preferred locations for detailed ground follow-up studies for siting water boreholes would be
near the linear features interpreted from aeromagnetic maps.
(iv) Digital elevation models from airborne surveys
A further potentially useful development is the use of GPS heights for the aircraft, coupled with
the radar-altimeter distance between aircraft and ground surface, to give accurate ground
heights along each flightline of a survey. This information can be converted into a digital
elevation model of the survey area, with terrain blocks (pixels) of dimension about 50 m and an
accuracy of better than 2 metres. Such information, particularly in areas that are poorly sur-
veyed, can be most useful in groundwater studies, not only in providing a digital terrain model
that may be important for modelling aquifer recharge, but also in determining reasonably
accurate heights for the collar of boreholes.
6.2.3 A case-study of aeromagnetic surveying for aquifer exploration
As an example, a case study from Botswana is presented in outline. Most of eastern Botswana is
underlain by Precambrian rocks which are relatively unpromising for large-scale water supply.
However, Palaeozoic/Mesozoic rocks of the Karoo Supergroup overlie the Precambrian rocks in
some areas and contain a massive porous sandstone which has proved to be a very productive
aquifer, particularly where it is capped by the relatively impermeable basalt and which forms
the topmost (Stormberg) stage of the Karoo Supergroup. The extensive Kalahari sand-cover
makes mapping the geology difficult, so outlining the areas of sub-outcropping Karoo rocks
requires innovative approaches.
When data from a low-level magnetic survey are presented as an enhanced image (Fig-
ure 6.4) the ripple signature of areas underlain by basalt are distinctive and the faulted margins
of areas of different sheet thickness are clearly displayed. Areas devoid of basalt show a
comparatively smooth glossy texture.
Extremely strong anomalies from highly magnetic rocks in the Precambrian sequence
below the Karoo tend to swamp the rather more subtle basalt anomaly patterns in the east of the
study area. Modelling of specific step-fault anomalies, as illustrated in Figure 6.1, enables the
thickness of the basalt to be estimated and the areas where the sandstone aquifer is confined by
an appropriate thickness of basalt can be mapped out in an interpretation map such as Fig-
ure 6.5. With support from suitable ground geophysical studies and indications from simul-
taneous interpretation of satellite imagery, plans can be made for drilling and testing the
potential for the sandstone aquifer where it is confined (unshaded, Figure 6.5). Less promising
areas, where the sandstone is unconfined or where Karoo rocks are totally absent can be
avoided.
6.3 Electromagnetic (EM) methods
EM techniques are used to survey the electrical properties of the subsurface materials. One of
the key feature of the EM techniques is that most applications do not require a galvanic contact,
i.e. there is no need for electrodes making a physical contact with the ground. This allows work
to be done over highly resistive top surface material (such as lateritic cuirasse, duricrust, caliche
or permafrost), to survey from airborne platforms (helicopter or fixed wing aircraft) and nor-
mally it permits a rapid rate of survey progress because the time required to install/remove elec-
trodes is eliminated. Usually EM and Resistivity techniques (see Section 6.4) are complementary.
In specific cases EM techniques may be more appropriate than resistivity and vice versa. There is

140
Figure 6.4 Magnetic anomalies of the Palla Road area, eastern Botswana, processed to second vertical
derivative and presented as a shaded-relief grey-scale image (white high and black low) with
illumination from the south. Note the positive anomalies over the north edges of the basalt sheet
(e.g. positions A and B) and negative anomalies over the south edges (position C) and where
faulting thins the sheet to the south (as at D). The ripple signature of the basalt (as at E) is in
contrast to the areas such as at F, where basalt is absent (Courtesy of the Dept. of Water Affairs,
Government of Botswana)
a wide variety of practical implementations for the EM techniques, but only a few commonly
applied techniques will be discussed in this Section.
6.3.1 Principles and survey techniques
EM surveying is based on the principle that an alternating current running through a loop wire
generates an electromagnetic field with the same frequency, and vice versa. The fundamentals
are properly described by Maxwells Equation. The Slingram or Horizontal Loops (HLEM) EM
system illustrates the concept (Figure 6.6). On one side a current-carrying loop with a vertical
axis produces, when energised, a magnetic field time varying over time. This loop is called the
EM transmitter (Tx), the magnetic field is called the primary field. If within the space around the

141
Groundwater studies
Figure 6.5 Interpretation of Figure 6.4, showing areas devoid of basalt (shaded), fault lines and basalt
thickness (Courtesy of the Dept. of Water Affairs, Government of Botswana)
transmitter a body with good electrical conductivity is present, the primary field will induce an
electro-magnetic force (emf) in that body and an alternating current will flow. This induced current
will in turn create an electromagnetic field in and around this body. This field is called the
secondary field. Primary and secondary fields and the induced current are all of the same
frequency. At the surface and at a given distance from the EM transmitter, another loop of wire
with a vertical axis is located. This loop detects the vector sum of the primary and secondary em
fields and is called the receiver coil (Rx). If we move the Tx-Rx assembly as a unit over the earth,
i.e. maintaining a constant Tx-Rx separation and orientation, the measured resultant field will
follow the variation in conductivity of the subsurface. In effect, the strength of the secondary
field depends on the electrical (i.e. conductivity) properties of the subsurface. Therefore we can

142
Figure 6.6 Principle of the Slingram EM technique. An electromagnetic field is generated

by a transmitter coil. Through induction this primary field generates a secondary
electromagnetic field, which is detected by a receiver coil (From McNeill, 1990)
detect and localise conductive features without any electrical contact with the ground. Because
the receiver observes the total strength of primary and secondary fields, the transmitter is fitted
with a reference coil which detects the strength of the primary field. To establish the secondary
field, the primary field is subtracted from the total field. In practice the receiver is designed in
such a way that it measures not only the magnitude of the resultant field but also its delay
relative to the field produced by the transmitter. This delay is called the phase shift. To analyse
this delay, the receiver measures two phase components of the resultant field: one in phase
with the primary field and the second the out of phase or quadrature which is delayed by
of a full (sines) cycle relative to the primary field. This information will be useful at the inter-
pretation stage in order to determine a conductors position and depth and its rating as a
conductive body. The principles of EM are well described by McNeill (1990).
(i) Electromagnetic properties of rocks
The significant rock properties for EM surveying are the electric conductivity and the magnetic
permeability . The magnetic permeability is often very nearly equal to that of vacuum ( 0).
Even when a formation contains a large concentration of a magnetic mineral such as magnetite,
the magnetic permeability is usually less than 10 times 0. The rocks electric conductivity is
mostly a function of the rock mineralogy, its porosity and its pore waters conductivity. The most
common conducting minerals are the clay minerals, the metallic oxides and sulphides and
graphite. The relationship between the rock conductivity, or rather resistivity, and its pore
contents is described in Section 6.4.
The dependence of the rock properties and are subject to the condition that >> 2f

143
Groundwater studies
(in which is the electrical permittivity or dielectric constant of the medium and f the
frequency). This is usually the case when f is below 100kHz. At higher frequencies, so-called
displacement currents are generated. Most EM systems operate at frequencies below 100 kHz
and the condition is met.
The depth capability (or investigation depth) of EM systems is a function of the geom-
etry of the EM transmitter-receiver pair (coil separation) and of the resistivity of the medium in
which the magnetic field propagates. Amplitude (and energy) of the EM waves are attenuated in
a conductive medium: the lower the resistivity and/or the higher the frequency, the higher the
attenuation and the shallower the penetration of this medium. This effect is called the skin depth
limitation. The skin depth is defined as the depth at which the amplitude of the EM field is
reduced (attenuated) to ~1/3 (1/e exactly, with e = 2.7....) of the value it would have if the
material were a perfect insulator. At the relative low frequencies normally used with EM
systems, its numerical value is equal to:
= 503
.f
where is the skin depth (in metres), is the electrical conductivity of the medium (in
Siemens/m [= inverse of resistivity]) and f is the frequency (in Hertz).
(ii) EM induction: frequency and time domain
The previous paragraph described the generic operation of an EM instrument. Variations relate
to the size of and distance between the transmitter and receiver coils used.
The now almost obsolete TURAM technique uses a large (several hundreds of metres in a
square) transmitter coil with a small portable roving receiver measuring strength and dip angle
of the secondary field. The selected frequency will control the depth investigation.
The so-called VLF system uses existing EM transmitters located at a very large distance.
VLF refers to the very low frequency (15 kHz to ~ 30 kHz) of submarine radio communication
stations, which are distributed all over the world. The field strength of these stations is large
enough to generate a detectable secondary field thousands of kilometres away. Modern VLF
instruments can also use the higher frequencies of long wave (LW) frequency civil broadcasting
stations. The VLF system is a simple and fast technique but with a relatively shallow depth of
penetration. Many VLF receivers measure selected geometrical characteristics of the resultant
field (dip angle, ellipticity etc.) rather than its amplitude, which is a function of various external,
uncontrollable, factors including the wave propagation conditions.
The SLINGRAM systems use relative small, equally sized and portable transmitter and
receiver coils. Frequencies are selectable in the range of 200 Hz to ~ 56 kHz, allowing for a
variety of depth investigations. They will display either in phase and quadrature response or,
for particular configurations, directly the electric apparent conductivity. Some instruments
display both. A setting where the target (aquifer) has distinctive electric properties and is
located inside a complex stratification of alternating resistive and conductive layers, can be
resolved with Slingram systems, provided a range of frequencies and coil separations can be
used. This may prove to be a rather time consuming task.
The so-called Time Domain system or Transient EM (TD or TEM) has been specifically
designed to cope with the more complex situations and to replace the Slingram system. The
principle is based on the transmission of an impulse or signal with a step function (Figure 6.7).
When the transmission is switched off, the induced emf will decline in time. With this, the sec-
ondary field will also decay in time. The time span in which these effects take place is very
short, usually a period of microseconds. The rate of decay and the duration depend on the
electrical properties of the earth. The decay is recorded by measuring the response in time slices

144
Figure 6.7 Principle of time domain EM. During the off-time of the transmitter, the decay of the induced
secondary field is recorded in time slices or time windows (From McNeill, 1990)
or time windows. The effects of shallow layers are represented in the early slices, while the
deeper layers are presented by the late slices. The geometry of the Tx-Rx systems is often dif-
ferent from FD systems (such as VLF and SLINGRAM systems), although there is considerable
overlap between FD and TD systems geometry. Time domain systems are often implemented
with medium to large transmitter coils (a square of tens to hundreds of metres per side) and
little or no separation between Tx and Rx. The controlled variables are the loop size and some
characteristic of the basic timing such as the loop current pulse width. Usually the loop size is
optimised according to the expected local geoelectric section. Time domain systems often use a
moving square loop laid out on the ground. The secondary magnetic field is detected by a loop
or a magnetic sensor or with an assemblage of three orthogonal sensors.
(iii) EM survey techniques
Both TURAM and SLINGRAM techniques are used in a profiling or mapping fashion, depend-
ing on the objective of the survey. The TURAM technique is presently not very favourable, as it

145
Groundwater studies
needs the layout of a very large transmitter loop, which limits speed of operation and flexibility.
Profiling is carried out by making observations at more or less regular intervals (stations).
Mapping consists of series of more or less parallel profile lines, preferably equally spaced. This
allows a similar processing and presentation as in magnetic surveying. Profiling and mapping
will result in an overview of the horizontal variation of the EM response (and consequently
the earths conductivity) in a surveyed area. These techniques are therefore comparable with
resistivity profiling, as discussed in Section 6.4.
The SLINGRAM technique can also be applied in a depth sounding technique (fre-
quency sounding), by making in one position a series of observations with different frequencies
and coil separations. Processing of this data will result in a geoelectric cross section of the sur-
veyed location in terms of layer thickness and layer conductivity.
The TD systems can both be used in a profiling and a depth sounding mode. For a
while, clear distinction was made between profiling and sounding work. Currently many
surveys are done in a hybrid mode: the survey is run in a profile mode but the quantity of
information recorded at each station is somewhat equivalent to a sounding. The depth sounding
mode requires observations at a range of selectable frequencies. Modern instruments will collect
the data automatically according to a programmed sequence. EM survey techniques both FD
and TD can also be performed in an airborne mode. To this end transmitters and receivers are
mounted either on a fixed wing aircraft or on a helicopter. As with other airborne geophysical
techniques, large areas can be covered in a relatively short time. Airborne EM surveys must be
conducted at a low ground clearance in order to maintain response levels. Also the line spacing
must be minimised. As a consequence, this type of surveying is more expensive than airborne
magnetometry. The use of a helicopter is particularly expensive.
Simple interpretation tools are available in the form of interpretation diagrams. They are the
result of the compilation of analogue scale modelling data. Responses of scale models have been
collected for popular EM systems, such as the Slingram. Discrete sheet-like conductors have
been modelled in particular. In such cases, the effects of variation of azimuth, position, depth,
dip and conductance (sheet thickness times sheet conductivity) of a hypothetical target have
been studied. The implementation of analogue scale models has been reviewed by Frischknecht
(1971).
Simple interpretation schemes for a TD EM system include the determination of the
apparent resistivity from the decay curves characteristics, and of the conductance of a discrete
conductor using a simplified conductors shape and dimension together with the response time
constant. For a data set from a tri-axial receiver, the dip of the secondary field may be calculated
for several stations along a profile and the migration of an equivalent induced current axis is
mapped as a function of delay time after the termination of the primary pulse; this gives an
indication of a conductors location, depth and attitude or dip.
Numerical modelling of common FD and TD systems is widely available for the 1D case
and to a lesser extent for 2D and 3D cases. A significant contribution in this domain was made
by Anderson (1979). His routine is used in a number of application programs. Typical appli-
cation programs include a forward modelling mode, which calculates the response of a given
model for a specific EM system, and a data inversion mode, where the forward modelling
component is included in a least squares inversion routine to find the best fitting model to a set
of field data. This combination is particularly useful for the data interpretation stage, as well as
for the survey planning and execution stage. Various programs are available for data entry,
editing, reduction and display. These programs includes text editors, general purpose numerical
spreadsheet software packages which include a graphic display capability, programs specially

146
designed for geoscientific mapping (e.g. GEOSOFT) and EM numerical modelling packages .
Low cost numerical EM modelling program are available from the USGS, Denver, CO, United
States of America (Anderson, 1979).
(i) Case studies
Figure 6.8 shows an example of ground surveyed frequency domain EM as applied to a ground-
water survey in Botswana (Palla Road project). The geology can be roughly described by a
sequence of a few tens of metres of Kalahari Sands covering 100 to 200 m of basalt, under which
silty and sandy sediments occur. The survey was carried out along a profile, taking multiple
frequency readings at every station. Data processing and interpretation show that a thin
conductive layer occurs above the basalt, probably representing saline or brackish groundwater.
In several places the data cannot be interpreted as a layered earth model due to the presence of
faults or fault zones. Several zones are indicated, with their conductivity thickness, depth and
dip. These zones are potential locations for borehole drilling.
Figure 6.8 Example of a frequency domain EM profile over dyke intrusions and faults in basalts (Palla Road
area, Botswana), made with a MaxMin horizontal loop instrument. The dyke intrusions generate less
obvious anomalies than the faults, which can be recognised by a typical sequence of relative high
response (shoulder) flanking a low response (trough). The geometry and magnitude of the
response allows interpretation of the more conductive faults in terms of conductivity thickness and
establishment of dip and dip direction. Note that the resistive dykes are much better defined in the
resistivity profile of Figure 6.4.3 (Courtesy of the Dept. of Water Affairs, Government of Botswana)
147
Groundwater studies
Helicopter-borne EM systems are usually multi-configuration (both coaxial and coplanar coil
geometry) and multi-frequency implementations on a bird (rigid frame) towed 15 to 20 metres
below the helicopter. Sengpiel (1986) published an example of such a survey with an operating
frequency of around 900 Hz, using a coaxial and coplanar coil configuration over the sandy
island Spiekeroog, off the German North Sea coast. The data was used to calculate the thickness
of a fresh water lens. The survey was flown at a 100 metres line spacing and totalled 40 line-km.
The interpretation, supported by vertical electric sounding, showed the largest thickness of the
lens to be some 56 m. Similar surveys were carried out in Pakistan.
Several airborne VLF systems have been used on a commercial basis. From a ground-
water exploration point of view the systems from which the dip of the resultant field could be
determined are particularly interesting. Such a system was used in the Serowe project in
Northern Botswana (Bromley et al., 1994). A regional sandstone aquifer is covered with a basaltic
layer also providing an excellent magnetic marker horizon in a region of horst and graben
covered by the Kalahari beds (sand, silcretes and calcretes). Weak VLF elongated conductors
were mapped as a result of 7,500 line km multiparameter airborne survey. These VLF conductors
(Figure 6.9) correspond to fractured zones which give the highest yields in the investigated area.
Figure 6.9 Example of the interpretation of airborne VLF data from a survey over Serowe, Botswana.
Interpretation shows position of major fractures, which have potentially the highest yield
of groundwater (From Bromley et al., 1994)

148
The interpretation strategy also involved aeromagnetic interpretation in a very effective

structural and fracture mapping technique.
Figure 6.10 shows the results of a high density sampling of conductivity, using a Geonics
EM 31 conductivity meter. This instrument presents direct reading of the apparent conductivity,
using a single frequency of 9.6 kHz and a coil separation of 3.6 metres. The depth investigation
is about 6.5 m. The survey was conducted over a dolomitic terrain south-east of Moura,
Portugal, with shallow karst dissolution features and is an adaptation of airborne techniques
brought down to the ground. Sampling was done at a 1m interval and with a line spacing of
10 m. Conductivity readings would vary from 8 mS/m to 60 mS/m over a distance of 20 m or
less. The high conductivities probably correlate with karst zones which represent areas with
enhanced rainwater infiltration.
Figure 6.10 Results of a single frequency horizontal loop EM survey (Geonics EM31) in the Moura area,
Portugal. Areas with high conductivity are associated with shallow karst features and faults,
which represent areas with enhanced infiltration of precipitation
(ii) Recent developments
The availability of convenient digital recorders and key electronic components and the improve-
ments in instrumentation performance have led to EM systems with progressively higher
resolution. From a spatial point of view readings may now be made relatively efficiently at a
high sampling density. This is especially valuable for detailed shallow investigations. Another
development is the extension of the equivalent frequency range and recording of a large number
of time windows (20 or more) in TEM. Work is presently being done to bridge the gap between
traditional EM frequencies and radar frequencies: New instrumentation and interpretation tools

149
Groundwater studies
are being developed along these lines (Pellerin et al., 1994). The interpretation tools will no
longer be limited to the quasi-static approximation when the condition >> 2 f is met.
6.4 Resistivity method
6.4.1 Resistivity of rocks
In resistivity surveying, the electrical properties of rock materials are established by measuring
the resistance to a galvanic current flow (direct current, DC). Current transport in rocks can take
place by means of free electrons, as in pure metals, by non-compensated electrons within
imperfect mineral crystals or by free ions within mineral crystals, which could be regarded as
solid electrolytes.
The resistivity of earth materials is a material property. Usually the Greek letter (rho)
is used for this property. The dimension of is Ohmmeter or Ohmm (m). The inverse
dimension for is conductivity (Siemens/m: = 1/).
Rocks usually have a high resistivity because they consist of an agglomeration of
minerals of which the bulk is resistant. High resistant minerals are quartz, feldspars, calcites and
dolomites. Many minerals are conductive, such as clay minerals, magnetite, hematite, graphite,
pyrite, pyrrhotite and many sulphidic minerals. However, these minerals occur only in a
relatively low concentration in rocks and do not make these rocks conductive. Table 6.2 lists a
few rock types with their resistivities.
Table 6.2 Resistivities of some rock types and fluids
Fluids Resistivity (ohmm) Rocks Resistivity (ohmm)
Oil, gas, air Sandstones 10010,000

Rain water 301,000 Shales 101,000
Soil moisture 1100 Limestones 50 10 5
Sea water ~ 0.2 Dolomites 10010 5
Clays 5100 Metamorphic rocks 10010 6
(River) sands 5010,000 Gneisses 1,00010,000
Lavas 30010,000
The electrical resistivity of rock is controlled by two components: the solid particles
(with the binding cement) and the pore fillings. If the pores are filled with water the rock
generally shows a relative low resistivity. The relation between the resistivity of the bulk of the
rock f and other rock parameters, also known as Archies Law (1942) is:
m
f = a . w . . s 2
in which:
a = texture factor, ranging from 0.5 to 1.5;
w = resistivity of the pore water;
= porosity as a volume fraction;
m = cementation factor, ranging from 1.3 to 2.6; and
s = saturation of the pores (as a volume fraction).

150
This relation is also presented by:
f = F . w
in which F is called the electric formation factor or formation resistivity factor. These two
expressions play an important role in groundwater exploration, as they can be used to correlate
formation resistivities with the hydraulic properties of aquifers of sedimentary formations.
Archies Law is an empirical one and based on studies of sandstones. The presence of clay
minerals in the rock matrix leads to serious deviations from this relation. The resistivity of pore
water w is an important parameter, as it is directly related to the TDS (Total Dissolved Solids) in
water and therefore gives an impression of the quality of the pore water.
6.4.2. Resistivity measurements
(i) Principles and instrumentation
The resistivity of rocks is usually determined in the field using an array of four electrodes,
which are placed in a straight line (Figure 6.11a). The instrumentation consists of a power
supply (which can be an arrangement of batteries or a generator), current and potential elec-
trodes, cables and a sensitive and accurate voltmeter and current meter.
Electrodes A and B are used to run the electrical current I into the ground (current
electrodes). By running the direct current (DC) into the ground, an electric potential field is
generated. The M and N electrodes measure the difference V between the potentials at the
Figure 6.11a Principle of the resistivity method, using two current electrodes and two potential electrodes.
Strata with different resistivities affect the current distribution in the earth
151
Groundwater studies
Figure 6.11b, c, d Layout of different electrode arrays: Wenner, Schlumberger and Dipole-Dipole
positions of these electrodes (potential electrodes). The potential difference varies with the position
and the geometry of the four electrodes. The resistivity is calculated from the relation:
= K.V/I, in which K is called the Geometrical Factor.
Because the earth is not homogeneous and isotropic, the measured resistivity is generally
addressed as apparent resistivity a: the resistivity appears to belong to a homogeneous earth,
which is not the case.
The variation of resistivity for an inhomogeneous earth is caused by the distortion of
current flow lines and with that, of the electric potential field around the potential electrodes
(Figure 6.11). For a stratified earth, the distortion is systematic, with the consequence that
through processing and interpretation the stratification of the earth can be derived from a
systematic collection of resistivity data at one location.
(ii) Electrode configurations
The most common arrays are the Wenner, the Schlumberger and the Dipole-Dipole (Figure 6.11b,
c and d).
In the Wenner array, the electrodes are placed at equal distances a. In the Schlumberger
array the distance AB must be at least 5 x MN. AB can be 100 x MN or more, depending on the
available power supply. By this layout the potential electrodes measure the potential gradient. In
the Dipole-Dipole array, the distance a between A and B is equal to the distance between M and
N, while the distance between the two dipoles is a multiple of a. For special type of surveys, the
Dipole-Dipole array can be used in a variety of non-linear configurations.

152
Each of the different arrays has its advantages and disadvantages and should be used
according to the technical and economic requirements of the planned survey. Sometimes one
current electrode and/or one potential electrode are placed at an infinite distance, resulting in
so-called half-Wenner or half-Schlumberger arrays (3 electrodes) or pole-pole arrays (2 elec-
trodes).
In resistivity profiling, or trenching, apparent resistivity observations are made with a
fixed electrode array at different locations (stations) along a survey line. The interval between
stations is preferably kept constant. The profiling technique allows the variation of earth resis-
tivity along the profile line to be studied, while the depth of investigation is more or less
constant. In practice, often two, three or more observations per station are made with increasing
current electrode separation, so that, to a limited extent, also the variation of the resistivity with
depth is recorded. An example of profiling is shown in Figure 6.12. The two and three electrode
arrays can be particularly useful in mapping structural features like faults and contacts by
profiling. A detailed discussion of responses can be found in Telford et al. (1990) and in Keller
and Frischknecht (1966).
Resistivity mapping is a variation on the profiling technique. It consists of a number of
not necessarily parallel and regularly spaced profile lines, so that apparent resistivity data is
obtained over the whole of the survey area. These mapped resistivities can be contoured, thus
yielding an overview of anomalous levels of apparent resistivity.
Generally the three survey techniques can be used with any of the electrode configur-
ations and very often a survey consists of a combination of depth sounding and profiling or
mapping.
(iv) Data processing and interpretation
The processing of profiling data is usually minimal and consists of presenting the data
graphically. Figure 6.12 shows part of the results of a full Schlumberger array profile over an
East-West striking resistive dyke in Botswana. The anomaly shows two resistive peaks,
separated by a minimum. The response is caused by a single dyke with a width approximately
equal to the current electrode separation (AB/2 = 100 m). It is interesting to notice that this dyke
would be less clearly present in an EM survey with moving coils run along the same profile line.
Depth sounding data requires more preparation before being interpreted in terms of the
thickness and the resistivity of layers. A traditional way of interpretation is the curve-matching
technique, but such a process is tedious, time consuming and not very accurate for multi-layered
field curves. Fortunately, it is now common to use a microcomputer. Simple programs allow
curve matching on the screen. More sophisticated ones allow fully automatic direct inter-
pretation of field data, or controlled optimisation of layer parameters through the use of inverse
modelling techniques. Use of such computer programs also make it possible to handle large
numbers of VES data in so-called batch processing. Figure 6.13 shows an example of the output
of the RESIST program that can handle Wenner, Schlumberger and Dipole-Dipole array data
(Van der Velpen and Sporry, 1992).
The interpreted earth models can be used to construct resistivity cross sections that can be
correlated with geology and lithology to model the survey area. Cross sections can be combined
into fence diagrams, presenting a subsurface model in three dimensions. Such a model can be
used to support the aquifer modelling in terms of extent, thickness, depth and variation of
lithology (read: permeability) in the aquifer.
(v) Resistivity interpretation and aquifer evaluation
An important step in aquifer evaluation is the correlation between geophysical and hydrogeo-

153
Groundwater studies
Figure 6.12 Example of resistivity profiling over a sequence of Kalahari sands and basalts in Botswana.
The high resistivity peaks around station 500 are caused by a sub-outcropping resistive dyke
intrusion (Courtesy Dept. of Water Affairs, Government of Botswana)
logical parameters. During the 1970s and 1980s quite a few researchers worked on this problem
(Kosinsky and Kelly, 1981; Mazac et al., 1985; Sporry et al., 1991). The importance of such a
correlation (if successful) is that hydraulic transmissivity values can be predicted for every depth
sounding location, thereby dramatically improving the control of the groundwater model.
Figure 6.14a presents an example of such a correlation for the La Paz-El Carrizal Graben south of
La Paz in Baja California, Mexico. The resulting transmissivity Contour Map is presented in
Figure 6.14b.
(vi) Recent developments
Availability of low cost electronics has led to the development of equipment and procedures
based on multicore cables and multi-electrode arrays, which allow the operator to switch
through a complete sequence of electrode separations and positions without leaving his central
operating position. Computer controlled instruments can store the acquired data automatically
on magnetic diskettes, for processing later. Recently, instrumentation for continuous profiling
and/or depth sounding has been developed. In-field processing is at present not an exception. A
disadvantage is that the increased cost of these instruments vis--vis the traditional ones
cancels out the gain made in production and in the reduced number of manpower.
The data of Figure 6.13 was collected using a multicore Barker cable (Barker, 1981) in the
Offset Wenner array. In principle, this sounding can be made in the field by one single operator,
although work would proceed much more efficiently with the help of a field assistant.

154
Figure 6.13 Example of a VES in the Wenner array, representing apparent resistivity data (crosses),
the interpretation in a block diagram (dotted line) and the model curve based on the layer
parameters (table). The fit between the model curve and the field data is displayed in terms of
RMS error. This depth sounding was performed using the Wenner Offset system over a sequence
of fresh water (38.7 Ohmm), brackish water (10 Ohmm) and saline water (1.6 Ohmm).
Location: Kapelle-Biezelinge, the Netherlands
Figure 6.15 shows an example, from The Netherlands, of the great detail that is possible
with a continuous sounding system (Van Overmeeren and Ritsema, 1988).
6.5 Induction polarisation
The Induction Polarisation method or IP is an electrical method combining the physical effect
called induction polarisation, with the principles of the resistivity method. Consequently the
technique of data collection is similar, but more complex equipment is used. As with the EM

155
Groundwater studies
Figure 6.14a Correlation of hydraulic transmissivity values with normalised transverse resistance T N ,
calculated from aquifer formation resitivities for the aquifer of the La Paz-El Carrizal Graben,
Baja California, Mexico. Two different correlations (AB and CD) can be observed, valid for the
locations A, B, C and D on the transmissivity map of Figure 6.14b (From Sporry et al., 1991)
methods, this method originally was developed for mineral exploration, but eventually it was
realised that groundwater exploration could benefit from it too.
6.5.1 Principles and procedures of IP
An IP survey is in principle carried out with resistivity equipment, normally in the Dipole-
Dipole configuration, although the Gradient array (a variation on the Schlumberger array) is
also used. It therefore requires current and potential electrodes, connecting cables, a power
supply and the instrumentation to measure current and potential difference. While using the
equipment to measure the IP effect, the apparent resistivity is also measured in the conventional
way. When a current is run into the ground and switched off, a decaying potential difference
may be observed after the moment of current switch-off. This effect is called induction polarisation.
The observed potential at the potential electrodes will thus show a response as illustrated
in Figure 6.16. The decay of the voltage may last milliseconds to a few seconds, basically 1,000
times longer than in EM methods. The level of this effect and the period of decay depends on
the IP properties of rocks.
(i) Sources of the IP effect
IP is caused by two major effects, called the membrane effect and the electrode polarisation effect.
Both are electro-chemical processes and are caused by the interaction between the minerals of
the rock and the electrolytic behaviour of the pore water. When current flows through or along
the boundaries of the solids, the transport of electrical charges is hampered by reduction or

156
Figure 6.14b Hydraulic transmissivity contour map, based on the transmissivity values derived from aquifer
formation resistivities using the correlation chart of Fig. 6.14a. The N-S oriented graben is broken
up into blocks by two faults running E-W (not shown) (From Sporry et al., 1991)
157
Groundwater studies
Figure 6.15 Example of a continuous depth sounding survey by electrode array scanning, imaging
the shallow subsurface. Sand-infilled frost cracks representing hydraulic connections
between surface sands and deeper sandy formations can clearly be identified
(Van Overmeeren and Ritsema, 1988)
oxidation processes as well as by different mobilities of ions. As a result, electrical charges will
locally build up (polarisation) and diffuse, or accumulate with a certain time lag in relation to
the inducing current. A formation containing such polarisable grain surfaces will show a time-
dependent current/voltage relation, which can also be presented as a frequency-dependent
specific resistivity combined with a capacity-like phase shift.
The membrane potential is caused by an ion cloud in the pore water. This cloud moves
under the influence of an applied electric field. Some non-metallic minerals (clays) possess a
negative surface charge, which is neutralised by free positive ions in the electrolyte, thus
forming an electrically charged membrane, consisting of alternating layers of negative-
positive-negative charges. When a current is applied, the neutral membrane is disrupted and
when the current is switched off, a movement of ions due to redistributions takes place. The
ionic movement creates a weak current flow, which is detected at the potential electrodes.
The electrode polarisation occurs when mineral particles block the flow of ions, carried by
the electrolyte. The current flow is then carried by electrons within the mineral grains, which

158
Figure 6.16 Principle of measuring the chargeability in the time-domain IP technique. The induced
primary current Ip is detected as a primary voltage Vp. At current turn-off, the voltage drops
to a secondary level Vs and the transient voltage Vt decays with time. A theoretical measure
of chargeability M is: M = Vs/Vp (From Sumner, 1976)
results in a current-opposing build-up of electrical charges at the interface of particles and

electrolyte (pore water). when the current is switched off, the bound ionic charge does not
immediately disappear, but decreases in time due to diffusion back into the pore water. The
electrode polarisation is generally the strongest effect and is particularly generated by sulphidic
minerals. The membrane polarisation, however, is particularly generated by clay particles and is
therefore useful in groundwater exploration. Where the resistivity and EM methods are not
capable of differentiating between conductive clay-containing formations and groundwater
formations, IP can indicate the presence of clay.
(ii) IP techniques
The IP effect can be measured in three modes: the Time-Domain method, the Frequency-Domain
method and the Phase method. In the Time-Domain technique, the decay of the IP voltage is
measured in time slices or windows, as in Time Domain EM. The number of windows is in
the order of 4 to 12. Because of the short period in which the decay takes place (less than 1 or 2
seconds), the voltages are measured and stored in a memory by a microprocessor. The physical
property measured is called the chargeability. The chargeability M is defined by:
t2
Vc
M = 1 V(t)dt
t1
where M is the apparent chargeability, Vc is the potential value near the end of the on-time

159
Groundwater studies
interval, t1 and t2 define start and end of the time window interval and V(t) is the decay
potential (Figure 6.16).
In the Frequency Domain technique, two frequency excitation modes are used, usually
in the range between 0.1 and 10 Herz. Due to the IP response the apparent resistivity will be
lower for the highest frequency. These different resistivity values are used to calculate the so-
called frequency effect, FE, and the metal factor, MF.
In the Phase Mode, the phase shift between the excitation current and the IP potential is
measured. This phase shift is directly proportional to FE. In this approach only one excitation
frequency is needed.
(iii) Field procedures
In IP surveying the normal apparent resistivity is measured (during on-time), together with the
IP effect (during off-time). The process is repeated (in the Dipole-Dipole array) with increasing
separations of the dipoles. The dipole separations run from 1 x a to 6 x a (a is the electrode sepa-
ration within the dipole). Investigations to different depth can be performed using differently
sized dipoles.
Although the apparent resistivity part of the survey is fairly straightforward, the
observations of the IP effect are rather time-consuming, which makes the IP method a less
popular survey method.
The observed apparent resistivities are plotted in a pseudo section. Conductive zones show up
clearly in such a section. The chargeability values are also plotted in pseudo sections. A quali-
tative evaluation will reveal the presence of conductive fault or fracture zones, which may be
potential (small) groundwater sources. More importantly, IP data can be helpful in differen-
tiating between thick or thin layers containing clay, which is useful information in establishing
the transmissivity of formations. Also, the IP method can be helpful in detecting pollution by
organic compounds, such as oil and organic solvents.
Modern computer techniques have brought about the possibility to perform quantitative
interpretation through the application of the finite element method in modelling. This entails
varying the model parameters, in an attempt to get the model response to fit the field
observations.
6.6 Seismic method
6.6.1 Principles and instrumentation
The seismic method is based on measuring the propagation velocity of elastic waves induced in
rock formations by an explosion or impact. The velocity V is calculated from the travel time T of
a wave over distance X between a source and a sensor: V=X/T. The waves propagate through
the earth and are subject to reflections and refractions at interfaces between rock strata with
different physical properties. The seismic signals which return to the ground surface are regis-
tered by arrays of sensors, called geophones when used on land or hydrophones when used in
water. Cables connect the geophones to a recording device, which is the seismograph. The geo-
phones consist of a magneto-dynamic element, which converts vibrations picked up from the
earth into a small current, which is relayed to the seismograph. Modern seismographs consist of
signal amplifiers and analogue-to-digital converters. The signal is digitally recorded on a

160
magnetic storage device. The recorded signal can be displayed on a CRT or LCD screen and it
can be printed on a paper printer for quality control. Multichannel recorders allow the simul-
taneous use of 12, 24 or more geophones. Because of the digital data storage techniques,
computers can be used to process the field data efficiently.
A large variation of sources can be used. The smallest is a heavy hammer striking a steel
plate on the ground. Other possible sources are dropping weights of a variety of sizes and from
various heights, explosives and a range of mechanical or electromechanical devices generating
impacts or vibrations. Each source type generates energy with a typical frequency bandwidth,
which may be crucial to the success of the survey. Some of these sources put high demands on
the skills, training and organisation of the field crews. The use of explosives is generally subject
to licences for storage, transport and use. Sources that are cheap and easy to maintain or to
replace (like a hammer and plate) have the disadvantage of a limited level of energy output.
The propagation of elastic waves through the earth is subject to several processes. It is not
only the wave velocity that is important, but also the frequency content, attenuation (reduction)
of the seismic energy, refraction, reflection and diffraction. play an important role.
Table 6.3 Elastic wave velocities in some media
Medium Velocity (m/sec) Medium Velocity (m/sec)
Air ~ 330 Anhydrite 3,5005,500

Water ~ 1,500 Limestones 3,4007,000
Soil ~ 3,400 Sandstones 2,0004,500
Sands 50300 Shales 3,9005,500
Igneous rocks:
Chalk 2,1004,200 Granite 6,5007,000
Dolomite 3,5006,900 Gabbro 6,5007,000
Gypsum 2,0003,500 Ultra mafic rocks 7,5008,500
(i) Wave velocities
The seismic wave velocity is expressed in m/sec. Two different groups of waves can be
observed: body waves and surface waves. Most important wave types are the Compressional or
P-waves and the Shear or S-waves, which are both body waves. Raleigh waves and Love waves
are surface waves. In groundwater investigations, usually only P-waves are used. P-waves can
travel through any medium: solids, liquids and gases. S-waves occur only in solids.
The velocity of a seismic wave is related to the elastic properties and the density of the
medium through which it travels. It has been found that Vp > Vs. Table 6.3 presents velocities of
some common rock types.
Generally, it can be observed that for sedimentary rocks the velocity increases with
geological age and depth of burial. All igneous rocks show high velocities in a fresh and
unweathered state. The presence of groundwater affects the velocity of rocks strongly,
depending on the porosity.
(ii) Wave frequencies
The seismic energy usually contains a wide range of frequencies, from as low as 1 Hertz to
1,000 Hz or more. It is found that high frequencies also lead to higher resolution, i.e. thinner
individual beds can be detected (in reflection seismics) than with low frequencies. On the other

161
Groundwater studies
Figure 6.17a,b In a layered medium three different waves can be observed: the direct wave, the refracted
wave and the reflected wave (a). Each wave has a different travel time, depending on the
geometry of the layout, depth of the interface and the velocities of the individual layers (b)
hand, high frequencies suffer from a stronger attenuation than low frequencies. The result of this
is that with an increased depth of investigation, the resolution also reduces.
6.6.2 Reflection and refraction of seismic waves
As the earth is composed of a sequence of sub-horizontal or dipping layered formations, the

boundary effects at the interface of layers with different physical properties will result in the
seismic energy being separated into a reflected part and a refracted part. These effects are
controlled by the same laws as in optics. When a wave is reflected from an interface between
two media, the angle of reflection equals the angle of incidence. Refraction of a seismic wave is
controlled by Snells Law: V1/sin1 = V2/sin2. V1 and V2 are the velocities of the two media and
1 and 2 are the angle of incidence and the angle of refraction respectively. Under specific

162
circumstances (V2 >V1) 1 can be such that 2 = 90. The refracted wave will then travel along the
interface and is said to be critically refracted. Angle 1 is then called the critical angle.
(i) Travel times of seismic waves
Figure 6.17a shows that in a layered medium three waves can be observed: a direct wave, a
refracted wave and a reflected wave. The travel paths are different for each of these waves,
resulting in different arrival times at the geophone. This effect has led to the development of two
main techniques: refraction seismics and reflection seismics. Figure 6.17b shows schematically
the travel time curves of the three waves for a two-layered earth model. It is the result of
measuring the travel times for different distances between source and receiver. Note that the
direct and refracted waves show a linear response, while the reflected wave shows a hyperbolic
response. Such a presentation is called a TX diagram or time-distance diagram.
In refraction seismics the boundary between two successive layers can only be detected if
the seismic wave velocity is higher in the deeper layer than in the overlying layer. If the seismic
velocity in the deeper layer is lower than the upper layer, no critical refraction will take place
and the interface will not be detected. In addition, a relatively thin layer may not be detected by
refraction seismics.
With reflection seismics there is no problem with detecting a low velocity layer, but here
too, a thin layer may not be detected. Velocity inversion and thin layers will inevitably lead to
errors in depth interpretation of refraction seismic data.
(ii) Refraction seismic surveying and data processing
In refraction seismic surveying a seismic line is made up of a sequence of one or more seismic
spreads. Each spread consists of a series of geophone positions, usually 12 or a multiple of that
number. A basic requirement is to fire at least two shots, one at either end of the spread: the
forward shot and the reverse shot (the reciprocal method). From the records (Figure 6.18) the
first arrivals are marked, listed and plotted in a TX-diagram, which forms the very core of the
processing sequence. The first arrivals are always related to direct or refracted waves and are
used to calculate the wave velocity for each layer, the intercept times for each geophone position
and the depth of the refractors below each geophone. Relatively simple procedures, such as the
Hawkins method (Hawkins, 1961) or Hagedoorns Plus-Minus Method (Hagedoorn, 1959), to
make these calculations can be performed in an organised table, or better still, using a
spreadsheet program.
The present day approach of refraction seismic data processing is the General Reciprocal
Method (GRM) (Palmer, 1980). It uses a minimum of 5 (sometimes 7) shots for each spread. One
shot is always in the middle of the spread. The approach leads to a very consistent arrangement
of geophonees with shots at regular intervals all along the seismic line. The GRM allows a
detailed velocity analysis but is rather elaborate to process manually. Therefore the data is
processed on a computer. Apart from resulting in a continuous depth section, the procedure will
result in a detailed velocity analysis, presenting variations of the wave velocity along the
refractor. This information can indicate the degree of weathering along a refractor, giving an
indication of the presence of a potential aquifer. Changes in wave velocities may also indicate
the presence of faults or fault zones covered by the seismic line.
In all processes based on forward and reverse shooting, the depth of a refractor is
calculated from: h = . t i . (V1/cosc), in which t i is the intercept time (travel time for X = 0).
The intercept time can be calculated for every geophone position. Cos c is calculated from the
velocities V1 and V2 (in a two layer case). For more than two layers, the depth of other refractors
needs to be calculated in sequence of increasing depth.

163
Groundwater studies
Figure 6.18 Example of a seismograph recording with a 24-channel system. The positive amplitudes
have been shaded black (variable area display VAR) to enhance the presentation of wavelets
Figure 6.19 shows an example of a refraction seismic section across a narrow valley, in
which fault zones are marked by increased depth of the refractor, due to weathering along the
fault. The weathered zones are potentially water-bearing.
(iii) Reflection seismic surveying
Until the early 1980s, reflection seismics was not very popular in the relatively shallow appli-
cations that groundwater surveys represent. However, with the availability of cheap micro-
processors and PC-compatible computers, this technique has proven to be a valuable tool for
groundwater applications. Downscaling of the geometry, reduction of the amount of equipment
and development of small, but efficient seismic sources has made its operation by a relative
small crew of 5 or 6 men possible. The data can be processed on PC-compatible computers, with
relatively cheap software, thus bringing the costs to acceptable levels for groundwater explo-
ration. It must be realised that of all the various geophysical methods, reflection seismics is
potentially capable of the highest possible resolution in subsurface mapping.

164
Figure 6.19 Example of a refraction seismic profile across a valley, showing the positions of three fault zones.
The fault zones can be recognised by an increased depth to the refractor (bedrock) and locally
lower velocities. Location: Santa Eufemia, Spain
In reflection seismics, use is made of the well established procedure of multifold coverage
Common Depth Point or Common Midpoint technique. Figure 6.20 illustrates the basic concept,
which starts with the data acquisition in the field. During this process, the seismic spread and
the shotpoint move up one shotpoint interval (usually equal to one geophone interval) along
the seismic line. Shotpoint intervals can be as small as 2 m, leading to a subsurface interval of
reflection points of 1 m only. Reflections from sequential interfaces can be observed as
hyperbolic events in the unprocessed field record (Fig. 6.18). Processing of this data is based on
the correction of the Normal Move Out or NMO, by applying the NMO-velocity, also called the
stacking velocity. NMO is the difference in travel time between a wave that travels along per-
pendicular to a reflector and a wave that travels at an angle to the perpendicular, as is usually
the case when shotpoint and geophone are not in the same place.
The depth of a reflector can be calculated from the expression Z = Vnmo . t 0 /2, in which t 0
is the travel time along the perpendicular to the reflector and Vnmo the average (RMS) velocity
of the wave to that interface. Vnmo is also called the stacking velocity. The processing sequence
in reflection seismics includes trace sorting, velocity analysis, NMO correction, stacking, as well
as frequency spectrum analysis, frequency filtering, corrections for surface level variations (static
correction), deconvolution processes (to compensate for filtering effects of the earth) and

165
Groundwater studies
Figure 6.20 Principle of the CDP technique in reflection seismics. Through this survey technique t
he same depth point (reflection point) is recorded 6 times (6-fold coverage), using an array
of 12 geophones. Processing of this data will lead to an improvement of the signal-to-noise
ratio as compared to a single coverage (Adapted from Kearey and Brooks,1991)
migration (to correct spatial positioning of reflectors caused by steep dips. The reader is referred
to excellent publications on these subjects by Yilmaz (1987) and by Sheriff and Geldart (1982)
and many others. A full discussion in this chapter would take up far too much space.
Figure 6.21 shows an example of a reflection seismic section in a geohydrological
application. It demonstrates the continuity of an aquifer formation, which was suspected to
wedge out, but actually continues at a reduced thickness. This example gives an impression of
the excellent detailing possible with shallow reflection seismics though at a cost. One of the
minor drawbacks of shallow reflection seismics is that, with a few exceptions, the first 20 metres
remain out of focus.
Many examples of shallow reflection seismics have been published. Good examples are
Hunter et al. (1984) and Steeples and Miller (1990).

166
Figure 6.21 Example of a high resolution seismic reflection section after full processing of the data,
as an application in groundwater surveying. The grey zones mark an aquifer which shows
a decreased thickness west of the 300 m point along the line. The section is combined with
gamma logs, recorded in two boreholes in the section. Strong reflections correspond with the
top of clay layers, which have a high level of natural gamma radiation (Meekes et al., 1990)
6.6.3 Recent developments
One recent development in shallow reflection seismics is the modification of 3D techniques.

These, however are rather costly, requiring the use of seismographs with at least 96 channels to
record on, and an even larger number of geophones. The investment in hardware required to
use 3D seismics in groundwater exploration is prohibitive. However, the accelerated devel-
opment of hardware and software may remove the price impediment.
Another development is seismic tomography. In this application recordings are made with
the seismic source either at the surface or placed in a borehole and with the geophones in a
borehole. Processing of the data allows imaging of the subsurface. Another new development is
the study of surface layers, using shear waves .
6.7 Gravity surveying
Although seldom employed as a primary tool for groundwater exploration, the gravity method
has proved to be extremely useful, both in the delineation of significant regional structures and
in the direct detection of aquifer formations such as paleochannels, shallow karst cavities and
fault zones.
6.7.1 Principles and field procedures
In gravity surveying, the quantity that is measured is the difference in acceleration due to
gravity between one station and another at the earths surface. The acceleration g is expressed

167
Groundwater studies
in cm/sec2 or in Gals. Practical units are milliGals and the gravity unit (g.u.), which is equal to
0.1 milliGal. The acceleration is measured with a gravimeter. The operating principle is based on
the variation in length of a spring, which compensates the weight of a small mass. Modern
instruments make use of electronic compensation of the weight. The differences in acceleration
are caused by:
Earth tides: the gravitational attraction of the sun and the moon impose a time-and
position-dependent variation on the Earths gravitational field.
Instrumental drift: instruments of older design suffer from drift of the readings. Modern
instruments are self-correcting.
Latitude: the distance between the surface and the gravitational centre of the Earth is
smaller at the poles than at the equator, causing a latitude-dependent increase of
gravity from the equator to the poles. This effect is enhanced by the Earths rotation.
Corrections must be applied using the International Gravity Formula 1967 (IGF).
Elevation: topographic variations also cause differences in measured gravity.
The differences are due to: a) a decrease in gravity for stations at higher altitudes due to
the increased distance to the gravitational centre and b) an increase in gravity for
higher stations due to the extra downward attraction of the slab of rock between the
station level and the survey datum level. Both effects may be removed from the obser-
vations by applying Free Air and Bouguer corrections respectively.
Terrain surrounding the survey station: hills and valleys within the vicinity of a station
exert an upward attraction on the spring-balance mass of the gravimeter, thus reducing
the observed gravity. This effect can be significant in hilly terrain and extremely
mountainous areas. Corrections can be applied using topographical information (maps
or digital terrain models).
Lateral variations of rock density: Once corrections for the above factors have been made,
the resulting Bouguer anomalies may be interpreted in terms of variations in rock
density between the observation platform (surface) and the centre of the earth. It is this
very property from which geological units can be modelled. Table 6.4 presents density
values of some common rocks.
Table 6.4 Density values of some common rocks
Rock type Density (t/m3) Rock type Density (t/m3)
Sands and gravels 1.42.2 Granites 2.62.8

Clays 1.72.5 Lavas 2.83
Marls 1.82.6 Basalts/Gabbro ~3
Sandstones 22.6 Seyenite 2.75
Limestones 2.22.8 Diorite 2.85
Dolomites 2.22.8 Peridotite 3.3
Anhydrite 2.83 Schists 2.42.8
Gneiss 2.63
Field procedures in gravity surveying essentially consist of making regularly spaced obser-
vations along a survey line or grid, during which station elevations also are levelled. Careful
notes must be made of the elevation of the surrounding terrain and where possible, rock
samples (preferably unweathered) must be collected so that the rock densities can be measured.
Since the acceleration due to gravity is proportional to the inverse square of the distance
between two attracting masses, density variations closer to the observation platform have a pro-

168
portionally greater effect on Bouguer anomaly values. However, deeper sources should not be
neglected in the interpretation, particularly as they often contribute regional gradients to gravity
data. For a full discussion of the principles of gravity measurements and data reduction, refer-
ence is made to standard geophysical textbooks (Telford et al., 1990; Kearey and Brooks, 1991;
Milsom, 1989). Essentially, the Bouguer anomaly values, g ba, are obtained by first correcting the
measured gravity for temporal variations (tidal effects, instrument drift) to give the observed
gravity, g0, and then by correcting for all non-geological factors:
g ba = g 0 g th g fa + g b + g t mGal (m.s2 x 105)
where:
g th is the theoretical gravity value on the geoid at the given latitude of the gravity station,
g fa is the free-air correction,
g b is the Bouguer correction,
g t is the terrain correction.
The accuracy of the final Bouguer Anomalies are dependent on the accuracy with which
the corrections are calculated. Excluding instrument malfunction, the largest error would be
caused by inaccuracies in station heights. The combined free-air and Bouguer correction is, on
average about 0.2 mGal per metre, using an average crustal density of 2.67 t/m for the Bouguer
correction. Consequently, a height accuracy of 5 m, as might be achieved using barometric
levelling techniques, would lead to an average accuracy of 1 mGal in the Bouguer anomaly.
Since the amplitude of Bouguer anomalies depends on the size, depth and density con-
trast of features of interest to the survey, it is not wise to generalise on these values; anomalies
between 0.1 to tens of mGals may be important. It is important to consider carefully the
expected anomaly amplitude and wavelength. It would not be feasible, for example, to detect
sub-mGals in mountainous areas, due to the inaccuracies in terrain correction estimates.
6.7.2 Interpretation of gravity anomalies
Bouguer anomaly maps may be interpreted qualitatively in terms of structural and lithological
features. For example, a strong gravity gradient may represent the faulted juxtaposition of geo-
logical units with differing densities, or the edge of an intrusive body.
A gravity low may be caused by a siliceous intrusion or a sedimentary basin, since both
granitic rocks and sediments are generally less dense than the remaining commonly occurring
rock types. As in all methods based on potential field measurements, the non-uniqueness of
possible solutions require a priori geological and physical constraints in order to arrive at
acceptable model solutions. Also, the cumulative effect of lateral density contrasts at different
depths will combine to produce a single observed anomaly. In the case of Figure 6.22 the upper
sediment-bedrock interface was resolved by a seismic survey, the lower one was based on bore-
hole information.
The recognition of regional gravity gradients plays an important role in the quantitative
interpretation The regional gradient is generally due to a deeper or broader geological feature
and removal or simultaneous modelling is essential for an accurate modelling of the target body
of interest.
Systematic density measurements of all rock types in a survey area is often not a practical
or economic solution. Usually a combination of sample measured values and standard ranges
for given rock types is used, even realising that density variation may occur within one geo-
logical unit. Complex models to fit the observed gravity anomaly can only be justified if
sufficient a priori information is available.

169
Groundwater studies
Figure 6.22 Example of the combined effect on a gravity anomaly by a shallow and a deep structure.
The circles represent the observed residual gravity anomaly over a buried graben structure.
The upper graph represents the calculated anomaly caused only by the displacement
of the thin slab of high density material, whereas the lower graph shows the computed total
gravity effect from both the shallow and the deep structure; the latter anomaly accords well
with the observed data. The contrast in specific density was assumed to be 0.6 g/cm3
for both the upper and lower contrast (Overmeeren, 1980)
6.7.3 Applications to groundwater exploration
A common application of gravity surveys to groundwater exploration is in the delineation and

modelling of regional features such as sedimentary basins and horst-graben structures (Ghazala,
1994; Bourgeois et al., 1994; Alvarez, 1991). Correlation of the Bouguer anomalies with, for
example, magnetic, magnetotelluric, VES and borehole data may confirm or refute gravity inter-
pretations as demonstrated in Figure 6.23. In this example the regional gradient was insuf-
ficiently defined (too short profile) to calculate sedimentary thickness. Both gravity and
magnetic anomalies could be interpreted as basement highs (<100 m) in the extreme west of this
profile. AMT (audiomagnetotelluric) and borehole data refute this conclusion. The magnetic and
gravity highs are interpreted as probably due to the presence of an intrasedimentary dolerite sill
at this location. Although the gravity and magnetic data in this case was not used for quan-
titative modelling, the fact that the Sethunya horst was not derived from the initial modelling of
the AMT and VES results is an example of the value of a multi-method approach. Additional

170
Figure 6.23 Gravimetry used for a structural study of a sedimentary basin, supported by AMT
(audiomagnetotelluric), magnetic and resistivity (VES) surveying. (Bourgeois et al., 1994).
Gravity and magnetic anomalies in the western part of the profiles could be interpreted
as basement highs. AMT and borehole data refute this conclusion (Bourgeois et al., 1994)
constraints are also required to avoid potential confusion between sediment-basement

topography and intra-basement density contrasts (Van Overmeeren, 1981).
A gravity low may be caused by a siliceous intrusion or a sedimentary basin, since both
granitic rocks and sediments are generally less dense than the remaining commonly occurring
rock types. As in all methods based on potential field measurements, the non-uniqueness of
possible solutions require a priori geological and physical constraints in order to arrive at

171
Groundwater studies
acceptable model solutions. Also, the cumulative effect of lateral density contrasts at different
depths will combine to produce a single observed anomaly. In the case of Figure 6.22 the upper
sediment-bedrock interface was resolved by a seismic survey, the lower one was based on
borehole information.
The recognition of regional gravity gradients plays an important role in the quantitative
interpretation The regional gradient is generally due to a deeper or broader geological feature
and removal or simultaneous modelling is essential for an accurate modelling of the target body
of interest.
Systematic density measurements of all rock types in a survey area is often not a practical
or economic solution. Usually a combination of sample measured values and standard ranges
for given rock types is used, even realising that density variation may occur within one geo-
logical unit. Complex models to fit the observed gravity anomaly can only be justified if
sufficient a priori information is available.
Figure 6.24 demonstrates how small (sub-mGal) anomalies due to faulted offset of the
basement-sediment interface can be detected by careful surveying and data reduction. To
achieve the required accuracy it necessitated the levelling of gravity stations to a precision of
0.1 m and careful consideration of terrain corrections. The resulting Bouguer anomaly profile
showed a noise envelope of 1 mGal. A smoothing filter was applied to remove short-wavelength
noise and to highlight the geologically significant anomalies. Correlation with faults derived
from aerial photo interpretation and with VES results added confidence to the geological model.
Note here that the regional gradient is modelled as the broad N-S slope of the density interface;
outcrop evidence is in agreement with this conclusion.
High sensitivity semi-regional gravity surveys have been successfully applied to the direct
detection of paleo-channel aquifers (Carmichael and Henry, 1977; Angelito et al., 1991) giving
typical anomalies of 0.2 to 2.0 mGal. Secondary porosity in karst terrain can be sometimes
detected by microgravity surveys (Butler, 1984), but there is a depth limit, and the expected
results should be careful considered before embarking on such a high-resolution survey. Gravity
data alone can not distinguish between water-filled and air-filled cavities, but under favourable
circumstances a combination with resistivity data may resolve this ambiguity.
6.8 Ground penetrating radar
The ground penetrating radar method (GPR) has been in use for a number of years, mainly for
engineering geology applications and for shallow investigations (less than 10 m). Recent devel-
opments, again stimulated by advances in electronic and computing technology, have made it an
interesting and extremely useful tool for detailed groundwater investigations.
GPR uses electromagnetic waves of frequencies between 25 to 200 MHz, the domain of
radar waves. When radar waves transmitted by a radar antenna penetrate the earth, reflection
and transmission processes take place, similar to those of reflection seismics. Reflection takes
place at the interface between layers with contrasting electric properties. Reflected waves are
received by a receiver antenna which registers travel times (in nanoseconds) and amplitudes in
the receiver system. The transmitter and receiver antennas can be used in a fixed mode, or in a
mode in which one of the antennas is moved away from the other at regular intervals.
Figure 6.25 demonstrates the principle of the method.
(i) The di-electric constant
The degree of reflection is determined by the contrast of the dielectric constants of the materials
on either side of the interface. The dielectric constant varies from a value of 4 for dry sands to 65

172
Figure 6.24 Careful surveying allows the detection of faulted offset of the basement-sediment interface.
Intersecting faults show up as small gravity anomalies. Location: Moura, Portugal
Figure 6.25 The sounding mode for GPR. The principle is similar to the one used for reflection seismics
(CMP or CDP technique) and allows for the calculation of wave velocity in each layer
and the calculation of depth to each reflecting interface
for peat (Van Overmeeren, 1994). Fresh and saline water have a dielectric constant of 80. The
high contrast between dry and water-saturated materials has the consequence that a water table
generally acts as a good reflector for radar waves. Crystalline and metamorphic rocks usually
show relatively low values of 5 to 15. Only gabbro may have a value as high as 40 (Telford et al.,
1990).

173
Groundwater studies
(ii) Depth penetration, damping
The depth penetration of radar waves depends on the resistivity of the medium and on the
frequency. As a consequence, the penetration depth of radar waves is rather limited in a low
resistive environment. Van Overmeeren (1994) reports that in a sandy formation the penetration
depth was reduced from some 40 m with 25 MHz to around 10 m with 200 MHz. The atten-
uation coefficient a of radar wave amplitudes is expressed in dB/m.
(iii) Radar wave velocity
Although difficult to measure accurately, it has been found that radar waves show considerable
differences in propagation velocities for different earth materials. To present these high velocities
in reasonable figures, they are presented in centimetres per nanosecond (cm/ns). In free air this
velocity is 30 cm/ns, in earth materials it can range from 15 to 3 cm/ns. The variation in velocity
makes it possible to apply seismic reflection techniques to obtain a radar section. Applying
the Common Mid Point technique, but in a single coverage mode, results in the registration
of hyperbolic reflectors. Using processing techniques similar to those used in seismics, wave
velocities can be derived and NMO corrections can be applied to obtain the section. Once the
wave velocities are known time sections can be converted into actual depth sections.
Figure 6.26 presents an example of a radar section obtained over a sandy formation (push
moraine) in the Netherlands (Van Overmeeren, 1994). The section clearly registers steps in the
water table attributable to small faults.
Figure 6.26 Example of a GPR section of push moraine deposits, showing groundwater steps at 286 m
and at around 330 m (Overmeeren, 1994)
6.9 Geophysical borehole logging
6.9.1 Principles and instrumentation
As long ago as the 1920s the petroleum exploration industry developed techniques to perform
geophysical observations in boreholes. These techniques were called wireline logging after the
principle of lowering a measuring probe into the borehole, suspended from a cable, but they are
generally referred to (geophysical) borehole logging or well logging.

174
Logging is applied in geohydrology to accurately determine the sequence and thickness of

rock formations, their physical properties, and to establish porosity, permeability, water flow and
groundwater salinity (Repsold, 1989). In addition, it is used to check and control the interpre-
tation of geophysical data recorded from the surface. Typical parameters, measured in boreholes
are: Resistance, Spontaneous Potential (SP), Formation Resistivity (in various modes), Formation
Conductivity (Induction Logging), Fluid Conductivity, Nuclear Radiation (in various modes),
Formation Temperature, Seismic Wave Velocity (sonic, acoustic), Borehole Diameter (calliper)
and Formation Fluid Flow.
(i) Instruments and tools
The set up of borehole logging equipment is shown in Figure 6.27. The basic components are the
measuring probe, which is suspended in the borehole by the logging cable, which is wound onto a
winch. The logging cable is connected to the actual measuring instrument, which controls the
logging speed and direction, signal levels and calibration, and supplies downhill power to the
probe, if needed. The measuring instrument also contains a paper and pen recorder, which
registers the signal level of the probe and continuously records the depth of the probe. Modern
instruments are usually fitted with a digital recording systems, which allows the data to be
redisplayed and processed in the office. Modern winches are driven electrically and conse-
quently need a generator to supply the power.
Normally there is a specific probe for every parameter. Therefore the probe can easily be
removed from the cable head and exchanged for the appropriate one. For some parameters it is
possible to combine several tasks into one probe.
The logging cable is made of steel wiring on the outside to take the weight of the cable
and probe in the borehole and has a core of electrical leads to transmit different signals, but also
to supply electric power to the probe. The logging cable is partly or completely covered by an
insulating sheath, to eliminate unwanted effects on electrical measurements. At the surface a
measuring sheave records the displacement of the cable via a stepmotor or an optical-electrical
device. The displacement information is linked to the recording instrument.
The recorder is fitted with a number of measuring modules, which can be switched on to
perform their task. Every module has its specific controls for calibration, signal zeroing, signal
level (recorder deflection) sensitivity and time constant. Some logging instruments particularly
in use in geohydrology are made very light and compact, to make them portable. Measuring
modules can then easily be removed and replaced, thus saving space and weight.
What is important in logging applications is that many logging techniques can only be
used in open, uncased boreholes. Most of the electrical techniques fall in this category, as well as
the calliper log and the sonic log. Other techniques, such as the nuclear logging, temperature
and fluid conductivity logging and flow meter measurements, can be recorded in completed
(cased) wells.
6.9.2 Logging physical parameters
The most common methods in borehole logging are the electrical methods. The probes to be
used are rather simple and inexpensive, with the exception of Induction Logging. Frequently used
in combination with the electrical methods are Natural Gamma Ray logging and Calliper logging.
(i) Electrical logging
Spontaneous Potential or SP is a parameter generated by the presence of drilling liquid, commonly

a mixture of mud and water. The SP is caused by two effects, an electrochemical potential and a

175
Groundwater studies
Figure 6.27 Basic components of geophysical borehole logging (Keys and MacCary, 1971)
streaming (electrokinetic) potential. The streaming potential is generated by the flow of water with
ions dissolved, through the narrow pores of a rock formation and can have a maximum value in
the order of 10 to 15 mV. The streaming potential across the mudcake on a sand layer is gener-
ally negative. The electrochemical potential is generated when two fluids with different salt con-
centrations are in contact with each other, either directly or via a semi-permeable membrane.
Lynch (1962) gives a detailed description of the mechanism. The electrochemical potential can be
negative or positive, depending on the salinity of the formation fluids.
The SP voltage is a relative value, whose sign depends on the formation fluids, the drill-
ing mud and the nature of the drilled formation, i.e. sand or clay (shale). In the interpretation,

176
the SP level at clay (or shale) formations is taken as a baseline. The SP will be relatively positive
across fresh water-bearing sands. The response may shift to a relatively negative deflection
across sand containing saline water. In a borehole the baseline can shift in position due to vari-
ations in (clay) lithology and groundwater composition. As a consequence, SP measurements are
only suitable for qualitative interpretation. Generally it is useful to determine layer boundaries,
to identify thin beds, lithology and a quantitative indication of saline groundwater.
(ii) Resistivity logging
Resistivity logging makes use of a similar arrangement of electrodes as in surface resistivity

surveying. One current electrode (B) and one potential electrode (N) are placed at the surface,
while the others (A and M) are positioned on the probe in the borehole. A potential field is
generated between A and B, while M measures the potential field around A (Figure 6.28). This
technique is also referred to as the Normal Device. The resistivity is calculated from : r = KV/I
, with the geometrical factor K = 4 AM.
Figure 6.28 Electrode configurations for single point resistance logging, normal resistivity logging and latero
logging. The single point resistance configuration can also be used for SP logging, when no power
is supplied to run a current through the formations
The depth investigation of a Normal Device depends on the electrode separation AM. A
small separation results in a shallow penetration (approximately 2AM), which is consequently
strongly affected by drilling mud in the borehole and by drilling fluid which has penetrated into
the borehole wall. A small separation also results in a better resolution of thin beds. A large
electrode separation suffers less from the borehole effects and therefore gives a better depiction
of the true formation resistivity f of the formation, with the disadvantage of lesser resolution
for thin beds. This entails fitting a resistivity probe with two potential electrodes M1 and M2.
Records with both a large and a small separation are called Long Normal and Short Normal
resistivities (LN and SN). Figure 6.29 shows an example of SN and LN measurements. Typically
a ratio of four is used for the LN and SN separations. In geohydrological applications, sepa-
rations are often 25 and 100 cm or 50 and 200 cm.

177
Groundwater studies
Figure 6.29 Example of a borehole log as used in groundwater surveys. The left graph shows the calliper
(borehole diameter) and SP log. The middle graph shows the long normal (LN) and short normal
resistivity logs, while the right graph represents the natural gamma radiation log with a lithological
interpretation. The resistivity logs show fresh water formations below ca. 130 m. Below this depth
the salinity increases until a depth of 180 m, to remain constant at larger depths. The gamma log
shows a high level between 155 and 160 m, correlating with resistivity (and SP) lows; this indicates
the presence of a clay layer (Courtesy TNO-Applied Geosciences, presently Netherlands Institute
for Applied Geoscience NITG)
Since the measurements are affected by the drilling mud, borehole diameter and the
invasion zone they must be corrected for these effects so that true formation resistivities can be
determined. To this end correction diagrams have been prepared, or computer techniques can be
applied.

178
(iii) Variations on the resistivity probe
The drive to obtain a more reliable value for the formation resistivity resulted in the devel-
opment of the Laterolog Device. In this variation, both M and N electrodes are placed on the
probe, with the A current electrode relatively far away (18 feet) from the midpoint of M and N.
This distance is called the spacing. Because of this array, the current is forced deeper into the
formations and the investigation depth is much greater than with a Normal Device.
(iv) Microresistivity Log
This tool, called a Dipmeter was developed to investigate the shallow invasion zone of the bore-
hole wall. It has all four electrodes closely spaced on a pad, which is pressed against the bore-
hole wall for good physical contact, but also to limit disturbing effects from the environment.
Because of the small electrode separation the tool can also register very thin beds within the
formation. The probe can also combine three or four pads, mounted at angles of 120 or 90. The
local dip of the thin layers can be established by correlation of the Microresistivity curves.
A variation on the microresistivity device is the Microlatero Log. It combines the spec-
ifications for the microlog with the laterolog.
(v) Single Point Resistance
This is a simplified version of resistivity logging. It uses one electrode at the surface and one in
the borehole (Fig. 6.28). Both act as current and as potential electrodes at the same time. The
recorded resistance response is not a linear one, therefore the data cannot be used for quantitative
interpretation. Still, the results are useful for qualitative interpretation, and very resistive, thin
layers can be detected very well without the problem of response inversion, as is the case in
normal logging. The technique requires relatively cheap instrumentation and a single core cable
and is often found in small and basic logging units.
(vi) Induction logging
This is an electric method, based on the electromagnetic principles as used in surface EM

methods. Transmitter and receiver coil are mounted in the probe. This tool was developed in
order to log in non-conductive drilling fluids (oil and gas). Typically this tool has a large
potential for use in geohydrology to log in boreholes above a low water table, since it does not
require physical contact between the probe and the borehole wall. It can also be used in PVC
cased boreholes. At present, induction loggers specifically designed for use in groundwater
exploration are available.
(vii) Calliper logging
As discussed before, it is often necessary to correct recorded data for the borehole diameter
effect. Irregularities in borehole diameter can be caused by material from the borehole wall
washing out during drilling, and by the drilling system. The well diameter can be measured
directly, by three or four feelers attached to the probe. The expansion of these feelers is ruled
by the borehole diameter and translated into a signal for the recorder. Results of a calliper
measurement are shown in Figure 6.29.

179
Groundwater studies
(viii) Nuclear logging
Nuclear logging concerns a group of methods, based on recording the natural radiation of
formations or the effect of formations on radiation transmitted from a source in the probe.
Natural Gamma Radiation logging is based on the emission of gamma radiation (photons)
by radioactive minerals, such as uranium, thorium and the K40 isotope of potassium. As well as
to gamma radiation, alpha and beta radiation are also generated, but these do not have sufficient
power to penetrate the ground. K40 is a very common natural component in earth materials and
occurs more frequently in clays and shales than in sands or chalks and dolomites. Therefore the
level of radiation is an indication of the type of lithology. The gamma radiation is measured with
a scintillometer, often containing an NaI crystal, which emits light when hit by a photon
(scintillation). The emitted light is converted by the photocathode of a photomultiplier tube into
emission of electrons. The photomultiplier amplifies the electron flow to a recordable level. The
gamma radiation is a statistical process, which means that the intensity varies with time. To
compensate for the irregular emission, the radiation counts are averaged over a time interval,
the time constant TC. Usually the TC is in the order of 2 to 10 seconds. The time constant is set by
the operator. A small TC will show larger statistical variations, i.e. the repeatability of a gamma
log will be small. A large TC will have less statistical variation, but if the logging speed of the
tool is too high, it may not produce the true radiation level of a particular formation. It is clear
that thin layers would then not be detected through their radiation level. Logging speed in
gamma radiation logging is typically of the order of 1 to 2 metres per minute. Calibration of a
gamma radiation tool opens up ways of quantitatively interpreting geohydrological parameters.
Gamma Gamma logging or Density Logging uses a gamma radiation source, which is built
into the probe. The emitted radiation penetrates into the surrounding formation and is scattered
by it and partly absorbed. The backscattered radiation is recorded by the probe. The degree of
absorption is related to the density of the formation. Thus the gamma-gamma log represents
variation of density with depth. This tool is particularly useful for identifying relatively dense
anhydrite and barite. Again the method can be used for quantitative analysis if the tool is
calibrated.
Neutron logging includes two variations: the Neutron-Gamma log and the Neutron-Neutron
log. The neutron-gamma log uses a fast neutron source in the probe, which radiates thermal
neutrons (with high energy level) into the formation. When the neutrons collide with nuclei in
the rock, gamma radiation is emitted, which in turn is recorded by the probe. This process is
particularly sensitive to the presence of hydrogen in the formation, i.e. water and hydrocarbons.
It is therefore an effective tool to determine the relative amount of pore water in a formation.
The neutron-neutron log also uses a neutron source in the probe, but records the remaining
(backscattered) neutron with a neutron counter. The physical effect of hydrogen is the same and
the curves obtained are similar. Calibrated tools allow for quantitative analysis of porosity in the
saturated zone. Above the water table the moisture content can be determined.
Logging techniques using radiation sources are less commonly used, since the use of these
instruments requires more skilled operators than the other techniques. Protection personnel
against health hazards and permits for storage and transport of equipment require higher
expenditure than for normal equipment.
(ix) Sonic logging, or velocity logging
This method is used to measure acoustic wave velocity in the formation around the borehole.
The velocity is related to density and the elastic properties of the rock. Measurements are made
with a probe, which contains a wave transmitter and two or more receivers placed at different
distances from the source. The tool records the travel time of the sonic wave between transmitter

180
and receiver. The transmitter and receivers are acoustically insulated, to prevent the signal from
travelling directly through the tool itself. The probe is centred in the borehole, which means that
the wave travels partly through the drilling mud. The distance travelled in the borehole is
eliminated by the use of at least two receivers. The travel time is usually recorded in msec/foot.
Velocity calculations are based on first arrival recording of the compressional wave. The porosity
of a formation can be calculated from the empirical time-average equation, developed by Wyllie,
Gregory and Gardner (1958). These porosity calculations are useful in consolidated rocks, but for
unconsolidated formations it is very difficult to make reliable calculations. A compaction factor
1/Cp must then be applied to the above expression.
Generally not much use is made of the density and sonic logs in geohydrology, since the
logging is rather expensive and the processing is based on computer applications. The sonic logs
however, provide additional information for reflection seismic interpretation. Density and veloc-
ity parameters can be used to prepare synthetic seismograms, which are used for correlation
with the processed field data. The nuclear methods can be used in cased and open boreholes.
(x) Flow meter measurements
Flow meter measurements are made with a probe in the cased borehole to measure the water
flow from aquifer formations during production. In principle an impeller is used to monitor the
flow, but more sophisticated techniques based on mass movement are also used. This yields an
impression of the permeability of different parts of the formation.
(xi) Fluid conductivity
Fluid conductivity can be measured with a water conductivity meter attached to the logging
cable. Very often it is combined with the flow meter. It presents a continuous water conductivity
profile of the drilled formation.
(xii) Temperature logging
Temperature logging results in a temperature profile of the borehole, in which anomalous high
or low segments are indicative of the origin of inflowing water. This recording is usually made
after one or two weeks after completion of the well, in order to prevent effects of the drilling
activities on the natural thermal balance of the location.
The main value of borehole logging lies in correlation. Logs from different boreholes in a survey
area or wellfield are carefully checked for similarity in response, which can lead to absolute cor-
relation or to a correlation of a certain sequence of layers. Invaluable stratigraphic insight is
obtained in this way
In the previous sections it was already mentioned that much of the borehole log infor-
mation can be used to do a quantitative interpretation, allowing true formation resistivity, poros-
ity, density and saturation to be established. Since the boreholes are logged almost continuously,
the enormous amount of data calls for computer processing. In this aspect, the applications in
geohydrology are approaching those of petrophysics as applied in oil exploration.
Most important in data processing and interpretation is the correlation of the physical and
geophysical parameters with the hydrological parameters of an aquifer. Potentially this allows
detail in modelling an aquifer to be improved, which in turn may lead to better groundwater
management.

181
Groundwater studies
ALVAREZ, R., 1991. Geophysical Determination of Buried Geological Structures and Their
Influence on Aquifer Characteristics. Geoexploration, Vol. 27, pp. 124.
ANDERSON, W. L., 1979. Numerical Integration of Related Hankel Transforms of Orders 0 and 1
by Adaptive Digital Filtering. Geophysics, Vol. 44, pp. 1287305.
ANGELITO, V.; CERVERA, G.; CHAPELLIER, D., 1991.La gravimtrie expditive applique la
recherche daquifres en zone aride; cas de la nappe alluviale du Teloua (Agadez, Niger).
Geoexploration, Vol. 27, pp. 17992.
ARCHIE, G. E., 1942. The Electrical Resistivity Log an Aid in Determining Some Reservoir Char-
acteristics. Trans. A.I.M.E., Vol. 146, pp. 5464.
ASTIER, J. L.; PATERSON, N. R., 1989. Hydrogeological Interest of Aeromagnetic Maps in Crys-
talline and Metamorphic Areas. Paper 59. In: Garland, G. D. (ed.), Proc. Exploration 87:
Third Decennial International Conference on Geophysical and Geochemical Exploration for
Minerals and Groundwater, Ontario Geological Survey, Special Vol. 3, 960 pp.
BARKER, R. D., 1981. The Offset System of Electrical Resistivity Sounding and Its Use with a
Multicore Cable. Geophysical Prospecting, Vol. 29, pp. 12844.
BOURGEOIS, B.; MATHIEU, F.; VACHETTE, C.; VAUBURG, P., 1994. AMT Measurements Compared with
Gravity and Magnetometry for Structural Study of a Sedimentary Basin. Letthakeng-
Bothlapatlou Groundwater Project, Botswana. J. Appl. Geophysics, Vol. 31, pp. 725.
BROMLEY, J.; MANNSTRM, B.; NISCA, D.; JAMTLID, A., 1994. Airborne Geophysics Application to a
Ground-Water Study in Botswana. Groundwater, Vol. 32, pp. 7990.
BUTLER, D. K., 1984. Microgravimetric and Gravity Gradient Techniques for Detection of
Subsurface Cavities. Geophysics, Vol. 49, pp. 108496.
CARMICHAEL, R. S.; HENRY, G., 1977. Gravity Exploration for Groundwater and Bedrock Topo-
graphy in Glaciated Areas. Geophysics, Vol. 42, pp. 8509.
FRISCHKNECHT, F. C., 1971. Electromagnetic Scale Modelling. In: J. R. Wait (ed.), Electromagnetic
Probing in Geophysics. The Golem Press, Boulder CO, pp. 265320.
GHAZALA, H. H., 1994. Structural Interpretation of Bouguer and Aeromagnetic Anomalies in
Central Sinai. J. African Earth Sci., No. 19, pp. 3522.
HAGEDOORN, J. G., 1959. The Plus-Minus Method of Interpreting Seismic Refraction Results.
Geophysical Prospecting, Vol. 7, pp. 15882.
HAWKINS, S. L. V., 1961. The Reciprocal Method of Routine Seismic Refraction Investigations.
Geophysics, Vol. 26, pp. 80619.
HUNTER, J. A.; PULLAN, S. E.; BURNS, R. A.; GAGNE, R. M.; GOOD, R. L., 1984. Shallow Reflection
Seismic Mapping of the Overburden-Bedrock Interface with the Engineering Seismo-
graph Some Simple Techniques. Geophysics, Vol. 49, pp. 13815.
KELLER, G. V.; FRISCHKNECHT, F. C., 1966. Electrical Methods in Geophysical Prospecting. Pergamon
Press.
KEYS, W. S.; MACCARY, L. M., 1971. Application of Borehole Geophysics to Water Resources
Investigations. Techniques of Water Resources Investigations of the US Geological Survey, Book
2, Chapter E1. Washington, D.C., United States Government Printing Office.
KOSINSKY, W.; KELLY, W. E., 1981. Geoelectric Sounding for Predicting Aquifer Properties. Ground-
water, Vol. 19.
LYNCH, E. J., 1962. Formation Evaluation. Harper and Row, New York.
MAZAC, O.; KELLY, W. E.; LANDA, I., 1985. A Hydro-Geophysical Model for Relations between
Electrical and Hydraulic Properties of the Aquifers. J. Hydrol., Vol. 79, pp. 119.
MCNEILL, J. D., 1990. Use of the Electromagnetic Methods for Groundwater Studies. In: S. H.
Ward (ed.), Geotechnical and Environmental Geophysics, Vol. 1: Review and Tutorial. Tulsa,
Society of Exploration Geophysicists, pp. 191218.

182
MEEKES, J. A. C.; SCHEFFERS, B. C.; RIDDER, J., 1990, Optimization of High-Resolution Seismic
Reflection Parameters for Hydrogeological Investigations in The Netherlands. First Break,
Vol. 8, pp. 26370.
MILSOM, J., 1989. Field Geophysics. Geological Society of London Handbook. Publ. Open
University Press.
PALMER, D., 1980. The Generalized Reciprocal Method of Seismic Refraction Interpretation. In:
K. B. S. Burke (ed.), Soc. Explor. Geophs., Tulsa.
PELLERIN, L.; LABSON, V. F.; PFEIFFER, M. C., 1994. VETEM A Very Early Time Electromagnetic
System (abstract). The John S. Sumner Memorial International Workshop on Induced
Polarization in Mining and the Environment, Tucson, Arizona.
REEVES, C. V., 1992. New Horizons in Airborne Geophysical Mapping. Exploration Geophysics,
Vol. 57, pp. 9569.
REPSOLD, H., 1989. Well Logging in Groundwater Development. IAH Contrib. to Hydrogeol., Vol. 9,
Heise, Hannover, 136 pp.
SENGPIEL, K. P., 1986. Groundwater Prospecting by Multifrequency Airborne Electromagnetic
Techniques. In: G. J. Palacky (ed.), Airborne Resistivity Mapping, Geological Survey of Canada
Paper 8622, pp. 1318.
SPORRY, R. J.; ARZATE, J. A.; YOUSSIF, M. S., 1991. Geophysical methods in groundwater modelling.
Proc. A. E. G. Seminar, Vol. 2., Hyderabad.
SHERIFF, R. E.; GELDART, L. P., 1982. Exploration Seismology, Vol. 1: History, Theory and Data Acqui-
sition; Vol. 2: Data-processing and Interpretation. Cambridge University Press, Cambridge.
STEEPLES, D. W.; MILLER, R. D., 1990. Seismic Reflection Methods Applied to Engineering,
Environmental and Groundwater Problems, In: Stan Ward (ed.), Soc. Explor. Geophs.,
Investigations in Geophysics, No. 5, Volume on Environmental Geophysics, pp. 130.
SUMNER, J. S., 1976. Principles of Induced Polarization for Geophysical Exploration. Elsevier Science
Publ. Co., 277 pp.
TELFORD, W. M.; GELDART, L. P.; SHERIFF, R. E., 1990. Applied Geophysics. Second Edition,
Cambridge University Press.
VAN DER VELPEN; SPORRY, R. J., 1993. RESIST A Computer Program to Process Resistivity
Sounding Data on PC-Compatibles. Computers and Geosciences, Vol. 19, pp. 691703.
VAN OVERMEEREN, R. A., 1981. A Combination of Electrical Resistivity Seismic Refraction and
Gravity Measurements for Groundwater Exploration in Sudan. Geophysics, Vol. 46,
pp. 130413.
, 1994. Georadar for Hydrogeology. First Break, Vol. 12, pp. 401 8.
VAN OVERMEEREN, R. A.; RITSEMA, I. L., 1988. Continuous Vertical Electric Sounding. First Break,
Vol. 6, pp. 31324.
WYLLIE, M. R. J., GREGORY, A. R.; GARDNER, G. H. F., 1958. An Experimental Investigation of
Factors Affecting Elastic Wave Velocities in Porous Media. Geophysics, Vol. 23, pp. 45993.
YILMAZ, ., 1987. Seismic Data Processing. Soc. Expl. Geophys., Tulsa.

183
7 We l l d r i l l i n g a n d d e s i g n m e t h o d s
7.1 Introduction
The first use of groundwater as a source of water supply is lost in antiquity. In fact, as soon as
the primitive man learned to domesticate and rear cattle and sheep in the valleys of the Tigris-
Euphrates, Yellow, and Indus rivers (10000 BC) the well became his most important possession.
Since that time, virtually all homes with their own water supply have wells and use ground-
water. Furthermore, groundwater is increasing in importance in world water supplies, in part as
a response to the growing costs and other constraints in storing and treating surface water and
partly because the advantages of groundwater are now better understood. Today, more than half
of the worlds population depends on groundwater for survival (UNESCO, 1992).
The use of wells for domestic purposes (human and animal consumption) usually has the
highest priority, followed by industrial requirements, and finally agricultural use. Today, wells
provide 75 percent of drinking water supplies in Europe and more than 50 percent in the United
States (Cohen, 1985; Driscoll, 1989; Rebouas, 1976; Solley et al., 1993). During the last 25 years
about 300 million wells have been drilled for groundwater withdrawal all over the world, and in
the United States of America alone, 100 million are currently in use (Carpenter, 1994; Fetter, 1993;
Rebouas, 1991, Solley et al., 1990; UNESCO, 1992). Most of these are generally 150 mm (6 in) in
diameter and less than 150 m deep; only a small proportion are over 250 m in depth. However,
in the large sedimentary basins there are many wells deeper than 500 m that reach the confined
aquifers; e.g. wells in the huge sedimentary basins, such as Great Australian Artesian Basin,
Paran and Marano Basins in Brazil (some of more than 2000 m depth and with diameters of
the pump chamber casing of 762 mm) yielding from tens to hundreds of thousands of cubic
metres a day (Habermehl, 1985; Margat, 1990; Rebouas, 1988; Rebouas, 1991).
Discussion in this chapter is limited to drilled vertical wells for producing or monitoring
groundwater. Because it is not possible to describe in detail the numerous drilling and well
construction techniques, the reader is urged to consult the major sources cited in this chapter.
7.2 The development of well drilling techniques
Shallow, hand-dug wells and crude water-lifting devices marked the early exploitation of
groundwater. The oldest dug-wells known were found in the Middle East and are dated from
8000 BC. The art of drilling wells by lifting and dropping a string of tools suspended on a cable
was invented, perfected, and extensively practised by the ancient Chinese 5000 before present.
The early Chinese could drill wells to depths ranging between 1,200 and 1,500 metres (Hard-
castle, 1987; Meinzer, 1934; Tolman, 1937).

185
Groundwater studies
Percussion methods of well drilling were developed much later in Greece and Rome
(500 BCAD 500). Medieval advances (AD 5001500) were made in Western Europe, largely after
the discovery of flowing wells, first in France about 1100 AD, and a few decades later in eastern
England and northern Italy. The wells in the region of Artois, discovered in 1126, became so
famous that flowing wells are often called artesian wells after the name of the region (Tolman ,
1937).
The emergence of Western technology (15001750) and the birth of steam power, rapidly
improved the development of the cable tool method of drilling, often referred to as the standard
method, or percussion method. The Industrial Revolution (17501900) brought advances in
power technology, such as development of steam power, internal combustion engine, and
electric power, which induced the most impressive development of the cable tool method of
drilling, as a result in part to knowledge borrowed from the oil and gas industry (Bowman, 1911;
Campbell and Lehr, 1973; Driscoll, 1989; Hardcastle, 1987; Lehr et al., 1987; Meinzer, 1934).
The hydraulic rotary method of well drilling was developed in the oil fields of Louisiana
(United States of America) about 1890. Since that time it has replaced the cable-tool method of
oil well drilling, and even of water well drilling to a great extent (Bowman, 1911; Campbell and
Lehr, 1973; Driscoll, 1989; Gordon, 1958; Speedstar Division of Koering, 1967). Later, the devel-
opment of the combined hydraulic rotary-percussion methods of well drilling and of the electric-
driven pumps made possible the recovery of groundwater in large amounts (up to 1,000 m3 per
hour), and at increased depths (up to more than 2,000 metres). At the present time, there are no
technological limitations to reaching the deeper confined aquifers in all over the world
(Anderson, 1992; Driscoll, 1989; Lehr et al., 1987; Rebouas, 1994; Solly et al., 1993).
If in the early times of civilisation the well was the base for development in the arid and
semi-arid regions, recently groundwater has become an important source of water supply for
domestic, industrial and agricultural purposes in temperate and even in the humid tropical
climatic regions, in view of its long range economic feasibility. Currently, most of the drilled
wells are generally 15 cm in diameter and less than 150 m deep; only a small proportion are over
200 m in depth, but in the large sedimentary basins there are many wells deeper than 1,000 m
that reach the best aquifers (Driscoll, 1989; Margat, 1990; Rebouas, 1991; 1994, Smith, 1980).
7.3 Water quality protection for wells
In the past, the purity and sanitary quality of groundwater was assumed, and even when
groundwater resources were used for drinking water supplies, little or no treatment was thought
to be required. But in recent years, there is growing evidence that this resource is being
contaminated locally. The main pollution sources are municipal and industrial wastes, sewage
treatment and disposal, spills and leaks from storage and transport of liquids, well injection of
liquid wastes, agricultural activities, and mining activities.
Groundwater contamination is so severe in certain localities that continued use of the
water could lead to serious health problems. Protection of water quality partly depends on
the well design, the drilling methods and the materials used to construct the production and
monitoring wells. When those wells are badly designed, constructed, and left uncontrolled, they
invite groundwater quality degradation (Fetter, 1993; Rebouas, 1991; EPA, 1975, 1980). As a
result, more than ever, the drilling contractor is faced with the challenge of complying with new
legislation that addresses groundwater quality and the growing number of groundwater
contamination problems. These challenges and problems can be handled satisfactory only
through greater understanding of the water well construction standards for all new and
reconstructed wells intended or used for supplying water for human consumption, preparation
of food products, and monitoring groundwater quality.

186
Well drilling and design methods
Collector wells and infiltration galleries are sometimes constructed where special
hydrogeological conditions exist (Wright et al., 1987).
7.4 Standards for water and monitoring wells
Many countries, states and organisations of drilling contractors have established minimum
standards covering such aspects of water well construction as: location with respect to possible
pollution sources, casing types and weights, sanitary seals, the protection of aquifers, and the
sterilisation of wells (ASTM, 1992; Driscoll, 1989; Fetter, 1993; Lehr et al., 1987; OECD, 1989; EPA,
1975).
Water well standards should assure that:
the completed well will be constructed of material that will be compatible with the
environment and will give an adequate well life time,
workmanship will be of an acceptable standard,
the well will be of adequate size and design to yield the desired discharge with
maximum efficiency.
Standards and well specifications should:
be reasonably flexible,
foster competitive bidding,
permit the preparation of cost estimates,
be a guide to those responsible for supervising the construction of the well.
For monitoring work, many of the objectives of a drilling programme are similar to those for a
water well, but some of the construction steps must be made with greater care, to ensure that the
water quality is protected and reliable water samples can be obtained (Aller et al., 1989; Driscoll,
1989; Fetter, 1993; EPA, 1975; USDI, 1981).
Groundwater protection policy and strategy should be based on the concept that pre-
vention of pollution is always less expensive than aquifer rehabilitation, which is a costly, time
consuming and technically demanding task.
7.5 Well drilling techniques
Shallow low-capacity production wells are usually dug, driven, or jetted, and high-capacity
production wells are usually drilled by cable-tool rigs, direct circulation rotary rigs and
variations thereof such as reverse circulation rotary drilling, air drilling systems, and inverse
drilling methods. In hard rocks down-the-hole-hammer techniques a combination of
percussion and rotary drilling are used.
Numerous drilling techniques have been developed to cope with the wide range of geo-
logical conditions, from hard rock such as granite and dolomite to completely unconsolidated
sediments such as alluvial sand and gravel. Moreover, the drilling contractor may vary the usual
drilling procedure depending on the depth and diameter of the well, type of geological
formation to be penetrated, sanitation requirements, and principal use of the well, such as
production, recharge or monitoring.
In this chapter only the most frequently used well drilling techniques will be described.
7.5.1 Cable tool method and variations
The cable tool method of drilling, often referred to as the standard method, or percussion
method, is one of the oldest and simplest drilling methods (Figure 7.1). The cable tool

187
Groundwater studies
Figure 7.1 Typical truck-mounted cable tool equipment for drilling wells. Regular bailer with flat valve bottom
(After Driscoll, 1989)
equipment is probably the most versatile of all rigs in its ability to drill satisfactorily under a
wide range of environmental and hydrogeological conditions. It may be the best, and in some
cases the only, method to use in coarse glacial till, boulder deposits, or rock strata that are highly
disturbed, broken, fissured, or cavernous.
The cable tool machine is usually compact, and powered with a diesel engine placed at
the front end of the frame. It can be easily moved in rough terrain. The simplicity of design,
ruggedness, and easy maintenance of the rig and tools are particularly advantageous in isolated

188
areas. Moreover, it generally requires less skilled operators and a smaller crew than other rigs.
The low horsepower requirements are reflected in lower fuel consumption, an important aspect
where fuel costs are high or sources of fuel are remote. Cable tool rigs are generally limited to
drilling maximum hole diameters of 600 to 750 mm (24 to 30 inches) and to depths of less than
600 m (2,000 ft).
The hole is drilled by raising and lowering a heavy bit on the end of a steel cable which is
threaded over sheaves at the top of the mast and down to the drill-line drum. For effective
penetration the drilling motion of the cable tool machine must be synchronised with the gravity
fall of the full string of cable tools. The drill bit breaks consolidated rock into small fragments,
but when drilling in unconsolidated sediments it primarily releases the material. The broken and
crushed material in the bottom of the hole is removed by means of a bailer (Figure 7.1).
In stable rock, an open hole can be drilled, but in unconsolidated formations, casing must
be driven down the hole during the drilling. Above the water table or in otherwise dry
formations, water is added to the hole to form a slurry of the cuttings so they may be readily
removed by a bailer. However, much less water is required for drilling than with most other
commonly used rigs, an important consideration in arid and semi-arid zones. Also, sampling
and geological logging are simpler and more accurate with the cable tool rig. Table 7.1 presents
the major advantages and disadvantages of the cable tool method (Anderson, 192; Bowman,
1911; Campbel and Lehr, 1973; DAO, 1965; Driscoll, 1989; Lehr et al., 1987; NWWA, 1981, 1985;
Roscoe, 1985; USDI, 1981).
Table 7.1 Advantages and disadvantages of cable tool method
Advantages Disadvantages
1. Rigs are relatively cheap, require little maintenance, 1. Relatively slow rate of progress
are readily moved in rugged terrain or where space compared to other rigs of similar capacity.
is limited.
2. Machines have low energy requirements and can 2. Economic and physical
drill in areas where little make-up water is available. limitations to depth and diameter.
3. Reliable for a wide variety of geological conditions, 3. Necessity of driving casing in

and the samples are not contaminated by drilling mud. unconsolidated materials and difficulty
of pulling back long strings of casing.
4. Generally require less skilled operators and 4. Each cable tool driller can complete only
smaller crew than other rigs of similar capacity. a limited number of holes per year.
The major disadvantage, compared to other types of rigs, is its lower rate of progress and
its depth limitations. A further disadvantage of the cable tool rig is the necessity of driving
casing concomitant with drilling in unconsolidated materials. This precludes the geophysical
well-logging, although the logging is desirable in many instances. However, gamma logs may be
taken inside the casing. The driving of casing necessitates a heavier wall pipe than would other-
wise be required in some installations. Screens often must be set by pullback or bail-down
methods. The pullback method in deep or large diameter wells is sometimes extremely difficult,
and the bail-down method may give rise to problems of alignment (AWWA, 1984; Bowman,
1911; Driscoll, 1989; Gordon, 1985; Lehr et al., 1987; NWPA, 1981; OECD, 1989; USDI, 1981).
Every cable tool machine has certain interdependent limits to borehole depth and
diameter. For example, if a hole is relatively small in diameter, it may be drilled to relatively
great depth. In larger-diameter holes, the weight of the drill string and cable may become so
excessive that the machine cannot function, thereby limiting well depth to a smaller diameter.

189
Groundwater studies
A variation of the cable tool drilling technique, called the full-hole cable tool method, has
been used in Brazil since the 1960s. With the open hole full of water or drilling fluid, the string
of cable tools is lifted and dropped to cut the borehole. In this case the hole can be drilled open-
hole, even in completely soft or unconsolidated sediments, because the hydrostatic water or
drilling fluid pressure prevents the borehole walls from caving in. Wells between 150 to 250 mm
(6 to 10 inches) in diameter have been drilled in this manner to depths of 100 m in surficial
aquifers which occur in Tertiary unconsolidated deposits, in crystalline rocks with deeply
weathered mantle or detrital cover, and in shallow confined or semi-confined sandy aquifers
(Rebouas, 1988).
Another cable tool drilling technique, used for many years in Japan and more recently in
the Western United State, is called the open-hole or reverse cable tool method. With the borehole
full of water or drilling fluid, heavy sand pumps or bailers are operated inside the casing to cut
the borehole. Using this method wells to about 600 mm (24 inches) in diameter and to about
30 m in depth for irrigation have been drilled and screened within one day (USDI, 1981).
Jet drilling is basically a percussion method combined with a pressure pump. The drill
pipe is lifted and dropped, which chops up material at the bottom of the hole. The water helps
to jet the broken material loose and caries the cuttings up the hole where they are discharged
(Campbell and Lehr, 1973; Driscoll, 1989).
The full-hole cable tool method and jet drilling are readily adapted to drilling holes 50 to
100 mm (2 to 4 inches) in diameter in soft formations such as sand and silt. These holes are
Figure 7.2 Major components of a rotary drilling rig (After Driscoll, 1989)

190
Figure 7.3 Diagrams of a direct (a) and a reverse (b) rotary circulation system; arrows indicate the direction
of the mud circulation
useful in installing aquifer test observation wells, monitoring wells and small capacity water
wells.
Both the jetting and full-hole cable tool variation methods are disadvantageous in regard
to sampling of formations and water during the constructing operation. Also, if the permeability
of the penetrated sediments is much greater than about 10 m per day, injected water may disturb
the water characteristics at some distance from the wall of the hole (Campbell and Lehr, 1973;
Driscoll, 1989; Lehr et al., 1987; Roscoe, 1985).
7.5.2 Direct circulation rotary drilling and variations
The hydraulic direct circulation rotary method of drilling was first used in the oil fields of
Louisiana (United States of America) about 1890. Since that time it has become important for the
drilling of water wells, especially in deep sedimentary confined aquifer systems (Anderson,
1992; Driscoll, 1989; Lehr et al., 1987; Roscoe, 1985; USDI, 1981). The direct circulation rotary rig
drills by rotating a bit at the lower end of a string of drill pipe, which transmits the rotating
action from the rig to the bit (Figure 7.2). As the bit is turned, water-based drilling fluid (mud)
is pumped down the pipe to lubricate and cool the bit, to pick up material from the bottom of
the hole, and to clean the hole by transporting the cuttings to the surface in the annular space
between the uncased hole wall and the drill pipe. The water based drilling fluid also forms a
thin layer of mud on the wall of the hole which reduces seepage losses and, together with the
hydrostatic head exerted by the mud column, holds the hole open. At the surface, the fluid is
channelled into a settling pit or pits where most of the cuttings drop out. Clean fluid is then
picked up by the pump at the far end of the pit or from the second pit and is recirculated down
the hole (Figure 7.3) Drilling fluids used in the water well industry include water-based and

191
Groundwater studies
Table 7.2 Major types of drilling fluids used in the water well industry
Water-based Air-based
1. Clean, freshwater 1. Dry air
2. Water with clay additives 2. Mist: Droplets of water entrained in the air stream
3. Water with polymeric additives 3. Foam: Air bubbles surrounded by a film of water
containing a foam stabilising surfactant
4. Water with clay and polymeric additives 4. Stiff foam: Foam containing film strengthening
materials such as polymers and bentonite
air-based systems; the major types are shown in Table 7.2 (Anderson, 1992; Driscoll, 1989;
NWWA, 1985; USDI, 1981).
The type of drilling fluid selected will depend principally on the rock formation expected
to be penetrated, the equipment available, and occasionally on environmental regulations, and
the experience of the drilling crew. For example, water-based drilling fluid with clay or poly-
meric additives is commonly used in unconsolidated formations, while air-based fluid is used in
consolidated rocks and sediment; and clean water is used by reverse rotary drilling equipment
for drilling large diameter wells in unconsolidated or semi-consolidated sediments. Regardless
of which drilling fluid type is used, its effectiveness will depend upon the crews ability to
anticipate the chemical and physical changes taking place during drilling and to modify the
fluid as required.
At a minimum, all managing staff on a rotary rig should be able to measure drilling fluid
density and viscosity, and understand the relationship of these properties to hole stability,
cuttings removal, and fluid-loss control (Driscoll, 1989; NWWA, 1981; USDI, 1981).
Because of limitations in pump capacity and therefore effective cuttings removal, most
direct rotary equipment used to drill water wells is limited to boreholes with maximum
diameters of 550 to 600 mm (22 to 24 inches).
These sizes may not be sufficient for high-capacity wells, especially those that must be
filter packed. As hole diameter increases past 600 mm (24 inches), the rate of penetration
becomes less satisfactory. Under these conditions, telescopic well designs have been adopted, as
shown in Section 7.7. Direct rotary drilling, the most common method, offers advantages and
disadvantages, as presented in Table 7.3 (AWWA, 1984; Driscoll, 1989; Lehr et al., 1987; USDI,
1981).
7.5.3 Reverse circulation rotary drills
Reverse circulation drills have been designed to overcome the limitation on hole diameter and
drilling rate. Originally they were used only in unconsolidated formations. Recently, the reverse
circulation rotary drills have been used both in fairly consolidated rocks such as sandstone and
even in hard rocks, using both water and air as the drilling fluid. This machine is probably the
most rapid drilling equipment available for unconsolidated formations, but requires a large
volume of water which must be constantly replenished since drilling mud is rarely used.
In Table 7.4 are presented the advantages and disadvantages of the reverse circulation
rotary method (AWWA, 1984; Driscoll, 1989: Lehr et al., 1987; NWWA, 1981; EPA, 1975; USDI,
1981).

192
Table 7.3 Advantages and disadvantages of the direct circulation rotary drilling method
1. Penetration rates are relatively higher 1. The cost of a rotary rig is much higher
in all types of materials, and drilling depth than a cable tool rig of equal capacity, and
capacities are greater maintenance and repair are more complex
2. Minimal casing is required during the 2. Requires much more skilled and
drilling operation, and permits use of larger crew with drilling fluid knowledge
most geophysical well-logging equipment and experience
3. Rig mobilisation and demobilisation 3. Mobility of the rigs may be limited by the
are rapid land surface, mostly the slope or wetness
conditions
4. Well screen can be set easily as part of 4. Collection of accurate samples requires
the casing installation special procedures and use of drilling fluids
may cause plugging of certain formations
Table 7.4 Advantages and disadvantages of the reverse circulation rotary method
1. Large-diameter holes (400 to 1,800 mm) 1. It requires a large volume of water can be drilled
quickly and economically which must be constantly replenished
2. The porosity and permeability of the 2. The rigs and components are usually larger
penetrated formations are relatively maintenance more complex, and thus more
undisturbed near the borehole expensive
3. Well screens can be installed easily 3. More larger and skilled crew are generally
as part of the casing installation required
4. Most sedimentary formations can be 4. Larger mud pits are required than for other
easily drilled drilling methods
5. This drilling method is probably the 5. Igneous and metamorphic rocks can not be
most rapid drilling equipment available drilled
for unconsolidated formations
The reverse circulation drill has essentially the same design as the direct rotary drill except that
most pieces of equipment are larger. Also, the operation is essentially the same as a direct cir-
culation rotary rig, except that the water is pumped up through the drill pipe rather than down
through it (Figure 7.3). The fluid returns to the borehole by gravity flow, and moves down the
annular space between the drill pipe and borehole wall to the bottom of the hole, picks up the
cuttings, and re-enters the drill pipe through ports in the drill bit. The discharge is directed into
a large pit in which the cuttings settle out. To prevent the hole from caving in, the fluid level
must be kept at ground level at all times, even when drilling operation is suspended tem-
porarily. Much reverse drilling equipment is equipped with air compressors to aid in circulating
the drilling fluid. The hydrostatic pressure of water column plus the velocity head outside the
drill pipe support the borehole wall (AWWA, 1984; Driscoll, 1989; Lehr et al., 1987; NWWA,
1981; USDI, 1981).

193
Groundwater studies
In the reverse circulation rotary method, the drilling fluid can best be described as muddy
water rather than drilling fluid, and additives are rarely mixed with the water to make a viscous
fluid. Suspended clay and silt that recirculate with the fluid are mostly fine materials picked up
from the penetrated geological formations.
7.5.4 Air drilling systems
The air rotary drilling method was developed primarily in response to the need for a rapid
drilling technique in mining exploration and geotechnical surveys and more recently in hard
Figure 7.4 Basic components of an air rotary drilling system (After Driscoll, 1989)

194
rock hydrogeology. Two different drilling methods use air as the primary drilling fluid: direct air
rotary method and down-the-hole air hammer. The direct air rotary equipment is essentially the
same as for direct circulation rotary drilling, except the fluid channels in the bit are of uniform
diameter rather than jets, and the mud pump is replaced by an air compressor (Figure 7.4). Air is
circulated down the drill string to cool the bit and to blow the cuttings to the surface (Driscoll,
1989; Lehr et al., 1987; Roscoe, 1985).
Initially, the air rotary method was used for relatively small diameter such as 50 to
150 mm (2 to 6 inches) boreholes in hard rocks. Larger holes have become possible through use
of foams and other air additives, and diameters up to 800 mm (36 inches) have been drilled
successfully. The equipment is mostly used in hard rock terrain where water is encountered in
fractures or similar openings and wells are completed as open holes. Penetration rates are often
faster and the bit life longer when using air as compared with water-based drilling fluids. Better
bottom-hole cleaning is partly responsible for this difference in performance. If too much water
from the formation comes into the hole during drilling, however, the penetration rate is no
better than when drilling with water-based drilling fluids (Driscoll, 1989; USDI, 1981).
Tricone rock bits, up to about 305 mm (12 inches) diameter, similar to those designed for
drilling with water-based fluids, are commonly used. In many areas bottom bits, with sintered
tungsten-carbide inserts in the perimeter of the steel rollers, are used successfully (Driscoll, 1989;
Lehr et al., 1987; USDI, 1981). Figure 7.5 shows the guide for the use of bit types in air drilling
systems (Driscoll, 1989).
Because of the time lag before cuttings arrive at the surface, it is not advisable to base well
design on samples of cuttings.
Shortly after the development of air rotary drilling, the down-the-hole hammer bit was
Figure 7.5 Guide for the use of bit types in air drilling system (After Driscoll, 1989)
195
Groundwater studies
developed. This second direct rotary method using air is called the down-the-hole drilling
system. A pneumatic drill operated at the end of the drill pipe rapidly hammers the rock while
the drill pipe is slowly rotated. The rate of penetration in several rock types is higher than that
obtained by other drilling methods or other types of tools. Cuttings are removed continuously
by the air used to drive the hammer. This arrangement efficiently combines some of the
advantages of the cable tool and rotary rig. Advantages and disadvantages of the air rotary
drilling systems are presented in Table 7.5 (Driscoll, 1989; Lehr et al., 1987; USDI, 1981).
Table 7.5 Advantages and disadvantages of the air rotary drilling systems
1. Penetration rates are high, especially 1. Restricted to consolidated or with semi-

down-the-hole hammers in high resistant consolidated rocks
rocks such as granite, dolomite
2. Cuttings removal is extremely rapid 2. Maintenance costs are high
3. Estimates of the yield from 3. Initial cost of large air compressor is high
a particular aquifer feature penetrated
can be made during the drilling
Drilling depth of rotary air and down-the-hole hammer systems in saturated materials is
limited by the available air pressure which must be greater than that exerted by the column of
water in the hole, otherwise the rig stops working if the rig is to function.
7.6 Auger hole drilling
Although geotechnical and geological exploration has traditionally used augers, their use in the
water industry has been quite limited until recently. Currently, many well contractors bore with
earth augers when installing of groundwater monitoring wells and aquifer clean-up wells.
Although monitoring wells can be drilled by virtually any drilling method, auger techniques are
used to install the overwhelming majority of monitoring wells in the world, because of the
availability and relative ease of collecting formation samples (ASTM, 1992; Driscoll, 1989;
NGWA, 1965; EPA, 1975).
Three principal types are used commonly: (1) hand auger, (2) solid stem auger and
(3) hollow-stem auger (Figures 7.6, 7.7 and 7.8). These earth augers have the important
advantage that drilling fluids and mud are normally not required. Another advantage is the
ability to provide accurate information on all geological materials being drilled, either by
cuttings, split-spoon or barrel samples, and side wall samples.
7.6.1 Hand augers
Hand augers have generally been used to install shallow production or monitoring wells up to
10 metres in depth and with casing diameters of 100 mm or less. A typical hand auger advances
by turning into the soil until the auger is filled (Figure 7.6). The device is then removed and the
sample is dumped from the auger. Motorised units for one or two operators are available.
Generally, the borehole cannot be advanced below the water table for more than one or
two metres, because of the risk of borehole collapse. It is often possible to avert borehole
collapse below the water table by driving a wellpoint into the augered hole, thereby advancing

196
Figure 7.6 Diagram of a hand auger Figure 7.7 Diagram of a solid-stem auger
(After Aller et al., 1989) (After Aller et al.,1989)
the wellpoint below the water table. The wellpoint can then be used to pump water or to
measure water levels and to provide access for water quality samples
Table 7.6 lists the advantages and disadvantages of the auger methods used to install
production and monitoring wells (Aller et al., 1989; ASTM, 1992; Driscoll, 1989; Fetter, 1993;
EPA, 1975; USDI, 1981).
7.6.2 Solid-stem augers
The solid-stem auger is advanced by a rotary drive at the surface and forced downward by a
hydraulic pulldown device (Figure 7.7). Augers are available in diameters up to about 600 mm
(24 in). Where used for monitoring well installation, available auger diameters typically range
from 150 to 350 mm (6 to 14 in) in external diameter.

197
Groundwater studies
Table 7.6 Major advantages and disadvantages of the auger techniques
Type Advantages Disadvantages
1. Good soil sampling 1. Limited to very shallow depths

2. Aquifer zones identification 2. Unable to penetrate below
Hand augers the water table
3. Easy installation of shallow 3. Unable to drill very dense
water wells or monitoring wells formations
1. Good soil sampling with 1. Soil samples limited to areas and depth
split-spoon or thin-wall samplers of predominant stable formations
Solid-stem 2. Able to penetrate below the 2. Unable to work in most
augers water table unconsolidated formations
3. Fast and mobile 3. Depth capacity decreases quickly
as diameter of auger increases
1. Good soil sampling with 1. Can be used only in

split-spoon or thin-wall samplers unconsolidated formations
2. Easy drilling and installation 2. Limited to depths of 30 to 50 m.
in all unconsolidated formations
Hollow-stem 3. Water-quality sampling 3. Possible problem in controlling
augers during drilling heaving sands
4. No drilling fluid is used 4. Limited diameter of augers limits
casing size
5. Usually less expensive than 5. Cost of hollow-stem augers is
rotary or cable tool drilling relatively high
Many of the drilling rigs used for installing monitoring wells in stable unconsolidated
material can reach depths of approximately 40 metres with 350 mm augers and approximately
50 metres with 150 mm augers. As the auger column is rotated into soil, cuttings are retained on
the flights. The auger is removed from the hole and rock formation samples are taken from the
retained earth. (Aller et al., 1989; ASTM, 1992; Campbell and Lehr, 1973; Driscoll, 1989; EPA,
1975).
7.6.3 Hollow-stem augers
The hollow-stem augers are more effective than solid-stem augers because they can be used
as temporary casing to prevent caving-in and abandonment of the borehole (Figure 7.8). The
hollow-stem method is a fast and efficient techniques of drilling and completing small-diameter
wells to moderate depths. Casing pipe and screens can be installed and gravel-packed without
using temporary casing or drilling fluids.
The most widely available hollow-stem augers are 158.8 mm (6.25 in) external diameter
auger flights with hollow stems of 81.25 mm (3,25 in) internal diameter The equipment most
frequently available to power the augers can reach depths of 45 to 50 m in clayey or silty sand
formations. Because of their availability and the relative ease of formation sample collection,
hollow-stem augering techniques are used for installing the overwhelming majority of
monitoring wells in the United States of America.

198
Figure 7.8 Diagram of a hollow-stem auger (After Aller et al., 1989)
7.7. Well design and construction
7.7.1 General
The controlling factors in permanent well installations are usually sanitation, stability, and an
estimated minimum useable well life of 25 years. Selection of the design features of a well and a

199
Groundwater studies
particular method of construction depends upon the wells objective, the quantity of water
desired, economic factors, and hydrogeological conditions. Furthermore, experience has shown
that well design features and construction practices have measurable effects on well perfor-
mance and operating life and on the economic utilisation of the well.
It is obviously not good engineering to design a 200 m3/h well to serve an industry when
3
20 m /h will suffice. It is equally poor practice to choose materials of inferior quality, merely to
cut initial costs. This would only reduce the useful life of the well and increase operation and
maintenance costs (Campbell and Lehr, 1973; Driscoll, 1989; Lehr et al., 1987). Nevertheless,
despite these relationships, the engineering and scientific aspects of well design have received
little attention. As a result of this and other factors, water well design has commonly been based
on the experience, observations, and judgement of the designer and driller.
A number of commonly used design standards that may be of interest to users have been
published by governmental, private or professional organisations, but there are so many that it
is impossible to describe them all here (Ahrens, 1987; Anderson, 1992; ASTM, 1992; Bowman,
1911; Campbell and Lehr, 1973; Driscoll, 1989; EPA, 1975; USDI, 1981).The guidelines presented
below are affected by governmental regulations and by site-specific geotechnical, hydrogeo-
logical, and subsurface chemistry conditions.
7.7.2 Particulars of design
Well design is the process of specifying the drilling techniques, materials and physical dimen-
sions for a production or monitoring well. The design guidelines presented below focus mainly
on the design of municipal, industrial and irrigation wells (Ahrens, 1970; Aller at al., 1989;
Anderson, 1992; ASTM, 1992; Lehr et al., 1987; NGWA, 1965; EPA, 1975).
The principal objectives of good design should ensure the following:
durable and reliable construction,
production of the desired yield or the highest yield available from the aquifer with the
highest efficiency,
accurate water level measurement and/or extraction of representative groundwater
quality samples,
efficient hydrogeological site characterisations.
The basic information desired for the design of efficient production well includes:
aquifer types (water table or confined aquifers),
current depth and trends of the water levels,
stability characteristics and rock formations of the unsaturated zone of the water table
aquifer or at the top of a confined aquifer,
thickness, nature of the porosity (interstitial or resulting from secondary voids), and
the degree of confinement of the aquifer,
grain size and sorting of the water-bearing formations,
available casing and screen materials,
desired yield and pump size patterns,
design and construction features of wells previously drilled in the area,
state and local statutes and regulations,
operation and maintenance history of previously constructed wells (control of
corrosion, encrustation and contamination),
Unfortunately, the information desired is rarely all available, even in a developed area. In
an undeveloped area, all that may be known initially is the approximate location of the
proposed well and the desired yield.
Most frequently, for major wells (yields greater than 20 m3 per hour) the available data
should be supplemented by a hydrogeological investigation and the drilling of a pilot hole. The

200
pilot hole will provide an accurate lithological log, sediment samples for mechanical analyses,
water samples for chemical analyses, and information on static water levels and the type of
aquifer present.
If the site is found inadequate for any reason, the pilot hole can be abandoned without the
major cost of drilling a production well. If the site is found to be satisfactory, a good well design
can be prepared and specifications defined. Furthermore, the contractor can foresee many
problems, minimising the unknown factors, the risks during construction, and even saving the
cost of the pilot hole.
For minor wells, the well drilling costs may be about the same as for a pilot hole; thus, the
latter may be uneconomic. As a result, the recommended procedure for the design of such wells
is to make a preliminary design based on the available information. Additions or changes to the
preliminary design such as in casing diameters and screen slot size and setting can be made on
the basis of information obtained during the drilling (Ahrens, 1970; Campbell and Lehr, 1973;
Driscoll, 1989; Lehr et al., 1987; EPA, 1975).
The particular design of monitoring wells will depend on:
how the well is to be used, whether for taking water samples, for measuring the
elevation of the water table, or for recovering contaminants,
the hydrogeological context,
the chemical reactivates/inertness of the contaminants,
cost of casing and screen materials.
The terminology used for various components of a well is not standardised. Consequently,
various terms are used for similar components in different publications and in different parts of
each country. In the following sections, the terminology is based on the cited bibliography.
7.7.3 Main components of a water well design
Standard design procedures involve choosing the casing diameter and material, estimating well
depth, selecting the length, diameter and material for the screen, determining the screen slot
size, and choosing the completion method. Whether a well is naturally developed or filter
packed depends on the grain-size distribution curve of the water-bearing formation. Coarse-
grained non-homogeneous material can be developed naturally, whereas fine-grained homo-
geneous materials are best developed using a filter pack. Well screen slot openings for either
method are selected from a study of sieve-analysis data for samples representing the aquifer
materials.
Every well consists of four main parts: surface casing, the pump chamber casing, the riser
pipe to conduit water upward from the aquifer intake portion to the pump intake, and the
screen assembly (Figure 7.9 and 7.10).
In consolidated rock aquifers, the borehole length serving as the water conduit and the
intake portion of the well may be left uncased. Some consolidated water-bearing formations,
however, such as sandstone, may deteriorate over time because high flow rates remove the
cement that holds sand grains together, thus causing a slow collapse of the borehole wall. There-
fore, screens are often used to protect pumps from loosened formation particles, and to stabilise
the water-bearing formation in many consolidated rocks, especially sandstone, limestone, and
sometimes granite (Anderson, 1992; Campbell and Lehr, 1973; Driscoll, 1989; Lehr et al., 1987;
Roscoe, 1985; USDI, 1980).
(i) Surface casing
Surface casing is generally installed to facilitate drilling a well by supporting unstable materials
so they will not cave in and fall into the hole, to reduce loss of drilling fluids, and to facilitate

201
Groundwater studies
Figure 7.9 Examples of well construction in unconsolidated (a) and consolidated (b) aquifer formations
installing or pulling back other casing. It is installed from near the ground surface through
unstable, unconsolidated, or fractured rocks and up to a short distance into a firm, stable, or
massive and, if possible, relatively impermeable layer. The surface casing may be temporary and
removed when completing the well, or it may be a permanent part of the structure.
When permanent surface casing is used, the first operation in construction of the well is to drill

202
Figure 7.10 Diagram showing the main parts of a deep water well design
203
Groundwater studies
an oversized hole and to install, centre, and grout the casing pipe. Permanent surface casing is
frequently used to seal out undesirable surface or shallow groundwater. In fact, it provides a
better degree of protection against pollution infiltration, and facilitates placing a sanitary seal.
Table 7.7 gives the recommended minimum and optimum diameters of permanent surface
casing for various well yields, taking into account the two main alternatives, a naturally
developed well or a gravel packed well (AWWA, 1984; Campbell and Lehr, 1973; Driscoll, 1989;
Lehr et al., 1987; USDI, 1981).
Table 7.7 Recommended diameters for pump chamber and permanent surface casing
Surface casing diameter for

Desired Pump chamber naturally packed wells Surface casing
well yield casing diameter developed wells diameter for gravel
(m3/h) mm (inches)* mm (inches) mm (inches)
Up to 20 127152 (56) ID 152203 (68) 406 (16)

20 to 40 152 203 (68) ID 203 254 (810) 457 (18)
30 to 80 203254 (810) ID 254305 (1012) 508 (20)
70 to 160 254 305 (10 12) ID 305356 (1214) 559 (22)
100 to 230 305 356 (12 14) ID 356406 (1416) 610 (24)
180 to 400 356 406 (1416) OD 406457 (1618) 660 (26)
300 to 700 406508 (1620) OD 457559 (1822) 762 (30)
450 to 860 508 610 (2024) OD 559460 (2226) 864 (34)
700 to 1,400 610 762 (24 30) OD 660813 (2632) 1,016 (40)
900 to 1,800 711762 (28 32) OD 762864 (3034) 1,067 (42)
* The size of the pump chamber casing is based on the outer diameter of the bowls for vertical turbine pumps and
on the diameter of either the pump bowls or the motor for submersible pumps. The increasingly sophisticated
technology of the pumping equipment, makes it advisable for the well designer to contact a pump supplier, about
the desired yield, the head conditions and the required pump efficiency.
For nominal diameters ranging between around 203 mm (8 in) and 457 mm (18 in) and to
maximum depths between 10 and 50 metres, the ASA schedule number or class 10 weight pipe
and wall thickness of 0.25 inches is usually adequate for permanent surface casing with welded
joints. For setting depths up to 150 m, and for diameter up to around 813 mm (32 in) the class
20 pipe with wall thickness ranging between 0.35 and 0.5 inches, is generally used. In gravel-
packed wells of the depth and diameter of most water wells, the gravel pack is usually placed
through nominal 51 mm (2 in) to 102 mm (4 in) coupled pipes. To permit insertion of these
pipes, there must be sufficient annular space between the surface casing or the wall of the hole
and pump chamber casing to allow the pipe couplings to pass through (Anderson, 1992;
Driscoll, 1989; Lehr et al., 1987; Koering, 1967; USDI, 1981).
After the surface casing has been installed and the grout has set, the well is drilled deeper
through the bottom of the cased hole. The borehole drilled is usually about 50 mm (2 in) smaller
in diameter than the outside diameter of the surface casing.
(ii) Pump chamber casing
The pump chamber casing furnishes a direct connection between the ground surface and the
aquifer intake part of the well, and when permanent surface casing is not used, it seals out

204
undesirable surface or shallow groundwater and supports the sides of the hole. The design of
the pump chamber casing, the water conduit casing and the screen assembly requires careful
consideration of the hydraulic factors that influence well performance (Ahrens, 1970; Driscoll,
1989; Lehr et al., 1987; Roscoe, 1985; EPA, 1975).
Depth of setting of a pump, hence depth of the pump chamber casing, is determined by
estimating projected pumping levels and considering the following factors:
present static water level and long-period fluctuations of levels. Probable drawdown at
desired yield,
possible interference by other wells or boundary conditions. Required pump submer-
gence,
the presence of any telescoping overlap.
Pump chamber casing should be always installed plumb and straight. Deviation from the
vertical should not exceed two-thirds of the inside diameter of the casing per 30 m of depth
approximately.
The diameter of the pump chamber casing must be large enough to accommodate the
pump required for the desired yield, with enough clearance for installation and efficient oper-
ation. Sometimes, technological and economical constraints may impose several reductions in
size of the pump chamber casing, so that the desired diameter can be set at the required depth
by telescoping. Unless the inner casings must be cut off to accommodate the pump bowls, or the
value of the casing makes retrieval worthwhile, it is a good idea to leave all casing strings in the
hole. Under these conditions one eliminates the need for much of the grouting, the risk of leaks
between casing strings, and the chance that overlapping inner casings will be damaged during
cut-off. If casing is cut off, this should always be done after development of the well (Driscoll,
1989; Lehr et al., 1987; EPA, 1975).
Pump chamber casing should be grouted except when it is run inside a grouted surface
casing. Grout must then be placed in the annulus between casings if the aquifer material will
expand or if the two casings are set in water-bearing formations that have different static water
levels.
(iii) Riser pipe
The riser pipe comprises all casing for water flowing upward from the aquifer portion intake to
the pump intake. In single string construction of uniform diameter, blank pipe may extend from
the top of the screen up to the pump chamber casing to which it may be attached by welding, a
coupling, or a reducer, or the riser pipe may be telescoped into the pump chamber casing for
one or several meters. Where the extension is relatively short, it is sometimes referred to as a
flush tube extension or overlap pipe .
Under these conditions, the diameter of the riser pipe must be sufficient to assure that the
uphole velocity is 1.5 m/sec (5 ft/s) or less. If the pump is set below any screened section, there
will be sufficient area around the bowls to allow water to pass downward with minimum head
loss to the pump intake. However, heat build-up can be a problem for a submersible pump set in
a blank casing section beneath the screen, because the intake portion of the submersible pump is
located above the motor.
In deep wells that have both high static and high pumping water levels, the casing
diameter can be reduced at a depth below the lowest anticipated pump setting, to reduce
drilling and material costs. Furthermore, drilling conditions, drilling methods, or economic
factors sometimes make it necessary to complete the lower portion of the well with smaller
diameter casing or screen. More than one inner casing can be telescoped depending on well
depth. In that case, the casing diameters for individual segments can be chosen from the bottom
of the well upward, to accommodate the driving conditions, the overlap requirements, and the

205
Groundwater studies
annular space necessary for grouting or filter packing. Ideally, each casing string should be
anchored in a poorly permeable bed. If the telescoped casing section ends in sand, water should
be run continuously into the annulus between the two strings. If the casing segments are
anchored in thick clay beds, no grout may be necessary. In the design process, casing diameters
for individual segments must be chosen taking into account that grouting regulations usually
require at least a 51 mm (2 in) annular space (Ahrens, 1970; Driscoll, 1989; Lehr et al., 1987; EPA,
1975; USDI, 1981). This is done in many wells completed in confined aquifers where pressure
is relatively high, such as the case in the huge sedimentary basins in Australia, Brazil, and
Russia (Habermehl, 1985; Margat, 1990; Rebouas, 1976, 1994b). Upward velocity in any smaller
diameter casing beneath the pump bowls should be 1.5 m/sec (5 ft/sec) or less. Table 7.8 lists
maximum discharge rates for various water conduit casing sizes that produce moderate friction
losses (Driscoll, 1989; Lehr et al., 1987; USDI, 1981).
Table 7.8 Maximum discharge rates for optimum diameter of riser pipe; based on an upward velocity
of 1.5 m/sec
Desired Optimum diameter of Optimum diameter

well yield pump chamber casing of riser pipe
(m3/h) mm in mm in
Up to 50 203 8 102 4
70 254 10 127 5
100 305 12 152 6
180 356 14 203 8
280 406 16 254 10
400 438 18 305 12
500 508 20 337 14
650 508 20 387 16
800 610 24 438 18
1,000 711 28 489 20
1,500 762 30 591 24
To meet certain pressure requirements, pipe is manufactured in three general weight

classes: standard, extra heavy, and double extra heavy. Since 1935 the American National
Standards Institute (ANSI) has assigned schedule numbers to classify wall thickness for different
pressure applications. Depending on the wall thickness, however, the inside diameter may be
less than or greater than the number indicated. To complicate matters further, pipe manufac-
turers have developed their own products with specifications that only approximately follow the
design guidelines established by ANSI (Driscoll, 1989; Lehr et al., 1987; USDI, 1981). Under these
conditions, it is difficult to present a simple analysis of the various types of casing used in the
water well industry.
The term nominal is used to designate the inside diameter, because the actual sizes vary
somewhat above or below the standard size. Data presented in Table 7.8 show that in terms of
hydraulic factors, string construction of uniform diameter is usually possible for minor wells.
At the present time, most water well drolleries use pipe constructed to ASTM standards
rather than to API standards. In the huge sedimentary basins, however, pipe used for water
wells is generally constructed to API standards.
Pipe used for constructing water wells is built conforming to major standard spec-
ifications: ASTM standards, and API standards. In general, ASTM standards steel pipe A-53 or

206
A-120 are recommended for most drilling situations and typical water quality conditions. API
standard 5L, either grade A or B, is commonly used for exceptionally corrosive water, for deep
wells, and for difficult drilling conditions, which are characterised by very dense water-bearing
formations and deep boreholes having a large diameter. In every case, the wall thickness must
be sufficient to support full hydraulic loading if the casing is pumped dry (Anderson, 1992;
Driscoll, 1989; Lehr et al., 1987; USDI, 1981).
Western European manufacturers sometimes produce pipe according to other spe-
cifications that is similar, but not identical; in few cases European pipe is made to metric
measurements rather than to dimensions in the Imperial system.
Aggressive corrosion, heavy encrustation, or both, often require the use of non-metallic
pipes. In most cases, thermoplastic pipe such as polyvinyl (PVC), polyethylene, acrylonitrile-
butadiene-styrene (ABS), or rubber modified polystyrene (SR) is the most satisfactory and econ-
omical. Standardisation of thermoplastic well casing is covered under ASTM Standard F480
(ASTM, 1992; AWWA, 1984; NGWA, 1965; EPA, 1975). The major advantages are light weight,
ease of installation, corrosion resistance, and low price. In water wells drilled in consolidated
rock to depths in excess of 250 m, thermoplastic pipes of diameters up to 203 mm (8 in) have
been used successfully to supply industrial plants and domestic water supply systems. How-
ever, these materials lack the tensile and impact strengths. Thermoplastic pipe of suitable wall
thickness and in diameters up to 254 mm (10 in) or 305 mm (12 in) is manufactured for a setting
depth of about 100 metres in unconsolidated water-bearing formations. Fibreglass reinforced
thermoplastic pipe up to 254 mm (10 in) in diameter with 4.6 mm (0.18 in) to 5.1 mm (0.2 in)
wall thickness has been used extensively in some areas in water wells to depths of about 150 m.
However, the conditions are exceptional, and collapse under normal development procedures
has been reported to be a problem (Anderson, 1992; Lehr et al., 1987; NWWA, 1981; USDI 1981).
(iv) Screen assembly
Well screens are required in all unconsolidated and most semi-consolidated formations and
occasionally in consolidated aquifer rock. Three main factors govern the selection of material
used to fabricate well screens: (1) encrustation/corrosion factors, (2) strength requirements of the
screen, and (3) cost factors. The least expensive and most commonly available screens are made
of low carbon steel. Those made from non-ferrous metals and alloys, thermoplastics, and exotic
materials are used in areas of aggressive corrosion and encrustation, to prolong well life and
efficiency (Anderson, 1992; Driscoll, 1989; Lehr et al., 1987; EPA, 1975).
The screen type may range from pipe perforated in place to carefully fabricated cage-type
wire-wound screen with precisely sized openings. Perforated casing made by sawing, machin-
ing, or torch cutting were commonly used until recently. Slot openings usually range from about
0.25 to 5.0 mm; the maximum percentage of open area is about 10 percent for the larger slots.
Another screen type of low efficiency has been manufactured with stamped perforations. Slots
range from 1.5 to 5.0 mm and maximum percentage of open area is about 20 percent. However,
the use of low efficient screen types is rapidly because more efficient screens consisting of a
continuous winding of round or specially shaped wire mounted on a cage of vertical rods are
becoming cheaper. Almost any slot size is readily available from 0.1 to 6.0 mm, usually in
increments of about 0.1 mm. Open areas are the largest obtainable and slot sizes can be closely
matched to aquifer gradations. Although such screens are more expensive initially than other
types of screen, they are usually more economical, especially when used in thin but highly
productive aquifers, and in the deep, confined aquifers (Anderson, 1992; Driscoll, 1989).
The main purposes of a screen assembly are to (1): stabilise the wall of the hole, (2) control
sand entrance into the well, and (3) facilitate flow into and within the well. The screen section
may consist of only a screen length, or of a screen assembly with a combination of slot sizes

207
Groundwater studies
desired, or associating blank casing sections. It is advisable to install a bottom sump consisting
of 2 to 5 metres of blank casing set below the bottom of the lowest screened section. The sump
provides a storage space for sand brought into the well which settles to the bottom and prolongs
the effective operation of the total screen length. Screen assemblies set in unconsolidated
materials should have a bottom seal or a concrete plug. The seal may consist of a steel plate
welded or coupled to the bottom. This seal not only precludes materials heaving up into the
well under certain circumstances but also provides a bearing area for support of the screen
assembly (Ahrens, 1970; Driscoll, 1989; Lehr et al., 1987; EPA, 1975; USDI, 1981).
Currently, most screens are made in lengths ranging from 1.5 m to 6.0 m which can be
jointed by welding or couplings to give the desired length of screen. To avoid electrolytic
corrosion, couplings and welding materials should be composed of the same materials as the
screen.
The optimum length of well screen is based on the thickness of the aquifer, available
drawdown, and nature of the stratification of the water-bearing formation. In virtually all
aquifers, certain zones will be more productive than others. Thus, the screened sections of the
well must be placed in front of the zones that have the highest hydraulic conductivity. These
layers can be selected on the basis of an interpretation of the following information:
the lithological well log and driller comments on drilling operation such as fluid loss,
penetration rate, etc.;
the geophysical borehole logs such as calliper, resistively, spontaneous potential,
natural gamma ray radiation, gamma-gamma ray radiation, and neutron logging;
grain size distribution curve of the samples of water-bearing formations;
hydrogeological site evaluation.
In uniform unconfined aquifers a continuous length of screen is usually installed. Theoretical
considerations and experience have shown that screening of the bottom one-third to one-half of
an aquifer provides the optimum design for aquifers less than 50 m thick. Where aquifers are
deeper and thicker, judgement should be used to determine the most economical combination of
penetration depth and screened section length. In this case the well screen is usually placed at
the bottom of the well and its length should cover not less than 35 percent of the estimated
thickness of the penetrated part of the aquifer (Driscoll, 1989).
For non-homogeneous unconfined aquifers, the screen sections must be positioned in
front of the most permeable layers of the lower portions of the water-bearing formation,
separated by blank pipes opposite the poor aquifer materials. If possible, the total screen length
should cover not less than 35 percent of the aquifer layers penetrated by the well, so that
maximum drawdown is available.
In confined aquifer conditions, theoretical considerations and experience have shown that
full penetration and maximum percentage of screen are recommended where aquifer depth and
thickness make such construction economically feasible. Usually, 80 percent of the thickness of
the homogeneous and uniform water-bearing sediment, or 80 percent of the most permeable
layers of the non-homogeneous sediment should be screened to obtain between 90 and 95 per-
cent of the total specific capacity of the well that could be obtained by screening the entire
thickness of the aquifer. An exception to this rule is made when poor-quality water is found in
part of an aquifer. In this case, the well should be completed to a depth that will avoid the
undesirable water. Any part of the hole drilled into a portion of the aquifer containing poorer
quality water should be isolated.
Pump chamber casing sometimes has perforated or screened sections in it, above the
pump bowls. This practice should be avoided, because if drawdown increases to depths below
the screened section, cascading results in air getting into the water which induces cavitation and
may result in other adverse effects such as the inflow of undesirable shallow groundwater. Also,

208
periodic exposure of a screen to the atmosphere may contribute to corrosion and/or to growth of
organisms such as iron bacteria.
Screens are available in diameters ranging from 32 mm (1 1/4 in) to 1,524 mm (60 in).
Screen diameter should be selected on the basis of the desired yield from the well and thickness
of the aquifer. Taking into account that the entrance velocity of the water generally must not
exceed the design standard of 0.03 m/sec (0.1 ft/sec), the screen diameter can be adjusted within
narrow limits after length of the screen and size of the screen openings have been selected in
accordance with the thickness of the aquifer and with the gradation of the sediments or the size
of the filter pack, respectively.
In a naturally developed well, the usual approach is to select a slot screen which will
allow 5040 percent of the material to pass, retaining 5040 percent. This practice results in
creating a zone of graded grain distribution extending over a distance of 0.3 to 0.6 m near the
borehole. The finer formation materials brought into the screen are pumped from the well
during development.
In filter-packed wells, the annular space between the screen and the borehole wall is filled
with specially graded material. The principal functions of a gravel pack are: (1) to stabilise the
aquifer formation and minimise sand pumping, (2) to permit use of the largest possible screen
slot with resultant maximum open area, and (3) assure good hydraulic conductivity, thus
increasing the effective radius of the well and the yield (Anderson, 1992; Driscoll, 1989; Lehr et
al., 1987; Roscoe, 1985; USDI, 1981).
The average entrance velocity is calculated by dividing the desired well yield by the total
area of the screen openings. If the velocity is greater than 0.03 m/sec, the screen length and/or
diameter should be increased to provide enough open area so that the entrance velocity is
0.03 m/sec or less. Table 7.9 gives open areas in square inches per foot of length of some
representative screens of various slot sizes and diameter (Driscoll, 1989; Lehr et al., 1987; USDI,
1981).
The transmitting capacity of a well screen section, expressed as US gallons per minute
(gpm)/ft of screen (0.7 m3/h per metre screen) at the recommended velocity of 0.1 ft/s (0.03 m/s),
Table 7.9 Open areas for some representative screens
Screen Slot Continuous Bridge Mill Plastic Plastic

diameter * size slot slot slotted continuous slotted
mm (in) 10 3 in in/ft % in/ft % in/ft % in/ft % in/ft %
102 ID 20 44 25 22 13
(4) 60 90 52 19 12 8 5 52 30 18 11
203 ID 30 80 25 57 18 26 8
(8) 60 135 41 17 6 15 5 93 29 47 14
95 165 51 23 7
305 ID 30 77 16 12 3
(12) 60 135 28 33 7 21 5 52 11
95 182 38 32 7
125 214 45 68 14 43 9
406 OD 30 97 16 16 3 52 9
(16) 60 169 28 35 6 27 5
95 228 38 41 7
125 268 45 78 13 55 9
* For example, an 8-in continuous-slot well screen has slot openings of 0.030, 0.060 or 0.095 in.

209
Groundwater studies
is calculated readily from the open area figures shown in Table 7.9. Multiplying the number of
square inches of open area by a conversion factor of 0.31 gives the transmitting capacity at a
velocity of 0.03 m/s (0.1 ft/s). The unit conversion factor of 0.31 results from specifying an
entrance velocity, V, of 0.03 m/s (0.1 ft/s) in the equation Q = VA. For example, the transmitting
velocity of a 12-in diameter screen may vary from 77 x 0.31 = 24 gpm/ft up to 214 x 0.31 =
66 gpm/ft, or 16 m3/h per metre up to 45 m3/h per metre.
7.8 Well design in deep confined aquifer systems
Special well designs have been developed in certain areas because of particular hydrogeological
conditions. These advanced methods include the type of drilling equipment used, the filter pack
injection techniques, and economic aspects of the wells. The technologies and procedures have
been developed because more water is needed than could obtained with standard design criteria
and because favourable cost/benefit ratios have been realised. Although developed locally, these
technologies can be used successfully in areas with similar social, economic and hydrogeological
conditions. Several examples of alternative well designs are described below.
7.8.1 Hydrogeological conditions
Underlying most of the developed region of South America, covering about 1,000,000 km2 in
Brazil, 100,000 km2 in eastern Paraguay, 100,000 km2 in north-western Uruguay and 400,000 km2
in northeastern Argentina is the Paran Sedimentary Basin (Rebouas, 1976, 1994b). The basin is
up to 5,000 m thick, and forms a large intracratonic structure uplifted and exposed along its
eastern margin and tilted southwest (Figure 7. 11).
The sedimentary sequence from the Silurian to Cretaceous is almost undisturbed, com-
prising a multi-layered aquifer system with gentle dips towards the centre of the basin. Regional
and local faults served as channels for extruding basalt flows during the Cretaceous period.
Currently the basaltic package extends over about 1,000,000 km2 and has a maximum thickness
of up to 2,000 m in the centre of the sedimentary basin. Groundwater occurs within the interflow
zones and along cooling joints; interflow sediment deposits greatly increase these local and
occasional basalt water-bearing conditions. Most commonly the vertical permeability is very
small in comparison to the horizontal permeability; indeed the basalt flow package is the
hydrogeological substratum for groundwater stored in the overlying Cretaceous sandstone
formations.
In the Paleozoic sequence, with aquifers occurring in quartzose sandstones of continental
and marine origin, and confining beds of siltstone and marine argillaceous sediments, ground-
water usually has a poor quality.
More important as a source of good quality groundwater is the Botucatu aquifer system
which includes the Triassic deposits (Piramboia and Rosario do Sul formations in Brazil and
Buena Vista in Uruguay) and Jurassic sandstones accumulated by aeolian processes under desert
conditions (Botucatu formation in Brazil, Misiones formation in Paraguay, and Tacuaremb
formation in Uruguay and Argentina). This aquifer system covers about 1,200,000 km2
(840,000 km2 in Brazil, around 70,000 km2 in Paraguay, 230,000 km2 in Argentina, and 60,000 km2
in Uruguay). As a result of the differential subsidence, the total thickness of both the fluvial-
lacustrine and aeolian deposits varies widely; from more than 800 m near the southwestern
Brazilian border to complete absence in some limited areas in the subsurface. But in general
terms, thicknesses of more than 500 m tend to prevail along a NNE-SSW axis which is followed
by the Paran and Uruguay rivers today. The Botucatu aquifer system is confined by the
Cretaceous basalts of the Serra Geral Formation and by underlying Permo-Triassic deposits of

210
Figure 7.11 Fence diagram of the Paran sedimentary basin (Rebouas, 1994b)
211
Groundwater studies
low permeability resulting in artesian conditions over about 70 per cent of the total area it
underlies. Jurassic aeolian sandstones form the best part of the aquifer system (hydraulic
conductivity ranging between 0.3 to 5.0 m/day), whereas more argillaceous fluvial-lacustrine
Triassic deposits are notably inferior, with hydraulic conductivity values an order of magnitude
less. Not surprisingly in view of the great depths it reaches (up to 3,000 m) and the thick con-
fining basaltic cover, the water stored in the Botucatu aquifer has temperatures exceeding 55C
in its deepest parts.
At the end of the Cretaceous the present structural configuration was largely established
and, since then, the aquifer has almost everywhere been flushed by infiltrated rain water.
Natural direct recharge occurs in the narrow eastern and western uplifted marginal unconfined
zone. In the confined context indirect recharges and discharges may occur through the basaltic
package, in view of the hydraulic head balances. Groundwater flow patterns and the continuing
recharge from geological to modern times have been confirmed by environmental isotope and
hydrochemical studies.
Throughout the vast area of this aquifer system, most of its water is potable. Few
exceptions occur where water quality is affected by enrichment with fluorine. The highest values
reaching 12 mg/l are related to areas of stagnant water or to the lacustrine deposits of the
Piramboia formation. As a result, it is frequently recommended that wells penetrate only
partially, because high fluorine contents may occur in the lowest sections of the aquifer system,
and because there may be some flow of low quality water from the basal Permo-Triassic
sequence into the Botucatu aquifer system, which may be increased by the pumping effects
(Rebouas, 1976, 1991, 1994a, 1994b).
7.8.2 Typical well drilling procedures
Despite being situated in a region with abundant surface water resources, the Botucatu aquifer is
increasing in importance as a source of regional water supplies, partly as a response to the
growing costs and other constraints in storing and treating surface water and partly because the
advantages of groundwater are now better understood. Currently, about 1,000 wells pump
groundwater from the Botucatu aquifer system for domestic water supply, industries and
thermal water recreational clubs.
The typical drilling problems result from the accumulation of particles in the fluid system
and at the bottom of the borehole, slumping or expansion of active clays which occurs at the
interflow zones of the thick basaltic package, and caving-in of the inter-basalt sand deposits.
Drilling efficiency depends largely on the drillers experience, drilling fluid control, and well
completion steps that are taken during installation of well screens, or immediately thereafter,
including filter pack installation, grouting the casing, and developing the well. A good filter
packing installation is especially important because the aeolian sediments of the Botucatu
Formation are very uniform and fine grained, and because the small slot size dictated by natural
development limits the transmitting capacity of the screen, so that the desired yield cannot be
obtained. In view of the depths of the wells (up to 2,000 m) and the high artesian pressure, filter
packing installation is a laborious and highly costly operation. Figure 7.12 shows various well
designs.
A typical procedure for constructing a high capacity well in the Botucatu aquifer system
includes the following steps:
drill a pilot hole at the chosen site, keeping a detailed drilling log,
conduct geophysical logs (typically SP, resistivity, natural gamma ray, and calliper) in
the pilot hole to determine the physical features of the basaltic package, the presence of
interflow sandstones, and the hydrogeological characteristics of the Botucatu sand-
stone. Depending on the physical features of the basalt section between the bottom of

212
Figure 7.12 Typical water well designs to tap the confined Botucatu aquifer system in Paran basin, Brazil
213
Groundwater studies
the recommended pump chamber casing and the top of the aquifer the well will be
either left uncased or lined,
after the well design has been chosen from analysis of the drillers logs, and geo-
physical logs, an open hole is drilled to receive the surface casing which is placed from
near the ground surface, through the unstable Bauru sandstone deposits and a short
distance into a firm and impermeable basalt, and grouted,
at least 24 hours thereafter the drilling should be continued with a diameter suitable to
receive the pump chamber casing which is set in the hole and grouted,
the borehole section between the bottom of the pump chamber casing and the top of
the aquifer is then drilled according the recommended diameters, to reach the desired
well yield and/or to accommodate the proper riser pipes,
to accommodate a proper thickness of filter pack between the screen and formation,
the drilled thickness of the aquifer is under-reamed to obtain a 76 to 203 mm (3 to 8 in)
annulus space,
occasionally, a riser pipe assembly is telescoped into the open hole, between the
bottom of the pump chamber casing and the top of the aquifer, and then grouted,
the screen assembly is then telescoped through the casing into the under-reamed
aquifer section, and the filter pack is introduced into the well by reverse circulation
methods,
after the filter pack has been installed, development work is continued to remove fine
sediment from the filter pack and to clean the contact surface between the filter pack
and the water-bearing formation.
In view of the varied diagenetic processes affecting the Botucatu water-bearing formation,
the specific capacity of the wells ranges from 5 to 25 m3/h per metre of water-level drawdown.
Currently, the production cost per cubic metre of water varies from US$ 0.10 to US$ 0.50,
assuming a discharge of about 500 m3/hour and a pumping regime of 16 hours per day.
AHRENS, T. P., 1970. Basic Considerations of Well Design. Water Well Journal (United States of
America), Vol. 24, Nos. 4, 5, 6 and 8.
ALLER, L.; BENNETT; T. W.; HACKETT; G.; PETTY, R. J.; LEHR, J. H.; SEDORIS, H.; NIELSEN, D. M., 1989.
Handbook of Suggested Practices for the Design and Installation of Ground-Water Monitoring
Wells. National Water Well Association, Ohio, 380 pp.
ANDERSON, K., 1992. Ground Water Handbook. 401 pp.
ASTM, 1992. Standards on Ground Water and Vadose Zone Investigations. American Society for
Testing and Materials, 166 pp.
AWWA, 1984. Standard for Water Wells. American Water Works Association, Denver, CO.,
A 10084, 75 pp.
BOWMAN, I., 1911. Well-drilling methods. U.S. Geol. Survey, Water Supply Paper 257, pp. 2330.
CAMPBELL, M. D.; LEHR, J. H., 1973. Water Well Technology. McGraw-Hill, New York, 681 pp.
CARPENTER, D., 1994. Wellhead Protection Anthology. NGIC, 13 papers.
COHEN, PH., 1985. Ground-Water Development in the United States of America. Proc. IAH 18th
Congress, Vol. XVIII, Part 1, Cambridge, United Kingdom, pp. 1730.
DAO, 1965. Technical Manual 5297, Well Drilling. Department of the Army Operations,
Washington, D.C.
DRISCOLL, F. G., 1989. Ground Water and Wells. Johnson Filtration Systems Inc., Second Edition,
United States of America, 1089 pp.

214
EPA, 1975. Manual of Water Well Construction Practices. EPA 570/9-75-001, U.S. Environmental
Protection Agency, Office of Water Supply, Washington, D.C., 156 pp.
, 1980. Proposed Ground Water Protection Strategy. U.S. Environmental Protection Agency,
Office of Drinking Water, Washington, D.C., 540 pp.
FETTER, C. V., 1993. Contaminant Hydrogeology. Macmillan, United States of America, 458 pp.
GORDON, W. R., 1958. Water Well Drilling with Cable Tools. Bucyrus-Erie Co., United States of
America, 256 pp.
HABERMEHL, M. A., 1985. Groundwater in Australia. Proc. IAH 18th Congress, Vol. XVIII, Part 1.
Cambridge, United Kingdom, pp. 31 52.
HARDCASTLE, B. J., 1987. Wells Ancient and Modern; An Historical Review. Quarterly Jour. of Eng.
Geol., pp. 231238.
KOERING, 1967. Well Drilling Manual. Speed Star Division of Koering Co., Enid, Okla.
LEHR, J.; HURLBURT, S.; GALLAGHER, B.; VOYTEK, J., 1987. Design and Construction of Water Wells; A
Guide for Engineers, NGWA, United States of America, 256 pp.
MARGAT, J., 1990. Les gisements deau souterraine. La Recherche, No. 221, Paris, pp. 5906.
MEINZER, O. E., 1934. The History and Development of Groundwater Hydrology. Jour.
Washington Acad. Sci., Vol. 24, pp. 623.
NGWA, 1965. Well Drilling Operations by U.S. Dept. of Army/Air Force. National Groundwater
Association, United States of America, 249 pp.
, 1985. Water Well Drillers, Beginners Training Manual. National Water Well Association,
Columbus, Ohio, 245 pp.
NWWA AND PPI, 1981. Manual on the Selection and Installation of Thermoplastic Water Well Casing.
National Water Wells Association and Plastics Pipe Institute, Worthington, Ohio, 64 pp.
OECD, 1989. Water Resource Management, Integrated Policies. Organisation for Economic Co-
operation and Development, Paris, 199 pp.
REBOUAS, A. DA C., 1976. Recursos hdricos subterrneos da Bacia do Paran; anlise de pr-viabilidade.
USPO Tese de Livre Docncia, So Paulo, 143 pp.
, 1988. Groundwater in Brazil, Episodes, Vol. 11, No. 3, pp. 20914.
, 1991. Desarrollo y tendencias de la hidrogeologa en Amrica Latina. 25th Anniversary of the
Curso Internacional de Hidrologa Subterrnea, Barcelona, pp. 42953.
, 1994a. Gesto sustentvel dos grandes aquferos. Anais 8 Cong. Bras. guas Subterrneas,
ABAS-Recife, Brazil, pp. 1319.
, 1994b. Sistema aqufero Botucatu no Brazil. Anais 8 Cong. Bras. guas Subterrneas,
ABASRecife, Brazil, pp. 5009.
ROSCOE, M., 1985. The Engineers Manual for Water Well Design. NGWA, United States of America,
242 pp.
SMITH, A. J., 1980. Design as the Key to a Good Recharge Well. Johnson Drillerss Jour. Johnson
Division, MN, United States of America.
SOLLEY, W. B.; PIERCE, R. R.; PERLMAN, H. A., 1993. Estimated Use of Water in the United States in
1990. U. S. Geological Survey Circular 1081.
TOLMAN, C. F., 1937. Ground Water. McGraw-Hill Book Co., New York, 593 pp.
UNESCO, 1992. Ground Water. United Nations Educational, Scientific and Cultural Organization,
Paris.
USDI, 1981. Ground Water Manual. US Departement of the Interior, Water and Power Resources
Service, A Wiley-lnterscience Pub., 480 pp.
WRIGHT, E. P.; HERBERT, R.; KITCHING, R.; MURRAY, K. H., 1987. Collector Wells in Crystalline
Basement Aquifers; Review of Results of Recent Research. Proc. XXl Congress of IAH, Rome,
Italy, pp. 127.

215
D e t e r m i n i n g h y d ro d y n a m i c
8 and contaminant transfer
p a r a m e t e r s o f g ro u n d w a t e r f l o w
8.1 Introduction
This chapter briefly describes of the methods to determine and quantify the hydraulic charac-
teristics of homogeneous and isotropic aquifers by pumping tests (section 8.3 and 8.4) and by
other methods (section 8.5) and to determine the hydraulic characteristics of fractured aquifers,
volcanic rocks and limestones (sections 8.6, 8.7, and 8.8). Diffusion, dispersion and macrodis-
persion are discussed in section 8.9.
8.2 Parameters of flow in homogeneous aquifers
In the following sections a broad distinction will be made between the aquifers with primary
porosity and those with secondary porosity. Primary porosity is porosity related to the sedi-
mentation process and hence related to the diameter of the pores and their distribution in the
original sedimentary deposits, or it is related to the formation of igneous and volcanic rocks.
Secondary porosity is related to post sedimentary or post formation processes, like folding,
fracturing, faulting and shrinkage; a particular group of hard rock aquifers are those in karstic
limestone in which solution porosity may play a major role.
The commonly used parameters of fresh water flow (see also Section 1.4) are:
the hydraulic conductivity K [L1 T 1];
the transmissivity T = KM [L2 T 1], where M is the saturated thickness of the aquifer;
the transmissivity for vertical flow through an aquitard Tv = Kv/M [T1]; the reciprocal
of this value is known as the hydraulic resistance of the aquitard: c [T];
the storage coefficient of a confined aquifer SA [dimensionless],
the specific yield of an unconfined aquifer SY [dimensionless].
8.3 Determination of hydraulic characteristics by pumping tests
8.3.1 The pumping test
Pumping tests are a commonly used method to determine the transmissivity (and, if the thick-
ness of the aquifer is known, the hydraulic conductivity) and the storage coefficient of an
aquifer. The hydraulic resistance of a semi-confining layer of a leaky aquifer can also be cal-
culated from pumping test data.
The principle of a pumping test is that if we pump water from a well and measure the
discharge of the well and the resulting drawdown in the well and in piezometers at known

217
Groundwater studies
distances from the well and at a specific time, we can substitute these measurements into an
appropriate well flow equation and calculate the hydraulic characteristics of the aquifer
(Fig. 8.1).
Figure 8.1 Drawdown in a pumped confined aquifer
Before a pumping test is conducted, the following geological and hydrological infor-
mation should be known:
the geological characteristics of the subsurface (i.e. all the lithological, stratigraphical
and structural features that may influence the flow of groundwater);
the type of aquifer and the presence or absence of horizontal recharge boundaries, e.g.
where percolating rain or irrigation water causes the water table of an unconfined
aquifer to rise, or where water passes through an aquitard and recharges a semi-con-
fined (leaky) aquifer;
the thickness and lateral extent of the aquifer and confining beds as found by geo-
physical surveys or from borehole data;
the boundary conditions. The aquifer may be bounded laterally by barrier boundaries
of impermeable material (e.g. the bedrock sides of a buried valley, a fault, or simply
lateral changes in the lithology of the aquifer material). Analogous effects may occur
when the aquifer has one or more recharge boundaries (e.g. a deeply incised perennial
river or canal, a lake, or the ocean). These conditions have to be taken into account if
the drawdown cone around a pumping test reaches one or more lateral boundaries;
data on the groundwater flow system (e.g. water table gradients and regional trends in
groundwater levels);

218
Determining hydrodynamic and contaminant transfer parameters of groundwater flow
Any existing wells in the area. From the logs of these wells, it may be possible to
derive approximate values of the aquifers transmissivity and storativity and their
spatial variation. It may even be possible to use one of these wells for the test, thereby
reducing the cost of fieldwork. Often, however, such a well produces unreliable results
because no details are available on its construction and condition.
8.3.2 The well
The pumping test site should be selected taking into account the following points:
The hydrogeological conditions should be representative of most, if not all of the area;
The site should be away from railways and motorways where passing trains or heavy
traffic might produce measurable fluctuations in the hydraulic head of a confined
aquifer;
The site should not be in the vicinity of existing discharging wells;
The water produced during the test should be discharged in a way that prevents its
return to the aquifer;
The gradient of the water table or piezometric surface should be small;
Manpower and equipment must be able to reach the site easily;
The well should in principle be drilled to the bottom of the aquifer and be screened
over 80 percent of the aquifer thickness This makes it possible to obtain about 90 per-
cent of the maximum yield. Moreover with this screen length the groundwater flow to
the well can be assumed to be horizontal, an assumption that underlies almost all well
flow equations (Fig. 8.2). Exceptions to these rules are:
In unconfined aquifers, it is common practice to screen only the lower half or lower
one third of the aquifer because, if appreciable drawdown occur, the upper part of a
longer well screen would fall dry;
Figure 8.2 Well penetration: a) a fully penetrating well; b) a partially penetrating well
219
Groundwater studies
In a very thick aquifer it will be obvious that the length of the screen will have to be
less than 80 per cent of the aquifer thickness. Such a well is said to be partially
penetrating. It induces vertical-flow components, which can extend outwards from
the well to distances roughly equal to 1.5 times the thickness of the aquifer
(Fig. 8.2).
After the well has been drilled, screened and gravel-packed, as necessary, a pump and
power unit are installed. They should be capable of operating continuously at a constant dis-
charge for a period of at least a few days. An even longer period may be required for uncon-
fined or leaky aquifers and especially for fractured aquifers. The same applies if drawdown data
from piezometers at great distances from the well are to be analysed. The capacity of the pump
and the rate of discharge should be high enough to produce good measurable drawdown in
piezometers as far away as, say, 100 or 200 m from the well, depending on the aquifer
conditions.
8.3.3 The piezometers
A piezometer is an open-ended pipe installed in a borehole that has been drilled to the desired
depth below surface. The bottom tip of the piezometer is fitted with a perforated or slotted
screen of 0.5 to 1 m length to allow the inflow of water. The annular space around the screen
should be filled with a gravel or uniform coarse sand to facilitate the inflow of water. The rest
of the annular space can be filled with any material available, except where the presence of
aquitards requires a seal of bentonite clay, very fine sandy clay or cement grouting to prevent
leakage along the pipe. The water level measured in the piezometer represents the average head
at the screen of the piezometer.
The number of piezometers to be installed depends on the amount of information needed,
and especially on its required accuracy, but also on the funds available. The distance at which
piezometers should be installed depends on the type of aquifer, its transmissivity, the duration
of pumping, the discharge rate, the penetration rate of the screen, and whether the aquifer is
stratified (Table 8.1).
The depth at which a piezometer is installed depends on the aquifer type and the
homogeneity of the aquifer. In an isotropic and homogeneous aquifer the piezometer should be
Table 8.1 Preferred distance between pumping well and piezometer in relation to characteristics
of the pumping test
Distance Near <> Far
Type of aquifer Unconfined Confined

Transmissivity Low High
Pumping time Short Long
Discharge rate Low High
Penetration rate Full Partial 1
Stratification Little 2 Strong
Notes:
1. The drawdown measured at a distance less than 1.5 times the thickness of the aquifer must be corrected for the
influence of the vertical flow components close to the well.
2. As a consequence of the differences in transmissivity at different depths, the drawdown observed close to the well may
differ at different depths within the aquifer. With increasing pumping time and increasing distance to the well, the
effect of stratification upon the drawdown diminishes.

220
Figure 8.3 Cluster of piezometers in a heterogeneous aquifer intercalated with aquitards
installed at a depth that is half the length of the well screen. For aquifers made up of sandy
deposits with intercalations of less pervious layers it is recommended to install a cluster of
piezometers (Fig. 8.3). If the aquifer is overlain by a partly saturated aquitard (leaky aquifer)
piezometers should also be installed in the aquitard to check whether the water table is affected
when the underlying aquifer is pumped.
8.3.4 The measurements
During a pumping test, measurements are taken of the water levels in the well and the piezo-
meters, of the discharge rate, and of the quality (e.g. the Electrical Conductivity) of the pumped
water.
Measurements of the water levels (e.g. twice a day) should start a few days before the
start of the pumping test, to establish the natural changes in hydraulic head in the aquifer. This
applies to both long term trends as well as to short term variations (e.g. tidal movements in
coastal aquifers, or day and night variations). When a test is expected to last one or more days,
measurements should also be made of the atmospheric pressure, the levels of nearby surface
waters, and of any precipitation or irrigation. Try to avoid the disturbing influence of nearby
production wells.
During the test, the water levels in the well and in the piezometers must be measured
many times. Because the water levels fall rapidly during the first one or two hours of the test,

221
Groundwater studies
the readings should be made at frequent intervals. Tables 8.2 and 8.3 give ranges of readings in
the well and in the piezometers close to the well, respectively.
For piezometers further from the well and for those in confining layers above or below
the aquifer, the intervals in the first minutes need not be so short.
After the pump has been shut down, the water levels in the well and the piezometers will
start to rise; rapidly in the first hour, but more slowly afterwards. The rising water levels are
measured as residual drawdown during a period that is as long as the pumping period. These
data can be used to check the calculations made on the basis of the date collected during
the pumping period. The schedule for the recovery measurements should be the same as that
adhered to during the pumping period (Tables 8.2 and 8.3). The most accurate recordings of
water level changes are made with fully-automatic microcomputer-controlled systems that use
pressure transducers or acoustic transducers for digitised recordings; the data are automatically
stored on micromemory from which they can be retrieved by a laptop computer. A good
alternative is the conventional mechanical recorder, which produces a continuous recording of
water-level changes. Fairly accurate water-level measurements can be taken by hand, using e.g.
an electrical sounder. For detailed descriptions of automatic recorders see e.g. Driscoll (1986),
Genetier (1984), Groundwater Manual (1981).
Table 8.2 Frequency of water level measurements in the pumped well
Time since start of pumping Time intervals
05 minutes 0.5 minutes

560 minutes 5 minutes
120 minutes shutdown of the pump 60 minutes
Table 8.3 Frequency of water level measurements in piezometers
Time since start of pumping Time intervals
02 minutes approx. 10 seconds

25 minutes 30 seconds
515 minutes 1 minute
100 minutes5 hours 30 minutes
548 hours 60 minutes
48 hours shutdown of the pump 3 times a day
The discharge rate can be measured with a commercial water meter of appropriate
capacity. If the water is discharged through a small ditch, a flume can be used to measure the
discharge. Other fairly simple methods are: container of known capacity in combination with a
chronometer, orifice, orifice bucket or jet-stream method. For details and tables see Driscoll (1986).

222
8.4 Analysis of pumping test data on homogeneous and isotropic aquifers
8.4.1 Data analysis
It is relatively easy to calculate hydraulic characteristics if the aquifer system (i.e. aquifer plus
well) are precisely known. This is generally not the case, so interpreting a pumping test is
primarily a matter of identifying an unknown system. System identification includes the con-
struction of diagnostic plots and specialised plots. Diagnostic plots are log-log plots of the
drawdown versus time since pumping started. Specialised plots are semi-log plots of drawdown
versus time, or drawdown versus distance to the well; they are specific for a given flow regime.
Both plots must be constructed, because the diagnostic value lies in the typical com-
bination of the log-log and semi-log plots (Fig. 8.4). The choice of a theoretical model is a crucial
step in the interpretation of pumping tests. If the wrong model is chosen, the hydraulic charac-
teristics calculated for the real aquifer will not be correct. Unfortunately the theoretical solutions
of well flow problems are not unique. Some models developed for different aquifer systems,
yield similar responses when required to handle a given stress. This means that besides the log-
log and semi-log plots of the drawdown versus time, all other relevant hydrogeological infor-
mation, e.g. lithology, boundary conditions, should be taken into account.
8.4.2 The well flow formula for confined aquifers
Theis (1935) noted that when a well penetrating an extensive confined aquifer is pumped at a
constant rate, the influence of the discharge extends outward with time. The rate of decline in
head, multiplied by the storativity and summed over the area of influence, equals the discharge.
Theis equation, which was derived from the analogy between the flow of groundwater and the
conduction of heat, is written as:
Q
s = W(u) (8-1)
4 KM
where:
s = the drawdown in m measured in a piezometer at a distance r in m from the well;

Q = the constant well discharge in m3/d;
KM = the transmissivity of the aquifer in m2/d;
W(u) = is a particular exponential integral, with u as the argument, which in this usage is
read as the well function of u.
r2 S 4KMtu
u = and consequently S = (8-2)
4K Mt r2
S = the storage coefficient;

t = the time since pumping started.
W(u) can be written as a series:
u2 u3 u4
W(u) = 0.5772 ln u + u + + ... (8-3)
2.2! 3.3! 4.4!

223
Groundwater studies
Figure 8.4 Log-log and semi-log curves of drawdown versus time; A and A: confined aquifer,
B and B: unconfined aquifer, C and C: leaky (or semi-confined) aquifer;
D and D: effect of partial penetration, E and E: effect of well-bore storage (large diameter well),
F and F: effect of recharge boundary, G and G: effects of an impervious boundary

224
In fact the equation of Theis describes the flow to a well in an aquifer system with the following
characteristics:
a confined aquifer of seemingly infinite areal extent that is homogeneous and isotropic,
and uniformly thick over the area influenced by the test;
a piezometric surface that is horizontal (or nearly so) prior to pumping, over the area
that will be influenced by the test;
the aquifer is pumped at a constant discharge rate;
the well penetrates the entire thickness of the aquifer and thus receives water by
horizontal flow;
the water removed from storage is discharged instantaneously with decline of head;
the diameter of the well is small, i.e. the storage in the well can be ignored.
From Eq. (8-1), it will be seen that, if s can be measured for one or more values of r and
for several values of t, and if the well discharge Q is known, S and KM can be determined.
8.4.3 Theiss curve-fitting method
The presence of the two unknowns and the nature of the exponential integral in Theis equation
makes it impossible to effect an explicit solution; however, a solution can be found by a curve
fitting method. Eq. (8-1) can be written as:
log s = log (Q/4KM) + log (W(u))
and Eq. (8-2) as
log (t/r2) = log 4KM/S + log (1/u)
Since Q/4KM and 4KM/S are constants, the relation between log s and log (t/r2) must be
similar to the relation between log W(u) and log (1/u). Values for W(u) as 1/u varies have been
tabulated and are widely published, e.g. by Kruseman et al. (1991).
The curve fitting method is based on the fact that if s is plotted versus t/r2 and W(u) versus
1/u on the same log-log paper the resulting curves (the data curve and the type curve,
respectively) will be of the same shape, but will be horizontally and vertically offset by the
constants Q/4KM and 4KM/S. The two curves can be made to match, and the coordinates of an
arbitrary matching point A are the related values of s, t/r2, 1/u, and W(u), which can be used to
calculate KM and S with Eqs 8-1 and 8-2, respectively (Fig. 8.5).
8.4.4 Jacobs straight line method
Jacobs method is based on the Theis formula (Eq. 8-3):
Q u2 u3
s = ( 0.5772 ln u + u + ....)
4KM 2.2! 3.3!
From u = r2S/4KMt it follows that u decreases as the time increases and the distance from the
well r decreases. Accordingly, for drawdown observations made in the near vicinity of the well
after a sufficiently long pumping time, the terms beyond ln u in the series become so small that
they can be ignored. So, for small values of u (u<0.01), the drawdown can be approximated by:

225
Groundwater studies
Figure 8.5 Analysis of data from pumping test (Q = 788 m3/d) in Oude Korendijk (The Netherlands)
with Theiss curve-fitting method
Note:
For point A: W(u) = 1 and i/u = 10 while s = 0.16 m and t/r 2 = 1.5 x 10 3 min/m2 = 1.5 x 103/1440 d/m2. Introducing
these values and the value of Q into equations 8-1 and 8-2 yields: K M = (Q/4 s)W(u) = = 788/(4 x 3.14 x 0.16) =
392 m/day.
Q
s = (0.5772 ln (r 2 S/4KMt)
4 KM
or
s = (2.30Q/4KM)log(2.25KMt/r 2 S) (8-4)
As Q, KM and S are constant, a plot of drawdown s versus the logarithm of t forms a straight
line (Fig. 8.6). If this line is extended until it intercepts the time axis where s = 0, the interception
point has the coordinates s = 0 and t = t0. Substituting these values into Equation 8-3 gives:
0 = (2.30Q/4KM) log (2.25 KMt 0/r2S)
and because 2.30Q/4KM 0, it follows that 2.25KMt0/r2 S = 1 or
S = 2.25KMt 0/r2 (8-5)
The slope of the straight line in figure 8.6, i.e. the drawdown difference s per log cycle of time
log t/t0 = 1, is equal to 2.30Q/4KM. Hence
KM = 2.30Q/4 s (8-6)

226
Figure 8.6 Analysis of piezometer (r = 30 m) data from pumping test (Q = 788 m 3/d) in Oude Korendijk
(The Netherlands) with Jacob straight line method. Note that t 0 = 0.25 minutes or 0.25/1,440 days
and that s = 0.375 m
Introducing these values into the Eqs 8-6 and 8-5 yields:
KM = (2.30 x 788)/(4 x 3.14 x 0.375) = 385 m2/d, and
S = (2.25 x 385 x (0.25/1440))/302 = 1.7 x 10 4.
This method can be applied if the pumping test satisfies the assumptions and conditions
mentioned above for Theiss method plus the condition:
The values of u are small (u<0.01), i.e. r is small and t is sufficiently large.
8.4.5 Recovery analysis
The analysis of the data collected during the pumping period of single well tests (section 8.5.1)
and pumping tests can be checked by the analysis of the recovery data collected after the
pumping has stopped. The analysis is based on the principle of superposition, i.e. it is assumed
that after the pump has been shut down, the well continues to be pumped at the same discharge
as before, and that from the time pumping ceased, the well receives an imaginary recharge equal
to the discharge (Fig. 8.7). The recharge and the discharge thus cancel each other out, resulting in
an idle well, as is required for the recovery period.
The residual drawdown, s, after a pumping test with a constant discharge is:
Q
s = {W(u) W(u)}
4 KM
where W(u) is the well function for the assumed recharge.

When u and u are sufficiently small and the storage coefficient during the recovery

227
Groundwater studies
Figure 8.7 Time drawdown and residual drawdown
period can be assumed to be equal to the storage coefficient during the pumping period, this
equation can be written as:
Q 4KMt 4KMt 2.30Q t

s = (ln ln ) or s = log ( )
4 KM 2
r S r 2S 4KM t
where t is time since the start of pumping and t is the time since the cessation of pumping.
A plot of s versus t/t on single log paper (t/t on the logarithmic scale) will yield a
straight line with a slope:
2.30Q 2.30Q
s = , and hence KM =
4KM 4s
The recovery data does not allow the calculation of the value of S.
8.4.6 Well flow formula for other conditions
For other aquifer conditions (e.g. leaky aquifers, unconfined aquifers, anisotropic aquifers,
aquifers close to hydraulic boundaries, wedge-shaped aquifers), or pumping conditions (e.g.
large-diameter wells, variable discharge, partially penetrating wells) well flow equations have
been developed that can be solved in an analogous manner, provided that the well function has
been tabulated. As a matter of course the well functions become more complicated when more
parameters are required to describe the flow to the well under the particular well flow
conditions. Kruseman and De Ridder compiled the pumping test analysis methods for a wide
variety of conditions (Kruseman et al., 1991; Genetier, 1984; see also Kruseman and De Ridder
1973, 1974 and 1975 that are translations in German, French and Spanish of Kruseman and De
Ridder, 1970, the predecessor of Kruseman et al., 1990).
Specific yield from pumping tests
A remark must be made concerning the calculation of the specific yield, SY, of an unconfined

228
aquifer. In principle it can be calculated from pumping test data, but in general the results are
very unreliable. SY is usually estimated from lithological data (Section 8.5.5).
8.5 Determination of hydraulic characteristics by other methods
8.5.1 Single-well tests
In many instances the hydraulic parameters have to be determined when there are no
piezometers and the water-level changes are measured only in the pumped well. The drawdown
in a pumped well, however, is influenced by well losses and well bore storage. In the hydraulics
of well flow, the well is assumed to have an infinitesimal radius so that the well bore storage can
be ignored. In reality, well bore storage is large by comparison with the storage in an equal
volume of aquifer material, therefore in a single-well test, well bore storage must be considered
when analysing the drawdown data. This is especially true when pumping tests are carried out
in large-diameter dug wells. On the other hand, the influence of well bore storage on the
drawdown in a well decreases with time and becomes negligible at t >25 rc2/KM, where rc is the
radius of the blank casing in which the water level is changing.
To determine whether the early time drawdown data are dominated by well bore storage,
a log-log plot of drawdown sw versus pumping time t should be made. If the early time
drawdown plot as a unit-slope straight line, we can conclude that well bore storage exists.
Curve-fitting methods and straight line methods have been developed to analyse single-well
tests, even when the early time data are affected by well bore storage (Kruseman et al., 1990).
8.5.2 Flowing well tests
When a confined aquifer whose piezometric level is well above the land surface is tapped by a
well, this well will be free-flowing or artesian. That means that the well has a constant draw-
down, i.e. the difference between the static piezometric head after shutdown and the outflow
opening of the well. When the well is opened up after a period of shut-in, the piezometric level
in the well drops instantaneously to the outflow level of the well, establishing a constant
drawdown, while the well starts flowing with a decreasing discharge rate. When the changes of
the discharge rate with time are measured, a curve-fitting method developed by Hantush
(Kruseman et al., 1990) can be used to calculate the diffusivity of the well (KM/S).
8.5.3 Slug tests
In a slug test, a small volume (or slug) of water is suddenly removed from a well, after which
the rise of the water level in the well is measured. Alternatively, a small slug of water is poured
into the well and the rise and subsequent fall of the water level are measured. From these
measurements, the aquifers transmissivity can be determined. If the aquifers transmissivity
is more than, say, 250 m2/d, the water level will recover too quickly for accurate manual
measurements and an automatic recording device will be needed. For conventional slug tests
curve-fitting methods have been developed (Kruseman et al., 1990).
8.5.4 Tidal movements
In estuaries and coastal deltas where the inland surface water level fluctuates under influence of
the tidal movements of the sea, the hydraulic parameters of semi-confined aquifers can be
calculated from data on the propagation of the tidal wave in the aquifer. This requires

229
Groundwater studies
piezometers in a line perpendicular to an open water course as well as a surface water level
recorder on the same line. Bosch (1951), as quoted by Van Eyden et al., (1963), studied the
situation depicted in Fig. 8.8.
Figure 8.8 Geohydrological conditions to which the formula of Bosch may be applied
They derived the following formula:
x,t = h + A 0.e x . sin (t x) (8-7)
1
2 2 = (8-8)
KM . c
.S
. = (8-9)
2 KM
where:
: Piezometric level in the aquifer [L]
x : Distance to the open water [L]
t : time since observations started [T]
h : constant level of the phreatic surface in the semi-pervious confining layer [L]
A 0: amplitude of the tide in the estuary [L]
: attenuation factor [L1]
: retardation factor [L1]
: 2 /T [radians.L1]
T : duration of a tidal period [T]
K : hydraulic conductivity of the semi-confined aquifer [LT1]
M : thickness of the semi-confined aquifer [L]
c : hydraulic resistance of the confining layer [T]
S : storage coefficient [-]
To calculate the geohydrological characteristics the attenuation Ax/A0 for each piezometer
is plotted versus the distance x from the open water on semi-logarithmic paper (Ax/A 0 on the
logarithmic scale). This gives a straight line. The slope of this line corresponds to the attenuation
factor . Subsequently, the retardation of the minimum and maximum values of the piezometric

230
values with reference to the moments of Low Tide and High Tide, respectively, is plotted on
graph paper versus the distance x from the open water. The slopes of the straight lines give the
values of the retardation factor for LT and HT, respectively. The average of the two values is
used in the calculations.
With the values of and known, the values of KMc and S/KM are calculated from the
equations (8-8) and (8-9).
8.5.5 Lithology and grain size analysis
There have been numerous attempts to determine the hydraulic conductivity from the lithology,
or more precisely from the grain size analysis. The formulas developed for this purpose have
only local significance and are only applicable to very uniform sand strata. Such formulas have
to be calibrated by other methods, e.g. pumping tests.
The range of variation of the specific yield, SY, of an unconfined aquifer is rather small;
from 0.05 for clayey fine sand to 0.30 for well rounded, perfectly sorted very coarse sand. The
specific yield is therefore usually estimated from the lithological data, using tables such as
presented by Driscoll (Table 8.4).
Table 8.4 Range of variation of the specific yield (Driscoll, 1986)
Material Sy Material Sy
Coarse gravel 23 Limestone 14

Medium gravel 24 Dune sand 38
Fine gravel 25 Loess 18
Coarse sand 27 Peat 44
Medium sand 28 Schist 26
Fine sand 23 Siltstone 12
Silt 8 Silty till 6
Clay 3 Sandy till 16
Fine-grained sandstone 21 Gravelly till 16
Medium-grained sandstone 27 Tuff 21
We consider the values mentioned in Table 8.4 as maximum values for perfectly sorted
and perfectly rounded material. Values half these may be encountered in the field.
8.5.6 Laboratory tests
Laboratory tests are generally considered as completely unreliable for determining the hydraulic
conductivity of an aquifer.
8.5.7 Water balance and groundwater flow simulation model
It is only possible to calculate the transmissivity from a water budget if the groundwater flow
term is known accurately. This is only so if the groundwater recharge and discharge terms are
known accurately, but these are terms that are notoriously difficult to determine.
The groundwater flow simulation model is little else than the integration of the simul-
taneous water balance calculation in each of the nodal areas in which the study area has been
discretised (Anderson and Woessner, 1991). This has the advantage that all available knowledge
on the aquifer and groundwater conditions can be integrated and that anomalies are clearly
visible in the form of aberrant calculated groundwater levels. During the calibration process all

231
Groundwater studies
parameters are adjusted within physically acceptable limits, to reproduce the historical ground-
water levels. The accuracy of the calibrated transmissivity values depends on how accurate the
other parameters are.
8.6 Determination of hydraulic characteristics of aquifers in fractured rocks
Fractures in a rock formation (Fig. 8.9) strongly influence the fluid flow in that formation. Con-
ventional well flow equations were developed primarily for homogeneous and isotropic aquifers
(e.g. uniformly well sorted sand layers), therefore they do not adequately describe the flow in
fractured rocks. An exception occurs in hard rocks of very low permeability if the fractures are
numerous enough and are evenly distributed throughout the rock; then the fluid flow will only
occur through the fractures and will be similar to that in an unconsolidated homogeneous
aquifer (Fig. 8.9-B).
Figure 8.9 Porosity systems: a) single porosity, b) microfissures, c) double porosity
A complicating factor is the fracture pattern, which is seldom known accurately. This
means that based on the geological data, a fracture model must be assumed . In recent years,
many theoretical models have been developed, but only the well functions for the double
porosity model (Figure 8.10), the single vertical fracture and the single vertical dike model have
been tabulated (Kruseman et al., 1990).
8.6.1 Double-porosity model
The double-porosity concept regards fractured rocks as consisting of matrix blocks with a
primary porosity and low hydraulic conductivity, separated by fractures with a low storage
capacity but a high hydraulic conductivity. This concept assumes that there is no variation of
head in the matrix blocks, i.e. the interporosity flow is in pseudo-steady state. The flow trough
the fractures will be radial and in no-steady state when the aquifer is pumped by a well (Fig-
ure 8.11).
Curve fitting methods and straight line methods have been developed to analyse
pumping tests in double porosity aquifers.

232
Figure 8.10 Semi-log time-drawdown plot for an observation well in a fractured rock formation
of the double-porosity type
8.6.2 Single vertical fractures
If a well intersects a single vertical fracture, the aquifers unsteady drawdown response to
pumping differs significantly from that predicted by the Theis solution. This well flow problem
has long been a subject of research in the petroleum industry, and various solutions have been
proposed; most have produced erroneous results.
A major step forward was taken when the fracture was assumed to be a plane, vertical
fracture of zero width (zero storage), relatively short length and infinite hydraulic conductivity
(Figure 8.11).
If an aquifer that satisfies this model is pumped by a well that intersects the plane
fracture midway, the assumption of infinite (or very high) hydraulic conductivity means that the
drawdown in the fracture is uniform over its entire length (i.e. there is no hydraulic gradient in
the fracture). The rock around the fracture is poorly permeable, and when the well is pumped
the drawdown in the fracture induces a flow from the rock matrix into the fracture. At early
pumping times this flow is one-dimensional (i.e. it is horizontal, parallel and perpendicular to
the fracture). In unconsolidated sediments this flow pattern occurs when there is a constant
groundwater discharge into an open channel with constant head that fully penetrates the
aquifer.
As pumping continues, the flow pattern changes from parallel flow to pseudo-radial
flow, regardless of the fractures hydraulic conductivity, and the aquifer reacts as an isotropic
aquifer. The time required to attain pseudo-radial flow may be excessively long. Curve-fitting
methods have been compiled to analyse pumping tests performed under the above mentioned
conditions.
8.6.3 Single vertical dikes
Dikes used to be regarded as impermeable, but they can become highly permeable as a result of
jointing when the magma cooled, or by fracturing as a result of shearing, or by weathering. If a
single, permeable, vertical dike bisects a country-rock aquifer whose transmissivity is several

233
Groundwater studies
Figure 8.11 A well that intersects a single plane fracture of finite length and infinite hydraulic conductivity:
a) the well-fracture system, b) the parallel flow system at early pumping times, c) the pseudo-
radial flow system at late pumping times
times less than that of the dike, a specific flow pattern will be created if the well is pumped. The
dike is assumed to be infinitely long and to have a finite width and a finite hydraulic con-
ductivity. The dikes permeability stems from a system of uniformly distributed fractures,
extending downward and dying out with depth. Below the fractured zone, the dike rock is
massive and impermeable. The upper part of the dike is also impermeable because of intensive
weathering or a top clay layer (Figure 8.12). The water in the fractured part of the dike and in
the aquifer in the country rock is thus confined. When the well is pumped at a constant rate, a
trough of depression will be formed, instead of a cone of depression. The time drawdown curve
shows three characteristic time periods: at early times all the water pumped originates from
storage in the dike and none is contributed by the country-rock aquifer. At medium times, all the
water pumped is supplied by the aquifer and none is contributed from storage in the dike. At
late times, the flow in the aquifer is pseudo-radial. The publication of Kruseman et al. (1990)
includes curve-fitting methods developed by Boehmer and Boonstra.

234
Figure 8.12 Composite dike/aquifer system: a) cross section showing an aquifer of low permeability
in hydraulic contact with the highly permeable, fractured part of a vertical dike,
b) plan view of flow in a pumped dike at early and late pumping times
8.7 Determination of hydraulic characteristics of volcanic rocks
Volcanic rocks show the characteristics both of sedimentary rocks (e.g. stratification of lavas, tuff
and breccias, sometimes with alluvial intercalations, or with fossil soils on top of weathered
parts of older volcanic deposits), and of fractured rock (the best known example probably being
polygonal cooling joints in basalt).
The analysis of geological data in combination with data on the hydrological behaviour of
the volcanic sequence must result in the selection of a model for the groundwater flow and
hence in the selection of the model that must be used to analyse pumping tests or to provide
other quantitative data on the groundwater flow.
8.8 Determination of hydraulic characteristics of limestone
In general, limestones have a low primary porosity and hence a low primary permeability.
Fracturing will provide a secondary permeability that puts the limestones in the same league as
other fractured rocks (Section 8.6). On the other, hand the fracturing is sometimes so intense that
the whole rock is criss-crossed by tiny fractures, resulting in a pseudo-isotropic aquifer.

235
Groundwater studies
In thoroughly karstified limestone (mature karst), the saturated groundwater flow

becomes restricted to the flow through large solution channels. This karst channel flow has more
of the characteristics of surface water flow than of groundwater flow.
8.9 Diffusion, dispersion and macrodispersion
Diffusion, dispersion and macrodispersion are transport phenomena additional to advective

groundwater flow. Diffusion and dispersion are mathematically combined into the disperse tran-
sport term in traditional groundwater transport models. The dispersive flux can be attributed to
two processes: molecular diffusion and hydrodynamic dispersion. The first has to be attributed
to Brownian motion of species and the second to differences in velocities in porous media due to
friction along the edges of the pores and to the tortuosity of stream lines along individual grains
(Fig. 8.13a,b). It should be stressed that these two processes at a microscopic level are the one
and only two true physical processes that cause dispersive transport. All other due to the
mathematical treatment of the phenomenon of dispersion and the method of observation in the
field or the laboratory (Domenico and Schwartz, 1990).
The two physical processes are combined to yield the dispersion coefficient, D, that is
used in the description of groundwater transport:
D = Dmp + . v
where Dmp is the aqueous diffusion coefficient in a porous medium, a is the dispersivity and v is
the average linear pore water velocity. It can be seen from this equation that molecular diffusion
becomes increasingly important with decreasing flow rate. The diffusion coefficient in a porous
medium is four times lower or more than that in a free liquid. This can be obtained from tables;
for aqueous solutes at 25C the value is about 1.0 x10 5 cm2/s.
Values for the dispersivity, , are more troublesome to obtain. The best procedure is to
create a breakthrough curve by a column experiment using a representative sample. However,
this can be unachievable. An estimate can be made from the grain size analysis, because it is
well known that the value for the dispersivity increases with increasing average grain size and
decreasing sorting. The following expression provides a good estimate for geohydrologically
homogeneous conditions.
= (d60 /d10) . d50
where di is the grain diameter below which i% of the grain size distribution falls.
If this approach is unrealistic due to the heterogeneous nature of the porous medium, an
upper limit for the value of the dispersivity can be found if one assumes that the value is related
to the length of the flow domain studied (L):
= L/10
In fact the dispersion is no longer considered as a true physical process but as a bundling of
flow tubes, each with its own travel time for advective flow at a macroscopic level which is
caused by the heterogeneous nature of the subsurface (Fig. 8.13c). This phenomenon is usually
referred to as macrodispersion. The principal difference between the two is that dispersion is
irreversible in its nature whereas macrodispersion is reversible. The combined act of hydro-
dynamic dispersion and molecular diffusion perpendicular to the groundwater flow direction
together with macrodispersion leads to the partly irreversible nature of macrodispersion as well.

236
Figure 8.13 The principal processes of hydrodynamic dispersion and macrodispersion
ANDERSON, M. P; WOESSNER, W. W., 1991. Applied Groundwater Modeling. Academic Press, Inc. San
Diego, 381 pp.
BOSCH, H. (1951) Geohydrologisch onderzoek te Bergambacht. Unpublished research report. (In
Dutch.)
DOMENICO, P. A.;. SCHWARTZ, F. W, 1990. Physical and Chemical Hydrogeology; Second Edition. John
Wiley and Sons, New York.
DRISCOLL, F. G. (ed.), 1986. Groundwater and Wells. Second Edition. Johnson Division, St Paul,
Minnesota, 1 089 pp.
FETTER, C. W. (1993). Contaminant Hydrogeology. Prentice Hall, Upper Saddle River, NJ.
GENETIER, B., 1984. La pratique des pompages dessai en hydrogologie. BRGM, manuels et methodes
No. 9, Orlans, 132 pp.
GROUNDWATER MANUAL, 1981. A Water Resources Technical Publication. U.S. Department of the
Interior, Water and Power Resources Service, U.S. Government Printing Office, Denver,
Colorado, 480 pp.

237
Groundwater studies
KRUSEMAN, G. P.; DE RIDDER, N. A., 1970. Analysis and Evaluation of Pumping Test Data. ILRI Bull. 11,
Wageningen, 200 pp.
(traduit par A. Meilhac), 1974. Interprtation et discussion des pompages dessai. ILRI Bull. 11-F,
Wageningen, 213 pp.
(bersetzt von A.W. Uehlendahl), 1973. Untersuchung und Anwendung von Pumpversuchsdaten.
Verlagsgesellschaft Rudolf Mller, Kln-Braunsfeld, 191 S.
, (traduccin M. Donezar), 1975. Anlisis y evaluacin de los datos de ensayos por bombeo. ILRI
Bol. 11-S, Wageningen, 212 pp.
KRUSEMAN, G. P., DE RIDDER, N. A.; VERWEIJ, J. M., 1991. Analysis and Evaluation of Pumping Test Data
(Completely Revised Second Edition). ILRI Publ. 47, Wageningen, 377 pp.
THEIS, C.V., 1935. The Relation between the Lowering of the Piezometric Surface and the Rate and
Duration of Discharge of a Well using Groundwater Storage. Trans. Amer. Geophys. Union,
Vol. 16, pp. 51924.
VAN EYDEN, W. A. A.; KUPER, W. W.; SANTEMA, P., 1963. Some Methods Used in the Geo-Hydrological
Survey of the South-Western Deltaic Area in the Netherlands. IAHS Publ. 63, pp. 52857.

238
9 Isotopes techniques in
g ro u n d w a t e r i n v e s t i g a t i o n s
9.1 Introduction
A comprehensive understanding of a groundwater system is necessary if thid resource is to be

developed sustainably without adverse effects on the environment. In general the most crucial
of the many questions which hydrogeologists may concern the safe yield and the source of
water in a groundwater system. In order to provide a satisfactory answer, many associated
extremely important, and often difficult, questions need to be answered. Isotope techniques are
effective tools for obtaining a variety of hydrological information such as: the origin of
groundwater, determination of its age, velocity and direction of flow, interrelations between
surface waters and groundwaters, possible interconnections between different aquifers, aquifer
characteristics such as porosity, transmissivity, dispersivity, etc. The cost of such investigations is
often relatively small in comparison to the cost of classical hydrological techniques, and in
addition isotopes provide information which sometimes cannot be obtained by other techniques.
Applications of isotopes in hydrology are based on the general concept of tracing, in
which either intentionally introduced isotopes or naturally occurring (environmental) isotopes
are employed. Environmental isotopes (either radioactive or stable) have the distinct advantage
over injected (artificial) tracers in that they facilitate the study of various hydrogeological
processes on a much larger temporal and spatial scale through their natural distribution in a
hydrological system. Thus, environmental isotope methodologies are unique in regional studies
of water resources to obtain time and space integrated characteristics of groundwater systems.
Artificial tracers are generally effective for site-specific, local applications.
In this chapter, the basic principles of isotope techniques applied to groundwater studies
are discussed. References provided include more details of different applications and examples
of field studies. More specific advice on isotope applications in hydrological and environmental
studies is available from the Isotope Hydrology Section, International Atomic Energy Agency
(IAEA), Wagramerstrasse 5, P.O. Box 100, A-1400, Vienna, Austria.
9.2 Environmental isotopes
Environmental isotopes, both stable and radioactive, occur in the atmosphere and the hydro-
sphere in varying concentrations. So far, the most frequently used environmental isotopes
include those of the water molecule, hydrogen: 2H or D, deuterium, and 3H, tritium and oxygen:
18O, as well as of carbon: 13C and 14C, radiocarbon or carbon-14 occurring in water as con-
stituents of dissolved inorganic and organic carbon compounds. 2H, 13C and 18O are stable iso-
topes of the respective elements whereas 3H and 14C are radioactive isotopes. The stable isotopes

239
Groundwater studies
are usually measured, with an isotope ratio mass spectrometer, to find the isotope ratios of the
less abundant to more abundant isotope, e.g., 2H:1H and 18O:16O (1H and 16O being the number
of atoms of the most abundant isotopes of the respective elements). The radioactive isotopes are
measured either by counting their radioactive decays (low-level counting), e.g., by Liquid
Scintillation Spectrometry) or, using Accelerator Mass Spectrometry, by counting the number of
atoms in a given sample.
9.2.1 Stable isotopes
(i) Stable isotope ratios
Variations in stable isotope ratios of natural compounds are governed by chemical reactions and
phase changes due to the energy difference between chemical bonds involving different isotopes
of an element. Such energy differences are caused by the relative mass difference between
isotopes. The stable isotopes of the light elements show the widest variations because they have
the largest relative mass difference. For example, variations in 2H:1H ratio are greater than
18O:16O because the ratio of mass difference between 2H and 1H is 2:1 while the same for 18O and
16O is 1.1:1. In addition to the 2H:1H and 18O:16O ratios in water, various other stable isotopes are
used in groundwater studies (Table 9.1).
Table 9.1 Environmental stable isotopes used in groundwater studies
Natural abundance Natural abundance

Isotope Ratio (atoms/atoms) Isotope Ratio (atoms/atoms)
2H 2H/1H 1.55 . 10 4 18O 18O/16O 2.04 . 10 3

3He 3He/4He 1.38 . 10 6 34S 34S/32S 4.22 . 10 2
6Li 6Li/7Li 7.50 . 10 2 37Cl 37Cl/35Cl 0.242
11B 11B/10B 0.80 81Br 81Br/79Br 0.493
13C 13C/12C 1.11 . 10 2 87Sr 87Sr/86Sr 0.709939
15N 15N/14N 3.66 . 10 2
The interest of an isotope hydrologist lies in the relative deviation of the ratio of less
abundant heavy isotope to more abundant lighter isotope (exceptions: helium, lithium and
boron isotope ratios) with respect to a reference rather than in the absolute isotope ratio of a
given sample. Therefore, a material is selected as a primary standard, and its stable isotope ratio
defines the zero point of a relative conventional scale. For convenience the measurements are not
reported as isotope ratios, but given as relative deviation from the isotope ratio of a standard
expressed as (delta) in permil (), defined as
R sampleRstandard (9.1)
() = Rstandard
x 1,000
where, (e.g., 2H, 13 C, 15N, 18O, 34S) is the normalised difference of the isotope
concentration ratios R (2H:1H, 13C:12C, 15N:14N, 18O:16O, 34S:32S) of a sample and a standard. For
example, hydrogen and oxygen isotopes are presented with reference to the hypothetical
Standard Mean Ocean Water (SMOW) (Craig, 1961a), which has been superseded by the Vienna
Standard Mean Ocean Water (VSMOW) standard which for all practical purposes is close to
SMOW.

240
Isotope techniques in groundwater investigations
(ii) Stable isotopes of hydrogen and oxygen
Most of the applications of stable isotopes of hydrogen and oxygen in groundwater studies
employ regularities in their variations in atmospheric precipitation, i.e., in the input to a hydro-
geological system under study.
Many natural processes cause variations in isotope compositions of natural waters. The
most important are evaporation and condensation. During evaporation, the light molecule of
water, H216O, is more volatile than the heavy molecules, i.e., 1H2H16O or H218O. Therefore,
vapour which evaporates from an ocean is depleted with respect to the ocean water by about
1215 in 18O and 80120 in deuterium. When this atmospheric water vapour undergoes
successive cooling and condensation with production of clouds and precipitation, the less
volatile (heavy) water molecules condense preferentially, leaving a residual vapour more and
more depleted of 2H and 18 O. As a result, successive precipitations derived from the same initial
vapour mass will be more and more depleted of heavy isotopes. Because the degree of con-
densation of a vapour mass depends on temperature, a relation between isotope composition of
precipitation and its temperature of formation should be expected: as the formation temperature
decreases, the -values of precipitation decrease (Dansgaard, 1954). This has been observed
directly in Antarctic precipitation (Picciotto et al., 1960) and a world-wide relation between 18O
of precipitation and mean annual air temperature has been reported (Fig. 9.1).
This dependency on temperature produces seasonal isotope variations of precipitation
(there is more heavy isotope depletion in winter precipitation than in summer precipitation),
latitude effect (high latitude precipitation has been depleted more than low latitude pre-
cipitation), and altitude effect (heavy isotope content of precipitation decreases with increasing
altitude) (Friedman et al., 1964; Moser and Stichler, 1970; Rosanski et al., 1993). These effects
constrain the use of these isotopes to delineate various hydrogeological processes, to indicate
past and present climate changes and to identify palaeowaters.
Through long-term observations made within the framework of the IAEA/WMO Global
Network for Isotopes in Precipitation (GNIP), a linear correlation between 2H and 18O in
precipitation samples collected from a world-wide network of stations has been established
(Rozanski et al., 1993), which is close to the so-called Global Meteoric Water Line (GMWL)
defined by Craig (1961b) and represented by the relationship:
2H = 8 . 18 O + 10 (9.2)
Precipitation which has undergone significant evaporation during its fall does not obey
Equation (9.2). Evaporation tends to enrich both of the heavy isotopes in water, but not in the
same relative proportion indicated by the above relationship (Craig, 1961b; Ehhalt et al., 1963;
Woodcock and Friedman, 1963).
When precipitation infiltrates to recharge groundwater, mixing in the unsaturated zone
and selective infiltration of precipitation result in attenuation of seasonal isotopic variations in
precipitation. In most aquifers, isotopic composition of water does not change further unless
exchange with the oxygen in rocks occurs. This process of exchange is significant and important
for high temperature geothermal systems only. The isotopic composition of groundwater is thus
related to that of precipitation in the recharge area of an aquifer at the time of recharge.
Groundwater may be very old, and the climatic conditions of the region at the time of recharge
may have been different from those of today. Due to the correlation between -values and
temperature (Gat, 1971) this implies that the isotopic composition of precipitation could have
been different from the present one.
Groundwater may also be recharged by seepage from surface waters, such as rivers and

241
Groundwater studies
Figure 9.1 The annual mean 18O of precipitation as a function of the mean annual air temperature at surface.
The figures in parentheses indicate the total thickness (in cm) of the investigated snow layers
(Modified after Dansgaard, 1964)
lakes. If most of the recharge is from seepage, the groundwater should reflect the mean isotopic
composition of the river or the lake instead of that of local precipitation which could be rather
different. The river may collect water which originates from precipitation in a completely
different area, for instance in a high mountain region; in this case its heavy isotope content
would be depleted compared to that of precipitation in the plain, due to altitude effect.

242
In the case of lakes or ponds, the water may be considerably enriched in heavy isotopes
through evaporation. The enrichment is limited by direct isotopic exchange with atmospheric
moisture. Thus, this enrichment is higher where evaporation is more intense with respect to the
total volume of water, e.g., in closed lakes and ponds or lakes and rivers in arid areas.
Groundwaters which have undergone evaporation before, during or after recharge
are easily recognised by their isotopic composition. Their heavy isotope content is higher than
that of non-evaporated waters in the region and they do not obey the relationship in the
Equation (9.2) (Craig and Gordon, 1965; Fontes and Gonfiantini, 1967; Diner, 1968).
9.2.2 Radioactive isotopes
Among the environmental radioisotopes, tritium and carbon-14 have found the widest appli-
cation in groundwater studies. Radioactive isotopes (also called radioisotopes) occurring in
groundwater originate from natural and/or artificial nuclear processes (Table 9.2).
Table 9.2 Cosmogenic and anthropogenic radioisotopes used in hydrology
Atmospheric production In situ production
Initial content Equilibrium content

Radioactive Half-life in 1 litre of in 1 litre of
isotope (a) Reaction groundwater Reaction groundwater
85Kr 10.76 Anthropogenic 2.5 * 10 6 Bq None 0

3H 12.43 14N (n,t) 1 Bq 6Li (n,) 2 * 10 5 Bq
39Ar 269 40Ar (n,2n) 8.5 * 103 atoms 39K (n,p) 0.22 * 10 3 atoms
14C 5,730 14N (n,p) 3 .10 3 Bq 17O (n,) 3 * 10 5 Bq
81Kr 2.1 * 10 5 80Kr (n,) 1,000 atoms None 0
36Cl 3.01 * 10 5 35Cl (n,) 10 5 10 8 atoms 35 Cl (n,) 5 * 107 atoms
129I 16 * 10 6 Spontaneous 5 * 10 4 atoms Spontaneous 3 * 105 atoms
fission fission
Note:
'Bq' in columns 4 and 6 implies measurement by decay counting using liquid scintillation spectrometry whereas 'atoms'
implies measurement of atoms using accelerator mass spectrometry. 14C could also be measured by AMS techniques.
Cosmogenic radioisotopes are produced in nuclear reactions between the nucleonic

component of cosmic radiation and the atmosphere. Anthropogenic radioisotopes are produced
in nuclear bomb tests and nuclear reactors. The concentrations of all these radioisotopes in
groundwater are very low and usually measured by counting their decay rate A in a given
sample. The number of atoms N in a sample can be derived from A by the relationship:
A=.N (9.3)
where, , the decay constant, is related to half-life T1/2 by the equation = ln 2/T. For
long-lived radioisotopes such as 36 Cl and 129I, the decay rate becomes unmeasurably small. In
these cases the number of atoms has to be measured directly, which is possible using accelerator
mass spectrometry (Elmore and Phillips, 1987). This AMS technique is superior to the con-
ventional decay counting for 14 C, since AMS requires a very small sample size (up to 1,000 times
less than conventional requirement) for analysis.

243
Groundwater studies
(i) Tritium
Tritium (3H), the radioisotope of hydrogen, emits low-energy radiation (Emax. = 18 keV).
Tritium content in water is expressed in Tritium Units (TU). 1 TU is equal to 1 atom of 3H per
1018 atoms of 1H, which is equivalent to 0.118 Bq or 3.193 pCi per litre of water. The half-life of
tritium is fixed as 12.43 a (Taylor and Roether, 1982). The concentration of tritium in natural
waters is generally very low. In hydrological studies, therefore, electrolytic enrichment of tritium
is often carried out prior to decay counting with liquid scintillation or proportional counters.
Environmental tritium occurs in precipitation from both natural and anthropogenic
sources. The natural production results from cosmic-ray produced neutrons interacting in the
upper atmosphere with nitrogen atoms,
14 1 3 12
N+ n H+ C (9.4)
7 0 1 6
Tritium oxidises rapidly to HTO and enters the global hydrological cycle. The natural
content of tritium in precipitation is estimated to be about 25 TU (Craig and Lal, 1961; Lal and
Peters, 1962). The second source of tritium is from atmospheric detonation of thermonuclear
devices from 1952 to 1962, and minor releases from industrial nuclear facilities. The atmospheric
testing injected periodic pulses of tritium into the stratosphere and in the Northern Hemisphere
in 1963 its concentration in precipitation had increased by more than three orders of magnitude
above that arising from cosmic-ray source (~5 TU). An increase in concentration was also noted
in the southern hemisphere but by only two orders of magnitude because of the lower ratio of
land to ocean in that hemisphere and the poor trans-equatorial mixing of the air masses. The
history of tritium content in precipitation is well known, owing to the global network of stations
jointly established by the IAEA and the WMO (Fig. 9.2).
Figure 9.2 Long-term tritium concentration in precipitation at Ottawa (Canada ) and Kaitoke (New Zealand)

244
The International Atomic Energy Agency, Vienna, publishes data on the concentration of
stable isotopes and tritium in precipitation samples collected at a large number of stations
around the globe, from which it is possible to estimate the tritium deposition at most places of
interest. The detailed information, references and data can be found on the Internet site:
http://www.iaea.org/programs/ri/gnip/gnipmain.htm.
Over the past decades, groundwater studies have considerably benefited from the
transient behaviour of bomb-tritium. Shallow groundwater was naturally labelled by the high,
impulse-generated tritium concentration of atmospheric precipitation. The detection of bomb-
tritium in shallow groundwater is a fingerprint for a component of recent recharge in the
groundwater. Indication of recent replenishment of groundwater is very important for the
management of these resources, especially in semi-arid regions where groundwater is often a
non-renewable resource. A more quantitative treatment of tritium data in shallow, unconfined
aquifers allows of the residence time distribution in groundwater to be determined, from which
relevant parameters of the groundwater system can be estimated, especially the recharge rate.
Presently, tritium concentration in precipitation is approaching the natural level, which makes
such evaluations of tritium data more difficult. It is now often impossible to use tritium for
evaluating groundwater recharge in the southern hemisphere, where the bomb-tritium in
precipitation is much lower than in the northern hemisphere.
However, the combined measurement of tritium and its decay product, helium (3He),
still provides a powerful tool for estimating groundwater residence time and recharge rate, and
for characterising flow and dispersion regimes in shallow aquifers. Assuming that a water mass
travels like a parcel (without mixing and dispersion) along the flow path and that there is
neither a 3He loss nor a contribution from sources other than tritium decay within the parcel, the
radioactive decay law can be employed to derive the residence time (transit time from the point
of recharge to the sampling site):
1 + 3He(t)
t t0 = 1 ln 3 (9.5)
H(t)
where, 3H (t) is the tritium concentration at the time of sampling t, 3He (t) is the tritiogenic 3He
concentration (given in the same measuring unit), and t0 is the time at which a water parcel
reached the water table. The tritiogenic 3He concentration is obtained from the measured total
dissolved 3He in a sample and subtracting the atmospheric and subsurface contributions. This
requires additional measurements of neon and helium-4 in the sample (Schlosser et al., 1989).
Note that in equation (9.5) initial tritium concentration 3H (t 0) is not explicitly included. This
quantity can, however, be reconstructed through the combined measurement of 3H and
tritiogenic 3He according to the equation:
3H(t = 3H(t) + 3He(t)

n) (9.6)
Thus, with this approach, the transient behaviour of bomb-tritium can be exploited to advantage
even if most of this tritium has already decayed.
(ii) Radiocarbon (14C)
Radiocarbon is produced in the transitional region between stratosphere and troposphere by

cosmic ray neutrons reacting with nitrogen atoms (Libby, 1965),
14 1 14 1
N+ n C+ H (9.7)
7 0 6 1

245
Groundwater studies
14C oxidises to carbon dioxide and becomes a part of the atmospheric carbon dioxide
reservoir and subsequently enters the biosphere and the hydrosphere. 14C has also been added
to the atmosphere as a result of the testing of thermonuclear devices since 1952 (Fontes, 1983;
IAEA, 1983).
14C emits radiation (E
max. = 156 keV) and has a half-life of 5,730 a (Godwin, 1962). For
hydrogeological applications, the 14C concentration is expressed as a percentage of modern (pre-
bomb) carbon activity of biosphere material used (percent of Modern Carbon or pMC). This
refers to the ratio of the activity of a sample to 95% of the activity (in 1950) of the accepted oxalic
acid standard from the US National Institute of Standards and Technology (NIST). The 100%
corresponds to specific 14C activity of 13.56 0.07 disintegrations per minute per gram of carbon
(Olsson, 1968). The activity of 14C is generally measured by decay counting using liquid
scintillation spectrometry or, more recently, by measurement of atoms using AMS.
The addition of man-made radiocarbon to the atmosphere has increased the natural levels;
the maximum increase was about 100% in the Northern Hemisphere (achieved in 1963) and 70%
in the Southern Hemisphere (achieved in 1965) (Fig. 9.3). This effect is important when dealing
with very young groundwaters near the recharge areas.
Figure 9.3 Carbon-14 distribution in atmospheric CO2 (Levin, 1978)
14C determinations are generally carried out on the total dissolved inorganic carbon
(TDIC), i.e., dissolved carbonate species CO2 (aq.), HCO31 and CO32 , the bicarbonate ions
being the species by far prevailing at the pH values normally encountered in groundwater. The
major part of carbon enters the groundwater through soil CO2, during the infiltration processes.
However, chemical and biochemical reactions along a flow path can modify the initial 14C
content of groundwater. Therefore, in application of radiocarbon dating of groundwater,
appropriate adjustments for effects of these reactions should be made in order to determine the
initial activity of 14C in the groundwater. Several models have been developed that consider the
13C and other physico-chemical parameters for the required corrections.
The principles for dating groundwater are based on the decay law and the assumptions
made with regard to the origin of the TDIC in groundwater. However, the behaviour of carbon
species in the solution is not conservative and several correction models have been developed to
account for the chemical interaction between the rock matrix, and other possible gaseous phases
in the system. The detailed information on various corrections or models can be found in the
monographs or textbooks (e.g., Fritz and Fontes, 1980; Clark and Fritz, 1997).

246
The other radioisotopes compiled in Table 9.2 have potential to date groundwater in an
age range of the order of their half-lives, i.e., many tens to some hundreds of thousands of years.
9.2.3 Applications in groundwater studies
(i) Groundwater recharge
A qualitative and quantitative characterisation of groundwater recharge is essential for sustain-

able development and management of groundwater resources. When withdrawal rate exceeds
recharge rate (over-exploitation), considerable lowering of the water table may be observed.
Subsequently, the manifestation of the phenomenon could be sea water encroachment in coastal
aquifers, induced leakage from adjacent saline aquifers into fresh aquifers or contamination of
aquifers from polluted surface water bodies
Aquifers which receive little recharge will exhibit only small fluctuations in groundwater
levels, so that a reliable estimate of recharge rate cannot easily be obtained on the basis of
classical approaches such as water level monitoring alone. In many recharge rate estimations
isotope techniques have provided a cost-effective tool. These techniques constitute virtually the
only approach for identification and evaluation of groundwater recharge particularly under arid
and semi-arid conditions.
a. Recharge from precipitation
It is possible to identify modern recharge, which occurred during the past forty to fifty years by
using bomb-tritium data measured in soil water in an unsaturated zone and in groundwater
from shallow, unconfined aquifers and springs. An example of the use of tritium from the
unsaturated zone is from Louga, Senegal, where the tritium content of samples taken in 1990
from an unsaturated zone profile showed the 1963 tritium peak at 12 m below the land surface.
The depth to the water table was 34 m. By applying a simulation model, the rate of modern
recharge was estimated to be 28 mm per year, corresponding to about 10% of the annual rainfall
(Tandia et al., 1993). Note that Louga is located in the Sahel region where precipitation events,
although sporadic, have a regular recurrence.
Similar investigations have been carried out using the 36Cl bomb-peak in precipitation
(e.g. Liu et al., 1995). The maximum 36Cl concentration occurred in precipitation around 1955,
almost 10 years prior to the tritium peak in precipitation. Presently, the tritium and 36Cl peaks
can only be found in the unsaturated zone if the recharge rate is low and the unsaturated zone
sufficiently thick (e.g., sand dunes in arid and semi-arid zones). Under more humid conditions,
these peaks have already reached the water table.
For a shallow unconfined aquifer recharged by vertical infiltration of precipitation, the
vertical penetration depth H of the thermonuclear tritium peak below the water table can be
determined and the average recharge rate R estimated using the relationship:
R = nH
tt (9.8)
0
where, n is the effective porosity, t the sampling time and, in the case of bomb-tritium, t 0 = 1963.
The applicability of the method is limited to the upper region of shallow aquifers, ideally for the
top layer of a water table where vertical movement of water is dominant (Andres and Egger,
1985). Equation (9.8) can also be used when the water samples are collected from supply wells
tapping the whole thickness of an unconfined aquifer. In that case, H stands for the aquifer
thickness (assumed to be constant along flow direction) and the (tt 0) term is replaced by mean
residence time derived from the measured tritium content of a sample (Zuber, 1986). In a more

247
Groundwater studies
comprehensive demonstration of this approach, tritium concentrations of water samples

collected from an unconfined aquifer near Sturgeon Falls, Ontario, Canada were evaluated using
a one-dimensional advection-dispersion model (Robertson and Cherry, 1989). A recharge rate of
150 mm/a (equivalent to about 16% of the annual precipitation) was obtained.
The 3H3He method represented by equations (9.5) and (9.6) has enhanced the potential of
bomb-tritium to estimate groundwater recharge rates through residence time determination. In a
shallow, sandy aquifer system in the southern New Jersey coastal plain, the apparent 3H/3He
ages were found to be in good agreement with results obtained by other approaches like
chlorofluorocarbon (CFC) dating and model-simulated flow path analysis (Szabo et al., 1996). In
this case, the vertical flow velocity was found to be of the order of 1 m/year.
Nevertheless, even in cases of low vertical flow velocities, the identification of the
3H 3He peak can be used for dating and thus, for recharge rate estimation (recharge rate =
porosity x vertical flow velocity). In addition, the mere presence of tritiogenic 3He in ground-
water with no measurable tritium provides evidence of modern recharge. For example, an 3He
content of about 3 TU in groundwater with no tritium can result in an aquifer where there is
modern recharge with water that has a tritium content of about 3 TU near the water table.
Although the rate of recharge in such an aquifer would be very low, it must be present in order
to provide a continuous source of 3He and to overcome diffusive 3He losses. The tritium and 3He
data can be modelled to estimate recharge and transport parameters in unsaturated and
saturated zones (Dillon and Aggarwal, 1999).
Under certain circumstances, the residence time and thus recharge rate of young
groundwater can also be estimated from measurements of the seasonal 2H/1H and 18O/16O
variations. It has been shown by Maloszewski et al. (1983) that the mean residence time of
groundwater is a function of attenuation of the amplitude of its seasonal variation relative to the
seasonal variation in local precipitation replenishing the groundwater. The applicability of this
method is limited to those areas where precipitation shows a pronounced seasonal variation,
e.g., in mountainous areas. The method was applied by Ramspacher et al. (1992) to estimate the
recharge rate of an Alpine groundwater system in Austria.
Many karst springs along the northeast Adriatic coast of Italy show seasonal changes in
the stable isotope composition of groundwater. The temporal pattern showed the most depleted
values during the summer months, suggesting a six-month time lag between the signal observed
in precipitation and the values measured in the springs. These results were unexpected,
considering that karst aquifers usually show a rapid transit time from recharge area to the
sampling point. On the basis of the location of the springs it was possible to prove that during
the winter months, the dominant recharge area is located closer to the coastal area, while during
the summer period, the precipitation at higher elevation (and therefore, isotopically more
depleted) was preferentially contributing to the recharge of the karst system (Flora and
Longinelli, 1989).
b. Recharge from surface water
Groundwater often consists of a mixture of recharge from surface water (lakes or rivers) and
local precipitation. It is important to know the proportions of these recharge components in
order to (1) increase the sustainable supply of drinking water through bank infiltration (ground-
water enrichment, induced recharge) and/or (2) prevent drinking water pollution by infiltration
of water from a contaminated surface water source. Different recharge components can be
identified from the stable isotope compositions of groundwater because evaporation of water in
surface water bodies, in particular under semi-arid and arid conditions, leads to an enrichment
of the heavy isotopes 2H and 18O. A simple isotopic balance equation can then be used to
estimate the relative proportions of surface water and precipitation in recharge. The accuracy of

248
this determination generally depends on the magnitude of the difference in isotopic com-
positions of the two components and under ideal conditions is in the order of a few percent.
Measuring the isotopic composition of water samples from the High Dam Lake, Egypt and
adjacent groundwater, Aly et al. (1993) found that infiltration of lake water was limited to a
distance of about 10 km from the lake.
Isotopes can also be used to identify the lack of recharge to an aquifer from surface water
bodies. Kulkarni et al. (1999), by employing environmental isotopes and hydrochemistry,
concluded that a limestone aquifer in the semi-arid part of Rajasthan State, India, did not receive
recharge from the local lakes.
Rivers carrying water derived from precipitation at much higher altitude are isotopically
depleted compared to local precipitation. Moreover, river water can show a seasonal variation,
which usually is observed in wells near the river with reduced amplitude and after a time lag.
The time lag as well as the change in the mean isotopic composition gives the minimum time
(transit time) required for river water and possibly its dissolved pollutants to reach a ground-
water supply well and the fraction of river water (possibly polluted) relative to other recharge
sources. By measuring 18 O in a bank infiltration along a section about 100 m from the River
Neckar near Heilbronn, Germany, estimated.a mean transit time of 7 to 42 days and fractions of
river water from about 100% to 30%.
c. Palaeorecharge
Groundwaters in shallow aquifers typically have residence times of decades to hundreds of

years. In contrast, deeper and less permeable aquifers that extend for many kilometres can have
through-flow times of thousands of years. If the flow regime is simple and mixing is minimal,
such aquifers can serve as archives of information about environmental conditions at the time of
recharge. The stable isotopes of hydrogen and oxygen of the palaeowater reflect the air surface
temperature and the air mass circulation at the time when it was formed by infiltrating precip-
itation (Rozanski et al., 1993). Palaeotemperatures derived from noble gas analyses complement,
and are sometimes more meaningful, than those derived from oxygen-deuterium analyses.
In arid climates, river water may be enriched in 2H and 18O relative to groundwater, if it
was replenished under former more humid conditions. For example, groundwater adjacent to
the river Nile along its course in Sudan and Egypt is such palaeowater significantly depleted in
2H and 18 O relative to the present Nile water. Making use of these differences in the isotopic
compositions of the mixing components, Vrbka et al. (1993) have estimated the fraction of Nile
water in groundwater collected along the river Nile between Khartoum and Atbara, Sudan.
Assuming a mixture between palaeowater and Nile water, they found Nile water fractions
between 40% and 80% at the different sampling sites. Palaeorecharge in the aquifers of arid
regions of Jalore and Barmer districts (a part of the Thar Desert) in western Rajasthan, India was
identified using environmental isotopes (Rao and Kulkarni, 1997). It was observed that deeper
palaeowaters were depleted of stable isotope compositions compared to present-day precip-
itation. The shallow aquifers, on the other hand, were replenished by infiltration of water
through river channels during episodic floods caused by sporadic rain events.
Stute et al. (1992) studied isotopic compositions of groundwater in the Carizzo aquifer in
south Texas and found a 5C to 6C reduction in temperature during the glacial period, consis-
tent with the snow-line estimates. Palaeohydrology studies based on isotope methods also
indicated that long-term climate change has strongly influenced modern groundwater flow
systems through changes of their boundary conditions, e.g. influence of the continental ice
shield on groundwater recharge (Siegel, 1990 and 1991). Love et al. (1994) have demonstrated
the effects of glacial/interglacial climate change on recharge in South Australia and Fontes et al.
(1991) have used environmental tracers to show how shifts in the position of the Niger River

249
Groundwater studies
since the end of the last glaciation have affected the distribution of groundwater recharge. Stute
and Talma (1998) found that during the last glacial maximum southern Africa was characterised
by 5C to 6C lower temperatures and by an expansion of the winter rainfall region that received
its moisture from the Atlantic Ocean.
(ii) Groundwater transit time
The radioactive decay of environmental radioisotopes (Table 9.2) and the transient nature of
some of them (bomb- 3H, anthropogenic 85Kr, bomb- 14 C and bomb - 36 Cl) make these isotopes a
unique tool for determining the groundwater residence time (age), i.e. the length of time the
water has been isolated from the atmosphere (Davis and Bentley, 1982). In unconfined aquifers
there is preferentially a vertical gradient of the isochronal surfaces (surface of constant residence
time of the water), while in confined aquifers the dominating feature is a lateral gradient. It can
be shown (Jordan and Froehlich, 1990) that in the former case this gradient is approximately
proportional to the inverse of the recharge rate (cf. Equation (9.8)), while in the latter case the
gradient is inversely related to the flow velocity. Therefore, the hydrogeologically relevant para-
meters primarily addressed by groundwater dating with radioactive isotopes are the recharge
rate and flow rate of groundwater in unconfined and confined aquifers, respectively.
One of the approaches to determine groundwater flow rate by a suitable radioisotope is
based on measuring the decrease in radioisotope concentration along the flow path. If radio-
active decay alone is governing the decrease in radioisotope concentration, then the radioactive
decay law yields for groundwater flow time t2 t1 between two wells (well 1 and 2):
T1
t2 t1 = 2 ln 1
C
ln2 (9.9)
C2
where C1 and C2 are the radioisotope concentration at wells 1 and 2, and T is the half-life of a
radioisotope used. Consequently, the groundwater flow velocity (v) is given by:
x2 x1 (9.10)
v=
t2 t1
where, x2 x1 is the distance between the two wells. With porosity (n) of an aquifer, the ground-
water flow rate (Vf) (volume of water flowing per unit time across the unit cross section of an
aquifer) can be estimated by:
Vf = n . v (9.11)
This simple approach requires access to at least two wells along the flow path in an aquifer.
Often, however, the flow direction is not well known or only one well is available for sampling.
In these cases the absolute age (t) needs to be determined rather than the difference of ages at
two wells (flow time). The respective relationship for absolute groundwater age is obtained
from Equation (9.9) by replacing the C1 by C0, setting C 2 = C , t 2 = t and t1 = t 0 = 0. Therefore,
this approach requires knowledge of the initial concentration C0 of the respective radioisotope.
The reconstruction of C0 is one of the major problems in groundwater dating with radioisotopes.
Under natural conditions, groundwater movement is generally very slow, often a few
metres per year. Therefore, after a distance of a few kilometres along the flow path, the ground-
water is already old and its age is beyond the dating range of 3H, 3H/3He and CFCs. The most
common radiometric approach to determining groundwater residence times in large aquifers
has been 14C. Its half-life of 5,730 a makes it a suitable tool for the dating of groundwater in an
age range from about 1 to 40 ka (Geyh, 2000).
An integrated mass balance approach may be used to understand geochemical and

250
isotopic evolution of groundwater and factors affecting the use of 14C for groundwater dating.
Plummer et al. (1994) developed an interactive geochemical model NETPATH (available through
Internet on the USGS site) for 14C dating, based upon the geochemistry along a flow path and
boundary conditions provided by the user. However, such geochemical models are based on the
assumption that CO2 produced in the root zone alone is the source of 14C in infiltrating water.
The 14C content of this CO2 is assumed to be nearly in equilibrium with the atmosphere. Recent
studies (Aggarwal and Dillon, 1998; Affeck et al., 1998; Bacon and Keller, 1998) indicate that
carbon dioxide may also be produced in deeper parts of the unsaturated zone and near the
capillary fringe. The 14C content of this CO2 would be much different, and generally lower than
that in the root zone. The initial 14C content of groundwater, therefore, can be much lower than
that at atmospheric equilibrium without radioactive decay or water-rock interaction. Under such
conditions, the usual corrections for dating purposes are invalid.
Radiocarbon dating of groundwater using 14C of dissolved organic carbon (DOC) has also
been developed (Murphy et al., 1989). The DOC of groundwater is derived in the soil zone, but,
unlike the TDIC, the DOC remains unaffected by dilution through carbonate reactions, and so
provides an independent method of 14C groundwater dating. Wassenaar and co-workers (1992)
examined the advantages and disadvantages of the DOC against the TDIC method. Murphy et
al. (1989) presented a detailed analysis of the interrelationship between carbon isotopes, carbon
species chemistry, and microbial activity on the dissolved organic and inorganic components in
the Middendorf aquifer. More recently, Artinger et al. (1995) discussed further methodological
aspects, emphasising that DOC dating should be focused on the fulvic acid component of DOC.
In groundwater with a high DOC concentration (>100 mg/L), they also found that 14C-free fulvic
acid could be released from sedimentary organic carbon (e.g. lignite) to dilute soil-derived fulvic
acid. It is expected that improved organic geochemical methods may in future be available to
correct 14 C for subsurface contributions.
Very slow moving groundwater in deep confined aquifers extending over tens and, in
some cases, several hundreds of kilometres, can reach ages of tens and even hundreds of
thousands of years. These ages are beyond the dating range of 14 C and require the use of very
long-lived radioisotopes. Of the three long-lived radioisotopes 81Kr, 36Cl and 129I (Table 9.2), only
36Cl has found wider practical use so far. Interpretation of 36Cl data in terms of groundwater age
is often hampered by insufficient knowledge of in situ production (Andrews and Fontes, 1992).
36Cl combined with other dating tools has been used to study the Great Artesian Basin aquifer
system in Australia (Torgersen et al., 1991; Love et al., 2000). Application of the other two radio-
isotopes is still in an exploratory stage. Recently, Collon et al. (1999) reported the first appli-
cations of 81Kr to groundwater dating in the Great Artesian Basin of Australia. They obtained
mean groundwater residence times of 225,000 to 400,000 year. The interpretation of these values
in comparison with available 36Cl dates is still being worked on. The authors concluded that the
results confirm 81Kr as a reliable tool to date old groundwaters.
Among the various radionuclides of the natural decays series, combined use of the
uranium isotopes 238U and 234U has shown some potential to date old groundwater (Ivanovich
and Harmon, 1992). Froehlich and Gellermann (1987) reviewed various attempts made by
several authors in using these isotopes, especially the 234U/238U activity ratio, as a dating tool.
They concluded that the major constraint of this approach comes from sorption-desorption
reactions, which can reduce the lifetime (the reciprocal value of the radioactive decay rate ) of
uranium in the liquid phase more than one order of magnitude below the radioactive lifetime
of 234U (2.45 10 5 a). Therefore, the applicability of the uranium isotope dating depends on the
geochemical conditions in a groundwater system. One of the practical applications of the
234U: 238U disequilibrium method is found in the Milk River aquifer study (Ivanovich et al.,
1991). The ages derived from the uranium isotopes were in reasonable agreement with the 36Cl
dates and the hydraulically calculated values (Froehlich et al., 1991). Musgrove and Banner

251
Groundwater studies
(1993) have used 87 Sr, 234U: 238U and other heavy element isotope analyses, more commonly
associated with rock than water, to trace the sources and rates of a deep groundwater flow
system in the United States of America.
Among the noble gases directly or indirectly produced in groundwater and in the aquifer
material by radioactive decay, 4He has proved to be the most useful for groundwater studies.
This is mainly due to the relatively rapid rate of 4He production compared to other noble gases.
4He is formed by neutralisation of particles produced in -decay processes in natural decay
series (238U, 232 Th and 235U). A fundamental advantage of using decay products such as 4He for
dating is the fact that as time progresses, more 4He will be present in groundwater and its
measurement will become easier. This is in contrast with the atmospheric radionuclides, which
will become more difficult to detect as the age of groundwater increases. The half-life of most -
producing radionuclides is more than two orders of magnitude than the age of the oldest
groundwater which might be dated. Thus, groundwater in homogeneous, confined aquifers
often experiences a nearly linear increase in 4He with time. However, the measured 4He concen-
trations usually exceed the values expected from this in situ production P (typical values in
sandstones are between 10 12 and 10 11 cm3 STP 4He/ga). This is mainly due to diffusion of
crustal helium produced in strata above and below an aquifer in question. Assuming a confined
aquifer of the thickness H, effective porosity n, and a crustal 4He flux F (often between about
1 and 6 10 6 cm3 STP 4He/cm3 a), the groundwater age t follows from the equation:
[4He] = ( P + F t ) (9.12)
nH
where, [4He] is the concentration of 4He in the groundwater. A typical field study using this tool,
was performed in the Great Artesian Basin, Australia (Torgersen et al., 1991).
It can be seen that there are various radioisotopes available to date groundwater and thus
study its flow dynamics. Given the uncertainties inherent in each of the methods for ground-
water dating, in groundwater studies as many methods should be used, as possible.
(iii) Interconnections between aquifers
Both the groundwater dynamics within a system of aquifers and the risk of groundwater con-
tamination can be influenced by hydraulic interconnections between aquifers. Environmental
isotopes, especially the stable isotopes, can be used to investigate such interconnections,
provided the isotopic composition of the groundwater in the aquifers is different.
A typical example of this approach is the identification of a hydraulic connection between
the Continental intercalaire, a large Lower Cretaceous aquifer in northwestern Africa, with
the overlying Complexe terminal aquifer near the Gulf of Gabes in Algeria (Gonfiantini et al.,
1974a). The 18O and 2H values of the groundwater in the confined part of the Continental
intercalaire were found to be relatively uniform; they were respectively about 3 and 20 lower
than in the Complexe terminal. In two areas, however, these stable isotopes provided a clear
fingerprint that the Continental intercalaire discharged into the overlying aquifer, in spite of the
fact that the two aquifers are separated by a clayey formation several hundreds of metres thick.
It was found that in these areas the groundwater of the Complexe terminal had the same
isotopic composition as the Continental intercalaire. This finding confirmed earlier expectations
that the El Hamma fault system provides a hydraulic connection through the rather thick
aquiclude.
Intense exploitation of aquifers can cause leakage from overlying and underlaying
aquifers. Payne et al. (1980) reported a typical case study related to a vertically stratified aquifer
system at Hermosillo, northwestern Mexico. Here, an alluvial aquifer is underlain by a low-
permeability clay formation of about 100 m thickness. Below this clay layer is a confined aquifer

252
with a piezometric head about 30 m above the water table of the overlying alluvial aquifer. The
groundwaters of these two aquifers had different stable isotope compositions. Heavy exploi-
tation of the alluvial aquifer caused a vast cone of depression and induced leakage from the
deep aquifer into the shallow one. The stable isotope data indicated that contribution of deep
groundwater into the shallow aquifer was up to 20%.
Isotopes can also be used to prove a lack of hydraulic interconnections between aquifers.
Ho et al. (1991) applied the stable isotope approach in a study of the Mekong Delta aquifer
system, Vietnam. They found that the upper Pleistocene aquifer was hydraulically isolated from
the lower Pliocene aquifer.
(iv) Groundwater salinisation
In many situations it is necessary to identify the mechanism of salinisation in order to prevent or

alleviate the cause. Salinity may be used as a practical tool in water resource investigation to
estimate recharge and discharge, to investigate palaeohydrology, and in the understanding and
management of groundwater in coastal regions (Edmunds and Droubi, 1998). Isotope techniques
can be used to distinguish the importance of the following processes which may lead to
salinisation of groundwater:
leaching of salts by percolating water; the origin of salts may be evaporite deposits,
aeolian salts, or the products of weathering,
intrusion, present or past, of salt water bodies such as sea water, brackish surface
water, brines and connate water,
concentration of dissolved salts by evaporation.
For example, the stable isotope composition of salt springs of the Sarvistan and Caspian
Gates in Iran was found to be the same as that of local groundwater. It was concluded that the
salinity originated through dissolution of rock salt beds by infiltrating recharge waters (Zak and
Gat, 1975). Isotope geochemical methods have been used extensively to demonstrate the
evaporative origins of saline groundwaters in inland basins, for example in chotts of Algeria
(Gonfiantini et al., 1974b), in the alluvial aquifers of the Indo-Gangetic Plain in Haryana, India
(Kulkarni et al., 1996). Combinations of geochemical techniques and various ratios may be
used to determine salinity of different ages and origins. The isotope geochemical approach was
applied to study the origin of salinity in the coastal, alluvial, multi-aquifer system in the
Mahanadi Delta, India (Kulkarni et al., 1998). The findings indicated that the groundwater
salinity in aquifers originated due to transgressions of sea during the Late Pleistocene and the
Holocene.
The origin of chloride in groundwater can be traced by the stable chlorine isotope ratio,
which also can be applied to investigate the evolution of deep formation waters (Eggenkamp,
1994). A multi-component approach using concentrations of major, minor, and trace elements
along with isotope compositions (2H, 3H, 11B, 13C, 14C, 18O, 34S, 37Cl, 87Sr, etc.) would be best
suited for studying the origin of salinity. However, some of these techniques are still in an
exploratory stage.
(v) Groundwater pollution
Pollution of aquifers by anthropogenic contaminants is one of the problems of high concern in

the management of water resources. Environmental isotopes can be used to trace the pathways
and predict spatial distribution and temporal change of pollutants in aquifers. This information
is critical in order to be able to understand the source of contaminants, assess their scale and
migration, and plan for remediation.
Measurements of concentration and stable isotope composition of sulphate and nitrate in

253
Groundwater studies
groundwaters have been widely used to identify the sources of sulphate/nitrate pollution as
well as the microbial sulphate reduction and denitrification processes, respectively. For example,
Kendall (1998) studied the sources of nitrate in catchment areas during the snow melt season in
spring. In many of the catchment areas an increase of nitrate concentration was observed in the
early spring. The nitrogen isotope data provided evidence that the soil nitrate is a major source
of the observed increase rather than atmospheric deposition, as previously assumed. Boettcher
et al. (1990) quantitatively assessed the microbial denitrification in a sandy aquifer in northern
Germany. They found a linear relationship between the 18O and 15N values of residual nitrate
of groundwater where denitrification was identified. Various applications of sulphur isotopes
have been discussed in detail by Pearson and Rightmire (980) and by Krouse (1980).
Concentration and stable isotope composition of hydrocarbons and their degradation
products can together provide a powerful tool for pollution assessment and remediation. For
example, Aggarwal and Hinchee (1991) used the 13C of soil CO2 to substantiate aerobic
biodegradation of hydrocarbons at three sites contaminated with jet fuel in the United States of
America. Kelley et al. (1997) measured the concentrations and 13C values of benzene, toluene,
ethylbenzene and xylene under bioremediation conditions at a gasoline-contaminated site in
southern California. Their data suggest the presence of at least two sources of gasoline con-
tamination (leaded and unleaded gasoline) at the investigation site. Aggarwal et al. (1997) were
the first to demonstrate in a laboratory study the combined use of oxygen and carbon isotope
analyses of carbon dioxide and molecular oxygen, respectively, for monitoring the biode-
gradation of fuel hydrocarbons. They concluded that the combined use of stable carbon isotope
composition of carbon dioxide and oxygen isotope composition of the electron acceptors
(molecular oxygen, nitrate, or sulphate) provides a robust tracer for the verification and
quantification of microbiological processes associated with hydrocarbon-contaminated
groundwater.
9.2.4 General remarks on environmental isotopes
Environmental isotopes are a supplementary tool for hydrogeological investigations and should
be employed as an integral part of studies. The amount and quality of information that they are
able to provide is strictly dependent on the degree of hydrogeological knowledge about the
area under investigation and also on the knowledge of the investigator. Generally, an integrated
approach using isotope, hydrogeological, and hydrochemical data will lead to the optimum use
of these techniques and to a logical interpretation.
In studying the isotope content of groundwater, a major requirement is to obtain a repre-
sentative and uncontaminated sample. For stable isotopes and tritium, relatively small samples
are required: 20 mL are sufficient for stable isotopes, 500 mL for tritium. Containers must be
airtight to avoid evaporation and exchange leading to alteration of isotope composition. Poly-
ethylene or polypropylene bottles with inner cones in their screw caps are generally used.
Bottles should be filled completely and care taken to minimise exposure of samples to the atmo-
sphere. 14C sampling is carried out by separation of dissolved carbonate species from a large
volume of water (usually 50100 L) to yield a few grams of carbon, preferably at the sample site.
The details of the sampling procedures, storage and precautions are discussed by Clark and
Fritz (1997).
Furthermore, isotopes can trace dispersion and infiltration of pollutants in landfills as well
as quantify degradation and migration of pollutants. Isotopes are also applied extensively
to study: leakages from dams and reservoirs, stream flow measurements, effluent dispersion,
suspended sediment and bedload movement in ports and harbours, lake dynamics, sedimen-
tation in lakes and reservoirs, geothermal systems, glaciology, etc. Various monographs and
textbooks describe these applications.

254
9.3 Artificial isotopes
Artificial radioisotopes can be measured in extremely low concentrations and often in situ,
making possible the design of convenient and efficient field experiments. However, safety
precautions, special equipment and trained personnel are required for handling radioisotopes.
Therefore, artificial radioisotope should be considered when other tracers such as fluorescent
dyes would not meet the needs of a problem. Fluorescent dyes became the first option for many
surface water and karst system studies. The main limitation for their use is that water samples
need to be measured outside the system, either with portable equipment or in the laboratory.
Some recent developments may overcome this difficulty. Artificial radioisotopes have added
advantage in that they can be used in a variety of conditions and environments and the quantity
required is very small.
Radioactive tracers can be used to determine the following parameters:
1. aquifer characteristics: (a) porosity, (b) transmissivity, (c) dispersivity;
2. direction and velocity of groundwater flow;
3. stratification of aquifers.
9.3.1 Radioactive tracers
When choosing radioactive tracers the following points should be taken into consideration:
The isotope should have a half-life compatible to the presumed duration of the
observations. The unnecessary use of long-lived isotopes will leave these isotopes in
the water for a long time, creating a persistent health hazard and potential interference
if the experiment needs to be repeated.
The isotope should not be adsorbed by the inorganic or organic solids in the aquifer.
It is advantageous to measure the concentration of radioactive tracers in the field. For
this reason -emitters are generally preferred.
The isotope must be available at a reasonable cost when and where required.
Frequently used radioactive tracers are listed in Table 9.3.
The chemical form of the radionuclide used in water tracing experiments plays an
important role. Anions are very useful as water tracers because usually they are not adsorbed
onto a matrix. In general, cations are not good tracers. In the form in which they are commonly
delivered, they may be subject to strong adsorption, interfering with the validity of their tracing
of the water mass. To minimise this effect within an aquifer, cationic tracers need to be
transformed into neutral molecules or anionic compounds. For this, they are combined with
complexing agents. The most commonly used in hydrological practice is a chelated metallic
compound formed by means of ethylene diamine tetracetic acid (EDTA).
Although not fully meeting the half-life and radiation requirements mentioned above,
tritium is an ideal marker for water since it is incorporated in the water molecule. Hence it is
used in spite of the fact that in situ detection is not possible and samples must be analysed in the
laboratory subsequent to an experiment. In certain cases, tritiated water (HTO) may suffer some
delay in comparison with water bulk velocity as a result of ion exchange with clays (Kaufman
and Orlob, 1956; Knutsson and Forsberg, 1967). Minimal amounts of artificial tritium should be
used to minimise interference with groundwater resource studies based on environmental
tritium.

255
Groundwater studies
Table 9.3 Radioisotope tracers commonly used in groundwater investigations
9.3.2 Techniques
The techniques most commonly used to determine aquifers parameters are (Margrita and
Gaillard, 1991):
Single-well methods in natural flow conditions;
Multi-well methods in natural flow conditions;
Multi-well methods in pumping conditions.
In the first case, the parameters that can be quantified are Darcy velocity, its vertical
profile and vertical velocity.
The introduction of tracer into a borehole may be done by pouring through a thin pipe, by
crushing an ampoule at the depth of interest, or by using a special injection device. Injection can
be performed at one or several depths to facilitate mixing of the tracer solution with the
standing water of the borehole. After the release and mixing, the radioactivity in the borehole is
measured by a probe, usually a scintillation counter but sometimes a robust Geiger counter,
inserted at the chosen depth. Specialised probes have been constructed which seal off a defined
volume of the borehole by inflatable packers, release radiotracer into this volume, and measure
the radioactivity by detectors in or immediately above the release volume. Figure 9.4 shows a
probe that can measure both tracer dilution with time and direction of flow (Drost, 1989).
In the Single-well Dilution Technique (Halevy et al., 1967), the dilution rate of the tracer
by the natural flow of water through the borehole is observed. In the Single-well Pulse Tech-
nique (Borowczyk et al., 1965), tracer is forced into the aquifer by pumping water into the well
and then retrieved by pumping water out of the well.These single-well techniques have been
widely used in the past decades to obtain local parameters of aquifers. (Drost and Neumaier,
1974; Tazioli, 1977; Plata Bedmar, 1981, 1983; Mandel et al., 1985; Drost, 1989).
When movement of water between boreholes is to be observed (this is the Multiple-well
Technique) tracer is injected in a borehole and detected in another one. Under natural flow
conditions, direction of groundwater flow needs to be known. For this, several holes are drilled
around the injection point to intercept tracer movement. The parameters to be obtained in this

256
Figure 9.4 Tracer probe for dilution and direction logging (Modified after Drost, 1989)
test are velocity, longitudinal and transverse dispersion coefficients, and kinematic porosity. The
time scale of such an experiment is usually rather long when groundwater velocity is low. The
pumping method is usually preferred in this case. The observation well is pumped and the
radioactivity of the pumped water monitored at the surface (Halevy and Nir, 1962). It is
important that such pumped water be disposed of in such a way that it cannot return inad-
vertently to the system under investigation. In some experiments, water pumped from the
observation well is returned to the injection well intentionally to establish a closed circuit. The
pumping flow must be chosen so that the velocity induced by pumping is greater than the
natural velocity of the aquifer. It is assumed that the isochrone line around the pumping
piezometer is circular (Margrita and Gaillard, 1991).
9.3.3 Applications
(i) Infiltration studies in the unsaturated zone
Information on the movement of water in the unsaturated zone is needed whenever one is
trying to set up a water balance, either for the unsaturated zone itself or for the groundwater in
the saturated zone beneath it.
Radioactive tracers are applied to study the movement of infiltrating water, and some-
times of pollutants, in the unsaturated zone. The technique includes the injection of one or more
tracers at a certain depth, either as a point source or in a horizontal plane, and the measurement

257
Groundwater studies
of the displacement of the labelled horizon after a certain time period. To determine the
infiltration rate and other transport parameters, it is required to determine the soil moisture
content and the tracer concentration in the samples collected along a depth profile.
Tritiated water and other radiotracers have been widely used to determine recharge rates
under different climatic conditions (Zimmermann et al., 1966, 1967; Datta et al, 1978; Nair et al.,
1979). Athavale et al. (1980) studied the variability of recharge in several regions in India under
different climatic conditions and soil types. If the main soil types are properly represented, this
approach can be used to determine the regional water balance at local or regional scales.
Multi-tracer experiments can be used to study the movement of pollutants, by deter-
mining the retardation factor of different solutes compared to infiltration rate of water. In some
cases, organic compounds (pesticides) or fertilisers are labelled with other artificial isotopes like
15N, 14C, 32P, 22Na, etc. (Nicholls et al., 1982).
(ii) Effective porosity of an aquifer
The principle of the method of porosity determination in the saturated zone is based on the
approximate equality of porosity (Void volume/Total volume) and partial volume of water s
equals Volume of water/Total volume.
Tracer is introduced into a well and a second well, at a distance r, is pumped. Ignoring the
dispersion of tracer on its path between wells, its arrival at the second well signifies that the
volume of water pumped, V, is
V = r 2 bs (9.13)
where b = aquifer thickness, r = distance between wells and s = effective porosity. The prime
requirements are that the distance between the wells be large compared to the aquifer thickness,
(r >b), radial pumping velocities be larger than the natural velocities, and the cone of depression
at the pumping well be small compared to volume of water pumped. A typical experiment of
porosity determination has been fully described by Halevy and Nir (1962).
(iii) Transmissivity
The transmissivity T, first introduced by Theis in his non-equilibrium equation, characterises

the ability of an aquifer to transmit water (Hantush, 1963). Along with the storage coefficient,
hydraulic conductivity, specific storage and specific yield, transmissivity is one of the formation
parameters characterising basic hydraulic properties of an aquifer. It is related to the hydraulic
conductivity K (mean filtration coefficient) and aquifer thickness, b, as,
T = bK (9.14)
Mercado and Halevy (1966) have found that the volume of water pumped during the
experiment referred to in section 9.3.3.1, measured until the tracer peak arrives at the next well,
is inversely proportional to the value of transmissivity. They have determined transmissivity
of two layers of the same aquifer by means of two injections and one observation well.
Transmissivity then was found from the relationships:
V1 = r12 b1 s1 T/T1
V2 = r22 b 2 s2 T/T2 (9.15)
T1 + T2 = T

258
where T is total transmissivity; T1,2 is partial transmissivity of the layers 1 and 2; r1,2 are distance
between injection and observation wells; b1,2 are thickness of the layers 1 and 2; s1,2 are effective
porosity of the layers 1 and 2.
(iv) Dispersivity
In problems of artificial recharge of aquifers by water of a lower quality, or the disposal of

wastes into the groundwater, a subject of interest may be the coefficient of dispersivity, D,
occurring in the tracer transport equation (Halevy et al., 1967). This value characterises the
mixing property of an aquifer. It is commonly estimated by finding the theoretical curve which
best fits the experimental tracer breakthrough curve in an observation well. Parameters on such
theoretical curves are dispersion coefficients derived from the mathematical model of the tracer
transport (Lenda and Zuber, 1970).
(v) Groundwater flow velocity
Groundwater flow velocity under the natural hydraulic gradient is generally evaluated from
Darcys formula. Using a tracer, it is possible to measure directly the groundwater velocity, and
from this to evaluate aquifer permeability. The method consists of injecting tracer into the well
and then following its concentration decrease with time (Fig. 9.5).
Figure 9.5 Schematic diagram showing the tracer concentration changes with time in a borehole dilution
experiment (Moser et al., 1963)
The principle applies for any tracer, but again radioactive isotopes are easier to detect in situ
with a high degree of accuracy, at very low concentrations (Moser et al., 1957, 1963; Mairhofer,
1963; Guizerix et al., 1963). The disadvantages of the method are that the results are valid only
in a restricted area neighbouring the well. Several measurements in different wells and at dif-
ferent depths are often necessary to establish a good picture of the groundwater flow in a given
aquifer.

259
Groundwater studies
The horizontal filtration velocity Vf of groundwater (also called Darcy velocity) is given
by:
[C]
Vf = V ln
Ft [C0] (9.16)
where, V is the measuring volume (the borehole volume in which dilution takes place), F is the
cross section of the measuring volume perpendicular to the direction of the undisturbed
groundwater flow, and t is the time interval between measurement of concentrations [C 0] and
[C]; is the correction factor accounting for the distortion of flow lines due to the presence of a
borehole. The calculation of the value of on the basis of well construction data is discussed by
Halevy et al. (1967). The above expression is a particular solution of the differential equation
describing dilution rate of the tracer.
In practice, several readings are required to derive results not affected by incomplete
mixing processes and residual currents. The lower limit of velocity, mainly dictated by the
diffusion rate, is 10 7 m/s.
If effective porosity, s, is known one can estimate the bulk velocity, Vbulk, as:
Vf
Vbulk = (9.17)
s
Potential (hydraulic) gradients, dh/dx, can be determined from the vertical distribution of
radioactivity in the observation wells if these boreholes are entirely cased (Weinberger, et al.,
1967). Knowing the potential gradient, one can also deduce the hydraulic conductivity, K,
applying the Darcy law:
Vf
K=
dh (9.18)
dx
Knowledge of vertical velocity component in a borehole is important in such problems as deter-

mination of the amount of water entering or leaving the well from different permeable strata, or
amount of interchange between zones. A single-well dilution technique may be applied. A series
of detectors placed at different depths is commonly used for such measurements (Fig. 9.6).
The method is specially recommended when expected vertical flow velocities are less than
2.102 m/s; in this region, due to the influence of friction, tracer methods are more efficient than
the use of mechanical current meters. The lower velocity limit is estimated at 10 7 m/s. In tur-
bulent flow, accuracy of measurement is reported to be better than 10% but less is known about
the accuracy in laminar flow (Halevy et al., 1967).
(vi) Direction of groundwater flow
The multi-well technique can be used to determine the direction of groundwater flow, especially
when there is radial symmetry in the well location, or if there are observation wells in the
approximate direction of flow. The direction of flow, however, may also be determined by means
of a single-well technique. Tracer is added to a segment of a borehole and is carried away in the
direction of the flow. Activity is then detected by a special directionally oriented probe rotated
by means of a stiff metallic rod for rather shallow wells, or in more sophisticated devices, by
means of a motor included in the probe. A diagram like that of Fig. 9.7 is then obtained, which
shows the direction of groundwater flow. In this case, contrary to the flow velocity measure-
ments, radioisotopes that are relatively rapidly absorbed by the formation are the most
appropriate ones to use.

260
Figure 9.6 Arrangement of a set of detectors for determining vertical flow in a borehole (Moser et al., 1963)
198Auin the form of chloride (198AuHCl4) and 51Cr in the form of chloride (51CrCl3) are
commonly used. In most laboratory tests the difference between measured and true flow
direction was less than 3. The reproducibility of the method in the field is better than 10%. A
simplified version of the method has been used by Wurzel and Ward (1965). A cylindrical metal
gauze placed in the well at the injection point adsorbs radioactive tracer (51CrCl3). The gauze is
removed, cut in many sections parallel to the axis and the activity of each section is measured in
the laboratory. The section giving the highest activity indicates the direction of water flow.
9.3.4 Practical considerations
In the past decades, the use of isotope tracer methods for the determination of parameters of
groundwater flow has gradually increased. The most fruitful field of application has been the
solution of local problems in civil engineering, in feasibility studies, site investigations, ground-
water pollution studies, and in the design, maintenance and control of hydraulic works (Drost,
1989).
The results of isotope tracers compare well with other conventional techniques and they
can give more detailed information. Values obtained from pumping tests represent the mean
hydraulic properties over the whole aquifer thickness. Dilution tests may arrive at the variations
of hydraulic conductivity with depth (Drost, 1989). As a result of dilution logs and pumping
tests in more than 20 unconfined porous aquifers, the hydraulic conductivities derived from
dilution logs, K (D), and those from pumping tests, K (P), fit a regression line (Fig. 9.8):

261
Groundwater studies
Figure 9.7 Diagram of water flow direction in a borehole. Activity injected, 0.6 mCi (Halevy et al., 1967)
K(D) = 1.15 K(P) + 0.0003, (r2 = 0.97) (9.19)
Since a commercial service in the application of these techniques is seldom available, it is

advisable to seek the co-operation of universities or research institutes in the initial use of
methods. Some difficulties in performing field measurements may be expected until the
investigator becomes experienced in the new techniques involved. The experimenter must
acquaint himself with all safety and legal requirements connected with the handling of
radioactive materials.

262
Figure 9.8 Unconfined porous aquifers. Relation of hydraulic conductivities determined by tracer dilution
techniques (K(D)) and by pumping tests (K(P)) (Drost, 1989)
AFFEK, H. P.; RONEN, D.; YAKIR, D., 1998. About Production of CO2 in the Capillary Fringe of a
Deep Phreatic Aquifer. Water Resour. Res., Vol. 34, pp. 98996.
AGGARWAL, P. K.; HINCHEE, R. E., 1991. Monitoring in Situ Biodegradation of Hydrocarbons by
Using Stable Carbon Isotopes. Environ. Sci. Technol., Vol. 25, pp. 1178 80.
AGGARWAL, P. K.; FULLER, M. E.; GURGAS, M. M.; MANNING, J. F.; DILLON, M. A., 1997. Use of Stable
Oxygen and Carbon Iisotopes Analyses for Monitoring the Pathways and Rates of
Intrinsic and Enhanced in Situ Biodegradation. Environ. Sci. Technol., Vol. 31, pp. 5906.
AGGARWAL, P. K.; DILLON, M. A., 1998. Stable Isotope Composition of Molecular Oxygen in Soil
Gas and Groundwater: A Potentially Robust Tracer for Diffusion and Oxygen Con-
sumption Processes. Geochim. Cosmochim. Acta, Vol. 61, pp. 57785.
ALY, A. I. M.; FROEHLICH, K.; NADA, A.; AWAD, M.; HAMZA, M.; SALEM, W. M., 1993. Study of
Environmental Isotope Distribution in the Aswan High Dam Lake (Egypt) for Estimation
of Evaporation of Lake Water and its Recharge to Adjacent Groundwater. Environ.
Geochem. and Health, Vol. 15, pp. 3749.
ANDRES, G.; EGGER, R. ,1985. A New Tritium Interface Method for Determining the Recharge Rate
of Deep Groundwater in the Bavarian Molasse Basin. J. Hydrol., Vol. 82, pp. 2738.
ANDREWS, J. A.; FONTES, J. CH., 1992. Importance of the in Situ Production of 36Cl, 39Ar, and 14C in
Hydrology and Hydrogeochemistry. In: Isotope Techniques in Water Resources Development,
International Atomic Energy Agency, Vienna, pp. 24569.

263
Groundwater studies
ARTINGER, R.; BUCKAU, G.; KIM, J. I.; GEYER, S.; WOLF, M., 1995. Influence of Sedimentary Organic
Matter on Dissolved Fulvic Acids in Groundwater: Significance for Groundwater Dating
with 14C in Dissolved Organic Matter. In: Isotopes in Water Resources Management, Vol. 1,
ATHAVALE, R. N.; MURTI, C. S.; CHAND, R., 1980. Estimation of Recharge to the Phreatic Aquifers
of the Lower Maner Basin, India, by Using the Tritium Injection Method. J. Hydrol.,
Vol. 45, pp. 185202.
BACON, D. H.; KELLER, C. K., 1998. Carbon Dioxide Respiration in the Vadose Zone: Implications
for Groundwater Age Dating. Water Resour. Res., Vol. 34, pp. 306977.
BOETTCHER, J.; STREBEL, O.; VOERKELIUS, S.; SCHMIDT, H. L., 1990. Using Isotope Fractionation of
Nitrate-Nitrogen and Nitrate-Oxygen for Evaluation of Microbial Denitrification in a
Sandy Aquifer. J. Hydrol., Vol. 114, pp. 41324.
BOROWCZYK, M.; GRAUCZAK, J.; ZUBER, A., 1965. Radioisotope Measurements of the Groundwater
Flow Direction by Single Well Method. Nukleonika, Vol. 10, pp. 1928.
CLARK, I.; FRITZ, P., 1997. Environmental Isotopes in Hydrogeology. Lewis Publishers, Boca Raton,
New York, 328 pp.
COLLON, P.; KUTSCHERA, W.; LEHMANN, B. E.; LOOSLI, H. H.; PURTSCHERT, R.; LOVE, A.; SAMPSON, L.;
ANTHONY, D.; COLE, D.; DAVIDS, B.; FAUERBACH, M.; HARKEWICZ, R.; MORRISSEY, D. J.;
SHERRILL, B. M.; STEINER, M.; PARDO, R. C.; PAUL, M., 1999. Development of Accelerator
Mass Spectrometry (AMS) for the Detection of 81Kr and First Application to Groundwater
Dating. In: Isotope Techniques in Water Resources Development and Management, International
Atomic Energy Agency, Vienna, CD-ROM.
COPLEN, T. B., 1993. Uses of Environmental Isotopes. In: Alley, W. D. (ed.), Regional Groundwater
Quality. Van Nostrand Reinhold, New York, pp. 22754.
CRAIG, H., 1961a. Standard for Reporting Concentrations of Deuterium and Oxygen-18 in
Natural Waters. Science, Vol. 133, pp. 18334.
, 1961b. Isotopic Variations in Meteoric Waters. Science, Vol. 133, p. 1702.
CRAIG, H.; GORDON, L. I., 1965. Deuterium and Oxygen-18 Variations in the Ocean and the
Marine Atmosphere. In: Stable Isotopes in Oceanographic Studies and Paleotemperatures, CNR-
Laboratorio di Geologica Nucleare, Pisa, pp. 9130.
CRAIG, H.; LAL, D., 1961. The Production of Natural Tritium. Tellus, Vol. 13, pp. 85105.
DANSGAARD, W., 1954. The 18 O Abundance in Fresh Water. Geochim. Cosmochim. Acta, Vol. 6,
pp. 24160.
DANSGAARD, W., 1964. Stable Isotopes in Precipitation. Tellus, Vol. 16, pp. 43668.
DAVIS, S. N.; BENTLEY, H. W., 1982. Dating Groundwater; A Short Review. In: L. A. Currie (ed.),
Nuclear and Chemical Dating Techniques Interpreting the Environment Record, American
Chemical Society, Washington, D.C., pp. 187222.
DATTA, P. S.; GUPTA, S. K. ; SHARMA, S. C., 1978. Simulation of Water Movement through Vadose
Zone. In: S. K. Gupta and P. Sharma (eds.), Current Trends in Arid Zone Hydrology. Today
and Tomorrow Printer and Publishers, New Delhi, p. 93.
DILLON, M. A.; AGGARWAL, P. K., 1999. A Combined Vadose Zone Saturated Zone Model for
Tritium and Helium Transport in Groundwater. In: Isotope Techniques in Water Resources
Development and Management, International Atomic Energy Agency, Vienna, Paper SM-
361/32, CD-ROM.
DINER, T., 1968. The Use of Oxygen-18 and Deuterium Concentrations in the Water Balance of
Lakes. Water Resour. Res., Vol. 4, pp. 1289306.
DROST, J. W.; NEUMAIER, F., 1974. Application of Single Borehole Methods in Groundwater
Research. In: Isotope Techniques in Groundwater Hydrology, International Atomic Energy
Agency, Vienna, pp. 24153.
DROST, J. W., 1989. Single-Well and Multi-Well Nuclear Tracer Techniques A Critical Review.

264
Technical Documents in Hydrology, International Hydrological Programme, UNESCO,

Paris, 96 pp.
EDMONDS, W. M.; DROUBI, A., 1998. Groundwater Salinity and Environmental Change. In: Isotope
Techniques in The Study of Environmental Change, International Atomic Energy Agency,
Vienna, pp. 50318.
EGGENKAMP, H. G. M., 1994. D37Cl The Geochemistry of Chlorine Isotopes. Geologica Ultraiectina
No. 116, Thesis, Univ. Utrecht, 151 pp.
EHHALT, D.; KNOTT, K.; NAGEL, J. F.; VOGEL, J. C., 1963. Deuterium and Oxygen-18 in Rain Water.
J. Geophys. Res., Vol. 68, pp. 377580.
ELMORE, D.; PHILLIPS, F. M., 1987. Accelerator Mass Spectrometry for Measurement of Long-Lived
Radioisotopes. Science, Vol. 236, pp. 54350.
FLORA, O.; LONGINELLI, A., 1989. Stable Isotope Hydrology of a Classical Karst Area, Trieste, Italy.
In: Isotope Techniques in The Study of The Hydrology of Fractured and Fissured Rocks.
FONTES, J. CH.; GONFIANTINI, R., 1967. Comportment isotopique au cours de lvaporation de
deux bassins Sahariens. Earth Planet. Sci. Lett., Vol. 3, pp. 25866, 386.
FONTES, J. CH., 1983. Dating of Groundwater. In: Guidebook on Nuclear Techniques in Hydrology
(1983 Edition), International Atomic Energy Agency, Vienna, pp. 284317.
FONTES, J. CH.; ANDREWS, J. N.; EDMUNDS, W. M.; GUERRE, A.; TRAVI, Y., 1991. Palaeorecharge by
the Niger River (Mali) Deduced from Groundwater Geochemistry. Water Resour. Res.,
Vol. 27, pp. 199214.
FRIEDMAN, I.; REDFIELD, A. C.; SCHOEN, B.; HARRIS, J., 1964. The Variation of the Deuterium
Content of Natural Waters in the Hydrologic Cycle. Rev. Geophysics, Vol. 2, pp. 177224.
FRITZ, P.; FONTES, J. CH. (eds.), 1980. Handbook of Environmental Isotope Geochemistry, The Terrestrial
EnvironmentA, Vol. 1. Elsevier, Amsterdam, 545 pp.
FROEHLICH, K.; GELLERMANN, R., 1987. On the Potential Use of Uranium Isotopes for Ground-
water Dating. Chem. Geol. (Isotope Geoscience Section and), Vol. 65, pp. 6777.
FROEHLICH, K.; IVANOVICH, M.; HENDRY, M. J.; ANDREWS, J. N.; DAVIS, S. N.; DRIMMIE, R. J.; FABRYKA-
MARTIN, J.; FLORKOWSKI, T.; FRITZ, P.; LEHMANN, B. E.; LOOSLI, H. H.; NOLTE, E., 1991.
Application of Isotope Methods to Dating of Very Old Groundwaters : Milk River Aquifer,
Alberta, Canada. Appl. Geochemistry, Vol. 6, pp. 46572.
GAT, J. R., 1971. Comments on the Stable Isotope Method in Regional Groundwater Inves-
tigations. Water Resourc. Res., Vol. 7, pp. 9803.
GEYH, M. A., 2000. An Overview of 14C Analysis in the Study of Groundwater. Radiocarbon,
Vol. 42, pp. 99114.
GODWIN, H., 1962. Half-Life of Radiocarbon. Nature, Vol. 195, pp. 9845.
GONFIANTINI, R.; CONRAD, G.; FONTES, J. CH.; SAUZAY, G.; PAYNE, B. R., 1974a. Etude isotopique de
la nappe continentale intercalaire et de ses relations avec les autres nappes du Sahara
septentrional. In: Isotope Techniques in Groundwater Hydrology, Vol. I, International Atomic
Energy Agency, Vienna, pp. 22741.
GONFIANTINI, R.; DINER, T.; DEREKOY, A. M., 1974b. Environmental Isotope Hydrology in the
Hodna Region, Algeria. In: Isotope Techniques in Groundwater Hydrology, Vol. I, International
Atomic Energy Agency, Vienna, pp. 29316.
GUIZERIX, J.; GRANDCLEMENT, G.; GAILLARD, B.; RUBY, P., 1963. Appareil pour la mesure des vitesses
relatives des eaux souterraines par la mthode de dilution ponctuelle. In: Radioisotopes in
Hydrology, International Atomic Energy Agency, Vienna, pp. 2535.
HALEVY, E.; NIR, A., 1962. The Determination of Aquifer Parameters with the Aid of Radioactive
Tracers. J. Geophys. Res., Vol. 61, pp. 2403 9.
HALEVY, E.; MOSER, H.; ZELLHOFER, O.; ZUBER, A., 1967. Borehole Dilution Techniques: A Critical
Review. In: Isotopes in Hydrology, International Atomic Energy Agency, Vienna, pp. 53164.

265
Groundwater studies
HANTUSH, M. S., 1963. Hydraulic of Wells. In: V. T Chow (rd.), Advances in Hydroscience, Vol. 1,
pp. 281432.
HO, D. H.; ARANYOSSY, J. F.; LOUVAT, D.; HUA, M. Q.; NGUYEN, T. V.; NGUYEN, K. C., 1991.
Environmental Isotope Study Related to the Origin, Salinisation and Movement of
Groundwater in the Mekong Delta (Vietnam). In: Isotope Techniques in Water Resources
Development 1991, International Atomic Energy Agency, Vienna, pp. 41528.
INTERNATIONAL ATOMIC ENERGY AGENCY, 1983. Isotope Techniques in The Hydrogeological Assessment
of Potential Sites for The Disposal of High-Level Radioactive Wastes. Technical Report Series
No. 228, International Atomic Energy Agency, Vienna, 151 pp.
INTERNATIONAL ATOMIC ENERGY AGENCY, 1989. Guide to Safe Handling of Radioisotopes in Hydrology.
IAEA Safety Series.
IVANOVICH, M.; FROEHLICH, K.; HENDRY, M. J., 1991. Uranium-Series Radionuclides in Fluids and
Solids, Milk River Aquifer, Alberta, Canada. Appl. Geochemistry, Vol. 6, pp. 40518.
IVANOVICH, M.; HARMON, R. S. (eds.), 1992. Uranium-Series Disequilibrium: Applications to Earth,
Marine, and Environmental Sciences. 2nd ed., Clarendon Press, Oxford, 907 pp.
JORDAN, H.; FROEHLICH, K., 1990. Groundwater Dating in Monitoring Groundwater Resources.
In: Groundwater Monitoring and Management, IAHS Publ. No. 173, pp. 13951.
KAUFMAN, W. J.; ORLOB, G. T., 1956. Measuring Groundwater Movement with Radioactive and
Chemical Tracers. J. Am Water Works Ass., Vol. 48, pp. 55972.
KELLEY, C. A.; HAMMER, B. T.; COFFIN, R. B., 1997. Concentrations and Stable Isotope Values of
BTEX in Gasoline-Contaminated Groundwater. Environ. Sci. Technol., Vol. 31, pp. 246972.
KENDALL, C., 1998. Tracing Nitrogen Sources and Cycles in Catchments. In: Isotope Tracers in
Catchment Hydrology. Elsevier, Amsterdam, pp. 51976.
KNUTSSON, G.; FORSBERG, H. E., 1967. Laboratory Evaluation of 51Cr-EDTA as a Tracer for
Groundwater Flow. In: Isotopes in Hydrology, International Atomic Energy Agency, Vienna,
pp. 62952.
KROUSE, H. R., 1980. Sulphur Isotopes in Our Environment. In: P. Fritz and J. Ch. Fontes (eds.),
Handbook of Environmental Isotope Geochemistry, The Terrestrial EnvironmentA, Vol. 1.
Elsevier, Amsterdam, pp. 43571.
KULKARNI, K. M.; KULKARNI, U. P.; SHARMA, S., 1999. Application of Environmental Stable
Isotopes in Mine Hydrology: A Field Example. In: Proc. 8th ISMAS Symposium on Mass
Spectrometry, Vol. II, Indian Society for Mass Spectrometry, Mumbai, pp. 6525.
KULKARNI, K. M.; NAVADA, S. V.; NAIR, A. R.; RAO, S. M.; SHIVANNA, K.; SINHA, U. K.; SHARMA, S.,
1998. Drinking Water Salinity Problem in Coastal Orissa, India Identification of Past
Transgressions of Sea by Isotope Investigation. In: Isotope Techniques in The Study of
Environmental Change, International Atomic Energy Agency, Vienna, pp. 293306.
KULKARNI, K. M.; NAVADA, S. V.; RAO, S. M.; NAIR, A. R.; KULKARNI, U. P.; SHARMA, S., 1996. Effect
of the Holocene Climate on the Composition of Groundwater in Parts of Haryana, India
Isotopic Evidence. In: Isotopes in Water Resources Management, Vol. II, International Atomic
LAL, D.; PATERS, B., 1962. Cosmic Ray Produced Isotopes and Their Applications in Problems of
Geophysics. Progr. Elementary Particle Cosmic Ray Phys., Vol. 6, pp. 174.
LENDA, A.; ZUBER, A., 1970. Tracer Dispersion in Groundwater Experiments. In: Isotope Hydrology
1970, International Atomic Energy Agency, Vienna, pp. 61941.
LEVIN, I., 1978. Regionale modellierung des atmosphrischen CO2 auf grund von C-13 and C-14
messungen, Diplomarbeit, Institut fr Umweltphysik, Univ. Heidelberg.
LIBBY, W. F., 1965. Radiocarbon Dating. University of Chicago Press, Chicago, Illinois, 175 pp.
LIU, B.; PHILLIPS, F. M.; HEINES, S.; CAMPBELL, A. R.; SHARMA, P., 1995. Water Movement in Desert
Soil Traced by Hydrogen and oxygen Isotopes, Chloride and Chlorine-36, Southern
Arizona. J. Hydrol., Vol. 168, pp. 91110.

266
LOVE, A. J.; HERCZEG,A. L.; LEANEY, F. W.; STADTER, M. F.; DIGHTON, J. C.; ARMSTRONG, D., 1994.
Groundwater Residence Time and Palaeohydrology in the Otway Basin, South Australia:
2H, 18O, and 14C Data. J. Hydrol., Vol. 153, pp. 411 42.
LOVE, A. J.; HERCZEG, A. L.; SAMPSON, R. G.; CRESSWELL, R. G.; FIFIELD, L. K., 2000. Sources of
Chloride and Implications for 36Cl Dating of Old Groundwaters, Southwestern Great
Artesian Basin, Australia. Water Resour. Res., Vol. 36, pp. 156174.
MAIRHOFER, H., 1963. Bestimmung der Strmungsrichtung des Grundwassers in einem einzigen
Bohrloch mit Hilfe radioaktiver Elemente. Atompraxis, Vol. 9, pp. 24.
MALOSZEWSKI, P.; RAUERT, W.; STICHLER, W.; HERMANN, A., 1983. Application of Flow Models in an
Alpine Catchment Area Using Tritium and Deuterium Data. J. Hydrol., Vol. 66, pp. 31930.
MANDEL, S.; BUGAYEVSKY, M.; BACHMAT, Y.; DROST, W., 1985. Estimation of Dispersivity and
Effective Porosity by a Single Well Tracer Technique. In: Scientific Basis for Water Resources
Management, IAHS, Wallingford, pp. 21929.
MARGRITA, R; GAILLARD, B., 1991. Use of Artificial Tracers for the Determination of Aquifer
Parameters. In: Use of Artificial Tracers in Hydrology, IAEA-TECDOC-601, International
MERCADO, A.; HALEVY, E., 1966. Determining the Average Porosity and Permeability of a Strat-
ified Aquifer with the Aid of Radioactive Tracers. Water Resour. Res., Vol. 2, pp. 52531.
MOSER H.; NEUMAIER, F.; RAUERT, W., 1957. Die Anwendung radioaktiver isotope in der
Hydrologie. Atomkernenergie, Vol. 2, pp. 22531.
MOSER H.; NEUMAIER, F.; RAUERT, W., 1963. New Experience with the Use of Radioactive Isotopes
in Hydrology. In: Radioisotopes in Hydrology, International Atomic Energy Agency, Vienna,
pp. 28395.
MOSER, H.; STICHLER, W., 1970. Deuterium Measurements on Snow Samples from the Alps. In:
Isotope Hydrology 1970, International Atomic Energy Agency, Vienna, pp. 4357.
MURPHY, E. M.; DAVIS, S. N.; LONG, A.; DONAHUE, D.; JULL, A. J. T., 1989. C-14 in Fractions of
Dissolved Organic Carbon in Groundwater. Nature, Vol. 337, pp. 1535.
MUSGROVE, M. L.; BANNER, J. L., 1993. Regional Groundwater Mixing and the Origin of Saline
Fluids: Midcontinent, United States. Science, Vol. 259, pp. 187782.
NAIR, A.; PENDHARKAR, R.; NAVADA, S. V.; RAO, S. M., 1979. Groundwater Recharge Studies in
Maharashtra Development of Isotope Techniques and Field Experience. In: Isotope Hydro-
logy 1978, Vol. II, International Atomic Energy Agency, Vienna, pp. 80326.
NICHOLLS, P. H.; BROMILOW, R. H.; ADDISCOTT, T. M., 1982. Measured and Simulated Behaviour of
Fluometuron, Aldoxycarb and Chloride ion in a Fallow Structured Soil. Pesti. Sci., Vol. 13,
pp. 47583.
OLSSON, I. U., 1968. Modern Aspects of Radiocarbon Dating. Earth Sci. Rev., Vol. 4, pp. 20318.
PAYNE, B. R.; QUIJANO, L.; LATTORE, C., 1980. A Case Study of Leakage between Aquifers in
Hermosillo, Mexico, Using Environmental Isotopes. In: Arid Zone Hydrology: Investigations
with Isotope Techniques, International Atomic Energy Agency, Vienna, pp. 11330.
PEARSON, F. J., JR; RIGHTMIRE, C. T., 1980. Sulphur and Oxygen Isotopes in Aqueous Sulphur
Compounds. In: P. Fritz and J. Ch. Fontes, (Eds.), Handbook of Environmental Isotope
Geochemistry, The Terrestrial EnvironmentA, Vol. 1, , Elsevier, Amsterdam, pp. 22758.
PICCIOTTO, E., DE MAERE, X. ; FRIEDMAN, I., 1960. Isotopic Composition and Temperature of
Formation of Antarctic Snows. Nature, Vol. 187, pp. 8579.
PLATA BEDMAR, A., 1981. Empleo de tcnicas isotpicas en el estudio de la problemtica de la
futura presa de San Marcos, Proyecto Tabacundo (Ecuador). Informe GANOP I-1A2/81,
CEDEX, Ministerio de Obras Pblicas, Madrid.
, 1983. Single Well Techniques Using Radioactive Tracers. In: Tracer Methods in Isotope
Hydrology, International Atomic Energy Agency, Vienna, pp. 1746.
PLUMMER, N. L.; MICHEL, R. L.; THURMAN, E. M.; GLYNN, P. E., 1993. Environmental Tracers for

267
Groundwater studies
Age Dating Young Groundwater. In: W. D. Alley (ed.) Regional Groundwater Quality, Van
Nostrand Reinhold, New York, pp. 25594.
PLUMMER, N. L.; PRESTEMON, E. C.; PARKHURST, D. L., 1994. An Interactive Code (NETPATH) for
Modelling Net Geochemical Reactions Along a Flow Path, Version 2.0. U.S. Geological
Survey Water Resources Investigations Report 94 4169, U.S.G.S., Reston, VA.
RAMSPACHER, P.; ZOJER, P. H.; FROEHLICH, K.; STICHLER, W., 1992. The Recharge of Large Springs
from a Carbonate Aquifer Near Innsbruck Applying Environmental Tracers. In: H. Hoetzl
and A. Werner (eds.), Tracer Hydrology. A. A. Balkema, Rotterdam, pp. 2517.
ROBERTSON, W. D.; CHERRY, J. A., 1989. Tritium as an Indicator of Recharge and Dispersion in a
Groundwater System in Central Ontario. Water Resour. Res., Vol. 25, pp. 1097109.
RAO, S. M.; KULKARNI, K. M., 1997. Isotope Hydrology Studies on Water Resources in Western
Rajasthan. Current Science, Vol. 72, pp. 5561.
ROZANSKI, K.; ARAGUS, L.; GONFIANTINI, R., 1993. Isotopic Patterns in Modern Global Precip-
itation. In: Climate Change in Continental Isotopic Records, American Geophysical Union,
Washington DC, pp. 136.
SCHLOSSER, P.; STUTE, M. C.; SONNTAG, C.; MNNICH, K. O., 1989. Tritiogenic 3He in Shallow
Groundwater. Earth Planet Sci. Lett., Vol. 94, pp. 24556.
SIEGEL, D. I., 1990. Sulphur Isotope Evidence for Regional Recharge of Saline Water During
Continental Glaciation, North-Central United States. Geology, Vol. 18, pp. 10546.
, 1991. Evidence for Dilution of Deep, Confined Groundwater by Vertical Recharge of
Isotopically Heavy Pleistocene Water. Geology, Vol. 19, pp. 4336.
STUTE, M.; SCHLOSSER, P.; CLARK, J. F.; BROECKER, W. S., 1992. Paleotemperatures in the South-
western United States Derived from Noble Gas Measurements in Groundwater. Science,
Vol. 256, pp. 10003.
STUTE, M.; TALMA, A. S., 1998. Glacial Temperatures and Moisture Transport Regimes
Reconstructed from Noble Gases and d18O, Stampriet Aquifer, Namibia. In: Isotope
Techniques in the Study of Past and Current Environmental Changes in the Hydrosphere and
Atmosphere, International Atomic Energy Agency, Vienna, pp. 30718.
SZABO, Z. D.; RICE, E.; PLUMMER, L. N.; BUSENBERG, E.; DRENKARD, S.; SCHLOSSER, P., 1996. Age
Dating of Shallow Groundwater with Chlorofluorocarbons, Tritium/Helium-3, and Flow
Path Analysis, Southern New Jersey Coastal Plain. Water Resour. Res., Vol. 32, pp. 102338.
TANDIA, A. A.; ARANYOSSY, J.-F.; GAYE, C. B., 1993. Evaluation de la recharge des nappes
phratiques sous climat semi-aride. Exemple du Nord-Sngal. In: Les ressources en eau au
Sahel, IAEA-TECDOC-721, International Atomic Energy Agency, Vienna, pp. 3951.
TAYLOR, C. B.; ROETHER, W., 1982. A Uniform Scale for Reporting Low-Level Tritium Measure-
ments in Water. Int. J. Radiat. Isotop., Vol. 33, pp. 377 82.
TAZIOLI, G. S., 1977. Groundwater Circulation in the Coastal Carbonate Aquifers of Apulia. In:
Hydrodynamic Diffusion and Dispersion in Porous Media, IAHR, Pavia, pp. 18799.
VRBKA, P.; JACOB, H.; FROEHLICH, K.; SALIH, M. K., 1993. Identification of Groundwater Recharge
Sources in Northern Sudan by Environmental Isotopes, J. Environment. Hydrol., Vol. 1,
pp. 816.
TORGERSEN, T.; HABERMEHL, M. A.; PHILLIPS, F. M.; ELMORE, D.; KUBIK, P.; JONES, B. G.; HEMMICK, T.;
GROVE, H. E., 1991. Chlorine-36 Dating of Very Old Groundwater, III. Further Studies in
the Great Artesian Basin, Australia. Water Resour. Res., Vol. 27, pp. 320114.
WASSENAAR, L. I.; ARAVENA, R.; FRITZ, P., 1992. Radiocarbon Contents of Dissolved Organic and
Inorganic Carbon in Shallow Groundwater Systems: Implications for Groundwater
Dating. In: Isotope Techniques in Water Resources Development 1991, International Atomic
WEINBERGER, W. A.; ELLIS, W. R.; SEATONBERG, B. W.; ANDREW, J. T. G., 1967. Radioisotopes as
Groundwater Tracers. J. Geophys. Res., Vol. 72, pp. 408191.

268
WOODCOCK, A. H.; FRIEDMAN, I., 1963. The Deuterium Content of Raindrops. J. Geophys. Res.,
Vol. 68, pp. 4477 83.
WURZEL, P.; WARD, P. R. B., 1965. A Simplified Method of Groundwater Direction Measurement
in a Single Borehole. J. Hydrol., Vol. 3, pp. 97105.
ZAK I.; GAT, J. R., 1975. Saline Waters and Residual Brines in the Shiraz-Sarvistan Basin, Iran.
Chem. Geol., Vol. 16, p. 179.
ZIMMERMANN, U.; MUNNICH, K. O.; ROETHER, W.; KREUTZ, W.; SCHAUBACH, K.; SIEGEL, O., 1966.
Tracers Determine Movement of Soil Moisture and Evapotranspiration. Science, Vol. 152,
pp. 3467.
ZIMMERMANN, U.; MUNNICH, K. O.; ROETHER, W., 1967. Downward Movement of Soil Moisture
Traced by Means of Hydrogen Isotopes. Am. Geophys. Union Monogr., No. 11, pp. 2836.
ZUBER, A., 1986. Mathematical Models for the Interpretation of Environmental Radioisotopes in
Groundwater Systems. In: P. Fritz and J. Ch. Fontes (Eds.), Handbook of Environmental
Isotope Geochemistry, The Terrestrial Environment - B, Vol. 2. Elsevier, Amsterdam, pp. 159.
269
10 H y d ro g e o l o g i c a l m a p p i n g
10.1 Introduction
The hydrogeological map has three main functions:

1. It is a tool for visualising and understanding the regional hydrogeological conditions
observed or deduced from various investigations and analyses. Insight into these con-
ditions is not obtained as instantly as insight into the conditions controlling surface
water. It requires the combination of a great amount of widely dispersed data of
various kinds, whose analysis is largely a matter of interpretation. Therefore mapping
is an important tool for coherently arranging coherently the available spatial infor-
mation; furthermore, it is an essential means of increasing our insight.
2. It is a means of communication between hydrogeologists, in the way the geological
map is for geologists. It is an indispensable supplement to monographs and, long
before the era of computerised information systems, it was already a veritable
pictorial database that was immediately and easily accessible. It also became a valued
teaching aid.
3. It is a means of communication between the hydrogeologist and various kinds of non-
specialists who require information about groundwater and its users.
Consequently, hydrogeological mapping serves two purposes: scientific and utilitarian.
10.2 The role and place of hydrogeological mapping
10.2.1 Maps among the methods of graphic representation
Hydrogeological maps have a central position among the methods of graphic presentation. They
depict only what can be presented in two horizontal dimensions in a given territory (Fig. 10.1).
By definition a map depicts space not time. However, it is not devoid of temporal
references: it represents the state of knowledge at a given moment, or data on variables over
time. The map may depict a situation at a given date, in principle the same for the whole map if
the variable is spatio-temporal (for example, piezometric levels within a single body of ground-
water). Alternatively it may depict several situations at different dates, so they can be compared;
or it may depict a mean situation for a specified reference period. Sometimes a cartogram can be
produced by superimposing on the map the graphs depicting a given variable at various
observation points.

271
Groundwater studies
Figure 10.1 Spatial graphics
10.2.2 Maps among other methods of storing and representing information
In presenting data or representing more complex and synthetic information, maps are closely
linked to other tools. They supplement data banks or databases, arrays of information and
simulation models (Fig. 10.2).
(a) Relation to data banks
The analytical map that shows point or line data precedes the data bank by delivering
entries to it, e.g. coordinates, digitisation, etc.; and follows it by presenting data bank
output, e.g. maps of selected points, and isolines of regional variables drawn by appro-
priate interpolation software.
(b) Relation to arrays of information
A map is especially useful for the analytical presentation of the water budget of an aquifer
system: notably to present the reference area, or its subdivisions, and the flow conditions
at the various boundaries.

272
Hydrogeological mapping
Figure 10.2 Relation between the tools and the modes of expression
(c) Relation to hydrodynamic simulation model

The input data for a model are mainly prepared as parametric maps that are derived from
data banks, and maps showing the delimitation of areas satisfying specific conditions that
may be explicit transformations of classic hydrogeological maps. The number of maps
increases according to the structural complexity of the studied aquifer system and the
complexity of the simulated phenomena. Ultimately, the output of the model will also be
expressed as maps.
10.2.3 Role of mapping in the development of hydrogeological insight into an area
Hydrogeological mapping plays different roles at two distinct stages in the accumulation of
knowledge:
At the stage of preliminary reconnaissance the collection of the data for the hydrogeo-
logical map is part of what is usually described as the inventory work. The require-
ments for the preparation of each map sheet dictates the framework and the work
programme of the fieldwork including field mapping and underground investigations.
The results are presented mainly (but not exclusively) as either reconnaissance maps or
as detailed regular maps that systematically cover the area.
At a more advanced stage, the hydrogeological maps are drawn using also the results
of local and regional studies, wellfield data and results from underground engineering
work, rather than data from systematic prospecting. It is a matter of synthesis and inte-
gration of scattered data. Therefore the hydrogeological maps are more a synthesis of
the recorded data than of a systematic coverage. They may be a part of national or
international syntheses that will serve scientific or didactic purposes, or they will
present a particular macro-economic orientation.

273
Groundwater studies
10.3 What can and must hydrogeological maps depict?
Hydrogeological maps show the data and information that can be rendered on a map. The
subjects of hydrogeological mapping are many and varied, and a list of them reads like the
menu of a complex theoretical cartographic legend.
10.3.1 Inventory
(i) First bifurcation, reflecting the subject and scope of hydrogeology, distinguishes two subjects, both
of which can be described and depicted cartographically:
(a) Rocks:
the nature, geometry and structure of geological formations, seen from the viewpoint
of their role vis--vis the presence and circulation of water in relation to its properties,
e.g. groundwater reservoirs or aquicludes,
formations that to some extent transmit or store water,
formations in contact with or isolated from the surface conditions.
The hydrogeological characteristics can be presented in different ways, e.g. on a lithological scale
portraying the nature of the rock, or on a geological scale with emphasis on the regional struc-
tures. Moreover they can be presented qualitatively (types) or quantitatively (parameters).
(b) Groundwater:
its presence, extent, and abundance,
its hydrodynamic character (hydraulic head or potential; flow direction; flow velocity),
its chemical and thermal characteristics, and
its hydraulic connection with surface water.
These characteristics can be described quantitatively or on various scales.
(ii) Artefacts
It is usually desirable to include on the map structures and sites that have a relationship to
naturally occurring groundwater or water in general. This may include:
ad hoc investigation sites (e.g. test holes),
permanent observation structures (piezometers and hydrometric stations),
other data-gathering stations,
exploitation structures (e.g., wells, boreholes, springs, injection sites, underground
engineering works) and
any sites of surface activity likely to influence the flow regime or the quality of the
groundwater (irrigation, slurry-spreading, drainage, etc.).
(iii) Typology and classification
Many of these items are characterised by qualitative, semi-quantitative (ordinal) or quantitative

typologies or classifications:
water-bearing/non-water-bearing rocks, or categories of permeability and transmis-
sivity;
porous or discontinuous formations (fissured, fractured, karstic);
different conditions at the boundaries of groundwater systems, or involving the
relationships between groundwater and surface water;

274
types of springs or categories of average rate of flow; hydrochemical or water-quality

categories;
types of construction.
(iv) Data and information
A whole range of information may be presented on maps: from directly observed relatively
simple field data to highly complex, processed and synthesised information geared to fore-
seeable questions. It is not always possible to show all these on the same map, however (see
Table 10.1).
Table 10.1 Mappable subjects, classified by nature and degree of conceptualisation
Subjects. Degree of Hydrogeological formations

conceptualisation and structures Groundwater Artefacts
Kind of outcropping Location of observation stations Location of

formations. (x, y, z). Depth of water table well, borehole,
Basic factual data Hydrogeological contour. drain. Depth
(measurements, Local depth of reservoir. of structure
observations) Upper or lower confining
bed. Local thickness of a
formation. Type of
overlying formation
Masked hydrogeological Local hydraulic potential (dated,

contours. Local elevation average). Variation of hydraulic
Results of simple of a reservoir's upper or head. Rate of flow (average).
data processing lower confining bed. Lines of equal potential
(simple calculations Isohypse of a formation's (piezometric surface)
and interpolation) upper or lower boundary.
Lines of equal thickness
(isopachs)
Hydrodynamic parameters Boundary of groundwater. Specific.

of the formation: Isobaths. Equal ionic or total Discharge
permeability, porosity, salinity lines (isohalines).
transmissivity (zoning; Boundary of unconfined/confined
surface conditions groundwater body. Site of
Results of complex controlling infiltration potential artesian flow. Flow paths
data processing (type and thickness of and mean local flow velocities.
superficial unsaturated Groundwater flow divides.
formations).Structural Degree of connection between
class (according to the groundwater body and river.
conditions at depth) Local fluxes or fluxes per inflow
and outflow sector for a defined
groundwater body
Accessibility. Probability of Water quality (normative), Probable

successful boring. Zones of distribution. Vulnerability to productivity
maximum productivity (in pollution. Sensitivity to changes or injectivity.
Information a multi-layered system) in surface conditions. Degree Withdrawal
of transformation of present state rate
as compared to a natural state
(e.g. depression cones)

275
Groundwater studies
10.3.2 Data of field surveys and subsurface investigations
Only part of the data needed for map-making can be collected by field observations, hydrogeo-
logical surveys (e.g. well inventory, groundwater levels, EC values, pumping data, spring inven-
tory, surface water discharge) and mapping of contours, in combination with the interpretation
of air photographs and remote sensing images. The additional information comes from the
interpretation of underground reconnaissance data, e.g. structural and hydrodynamic data from
drilling logs, data on the analysis of water and rock samples. The latter is usually mainly entails
compiling data from reports and documents
10.3.3 Shapes and dimensions
Geometrical, mappable objects can be clustered into three formal groups distinguished essen-
tially by the possibility of depicting them together on the same map:
points that can be represented by symbols,
linear (on the map) objects, and
zonal, two-dimensional (at least on the map) objects (see Table 10.2).
Table 10.2 Clustering of mappable objects
Subjects Formations and

Dimensions hydrogeological structures Groundwater Artefacts
Points Reference point. Spring. Well.

Location Swallow hole. Borehole.
Reference point Piezometer.
Hydrometric
station.
Groundwater
dam
Lines
Actual Hydrogeological Underground channel (karst). Drain.
contour. Watercourse; Littoral zone Gallery.
Line of tectonic (as specified limit) Canal.
disturbance, Buried
e.g. fracture, fault pipelines
Virtual Contour lines of structural Isopleths of hydrodynamic

variables, e.g. isohypses, variables, e.g. equipotentials,
isobaths, isopachs. lines of equal fluctuation,
Boundaries, e.g. of a concentration and depth.
zone of the same class Divide.
of transmissivity Streamlines (vectors).
Boundary of groundwater
body/confinement
Zones or regions Zones belonging Extent of defined ground-water Irrigated or

to a specific class, body. Area of artesian flow. Area drained area.
e.g. type of formation, of outflow or seepage. Zones Intensive
transmissivity, belonging to a specific class, e.g. farmed area
deep structure, average recharge, flow regime,
accessibility hydrochemical characteristics,
water quality

276
While points and lines or families of (differentiated) lines can to a large extent coexist on
the same map by virtue of the variety of possible modes of representation (shape, colour, etc.), it
is much more difficult to superimpose zonal representations, hence the need in group 3 to
choose a priority subject.
10.3.4 Mapping programme
The hydrogeological data and information that can be expressed on a map are not equally
significant whatever the scale. The scale is therefore in itself a technical selection factor, but it
may also be dictated by the choice of the principal subject to be mapped. Nevertheless, even at a
given scale, the abundance of subjects may be such that juxtaposing them would result in a map
so overloaded as to be unintelligible. A selection must therefore be made and the possibility and
advantage explored of distributing the subjects over several maps; thus programme selectivity
and a plurality of maps with distinct and complementary programs go together.
A mapping programme must try to reconcile:
the essential data and information that must be presented,
the groups of data or information that are complementary and mutually enlightening,
in particular couplings of data relating the terrain to the groundwater,
consistency of the degree of abstraction: priority being given to objective descriptions
with minimum interpretation or to the presentation of more complex and elaborate
scientific or practical information,
compatibility of presentations (linked to the dimensions of the subjects).
Finally, the need for a clear and understandable presentation limits the amount of information
that can be shown.
10.3.5 Hydrogeological map sensu stricto
This should satisfy three principles:
(i) Scientific orientation (cognitive)
The hydrogeological map must have a scientific content and therefore gives priority to an
objective and explicit presentation of the hydrogeological conditions. This does not exclude the
presentation of interpolations and structural hypotheses, Of secondary importance the pre-
sentation of derived information that the hydrogeologist can infer or that should be expressed
on supplementary maps. The map therefore will show:
hydrolithological classes, associated with average estimates of permeability or trans-
missivity rather than with productivity figures;
isohypses of significant surfaces such as the upper or lower confining bed of an aquifer
rather than isobaths;
the equipotential contours of the piezometric surface, rather than isobaths;
classes of salinity of groundwater rather than of normative quality.
(ii) Balance in the presentation of the hydrogeological structures and the groundwater characteristics
The hydrogeological map must represent the data on the aquifers, including discontinuities
and gaps, and the data on the dynamics and the physical or chemical characteristics of the
groundwater they contain. They must explicitly render the relationships between content and
container where those relationships are not obvious.

277
Groundwater studies
(iii) Balance between local characteristics and regional hydrogeological conditions
The hydrogeological map must represent both the spatial distribution of different continuous or
discontinuous variables of local significance such as the nature of surface formations or the
potential of a groundwater body and structural characteristics (contours and classes) giving
visual expression to hydrogeological structures that are significant at the given scale. A
hydrogeological map must in particular make it possible to identify the configuration and
boundary conditions of the principal water-bearing systems of the area it covers.
10.3.6 Presentation of surface water
Three options may dictate the choice of the surface hydrological data that will be presented on a
hydrogeological map:
(i) Minimum geographical option
The hydrographic network is shown on the topographic base map for planimetric purposes
only; it is presented in the traditional blue color. The hydrometric stations are shown as well,
including relevant data about surface runoff (average rate of flow, areas of upstream basins,
number of years of measurement, etc.).
(ii) Hydrological option
Watercourses are classified according to their average flow, deduced by interpolating the data
from the hydrometric stations, or by the base flow that shows mainly shows the magnitude of
the groundwater runoff collected by the watercourses.
(iii) Hydrogeological option
Presentation of watercourses according to the degree of hydraulic continuity with the under-
lying aquifers and the kind of interaction with the groundwater (whether draining, infiltrating,
none at all, etc.), if desired, combined with quantitative data on the interactive flows. This option
is suitable for the explicit mapping of aquifer systems and their boundary conditions. It requires
more complex, less classical modes of representation that differentiate shapes and sometimes
colours.
10.3.7 Specific purpose maps
Specific purpose groundwater maps are derived from the hydrogeological map and give
priority, or are exclusively devoted to the information required to respond to the identified
needs and questions of specific users (Table 10.3).
10.4 Classification of hydrogeological maps
The variety and abundance of mappable data means that they must be sorted and distributed
among several maps. The difference between the purposes for which the maps are made offer
criteria for the choice of a mapping programme. Moreover, the diversity of scales is at once a
cause and an effect of their multiplicity. Programme, purpose and scale are thus the essential
base of any classification of hydrogeological maps in the broad sense (Table 10.4).

278
Table 10.3 Users and information requirements
Users Useful mappable information
Planners of groundwater-capture Presence of aquifer: depth and thickness of the aquifer,

structures (wells and boreholes), depth to water.
drainage or underground Productivity: probable specific rate of flow, pumping level.
engineering work. Estimate of the zone of influence (calculation parameters).
Users of groundwater Probability of successful drilling.
Water quality
Expert on protection Conditions of recharge; extent of recharge area.

of capture infrastructure Direction of surface runoff and groundwater flow.
Proximity of active works upstream.
Protective capacity of the soil
Land use impact expert Water table elevation and depth below surface. direction
of runoff. groundwater flow trajectory.
Protective capacity of the soil cover (factors affecting
vulnerability to pollution)
Resource assessors; Extent of groundwater reservoir and boundary conditions.

Public administrators; Average inflow per unit of area; zones.
Planners Aquifer/stream connections.
Distribution of parameters determining accessibility
and probable exploitation costs.
Groundwater quality zones
10.4.1 Programme and purpose
Programme and purpose obviously go together. Similarly, the communication purpose is

inseparable from the users and their information needs. Those needs are explicitly expressed by
the users or they are attributed to them; that is, the supply of information may anticipate unfor-
mulated needs or it may proceed from a desire to inform. The simplest grouping of the users
(see also Table 10.3) is:
(a) specialists and professionals, e.g. hydrogeologists, hydraulic engineers and hydro-
logists, whose background is identical to that of those who designed and drew the
map. They require a mapping programme that gives priority to descriptive and objec-
tive data, explicitly separating fact and interpretation;
(b) the users of groundwater and other underground resources and underground space.
They include:
the users of groundwater for domestic and industrial purposes,
the users that may affect the groundwater by activities in the underground (borings,
underground infrastructural works, mines, quarries, hydraulic installations, etc.),
land owners undertaking construction work,
users of land and water resources for agricultural purposes (effects on the depth of
water, pollution risks),
environmental managers and planners and various intermediaries between
specialists and the public,
each with a need for a specific mapping programme that focuses on selected infor-
mation as material for replies to specific questions.

279
Groundwater studies
Table 10.4 Classification of hydrogeological maps
Scale
Purpose Large Medium and small
Hydrogeological map, Composite hydrogeological maps

detailed or 'regular'
Hydrogeological Provisional composite hydrogeological maps

reconnaissance map
Scientific Specialised hydrogeological Hydrogeological maps of basins and distinct

maps: parametric maps groundwater formations.
(structural, hydrodynamic, 'Geohydrodynamic' maps of groundwater systems.
piezometric, transmissivity, Hydrological parameter maps; maps of
etc.) underground flow; maps of aquifer inflows.
Maps of groundwater storage
Hydrogeochemical maps, Composite hydrogeochemical maps.

general or specific Maps of thermal or mineral waters
Works productivity map. Classification by various criteria

Maps of depth of groundwater of groundwater resources.
and of upper confining bed. Map of groundwater thermal potential
Practical Groundwater production
costs map
Map of general groundwater quality.
Specific quality map. Map of groundwater vulnerability
Map of groundwater to pollution in general
vulnerability to specific
pollution
10.4.2 Scale
Although above all a matter of convention, the usual definitions are:

large scale: 1:20,000 1:100,000;
medium scale: 1:200,000 1:500,000;
small scale: 1:1,000,000 and upwards.
10.4.3 Scientific maps and practical maps
The difference in purpose and therefore of the programmes has led to the design and pro-
duction of two families of maps:
(1) scientific hydrogeological maps in the strict sense of the term, whose primary purpose is
to represent natural conditions, and human activity and its effects in a given situation.
They are preceded by reconnaissance maps, supplemented by more specific maps (para-
metric, in particular) and even reduced for atlases.
(2) practical maps, whose principal purpose is to inform and whose programme must consist
of the information required to meet the assigned object. The utilitarian maps are the pro-
duct of selective interpretations of the scientific maps. The hydrogeological map is the
necessary basis of all practical maps, even though they may also use supplementary data.
The usual classification, presented in Table 10.4, shows these two families in relation to the usual

280
scales. The differences in scale correspond to differences in levels of information, detail of inter-
polations and classifications, and reliability and precision of local information available from the
map.
Table 10.5 reproduces Struckmeiers (1989) classification of maps that is more explicit
regarding levels of information and more technical in its distinction of uses.
Table 10.5 System for classifying hydrogeological maps (After Struckmeier, 1989)
Level of Low Advanced High

information (scarce and (+ systematic (+ hydrogeological
heterogeneous investigation systems analysis
data from programmes, more and groundwater
various sources) reliable data) models)
Possible use
Reconnaissance General hydrogeological Hydrogeological parameter Regional groundwater
and exploration map (aquifer map) maps (map sets, atlases) systems maps (conceptual
model representations)
Planning and Map of groundwater Graphic representation

development resource potential derived from Geographic
Specialised Information Systems
hydrogeological maps (maps, sections, block
(planning maps) diagrams, scenarios)
Management Map of groundwater

and protection vulnerability
Possible use static --------------------------------------- time-dependence -------------------------- dynamic

low ---------------------------------------------- reliability ------------------------------------ high
low ---------------------------------------- cost per unit area ------------------------------- high
large ------------------------------------ area represented ---------------------------------- small
Parameters of small ---------------------------------------------- scale --------------------------------------- large
representation
10.4.4 Terminology
The classification of maps according to different criteria is less an end in itself than a means of
highlighting their differences and specialisations. It is less important to classify maps according
to one system or another, than to give each type a clear name consistent with its content. Con-
sequently, it seems desirable to retain the name hydrogeological map (in a broad sense) as a generic
name, but to restrict the use of the name
hydrogeological map (sensu stricto) to maps that express scientific knowledge relating to
the geological structures that determine the occurrence and circulation of groundwater
and to the groundwater itself, in other words maps that describe and explain static and
dynamic hydrogeological conditions, and
groundwater resources maps to utilitarian maps with a more specific and practical
purpose.

281
Groundwater studies
10.5 Hydrogeological map making
10.5.1 The language of maps; properties and constraints
(i) Modes of expression
Six sorts of basic visual features may modify the signs that make up graphic language in general
and map language in particular (Bertin 1973):
size,
value (from light to dark),
shape,
grain (screen, pattern and ornament),
colour,
orientation (of axially asymmetric signs).
These features may be combined in a single sign: one shape may have several sizes or orien-
tations, a colour may have several values, etc. They are divided into two sorts:
The first are called quantitative or ordered and may express either:
(a) continuous gradations, of which size is the only absolute, direct mode of expression;
or
(b) quantitative or ordinal classifications, i.e. in a number of classes that are, however,
limited (by convention, value and size again).
The others, shape, grain, colour and orientation, are called qualitative, unordered or
selective and can express differences of type or category; e.g. colour in different shades,
and grain, with a large number of possible patterns. They can be continuous or not and
orientated or not and can in addition be differentiated by colour. They offer the
greatest range of options, particularly for filling in areas (zones).
(ii) Topology; features and dimensions
Map writing uses three types of signs, each with a different dimension:
points (dimensionless), located at x and y;
solid or broken lines (one dimension), which can be indicated in digital shape by sets of
points defined in terms of x and y;
zones expanses or areas (two dimensions), indicated by lines or contours or fuzzy
edges.
Five of the six visual features can modify the signs depicting these three types of features. Zones,
however, cannot have size, shape or orientation.
(iii) Construction; assembly and superposition
The production of a thematic map is a work of assembly and involves matching the signs that
express the data and information according to the adopted mapping programme. Therefore the
compatibility and superimposability of signs are essential selection criteria when the key is
prepared.
(iv) Degrees of perception
When the various visual features appear together they are not all perceived to the same degree.
The order of perceptibility of the signs must therefore match the order of importance of the
subjects, i.e. of the information to be presented.

282
Colour is the most immediately perceptible feature. It must therefore describe the most
important variables, especially for zones. In point or linear marking, only a few pure
colours are clearly discernible.
Value is, after colour, the most perceptible feature in zone marking, but it is barely
perceptible in point or linear marking.
Size is the most perceptible feature for points and lines.
Shape is less perceived, and its perception is subordinated to size. The shape of a point
sign is only noticed if the size exceeds 2 mm.
Grain is perceived first by its value and only secondly by its shape.
Orientation is the least immediately perceptible feature.
(v) Recommendations
The range of a feature must express only one physical variable, qualitative or quan-
titative. Its gradation must be consistent, particularly if it is an ordered feature size,
value, or scale of value of the represented variable (e.g. rate of flow, permeability.
The qualitative, unordered features (colour, shape and grain) must express qualitative
types and not quantitative classes. For example, classes of transmissivity or flow
should not be represented by different colours or grain.
It is recommended that superposition of the various signs and patterns is tested before
deciding upon a key that presents them separately.
10.5.2 Key
A key or legend is a table that shows the correspondence between definitions of mappable
objects (meanings) and the signs that represent them. Two general principles help to make this
correspondence clearer:
obvious dimensional analogies: point, and line data (contours), and areas of equal
value relating to expanses are quite naturally represented by points, lines and zones,
respectively;
similar subdivisions between on the one hand, the material to be shown (e.g. areas of
distribution of hydrogeological conditions or classes of defined variables; primordial
boundaries), and read when consulting the map (information on various parameters or
qualitative features); and on the other hand, between primarily visible (e.g. zonal
markings) and primarily legible signs (points and lines).
(i) Rules
There are five rules to be observed when choosing a key:

Consistency of the degree of perceptibility (visibility and legibility) with the order of
importance of the presented data and information;
Consistency of the features of the signs with the represented subjects: eloquent corres-
pondence facilitates immediate understanding. In particular, discontinuities in lines
or grain (patterns) best express actual discontinuities in space or time; binary
oppositions (black/white, empty/full, positive/negative, etc.) are very appropriate for
expressing the many conceptual dualities in hydrogeology (aquifer/non-aquifer, free/
confined groundwater, fresh/salt water, etc.);
Intelligibility: the map must be clear, particularly if it has been designed for non-
specialist users. This calls for superimposability of signs and patterns;

283
Groundwater studies
Conformity with recommended standards, particularly those that are international or

tested by previous use;
The standardisation of hydrogeological map keys and their derivatives should, how-
ever, allow for some flexibility and the possibility of change and enrichment deriving
from increased knowledge and conceptual innovation;
Printability: practical difficulties that could arise at the drawing and publication stages
should be minimised.
(ii) Structure
The key to a hydrogeological map is generally structured in four sections devoted respectively
to:
hydrogeological formations and structures;
groundwater;
hydrography;
artefacts (works, observation stations, and locations of significant human activity).
(iii) General keys
A general key is both a universal map programme, a menu to be chosen from, and a dictionary
of corresponding conventional signs. The first aim of a national or international general hydro-
geological key is the standardisation of modes of expression; the establishment of a common
language between map publishers, rather than standardisation of map programmes. The variety
of programmes is still clearly subordinate to the diversity of conditions and the multiplicity of
purposes. Nevertheless, a common programme is needed where a regional or national territory
is to be mapped by different authors, and still more so where an international map is to be
made.
The most universal general key is the international hydrogeological map key developed
and adopted by the IAHS and the IAH in 1967 and subsequently approved and recommended
by UNESCO (1970); a revised second edition was published by UNESCO in 1983. A third revised
edition was included in the Hydrogeological Mapping Guide prepared by the IAH and
UNESCO and published in 1995 (Struckmeier and Margat, 1995).
(iv) Keys to individual maps
The key to a particular hydrogeological map is the expression of a given map programme and is
therefore selective; it should include only those signs actually used on the map. Preparing the
key and drawing the map are mutually interactive activities.
(v) Derived special-purpose maps
The keys to specific purpose maps that give information about groundwater resources with
more practical objectives (Section 10.4 and Table 10.4), must observe the same rules. They are not
standardised, but they must be particularly simple and intelligible as they are designed for non-
specialist users. They should be easy to produce and frequently updated.

284
10.6 Producing hydrogeological maps
10.6.1 Preliminary choice
The preliminary choices, area mapped, topographic base maps, and scale, are interdependent:
The area may be imposed by the topographical sheet, e.g. if the map will be part of the
regular large-scale or medium-scale coverage of a national territory or, on a smaller
scale, by the sheet of an international map. It may also correspond to an administrative
or political district (a province or region); this applies to any national map. Or it may
be adapted to a natural physical domain such as a sedimentary or river basin or an
individual aquifer.
The available topographic base maps often have to be supplemented or revised, in
particular as regards the drainage pattern, and simplified.
The choice of scale, dictated first and foremost by the maps purpose, which suggests
either a large or a small scale, is also limited by the scale of the available topographic
base maps.
10.6.2 Programming
The four main phases in the production of a hydrogeological map have been shown in the flow
chart of Figure 10.3.
(i) Data collection and drawing the hydrogeological map
From the outset drawing a hydrogeological map may or may not be one of the objects of
regional hydrogeological inventory and prospecting operations. In any case, it is at this stage
that useful basic data are collected (see section 10.2.3). But the actual drawing of a hydro-
geological map may include more complete, specific observations and sometimes investigations
of the sub-surface. This always includes research: the use of any existing documents, published
or unpublished (literature, maps, files and so on), including of course earlier hydrogeological
reports and studies, not forgetting technical archives (connected with prospection for oil or with
mining operations).
(ii) Data analysis and data preparation
The preparatory phase involves transferring all the data gathered or interpreted from fieldwork
(e.g. classified water points, classified and checked hydrographic detail, hydrogeological
contours) on to work-maps in each subject at the same scale as the map in preparation. This may
include the drawing of parametric maps, e.g. piezometric maps, maps of structural isohypses or
transmissivity maps. During this phase gaps and uncertainties may appear, necessitating further
fieldwork or reinterpretation of basic data.
(iii) Preparing the dummy
This central phase begins with the choice of the mapping programme and the preparation of the
draft key on the basis of the prepared information, among which a choice may have to be made
according to the purpose of the map. As work proceeds, alterations or improvements to the
programme and the key may be desirable. It is advisable to make the dummy as far as possible a
facsimile of the projected map as regards signs and colours, so as to be able to judge the

285
Groundwater studies
Figure 10.3 Flow chart of the production of a hydrogeological map
compatibility and perceptibility of the whole. Material shown in the margin (inset maps, sections
and diagrams) should also be prepared at this stage.
(iv) Publication
This phase is largely the responsibility of others and neither at the drawing stage nor at the
printing stage is it specific to the production of hydrogeological maps. The role of the hydro-
geologist and draughtsman is limited to checking and correction at two stages:
checking the drawing before printing;
examining and correcting the proofs, particularly with regard to colour.
These two phases can partially be merged by the use of computer-assisted mapping techniques
that directly memorise the contours of the original dummy.
10.6.3 Explanatory note
The explanatory note to a hydrogeological map should fulfil three functions:

Supplementing and explaining the definitions in the key, following the same order. In

286
particular, indicating the degrees of approximation variables for the same datum
according to zone and the sources, e.g. the basic data.
Giving instructions as to how to use the map; explaining the use of certain data or
information.
Providing additional data or information: this may be mappable but not represented
on the main map (to avoid overloading), but appearing on maps appended in the note,
or it may be unmappable, e.g. historical data, cross sections, or explanatory, e.g. par-
ticulars on methodology used.
Including a selection of publications on hydrogeological maps or keys to national or inter-

national maps and a number of hydro-ecological maps concerning the vulnerability of ground-
water to pollution.
ALBINET, M.; MARGAT, J., 1971. Cartographie de la vulnrabilit la pollution des nappes souterraines.
Prsentation de la carte de France 1:1,000,000. IAHS, XVth General Assembly, U.G.G.I.,
Moscow, Russia.
ALLER, L.; BENNETT, T.; LEHR, J. H. ET AL., 1985. DRASTIC: A Standardized System for Evaluating
Groundwater Pollution Potential Using Hydrogeological Settings. EPA/600/2-85/018, Govern-
ment of United States of America, Washington D.C.
ANDERSEN, L. J.; GOSK, E., 1989. Applicability of Vulnerability Maps. Environ. Geol. Water Sciences,
Vol. 13, pp. 3943.
BARROCU, G.; BIALLO, G., 1993. Application of GIS for Aquifer Vulnerability Evaluation; Application of
Geographic Information Systems in Hydrology and Water Resources Management. IAHS Publ.
N211.
BERTIN, J., 1973. Simiologie Graphique. Gauthier-Villrs, 2e d., Paris et La Haye, 431 pp.
BRGM, 1992. Cartographie Hydrogologique de lAfrique. Hydrogologie, N 1-2, ditions BRGM,
Orlans, 122 pp.
CASTANY, G., 1968. Prospection et exploitation des eaux souterraines (Chap. 13: Cartographie des
eaux souterraines). Dunod, Paris, France, 717 pp.
CHAMPAGNE, L.; CHAPUIS, R. P., 1993. valuation et cartographie de la vulnrabilit la pollution
des formations aquifres de la MRC de Montcalm selon la mthode DRASTIC. Sc. et Tech. de
lEau, Vol. 26, N3.
COLLIN, J. J., 1989. An Attempt to Evaluate the Socio-Economic Impact of Hydrogeological Maps.
IAH, Mem. Intern. Symp.: Hydrogeological Maps as Tools for Economic and Social Development,
Hannover, Germany, pp. 31931.
GALOFRE, A., 1976. Mapas hidrogeolgicos. In: E. Custodio y M. R. Llamas, Hidrologa subterrnea,
Ediciones Omega, Barcelona, pp. 154557.
GILBRICH, W. H.; HEINDL, L. A.; KELLER, R.; NEWSON, W. D.; ROGOVSKAYA, N. Y.; S. BUCHAN, (eds.)
1977. Hydrological maps. UNESCO/WMO, Paris/Geneva, 204 pp.
GRIMMELMANN, W. F., 1989. Hydrogeological Mapping in the Federal Republic of Germany. IAH,
Mem. Intern. Symp. on Hydrogeological Maps as Tools for Economic and Social Development,
Hannover, Germany, pp. 75 8.
GRIMMELMANN, W. F.; KRAMPE, K. D.; STRUCKMEIER, W., (eds.), 1986. Hydrogeological Mapping in
Asia and the Pacific Region. IAH Intern. Contr. to Hydrogeol., Vol. 7, 410 pp.
IAH, 1989. Mem. Intern. Symp. on Hydrogeological Maps as Tools for Economic and Social
Development, Hannover, Germany. IAH, Hannover, 598 pp.
KARRENBERG, H.; STRUCKMEIER, W., 1978. The Hydrogeological Map of Europe. Episodes, IUGS
Geol. News, 1978/4, pp. 168.

287
Groundwater studies
KHOURI, I.; DROUBI, A.; ZEBIDI, H.; ZERYOUHI, H.; I.; HAWA; A., 1982. ACSAD Legend for the Water
Resources Map of the Arab Countries. ACSAD, Damascus, W.R./p-31, 29 pp.
KOVALEVSKY, V. S.; MAKSIMOVA, N. G., 1989. Maps for Predicting the Variability of Groundwater
Resources in the Future. IAH., Mem. Intern. Symp. on Hydrogeological Maps as Tools for Econ-
omic and Social Development, Hannover, Germany, pp. 8391.
LAU, J. E., COMMANDER, D. P.; JACOBSON, G., 1987. Hydrogeology of Australia; Map at Scale
1:5,000,000 and Text. BMR Bull., No. 227, Canberra, 21 pp.
MARGAT, J., 1966. La cartographie hydrogologique. Chron. dHydrogol., BRGM, Paris, Vol. 9,
pp. 732.
, 1978. Nouvelle lgende de carte hydrogologique. Doc. BRGM 78 SGN 473 HYD, 45 pp.
, 1981. La carte hydrogologique de la France 1 : 1 500 000. Systmes aquifres. Bull.
BRGM (2) III, Orlans, pp. 989.
MARGAT, J.; ROGOVSKAYA, N., 1979. Cartographie des ressources en eaux souterraines. IAH, Intern.
Sympos. Vilnius, U.S.S.R, Rapport gnral, thme VIII, Mm. IAH, Vol. XV-2. Moscow (Doc.
BRGM 79 SGN 783 HYD J.M.)/Orlans, 18 pp.
PALOC, H., 1975. Cartographie des eaux souterraines en terrains calcaires. In: Hydrogeology of
karstic terrains, IAH publ., IUGS, Srie B, No. 3, pp. 13748.
SARIN, A., 1989. What is a Hydrogeologic Map and for What Does It Serve. IAH, Mem. Intern.
Symp. on Hydrogeological Maps as Tools for Economic and Social Development, Hannover,
pp. 3140.
STRUCKMEIER, W.; ENGELEN, G. B.; GALITZIN, M. S.; SHAKCHNOVA, R. K., 1986. Methods of Repre-
sentation of Water Data. In: G. B. Engelen and G. P. Jones (eds.), Developments in the Analysis
of Groundwater Flow Systems, IAHS Publ. 163, pp. 4763.
STRUCKMEIER, W. 1989. Types and Uses of Hydrogeological Maps. IAH, Mem. Intern. Symp. on
Hydrogeological Maps as Tools for Economic and Social Development, Hannover, pp. 1730.
STRUCKMEIER, W. F.; MARGAT, J., 1995. Hydrogeological Maps; A Guide and Standard Legend.
IAH, Intern. Contr. to Hydrogeology, Vol. 17. Heise, Hannover, 177 pp.
UNESCO, 1970. International Legend for Hydrogeological Maps. UNESCO/IAHS/IAH/Inst. Geol.
Sci., London, 101 pp.
UNESCO, 1983. International Legend for Hydrogeological Maps, Revised Draft. UNESCO Techn.
Document, SC-84/WS/7, Paris, 51 pp.
VAN DUIJVENBOODEN, W.; VAN WAEGENINGH, H. G., 1987. Vulnerability of Soil and Groundwater
to Pollutants. TNO Com. Hydrol. Research, The Netherlands, Proceed and Inform., N 38,
pp. 4336.
VARTANYAN, G. S.; SHPAK, A. A.; GALITSIN, M. S.; OSTROVSKY, V. N.; SHAKHNOVA, R. K.; L. A. OSTROVSKY,
1989. Main Conception and Experience of Hydrogeological Mapping Development in the
USSR. IAH, Mem. Intern. Symp. on Hydrogeological Maps as Tools for Economic and Social
Development, Hannover, pp. 26976.
VILLUMSEN, A.; JACOBSEN, O. S.; SONDERSKOV, C., 1983. Mapping the vulnerability of groundwater
reservoirs with regard to surface pollution. Danm. Geol. Unders. Arbog. 1982, Copenhagen,
pp. 2738.
ZAPOROZEC, A.; VRBA, J., (eds.), 1995. Mapping of Groundwater Vulnerability. IAH Groundwater
Protection Commission, IAH, Intern. Contrib. to Hydrogeology, Vol. 16. Heisse, Hannover.

288
A s s e s s m e n t o f g ro u n d w a t e r
11 r e s o u rc e s a n d g ro u n d w a t e r
regime forecasting
11.1 Introduction
An assessment of the groundwater resources of an aquifer system involves developing of a

quantitative understanding of the flow processes which operate within the aquifer. Three
features must be considered:
how water enters the aquifer system,
how water passes through the aquifer system and
how water leaves the aquifer system.
Groundwater regime forecasting involves a study of modifications to the flow processes due to
changes in any of these features; these changes may be natural or man-made.
The assessment of groundwater resources requires a detailed understanding of the aquifer
system. Information gained from the techniques described in the preceding chapters is vital for
the development of an understanding of the aquifer system. If conditions in the aquifer have
been modified due to the withdrawal of water from the aquifer and the effects of this exploita-
tion have been monitored over a number of years, it is easier to identify the groundwater flow
processes and to predict further changes. The key to the assessment of groundwater resources is
the formulation, which involves the description of the essential features of the flow processes.
Groundwater problems are complex in three-dimensional space and usually vary in time; it is
rarely possible to have complete information about all features such as the variations in the
component permeabilities within the aquifer system. The primary aim of the formulation is to
identify the essential features which strongly influence the flow processes; the following
references contain valuable information about the formulation of groundwater problems (Bear,
1979; Bouwer, 1978; Balek, 1989; Driscoll, 1986; Freeze and Cherry, 1979; Marino and Luthin,
1982; de Marsily, 1986; McWhorter and Sunada, 1977; Rushton and Redshaw, 1979; Todd, 1980;
Walton, 1991).
To quantify the flow processes a model is introduced. The model may be simple, such as
the use of Darcys Law (Chapter 1) in one-dimension, or it may be a complex three-dimensional
time-variant representation; in practice it is advisable to use both simple and complex models.
For complex problems a model in the form of a computer package is often used to obtain
numerical results. However, it is essential first to carry out a detailed formulation and then
choose a suitable package, rather than selecting a package and modifying the physical problem
to fit the assumptions and idealisations inherent in the package. This issue is similar to the
frequent misuse of methods of pumping test analysis (Chapter 8) when a type curve method is
selected which has a similar shape to the field results even though the conditions for which the
type curve is devised are very different from those in the field.

289
Groundwater studies
When historical responses have been represented adequately by a model, the model
can then be used to forecast the likely consequences of changes in inflows, conditions within
the aquifer system or outflows. Typical examples of forecasting include a study of changes in
the exploitation of the aquifer system and the effect on the water table, stream flows or quality
of the abstracted water. If the changes being investigated result in aquifer conditions which are
significantly different from the current conditions, predictions must be treated with caution
The stages in an investigation are as follows:
(i) collect together all the hydrological and hydrogeological information and data for
the catchment,
(ii) identify the important flow processes and confirm the magnitude of the com-
ponents by carrying out a preliminary flow balance,
(iii) develop a preliminary model to see whether the model reproduces the general
trends of the aquifer response,
(iv) in the light of the initial modelling results, review all the hydrogeological infor-
mation and field data and develop an improved understanding of the flow pro-
cesses. This may involve further field investigations and the use of the model to
investigate its sensitivity to changes in various parameters,
(v) produce a revised model,
(vi) repeat the stages listed above until the model reproduces hydrographs of ground-
water heads and flows to a prescribed accuracy,
(vii) the model can then be used for predictive purposes.
Even if the investigation does not make use of a computer model, most of the stages listed above
should be used in an investigation.
11.2 Identifying and quantifying groundwater resources
11.2.1 Formulation
The formulation of a groundwater problem involves identifying the critical flow processes
including the inflows and outflows. This section considers the formulation of the physical
problem; the formulation in mathematical terms is considered in Section 11.3.2. An inadequate
understanding of conditions in an aquifer may lead to unwise exploitation, with serious
consequences for the reliability of the aquifer model. Three case studies representative of most
groundwater problems have been selected, to introduce how a physical formulation is carried
out; these examples from India are outlined below and are subsequently used to illustrate
procedures throughout this chapter.
The groundwater resources of the Dulapally Basin in the hard rock area of Central India
have been exploited for many centuries using large-diameter dug wells in the weathered zone of
a granitic aquifer. In an attempt to increase the yield, boreholes were drilled into the underlying
fractured zone, Fig. 11.1, and good yields were obtained. It was assumed that a new aquifer had
been found, primarily because the yield of the fractured zone was good and there was little
effect on the weathered zone during test pumping. However, following the drilling of many pro-
duction boreholes in the fractured aquifers, yields fell off rapidly. Dewatering of the weathered
zone also occurred, leading to a severe reduction in yields of the original large-diameter wells.
In assessing the groundwater resources, the investigators had failed to identify the inter-
connection between the weathered and fractured zones. This example shows that a study of a
multi-layered aquifer must allow for the possibility of transfer between the layers.
A similar situation occurs in many alluvial aquifers. A typical example is provided by the
Mehsana alluvial aquifer, Fig. 11.2. Falling yields of shallow wells in the near surface aquifers

290
Assessment of groundwater resources and groundwater regime forecasting
Figure 11.1a, b, and c Example of weathered-fractured aquifer in India: (a) pumped and observation wells,
(b) general form of drawdowns for pumping test, (c) numerical model to represent
aquifer system
prompted the drilling of deep boreholes through clay zones to tap the deeper aquifers. Often
these original trial boreholes tapped a plentiful supply of water and it was possible to identify a
distant recharge zone which appeared to supply the deeper aquifers, Fig. 11.2a. However, as
extensive exploitation of the aquifer occurred ,with many production boreholes drilled into the
deeper aquifers, the yields of boreholes decreased. In addition, in the water table of the upper
aquifers dropped.

291
Groundwater studies
Figure 11.1d and e Example of weathered-fractured aquifer in India, (d) detailed comparison between field
and modelled results, (e) main flow mechanisms

292
Figure 11.2a and b Example of alluvial aquifer: (a) simplified diagram of aquifer system,
(b) history of exploitation of the aquifer system
293
Groundwater studies
Figure 11.2c and d Example of alluvial aquifer: (c) response of shallow and deep observation wells,
(d) change of flow mechanisms from before to after exploitation

294
Figure 11.2e Simplified model of the aquifer system
295
Groundwater studies
Figure 11.2f Comparison of field and modelled groundwater head fluctuations
The third example refers to a Miliolite Limestone aquifer in coastal Western India in
which excellent yields occur under high water table conditions but the yield reduces
significantly as the water table falls, Fig. 11.3. Another issue is that encroachment of saline water
occurs within two kilometres of the coast, Fig. 11.3a. Artificial recharge was proposed as a means
of flushing out the saline water and improving the overall yield of the aquifer system. However,
pilot schemes indicated that neither of these goals would be achieved.
These examples demonstrate that it is often difficult to identify the flow mechanisms
resulting from the exploitation of aquifers, since these mechanisms may not be operating when
the initial investigations are carried out. When preparing an assessment of the groundwater
resources, the various components of the flow processes must be considered. These are
discussed below.
11.2.2 Groundwater resource components
(i) Inflows
Recharge is the major inflow to most aquifers; methods of estimating recharge are discussed in
Lerner et al. 1990, Simmers 1997. Recharge may result from precipitation and also occur due to
return flow from irrigation. In arid areas, the main recharge may be from wadis which flow for
only a short period most years. Other sources of recharge are permanent water bodies (such as
lakes, rivers, canals, flooded rice fields) and leaking water mains or sewers. It is important to
distinguish between the potential recharge which is available at the ground surface and the
actual recharge which reaches the body of underground water; the actual recharge may be less
than the potential due to factors such as the presence of poorly permeable zones between the
ground surface and the permanent water table, or the inability of the aquifer to accept all the
potential recharge.

296
Figure 11.3 Example of Miliolite limestone aquifer: (a) typical cross-section through aquifer system,
(b) information about drilling, (c) assumed variation of hydraulic conductivity and transmissivity
with depth
297
Groundwater studies
Flows from adjacent strata can be another source of inflow, but it is essential to perform an
approximate water balance of the adjacent strata to ensure that there is sufficient water
available. It is preferable to deduce the inflow from a water balance rather than using a Darcys
Law calculation based on uncertain values of transmissivity and groundwater head gradient to
estimate the cross-boundary flows.
Water from storage is a significant input when the outflows from the aquifer exceed the
recharge and lateral inflows. When the outflows exceed the inflows, water can only be supplied
by a fall in the water table thereby releasing water from storage. If the recharge exceeds the
outflows, the excess water will be taken into storage, leading to a rise in the water table.
(ii) Outflows
Discharge areas can occur in the lower areas of a groundwater catchment; it can be difficult to
determine the actual flow in discharge areas because of the diffuse nature of the outflow, which
cannot be measured; often some of the discharge is lost to evaporation.
Flows to adjacent aquifers; comments are similar to those concerning inflows from adjacent
strata.
Rivers and springs can provide natural outlets for water. One technique for identifying the
groundwater flow into rivers is baseflow separation. However, for investigations using models it
is possible to derive a relationship from which the flow to or from a river can be established. The
outflow is usually proportional to the difference between the groundwater head in the aquifer
and the elevation of the river or spring. Rivers can also act as sources of inflow when the
groundwater head falls below the river level; a more detailed discussion can be found in the
section on Flow Processes.
Rejected recharge may be significant in areas where the surface permeability of the aquifer
is low or where the aquifer is already full of water. In many hard rock aquifers, the infiltration
capacity is limited and much of the potential recharge is rejected and enters drainage channels
and rivers.
Abstraction from an aquifer is the one outflow which can usually be quantified. Even if the
pumpage is not measured directly, reasonable estimates of the volume of water withdrawn from
the aquifer can be obtained from an examination of the domestic, industrial or agricultural use
of the water.
(iii) Flow processes within the aquifer system
Nature of aquifer: information about the nature of the aquifer combined with the location and
magnitude of the inflows and outflows allows a picture to be developed of the flow processes
within the aquifer system. For example, if there are poorly permeable zones in the vicinity of the
water table, this may restrict the recharge, while very permeable zones can increase the yield of
boreholes.
Influence of very permeable zones: very permeable horizontal zones allow rapid horizontal
flow towards a borehole or a spring. However, the question which needs to be asked is: from
where does the very permeable zone collect water? Frequently the very permeable horizontal
zones attract water by upward or downward flow through poorly permeable zones.
Influence of poorly permeable zones: significant vertical flows can occur through poorly
permeable zones even though their permeability is at least three orders of magnitude lower than
the very permeable zones. For the poorly permeable zones the total vertical flow can be
estimated from the product of a small vertical velocity and a large plan area, whereas for a very
permeable zone the higher horizontal velocity is multiplied by a far smaller vertical cross-
section.

298
Fissures, fractures and faults can all have a strong influence on the flow processes. Fissures
and fractures often enhance yield from a well or borehole. Because of the convergent flow
towards a pumped well or borehole, there are high velocities within ten metres of the borehole
and there is a significant groundwater head loss in this distance. The higher permeability of
fissures and fractures can reduce the head losses in the vicinity of the pumped boreholes. The
influence of faults is more difficult to quantify; in some circumstances a fault with a large throw
isolates parts of an aquifer system. However, a series of smaller faults may enhance the
permeability.
River/aquifer interaction can provide a route for water to leave an aquifer and increase river
flows if the groundwater level is higher than the river elevation, or a river can recharge an
aquifer if the groundwater level is below the river elevation. A form of river/aquifer interaction
is illustrated in Fig. 11.4; in the following discussion the groundwater head can refer either to the
water table close to the river or to the groundwater head within the aquifer adjacent to or
beneath the river.
11.2.3 Examples of identification and formulation of aquifer flow mechanisms
(i) Weathered-fractured aquifer
The workers who carried out the initial analysis of the pumping test in the Dulapally weather-
fractured aquifer assumed that the fractured zone is a separate aquifer from the weathered zone.
Consequently when tube-wells were drilled into the fractured zone and cased out through the
weathered zone, it was believed that horizontal flow through the fractured zone would
dominate. This view was supported by good yields during the pumping tests which were
obtained with relatively small drawdowns. Furthermore the Theis method was used to analyse
the observation well readings obtained from the pumping test; in this analysis the fractured
zone was represented as being confined, with no hydraulic contact with the weathered zone.
However, some of the field results from the pumping test were ignored. Figure 11.1b
shows that, in addition to the rapid fall of the piezometric heads in the observation wells in the
fractured aquifer, the shallow observation wells in the weathered zone respond slowly, with the
fall of the water table continuing after pumping stopped!
To investigate this observed response, a conceptual model (Fig. 11.1c) was developed in
which there is a possibility of flow between the fractured and weathered zones. In fact the
difference between the drawdowns in the fractured and weathered zones (Fig. 11.1b proves that
vertical flows do occur! The difference in the groundwater heads increases to a steady value of
about 0.8 m; when pumping stops the difference decreases only slowly, demonstrating that
vertical flows continue from the weathered zone to the fractured zone to refill the fractured
aquifer. This example shows that the difference in groundwater heads between two observation
wells at different depths within the aquifer system is of crucial importance in identifying the
aquifer flow mechanisms. The quantification of these flows can be achieved using a model as
outlined in section 11.3.4.
(ii) Alluvial aquifer
During initial investigations when deep trial boreholes were constructed into the deeper zones
of the Mehsana alluvial aquifer, the groundwater heads reflected the higher heads of the
common recharge zone, Fig. 11.2a. These higher heads led to the assumption that a new aquifer
had been identified. However, as many production boreholes were drilled and the deeper
aquifers became heavily exploited and the groundwater heads fell; it then became possible to

299
Groundwater studies
Figure 11.4 River/aquifer interaction: (a)-(c) effect of different relative positions between the water table
and the river level, (d) assumed river/aquifer interaction

300
identify the nature of the flow mechanisms. Figure 11.2c shows the hydrographs of a shallow
piezometer and a deep observation borehole. An examination of the results shows that:
a. the groundwater heads in the shallow piezometer rise during the recharge period
and fall during the dry season,
b. the heads in the deep piezometer fall when heavy pumping for irrigation occurs
but rise during the hot season of March to June, when few crops are grown, and
continue to rise during the monsoon season,
c. there is a distinctive vertical gradient between the shallow and deep aquifers, indi-
cating that vertical flows are significant and that the vertical resistance to flow is
high.
These results allow the important flow mechanisms to be identified, Fig. 11.2d. Diagram (i)
shows that when the trial boreholes were drilled, the lateral flow from the common recharge
zone was intercepted and a good yield could be obtained for a small pumped drawdown; this
only occurred because there was no other pumping from the aquifer. However, when a large
number of deep tube wells were drilled, the lateral flows in the aquifer were insufficient to meet
the demand and therefore vertical gradients were set up to draw water from the water table in
the upper aquifer, diagram (ii). Figure 11.2b illustrates how conditions changed with time within
the aquifer system.
Were the conclusions of the initial investigation wrong? With hindsight it can be seen that
the flow mechanisms within the aquifer system changed as abstraction increased, with large
vertical flows becoming the most important mechanism. However, this could not be identified
from the initial tests. When the conditions in an aquifer change from those during the testing
stage, predictions of future behaviour can be incorrect.
(iii) Coastal limestone aquifer
The key to the response of the limestone aquifer, Fig. 11.3a, is that there are major solution
channels which have developed in the limestone. Figure 11.3b provides information about the
drilling of a well in the limestone; major solution channels were identified and mud loss
occurred when the drilling reached the first solution channel. These solution channels allow
substantial quantities of water to be drawn into a pumped well; a discharge of 5200 m3/d was
achieved in a 0.45 m diameter well for a drawdown of 0.15 m. However, during the dry season
when the water table fell below the level of the solution channels, an abstraction rate of 500 m3/d
could not be maintained because of the reduction in transmissivity.
It is not possible to measure the effective permeability of the solution channels but the
postulated distributions of the hydraulic conductivities, shown in Fig. 11.3c, explain the large
difference in the well yields. When the well water level is at 5 m below ground surface the
transmissivity is 720 m2/d, but when the well water level is 15 m below ground level and below
the lowest fissure the transmissivity falls to 100 m2/d. This varying transmissivity with water
table elevation is often the most important factor in understanding the response of aquifers with
karstic features (Rushton and Raghava Rao, 1988).
11.2.4 Important questions when identifying resources
The purpose of this section is to list questions which should be considered when identifying the
flow mechanisms within an aquifer system prior to quantifying the resources.
a. which flows are more important, horizontal flows, or vertical flows, or both combined?
b. does the aquifer respond quickly to recharge events or changes in abstraction (i.e.
within one or two months) or is the response much slower?

301
Groundwater studies
c. is there a significant change in the saturated depth (a change of more than 10% is
usually significant)?
d. does the actual recharge reaching the water table equal the potential recharge from
below the soil zone?
e. does the lowering of the water table cause the recharge to increase or decrease?
f. are conditions within the aquifer system likely to change with time (such as the
dewatering of the upper part of the aquifer or the cessation of groundwater flow into
streams and rivers)?
11.2.5 Selection of the co-ordinate system
A further important issue is the selection of the co-ordinate system to be used in describing the
problem. A number of alternatives are listed below, in each case the groundwater head is
expressed as a function of certain co-ordinates, z is always the vertical coordinate positive in the
upwards direction.
h(x), because the flow is predominantly in one horizontal direction, the flow is only a function of
the horizontal coordinate (this will not occur if boreholes are used to abstract water from an
aquifer).
h(r), radial flow; this assumption can apply to flow towards a borehole in a confined aquifer.
h(x,y), regional groundwater flow formulation; even though the groundwater head is a function
of the horizontal x-y coordinate system, vertical components of flow are indirectly included,
provided the length to thickness ratio of the aquifer system is greater than five and there are
no poorly permeable strata within the aquifer system (Connorton, 1985; Rushton and Rathod,
1985).
h(x,z), flow in a representative vertical section; this approach is important for problems such as
the flow through earth dams.
h(r,z), combined radial and vertical flow which occurs around a pumped borehole in a layered
aquifer.
h(x,y,z), this is a full three-dimensional formulation; for a layered aquifer such as an alluvial or
sandstone aquifer an alternative approach is to consider a series of permeable aquifer layers
interconnected by vertical flows through the less permeable strata.
An unsatisfactory selection of the coordinate system can mean that important flow mechanisms
are not represented; this can invalidate the analysis. Steady state conditions are unusual and
therefore the time variant response of the aquifer system should be included; the groundwater
head is then written as h(x,t), h(z,t), h(x,y,t), etc. For quick response aquifers it may be sufficient
to consider only two or three years but for slow response aquifers such as extensive alluvial or
sandstone aquifers, field information over a period of twenty or more years is desirable.
11.2.6 Preliminary flow balances
A vital step prior to the preparation of a mathematical model is the preparation of a water
balance which includes inflows, outflows and changes in storage. Reliable estimates can be
made of certain inflows (such as recharge) and outflows (such as abstraction) but for other
components, such as the change in storage, estimates are based on field measurements of the
groundwater head fluctuations multiplied by the estimated specific yield.
A careful selection of the periods over which the water balances are carried out can increase
the usefulness of the calculation. In cases where there is a recharge period each year it is helpful
to estimate the water balance three or four times each year. This will ensure that account is taken
of the balance between recharge and changes in storage It is unlikely that the components of the
water balance will sum to zero but any serious out of balances require careful investigations.

302
A preliminary water balance for the case study of the Miliolite Limestone aquifer can be
found in Table 11.1; the basis on which the estimates are made is explained in notes following
the table. It is inevitable that this form of water balance will involve uncertainties. Nevertheless
the balance will show whether there is an approximate understanding of the flow components.
As modelling proceeds it is possible to refine the flow balance.
11.3 Use of models for quantifying groundwater resources
11.3.1 Fundamentals of groundwater modelling
Any groundwater model is based on two principles:

(i) Darcys Law which relates the velocity of flow to the hydraulic conductivity and
the head gradient,
(ii) Continuity conditions, which ensure continuity of flow from inflows, through the
aquifer system to the outflows (water budget).
The three-dimensional time-variant equation for saturated groundwater flow based on these two
principles is:
K h + K h + K h = S h
x x x y y y z z z S t (11.1)
To study a particular problem it is necessary to specify the hydraulic conductivities, Kx, Ky

and Kz the specific storage, Ss throughout the study area, to identify the location of all the
boundaries and to specify flow or head conditions on each boundary. These boundary
conditions, which may be external or internal, represent inflows such as recharge or flows from
other strata or outflows such as rivers, springs or flows from pumped wells and boreholes.
Table 11.1 Approximate water balance for Miliolite Limestone aquifer; all quantities are in Ml/d
1982 1983
Notes Jan Apr MayAug SepDec JanApr MayAug SepDec
Recharge 1 0 130 190 0 490 0
Abstraction 2 120 70 10 220 60 80
Storage 3 +130 60 130 +240 400 +70
Springs 4 0 0 0 0 5 0
Coast 5 6 12 30 +3 10 30
Notes:
1. Recharge is estimated from a daily soil moisture balance.
2. Abstractions are estimated from the number of groundwater structures, their withdrawal rates, areas of crops grown
and the crop water requirements.
3. Storage changes are based on the estimated specific yield of 0.12 multiplied by the average change in water table
elevation; a rise in the water table means that water is released from storage to flow through the aquifer and is
indicated as +.
4. Spring flows are based on field observations.
5. Flows across the coast ( means an outflow to the sea) are estimated from the groundwater gradient at the coast and
the transmissivity; the transmissivity is higher for outflows since the flow occurs in the higher permeable zones.

303
Groundwater studies
There is no direct way of solving Equation (11.1) for practical situations. Two approaches are
possible:
(i) reduce the number of time and/or space dimensions until the simplified equation
can be solved using a suitable mathematical technique to give an expression for the
groundwater head distribution; this is called an analytical solution. The main
limitation of this approach is that the process of removing one or more of the
dimensions means that certain of the important flow processes may not be
included.
(i) use a numerical technique in which the groundwater head is specified only at nodal
points; the differential equation can then be written as a series of simultaneous
equations which can be solved using a digital computer. A commonly used method
is the finite difference method; other suitable techniques include the finite element
method and the boundary element method (Anderson and Woessner, 1991;
Boonstra and De Ridder, 1981; Liggett and Liu, 1983; Pinder and Gray, 1977;
Rushton and Redshaw, 1979).
11.3.2 Various types of models
The most direct models are based on a calculation using inflow, Darcys Law and outflow.
Alternatively, models may be complex three-dimensional time-variant simulations. In most
studies it is advisable to carry out preliminary direct calculations before embarking on complex
model studies.
Many mathematical models are based on different approximations. However, it is
essential to identify the type of model that is required from hydrological and hydrogeological
considerations. Consequently, the focus of this discussion will be the different types of model
available as summarised in Table 11.2. This is not a complete list but it does indicate a range of
models including analytical techniques in which exact solutions can be obtained and numerical
techniques that are usually more flexible. The analysis may assume predominantly one-
dimensional lateral or radial flow, two-dimensional flow, multi-layered flow or fully three-
dimensional flow.
There are many computer packages available which can be used by workers who have
little or no experience in numerical analysis. These packages usually correspond to one of the
types of model listed in Table 11.2. Nevertheless it is essential to carry out the formulation phase
of the study before selecting the method or package; if the essential features are not included in
the model, the model results may be totally misleading.
A model checks the consistency of the assumed components and parameters associated
with the inflows, outflows and flow processes. A model cannot be used to avoid the need to
collect and interpret field data, but it can be used to indicate the important processes and
parameters which should be considered in detail.
11.3.3 Important issues in developing models
Important information about the development of models can be found in a number of refer-
ences, including Anderson and Woessner (1991), Kinzelbach (1986), Pinder and Gray (1977),
Rushton and Redshaw (1979), Wang and Anderson (1982). The issues summarised below high-
light questions which must be resolved during the development of a model.
Size of area to be studied: the issue in selecting the size of the study area is whether to
investigate a large catchment or a smaller area of special interest. When studying a large
catchment there may not be enough detail in areas of interest. On the other hand, the area of
interest may not have boundaries on which hydraulic conditions are known. Often the best

304
Table 11.2 Different types of groundwater models
Type of model Purpose
i) Darcy's law calculation Applied in horizontal plane for flow in confined aquifer.
h(x) Applied in vertical plane for flow through poorly permeable layers.
h(z)
ii) One-dimensional flow Based on Darcy's law and continuity, a differential equation can be
h(x,t) derived,
K h = S h q
x x x t
with appropriate boundary conditions; S is the confined storage
coefficient or the specific yield; this equation can be integrated to obtain
steady state heads for horizontal one-dimensional flow. Alternatively
the Dupuit approach can be used.
Numerical solutions to this equation using finite differences can be
used to examine time-variant responses, the effect of varying saturated
depth, etc.
iii) Radial flow to well Analytical solutions such as the Theis equation and the many further
h(r,t) developments can be used to examine predominantly radial flow to
pumped boreholes in a wide variety of situations.
Numerical solutions can be obtained for the radial flow equation and
they can include all the conditions represented by the analytical
solutions and other factors such as well clogging, non-steady
abstraction, conditions changing between confined and unconfined.
iv) Two-dimensional flow Regional groundwater flow may be represented as two-dimensional
h(x,y,t) in plan with the equation
K h + K h = S h
x x x y y y S t
and appropriate boundary conditions. The transmissivity T may be a

function of h(x,y); Sy is the specific yield.
v) Two-dimensional flow in a vertical section
h(x,z,t)
K h + K h = S h
x x x y y y S t
with appropriate boundary conditions; Ss is the specific storage.

vi) Combined radial and vertical flow
h(r,z,t)
K h + 1 K h + K h = SS h
r r r r r r z z z t
with appropriate boundary conditions.
vii) Multi-layered flow Two zone radial flow model; radial flow models as described above
h1(r,t), h2(r,t) are inter-connected vertically to represent intermediate low
conductivity zones; a wide range of practical radial flow situations can
be studied using this approach.
h1(x,y,t), h2(x,y,t) Multi-layered systems such as alluvial and sandstone aquifers can be
represented as two-dimensional aquifers in plan interconnected
through poorly permeable aquitards, etc.
viii) Three-dimensional flow Full three-dimensional solutions in which variable hydraulic
h(x,y,z,t) conductivity distributions can be included; very extensive field
K h + K h + K h = S h
x x x y y y z z z S t
information is required to justify this approach.

305
Groundwater studies
approach is to develop a two-dimensional model for the whole groundwater catchment and
then study a smaller area as a three-dimensional model with boundary conditions taken from
the whole catchment model (Ward et al., 1987). In aquifers with a lower transmissivity too large
an area is often studied. Rushton and Redshaw (1979) introduced a method of estimating the
time taken for a change in conditions at a major feature (such as a river or a well field) to spread
a specified distance. Consider an unconfined aquifer with the following properties:
transmissivity, T = 40 m2/d,
specific yield, S = 0.03.
The time for the effect of a change in condition to spread for 2 km can be estimated from the
equation,
time, t = 2.5 L2 S/T = 2.5 x 2,000 2 x 0.03/40 = 7,500 days or 20.4 years.
If the simulations covers a period of about 20 years it is sufficient to model an area which
extends 2 km on either side of the location at which the changes in major features occur. The
model does not need to extend to features 10 km from the area of interest, since the effect will
take 510 years to have any influence!
Mesh spacing and length of time steps: when a numerical model is used in which heads are
calculated at discrete points and at specified time intervals, the mesh spacings and time
increments must be defined. The important criterion is that the mesh interval and time step
must be selected to give sufficient detail in areas of rapid change in groundwater head. Advice
about the choice of mesh interval and time step can be found in the standard references quoted
earlier.
Average conditions: it is usually unwise to attempt to model average conditions, since time
variant effects are generally of considerable significance. Average conditions (sometimes called a
pseudo-steady state) will not be appropriate if:
significant variations in saturated depth occur due to large groundwater head
fluctuations,
the annual recharge pattern is erratic,
abstraction patterns vary significantly during the year.
For most groundwater problems the time-variant fluctuations are an important part of the
study.
Starting (initial) conditions: it is essential to select suitable starting conditions. Specified
heads are usually unsatisfactory for starting conditions (Wang and Anderson 1982); if the
specified heads are not consistent with the assumed transmissivities, these specified heads may
be equivalent to unrealistic recharge values. The best approach is to use a dynamic balance in
which the simulation of a typical year is cycled a number of times until the annual dynamic
changes become regular.
Length of time to be studied: groundwater conditions change only slowly; if the model
simulation is too short it is possible that the changes in the aquifer conditions due to factors such
as increased abstraction may not become apparent during the simulation. Changes due to
increased abstraction from an unconfined aquifer often take tens of years to occur. There is no
convenient way of identifying the required length of a simulation; the best advice is to run the
simulation for several years more than initially planned, to see whether the effects of changes
have dissipated.
Unsaturated conditions: above the water table unsaturated conditions are certain to occur.
Whether the unsaturated conditions need to be considered depends on the nature of the aquifer
system. As rainfall recharge passes through an unsaturated zone a delay is likely, especially if
there are zones of low hydraulic conductivity in the unsaturated zone. Unsaturated conditions
can also be important beneath leaking canals. In many practical situations a factor to delay or

306
reduce the vertical movement of water in the unsaturated zone can be introduced, rather than
modelling the unsaturated conditions directly.
Quality issues: the above discussion has been concerned with groundwater heads and
groundwater flows, but the quality of the water and the movement of any contaminant in the
aquifer may be the crucial issue. Information about contaminant transport techniques can be
found in Anderson and Woessner (1991); new techniques of quality modelling can be found in
recent literature. However, one important consideration is that only when the understanding
and modelling of the groundwater flow is reliable will the predictions of the contaminant
movement be adequate.
11.3.4 Description of models used for case studies
The following section outlines the nature of the models used for the three case studies.
Field data for a pumping test in the weathered-fractured aquifer system is available;
consequently, the model used to represent this test is a radial flow model. The model must
include the following features, Fig. 11.1c:
fractured aquifer with horizontal hydraulic conductivity, specific storage and vertical
hydraulic conductivity;
weathered aquifer with horizontal hydraulic conductivity, vertical hydraulic
conductivity, specific yield and recharge;
the fractured and weathered zones are interconnected through the vertical hydraulic
conductivities;
water is only pumped from the fractured zone.
There are five unknown parameters (the vertical hydraulic conductivities are combined) which
are deduced by a sensitivity analysis using a two-zone model (Rathod and Rushton, 1991). The
aim of the modelling is to reproduce the pumping and recovery responses in the weathered and
fractured observation wells (Rushton and Weller, 1985). Because of the variability of a hard rock
aquifer, it is unreasonable to expect the field response to be represented exactly by a model
which cannot simulate the heterogeneous nature of the fractured aquifer. Nevertheless, the
general trends during the 7.5 hours of pumping and the subsequent recovery are reproduced by
the model, Fig. 11.1d. From the model results, the overall flow processes can be identified as
shown in Fig. 11.1e. The model parameters are deduced to be:
weathered aquifer, horizontal hydraulic conductivity = 0.5 m/d, vertical hydraulic
conductivity = 0.3 m/d, specific yield = 0.01,
fractured aquifer, horizontal hydraulic conductivity = 5.0 m/d, vertical hydraulic
conductivity = 3.0 m/d, confined storage coefficient = 0.0025.
Three types of model were used during the study of this alluvial aquifer:
(i) a vertical section model which represented the complex layering of gravel, sands,
silts and clays; this model confirmed the importance of the vertical flow
components when increased abstraction occurred from the deeper aquifers
(Rushton and Tiwari, 1989);
(ii) a large regional model which idealised the aquifer unit as a system of inter-
connected layers of higher and lower hydraulic conductivity;

307
Groundwater studies
(iii) a small regional model which concentrated on a region of plan area 4 km by 4 km

with a mesh spacing of 250 m; the computational work for this small regional
model was carried out on one of the earlier personal computers (Kavalanekar et al.,
1992). The complex layered aquifer is represented as consisting of the upper water
table aquifer, an intermediate poorly permeable clay zone and a lower aquifer from
which most of the abstraction occurs. Lateral flows are small compared to the
vertical flows and can be represented as small specified flows on a boundary of the
lower aquifer. The advantage of this small study area is that individual tube-wells
can be represented; this is not possible with a large regional model having a mesh
interval of 2 km. The adequacy of the simulation is demonstrated by comparisons
between the field and modelled groundwater heads in the shallow and deep
aquifers, Fig. 11.2f; the field information is very limited but it is sufficient to show
that the model does represent the major flow processes.
(iii) Limestone aquifer
The limestone aquifer can be represented by a two-dimensional regional groundwater flow

model, provided that the change of transmissivity with the groundwater head is included at all
nodal points. Relationships between transmissivity and water table elevation in the form of
Fig. 11.3c must be included at every node.
A second important feature of the mathematical model is the representation of periods of
high recharge. When the recharge is high, water moves rapidly down-dip and leaves the aquifer
through springs. The adequacy of the models is demonstrated by a comparison between the
field and modelled groundwater heads on a typical section during the pre-monsoon and post-
monsoon seasons, Fig. 11.3a. The significance of the springs during the post-monsoon period is
apparent from the diagrams since the groundwater head profiles intersect the ground surface.
11.4 Using forecasting to identify safe yields
11.4.1 Forecasting and prediction
During the study of groundwater systems, an investigator is often required to comment on

future groundwater conditions. These estimates of future groundwater conditions can be
grouped into:
forecasting the effect of natural changes and
predicting the effect of man-made changes.
Hydrodynamic methods based on distributed models are generally used for predicting conditions
in the disturbed groundwater regime, whereas statistical methods are often used for forecasting the
groundwater regime under natural conditions.
11.4.2 Forecasting*
A hydrogeological forecast has the following specific features:

a probabilistic aspect involving the determination of the confidence interval of
probable predicted values,
* This section has been based on extensive material provided by Professor V. S. Kovalevsky, Institute of Water Problems,
Moscow (Russia).

308
a temporal aspect and

a verification aspect involving the reliability of predicted values.
Forecasts may be made for a natural, slightly disturbed or disturbed groundwater regime. For
natural conditions, forecasting only requires data on the natural hydro-meteorological features;
for disturbed regimes, both the meteorological features and aquifer flow mechanisms must be
considered.
The following time scales apply to forecasts:
short term forecasts are made for a period of several days and are based on factors
which have already come into action (rainfall, flood water discharge etc.); they reflect
the time-lag of the responses to the external impacts,
seasonal forecasts which cover a period of several months are based both on factors
which are partially known at the time of forecasting and on long term inter-relations
between the regime components,
long term forecasts are often based either on the inertia of the system and the exposed
tendencies of its development or on the relations with other long term factors such as
solar activity and climate changes.
Forecasts may not relate directly to specific times, but may be in the form of possible magni-
tudes of variables without indication of the time of their occurrence. Furthermore, the predicted
data may relate to global, regional, basin or point scales. In terms of the predicted elements, fore-
casts can be divided into two types, one parameter (groundwater head, temperature, or specific
discharge for a certain period of time) or combined parameter forecasts (groundwater chemical
composition, intra-annual groundwater level distribution).
The choice of the forecasting method depends on the availability of initial information, the
duration of the forecast, the accuracy and the purpose. Available methods include the method of
hydrogeological analogy, the balance method, probabilistic statistical methods and dynamic
stochastic methods.
(i) Forecasting based on hydrogeological analogies
The hydrogeological analogy may be used for initial approximate estimates of the scale of
possible changes. Relationships in both time and space can be used by considering the hydro-
geological subdivisions of regions by groundwater regime formation conditions (i.e. distinguish-
ing regime types, sub-types, kinds and varieties) it is possible to:
extend the data on the groundwater regime and forecasts from an area which has been
studied to an area which has not been studied,
evaluate the probabilistic nature of the seasonal and long term variability in ground-
water heads, the variation in amplitude of the groundwater heads and
determine the most probable minimum and maximum groundwater heads, temper-
atures and chemical composition of the groundwater.
Quantitative relationships between pairs of parameters may be used for approximate forecasts.
Typical examples include:
relationships between the amplitude of the fluctuation of groundwater heads, hg, and
the depth of water in a river at a gauging station, hr, expressed as a fraction of unity,
(Fig. 11.5a);
relationships between the amplitude of the fluctuations of groundwater heads and the
water table position or the character of the inter-annual level regime in years with dif-
ferent water availability (probability) (Fig. 11.5b-c);
normal annual relations between groundwater flow and precipitation in the form of
coefficients, K, which equal the percentage ratio of precipitation to groundwater

309
Groundwater studies
Figure 11.5 Examples of forecasting: (a) relationships between groundwater levels amplitude/river stage
amplitude and the distance to the river, (b) between groundwater level fluctuation amplitude
and depth to ground water, (c) graph of the inter-annual groundwater regime, (d) graph of
groundwater discharge in years with different probability of water availability
discharge (K = 100 P(mm)/Q(mm)); the relation between the groundwater discharge

and the annual water availability (Fig. 11.5d) may be used for predicting the possible
temporal variability in groundwater discharge to streams.
Long series of observations can be used as an analogy to obtain statistical parameters for proba-
bilistic estimates for areas with a short series of observations. The similarity of groundwater
regimes in areas with short or long series of observations can be proved by the similarity of
physical conditions. Consequently the method of hydrogeological analogy can be used for areas
with insufficient hydrogeological information, to indicate the nature and scale of the processes.
(ii) Probabilistic statistical methods of forecasting
Probabilistic-statistical methods can be applied to forecast the natural and slightly disturbed
groundwater regimes (Brown et al., 1972; Bredehoft, 1982). Probabilistic techniques are required
because of the interaction between the groundwater regime and a large number of different but

310
essentially unpredictable factors, which are mostly climatic. For example, the magnitude of the
groundwater recharge depends on the magnitude and intensity of the precipitation and also on
the air humidity deficit, soil moisture content, freezing of the aeration zone, air temperature,
evapotranspiration and, if appropriate, snow melting and thaw periods. It may be impossible to
obtain all this information on a regional scale for the period of the forecast; consequently the
forecast is based on the most significant factors.
Often it is difficult to find a long series of observations in the region where the forecast is
required and therefore a probabilistic statistical technique is used. For forecasting in moderate
latitudes it is usually possible to predict mean annual and extreme groundwater levels and
discharge rates from:
depths to the water table at the start of the forecast,
total mean monthly precipitation for the spring, autumn and winter periods,
total mean monthly temperatures with an air moisture deficit for the spring, autumn
and winter periods,
duration of thaws or a total positive temperature for the winter thaw period.
The contribution of these factors is different in different regions and for various groundwater
level forecasts. It is necessary to identify the important factors for a particular forecast; this can
be achieved by correlation techniques and good judgement. More information can be found in
Brown et al. (1972).
The longer the period covered by the forecast, the lower the accuracy. The optimum
period for forecasts of the minimum levels is 3 to 4 months, and for the maximum spring levels
it is 2 to 3 months. As the duration of the forecast increases, the meteorological factors become
more important and the role of the preceding history becomes less important. Further infor-
mation can be found in Kovalevsky (1976).
(iii) Balance methods of forecasting
Balance methods of forecasting are based on the main groundwater components which deter-
mine the groundwater head response with time. The groundwater balance equation can be
written as
(Q Q2 ) + w t
h = 1 (11.2)
F
where
= water yield or saturation deficit of the soil or rock (the storage coefficient),
h = change in groundwater head in a flow element between wells,
t = time interval for the calculation,
Q1, Q2 = groundwater inflow and outflow in the flow element,
F = plan area of the flow element,
w = rate of infiltration recharge of groundwater from above.
The water balance equation for the soil-rock prism extending from the top of the confining layer
to the water surface is:
(Q Q2 ) t
C + n h = Wa + 1 (11.3)
F
where:
C = the change in moisture reserve in the zone of aeration,
n = total soil moisture capacity,
wa = water exchange between zone of aeration and atmosphere (evapo-transpiration).

311
Groundwater studies
Combining the above equations
h = [Wa t w t C] Vvol (11.4)
h Vvol
= 1 W t + C (11.5)
Wa t Wa t
where Vvol (= n ) is the soil moisture content within the capillary fringe.
These equations allow a forecast to be made of the rise in the water table, h, if data are avail-
able concerning the water exchange between the zone of aeration and the atmosphere and the
groundwater recharge. Such forecasts are only realistic if there are long series of data for these
components. Forecasts based on lumped parameter approaches are discussed by Anderson and
Burt (1985).
(iv) Forecasts based on the inertial properties of aquifers
The longer term response of an aquifer to meteorological influences is not direct. Because hori-
zontal groundwater velocities are typically tens of metres per year, the precipitation which
infiltrates into the aquifer at some distance from a river may take many years to reach the river.
As the groundwater moves towards the river it will receive additional recharge each year hence
the river response is the result of precipitation over a significant number of years. This may be
the reason for the weak correlation between precipitation in one year, and groundwater heads or
flows.
The capability of an aquifer system to integrate the effects of the precipitation over a
number of years and the degree of the inertia of the system is unique because it depends on the
seepage properties of the water-bearing rocks and the drainage area. This groundwater inertia
may be assessed by the correlation between groundwater heads and the summed effect of the
precipitation in preceding years.
It is advisable to use as an annual precipitation the total effective precipitation (the
precipitation for the cold season being the main source of groundwater recharge). The extremes
of the curve of these correlations gives the optimal number of years whose precipitation should
be taken into consideration in forecasting. This procedure is necessary when it is not possible to
use the values of preceding groundwater levels as one of the factors in which the precipitation of
the preceding stage is taken into account. Such variants emerge, for example, when forecasting
the groundwater regime and resources for the remote future, using various climate change
scenarios.
Although the integrating capability of aquifers to reflect the precipitation for the preced-
ing years may be large, the greatest contribution may be due to the recent history. All this
predetermines the possibility of using as the method of forecasting the extrapolation of regular
features established for the period of earlier observations, preserving the former conditions of
the groundwater regime for the period of forecasting.
Methods based essentially on the inertial properties of the aquifer include the following:
identifying and extrapolating linear and non-linear trends,
distinguishing significant harmonics in the structure of the series and extrapolating
future groundwater head and discharge variations as a total of these harmonics,
forecasting groundwater heads as a Markovian process.
Strictly speaking, forecasts based on these methods do not require a knowledge of the physics of
the processes, since they are based on properties of the time series. However, the reliability of
the forecast depends on the extent to which the trends can be substantiated physically. Linear

312
and non-linear trends in groundwater head and discharge variations may be caused by human
activities such as reclamation, urbanisation, construction of hydraulic structures, groundwater
abstraction, technological changes, or by natural climate changes. Linear or non-linear trends
should be identified and the statistical reliability of the trends should be considered.
(v) Probabilistic forecast estimates
Probabilistic forecasts of the groundwater regime may be made if calendar (time-based) fore-
casting is impossible or inadvisable. Probabilistic forecasts allow characterisation of extreme,
mean or other values with given probabilities, without an indication of the time of their onset.
Forecasts based on probability concepts depend on:
statistical analysis of long term groundwater observations,
computation of statistical properties of groundwater heads and discharge,
finding the law of groundwater head and discharge distributions,
using empirical and theoretically calculated probability graphs,
groundwater head and discharge values of 50% or 95% or other recurrence intervals
can be deduced from the probability graphs.
It is essential to estimate the possible errors in the computed values. The accuracy of all
statistical estimates depends on the length of the series. Higher accuracies result from longer
series. If the series are short, they should be extended using standard techniques or using series
analogues.
11.4.3 The use of models for predictions
The reason behind the development of a groundwater model is usually to obtain a means of
predicting the effect of exploiting an aquifer or modifying the exploitation to minimise the
effects of falling water tables. At the outset it must be stressed that the use of a model for
predictive purposes does not mean that precise forecasts will be made of the groundwater
conditions at a particular time in the future; the variability and unpredictability of the recharge
usually prevents precise forecasts. Nevertheless it is possible to consider the likely consequences
of any changes; the findings will be in terms of likely changes in groundwater levels, river flows
and changes in the quality of water. There are a number of important issues to be considered
when predicting future aquifer responses:
the purpose of the prediction may be to identify the safe yield of an aquifer; this can be
defined in terms of the dynamic storage, whereby the volume of water which can be
removed equals the volume of stored water which can be replenished by the end of the
annual recharge season. If the annual recharge shows significant variations it may be
acceptable to make use of some of the static storage to smooth out the fluctuations in
recharge. The general principle behind the safe yield concept is that, in the long term,
the withdrawal from the aquifer should be such that a long term water balance may be
maintained.
an alternative approach involves the mining of groundwater. In deep alluvial and
sandstone aquifers, the dynamic storage may be less than 1% of the total storage, con-
sequently some mining of the groundwater reserves may be acceptable. In semi-arid
areas with deep aquifers where groundwater is used for irrigation, mining of ground-
water is almost certain to occur. However, as the mining of groundwater proceeds, the
water table falls, the cost of lifting the groundwater from depth increases and the
efficiency of the wells and boreholes is likely to decrease so much that shallow wells
are abandoned and the drilling of new deep boreholes becomes necessary. These severe
consequences of the mining of groundwater must be considered.

313
Groundwater studies
there are also environmental consequences of changing the use of an aquifer. Examples
include the settlement of land due to the mining of groundwater (Ramnarong, 1983)
and the drying up of rivers and lakes. On the other hand, excess recharge due to irriga-
tion or urban development can lead to waterlogging; typical examples are the water-
logging in the Indus Basin or in cities such as Riyadh (Rushton and Al-Othman, 1994).
A further environmental consequence can be a deterioration in the quality of pumped
water due to recent or trapped saline water being attracted to the pumped boreholes.
11.4.4 Methodologies in the use of models for predictive purposes
This section outlines certain of the issues which need to be resolved when using a model for
predictive purposes.
What are the inputs? Recharge to the aquifer has to be specified and it is advisable to use
historical rainfall and evapotranspiration values so that the forecasting can show how aquifer
conditions would differ under alternative abstraction regimes. If the purpose of the study is to
predict the effect of climate change, then rainfall and evapotranspiration modified according to
the predicted climate changes should be used in the recharge estimation.
Do conditions in the aquifer system change? Care must be taken when the model is used for
conditions other than those for which the model was tested New mechanisms can become
important, as demonstrated by the case studies discussed in this chapter. Typical changes in con-
ditions include reductions in transmissivities, dewatering of fissures, changes between confined
and unconfined conditions, modifications to the recharge mechanisms, an alteration to the effi-
ciency of boreholes, or changes in the nature of rivers or streams.
What are the starting conditions? For certain situations it is acceptable to use the same start-
ing conditions as for a historical run. However, if mining of groundwater occurs with a steady
decrease in the water table elevation, the simulation should start from the current conditions, to
predict future trends. Unsatisfactory starting conditions are the cause of many misleading
predictive runs.
What are the outputs? Often it is helpful to plot differences between the predicted results
with changed conditions and a baseline which may be a historical run or a predictive run using
the current abstraction policies. modifications to the flow balance within the aquifer system also
provide useful insights.
For how many years should predictive simulations continue? For aquifers with large volumes
of stored water, changes in the conditions within the aquifer can be slow, often taking decades.
consequently, it is advisable to run predictive simulations for very long time periods until a new
equilibrium is approached.
11.4.5 Examples of predictions
Each of the models developed for the case studies has been used for forecasting.
The model described in Section 11.3.4 was developed to represent a pumping test with 7.5 hours
of pumping and the subsequent recovery. This model can also be used to obtain estimates of the
likely responses due to long term pumping. However, because the predictive runs are far longer
than the historical testing, the results should be used with caution. In predicting likely
responses, three situations have been considered to determine how the aquifer can be exploited
safely; the aquifer properties are recorded in Section 11.3.4.

314
(a) a large-diameter well in the weathered zone with a diameter of 5.8 m and a depth of
water at the start of the irrigation season of 8.0 m.
(b) a tube well with a diameter of 0.15 m tapping only the fractured zone with a sub-
mersible pump, thereby allowing large pumped drawdown,
(c) a dug-cum-bore well in which a bore is drilled into the fractured zone through the base
of the large diameter well of example (a); the pumped drawdown is restricted since
there must always be sufficient water in the bottom of the large diameter well to provide
suction for the surface pump.
The effective spacing between the wells is set at 400 m; consequently the area of aquifer from
which any of the structures can draw water is 160,000 m2. The aim is to pump water for
100 days to maximise the irrigation potential. The results are as follows:
(a) The yield of the large-diameter well is poor; if the duration of pumping is 7.5 hours, the
pumping rate must not exceed 150 m3/d if the abstraction rate is to be maintained for the
full 100 days of the growing season.
(b) If a submersible pump is used with the borehole taking water from the fractured zone,
the yield is greatly increased; in fact a discharge rate of 350 m3/d can be maintained
without difficulty for 15 hours per day for 100 days. The maximum pumped drawdown
would be 13.5 m but significant drawdown would also occur in the weathered zone. In
fact the weathered zone would almost be dewatered, with drawdown greater than 9.3 m.
This would mean that shallow dug wells in the weathered zone would become dry.
Another serious consequence is that the recharge needed to refill the weathered aquifer
during the next monsoon rains would be 94 mm; this intensity of recharge occurs only in
the wettest years. Prediction (b) represents precisely what happened when tube wells
were drilled in the fractured zone of the Dulapally aquifer. Because so much water was
pumped from the aquifer and because the spacing between the tube wells was often less
than 400 m, the tube wells failed after one or two years and also the weathered zone was
dewatered, causing many of the large diameter wells in the weathered zone to fail too.
The problem is that the submersible pumps in the tubewells are too powerful at
removing water from the aquifer.
(c) To use the aquifer efficiently, the ability of the fractured zone to transmit water should
be used but the drawdown should be limited; this can be achieved using a dug-cum-
bore well. When the abstraction is 350 m3/d for 7.5 hours, the maximum drawdown in
the well is 6.6 m, which is within the capability of a suction pump. The maximum
drawdown in the aquifer is 5.0 m close to the well; with this order of drawdown some
deepening of domestic dug wells would be necessary. In most years the recharge would
reach the 47 mm required to ensure that levels recover following the monsoons. Conse-
quently a dug-cum-bore well with a surface pump should be used with this form of
layered aquifer.
Extensive predictive runs have been carried out in the Mehsana alluvial aquifer; all have shown
that a significant reduction in the volume of water pumped from the deep aquifers is required to
stabilise the drawdown.
One particular concern is that, as the water table continues to fall, it moves into the zones
with a higher clay content having a lower specific yield. This condition of a decreasing specific
yield can be included directly in the model by modifying the specific yield from the standard
value of 0.08 to 0.01 when the water table in the overlying aquifer falls more than 50 m below
ground level. After six years the predicted groundwater heads in the main aquifer fall to 110 m

315
Groundwater studies
below the ground level; with this substantial decline in heads there would be a serious risk that
unconfined conditions would develop in the deeper aquifers.
(iii) Limestone aquifer
Canal irrigation has extended to an area close to the Miliolite Limestone aquifer; one predictive
scenario is to use the surplus canal water during the monsoon season for artificial recharge, both
to improve the yields of the existing wells and to flush out the saline water in the coastal zone.
Field experiments demonstrated that the wells are very efficient at accepting the recharged
water but the trial did not continue for long enough for the overall success of the scheme to be
proved.
The likely outcome of implementing an artificial recharge scheme was explored using
the mathematical model which had been verified against the historical field response. When
artificial recharge was simulated in the model:
the first benefactors were the farmers with wells downstream from the recharge sites;
the yield of their wells increased significantly,
farmers with wells upstream or at some distance from the artificial recharge sites
would receive little benefit from the recharged water,
the next stage is that much of the recharged water would escape from the springs,
there would be no reduction in the salinity of the groundwater within 2 km of the
coast; this saline water was drawn in by pumping and will not be flushed out by the
artificially recharged water which takes easy routes out of the aquifer rather than
moving to the coastal area to force out the poor quality water.
11.5 Concluding remarks
Extensive experience has been gained in the assessment of groundwater resources and the pre-
diction of likely future responses. The progress in recent years has been due to improved field
techniques combined with developments in groundwater modelling. The key to reliable assess-
ments and predictions is the imaginative use of the wide variety of hydrogeological techniques
combined with the skilful use of groundwater models. Groundwater assessment and prediction
requires a team effort in which skills and techniques are combined to lead to an understanding
of the flow processes within the aquifer system.
ANDERSON, M. G.; BURT, T. P., 1985. Hydrological Forecasting. Wiley, New York.
ANDERSON, M. P.; WOESSNER, W. W., 1991. Applied Groundwater Modelling. Academic Press, 381 pp.
ANDERSON, P. F.; MERCER, J. W.; WHITE, H. O., 1988. Numerical Modelling of Salt Water Intrusion
at Hallandale, Florida. Groundwater, Vol. 26, pp. 61930.
BALEK, J., 1989, Groundwater Resources Assessment. Developments in Water Science 38, Elsevier,
249 pp.
BEAR, J., 1979. Hydraulics of Groundwater. McGraw-Hill, 567 pp.
BOONSTRA, J.; DE RIDDER, N. A., 1981. Numerical Modelling of Groundwater Basins; A User-Oriented
Manual. ILRI Publ. 29, Wageningen, 226 pp.
BOUWER, H., 1978. Groundwater Hydrology, McGraw-Hill, 480 pp.
BREDEHOFT, I. D. (ed.), 1982. Groundwater Models, Vol. 1. UNESCO, Paris.

316
BROWN, R. H.; INESON, J.; KONOPLYANTSEV, A. A.; KOVALEVSKY, V. S., (eds.), 1972, Groundwater
Studies, UNESCO, Paris.
CONNORTON, B. J., 1985. Does the Regional Groundwater-Flow Equation Model Vertical Flow?
J. Hydrol., Vol. 79, pp. 27999.
DE MARSILY, G., 1986, Quantitative Hydrogeology, Academic Press, 440 pp.
DRISCOLL, F. G., 1986. Groundwater and Wells, Second edition, Johnson Division, St. Paul, Minn.,
1089 pp.
FREEZE, R. A.; CHERRY, J. A., 1979. Groundwater, Prentice-Hall Inc., 604 pp.
KAVALANEKAR, N. B.; SHARMA, S. C.; RUSHTON, K. R., 1992. Over-Exploitation of an Alluvial
Aquifer in Gujarat, India. Hydrol. Sci. J., Vol. 37, pp. 32946.
KINZELBACH, W., 1986. Groundwater Modelling. Developments in Water Science 25, Elsevier,
333 pp.
KOVALEVSKY, V. S. 1976. Long Term Fluctuations in Groundwater Levels and Discharge. Nauka,
Moscow. (In Russian.)
LERNER, D. N.; ISSAR, A. S.; SIMMERS, I., 1990. Groundwater Recharge, International Contributions to
Hydrogeology, IAH, Vol. 8, 345 pp.
LIGGETT, J. A.; LIU, P. L., 1983. The Boundary Integral Equation Method for Porous Media Flow. Allen
and Unwin, London. 255 pp.
MARINO, M. A; LUTHIN, J. N., 1982. Seepage and Groundwater. Elsevier, 489 pp.
MCWHORTER, D. B.; SUNADA, D. K., 1977. Groundwater Hydrology and Hydraulics. Water Resources
Publications, 290 pp.
PINDER, G. F.; GRAY, W. G., 1977. Finite Element Simulation in Surface and Sub-Surface Flow.
Academic Press, New York.
RAMNARONG, V., 1983. Environmental Impacts of Heavy Groundwater Development in Bangkok,
Thailand. Intern. Conf. on Groundwater and Man, Australian Government Publishing
Service, Canberra, Vol. 2, pp. 34550.
RATHOD, K. S.; RUSHTON, K. R., 1991. Interpretation of Pumping from a Two-Zone Layered
Aquifer Using a Numerical Model. Groundwater, Vol. 29, pp. 499509.
RUSHTON K. R.; AL-OTHMAN, A. A. R., 1994. Control of Rising Groundwater Levels in Riyadh,
Saudi Arabia. In: Groundwater Problems in Urban Areas, Thomas Telford, London,
pp.299309.
RUSHTON, K. R.; RATHOD, K. S., 1985. Horizontal and Vertical Flow Components Deduced from
Groundwater Heads. J. Hydrol., Vol. 79, pp. 26178.
RUSHTON, K. R.; TOMLINSON, L. M., 1979. Possible Mechanisms for Leakage between Aquifers and
Rivers. J. Hydrol., Vol. 40, pp. 4965.
RUSHTON, K. R.; WELLER, J., 1985. Response to Pumping of a Weathered-Fractured Granitic
Aquifer. J. Hydrol., Vol. 80, pp. 299309.
RUSHTON, K. R.; RAGHAVA RAO, S. V., 1988. Groundwater Flow through a Miliolite Limestone
Aquifer, Hydrol. Sci. J., Vol. 33, pp. 44964.
RUSHTON, K. R.; REDSHAW, S. C., 1979. Seepage and Groundwater Flow. Wiley, Chisester, 332 pp.
RUSHTON, K. R.; SRIVASTAVA, N. K., 1988. Interpreting Injection Well Tests in an Alluvial Aquifer.
J. Hydrol., Vol. 99, pp. 4960.
RUSHTON, K. R.; TIWARI, S. C., 1989. Mathematical Modelling of a Multi-Layered Alluvial Aquifer.
J. Inst. Eng. India, Vol. 70, part CV2, pp. 4756.
RUSHTON, K. R.; WELLER, J., 1985. Response to Pumping in a Weathered-Fractured Granitic
Aquifer. J. Hydrol., Vol. 80, pp. 299309.
SIMMERS, I. (ed.), 1997. Recharge of Phreatic Aquifers in (Semi-)Arid Areas. Intern. Contrib.
Hydrogeol., IAH, Vol. 19, Balkema, Rotterdam.
TODD, D. K., 1980. Groundwater Hydrology. John Wiley and Sons, New York, 535 pp.
WALTON, W. C., 1991. Principles of Groundwater Engineering. Lewis Publishers, Michigan, 546 pp.

317
Groundwater studies
WANG, H. F.; ANDERSON, M. P., 1982. Introduction to Groundwater Modelling, Finite Difference and
Finite Element Methods. W. H. Freeman & Co., San Francisco, 237 pp.
WARD, D. S.; BUSS, D. R.; MERCER, J. W.; HUGHES, S. S., 1987. Evaluation of Groundwater
Corrective Action of the Chem-Dyne Hazardous Waste Site Using a Telescopic Mesh
Refinement Modelling Approach. Water Resources Research, Vol. 23, pp. 60317.
ZEKTZER, I. S. (ed.), 1983. Principles of Hydrogeology, Vol. 2, Hydrogeodynamics. Nauka, Novosibirsk.
(In Russian.)

318
12 G ro u n d w a t e r m a n a g e m e n t
12.1 Introduction
12.1.1 Scope
As with any natural resource, the management of groundwater involves reconciling a supply-
based approach (managing the resource) with a demand-based approach (managing exploitation
and utilisation), in the light of the objectives of the management actors whose decisions are
based on a number of criteria subjected to various constraints and applicable on the basis of dif-
ferent methods. The following brief survey of groundwater management will therefore look at:
the concept of the resource to be managed, the physical conditions and exploitation strategies
for which a strategy is to be defined, and the socio-economic conditions. The latter are the mana-
gers and their objectives, management constraints and criteria, management decision methods
and aids, and management tools.
12.1.2 Groundwater resource management preliminary remarks
As is the case with water resources in general, the groundwater resource is a combined physical
and economic concept: it is water found underground which is seen from a utilitarian point of
view in terms of the possibility and benefit of drawing on it. It constitutes a natural supply to
meet the demand for water.
The groundwater resource is a multidimensional concept: it is defined by its location, its
occurrence over time, its size, properties, conditions of accessibility, the effort required to
mobilise it and, therefore, its cost, all of which are also to be considered in the context of
demand. The resource is thus assessed using a variety of criteria, and it would be simplistic to
measure it solely in quantitative terms on the basis of the flow. In terms of user criteria, the
groundwater resource as a source of water supply competing with others (surface water, in
particular) has different advantages as well as some disadvantages and limitations, which are
summarised in 12.1.

319
Groundwater studies
Table 12.1 Advantages and disadvantages of groundwater as a source of supply, compared to the use
of surface water
Spatial distribution An extensive resource facilitating catchment near points of use, thus minimising
conveyance costs, but to satisfy strong demand a number (how many depends
on output) of catchment areas will be required
Availability over time A permanent resource with steady discharge volume, less affected by the
weather than surface water, and therefore affording greater security of supply.
Natural storage which does not require regulation works
Resource evaluation An invisible resource whose precise evaluation may require the use of rather
sophisticated methods which are more expensive but also quicker than those
used for surface water resources
Natural properties Constant or little changing properties making it easy to correct certain
deficiencies for certain uses (potability, process water): hardness, Fe and Mn
content. Increased salinity at medium and great depth
Vulnerability to pollution Deep-lying water is generally free of risks, except those posed by faulty well
drilling and deliberate injection. Underground phreatic water is more
indirectly exposed to the risks of diffused or localised pollution than surface
water, but suffers from more lasting effects (being less resilient)
Production costs Investment and operating costs are lower on average than for surface water, but
fairly diversified, depending on the local characteristics of the aquifer (depth,
capacity). Operating costs reflect variations in energy costs (except in the case of
gravity catchment). Decreasing yield with increase in the overall rate of
exploitation of a given aquifer
Operational flexibility Possibility to install or close at any time additional production wells or water
abstraction equipment, and therefore better adapted to react to changing
demand than surface hydraulic development structures; investments can be
spread in time and returns will come quicker
The groundwater resource must be described, evaluated and managed within precisely
defined physical frameworks: the aquifer systems (see also Chapter 10), the natural units of
groundwater management whose extent, complexity and behaviour condition the exploitation,
conservation and management of the resources.
Because of the continuity that exists between the aquifers and the surface water systems
of a catchment basin, and because of the possible interactions between groundwater exploitation
and surface water management, the groundwater resources are not independent from the
surface water resources in terms of their evaluation or their management. Hence the need to
integrate groundwater management into overall water management. This integration does not
exclude the profitable utilisation of distinctive groundwater characteristics which often com-
plement those of surface water.
The groundwater resource, described as being a supply model, can rarely be evaluated
in terms of units (average producible quantities, costs, external effects, etc.). Its evaluation
cannot be dissociated from predictive management and the assessment of the feasibility of different
exploitation scenarios, taking different constraints and user criteria into account.
Unlike other underground resources, groundwater is not a mineral raw material like
any other. It is mobile and, therefore, should be considered not in terms of legal ownership but
rather of exploitation rights. As a result of the transmission of effects in the aquifers, the impact

320
Groundwater management
of actions on groundwater (withdrawals, excess of inflow, pollution) extends as far as the

physical boundaries of the aquifer system. The management of a unit of groundwater resources
should, therefore, be community based, involving all the stakeholders concerned.
12.2 Physical conditions of groundwater management
12.2.1 Flow management and storage management
As an aquifer system combines the functions of conveyance and storage, its management should
concern both flow and storage, but it can combine the two in very varying degrees.
In general, an aquifer system provides a renewable resource managed as a flow that is
regulated to a greater or lesser degree by storage variations. The possibility of amplifying these
variations provides a degree of freedom of action in relation to the aquifer. It makes the strategy
of withdrawals more or less independent of the natural inflow, i.e. more regular or irregular
than the latter on a perennial basis as well depending on variation in demand. In certain
cases, the aquifer storage itself may become a non-renewable resource, which may be mined.
The respective and more or less combined roles of flow and storage in groundwater
resource management (quantitatively speaking) can be expressed by the diagram of Figure 12.1.
The effect of an aquifers storage and the variation in storage on the rate of water pro-
duction is therefore of prime importance in the definition of the actual or planned exploitation
strategy.
Figure 12.1 The roles of flow and storage in groundwater management
12.2.2 Overall exploitation strategies of an aquifer
Depending on the degree and reference period of the dynamic balance established between an
aquifers natural flows and the flow imposed by exploitation depending, therefore, on the role

321
Groundwater studies
played by storage variation three types of exploitation strategy are conceivable and prac-
ticable, albeit under certain constraints.
(i) A dynamic balance strategy
A dynamic balance strategy, may be imposed by an average withdrawal rate that is less than or
equal to the average inflow (possibly increased by boundary effects), without prejudice to
possible seasonal or even interannual variations (e.g., increase during periods of drought).
The storage decreases during an initial phase of imbalance, but then stabilises and
assumes a regulatory role (on an annual or perennial basis). The natural regulatory function may
be subject to the requirement that a minimum flow rate is maintained at the discharge
boundaries (e.g., spring, flow or low-water level of main stream) or the need to preserve the
freshwater/saltwater balance in a coastal aquifer (Fig. 12.2).
Figure 12.2 Exploitation under a dynamic balance strategy after a necessary transitional phase: diagram
of changing average levels global indicator of the state of the storage and flows.
P is less than Q. The stabilisation of levels (on average) is deferred (in t2) in relation to that
of withdrawals (in t1), After a withdrawal phase from t0 to t2, there is a time lag until the levels
stabilise; the effect of the withdrawals in t1 results in the levels at t2
(ii) A temporary non-equilibrium strategy
A temporary non-equilibrium strategy, whether managed or unavoidable, has an average

withdrawal rate higher than the average inflow (even if augmented by boundary effects),
irrespective of whether the withdrawals are increasing or stabilised. There is a time lag before
equilibrium is restored.

322
Withdrawals from storage make a large, and sometimes dominant, contribution to water
production, for a period of variable length which is limited either by external constraints (strat-
egy 1) or as a result of the reduced capacity of the wells (e.g., drawdowns which are excessive
or restricted by the depth of the aquifers substratum). In a later phase of possible restoration of
equilibrium involving a reduction in withdrawals or sometimes stimulated by an artificial or
induced inflow increase groundwater storage can be partially recovered and stabilised (on
average), as in strategy 1 (Figure 12.3).
Figure 12.3 Exploitation under a prolonged imbalance strategy followed by a return to balance by reducing
withdrawals: diagram of changes in average level and flows, with P > Q. Imbalance strategy
occurs from t0 to t3 during the phases of growth, followed by stability then decrease
in withdrawals P, with dewatering from t0 to t2, despite the almost total reduction in residual
outflow q. There is a tendency towards restoration of balance in t3, with partial reconstitution
of groundwater storage, withdrawals now being at a rate close to the initial recharge flow.
Overexploitation would consist in a lowering of levels below the maximum acceptable depth
(iii) Depletion or permanent non-equilibrium strategy
Depletion or permanent imbalance strategy, that is the mining of groundwater at a withdrawal

rate that may or may not increase, and that from early on is above often well above the
average inflow rate.
Withdrawals from storage provide most of the water produced. In the long-term, the
exploitation is limited when the drawdown becomes excessive, but equilibrium is not regained
returning to a balanced strategy, because the recharge of the groundwater storage is too slow
and is sometimes hindered by the irreversible degradation of storage capacity trough com-
paction of the aquifer as a result of the pressure decrease (Fig. 12.4).

323
Groundwater studies
Figure 12.4 Mining of stored groundwater storage, under a depletion strategy; diagram of changes
in levels and flows (P > Q). The time limit of the exploitation, depends on the rate of withdrawal P
and the characteristics of changes in the levels. It is reached when the drawdown levels reach
the maximum depth which is practicable (given the position of the boundary of the aquifer),
or acceptable from the point of view of pumping costs
(iv) Overexploitation: surplus or excessive exploitation?
Rather than the neutral sense of temporarily excessive withdrawals from the average natural
flow of an aquifer, which is covered by the hydraulic concept of imbalance, overexploitation is
more usually defined in a derogatory way as excessive exploitation, a management error or
failing which has to be prevented or stopped. It is the kind of exploitation which has harmful
consequences, either for the exploiters themselves or for third parties (disregard of internal or
external constraints), whether or not the exploitation strategy is balanced in the shorter or longer
term. Quantitatively speaking, it is a kind of exploitation which can hinder the long-term repro-
duction of the resource and exceed the limits of resource renewal to the detriment of future
yields, without this being an intentional strategy. This type of exploitation can also lead to
unacceptable production cost overruns and quality degradation.
The term overexploitation should not be applied neither to temporary large-scale with-
drawals from an aquifer with an irregularly recharged aquifer that make judicious use of the
regulatory capacity of its storage. Nor should it be applied to the deliberate mining of an
aquifer not receiving recharge. Whether the term overexploitation has a negative connotation
always depends on the management objective for an aquifer and the relevant exploitation
strategy. Hence an exploitation strategy characterised by temporary or permanent imbalance
should only be described as overexploitation if the imbalance is unintentional and has harmful
consequences, and is therefore contrary to the chosen strategy.

324
12.2.3 Types of aquifer systems and management conditions
Of course, individual exploitation strategies cannot be indiscriminately applied to any given

type of aquifer. Whatever the development objective may be, each particular type of aquifer
requires a particular exploitation strategy (Table 12.2).
The variety of hydrogeological conditions and the resulting size, structure and com-
plexity of aquifer systems implies that there is a matching variety of physical management
conditions, as well as approaches to groundwater resources. This diversity can, however, be
narrowed down to the most common types, presented in Table 12.2. The two most important
defining characteristics for this typology are: the flow storage ratio and aquifer/surface stream
relationships. Both can show very wide variations:
The ratio of average flow and average storage, in other words the capacity of the
storage to regulate the overall discharge rate of an irregularly recharged aquifer,
expressed in terms of duration of overall replenishment, varies from less than one year
for thin and shallow alluvial aquifers, to the order of 10 4 to 10 5 years for deep-lying
aquifers containing what is often termed fossil water.
The aquifer/river(s) relationship ranges from a very strong, continuous and permanent
connection (thin alluvial aquifers that have an unclogged connection with the river), to
complete independence (deep-lying confined aquifers and certain coastal aquifers).
12.3 Socio-economic conditions management actors and objectives
12.3.1 Management actors
Depending on an aquifers size and accessibility, the usually numerous economic actors may
have the practical and financial means as well as the legal right land-owners rights to
exploit the aquifer and possibly find it to be a profitable source of water supply. They are either
direct users (households, farmers, industries, etc.) or producers and distributors of water as a
commercial economic product (the supply of public drinking water). In addition to the ground-
water operators, there are many more actors who can directly or indirectly influence the aquifers
exploitation strategy or its properties. They comprise:
occupants of (i) the land containing the aquifer whose impacts stem from their efflu-
ents and their use of the land, e.g., agriculture with the use of fertilisers, waste dis-
posal), or (ii) the users of the underground, e.g., quarries of sand and gravel deposits,
mining, underground storage.
engineers who dam, gauge, embank or otherwise influence the streams linked to the
aquifer.
A typology of these different people can be based on whether they are users and non-
users of the groundwater, exploiters or developers, or both combined (Table 12.3). The analysis
of the system of economic actors involved, of their individual objectives, modes of action and
actual or potential conflicts of interest is a preliminary exercise which is as important to
management as the analysis of the aquifer system itself.
12.3.2 Management levels
The fact of using or influencing one and the same aquifer unites numerous actors even though
their various acts of exploitation or activities impacting on the water strategy or properties have
different individual motivations. Usually, the actors are not aware of participating in the de facto
management of a common property, and when mutually harmful influences are felt, user

325
Groundwater studies
Table 12.2 Types of aquifer system and management conditions
Type of Relationship with Appropriate

aquifer system surface water Sensitivity exploitation strategy Restrictive factors
Alluvial system Strong Sensitivity to Strategy (1) External constraints:

(often unconfined) relationship regulation of Short-term balance. clogging of banks,
streams and land Over-balance conservation of
Limited size Aquifer/river occupation possible from river- minimal flow from
and storage exchanges induced recharge springs and in rivers
possible in both Vulnerable to
Weak regulatory directions pollution from the Flow management
capacity surface and streams common
Phreatic system of Often one-way Variable recharge: Strategies (1) and (2) External constraints:
plains and plateaux connection: Perennial balance. - conservation of
and karst systems river drainage Sensitive to drought spring discharges
(except for bank and to land occupation Management and minimal
Significant storage and of the flow and stream flow
storage(storage river loss Vulnerable to regulating storage
often > 10 annual in karst areas) pollution, especially
flows) with good when diffuse
regulatory capacity Unequal density
depending on
Strong hydro- hydrographic
dynamic inertia network
Multi-layered Unequal Sensitive to land Strategy (2) External constraints

system with connection occupation (especially (for near surface
interconnected restricted to the shallow Final equilibrium unconfined aquifers):
phreatic and semi- shallow unconfined aquifers) possible after - conservation of
confined aquifers unconfined prolonged spring discharges
aquifers Varying vulnerability dewatering phase. and minimal
Significant storage to pollution. Exposed stream flow
(Storage > 100 to to risks of exchange Management of
1,000 annual flows) between aquifers with flow and storage Internal constraints
partial regulatory water of unequal moving towards (for confined
capacity quality resulting of flow management aquifers):
drilling - maximum draw-
Variable inertia downs acceptable
Deep-lying Independence Insensitive to climatic Strategy (3) Internal constraints:

confined aquifer (except in case conditions, not - maximum
of discharge vulnerable to pollution Global imbalance acceptable
Large storage by an artesian from the surface. (final equilibrium draw-down,
( > 1,000 - 10,000 spring) Exposed to risks of in some areas) - risk of intrusion
or more annual exchange between of brackish water
recharge) but no aquifers as a result of More sustainable
regulatory capacity perforation of the dewatering with
casing stabilisation (or
It cannot be adjustment) of
mobilised withdrawals
without
dewatering Management of
storage and some-
Low inertia times management
of sectoral flow
(near open
boundaries)

326
Table 12.3 Classification of the actors in groundwater development, exploitation and management
Users, beneficiaries Intermediaries, non-users
Mining industries, users Authorities and public

of subterranean space, land bodiescontracting authorities
occupants, surface operators
Developers
who are not exploiters responsible for hydraulic
development, etc. impacting
on groundwater (strategy,
properties, etc.)
Businesses and associations Producers/distributors of drinking

of irrigators, industries with water, industrial water, irrigation water
their own water supply
Developers With development of subterranean dams, artificial recharge, etc.
who are also exploiters
Mining industries, users
of subterranean space,
dewatering operators
Users-exploiters of water Producers/distributors of drinking

(without development): water, irrigation water (without
Exploiters households, industries development).
who are not developers with their own water Exploiters of geothermal energy,
supply, irrigators. distributors (heat network)
Exploiters and users
of geothermal energy
conflicts arise. The aquifer is like an unrecognised condominium without any spontaneous ad
hoc regulation. By the same token, if the exploitation or use of an aquifer as a whole has harmful
effects on a third party for example, the users of dependent surface water no individual
among the community of exploiters acknowledges a collective responsibility. None of the
aquifers stakeholders assumes the role of manager, and, no individual acts at system level or,
hence, at the level of global management.
On the one hand, there is the physical unity of an aquifer system, and on the other a
multiplicity of various actors with different and, sometimes, diverging interests. This being the
case, the management of the aquifer requires that the behaviour of the actors is guided by
applying pressure or providing incentives in different forms. In contrast to the management of a
dam which is operated by its manager, the management of an aquifer is, of necessity, partially
indirect in character, and at two levels:
(a) that of the actions relating to the groundwater, which are carried out by a set of actors
specific to each aquifer system. Such actions may be direct and intentional or are the
indirect results of various activities,
(b) that of the actions aimed at the behaviour of the actors. This corresponds to a
management authority vested with appropriate powers.
This management at two levels applies to water management in general. The particularity and
difficulty of groundwater management lies in the fact that:
the actors at level (a) do not have resource management as a spontaneous objective
and their individual fields of action do not, in general, correspond to the relevant level

327
Groundwater studies
of an aquifer system, although they alone directly operate the groundwater and the
effects of their acts accumulate and are felt within the system as a whole;
the management authorities at level (b), when they exist, can define relevant and
collective management objectives and are well suited to the required level, but, on the
other hand, they do not, as a general rule, have any means of direct action on the
groundwater. The only way they can take action in regard to the exploiters and
developers is by using various indirect management tools.
12.3.3 Management objectives
The objectives and decision-making criteria at the two management levels differ.
(i) Individual actors
The individual actors are mainly interested in micro-economic production and consumption
objectives; groundwater is, at best, a factor, a raw material, a vector, or merely an object at the
receiving end of impacts and external effects. The direct exploiters of groundwater to secure
water for their own use or to produce water for distribution manage their water production
individually on the basis of their own criteria. They do not manage the resource itself as long as
production does not suffer because of its repercussions on the environment. They do not have
resource management objectives, but do have the implicit objective of individually preserving
the conditions sustaining their water production. The same is true of the actors who remove
groundwater not in order to use it, but to remove it from the underground that is to be occupied
or exploited (dewatering of mines, land drainage, etc.), and for whom the water is not a resource
but an obstacle to be removed in order to achieve the aim of mineral production or use of under-
ground space.
(ii) Management authorities
It is at the level of management authorities that decisions are taken concerning the aquifer as a
whole and the objectives of general interest (not all of which are necessarily compatible). Con-
sequently, genuine macro-economic management objectives can be defined and placed in order
of priority:
The prevention and arbitration of conflicts between individual co-exploiters (or people
whose activities have an impact) in the same aquifer system, while ensuring respect for
external constraints. This means the prevention of conflicts between the community of
groundwater users and a wider community of users of surface or subterranean water
and space: in short, the minimum and most common objectives of water policing, in
regard to groundwater.
The conservation of the capacities and accessibility of groundwater, including its
initial dynamic levels, which implies that the exploiters who were first to arrive have
an advantage.
The conservation of the potential resource in terms of both quantity and quality, in the
interests of the present and future community of exploiters, or the prevention of excess
exploitation that would damage the renewal possibilities or the properties of the water,
i.e. the prevention of overexploitation (see sub-section 12.2.2(iv)).
The allocation of the resource, with priority being given to certain present and future
demands (e.g., drinking water supply), which can imply reservations. This means
arbitration that is more planned or State-managed, in the general interest and on a
long term basis, or is brought about by a market mechanism for water rights.

328
Intensifying the use of the resource if it is considered to be underexploited, by using

groundwater, rather than other water resources to satisfy certain demands. Getting the
most out of the resource; that is, not leaving unused a production factor unused that
could contribute to socio-economic development if it could be linked to the objective of
allocating the resource, as a matter of priority, for the most beneficial uses.
Given that groundwater is usually an integral part of the overall water resource, these
different groundwater management objectives are themselves dependent on the water
management objectives in general, and must sometimes be reconciled with subsurface
management objectives (mineral resources, subterranean space), and land management
objectives. In the last analysis, they are dependent on the wider socio-economic objectives of the
community concerned.
The objectives may be incompatible, e.g., conservation management is not compatible
with the kind of management which promotes development and therefore requires transfor-
mations and redistribution. Consequently, the choice of priority objectives is the primary
management option.
12.4 Management constraints and criteria
Groundwater resource management is subject to constraints linked to the appreciation of the

internal and external effects of the exploitation, in the light of different criteria. The internal
effects, the inevitable repercussions of exploitation on the exploitability of the resource, engender
internal constraints stemming from cost/benefit analyses based on the criteria of the exploiter.
The external effects or impacts can engender external constraints stemming from arbitration
between the interests and objectives of the users of the water produced and those of other actors.
12.4.1 Internal constraints
The internal constraints determine the limits of overall or sector exploitation in keeping with
exploitability criteria based on the dimensions of the resource and the water demand mentioned
above: discharge, location, conditions of access and mobilisation, water properties, cost. But the
criteria of the exploiters and those of a management authority are not defined at the same levels,
and can differ:
for the exploiter, the most important considerations are local productivity, the imposed
practical methods of exploitation, the quality of the water produced, security of
production, direct production costs (to be minimised in a micro-economic perspective);
for the management authority, the most important considerations are the conservation
of the resource in terms of both quantity and quality, an equitable distribution of the
conditions of access and productivity, and the overall cost of exploitation (to be
optimised in a macro-economic perspective).
It is always necessary to classify and grade these criteria in order to determine management con-
straints. If, for example, the most important criterion is that of cost, in the form of the maximum
unit production cost acceptable, this means pumping from the maximum depths practicable
and, as a result, the fixing of a ceiling for withdrawals. The conservation of the dynamic levels of
groundwater without exceeding a maximum acceptable depth established for each individual
zone is the constraint most often generated by the criteria. Where the management of a non-
renewable resource is concerned, if the criterion of duration of production is paramount, it is the
chosen duration which determines the ceiling constraint on annual withdrawals per sector.

329
Groundwater studies
12.4.2 External constraints
The external constraints are based on the desire or the need to reduce or prevent external effects
which are detrimental to other economic actors, users of dependent surface water, land occu-
pants, or, more generally, to the environment (e.g., preservation of aquatic ecosystems, which are
dependent on outlets). They can stem directly from the obligation to maintain reserved
discharges on the surface. These constraints are the result of negotiated consensus and arbi-
tration between the objectives and sectional interests of the different parties concerned, based on
socio-economic criteria which are in the general interest. They can also lead to financial mecha-
nisms of internalisation. The constraints also very often impose the maximum admissible depths
of groundwater levels.
The external constraints combine with the internal constraints, but take priority: in this
way, they can impose additional limits on the direct use of groundwater. Both the external con-
straints and, sometimes, the internal constraints can find legal expression in regulations
(prohibitions, obligations). Conversely, groundwater managers can impose constraints in order
to conserve and/or protect the resource. These result in external constraints on other actors:
on developments and/or the use of surface water when a development or use is likely
to impact on the groundwater strategy or properties in a manner detrimental to its
users;
on various activities relating to land and subsoil occupation, e.g., prevention of sub-
sidence caused by dewatering, and prevention of pollution, particularly in the sanitary
buffers around catchments of water intended for human consumption.
12.5 Management decision methods and aids
The management decisions concerning a given aquifer depend on whether the management is
direct or indirect. The decisions may be strategic (choice of strategy and general plan of exploita-
tion) or tactical (choice of machinery and technical modes of exploitation and guidance of the
production), They can be based on predictive management techniques which are themselves pri-
marily based on hydrodynamic or optimisation modelling, and management control techniques.
12.5.1 Modelling and predictive management
The principles and ways of the hydrodynamic modelling of aquifer systems are dealt with in
Chapter 11 and are the subject of an abundant literature. The models are used as tools for study
and analysis, checking the consistency between the observed data, hypotheses and calculated
results, and can also help to optimise data collection (Pfannkuch, 1975). They also simulate
behaviour, so that forecasts can be made of the consequences of actions envisaged on the aquifer
system, including exploitation scenarios, and exercises in predictive management can be carried
out. The results obtained provide useful indications about feasibility and comparative data on
different variants of the exploitation projects in well-defined time frames.
The primary exploitation answers from such models concern forecasts of changing draw-
down rates of the piezometric levels, decreasing discharge of natural outlets, and sometimes the
inferred impact on salinity. Moreover the modelling results make it possible to estimate fore-
seeable production costs in terms of the depth variable, or to compare the calculated future
states of the system within the established constraints. Such models are thus outstanding aids to
management decision-making and their use has become commonplace (Pfannkuch, 1975;
Bachmat et al., 1980; Hogh Jensen, 1987). The model must be sufficiently representative to make
a reasonably valid simulation of variations in the state of the system, even when the situation

330
differs significantly from the situation used to calibrate the model. The conceptual schema-
tisation must be relevant particularly in the case of multilayered systems and the adopted
storage coefficients must be valid, because they have a large influence on the results of the
model. It is advisable to establish margins of uncertainty (confidence intervals) for the numerical
results.
A hydrodynamic simulation model can also constitute the infrastructure of an economic
management model, especially for the purposes of optimisation, and be generally integrated
into a basin management model. Lastly, if the models are periodically updated and adjusted in
the light of the improved knowledge resulting from the operations themselves, they can also
guide management by providing revised projections.
12.5.2 Forecasting unit production costs
Unit production costs and their evolution, in constant monetary units per cubic metre of water,
are an essential ingredient of the comparative feasibility studies of projected exploitation plans.
They are used in particular as the basis for predictive cost/benefit analyses. The direct costs
borne by the exploiters are calculated on the basis of fixed terms (investments) and of variables
linked above all to the operation:
average envisaged productivity per well, pumping station, or spring,
desired total yield, possibly in stages, and number of wells programmed to achieve
this result,
expected average life of wells and pumps,
costs of initial wells and their equipment and of the future replacement costs,
time allowed for amortisation of investments,
operating costs (fuel, maintenance, etc.), based in particular on pumping levels inferred
from dynamic depth levels calculated at different times.
12.5.3 Forecasting external costs
The exploitation of groundwater can have harmful external or internal effects or impacts, as
examined in Chapter 11. A forecast of the resulting costs should also be included in the
feasibility studies and cost/benefit analyses of projects, whether or not it is used to incorporate
such costs in the direct costs borne by the exploiters. Depending on the types of impacts, these
external costs are not all readily calculable in monetary units but their likely magnitude can be
indicated.
Decreases in productivity of previous drilled wells, in particular of traditional modes
of capture, can be evaluated on the basis of compensation costs, alternative solutions,
or the costs of the resulting losses in production (in agriculture, etc.).
Degraded water properties can be evaluated in terms of the costs of remedial
measures, or by those of solutions involving alternative sources of supply.
Land subsidence or drainage harmful to vegetation can be evaluated on the basis of
the costs of remedial measures or reductions in land value (depreciation of assets).
12.5.4 Optimisation methods
Simulation and predictive calculation methods enable the different exploitation options and
scenarios to be compared in terms of their costs and expected results. However, they do not deal
with the question of determining the optimal solution. The multicriteria methods used for
analysing management options provide a new and efficient approach (Tecle and Duckstein,
1992), that augments the function-objective and single-criterion based optimisation techniques,

331
Groundwater studies
such as linear and dynamic programming (Wanakule et al., 1986). Their principles and ways of
application to aquifer management, are not dealt with here (but see Duckstein et al., 1993; El
Magnouni, 1993; El Magnouni and Treichel, 1994).
The management of a given aquifer is rarely based on a single objective or criterion, for
example, to maximise discharge or duration of production, or to minimise cost. Instead a multi-
plicity of criteria are involved. In most cases, each criterion has to do with an aspect of the
aquifers behaviour; the criteria do not have a common scale (for example, of monetary values),
however. A hierarchy of preferences has to be established and formalised for the criteria. For
example, priority can be given to the quantity of water produced, but not at any cost or at the
risk of an unacceptable deterioration in quality or an extremely short duration of production.
Information on the preferences both inherent in and established between the criteria should
logically enable the establishment of an optimal solution, also known as the best compromise.
The interconnection between the simulation model of an aquifer and a model of the value
system of the decision-maker (constraints, criteria and preferences) also constitutes a multi-
criteria management model of the aquifer (El Magnouni, 1993). But that implies the existence of
a management authority in which the power to take decisions is a centralised.
12.5.5 Management control
To check whether the changes in an exploited aquifer are in keeping with forecasts and the pro-
gramme, the management authority should have regularly updated data on the condition of the
aquifer and on the amount of water withdrawn. This requires the monitoring of two types of
significant variables: variables of condition (e.g., piezometric levels, salinity of the water pro-
duced, natural discharge outlets if relevant) and decision variables (e.g., withdrawals, overall
state of wells).
(i) Monitoring of piezometric levels
The methods and techniques of monitoring groundwater levels by piezometric networks and the
analysis of the data obtained are dealt with in Chapters 3 and 10. In addition to providing
insight into aquifer dynamics, such monitoring provides essential information for management
control. Groundwater levels are, after all the most significant state variable; they indicate the
states of flow and of storage, and reveal all the impacts. In order to monitor the exploited
aquifer, a piezometric network should have reference observation points in zones not subject to
influence from the groundwater exploitation as well as at points close to the well fields.
In the case of deep aquifers where specially drilled piezometers are rare, because of the
high costs, the following can often be used as piezometers:
unexploited unproductive water wells,
boreholes used for mining or oil exploration which pass through the aquifer and can
be partly filled up and perforated at the desired depth.
The frequency of measurement should be adapted to the type of aquifer and operation strategy.
Monthly or more frequent monitoring can be useful in the case of a low-capacity aquifer which
is intensely exploited with seasonal variations. Annual monitoring is enough for high-capacity
confined and unconfined aquifers.
(ii) Monitoring of outlet discharges
The monitoring of the discharges from springs and gaining streams, identified as outlets of
the exploited aquifer, provides essential information, particularly for evaluating the possible

332
impacts of the exploitation on boundary flows. Such monitoring, using traditional hydrometric
techniques, should be specifically adjusted to low discharge periods.
(iii) Monitoring of the salinity of yields
It is advisable to sample and analyse the main physical and chemical variables of quality
(resistivity, chloride, sulphate content, etc., temperature) of yields annually from a considerable
number of the abstraction points.
(iv) Monitoring of water production
In cases of direct management, the operating authority regularly quantifies the amount of water
withdrawn. In other cases water production monitoring can be based on several methods, some
of which are more precise and costly:
the installation of meters which are read periodically (including the measurement of
artesian well discharges) is difficult to carry out in practice when there are many
wells or boreholes and these, provide moderate yields on average;
periodic surveys, in which indirect indicators (energy consumption, area irrigated, etc.)
are measured or recorded.
Making it mandatory for the operators to declare their withdrawals can facilitate a
quantifying of withdrawals. It does not exempt the management authority from some degree of
inspection, however.
(v) Data bank and information
A data bank of computerised geological references, possibly linked to a geographical infor-

mation system (GIS), is the modern way of filing all the data collected periodically and
validated. It should enable tables, graphs and updated maps to be produced, including reports
on levels and yields. these are the management control data on the operation of the managed
aquifers. The output could be published and distributed periodically as a news-sheet among
users as well as regional and national administrative bodies responsible for water, the environ-
ment and economic development.
12.6 Management instruments
Collective management systems similar to the surface hydraulic management systems used to
manage a catchment basin are rarely used for the direct technical management of an aquifer,
except for specific cases such as subterranean dams, artificial recharge installations, etc. The
actors directly exploiting or impacting on groundwater are not individually involved in resource
management. This management devolves on authorities, who set up indirect management
instruments which guide the behaviour of such actors so that their decisions will all be as
consistent as possible with the management objectives established in the general or community
interest. These indirect management instruments are primarily of two kinds: regulatory and
economic.
12.6.1 Regulation
Legislation consistent with the water laws of the country subject individual and collective
operators to certain obligations:

333
Groundwater studies
either a simple strategy for statistical purposes with a declaration of withdrawals and/or
pumping units being exploited, installed or abandoned by the user, and a kind of
registry kept by the management authority. There is usually a discharge or depth
threshold ;
or, in addition, a prior authorisation strategy requiring a licence to install a catchwork,
and/or to withdraw water; here again with application of thresholds.
Other legislation can restrict or prohibit acts by land or subsurface users which are likely
to lead to a deterioration in the natural properties of the groundwater, with possible zoning
(a particular case being the sanitary perimeters established to protect the public catchworks of
water for human consumption). A water policing authority or water or subsurface adminis-
trative body can be responsible for applying such legislation.
12.6.2 Financial incentives
Financial incentives, also established within a legal framework, can be used to guide and change
the behaviour of the private and collective economic actors operating or impacting on ground-
water, in the way desired by the management authority. Such incentives can function in two
ways:
Positive incentives or aid: stimulation of exploitation by investment credit, subsidies,
free insurance against prospecting and well-drilling risks, tax refunds on energy
sources for pumping; including aid given to actions to lessen pollution.
Negative incentives or taxation: restriction on or dissuasion from exploitation by the
establishment of royalties on withdrawals above a number of ceilings which are
adjustable according to zone, season or user category, in the context of a resource
allocation policy. Also, taxation of acts which generate pollution, on the basis of the
polluter pays principle..
Lastly, the effectiveness of the measures can be enhanced by providing all the actors
concerned with information, in order to ensure greater understanding of problems and explain
objectives which are in the common interest and the reasons behind the measures taken. In
general, it is advisable that the powers to make regulations and provide financial incentives and
information be vested in the same management authorities. In all cases, the remit of these
authorities should cover whole physical units of management (aquifer systems), irrespective of
the size of such systems and whether or not they are divided between several administrative
and political districts.
12.7 The future of groundwater management
Groundwater management is likely to become more of a reality in the future, to be more

integrated with water and land management, and, at times, to be more ambitious.
12.7.1 Towards groundwater management implementation
In many countries groundwater management still largely comprises statements of intention and
the installation, rather than the implementation of regulatory instruments. This is because the
applicability and application of these instruments require resources and motivation, but these
are often lacking. As long as groundwater management is only dealt with in terms of describing
such instruments and attributing the responsibilities for their application, and defining
administrative powers, it will remain hypothetical. It is foreseeable or in any case, desirable
that in future groundwater administration should put greater emphasis on actual water mana-

334
gement which, should be in some form co-management. This will involve more aware and
better informed managers, common and priority objectives, and management authorities
adapted to each unit formed by natural structures.
12.7.2 A more integrated management
The management objectives should not be restricted to those of the exploiters or direct users
of groundwater, but be consistent with groundwater management objectives in general
groundwater as resource and environment and with those relating to land and subsurface
occupation and use. This could imply a variety of priorities, depending on the individual case
and zone:
production of water by the maximised intensive exploitation of groundwater,
preferable to withdrawals from main streams, especially when such surface water is
poor in quality or governed by low-quality objectives; or, on the contrary, the long-
term conservation of current exploitation conditions;
production of high-quality water, implying controls on land occupation;
conservation of reserved surface discharges and preservation of the quality of
perennial surface water dependent on groundwater runoff: extension to groundwater
forming the upper waters of a reserved discharge stream, and quality objectives
assigned to the latter;
land use (agricultural production, urbanisation, etc.) with impacts on the groundwater
strategy and/or properties that can be tolerated up to a certain point;
final dewatering of a subsurface space, implying the permanent transformation of the
local groundwater strategy.
This will mean wider consultation, encompassing a variety of partners with, at times, con-
trasting motivations and interests, and, in general, wider political arbitration, with the particular
objective of taking account of long-term collective interests (future generations).
12.7.3 A more ambitious form of management
In certain cases, groundwater management will evolve towards a form of aquifer management
that is more ambitious than the mere extensive or intensive exploitation of natural flows, and
involves more active mobilisation of groundwater storage and actions relating to aquifer
recharge:
increasing use of the exploitation strategies of the second type described above, using
the multi-annual regulation capacities of certain aquifers, particularly to compensate
for surface shortfalls in runoff during dry years (replacement resource or minimal flow
support), sometimes based on spring tapping at different levels or on subterranean
dams. The regulatory function of aquifers containing fairly significant and manoeu-
vrable quantities of water can be put to greater use, together with surface dams, to
control irregular flows, especially in countries where the useful capacities of surface
reservoirs are rapidly deteriorating as a result of sedimentation;
development of artificial recharge, often combined with the preceding strategy;
utilisation and maintenance, even stimulation, of the self-purifying capacities of certain
aquifers in the management of the properties of water, particularly in systems which
reutilise waste water;
reservation of certain soils for the production of water of excellent quality. Hydro-
culture might take precedence over agriculture, taking over from it within the
framework of a land development policy and an appropriate agricultural policy;

335
Groundwater studies
exploitation of non-renewable resources, managed in a long-term perspective and a

view to the preparation of alternative sources. One way this could be achieved is by
investing some of the profits gained from using the abstracted water, to enable changes
in the water economy.
ANONYMOUS, 1991. Sustainable Use of Groundwater Problems and Threats in the European Communities.
E.C.E., RIVM-RIZA, Rep. No. 600025001 to Ministers Seminar on Groundwater, The Hague,
2627 Nov. 1991, 80 pp.
ASCE, 1972. Groundwater Management. ASCE Manuals and Reports on Engineering Practice,
No. 40, New York, 216 pp.
BEAR, J.; LEVIN, O., 1967. The Optimal Yield of an Aquifer. In : Management of Aquifers. In :
Alimentation artificielle et amnagement des nappes aquifres. Internat. Assoc. Hydrol. Sciences,
Proceed. Sympos. Haifa, IASH/UNESCO-DHI, Publ. IASH, No. 72.
BACHMAT, Y. ET AL. 1980. Groundwater Management. The Use of Numerical Models. Amer.
Geophys. Union, Water. Res. Monogr. 5, 127 pp.
CUSTODIO, E.; GURGUI, A., (eds.), 1989. Groundwater Economics. United Nations International
Symposium and Workshop, Barcelona, Oct. 1987, Elsevier, 625 pp.
DUCKSTEIN, L.; TREICHEL, W.; EL MAGNOUNI, S., 1993. Multicriterion Analysis of Groundwater
Management Alternatives. University of Paris-Dauphine, Cahier du LAMSADE No. 114,
Paris.
EL MAGNOUNI, S., 1993. Mthodologie daide la dcision pour lvaluation et la gestion
multicritre des ressources en eau souterraine. Doctoral thesis, Univ. Nancy, Nancy, 134 pp.
EL MAGNOUNI, S.; TREICHEL, W., 1994. A Multicriterion Approach to Groundwater Management.
Water Resources Research, Vol. 30, p. 188196.
ERHARD-CASSEGRAIN, A.; MARGAT, J., 1983. Introduction lconomie gnrale de leau. Masson, Paris,
361 pp.
HUBERT, P., 1984. Chap. XI : Aspects de la notion de ressource en eau, et Chap. XIV : Gestion des
eaux souterraines. In : Eaupuscule ; une introduction la gestion de leau. Ellipses, Paris, 192 pp.
HOGH JENSEN, K., 1987. The Value of Groundwater Models for Planners and Decision-Makers. UNESCO,
Tech. Doc. in Hydrology, Project IHP.III 2-4(b), Paris.
LLAMAS MADURGA, M. L., 1975. Motivaciones extraeconmicas en la utilizacin de las aguas subterrneas;
la hidroesquizofrenia. National conference on general and applied hydrology, Zaragoza,
Spain, 1975.
LLAMAS, R.; BACK, W.; MARGAT, J. , 1992. Groundwater Use; Equilibrium between Social Benefits
and Potential Environmental Costs. Applied Hydrogeology, Vol. 2, pp. 314.
MANDELS, S., 1979. Problems of Large-Scale Groundwater Development. Journal of Hydrology,
Vol. 43, p. 43943.
MANDELS, S.; SHIFTAN, Z. L., 1981. Groundwater Resources; Investigation and Development. In:
Criteria for the Regional Exploitation of Groundwater, Acad. Press, New York, 369 p.
MARGAT, J.; SAAD, K. F., 1983. Concepts for the Utilization of Non-Renewable Groundwater
Resources in Regional Development. United Nations Natural Resources Forum, Vol. 7,
pp. 37783.
MARGAT, J., 1985. Groundwater Conservation and Protection in Developed Countries. IHA, Report
18th Congress: Hydrogeology in Service of Man, Vol. XVIII, Part 1, Keynotes paper, Cambridge,
pp. 270301.
MARGAT, J,. 1992. The Overexploitation of Aquifers, in: Selected Papers on Aquifer
Overexploitation. I.A.H. Selected Papers on Hydrogeology, Vol. 3, pp. 2940.

336
PFANNKUCH, H. O., 1975. Study of Criteria and Models Establishing Optimum Level of Hydrogeologic
Information for Groundwater Basin Management. Water. Res. Research Center, Univ.
Minnesota, Minneapolis, Bull. 81, 165 pp.
SAHUQUILLO, A.; ANDREU, J.; ODONNELL, T., (eds.), 1989. Groundwater Management, Quantity and
Quality. Symposium 25 Oct., Benidorm. IAHS Publ. No. 188, 2 Vols., 1015 pp.
SCHNEIDER, H., 1977. Management of groundwater resources in arid zones. Conf. Different Strategies,
Development and Management of Deserts, Vol. 11, Sacramento, 25 pp.
TECLE, A.; DUCKSTEIN, L., 1992. Concepts of Multicriterion Decision-Making. In: Decision Support
Techniques for Integrated Water Resources Management, UNESCO, Paris.
UN/EEC, 1983. Groundwater Protection Strategies and Practices. UN/EEC, Seminar, Athens, Oct.
1983, Proceed. Publ. Minist. Energy and Natural Resources, Athens, 501 pp.
UNESCO et al., 1983. Groundwater in Water Resources Planning. Proceed. Internat. Symposium,
Coblenz, Aug.Sept., 3 Vols., UNESCO, IAH, IAHS, Nat. Comm. Fed. Rep. Germany.
WANAKULE, N.; MAYS, L. W.; LASDON, L. S., 1986. Optimal Management of Large-Scale Aquifer:
Methodology and Application. Water Resources Res., Vol. 22, No. 4.
WIENER, A., 1972. Role of Groundwater and Surface Water in the Management of a Basin. In: The
Role of Water in Development, McGraw-Hill, New York, pp. 385 93.
WRIGHT, C. E. (ed.), 1980. Surface Water and Groundwater Interaction. Studies and Reports in
Hydrology, UNESCO, Paris, Vol. 29, 123 pp.
Selected references on case studies of groundwater management
ANONYMOUS/K.F.U.P.M. 1987. Groundwater Resources Evaluation in Saudi Arabia and Long-Term

Strategic Plan for Fresh Groundwater Use. King Fahd University of Petroleum and Minerals,
Dahran.
ANONYMOUS/UNESCO, 1972. tude des ressources en eau du Sahara septentrional, Algrie-Tunisie.
UNESCO, Final Report REG 100, 7 vols., technical annexes, Paris.
COLLECTIVE WORK, 1981. Lexploitation des rserves deau souterraine en zones arides et semi-arides. Doc.
CEFIGRE, Sophia-Antipolis, Sept. and Bull. BRGM, Sect. III, 1982, No. 2, Orlans. This work
contains articles on a variety of case studies: aquifers in large sedimentary basins in Algeria-
Tunisia, Libya-Egypt, Senegal and Mauritania, Arizona and Australia.
COLLECTIVE WORK, 1983. Groundwater Protection Strategies and Practices. UN-ECE Seminar, Athens,
Greece. This seminars proceedings include reports on groundwater management in
different countries: Austria, Bulgaria, Canada, Federal Republic of Germany, Finland,
France, Greece, Hungary, Netherlands, Poland, Switzerland, Ukraine, USSR.
FOSTER, K. E., 1977. Managing a Finite Groundwater Supply in an Arid Area; the Santa Cruz Basin
Example in Arizona. In: Proceed. Conf. On Alternative Strategies for Desert Development, May,
Sacramento, Ca., 21 pp.
HABERMEHL, J., 1980. The Great Artesian Basin, Australia. B.M.R. Journal of Australian Geology and
Geophysics, Vol. 5, pp. 918.
MARGAT, J., 1990. La gestion des eaux souterraines/France. Prospective studies group, seminar,
valuation des politiques publiques de lenvironnement, meeting of 15 February 1990,
Paris, 18 pp.
SALEM, O. M., 1992. The Great Manmade River Project; A Partial Solution to Libyas Future Water
Supply. Water Resources Development, Vol. 8, pp. 2708.
SCHEFTER, J. E., 1985. Declining Groundwater Levels and Increasing Pumping Costs in Floyd County,
Texas; A Case Study. National Water Summary, 1984. Hydrologic Perspectives, USGS, Water
Supply Paper 2275, Washington.

337
Groundwater studies
SHAHIN, M., 1991. Groundwater Management. In: Assessment of Groundwater Resources in Egypt,
IHE, Report Series No. 23, Delft, 71 pp.
SOLIMAN, M. M., 1984. Groundwater Resources Planning and Management in the Libyan Sahara.
In: M. M. Soliman (ed.), Groundwater Management in Arid Regions, Contrib. for the Intern.
Course on Groundwater Hydrology, UNESCO/Ain Shams University, Cairo, 2 Vols., 275
and 136 pp.

338
(A) The influence of changes in
13 h y d ro g e o l o g i c a l c o n d i t i o n s
o n t h e e n v i ro n m e n t
( B ) G ro u n d w a t e r q u a l i t y
p ro t e c t i o n
13A The influence of changes in hydrogeological conditions

on the environment
13A.1 Introduction
Many artificial disturbances in the environment result from changes in hydrogeological con-
ditions. Prolonged fluctuations in groundwater levels and changes in the chemical composition
of groundwater can lead to:
substantial disturbances in the functioning of groundwater and surface ecosystems,
activation of geodynamical and engineering geological processes (e.g. karst, landslides,
subsidence, erosion, flooding, etc.),
changes in soil and cryogenic processes and surface water flow,
seismic activity.
Because these negative environmental impacts occur in regions where the groundwater
regime is disturbed, forecasting them is associated with forecasting the disturbed groundwater
regime. Therefore, the purpose of hydrogeological investigations is not only to study and fore-
cast external effects on the groundwater, but also the influence of changing hydrogeological
conditions on the environment (Kovalevsky, 1994). Understanding these consequences requires
monitoring of the groundwater system and other related environmental components. It is
essential to predict the development of critical and disastrous situations to the environment
caused by groundwater, and to apply preventive measures. Three types of disturbance of the
groundwater regime have been identified:
artificial discharges or increases in the outflow components of the groundwater
balance that are characterised by lowering of the groundwater levels;
artificial recharge of groundwater or increase in the inflow components characterised
by groundwater level rises, and
changes in the chemical balance of the groundwater, characterised by changes in the
groundwater quality.

339
Groundwater studies
High rates of groundwater abstraction and the resultant decline in groundwater heads
take place in regions of centralised water abstraction for domestic and industrial purposes, in
mine water discharge areas, in the regions of oil and gas fields, and in irrigation and land
drainage zones. First of all, the exploitation disturbs the groundwater regime and balance. The
groundwater level decline may lead to a decrease in groundwater recharge, infiltration and
evaporation. The formation of cones of depression may lead to a reduction in the yield of wells,
the disappearance of springs and a reduction in groundwater discharge to rivers. Under the
influence of groundwater withdrawal, salt water encroachment may occur, sea water intrusion
into aquifers is observed, and in desert areas salt water contours approach the fresh ground-
water lenses. Changes in the water temperature in cones of depression is also observed
(Kovalevsky, 1976). All these factors change the salt balance and the quality of the groundwater
and often make the groundwater unfit for use.
All these processes can be studied, and there are many types of models to forecast them
(e.g. Goldberg, 1976; Gavich, 1980).
13A.2 The environmental impact of groundwater withdrawal
13A.2.1 Effects on the relation between groundwater and surface water
Groundwater development changes the interaction of surface water and groundwater; ground-
water discharge to rivers either decreases or the rivers begin to recharge the aquifer. As a result
the surface water resources decrease. The flow of small rivers decreased by 70 80 percent in
Belarus, by 1015 percent in the Moscow region (Russia), and some rivers of Kazakhstan have
fallen dry. The water withdrawal in lakeland areas also leads to reduction of water depths,
drying, and the loss of recreational potential and fish production. For example , the decrease of
cold water discharge into Lake Pleshcheevo (Russia) has led to an increase in the temperature
of the bottom layer and the water of the lake. This has intensified the decomposition of silt,
enriched the lake water with organic matter, and reduced the oxygen content, resulting in a
decline of the productivity of some rare fish in this region. Marshland reclamation is also
accompanied by enrichment of water with organic matter, iron, manganese, and nitrogen
compounds. The acidity of water increases and the solubility of heavy metals, including toxic
ones, increases. The base flow decreases, and deflation and spontaneous ignition of peat land
may occur.
13A.2.2 Land subsidence
One widespread consequence of intense groundwater development is the consolidation of

sediments, resulting in land subsidence and fractures that cause the deformation of buildings,
roads, etc.; Konoplyantsev and Yartseva-Popova (1983) mention examples in United States of
America, Japan, Mexico, Italy, The Netherlands, and other countries. Subsidence rates range
from a few millimetres to 75 cm per year, and the resultant subsidence can be several metres
(e.g. up to 8.8 m in California, United States of America). The subsided volume in Texas (United
States of America) is 22 percent of the volume of groundwater pumped from sandy and clayey
sediments. Subsidence is accompanied by a change in the state of both water-bearing and
confining layers. Subsidence is caused by the release of the hydrostatic pressure. At first the
compaction is due to deformation by compression of the layers; this is succeeded by plastic
deformation that is accompanied by the displacement of rock particles. The interaction of par-
ticles is increased by the forces of molecular attraction, resulting in an increase of the amount of
bound water. These processes are irreversible and ground surface levels recover only partially

340
The influence of changes in hydrogeological conditions on the environment
when water withdrawal is terminated or decreased. Subsidence can be simulated by the

numerical solution of equations of two-dimensional unsteady flow and equations of loamy soil
deformation.
13A.2.3 The influence on karstification
Groundwater development in carbonate rocks very often results in the intensification of karst
processes, formation of sinkholes and collapse of the ground surface, destroying buildings and
damaging to the land and infrastructure. Lowering the piezometric levels of confined or semi-
confined groundwater bodies to below river level may result in the aquifer being recharged with
river water rich in free carbon dioxide, which may bring about the intensification of the dis-
solution of carbonates. Some pollutants and micro-organisms in the groundwater may act as
catalysts of the karstification process. For example, as a consequence of the exploitation of a
semi-confined aquifer in the Moscow area its piezometric level dropped by 1070 m below the
levels of the river and of the phreatic groundwater and turned this region into a recharge area.
The aggressive river water and polluted phreatic groundwater encroached on the semi-confined
aquifer and changed the hydrochemical situation:
the groundwater salinity increased 1.52 times,
the calcium content increased 1520 times,
the oxygen content increased 210 times,
the pH decreased from 7.08.0 to 4.95.5,
the free carbon dioxide reached values of 20 mg/l or more,
the salt discharge by the pumped water increased to 100 t/year per km2,
the strontium content, characteristic for the leaching process, increased from
0.10.21 mg/l to 1618 mg/l,
the residence time, estimated by the tritium method, decreased from about 100 years to
212 years,
the helium content decreased 1520 times.
The river water that reached the semi-confined aquifer caused an increase of the
number of micro-organisms in the groundwater to 14,000/ml divided over 30 species,
including:
four species of filamentous bacteria,
five species of diatoms,
seven species of green algae,
three species of green-blue algae,
three species of euglenes,
infusoria,
shell amoebae.
The carbon dioxide content reached 200 mg/l in zones with a high content of micro-
organisms that produce weak acid (Kovalevsky and Zlobina, 1988). The leaching of carbonates
in the presence of wastewater of a thermal power plant increased 45 times, because the high
groundwater temperature resulting from the thermal pollution promoted an increase of the
acidity and of the calcium deficiency (Zlobina, 1988). Thus, water withdrawal may create not
only hydrodynamic, but also thermal, hydrochemical, and hydrobiological anomalies. These
factors may intensify the karst processes, the appearance of sinkholes in areas where they were
unknown during the first decades of groundwater exploitation.
Karstification resulting from of groundwater exploitation has been numerically simulated
at the theoretical level only (Mercado, 1977).

341
Groundwater studies
13A.2.4 Effects on plants and animal life
Large scale pumping or drainage of groundwater inevitably results in a lowering of the water
table (or phreatic surface) and the capillary fringe, expansion of the zone of aeration, a change in
the soil moisture content and consequently an impact on the environment. This impact may be
positive or negative (Kovalevsky, 1994).
We know numerous examples of the dewatering of forest areas in regions with artificial
drainage of groundwater, resulting in a transformation of the vegetation and of the quality and
productivity of the forest. The moisture-loving vegetation (hydrophytes), e.g. willow, reed and
reed mace, are the first to perish in river valleys in areas that have an arid climate and deep-
rooting phreatophytes, e.g. salt cedar, honeysuckle and dog-rose show considerably depressed
growth. On the other hand, when the groundwater development leads to dewatering of wet-
lands in the humid zone, the yield of grasses, especially on flooded meadows can be substan-
tially increased because low yielding species are replaced by much higher-yielding ones. A few
examples:
On artificially dewatered lands of the Vologda Region (Russia) sedge (150300 kg/ha)
has been replaced by high quality meadow grass (20002200 kg/ha).
On flooded land only wilted bushes, low birch, and poor quality pine can grow, while
on dewatered lands mixed coniferous/deciduous forests of good quality can thrive.
The natural species succession of forests in the piedmont region of the Caucasus and in
the region adjoining the Aral sea is: willow/alder/poplar/oak and hornbeam. Drastic
artificial lowering of groundwater levels has caused these forests to disappear.
The correlation between the productivity of forests and other plant communities and the
depth to the water table or to the capillary fringe for areas with different climatic and landscape
conditions can be the basis for predicting the environmental consequences of water withdrawal
and land drainage. The decline in groundwater level and the resulting desiccation of the forest
litter is leading to locust plagues in the southern parts of the former USSR and to the prolifer-
ation of the encephalitic tick in the northern parts. On the other hand, better drainage leads to a
reduction of the population of certain rodent species carrying febrile diseases like tularaemia
and leptospirosis.
The groundwater development in small river valleys (e.g. the Ural river, Russia) may
result in a decrease in flood rates, and the cessation or reduction of the flooding of floodplains
and oxbow lakes. This influences the reproduction of fishes and adversely affects fishery
activities.
Drainage operations in Belarusia, Lithuania and Karelia have led to:
an increase in groundwater salinity (to 300 3,000 mg/l),
ablation of soils (up to 15 t/km per year),
appearance of ravines,
downcutting by rivers and watercourses,
shrinkage and spontaneous ignition of peat,
increase of karstification processes in the beds of canals carved into limestone..
13A.2.5 Influence on seismicity
One of the most controversial questions is the possible influence of changes in hydrogeological
conditions on seismic activities, comparable to the seismic activity that occurs in the wake of oil
and gas exploitations. Valeisha (1983) considers that one of the earthquakes in Tashkent is
related to the large thermal water development that resulted in the hydraulic head being
lowered by 150 m. The 1985 earthquake in Mexico may also be related to intense groundwater
exploitation. Shtengelov (1980) suggests that most of the severe earthquakes in the Crimea

342
(Russia) coincide with sharp natural drops of groundwater levels; he also found a relationship
between high seismic activity and long term pumping tests.
13A.3 The impact of man-induced groundwater level rise on the environment
13A.3.1 General
Artificial groundwater recharge is a widespread cause of a disturbed hydrological regime char-

acterised by rising groundwater levels and the water table subsequently stabilising at a higher
level than before the disturbance. The increased recharge may be due to:
field irrigation systems,
leakage from canals, drinking water mains and sewerage systems,
industrial wastewater disposal,
seepage losses from reservoirs, ponds and sludge tanks,
street flushing.
Another cause of the rise of the water table is deterioration of the natural subsurface
outflow. It may be caused by:
subsurface obstructions (e.g. tunnels, embankments, foundations),
filling up of gullies,
permanent rise of surface water levels (dammed rivers).
The effects of the rise of the water table depend on climatic, geological and other local
conditions, but they can be summarised as follows:
flooding of settlements and urban areas, agricultural land and forests,
waterlogging and salinization of soils,
changes in the groundwater and salt balance,
changes in the engineering properties of rocks that may result in:
changes in bearing capacity,
greater susceptibility to landslides, subsidence, seismicity, etc.,
changes in the ecological situation.
13A.3.2 Effects of waterlogging
Waterlogging is probably the most widespread and serious ecological effect, resulting in a
disturbance of the water, salt, air and temperature regimes in the root zone of the natural and
planted vegetation, causing a decrease in yield and productivity. The critical depth to water
depends on the type of soil, the climate and the sensitivity of the vegetation to high soil
moisture and increasing soil salinity. The impact of water table rise on the functioning of eco-
systems depends on the relation between the depth of the root system and the capillary fringe,
and the water requirements of plants at different vegetative stages. Each type of vegetation has
an optimal groundwater level, below this level the root system dries up and above it air
exchange deteriorates and the root system is asphyxiated or will rot away. The changes in
productivity and composition of the vegetation are an indicator of the soil moisture conditions.
Over-moistening and hydromorphism of the former automorphic soil layer results in an
increase of the acidity of the soil, soil gleying, change in humus content of the soil and in the
exchange of cations. This brings about a drop in soil temperature, because of the heat expen-
diture for phase transitions during the evaporation of surplus soil moisture and the melting of
frozen soils, etc. This leads to a change in seepage properties of water bearing formations
(Maslov, 1970). Changes in the water, air and thermal regimes of soils, in the related redox

343
Groundwater studies
conditions and in the chemical composition of the soil moisture, affect the structure of the soil.
Changes in the formation of humus, in the amount of organic substances present and in the
minerals and toxic compounds lower the biological activity of soil micro-organisms and soil
productivity. In turn, this favours the above mentioned transformations of surface ecosystems in
the flooding zone and their biological productivity.
Considerable changes may occur in soils of the humid zone of northern Russia where the
climate makes the environment especially vulnerable to disturbances of the groundwater
regime. In the permafrost areas, waterlogging will change the cryogenic situation. The warming
effect of the rising groundwater causes the degradation of the frozen ground and reduces the
thickness of the permanently frozen soil layers.
An undue increase of the moisture content of soils and rocks induces changes in their
physical and engineering properties. When the specific weight and consistency index increase,
this causes a change in the general deformation modulus, internal friction, binding angle and
shear stress parameter. This results in an increase of the sensitivity to vibration, clay plasticity,
and a decrease in the binding of soil particles, resulting in a decrease in the carrying capacity of
the soils, the compression of subsiding soils, fine sands turning into quicksand, and disturbance
of the structural bonds in soil.
13A.3.3 Effects on agricultural land
Land reclamation and especially surface water irrigation under semi-arid conditions strongly
affects the groundwater balance of the involved area; unavoidable results are:
changes of the annual groundwater regime,
changes in the hydrochemical composition ,
changes in the interrelationship between the phreatic and the (semi-) confined aquifers
and in the groundwater drainage conditions,
and in the later stages:
waterlogging and subsequent secondary salinization of the soil.
The methodology of predicting the hydrodynamic and hydrochemical groundwater regime
under irrigation is well known (Averyanov, 1978; Baron et al., 1981; Katz and Pashkovsky, 1985).
To counteract waterlogging and secondary soil salinization, artificial land drainage, e.g. by
a horizontal subsurface pipe drainage system (ILRI, 1974) is required when, at the end of the wet
season, the phreatic level is less than 12 m below the surface.
13A.3.4 Effects of surface water reservoirs
In the areas downstream of surface water reservoirs, groundwater levels may drop and con-
sequently meadow-alluvial soils may transform into podzolic and even dry-river-valley soils.
They will become overgrown with forests and bushes, the sod will be destroyed, dunes may
form and soil fertility will decrease. Upstream of the reservoir, where the groundwater levels
rise, comparable transformations will occur.
Reservoir construction usually contributes to the groundwater level rise and waterlogging
of the area adjoining the reservoir, where flooding, undercutting and reworking of the lower
part of the shore, and erosion of the old landslide slopes reduces the soil stability, reactivating
old landslides and sometimes causing new ones. Models for the simulation of the effect of
surface water reservoirs on the groundwater regime have been developed by Abramov (1978)
and Shestakov (1965).
The intensification of subsidence processes in waterlogging and groundwater discharge
zones are not rare. In the area of the Kakhasvskoe Reservoir (Russia), subsidence effects

344
occurred in a strip that is 100300 m, in some places 600 m, wide. Along tens of kilometres of the
shore landslides occurred in a strip 70100 m inland.
It should be noted that most negative ecological consequences of the changes in the
hydrogeological regime around reservoirs can be averted by draining the land properly, so as to
avoiding waterlogging.
13A.3.5 Effects on karstification
Changing groundwater levels in limestone results in changes into the depth of the zones where
karst formation is most intense, as can be observed in the areas around the Verkhnekamskoe and
Bratskoe reservoirs (Russia) (Pechorkin et al., 1980).
13A.3.6 Biological effects
Waterlogging may pose a health risk (Elpiner and Boer, 1990):

Areas with groundwater at or very close to the surface may become breeding places
for vectors of many diseases, e.g. malaria mosquito (anopheles), and ideal habitats for
the propagation of worms (ascarides, etc.);
Groundwater recharge by sewage may result in a deteriorating sanitary situation, in the
contamination of groundwater and soil, and in the outbreak of epidemics such as cholera;
People living in damp dwellings in areas with poor drainage show high rates of
respiratory diseases, tuberculosis, rheumatism and of stomach and intestinal diseases.
13A.3.7 Effects on seismicity
An increase in formation pressure may lead to high seismic activity. An example is the deep well
injection of 21,000 m per month of industrial wastewater in the aquifer at a depth of 3,671 m in
the Denver area (Colorado, United States of America). This injection caused earthquakes with
epicentres near the injection well. After the injection was stopped, the number of earthquakes
decreased sharply. A similar reaction to deep well injection has been observed on the Russian
Plateau in the Dimitrovgrad region. The storage in groundwater reservoirs of large masses of
water for irrigation in the agricultural areas of the northern Crimea has also resulted in seismic
activity (Lushchik et al., 1988).
The diversity of technical impacts on seismic activity may be explained by the specificity
of tectonic movements. The artificial increase in loading is in the descending block, where it
stimulates seismic activity, in the ascending block it restrains it. Consequently, water withdrawal
in the ascending block will promote this process. It must be remembered that not every deep
well abstraction or injection of water results in earthquakes, because the hydrodynamic factor is
the trigger rather than the cause of the seismic activity.
13A.3.8 Effects in urban areas
The consequences in urban areas are complicated and varied. High rate groundwater with-
drawal, groundwater outflow by subway and deep drainage structures, regulated storm runoff,
snow removal, interception by asphalt roads and roofs, may cause an increase of the output and
a decrease of the input components of the groundwater balance. On the other hand, ground-
water inflow via embankments and deep foundations, leakage from water supply mains and
sewage water disposal systems, industrial wastewater leakage, leakage from surface water
storage ponds, street washing, deep percolation of river water, etc. bring about an increase of the

345
Groundwater studies
groundwater balance input component and hamper the outflow. The summation of the effects
may give a different outcome, even for different parts of a single city.
Depending on the lithology of the soil and rocks, waterlogging may result in land sub-
sidence or in a rise of the surface as a consequence of swelling. Uneven subsidence on a small
scale may result in the deformation and eventual collapse of buildings and infrastructures.
Predictions of waterlogging development in urbanised areas are made both by modelling
and by extrapolation of stable trends in groundwater levels.
13A.4 References and additional reading
ABRAMOV, S. K., 1978. Predicting and Preventing of Groundwater Flooding of the Territories under
Construction. Stoiizdat, Moscow, 177 pp.
AVERYANOV, S. F., 1978. Control of Irrigated Land Salinization. Kolos, Moscow, 288 pp.
BARON, V. A.; BETSINSKY, T. A.; BOTOVA, K. D.; KATS, D. M.; PAFYONOVA, L. I.; PETROV, P. G., 1981.
Prediction of Groundwater Regime in Irrigated Areas. Nedra, Moscow, 246 pp.
ELPINER, L.I.; BOER, S. A., (eds.), 1990. Manual for Predicting Medical and Biological After-Effects of
Hydrotechnical Constructing. USSR Acad. Sc., Moscow, 172 pp.
GAVITCH, I. K., 1980. Theory and Practice of Using Simulation in Hydrogeology. Nedra, Moscow, 359 pp.
GOLDBERG, V. M., 1976. Hydrogeological Predictions of Groundwater Quality in Well Fields. Nerd,
Moscow, 152 pp.
ILRI, 1974. Drainage Principles and Applications. ILRI Publication No. 16, Wageningen, 4 Vols.
KATS, D. M.; PASHKOVSKY, I. S., 1985. Reclamative Hydrogeology. Nauka, Moscow, 147 pp.
KONOPLYANTSEV, A. A.; YARTSEVA-POPOVA, E. N., 1983. Land Surface Subsidence Due to Groundwater
Table Lowering. VIEMS, Moscow, 48 pp.
KOVALEVSKY, V. S., 1976. The Study of Groundwater Regime under Exploitation. Nedra, Moscow, 200 pp.
, 1994. The Impact of Groundwater Condition Changes on the Environment. Nauka, Moscow,
138 pp.
KOVALEVSKY, V. S.; ZLOBINA, V. L., 1988. Location of concealed areas with intensified karst-
suffosion processes using helium survey. IAH XXI Congress, Guilin, China, Proc. part II,
pp. 77985.
LUSCHIK, A. V.; LISICHENKO, G. V.; YAKOVLEV, E. A., 1988. Groundwater Regime Formation in the
Areas of Active Geodynamic Processes. Naukova Dumka, Kiev, 162 pp.
MASLOV, V. S., 1970. Groundwater Regime in Oversaturated Soils and its Regulation. Kolos, Moscow,
150 pp.
MERCADO, A., 1977. The Kinetics of Mineral Dissolution in Carbonate Aquifers as a Tool for
Hydrological Investigation. J. Hydrol., Vol. 35, pp. 37584.
MINKIN, E. L., 1973. Groundwater and Surface Water Interrelation and its Significance in Solving some
Hydrogeological Water Management Problems. Stroiizdat, Moscow, 103 pp.
PECHORKIN, I. A.; PECHORKIN, A. I.; KACHENOV, V. I., 1980. Theoretical Basis of Geological Process
Predictions on Reservoir Shores. PGU, Perm, 85 pp.
SHESTAKOV, V. M., 1965. Theoretical Basis for Assessment of Groundwater Lowering Aflux and
Drainage. MSU, Moscow, 250 pp.
SHTENGELOV, E. S., 1980. On the Impact of the Upper Jurassic Aquifer on the Crimea Seismicity. Water
Resources, No. 2, pp. 57 65.
VALEISHA, V. O., 1983. Formation of Hydrogeodynamic Regime under the Impact of Counter
Regional Loads on the Aquifer. Moscow, thesis MSU, 51 pp.
VERIGIN, N. N., 1975. Filtration from Ponds and Reservoirs, Kolos, Moscow, 303 pp.
ZLOBINA, V. L., 1988. The Impact of Groundwater Exploitation on Developing Karst-Suffosion Processes.
Nauka, Moscow.

346
Groundwater quality protection
1 3 B G ro u n d w a t e r q u a l i t y p ro t e c t i o n
13B.1 Introduction
Groundwater, a renewable and finite natural resource, vital for mans life, social and economic
development and a valuable component of the ecosystem, is vulnerable to natural and human
impacts. Earlier, little attention was paid to the protection of groundwater quality, mainly
because people were unaware of the threats to this hidden resource. The idea that groundwater
is well naturally protected by geological environment, separated by soil and unsaturated zone
from pollution impact, and thus not vulnerable to human activities survives a very long time
and has serious consequences on groundwater quality in many countries.
Groundwater vulnerability to pollution as a side effect of human impacts was recogn-
ized as a significant world wide environmental problem during the nineteenfifties. In the nine-
teensixties the conceptual approach to the groundwater resources protection based on moni-
toring, mapping, modelling and remedial technology was developed. In the nineteenseventies
groundwater protection policy and management were formulated, and appropriate pricing
policy based on the polluter pays principle was defined as part of water legislation. Sustainable
groundwater use and development and environmentally sound management of groundwater
protection became important in national water planning, policy and strategy in the course of the
ninteeneighties. A holistic concept in water resources policy and management emphasized on
the International Conference of Water and the Environment in Dublin (1992), significantly
influenced the approach to the development, use and protection of water resources in the nine-
teennineties. The implementation of a holistic approach is particularly important for the
management of transboundary aquifers. Their development and protection policy is a serious
international hydro-political problem; the solution of which will proceed in the twenty-first
century.
13B.2 Groundwater quality protection strategy
Groundwater protection strategies should be based on environmentally sound management

and on the concept that prevention of pollution is always less expensive than groundwater
rehabilitation, a costly, long-term and technically demanding task.
Groundwater protection should not be an emergency action only taken up when pollution
problems arise. The concept of a groundwater protection strategy is translated into a long-term
multidimensional programme that accommodates and integrates the environmental, technical,
economic and social aspects involved. This programme is implemented, coordinated, and
managed by governmental authorities. It is supported by relevant legislation, requires regular
supervision and inspection, includes the education and training of responsible experts and tech-
nical staff, and the education and information of the general public (Vrba, 1994). The principal
technical and capacity building criteria for the successful implementation of an aquifer pro-
tection strategy are:
the availability of comprehensive analyses of the aquifer systems based on good
knowledge of the groundwater systems unsaturated and saturated properties (the
geological, physical and chemical parameters, the flow net, aquifer boundary con-
ditions and the level of vulnerability);
the effective operation of groundwater quality monitoring systems on national, pro-
vincial and local levels and the analysis of the data collected;

347
Groundwater studies
the identification and listing of the historical, present and potential pollution sources
and the evaluation of their nature, extent and real or potential impact on groundwater
quality;
a research programme supporting the development , improvement and innovation of
methods and techniques for groundwater protection, quality conservation and
pollution remediation;
the existence of institutional structures that have the power and human and financial
resources for the creation, co-ordination, and implementation of a comprehensive
groundwater protection strategy;
a legislative basis, regulatory statutes, codes and standards for the implementation of a
groundwater protection and quality conservation programme;
an effective inspection and control system, based on relevant legislation, the polluter
pays principle and on implementation of repressive or stimulating financial instru-
ments;
availability of qualified, trained, experienced and motivated professional personnel to
implement groundwater protection programmes;
a public information education and involvement groundwater protection programme,
particularly regarding the consumer pays principle.
Groundwater protection strategy is effective if the above criteria are applied in an coherent
manner. Their partial or inconsistent application will not lead to a long term successful pro-
tection of the groundwater resources.
13B.3 Groundwater quality protection policy
Not all groundwater resources need to be protected comprehensively, because they possess
different environmental and socio-economic values and not all of them are of the same vul-
nerability. To protect all groundwater resources to the same extent would be economically
unsustainable, hydrogeologically pointless and unrealistic in terms of management and control.
Comprehensive protection throughout the national territory of a resource as diffuse and vast as
groundwater certainly is, would be too costly, would provoke conflicts in the exploitation of
other natural resources, and could lead to an undesirable restriction of economic development.
Groundwater protection policy should be constructive and not based on bans and restrictive
measures only.
Groundwater protection policy should be:
based on relevant legislation,
integrated with the protection of remaining components of the hydrological cycle and
other environmental elements,
co-ordinated with land use activities and industrial development,
based on the groundwater resources, their vulnerability and water supply needs,
linked with the social policy,
reflective to the cultural and historical traditions of the society.
The criteria supporting establishment of groundwater policy depend on:
the value (quantity and quality) of the groundwater and its vulnerability;
the current and expected demands for groundwater use and related protection in a
given area;
the importance of groundwater for the ecosystem (e.g. wetlands).
The classification of groundwater using the above criteria is always a complicated task, it
can be controversial and requires good knowledge of the hydrogeological, economic and social

348
aspects of the area concerned. An example of protection criteria is from the United States of
America (Table 13B.1) in which three classes of groundwater are defined.
Table 13B.1 Classes of groundwater in the United States of America (EPA, 1984)
Level of groundwater
Class Basic criteria protection Remarks
I. Special Highly vulnerable, Extremely high Significant water

groundwater irreplaceable or resources value
ecologically vital
II. Current and All other groundwater High to moderate Majority of usable
potential drinking use or available for prevention of sources of ground-water
water and waters drinking or other contamination based in the United States
having other purposes on technological of America
beneficial use remedies rather than
through restrictions
III. Groundwater not Heavily saline or Usually low migration Limited beneficial use
considered potential heavily polluted to class I or II ground-
sources of drinking water or discharge to
water and of limited surface water must be
beneficial use precluded
A three-level classification system for groundwater quality has also been applied in the
Czech Republic:
Class I Groundwater suitable as a source of drinking water without any treatment (only
disinfectants and mechanical acidification when necessary);
Class II Groundwater which requires treatment that is justifiable financially as a source of
drinking water (chemical acidification, removal of iron, manganese etc.);
Class III Groundwater unsuitable as a source of drinking water because of its natural
properties, or a high level of contamination due to human impacts. Improving its
quality is technically demanding and costly.
The policy of protecting groundwater as a drinking water source reflects the economic,
social and cultural level of a society and its environmental philosophy and thinking. Gone are
the days when groundwater protection policy focused only on the technical, economic and
financial aspects while ignoring environmental and social attributes. Therefore, the objective of a
modern policy on sustainable groundwater protection must be:
timely identification and analysis of potential conflicting issues and constraints;
formulation, quantification, analysis and validation of competitive factors with respect
to land use and groundwater values;
hierarchical screening with the aim of finding a balance between groundwater quality
protection and conservation, economic development and social and health implications
with a view to both the short term and long term prospects.

349
Groundwater studies
13B.4 Groundwater protection management
The objective of groundwater management is to ensure the quality, safety and sustainability of
groundwater used, in the first and foremost place as a drinking water source, but also as part of
the natural heritage and as a source of irrigation and industrial water. Groundwater protection
management is not a universal or anonymous process; it is carried out and applied to particular
aquifer systems having their specific natural properties, influenced by various human impacts,
and therefore it varies from place to place.
Groundwater protection management must take into account: a degree of ignorance,
uncertainty and unpredictability concerning the groundwater systems behaviour and pro-
perties, and the cumulative human impacts that are frequently difficult to identify. The effects
of the latter cannot be predicted accurately over a longer term. The uncertainties in defining
internal and external influences on groundwater systems are associated with certain risks.
Therefore risk assessment and risk analysis are part of groundwater protection management.
Sustainable groundwater protection management is based on a holistic approach, reflect-
ing the social and economic value of groundwater of groundwater and integrity of aquatic and
terrestrial ecosystems, considering the close connection between groundwater, surface water and
other components of the hydrological cycle in relation to land use. At the same time, attention
should be given to both quantitative and qualitative aspects in view of sustainable development
and environmentally sound protection of groundwater resources. Holistic groundwater pro-
tection management should be based on participatory approach, involving planners, policy and
decision makers, managers, water users and the general public.
Two categories of groundwater protection management can be considered: general
protection of groundwater resources and comprehensive protection around public groundwater
supplies.
13B.4.1 General protection of groundwater
General groundwater protection is based on the assumption that all effectively accessible
groundwater resources are, or could be, tapped for drinking or other purposes now or in the
future, and therefore their protection is desirable. The water authorities should bear the respon-
sibility for the protection of all usable groundwater resources.
Implementation of general protection of groundwater resources calls for the following
activities:
investigation of the groundwater system and determination of its vulnerability;
identification, and listing the existing and potential pollution sources;
operation of groundwater quality monitoring systems;
control over human activities, in particular in vulnerable areas of groundwater basins
and aquifers;
stipulation and implementation of legislative measures for the protection of ground-
water resources, and the control over the pollution sources.
In several European countries, recharge areas of important groundwater basins and aquifers
have been delineated and protected within the general protection of groundwater resources.
These areas cover several hundred square kilometres; human activities within these areas are
under control and partially restricted by law.
13B.4.2 Comprehensive groundwater protection
The main purpose of delineating groundwater protection zones is to protect drinking water
supply wells or well fields from pollution and provide the population with water that meets the

350
standards for drinking water. In many countries, groundwater resources used for public
drinking water supplies are comprehensively protected by protection zones. In the United States
of America. they are referred to as wellhead protection areas, usually comprising two or three
levels of protection. Wellhead protection is an obligatory part of groundwater protection mana-
gement programmes, and is based on the relevant legislation. Protection of public water supply
wells requires:
delineation of protection areas for each abstraction well or spring as protection zone I
and designation of protection zones II and III, which cover vulnerable and recharge
areas of water supply systems, that are sensitive to pollution or other human impact;
designation and operation of groundwater monitoring systems;
an inventory of all potential sources of groundwater pollution;
cooperation between local and provincial authorities, water work companies, land
users, (particularly farmers) and water users (public);
establishment of a specific plan of management of groundwater and land use activities
in protection zones, with the scope to maintenance a good quality of drinking water;
evaluation of economic losses due to the limitation of human activities (particularly
farming) in groundwater protection zones and determination of the amount and the
responsibility to compensate;
implementation of institutional, legislative, technical and control measures and
regulations in protection zones;
elaboration of an emergency plan for providing alternative sources of drinking water
in case pollution of water supply systems occurs.
First level protection zones protect the well and its immediate surroundings from physical
damage and direct pollution. Their extent is usually small=several tens of square meters is the
maximum and human activities are excluded.
The second and third level groundwater protection zones are mostly extensive (several
hundreds of square metres to several square kilometres) and include the discharge areas, the
cones of depression (zone of influence around pumping wells, the recharge areas and other vul-
nerable parts of the water supply system of interest. In several European countries second level
zones cover areas having a delay or residence time of 50 to 60 days. It should be emphasized
that the delay time has been determined so as to protect water supply wells from the risk of
microbial contamination of groundwater, but it may become inadequate for viruses and certain
persistent chemical pollutants. The third level zone protects groundwater quality in water
supply wells from persistent chemical pollutants. It is the most extensive zone but there are less
restrictive measures. Van Waegeningh (1985) recommends a 10 to 25 year residence time for the
delineation of groundwater protection zones.
13B.4.3 Delineation of groundwater protection zones
In general, the level of restrictions and prohibitions in groundwater protection zones decreases
with the distance from a well or wellfield. The second and third level protection zones cover
significant areas, frequently including arable land. Over-protection of water supply wells is
therefore not desirable because restrictive measures or exclusion of land from farming lead to
economic losses. On the other hand, under-protection of wells and well fields may cause
groundwater pollution requiring long-term and costly remedial action.
Although the environmental-social-economic system and the physical (soil and water)
system, are independent of each other, their mutual relations must be coordinated and inte-
grated with the objective of deriving benefits from utilisation of soil and water resources while
conserving the quality of the environment. Management of the soil-aquifer system requires a
specific approach in each region because the natural conditions and intensity of human activ-

351
Groundwater studies
ities, including farming, differ locally and regionally. Comprehensive protection and mana-
gement of groundwater quality achieved through stringent control over agricultural activities
will unfavourably affect farmers production targets.
Objective integration of the soil/water users interests and allocation of the benefits and
costs between the water and agricultural sectors is the key factor in the management strategy of
effective utilisation of soil and water resources. Under all circumstances the distribution of costs
and benefits is a very sensitive and critical point, and should be dealt with in a timely and
responsible manner.
It is widely held that protection zones should be as small as possible but as large as
necessary. Expressed in financial terms, higher input costs on accurate delineation of protection
zones will result in reduced operational costs for well and wellfield protection. Since water
supply systems operate for a long time, the operating costs involved in their protection should
be as low as possible. A sophisticated approach to protection zone definition is therefore
preferable. There is no question about groundwater protection zones having a positive influence
on the quality of groundwater (Table 13B.2).
Table 13B.2 NO 3 content in groundwater, in mg /l; regional monitoring system in central Bohemia
(Czech Republic)
NO3 content
1968 1986 Annual increase
Type of monitoring station mg/l
Wells and boreholes without

protection zones 24.0 43.0 1.1
Public wells and well fields with

protection zones 21.0 28.0 0.4
13B.5 Groundwater pollution control
13B.5.1 Influence of natural processes and human impacts
Groundwater contains a broad range of dissolved solids in varying amounts (major, minor and
trace constituents). The natural composition of groundwater is controlled by the rock medium in
which groundwater moves; contact time of water with rock; a complex of geochemical (dis-
solution, precipitation, hydrolysis, adsorption, ion exchange, oxidation, reduction), physical
(dispersion, advection, filtration) and microbiological (microbial metabolism and decomposition,
cell synthesis) processes; the presence of gases (especially carbon dioxide); soil properties;
climate and the impacts of other components of the hydrological cycle (e.g. sea water intrusion).
The potential solution and precipitation of minerals and rocks depend on the mechanism of
reactions between water and rock.
The properties of natural groundwater may not always meet the criteria for drinking
water, nor be suitable for farming or industrial needs, because groundwater quality can
deteriorate by natural processes and human impacts. Influence of natural processes on ground-
water quality is long term and difficult to control. Human impact can significantly influence and
accelerate natural processes in groundwater/rock systems. The most important causes of
groundwater quality deterioration are different kinds of pollution, and over-exploitation of

352
aquifers that change the groundwater flow dynamics (Chapter 13A). Groundwater pollution is
understood to be a process whereby due to human impact water suddenly or gradually changes
its physical, chemical or biological composition and ceases to meet the criteria set for drinking
water. If it contains hazardous or toxic compounds water becomes dangerous for people and
other living organisms and ecosystems. The vulnerability of a groundwater system to changes in
its quality depends on hydrogeological conditions, on the above mentioned chemical, physical
and biological processes in soil, rock, and groundwater, and the type, extent and intensity of
pollution.
Pollution of groundwater is a hidden and long term process and, due to the mostly slow
movement of the pollutant and the groundwater, a period of several weeks, months or some-
times even years will elapse between contaminant leakage and groundwater pollution. The
groundwater quality is most frequently affected by:
heavy metals present in liquid and gaseous wastes (leakage of industrial waste water,
mining water or processing water, waste water from uncontrolled landfills, etc.),
organic chemicals (particularly uncontrolled leakage of hydrocarbons, halogenated
hydrocarbons and chlorinated hydrocarbons),
inorganic and organic fertilizers as a main source of nitrate pollution of groundwater
under arable land,
different kinds of pesticides (herbicides, insecticides and nematicides) used in agri-
culture and forestry, and
pathogenic bacteria and viruses (leakage from septic tanks, sewerage systems, animal
sludge applied to farm land, etc.).
Various criteria are used for classification of groundwater pollution. The most commonly used
classification is based on the extent of pollution point (local), non-point (diffuse), line
regional or continental; kind of pollutants physical, chemical, biological, radioactive; and
source of pollution urban, industrial, agricultural, traffic, mining, land waste disposal. An
overview of groundwater pollution sources is given in Table 13B.3.
Among the most frequent point pollution sources with impact on groundwater quality are
industrial sites, mining areas and waste disposal sites. According to the EEA (1995) the potential
pollution of groundwater by point sources does not cover more than 1% of the European
territory. However, point pollution sources mostly occur close to urban areas or rural settlements
and have serious impact on the quality of public or domestic groundwater supplies. Oil
products, heavy metals and various organic compounds are the most prominent pollutants of
the groundwater system.
Urban and rural areas are sources of multi-point and heterogeneous pollution of
groundwater. Insufficient handling, treatment and management of household wastes and waste
waters, industrial effluents, uncontrolled waste disposal sites, rain and melt waters and salt
water intrusion in coastal areas are the main sources of multi-point pollution of municipal
groundwater (Jackson et al., 1980; Mathess 1982; Vrba, 1985; RIVM 1992 and others). In rural
areas the most frequent source of groundwater pollution are unsewered sanitation systems,
which may cause degradation of groundwater quality in domestic and public water supply
wells. Pathogenic micro-organisms, chlorides, nitrates, household detergents and disinfectants
are the main pollutants of groundwater in rural settlements.
The only widely occurring groundwater pollutants reported in respect to non-point or
diffuse pollution are nitrates originating from organic and inorganic fertilizers applied to arable
land. In Europe and United States of America the groundwater quality is much more affected by
diffuse nitrate pollution, than in other continents. In several areas with intensive farming
activities the nitrate levels in shallow aquifers are above 50 mg/l. According to Stannes and
Bourdeau (1995) the nitrate content reaches EU target value (25 mg/l) on 87%, and drinking
water standard (50 mg/l) on 22% of the shallow aquifers under agricultural soil in Europe. In

353
Groundwater studies
Table 13B.3 Sources of groundwater pollution
Extent of pollution Source of pollution Main pollutants
Point Industry Heavy metals (Pb, Zn, Cd, Cr), arsenic, phenols,
petroleum products and additives, high BOD, suspended
solids, PAHs, synthetic organic and organometalic
compounds
Mining Heavy metals, salts (chloride, sulphate), low pH, high
TDS, cyanide, PAHs, petroleum products
Waste disposal sites, Heavy metals, ammonium, sulphate, chloride, phenols,
including deep various biodegradable and non-biodegradable organics,
disposal wells faecal pathogens
Radioactive waste 3H - Tritium, 90Sr, 137Cs, 239Pu, 129I, 226Ra, toxic metal
Animal husbandry High suspended solids, BOD, total nitrogen, chloride,
faecal pathogens
Multipoint Urban areas Heavy metals (Pb, Zn), ammonia, chloride, sulphate,
petroleum products, chlorinated hydrocarbons,
surfactants
Rural settlements Ammonia, nitrate, chloride, sulphate, surfactants, iron,
manganese, faecal pathogens
Military areas Petroleum products, heavy metals
Non point (diffuse) Agriculture crop Fertilizers (organic and inorganic): nitrate, ammonia,
and root-crops farming, chloride, phosphate, sodium, potassium, faecal
cattle breading, pathogens, salinity
irrigation Pesticides: organochlorine compounds (aldrine,
heptachlore, carbamate insecticides (atrazine),
polyphosphate, organometallic compounds (fungicides)
Line Roads High suspended solids, salts, petroleum products,

solvents
Railways Petroleum products, organic chemicals
Oil pipelines Petroleum products
Sewerage systems High suspended solids, nutrients, chloride, high BOD,
faecal pathogens
Streams Nitrate, ammonia, iron, manganese, phenols
Regional (continental) Acid deposition Aluminium, low pH, nitrates, sulphates
Coastal areas Salinization Sodium, magnesium, chloride, sulphate, high salinity

and TDS
United States of America, particularly in the mid-continental United States Corn Belt, where
nearly 60% of nitrogen fertilizers of the whole United States of America are applied, high aquifer
pollution (150 mg/l NO3-N) can be found in many regions (Spalding and Exner, 1991). A NO3-N
content from 4060 mg/l is also reported in groundwater in several irrigated valleys in
California and other irrigated areas of the United States (Keeney, 1986). However, high contents
of nitrates in shallow wells are mostly a consequence of poor construction and siting of wells
located closely to animal corrals and cattle feeding areas.

354
Acid atmospheric emissions (sulphur dioxide SO2, nitrogen oxides NOx) are trans-
ported hundreds of kilometres across the continents and their chemically converted products
(sulphuric and nitric acids) are potential sources of regional transboundary acidification of the
soil and water bodies. The influence on the groundwater quality (lowering of pH values,
increasing content of aluminium sulphates and heavy metals) has been recognized in some
industrial regions in Europe (Holmberg, 1987; Stanners and Bourdeau, 1995).
Natural processes and man-made factors involving groundwater composition and quality,
are described in details by Matthess (1982), RIVM (1992) and others. Section 13B.6 comprises
references to selected case studies, as examples of different kinds of pollution and remediation
(Al-Agha, 1995; Terao et al., 1993; Ramon, 1993; Vellner, 1993; Hyde et al., 1993; Christe et al.,
1995; Fried, 1991; Vrba, 1991; Frchtenicht et al., 1995).
13B.5.2 Point pollution control of groundwater
No generally valid methods for identifying pollutants below the surface can be established,
because each pollution episode is highly individual in nature, and the hydrogeological envi-
ronment differs from case to case. Point pollution control based on mapping, monitoring and
modelling (simulation and forecasting of pollutant transport) includes the following steps:
identification of the pollution source and extent (plume and front); use of remote
sensing methods is very effective;
design and establishment of a site-specific monitoring system and programme;
determination of the composition, properties, amount and duration of the pollutant
spill;
assessment of hydrogeological conditions and delineation of the risk area.
Point pollution is usually a time-limited problem, due to the possibility of identifying and
insulating the pollutant source. Once the source is insulated, and pollution plume defined
remediation of pollution can be launched. Decisions for the selection and implementation of
clean-up techniques and rehabilitation processes are based on the properties and mobility of the
pollutant, the age and extent of the pollution (unsaturated zone only, part or all of the aquifer),
and the vulnerability and attenuation capacity of the affected groundwater system.
The most frequent source of point pollution of groundwater, i.e. oil hydrocarbons, can be
tackled with various physical, chemical and biological methods. When only the unsaturated
zone is contaminated, effective methods are: excavation of polluted soil, or injection through
boreholes or galleries of water enriched with oxygen, hydrogen peroxide, ozone or with
nutrients that stimulate the growth of autochthonous micro-organisms. In the case of soluble oil
compounds (petrol, aviation kerosene) washing the contaminated rock medium with hot water,
ventilation by injected air (to remove easily volatile hydrocarbons) or vacuum pumping (to
remove light oil hydrocarbons in the liquid phase) are clean-up methods that can be employed
successfully.
The remediation methods used for cleaning up a saturated aquifer depend on the
intensity of pollution and the solubility of the oil hydrocarbons. When the groundwater table is
covered by a layer of oil compounds, scavenge pumping is applied. When the content of
dissolved and emulsified hydrocarbons is low, the contaminant is collected from the ground-
water surface using a strip-type or mop-type mechanical skimmer. Some other methods that can
be used with success for remediation of oil polluted aquifers are: vacuum pumping (in the case
of oils with a low boiling point), stripping of light fractions of oil hydrocarbons, biodegradation
of oil hydrocarbons in an aerobic medium, anaerobic degradation of aromatic hydrocarbons. It
should be emphasised that the clean up of water systems polluted by oil hydrocarbons is always
a long, and technically and financially demanding process, with uncertain results.

355
Groundwater studies
13B.5.3 Non-point pollution of groundwater
Non-point pollution, also referred to as diffuse pollution, of groundwater is usually related to

agricultural activities, particularly the massive application of fertilisers and protective chemicals,
or irrigation return flow on farmland. At present the only widely occurring groundwater
pollutants, reported in respect to non-point agricultural contamination, are:
nitrates originating from organic and inorganic fertiliser applied to arable land;
potassium and phosphorous compounds derived from fertilizers and pesticides that
accumulate in soil and the unsaturated zone. Because of their lower solubility in water
and lower mobility, they constitute a potential threat to groundwater;
atrazine, which has been identified in many regions of the world, especially in areas
with fruit and vegetable plantations on sandy soils over shallow vulnerable aquifers.
Current fertiliser management practices result in the input of nitrogen significantly
exceeding nitrogen uptake. Data from several countries show that only about 50 percent of the
nitrogen applied as fertiliser is removed with the produced grain.
Incorrect fertilising regimes also lead to disturbances in the dynamic stability of the soil
organic matter, and accelerate mineralization and nitrification processes. Eventually they
percolate into the groundwater system as inorganic nitrogen compounds and organic carbon-
nitrogen compounds mainly fluvic acids and saccharides.
Nitrate contents in groundwater bodies are not stable. Short-term cyclic changes in nitrate
concentrations depend mainly on natural conditions (particularly precipitation). The long-term
upward trend in nitrate contents in groundwater reflects human, largely farming influences
(Fig. 13B.1). The vertical movement of the nitrate front to deeper parts of an aquifer is also
frequently observed in areas under intensively farmed arable land (Fig. 13B.2).
Pesticide management is one of the most important tasks of agriculture in the envi-
ronmental sphere. The effects of pesticides on the groundwater system are governed mainly by:
the kind, properties and amount of pesticides and their potential to leach through the
soil to the groundwater,
the meteorological conditions at the time of pesticide application and the application
techniques used,
the type of vegetation and its ability to take up water through the root zone,
the soils physical and chemical properties, its structure and texture, amounts of
organic matter and clay material, soil water potential and hydraulic conductivity,
the vulnerability of the groundwater system, thickness and permeability of the
unsaturated zone and type of aquifer (phreatic, semi-confined and confined), and its
properties.
Of special importance is the choice of the pesticide type. Physical and chemical charac-
teristics, toxicity, solubility, persistence, mobility, adsorption, breakdown products, dynamics of
residuals and adverse environmental effects should be evaluated when selecting the type of
pesticide to be applied.
13B.5.4 Non-point pollution control
Currently, non-point source contamination control systems, and control measures are concerned
chiefly with the pollution of groundwater by nitrogen compounds (Jackson et al., 1980;
Zwirnmann, 1981; Landreau, 1982; Vrba, 1985).
Because of the great areal extent of diffuse nitrate pollution, the application of isolate
source policy and subsurface clean-up techniques are ineffective. The protective measures
conducive to improvement of the groundwater quality consist of the management and control of
nitrogen input to the plant/soil system. These can be achieved by restricting or prohibiting

356
Figure 13B.1 Impact of nitrate pollution on groundwater in the Czech Republic
farming activities in recharge and vulnerable areas of the groundwater system, or by changing
agricultural practices. The latter may be related to the selection of suitable fertilisers and
determination of the doses, times and techniques of their application; selection of suitable plants
(crops); and replacing monocultures by a crop rotation system. Other agricultural activities may
have to be modified through regulation of animal production and extension of farmland, or
expansion of grassland at the expense of arable land. The control measures depend above all on
the steps taken in the agricultural sector. In the sphere of water management the control
measures can be focused on symptomatic actions only, not on eliminating the cause of pollution.
A modern approach to pest control lies in the preventive sphere, in the integrated pro-
tection of crops against diseases, weeds and pests. The preventive step in integrated protection
entails the policy of licensing new pesticides, permanent pest control, soil preparation, selection
of plants and sowing procedures, choice of highly resistant kinds of cereals and other crops, and
optimal nitrogen and other fertiliser doses. If a disease, pest, or weed has already been
identified, the time and dose of pesticide and the technique of its application should be defined
with regard to the type of crop. Before the pesticide is applied, the data recorded on weeds and
diseases in past years and the qualified forecasts of the potential occurrence and intensity of the
disease, pest, or weed in question should be evaluated. When selecting a pesticide, priority

357
Groundwater studies
Figure 13B.2 Changes in hydrogeological profiles in a shallow aquifer in 19841989; monitoring borehole HP-65
(Central Bohemia, Czech Republic)
should be given to those types that do not produce a need for additional chemical treatment; the
synergetic effects of pesticide mixtures should be utilised. Biological and biochemical substances
or rapidly degradable types of pesticide (predators, hormone stimulators, pheromones, etc.)
should be applied so as to reduce adverse affects on the groundwater system.
13B.5.5 Impact of groundwater on human health
The physical, chemical and biological composition of groundwater may have positive as well as
negative effects on human health. Entry of pollutants into a groundwater system through
human impacts may be the cause of alimentary diseases, which can grow to epidemic pro-
portions, with potentially fatal consequences. In this respect, groundwater quality is of parti-
cular importance, as groundwater accounts for nearly 100 percent of drinking water supply for
the population of less developed countries. In advanced industrial countries, as, for instance, in
Europe, groundwater usually provides more than 50 percent drinking water. Due to its mobility
and ability to transport, transform and absorb pollutants, groundwater is becoming one of the
most potentially dangerous contaminating media. It has been reported that in less developed
countries, polluted water may cause 80 per cent of diseases.
By comparison with surface water, groundwaters self-purification potential is markedly
lower and lessens with the aquifers depth depending on the declining amount of dissolved
oxygen. In spite of that, contact between groundwater and sediments when these have great ion-
exchange potential, may cause distinct changes in space and time in their physical and chemical
composition, biodegradation and microbial activity.
The adverse effects of pollutants on human health via the food chain, groundwater

358
included, impinge on social and economic spheres (sickness and death rates, migration of
population, lower working output, impact on peoples mental state, etc.). The health risks posed
by different kinds of pollutants in groundwater should therefore be the subject of continuous
control and evaluation, because they may assume enormous significance for present and future
generations.
Elevated nitrate levels in groundwater are particularly monitored in areas of intensive
agriculture. High doses of fertilisers may result in the accumulation of unmetabolised nitrates
in plants such as spinach, carrot, kohlrabi and beet which are especially liable. There is a well-
known and close relationship between high nitrate contents in groundwater consumed for
drinking and alimentary methaemoglobinaemia, which can be fatal for infants under six months
of age (through bacterial action nitrates are reduced to nitrites which cause the haemoglobin in
blood to change into methaemoglobin, which is unable to transport oxygen). The adverse health
effects (gastric cancer, birth defects, cardiovascular disease, effects on the thyroid gland) as a
consequence of long-term consumption of water with high nitrate contents are currently being
studied. Health risks can also arise when high nitrate contents are combined with pesticides, or
when their residues form carcinogenic nitrosamines. Exposure to nitrate in drinking water
cannot be implicated or excluded as a causative factor for certain types of cancer (Weisenburger,
1991).
In view of their effects on human health, pesticides should be monitored to ascertain their
persistence in soil, ecological-toxicological and neurotoxicological effects, penetration into the
food chain, and biotransformation (their intermediate products may be more dangerous for man
than the original compounds). At first, pests were combated using salts of arsenic, mercury,
copper and zinc, as well as sulphur and barium chloride; later dinitro-compounds, triocianates
and, most importantly, chlorinated hydrocarbons, especially DDT, HCH and its isomers lindane,
aldrin, dieldrin and heptachlor were introduced. Even low concentrations of the above sub-
stances are detrimental to the environment, and chronically toxic. The use of most chlorinated
hydrocarbons in farming has been banned because they can cause acute or chronic conditions
and pathological processes, and adversely affect the health of future generations. However, their
residues in groundwater persist, and have been identified locally and regionally in concen-
trations of 0.001 mg/l and higher, with polychlorinated biphenyl (PCB) occurring in concen-
trations up to milligrams.
A broad spectrum of organic compounds is observed in most groundwaters. So far, they
have only locally shown up in harmful concentrations in groundwater, in shallow vulnerable
aquifers beneath intensively farmed arable land or as a consequence of uncontrolled leakage.
Organic compounds of phosphorus pose a risk through their acute toxicity; organic fungicides
are a potential source of poisoning and, due to their persistence in soil, cause long-term
degradation of the biosphere and hydrosphere; carbamates are toxic substances that are poten-
tially carcinogenic and possess mutagenic and teratogenic properties. Mercurial fungicides are
also highly toxic. Triazine, carbamate and similar herbicides pollute groundwater, even at trace
concentrations.
The metal contents in groundwater resulting from human activities should be monitored
because of their toxic effects on the human organism (its liability to genetic damage), and the
ability of these compounds to accumulate in the human body due to their long biological half-
life. Cadmium enters the food chain through crops or water. The uptake of cadmium by roots
depends mainly on the soil pH. The presence of cadmium in groundwater has been monitored
in particular in certain intensively farmed regions. Enhanced levels of lead in groundwater have
been identified in areas affected by exhaust gases of dense road traffic and in areas where
metallic ores are mined. An increased intake of lead results in anaemia, fatigue and irritability.
Reports also mention leads effect on the growth of long bones during the development of
young animals.

359
Groundwater studies
Fluorides are frequently present in groundwater and their high content is associated with
calcium deficiency. Long-term consumption of groundwater containing more than 2 mg/l of
fluorides causes dental fluorosis particularly in children. In concentrations of less than 1 mg/l,
fluoride in drinking water reduces solubility of tooth enamel, supports the growth of teeth, and
has a beneficial effect on human health.
Uncontrolled spills of liquid human waste from septic tanks and sewerage systems,
animal sludges, and fertilising with farm manure (above all pig manure) over shallow, vulner-
able aquifers can be a source of bacterial, viral and parasitic contamination of groundwater.
Health risks are considerable in such areas, particularly from domestic and shallow wells in
rural areas. Diseases can grow into epidemics with potential fatal consequences. Intake of water
contaminated in this way may lead to alimentary diseases, above all typhoid, paratyphoid and
other salmonelloses, enterovirosis, yersiniosis, etc. The World Health Organization estimated in
1999 that water related diseases caused 3.4 million deaths in 1998, more than half of them
children. Other estimates (Van der Hoek et al., 1999) are even higher especially for diarrhoea
(5 million deaths).
13B.6 References and additional reading
AL-AGHA, M. R., 1995. Environmental Contamination of Groundwater in the Gaza Strip.

Environmental Geology, Vol. 25, pp. 10913.
CANDELA, L.; AURELI, A., 1998. Agricultural Threats to Groundwater Quality. UNESCO, CIHEAM,
Zaragoza, 251 pp.
CHRISTE, R.; FLURY, F.; HESSENAUER, M.; JEANMIN, P. J.; TURBERG, J., 1995. Groundwater Impact
Assessment for the N16 Motorway in Northwestern Switzerland. In: Groundwater Quality;
Remediation and Protection, IAHS Publication No. 225, pp. 45970.
EPA, 1985. Groundwater Monitoring Strategy. United States Env. Prot. Agency, Washington, 33 pp.
FRIED, J. J., 1991. Nitrates and Their Control in the EEC Aquatic Environment. NATO ASI series G:
Ecological Sciences, Vol. 30, pp. 312..
FRCHTENICHT, H.; HELD, T., 1995. Microbiological in Situ Remediation of an Aquifer Contaminated
with Waste Oil from a Refinery. IAHS Publication No. 225, pp. 41120.
HOLMBERG, M., 1987. Assessing Aquifer Sensitivity to Acid Deposition. In: Vulnerability of Soil and
Groundwater to Pollutants, Proceedings and Information No. 38, TNO, RIVM, The Hague,
pp. 37380.
HYDE, G.; PEDERSEN, K. E.; ANDERSEN, E., 1993. Aquifer-Wide Groundwater Restoration and
Protection; from Concept to Practice, an Example from Denmark. In: Groundwater Quality
Management, IAHS Publication No. 220, pp. 4718.
JACKSON, R. E., 1980. Aquifer Contamination and Protection. UNESCO, Paris.
KEENEY, O. R., 1986. Sources of Nitrate in Groundwater. Critical Review in Environmental Control,
Vol. 16, pp. 257335.
LANDREAU, A., 1982. Fertilisants azots et qualit des eaux souterraines. In: Cartographie des zones
vulnerables. B.R.G.M., Orlans.
MATTHESS, J., 1982. The Properties of Groundwater. Wiley-Interscience Publications, New York,
406 pp.
RAMON, S., 1993. Experiments on Agricultural Restrictions Carried Out in Protection Areas by
the Agence de lEau Rhin- Meuse. In: Groundwater Quality Management, IAHS Publication
No. 220, pp. 4338.
RIVM, 1992. The Environment in Europe: a Global Perspective. Report No. 481505001, RIVM, 119 pp.
SPALDING, R. F.; EXNER, M. E., 1991. Nitrate Contamination in the Contiguous United States. In:
Nitrate Contamination, NATO, ASI Series G, Vol. 30, pp. 1348.

360
STANNERS, D.; BOURDEAU, P., (eds.), 1995. Europes Environment. European Environmental Agency,
Copenhagen, 676 pp.
TERAO, H.; YOSHIOKOJ, Y.; KATO, K., 1993. Groundwater Pollution by Nitrate Originating from
Fertiliser in Kakamigahara. Selected Papers on Environmental Hydrogeology, IAH/Heise,
Hannover, Vol. 4, pp. 5162.
VAN DER HOEK; KONRADSEN, F.; JEHANGIR, W. A., 1999. Domestic Use of Irrigation Water; Health
Hazard or Opportunity? Water Resources Dev., Vol. 15, pp. 10719.
VAN WAEGENINGH, H. G., 1985. Protection of the Groundwater Quality in Porous Permeable
Rocks. In: Theoretical Background, Hydrogeology and Practice of Groundwater Protection Zones.
Intern. Contrib. Hydrogeol., Ed. Heise, Vol. 6, p 11121.
VELLNER, L., 1993. Problems of Groundwater Quality Management in Estonia. Groundwater Quality
Management. IAHS Publication No. 220, pp. 44960.
VRBA, J., 1985. Impact of Domestic and Industrial Wastes and Agricultural Activities on
Groundwater Quality. Memoirs XVIII IAH Congress, Cambridge, Part 1, pp. 91117.
, 1994. Groundwater Protection Strategy, Policy and Management. Proc. Congress Spain Group
of IAH, Madrid, Part II, pp. 36383.
; PEKN, V., 1991. Groundwater Quality Monitoring; An Effective Method of Hydrogeological
System Pollution Prevention. Environmental Geology, Vol. 17, pp. 913.
WEISENBURGER, D. D., 1991. Panel Consensus on Potential Health Consequences of Elevated
Nitrate Levels. In: Nitrate Contamination; Exposure, Consequence and Control. NATO ASI
Series G: Ecological Sciences, Springler-Verlag, Vol. 30, 511 pp.
WORLD HEALTH ORGANIZATION, 1999. World Health Report. Geneva.
ZWIRNMANN, K. H., 1981. Analysis and Control of Non-Point Nitrate Pollution of Municipal Water
Supply Sources. IIASA, Luxenburg, Austria.
361
14 H y d ro g e o l o g y
o f c a r b o n a t e ro c k s
14.1 Introduction
Carbonate rocks as such are usually poorly permeable and have a low specific yield. Only when
they are intensely fractured and particularly when the fractures are widened by solution (devel-
opment of karst phenomena) they do become water bearing and water transmitting, often on a
spectacular scale.
People have always been intrigued by carbonate rocks with karst phenomena, particularly
caves (Fig. 14.1). In prehistoric times, early man chose the caves in limestone as living areas, and
his explorations led him deep underground to sources of water. Even 2000 years ago, Greek and
Roman philosophers speculated on the origin and nature of caves, underground rivers and
springs The earliest hydrological concepts of the hydrological cycle, the source of water, its
occurrence, and quality were related to the hydrology of karstic limestone.
How important are carbonate rocks to man? Approximately 12 percent of the earths
continental surface is underlain by carbonate rocks (Fig. 14.1) that have produced diverse topo-
graphical landforms by weathering under varied climatic conditions (Milanovich, 1981). The
surface topography of some areas underlain by carbonate rocks is one of broad, rolling plains,
but elsewhere it is characterised by steep-sided bluffs, canyons, sinks, and valleys. In some
areas, carbonate rocks are overlain by fertile soils; in others areas they are bare.
The impact of carbonate rocks is great on man and of substantial interest financially, e.g.:
(i) At the International karst meeting in Guilin, China, in October 1988 (Yuan Daoxian,
1988), the Director of the U.S. Geological Survey stated that karst aquifers are a major
source of drinking water in the United States of America, providing 25 million cubic
metres of water per day in 1985 (Peck et al., 1988).
(ii) In the midwest of the United States of America, a large area underlain by limestone
is covered by a very productive, rich soil that is farmed to produce large quantities of
food. This area is literally the breadbasket of a nation.
(iii) John Newton (1986) reported that from 1950 to the early 1980s more than 6,500 sink-
holes or related features appeared in the eastern United States. He further stated that
the total cost of damage and associated protective measures resulting from these
induced sinkholes is unknown, however, at five dam sites alone the repair costs were
already in excess of US$ 140 million.

363
Groundwater studies
Figure 14.1 Location map of the major outcrops of carbonate rocks (After Ford and Williams, 1989)
(iv) A Panel on Land Subsidence of the U.S. National Research Council, in 1991, related
that six states had individually sustained in excess of US$10 million from damage
resulting from sinkholes, and an additional four states sustained from US$1 to
10 million in damage from the same cause. As a result, awareness programmes for
catastrophic subsidence areas, and insurance programmes applicable to sinkhole
problems have been developed in the United States of America.
(v) Finally, carbonate rock or karst areas are dynamic and environmentally sensitive. The
geological structure, the solubility of the rocks involved, and the climatic conditions
determine to a large degree how rapidly changes can take place. Therefore, it is
necessary to recognise the synergistic relation between circulation of water and
solution of the rock. The solution leads to progressive lowering of water tables that
will result in base level changes, progressive cave enlargement, and changes in karst
topography that can all take place within a relatively brief period of time and bring
about major environmental problems. One of the first studies relating geography,
geology, mining, and resources and their development and impact on man and his
environment was published as an atlas series for northern Alabama, United States of
America (LaMoreaux, 1975).
Carbonate rocks are a source of abundant water supplies, minerals, and oil and gas.
However, there are also many problems related to developing adequate water supplies, assuring
proper drainage, providing stable foundation conditions, and preventing serious pollution
problems. Because of the complexity of carbonate rocks, it is impossible to base the evolution of
concepts related to the movement and occurrence of groundwater, methods of exploration and

364
Hydrogeology of carbonate rocks
development of water, safe engineering practices in construction of all kinds, and adequate
environmental safety precautions on one uniform set of rules.
The complexity of hydrogeological systems in carbonate rock terrain, requires that a
thorough hydrogeological study be carried out to determine whether a specific site is, or can be
rendered suitable, e.g. for construction, for highway location, for a water supply, or for a land
disposal facility. Important components of hydrogeological studies are:
field mapping of structural and stratigraphical units;
interpretation of sequential aerial photographs;
test drilling and geophysical analyses;
fracture analyses;
monitoring seasonal variation in water levels and in water quality;
establishing the spatial variation of hydraulic characteristics of the aquifer and asso-
ciated aquicludes, and the velocity and direction of movement of groundwater within
the aquifers;
determination of the controls for recharge, discharge, and local base level;
evaluation of the effects of mans activities, such as pumping, dewatering and con-
struction.
14.2 Factors determining the occurrence of groundwater in carbonate rock
The hydrogeology of carbonate rocks can only be understood by careful observation of the
physical characteristics and distribution of the rocks. The fundamental geological approach to
the exploration, use, and conservation of water involves considering:
the composition of the rocks,
the shape of their units,
the variation of the composition of the unit throughout its extent,
the sequence of units and how it varies,
the deformation of the sequence and the influence of the deformation on the charac-
teristics of the rocks,
the units, the sequence, and the topographical and physiographical situation These
characteristics are discussed in the following sections on structure, fracture systems,
joints, hydrogeological features, porosity, permeability, groundwater flow, and hydro-
chemical character.
14.2.1 Structure
The structural and tectonic history of a region plays a significant role in determining the
behaviour of carbonate aquifers. The range of variation among these many interdependent and
independent factors is so wide that no two carbonate aquifers have identical characteristics. The
influence of geological structure on aquifer characteristics is also variable, so that it is not
possible to provide a set of simple rules governing the influence of joints, faults and folds on
groundwater movement or availability. Thus in carbonate rock areas, it is imperative to develop
a conceptual framework to unravel the structural controls of groundwater movement for a
particular study area.
The tectonically induced position of a body of carbonate rocks determines potential
recharge/discharge relationships. Folding, faulting, and fracturing of the rocks often determine
the porosity and permeability of the aquifer, as well as specific directions of groundwater flow
through the carbonate system. After a recharge/discharge system has been established, perme-
ability of the carbonate rocks must increase with time if the transmitted groundwater is

365
Groundwater studies
unsaturated in calcium and/or magnesium carbonate. Therefore, karst development is a progres-

sive process although its probably not cyclic in the sense described by Penck (1900) or Davis
(1901). However, the characteristics of carbonate aquifers continue to change as long as an open
groundwater system exists.
14.2.2 Fracture systems
In carbonate rocks that do not have a primary intergranular permeability, joints or fractures are
essential for initiation of downward percolation of water (Stringfield et al., 1979). The lateral, as
well as vertical routes along which groundwater flow is channelled prior to solution
modification, may also be controlled by fracture patterns (Kiersch and Hughes, 1952). Research
performed throughout the world has documented the critical importance of fractures in
controlling groundwater movement in carbonate aquifers of Paleozoic and Mesozoic age. In fact,
the relationship between fractures and permeability developed by solution in carbonate rocks is
so well documented that one almost presumes the presence of fractures if solution cavities exist
in Paleozoic or Mesozoic limestones. Bedding planes may also provide avenues for groundwater
movement (Palmer, 1977), but even in such cases, movement of groundwater between bedding
planes is often fracture controlled. Tertiary and Quaternary limestones frequently have
permeability due to solution development of their primary textural porosity (Stringfield, 1966),
but research by Vernon (1951) and Leve (1983) has demonstrated that in the Floridan aquifer
groundwater movement is locally influenced and modified by the presence of faults.
If a favourable recharge/discharge relationship exists for a carbonate unit, then water
can gain access to that unit and move through it. The groundwater movement will gradually
enlarge the fractures through which it moves by solution, and permeability of the unit will
increase with time.
Moore (1966) has postulated that pumping, oscillations from earth tides and distant
earthquakes may move water through joints and partings too narrow to allow much movement
of water under normal hydraulic gradients. Fractures in carbonate rocks commonly have an
uneven distribution, consequently groundwater movement often develops secondary perme-
ability in selective areas of closely-spaced fractures or along open fracture systems. Therefore,
according to LeGrand et al. (1971), moderately large openings tend to enlarge by solution action
while small openings not in the path of water flow generally enlarge only slightly. The net result
is a typically uneven distribution of permeability in aquifers where permeability is controlled by
fractures.
14.2.3 Joints
Joints are the most frequently occurring fractures found in carbonate aquifers. Typically, they are
nearly perpendicular to bedding. However, depending on conditions of formation, joints may
have any orientation with respect to bedding. For example, Grice (1968) found that joints near
Grand Rapids, Manitoba, Canada, were essentially parallel to bedding planes.
Joints may exert primary control over the direction of groundwater flow prior to devel-
opment of solution cavities in a carbonate aquifer. Therefore, non-vertical joints may exert a
considerable lateral influence on the movement of groundwater. Therefore, use of rose diagrams
or histograms to represent the orientation of joints may result in considerable error unless these
data are carefully used to estimate groundwater flow directions.
If recharge/discharge relationships allow groundwater to move through an open car-
bonate aquifer, then joints can become enlarged by solution of the limestone. The result may
lead to development of: (1) a very permeable and cavernous, unsaturated zone; (2) a zone of an

366
exceptionally high permeability in valleys; and (3) a rapidly decreasing permeability with
increasing depth below the water table, as postulated by LeGrand et al. (1976).
Figure 14.2 contains three examples of joint enlargement by solution. It also illustrates
movement of water along bedding planes and an apparent rapid decrease in permeability with
depth. Stearns (1977) has demonstrated that the openings in the middle Ordovician limestones
of the Percy Priest Dam area in central Tennessee are a box-work of bedding planes and joints.
Figure 14.2 Vertical section through limestone outcrops showing diminution of fissure widths with depth
(From Williams, 1983)
367
Groundwater studies
The porosity of the limestone is approximately 15 percent in the upper 3 m but decreases rapidly
to about 1.5 percent at a depth of 10 m.
14.3 Hydrogeological features of carbonate rocks
The primary porosity of carbonate rocks is the result of open spaces in the rocks that have
persisted throughout the period of deposition, diagenesis and lithification. Textural porosity and
intergranular porosity are common terms used as synonyms for primary porosity. Secondary
porosity is a term used to express the amount of open space in a rock that has been created by
post-lithification processes such as fractures (joints, faults, parting) or solution cavities. To be
able to solve problems in carbonate aquifers, the porosity characteristics of these aquifers must
be thoroughly understood. One of the textbooks on this subject is Guide to the hydrology of
carbonate rocks, published by UNESCO (LaMoreaux et al., 1984).
Hydrogeological parameters such as porosity, storage coefficient, permeability, and
leakance must be determined in carbonate aquifers and related to the structure of the voids and
fissures that predominate. To assess the water resources of an area the process of evaluation
must use a variety of geological, chemical, engineering and mathematical disciplines
The principal characteristics of carbonate rocks, namely their heterogeneity and aniso-
tropy, determine the porosity and permeability parameters at the spatial scale at which these
parameters are studied. Results of laboratory and field tests represent only certain local values
and cannot be generalised to the entire aquifer without considering the geological characteristics
such as lithostratigraphy, structure, and geomorphology (Milanovich, 1981; White and White,
1989; LaMoreaux et al., 1989; Burger and Dubertret, 1975).
14.3.1 Porosity
Porosity is the property of a rock having voids or interstices. The types of porosity of unconfined
or confined aquifers can be classified on the basis of the nature and proportion of voids, inter-
stices, microfissures and channels in the rock and in terms of the amount of groundwater
released gravitationally or retained. The characteristics of the voids, their shape, size, distribu-
tion, and volume, compared to the overall rock volume are the essential features necessary for
defining the porosity types. On the basis of void types, porosity may be microscopic or macro-
scopic (Table 14.1 and Figure 14.3).
Table 14.1 Classification of voids and porosity
Reference scale Void types Porosity types
Pores or interstices Intercrystalline Intercrystalline porosity
Microscopic Pores or interstices Intergranular Interstitial porosity
Microfissures Joints Microfissure porosity

Microfissures
Macroscopic Channels Channel porosity

Cavities

368
Figure 14.3 Porosity types in soluble carbonate rocks: (a) rock porosity; (b) microfissure porosity
14.3.2 Permeability
Permeability depends upon the void characteristics, mainly their shapes and opening sizes.
There are several types: intrinsic permeability, fissure permeability, and channel permeability,
but the hydrodynamic phenomena are the same. Nevertheless, it is convenient to distinguish
between the rock or intrinsic permeability (microscopic) of the rock mass and the regional or
formational permeability (macroscopic). The latter includes the added permeability supplied by
joints, fissures, and channels. Rocks intrinsic permeability is very small compared to the
regional or formational permeability.
Similar to the porosity, the permeability varies with time, and it usually increases with the
increase of the size of openings of the fissures. Therefore, an initial permeability and a transient
permeability resulting from the widening of the voids, may be recognised (White and White,
1989; Milanovich, 1981; LaMoreaux et al., 1984).
14.3.3 Groundwater flow
Considerable progress in understanding the hydrology of carbonate rocks has been made in
recent years. However, some investigators still disagree on the nature of the occurrence and
movement of water in carbonate rocks. Most investigators of groundwater in carbonate rocks
identify a zone of saturation (phreatic water) with either a water table under non-artesian
conditions or a piezometric surface under artesian conditions. A water table concept is useful
where the permeability is so poor that the water table is discontinuous or so great that water
moves through large artery-type openings. Special emphasis is placed on carbonate rocks as
conduits for the transmission of water because only where they act as conduits are they in
contact with moving water, and therefore, vulnerable to solution. A conduit containing water in

369
Groundwater studies
a carbonate rock may be under either water table or artesian conditions; in both cases, three
things are essential. It must have (1) an area of intake, and it must (2) transmit, and (3) discharge
water. If any one of these three requirements is not met, the limestone body is hydrologically
inert and cannot act as a conduit for water (LeGrand and Stringfield, 1966).
Theoretically, groundwater moves in arcuate paths following lines of flow that have
their origin at the top of the zone of saturation in recharge areas (Swinnerton, 1942). The flow
lines curve downward for some distance and then rise to an outlet or point of discharge.
Diagrammatically, these lines can be represented by a family of curves, most closely spaced near
the area of the outlet and becoming more widely spaced along the top of the zone of saturation
with increasing distance from the outlet. The water may be expected to have this arcuate pattern
of flow in uniformly permeable material where geological structure has only subordinate
influence on its direction of movement. However, even in the initial stages of groundwater
circulation in limestone and other carbonate rocks, lines of flow are modified. Circulation in the
discharge area, having greater velocities, may result in an enlargement of the outlet and a
consequent shallowing of the more arcuate paths. Solution openings along the more direct paths
will become larger than those along the less direct paths and will permit progressively larger
flows at the expense of other passageways.
14.4 Examples of groundwater flow systems in carbonate rocks
14.4.1 Karst hydrological systems
The functions of a karst hydrological system are controlled by stratigraphical features and
geological structure. Its extent varies from 1 km2 to thousands km2. The input of the system may
include direct inflow of precipitation through natural shafts, infiltration of precipitation through
epikarst zone and soil, and allogenic water from non-soluble rock areas nearby, and conden-
sation of water in caves In some cases, the cave condensation rate could be 0.5 l per m3 cave
volume per day, but depends on many local factors. The output of the system could be a spring
outlet or an underground stream, underflow beneath a sediment plain, submarine spring, etc. A
clear and quantitative understanding of the karst hydrological system must be the basis for
studying groundwater in soluble carbonate rocks.
The prerequisites for identifying a karst hydrological system are the comprehensive
mapping of:
the stratigraphical relation between carbonate rock and insoluble rock;
the pattern of geological structure;
the relief and surface drainage system (river, lake, sea, glacial, etc.);
the relationship with alluvium and other aquifers,
For instance, in Figure 14.4, pattern 1 is a tight folding of carbonate rock alternating with
insoluble rock along deep gorges, giving a series of karst hydrological systems in a narrow belt
with deep water circulation and small catchment area. Pattern 2 shows a gently folding hydro-
logical system that enjoys a broader catchment area. The water table is shallow because it is
perched on an impervious layer. Nevertheless, there is a deep gorge.
The basic approaches for studying the hydrology of carbonate rock areas are as follows:
mapping of the geomorphology of the hydrologically important surface karst features,
such as polje, doline, sinkhole, etc.;
survey of caves and geophysical prospection for siting underground conduits;
laboratory porosity analysis, applicable for young porous carbonate rocks;
hydrological, hydrochemical and isotopical response of the system to storm (Fig. 14.5).

370
Figure 14.4 The relationship between stratigraphy, folding pattern, relief and karst hydrological system;
1: tight folding, 2: gentle folding
Figure 14.5 The hydrological, hydrochemical and isotopic response to a storm of S31 karst spring near Yaji,
Guiin, China 1986 (Yuan Daoxian, 1988)
371
Groundwater studies
14.4.2 Characteristics of karst hydrological systems
Only a brief statement regarding the physical, hydrological, and geomorphological aspects of
carbonate rocks can be provided here. What all karst terrain have in common is that they are
heterogeneous, complicated, and dynamic. To summarise, a table of characteristics is provided
as an aid to understanding any type of development in a carbonate rock area (Table 14.2).
The nature of underground circulation depends largely upon the geological structure and
the relation of permeable rocks to groundwater discharge areas. Insofar as circulation is con-
cerned, a carbonate formation or body at any one place possesses at least one of the following
four types of hydrological zones (LeGrand and Stringfield, 1966):
Zone 1
Carbonates at or near the surface: The water table occurs in the karst rock. Water from
precipitation moves vertically downward to the water table and then laterally toward a
surface stream.
Zone 2
Carbonates buried beneath an impermeable bed, forming a homoclinal artesian system:
under artesian pressure water from a higher area moves through the rock towards a lower
discharge area.
Zone 3
Carbonates with no significant discharge facilities: either (a) a homoclinal artesian system
so deeply buried beneath impermeable beds that almost no water can escape, or be
(b) carbonates lying below the stream controlling the base level of erosion and perhaps so
faulted or folded as to nearly preclude the discharge of water.
Zone 4
Denuded carbonates elevated above subjacent valleys and sufficiently impervious locally
to preclude a continuous zone of saturation: no subsurface discharge of water occurs. This
zone is absent in many areas.
Although both the occurrence and movement of groundwater in carbonate rocks are
related to geological structures, the relationships are not direct and exclusive. We cannot
rightfully place such geological structures as synclines, anticlines, monoclines, and faults into
separate categories for an appraisal of their relation to the occurrence and movement of ground-
water. In some areas, faults may serve as avenues for movement of water, and in others they
may serve as barriers. An acceptable study technique requires the geological framework to be
studied in relation to the superimposed hydrological continuum. Is the hydrological continuum
real and significant for a particular setting? The answer may be negative if one of the following
conditions applies: (1) insignificant recharge facilities, (2) insignificant discharge facilities,
(3) insignificant permeability in the system, and (4) insignificant hydraulic head in the system.
Where carbonate bodies are thin and are compartmentalized by other rocks in complex
structural settings, the hydrology is commonly similar to that of the enclosing rocks. Under
these conditions some aspect of (a) discharge, (b) movement through the rock, or (c) recharge is
likely to be so restricted that a carbonate rock does not act as a separate hydrological system.
Such microcompartments, in which circulation of water is so slight as to restrict karst devel-
opment, should be distinguished from macrocompartments in which the breadth of carbonates
in a water-table circulation system may be tens of kilometres. Where carbonates occur as macro-
compartments, many features of karst hydrology may develop; yet, the hydrology will be
influenced by the hydrological and topographical features of the enclosing foreign beds.
Karstification occurs where circulation of water is not impeded. The circulation of water
is not evenly distributed in carbonate rocks, therefore the solution-developed openings repre-
senting major permeability features of karst rocks are also unevenly distributed.
Compartments of groundwater in carbonate rocks may come about as a result of aquifers

372
Table 14.2 Important characteristics of carbonate rocks (Hughes et al., 1994)
Stratigraphy Geological Structure Geomorphology

(Regional and local) (Regional and local) (Regional and local)
Stratigraphical column Nearly horizontal bedding Relief slopes
Thickness of each carbonate unit Tilted beds Density of drainage network
Thickness non-carbonate interbeds Homoclines Characteristics of streams
Type of bedding Monoclines Drainage pattern
Thin Folded beds Dendritic
Medium Anticlines Trellis
Thick Synclines Rectangular
Purity of each carbonate unit Monoclines Other
(Limestone or dolomite) Domes Perennial
Pure Basins Intermittent
Sandy Other Terraces
Silty Fractures Springs and/or seeps
Clayey Lineaments Lakes and ponds
Siliceous Locations Floodplains and wetlands
Interbeds Relationships with: Karst features - active
Geomorphic features historic
Overburden Karst features Karst plains
(Soils and sub-soils) Stratigraphy Poljes
Distribution Structural features Dry valleys, blind valleys
Origin Joint system Sinking creeks
Transported Joint sets Depressions and general
Glacial Orientation subsidence
Alluvial Spacing Subsidence cones in
Colluvial Continuity overburden
Residual Open Sinkholes
Other Closed Roof-collapse
Characteristics and variability Filled Uvalas
Thickness Faults Caverns, caves and caveties
Physical properties Orientation Rise pits
Hydrological properties Frequency Swallow holes
Continuity Estavelles
Hydrology Type Karren
Surface water Normal Other
Discharge Reverse Paleo-Karst
Variability Thrust
Seasonal Other Climate
Gaining Age of faults Precipitation
Losing Holocene Seasonal
Groundwater Pre-Holocene Annual
Diffuse flow Long-term
Conduit flow Activities of man Temperature
Fissure flow Construction Daily
Recharge Excavation Seasonal
Storage Blasting Annual
Discharge Vibration Long-term
Fluctuation of water levels Loading Evapotranspiration
Relation of surface water and Fill Vegetation
groundwater flow Buildings
Changes in drainage
Dams and lakes
Withdrawal of groundwater
Wells
Dewatering
Irrigation

373
Groundwater studies
being intercepted by impermeable or insoluble materials because of faulting or folding. Some

compartments occur in broad carbonate aquifer systems that are close enough to the ground
surface to be within the meteoric water circulation system. A good example of this type of
groundwater occurrence is present in the gold mining area of South Africa near Pretoria (Enslin,
1967; Buttrick, 1992). In such a system, a drainage network with a dominant line of flow lies on
the piezometric surface and depresses it. These drainage lines may be perennial surface streams
into which water from the carbonates discharges, or, in the case of mature karst areas where no
perennial surface drainage occurs, they may be lines of concentrated groundwater drainage that
lie at the base of air-filled openings, depressing the water table and leading to points of ground-
water discharge. These compartments are separated laterally by groundwater divides, which
do not necessarily coincide with topographical divides on the land-surface. Vertically below they
merge into the overall carbonate system.
Lack of uniform permeability may be attributed to a combination of the following con-
ditions. Separate parts of carbonate formations may have had (1) different original porosity-
permeability characteristics, (2) different opportunities for groundwater to circulate in them,
(3) different degrees of access to water undersaturated in calcium carbonate and (4) different
opportunities for erosion or, conversely, for preservation (Stringfield and LeGrand, 1966).
Karst hydrological systems can be divided into two types: (i) exposed karst and
(ii) buried karst (carbonate rock aquifer buried under impervious formations), or covered karst
(carbonate rock aquifer covered by loose sediments). Land surface topography has an ever
present influence on the position of the water table. Combinations of topographical and
permeable conditions control the positioning of the water table in important practical ways. The
following general statement forms a basis for useful generalisations: A water table deep below
land surface occurs where high relief and very permeable limestone are combined. Local
variations of considerable magnitude in permeability are common in carbonate terrain. They
contribute to the disappearance of surface streams into underground courses and to their reap-
pearance at other places.
The theory of groundwater motion in small drainage basins, which has been described
by Toth (1962, 1963), has application to broad carbonate regions on which local discharge areas
are superimposed. Where these compartments occur it is wise to evaluate the most likely
directions of flow of water. In many cases more than 90 percent of recharge is shunted out into
one of the compartments, leaving little water to reach the lower and major carbonate system.
The lack of opportunity for water to move out of the lower part of the carbonates allows this
part to stay full of water and to repel most of the available recharge. A conceptual model that
considers compartments of carbonate hydrology helps when evaluating the distribution of the
flow of water and of the contaminants this water may contain.
Where an artesian circulation system occurs, base-level considerations may be less
significant because the piezometric surface is above the aquifer. Zones of greater circulation and
solution may develop, perhaps near the top of the confined carbonate aquifer, but these zones
are not closely related to base level. Development of secondary permeability in an artesian
carbonate system may have considerable overall importance but is slower and less dynamic
than the development of secondary permeability in a water table carbonate aquifer. It should be
noted, however, that some carbonates in the artesian system were at some earlier time in a
water-table circulation system.
Under artesian conditions in inclined aquifers, water may move down dip where: (1) the
aquifer is overlain by relatively impervious beds, (2) the relative positions of the recharge and
discharge areas are favourable, and (3) the hydraulic gradient is in the same direction as the dip
(Stringfield and LeGrand, 1966). Such conditions are not widespread because much water
follows arcuate courses to concentrated discharge areas, and locally the direction of this water
movement cannot coincide with the dip. Where the permeable zones connecting recharge and

374
discharge areas are not homoclinal, as in folded aquifer systems typical of carbonates of the
southern Appalachians (United States of America), water may move down dip on one limb of
the structure and up dip to some degree on the other limb.
To obtain the most reasonable assessment of water resources in exposed karst, the time
series of precipitation data should be as long as possible, as this is helpful not only for a more
accurate assessment of the water resources, but also for a better understanding of the hydro-
logical function of the karst system. For instance, the 60-year-long records of the biggest karst
spring of North China, the Nyangziguan Spring, show 20-year cycles of dry and wet periods
(Fig. 14.6) and a 3-year time lag of spring discharge records compared to the precipitation
records.
Figure 14.6 Correlation between precipitation and discharge of Nyanziguan karst spring, China,
for the past 60 years (Zhang Fenngqi, 1992)
Because of the uneven distribution of rainfall in the mountainous region precipitation

gauges should be installed in the studied catchment area and water table gauges in the epikarst
zone, the soil water or vadose zone, and the general outlet of the system (Fig. 14.7). Given the
quick response of water table and discharge to precipitation (Fig. 14.8), it is highly recom-
mended to use automatic recorders for precipitation and water table gauges. For the observation
of a deep and greatly fluctuating water table a pneumatic water table gauge should be used
instead of a float gauge. If available, a satellite transfer multidata collecting platform powered
by solar energy should be used.
Discharge of water from carbonate aquifers is commonly less diffused than that from
other types of aquifers. Development of large solution openings through which there is prefer-
ential movement of water gives rise to large springs in many low places in carbonate terrain; the
springs may be obscure or somewhat indistinct where they occur in river channels and in
shallow seas. Discharge may be diffused where water passes from carbonate into sands or where
relatively impermeable rocks underlie the aquifer above the low areas (Burdon and Papkis,
1963).

375
Groundwater studies
Figure 14.7 Components of water flow and monitoring facilities in Yaji karst hydrological system, Guilin, China.
(Yuan Daoxian et al., 1990)
Notes: Q1 - slope runoff; Hydrological features of the springs

Q2 - epikarst; Max. discharge Min. discharge
Q3 - soil water. Spring (l/s) (l/s)
S31 7,000 0.1
S29 470 0
S29-1 120 0
S32 30 0

376
Figure 14.8 Hydrograph of spring S31 in Yaji karst system, Guilin, China, 1987 (Yuan Daoxian et al., 1990)
14.5 Hydrochemical character of carbonate rock aquifers
Hydrochemistry is an important part of water study for carbonate rocks. It can, on one hand,
reveal the mechanism of karst formation, the intensity and direction of the process, i.e., whether
it tends to dissolution or precipitation. The mixing corrosion on the coastal zone of Yucatan
Peninsula, which was revealed through detailed hydrochemical studies, is a good example. On
the other hand, the hydrochemical field in a carbonate rock aquifer, and the time variation of
hydrochemical features at the output from the aquifer are important information from which the
hydrological functions such as recharge, water storage and regulation, and the characteristics of
the aquifer can be deduced. Accordingly, hydrochemical studies are very often used as a tool for
hydrogeological exploration (Back et al., 1986).

377
Groundwater studies
14.5.1 Important hydrochemical features of carbonate rocks
The most essential hydrochemical characteristic of carbonate rocks is that it is a very active
triphase, disequilibrium, open system because of the involvement of CO2. The entering or
outgassing of CO2 will change not only the hydrochemistry of the water, but also the intensity
and direction of karst processes. Consequently, the hydrochemistry is very sensitive to many
environmental factors such as rainfall, temperature, photosynthesis, and other biogenic pro-
cesses as sunshine, turbulence, and depth of water. The hydrochemical features are different
even in different parts of the same water body. The traditional water/rock interaction approach
based merely on laboratory chemical analysis may quite often lead to serious errors (see the
special Issue of the Journal of Environmental Geology, JEG, 1995).
14.5.2. Environmental aspects and recommended references
The major environmental problems related to groundwater in carbonate rocks are drought,
flood, deforestation, surface collapse and water pollution (Memon and Prohic, 1989). Some of
these are caused by natural change of water regime, but most of them are induced by human
activities, i.e., inappropriate management on groundwater in carbonate rocks (see the special
issue of the Journal of Environmental Geology (JEG, 1993)). Different environmental problems
are sometimes related to each other, when they occur in the same karst hydrological system.
Accordingly, an overall understanding of the karst hydrological system is the basis for a proper
management of groundwater and environment (Newton, 1976; Holzer, 1991).
Human endeavours involving karst terrain are confronted with a much wider range of
conditions with poorly predictable surface and subsurface responses than occurs in most other
terrains. Sensitive environmental problems prevail, many of which cannot be predicted and
which have subtle and indirect cause-and-effect relations. Thus, certain legal aspects are very
prominent when land is exchanged or dewatered and when the consequences of an action on
one property affect neighbouring properties. It is not the purpose of this paper to describe the
principles of karst hydrogeology that can be found in many geological texts on the subject
(LaMoreaux and Newton, 1986; LaMoreaux et al., 1997).
A description of the structural setting is essential for understanding karst hydrology:
(1) development (water and land use), (2) interactions, and (3) problems that may result in legal
actions (LaMoreaux and Newton, 1986). Attention has been directed to these subjects in
numerous publications, for example those by Meinzer (1923), Herak and Stringfield (1972), and
Burger and Dubertret (1975). A summary of key principles described by LeGrand and
LaMoreaux (1975), and specific methods for describing the geology, lithology, and structural
setting in a karst area are discussed in great detail in the UNESCOs Guide to the Hydrology of
Carbonate Rocks (LaMoreaux et al., 1984). Detailed descriptions of the important physical
characteristics for siting a landfill in a karst terrain are given in Landfills in Karst Terrains (Hughes
et al., 1994).
Various regional symposia and colloquia on karst have been organised world-wide over
the past 20 years by IAH, IAHS, FAO and UNESCO within the International Hydrological
Decade (IHD) and International Hydrological Programme (IHP). The IHD included a com-
mission for the study of carbonate rocks in Mediterranean countries, and since 1970 there has
been a permanent commission for karst hydrogeology within IAH. There have been over
150 field trips, meetings, symposia, and associated documents from this IAH activity, including
Karst Waters and Environmental Impacts (UKAM, 1995). In southern France, one of the most
detailed series of studies of spring flow, water management, and development for municipal
use has been carried on by Professor Jacques V. Avias (1995) in his studies of the karstic spring
Source du Lez. This spring supplies the city of Montpellier, France.

378
One of the most comprehensive series of papers on carbonate rocks appeared as a series of
books on the symposium held under the leadership of Dr Barry Beck, first sponsored by the
Sinkhole Institute of Florida and beginning in 1995, by the firm of P. E. LaMoreaux and
Associates. These symposium papers, published by Balkema are: Beck, 1984, Beck and Wilson,
1987; Beck, 1989, 1993, 1995 and Beck and Stephenson, 1997. Since 1990, UNESCO and IUGS
supported the International Geological Correlation programme (IGCP) that organized three
projects to deal with the karst hydrological and environmental problems of the world, incl.
IGCP299 Geology, climate, hydrology and karst formation (19901994), IGCP379 Karst processes and
the carbon cycle (19951999) and IGCP448 World Correlation of karst ecosystems (20002004). A
series of books was published, including Global Karst Correlation, VSP, the Netherlands.
The reasons for the increased interest in the hydrogeology of carbonate rocks are the rapid
increase in population, the application of new technologies, the more rapid development of
natural resources, and the more extensive exploitation of hydrological systems. Karst occurs in
many parts of the world where groundwater supplies represent the sole or most important
natural resource to a population. Thus, understanding the behaviour of groundwater systems in
a karst area directly affects the local or regional social and economic development.
AVIAS, J. V., 1995. Active Management of Lez Karstic Spring (Hrault, France) 19571994.
Hydrogologie, No. 1, ditions BRGM, Orlans, pp. 11328.
BACK, W.; HANSHAW, B. B.; HERMAN, J. S.; VAN DRIEL, J. N., 1986. Differential Dissolution of a
Pleistocene Reef in the Ground-Water Mixing Zone of Coastal Yucatan, Mexico. Geology,
Vol. 14, pp. 137140.
BAKALOWITCZ, M., 1979. Contribution de la gochimie des eaux la connaissance de laquifre
karstique et de la karstification. Thse de doctorat, Lab. souterrain du C.N.R.S., Moulis,
France.
BECK, B. F. (ed.), 1984. Sinkholes: Their Geology, Engineering and Environmental Impact. Proceedings
of the First Multidisciplinary Conference on Sinkholes, Orlando, Florida, 1517 October
1984. A. A. Balkema, Rotterdam, 429 pp.
(ed.), 1989. Engineering and Environmental Impacts of Sinkholes and Karst. Proceedings of the
third multidisciplinary conference on sinkholes and the engineering and environmental
impacts of karst, St. Petersburg Beach, Florida, 24 October 1989. A. A. Balkema,
Rotterdam, 384 pp.
(ed.), 1993. Applied Karst Geology. Proceedings of the fourth multidisciplinary conference on
sinkholes and the engineering and environmental impacts of karst, Panama City, Florida,
2527 January 1993. A. A. Balkema, Rotterdam, 295 pp.
(ed.), 1995. Karst Geohazards, Engineering and Environmental Problems in Karst Terrain.
Proceedings of the fifth multidisciplinary conference on sinkholes and the engineering
and environmental impacts of karst, Gatlinburg, Tennessee, 25 April 1995. A. A. Balkema,
Rotterdam, 581 pp.
BECK, B. F.; STEPHENSON, J. B., (eds.), 1997. The Engineering Geology and Hydrogeology of Karst
Terrain. Proceedings of the sixth multidisciplinary conference on sinkholes and the
engineering and environmental impacts of karst, Springfield, Missouri, 69, April 1997. A.
A. Balkema, Rotterdam, 516 pp.
BECK, B. F.; WILSON, W. L., (eds.), 1987. Karst Hydrogeology: Engineering and Environmental
Applications. Proceedings of the second multidisciplinary conference on sinkholes and the
environmental impacts of karst, Orlando, Florida, 911 February 1987. A. A. Balkema,
Rotterdam, 467 pp.

379
Groundwater studies
BURDON, D. J.; PAPKIS, N., 1963. Handbook of Karst Hydrogeology, with Special Reference to the
Carbonate Aquifers of the Mediterranean Region. Inst. Geol. and Subsurface Research, Athens,
United Nations Special Fund, mimeo, 276 pp.
BURGER, A.; DUBERTRET, L., (eds.), 1975. Hydrology of Karstic Terrain. Int. Union Geol. Sci., Ser. B,
No. 3, 190 pp.
BUTTRICK, D. B., 1992. Characterisation and Appropriate Development of Sites on Dolomite. In partial
fulfilment of the requirements for the degree of Doctor of Philosophy in the Faculty of
Science, University of Pretoria, November 1992, 182 pp.
CHEN XIAOPING ET AL., 1998. South China Karst I. Zalozil (ZAL, ZBA, ZRC), Republic of Slovenia,
247 pp.
DAVIS, W., 1901. An Excursion in Bosnia, Herzegovina and Dalmatia. Bull. Geograph. Soc., Vol. 3,
No. 2, pp. 4750.
DREW, D.; HOETZL, H., (eds.) 1999. Karst Hydrogeology and Human Activities: Impacts, Consequences
and Implication A. A. Balkema, Rotterdam, 322 pp.
ENSLIN, J. F., 1967. Dolomitic Water Supplies in the Transvaal Republic of South Africa. Mem. IAH
Conference, Istanbul, Sept. 1967, p. 78.
FORD, D.; WILLIAMS, P., 1989. Karst Geomorphology and Hydrology. Unwin Hyman Ltd., London,
601 pp.
GRICE, R. H., 1968. Hydrogeology of the Jointed Dolomites, Grand Rapids Hydroelectric Power
Station, Manitoba, Canada. In: Engineering Geology Case Histories: No. 6, Geol. Soc. of Am.,
pp. 3348.
GULTEKIN GUNAY; JOHNSON, A. I., (eds.), 1997. Karst Waters and Environmental Impacts. Proc. 5th
international symposium and field seminar on karst water and environmental impacts,
Antalya, Turkey, 1020 September 1995. A. A. Balkema, Rotterdam/Brookfield, 525 pp.
HERAK, M.; STRINGFIELD, V. T., (eds.), 1972. Karst: Important Karst Regions of the Northern
Hemisphere. Elsevier, Amsterdam, 526 pp.
HOLZER, T. L. (ed.), 1991. Mitigating Losses from Land Subsidence in the United States, Panel on Land
Subsidence. U.S. Geological Survey, Menlo Park, California, National Academy Press,
Washington, D.C., 58 pp.
HUGHES, T. H.; MEMON, B. A.; LAMOREAUX, P. E., 1994. Landfills in Karst Terrains. Bull. Assoc.
Eng. Geol., Vol. XXXI, pp. 2038.
JAMES, N. P.; CHOQUETTE, P. W., (eds.), 1987. Paleokarst. Springer-Verlag, New York, NY, 416 pp.
JEG, 1993. Environmental Change in Karst Areas. Journal of Environmental Geology, special issue,
Vol. 21, No. 3, pp. 10782.
JEG, 1995. Geochemical Environments in Karst. Journal of Environmental Geology, Vol. 25, No. 1,
69 pp.
KIERSCH, G. A.; HUGHES, P. W., 1952. Structural Localisation of Groundwater in Limestone Big
Bend District, Texas-Mexico. Econ. Geology, Vol. 47, pp. 794806.
KLIMCHOUK, A. B.; FORD, D. C.; PALMER, A. N.; DRYBRODT, W., (eds.) 1999. Speleogenesis; Evolution of
Karst Aquifers. National Speleological Society, Inc., Huntsville, Alabama, 527 pp.
LAMOREAUX, P. E., 1975. Environmental Geology and Hydrology, Huntsville and Madison County,
Alabama. Geological Survey of Alabama, Atlas Series 8, University, Alabama, 119 pp.
LAMOREAUX, P. E.; HUGHES, T. H.; MEMON, B. A.; LINEBACK, N., 1989. Hydrogeologic Assessment
Figeh Spring, Damascus. Environmental Geology and Water Sciences, Vol. 13, No. 2,
pp. 73127.
LAMOREAUX, P. E.; NEWTON, J. G., 1986. Catastrophic Subsidence: An Environmental Hazard,
Shelby County, Alabama. Environmental Geology and Water Sciences, Springer-Verlag,
Heidelberg, Vol. 8, pp. 2540.
LAMOREAUX, P. E.; POWELL, W. J.; LEGRAND, H. E., 1997. Environmental and Legal Aspects of
Karst Areas. Environmental Geology, Vol. 29, pp. 2336.

380
LAMOREAUX, P. E.; WILSON, B. M.; MEMON,B. A., (eds.), 1984. Guide to the Hydrology of Carbonate
Rocks, Studies and Reports in Hydrology. UNESCO, Paris, 345 pp.
LEGRAND, H. E.; LAMOREAUX, P. E., 1975. Hydrogeology and Hydrology of Karst. In: A. Burger,
L. Dubertret, (eds.). Hydrogeology of Karst Terrains. Intern. Assoc. Hydrogeol., Paris,
pp. 919.
LEGRAND, H. E.; STRINGFIELD, V. T., 1966. Development of Permeability and Storage in the
Tertiary Limestones of the Southeastern States. Intern. Assoc. Sci. Hydrol., Vol. 9, No. 4,
pp. 6173.
, 1971. Development and Distribution of Permeability in Carbonate Aquifers. Water Resources
Research, Vol. 7, No. 5, pp. 128494.
LEGRAND, H. E.; STRINGFIELD, V. T.; LAMOREAUX, P. E., 1976. Hydrologic Features of United States
Karst Regions. In: J. W. Tolson, F. L. Doyle, (eds.), Karst Hydrology. AIH Memoirs, Vol. XII,
pp. 3146.
LEVE, G. W., 1983. Relation of Concealed Faults to Water Quality and the Formation of Solution
Features in the Floridan Aquifer, Northeast Florida, United States of America. In: W. Back,
P. E. LaMoreaux, (guest eds.), V. T. Stringfield Symposium: Processes in Karst Hydrology.
Journal of Hydrology, Vol. 61, pp. 25164.
MEINZER, O. E., 1923. Outline of Groundwater Hydrology with Definitions. U.S.G.S. Water Supply
Paper 494, 71 pp.
MEMON, B. A.; PROHIC, E., 1989. Movement of Contaminants in Karstified Carbonate Rocks.
Environmental Geology and Water Sciences, Vol. 13, pp. 313.
MILANOVICH, P., 1981. Karst Hydrogeology. Water Resources Publications, Fort Collins, Co., 434 pp.
MOORE, G. C., 1966. Introduction to Limestone Hydrology. Nat. Speleol. Soc. Bull., Vol. 28,
pp. 10917.
NATIONAL RESEARCH COUNCIL, 1991. Mitigating Losses from Land Subsidence in the United States.
NRC, Panel on Land Subsidence. T. L. Holzer and committee members (eds.). National
Academy Press, Washington, D.C., 58 pp.
NEWTON, J. G., 1976. Early Detection and Correction of Sinkhole Problems in Alabama, with a
Preliminary Evaluation of Remote Sensing Applications. HPR Report No. 76, Research Project
930070, prepared by United States Geological Survey in co-operation with Alabama
Highway Department and United States Department of Transportation, January 1976,
83 pp.
Newton, J. G., 1986. Development of Sinkholes Resulting from Mans Activities in the Eastern United
States. U.S. Geological Survey, Denver, CO, 54 pp.
PALMER, A. N., 1977. Influence of Geologic Structure on Groundwater Flow and Cave
Development in Mammoth Cave National Park, United States of America. In: J. S. Tolson,
F. R. Doyle, (eds.), Karst Hydrogeology. AIH Memoirs, Vol. XII, pp. 40530.
PALMER, A. N.; PALMER, M. V.; SASOWSKY, I. D., (eds.), 1999. Karst Modelling. Proceedings
Symposium held February 2427, 1999, Charlottesville, Virginia. Karst Water Institute
special publication No. 5, 265 pp.
PALOCK, H.; BACK, W.; HERMAN, J. S., (eds.), 1992. Hydrogeology of Selected Karst Regions. IAH
Contributions to Hydrogeology, Vol. 13. Verlag Heinz Heise, Germany, 494 pp.
PECK, D. L.; TROESTER, J. W.; MOORE, J. E., 1988. Karst Hydrogeology in the United States of
America. In: Proc. of the 21st Congress of the International Association of Hydrogeologists, Karst
Hydrogeology and Karst Environment Protection, Guilin, China. U.S. Geological Survey, Open-
File Report 88-476, 19 pp.
PENCK, A., 1900. Geomorphologische Studien aus der Hercegovina. Z. Deut. Osterreich. Alpenver,
Vol. 31, pp. 2541.
SAURO, U.; BONDESAN, A.; MENEGHEL, M., (eds.), 1991. Proceedings of the International Conference on

381
Groundwater studies
Environmental Changes in Karst Areas (Italy, 1527 September 1991). Quaderni del
Dipartimento di Geografia 13, Italy, 414 pp.
SMITH, D. I.; ATKINSON, T. C., 1977. Underground Flow in Cavernous Limestones, with Special
Reference to the Malham Area. Field studies, Vol. 4, pp. 597616.
STEARNS, R. G., 1977. Porosity and Hydrology of Jointed Middle Ordovician Limestones in the
J. Percy Friest Dam Area of Central Tennessee. In: J. S. Tolson and F. L. Doyle, (eds.), Karst
Hydrology. IAH. Memoirs, Vol. XII, pp. 4312.
STRINGFIELD, V. T., 1966. Artesian Water in Tertiary Limestone in the Southeastern States. United
States Geological Survey Professional Paper 517, Washington, D.C., 226 pp.
STRINGFIELD, V. T.; LEGRAND, H. E., 1966. Hydrology of Limestone Terraines in the Coastal Plain
of the Southeastern United States. Geological Society of America, Special paper, Vol. 93,
45 pp.
STRINGFIELD, V. T.; RAPP, J. R.; ANDERS, R. B., 1979. Effects of Karst and Geologic Structure on the
Circulation of Water and Permeability in Carbonate Aquifers. Journ. of Hydrol., Vol. 43,
pp. 31332.
SWEETING, M. M., 1995. Karst in China; its Geomorphology and Environment. Springer-Verlag, Berlin,
265 pp.
SWINNERTON, A. C., 1942. Hydrology of Limestone Terranes. In: O. E. Meinzer (ed.), Physics of the
Earth, pt. 9: Hydrology. Doves Publications, Inc., New York, pp. 65677.
TOTH, J., 1962. A Theory of Groundwater Motion in Small Drainage Basins in Central Alberta,
Canada. Journ. Geophys. Research, Vol. 67, pp. 437587.
TOTH, J., 1963. A Theoretical Analysis of Groundwater Flow in Small Drainage Basins. Journ.
Geophys. Research, Vol. 68, pp. 4795812.
UKAM, 1995. Karst Waters and Environmental Impacts. Proceedings of the Antalya symposium,
September 1020, 1995. UKAM, Haceteppe University, Ankara.
VERNON, R. O., 1951. Geology of Citrus and Levy Counties, Florida. Florida Geological Survey,
Bulletin 33, pp. 256.
WHITE, W. B.; WHITE, E. L., (eds.), 1989. Karst Hydrology, Concepts from the Mammoth Cave area, Van
Nostrand Reinhold, New York, 346 pp.
WILLIAMS, P. W., 1983. The role of the Subcutaneous Zone in Karst Hydrology. In: W. Back, P. E.
LaMoreaux, (eds.), V. T. Stringfield Symposium Processes in Karst Hydrology. Journal of
Hydrology, Vol. 16, pp. 4567.
YUAN DAOXIAN, (chief ed.), 1988. Karst Hydrogeology and Karst Environment Protection. Proceedings
of the IAH 21st congress, 1015 October 1988, Guilin. Geol. Publish. House, Beijing,
2 parts, 1261 pp.
YUAN DAOXIAN (ed.), 1991. Karst of China. Geological Publishing House, Beijing, 224 pp.
YUAN DAOXIAN; DROGUE, C. ET AL., 1990. Hydrology of the Karst Aquifer at the Experimental
Site of Guilin in South China. Journ. of Hydrol., Vol. 115, pp. 28596.
YUAN DAOXIAN; LIU ZAIHUA (eds.), 1998. Global Karst Correlation. Final report of IGCP299:
Geology, Climate, Hydrology and Karst Formation (19901994). Science Press, Beijing-
New York; VSP, Utrecht, 308 pp.
ZHANG FENNGQI; LI BOTAO, 1992. Artificial Adjustment and Recharge of Karst Springs, and Prospect of
Water Resources in Northern China. Proceedings of the Symposium on karst and karst
waters in north China, October 711, 1991. Jinan, Shandong; Guangxi Normal University
Press, pp. 8896.

382
15 H y d ro g e o l o g y o f h a r d ro c k s
15.1 Introduction
For the purpose of this chapter, hard rock aquifers are aquifers in non-carbonate, fractured rock
like crystalline basement complex and metamorphic rocks. Extensive, ancient volcanic terrains
are also included as a special case in this discussion.1 For example, in western India, hundreds of
nearly horizontal, basaltic lava flows form a thick pile that is known as the Deccan trap, and
covers around 500,000 km2. This sequence was not tectonically disturbed after consolidation and
due to the non-frothy nature of the lava a hand specimen does not show any primary porosity.
Due to the great thickness of the lava flows, it is not possible to drill through them to obtain
groundwater from the underlying strata. Hydrogeologically, the Deccan trap, therefore, acts as a
fractured basement complex.
The most significant features of hard rock aquifers are as follows:
(i) A topographical basin or sub-basin generally coincides with the groundwater basin,
the groundwater resources tend to be concentrated towards the central valley
portion, closer to the main stream and its tributaries.
(ii) The depth of groundwater occurrence, in useful quantities, is usually limited to a
hundred metres or so.
(iii) The aquifer parameters like Storativity (S) and Transmissivity (T) often show erratic
variations within small distances. The annual fluctuation in the value of transmis-
sivity is considerable, due to the change in saturated thickness of the aquifer from
the wet season to the dry season. When different formulae are applied to pump test
data from one bore well, a wide range of S and T values may be obtained. The
applicability of mathematical modelling is limited to only a few simpler situations.
(iv) The saturated portion of the mantle of weathered rock or alluvium or laterite over-
lying the hard fractured rock often makes a significant contribution to the yield
obtained from a dug well or bore well.
(v) Only a modest quantity of groundwater, in the range of 1.0 to 100 m3 per day, is
available at one spot. Drawdown in a pumping dug well or bore well is often
almost equal to the saturated thickness of the aquifer.
(vi) For sustainable groundwater development, it is advisable to take a sub-basin as a
unit and to plan for conjunctive use of surface water and groundwater. It is also
advisable to adopt soil and water conservation techniques for augmenting recharge
1. For a detailed discussion of the hydrogeology of volcanic rocks see Chapter 16.

383
Groundwater studies
to groundwater body, so as to minimise the adverse effects of increased pumpage

due to groundwater development. In many sub-basins underlain by fractured rock
aquifers, groundwater has a residence time of a few years only and the system is
very sensitive to droughts.
In hard rock aquifer areas, groundwater development has always played a secondary role com-
pared to that in the areas having high-yielding unconsolidated or semi-consolidated sediments
and carbonate rocks. This is because of the relatively poor groundwater resources in hard rocks,
the low specific capacity of wells, erratic variations and discontinuities in the aquifer properties,
and difficulties in exploration and assessment of the resource.
It should, however, be remembered that for the millions of farmers in developing
countries, with small farms in fractured basement or basaltic terrain, whatever small supply is
available from these poor aquifers is their only hope for upgrading their standard of living by
growing irrigated crops or by buffeting their rain-fed crops against the vagaries of rainfall. It is
also their only source for drinking water for the family and cattle. In many developing countries,
hard rock hydrogeologists have an important role to play. In the developed countries, interest in
hard rock hydrogeology, other than in relation to drinking water supplies to small communities,
has increased recently because of the prospects of using these low permeability rocks for the
storage of hazardous nuclear and chemical wastes. The study of groundwater flow through
faults, fissures and fractures is also of interest to scientists studying the migration of con-
taminants and to engineers engaged in tunnelling through hard rocks for water supply and
highway construction projects.
Hard rock hydrogeologists are therefore divided into two main groups: those interested
in obtaining groundwater for domestic, irrigation or industrial use by exploring fractured and
permeable zones in a relatively less permeable matrix of hard rock and those interested in
locating impermeable or very low permeability zones for storage of hazardous waste. Ironically,
for the first group even the most permeable zones are often not good enough to yield an
adequate water supply, while for the second group even the least permeable zones are often not
good enough for safe storage of hazardous waste over a prolonged period.
15.2 Occurrence of groundwater
Groundwater under phreatic condition occurs in the mantle of weathered rock, alluvium and
laterite overlying the hard rock, while within the fissures, fractures, cracks, joints and lava flow
junctions within the hard rock, groundwater is mostly in a semi-confined state. Compared to the
volume of water stored under semi-confined conditions within the body of the hard rock, the
storage in the overlying phreatic aquifer is often much greater. In such cases, the network of
fissures and fractures serves as a permeable conduit feeding this water to the well.
The recharge to groundwater takes place during the rainy season through direct
infiltration into the soft mantle overlying the hard rock and also into the exposed portions of the
network of fissures and fractures. In India and other Asian countries, the ratio of recharge to
rainfall in hard rock terrain is usually between 3 and 15%, depending upon the amount and
nature of precipitation, the nature and thickness of topsoil and weathered zone, and the topo-
graphical features of the sub-basin. Groundwater flow rarely occurs across the topographical
water divides; for planning the development of groundwater resources each basin or sub-basin
can be treated as a separate hydrogeological unit. After the rainy season, the fully recharged
hard rock aquifer gradually loses its storage, mainly due to pumpage and effluent drainage by
streams and rivers. The dry season flow of the streams is thus supported by groundwater out-
flow. The flow of groundwater is from the peripheral portions of a sub-basin to the central valley
portion, thereby causing dewatering of portions closer to the topographical groundwater

384
Hydrogeology of hard rocks
divides. In many cases, dug wells and bore wells yielding a perennial supply of groundwater
can only be located in the central valley portion,
The average residence time in a sub-basin of about 100 km2 is up to five years. The annual
recharge is thus a sizeable part of the total resource of an aquifer and the whole system is very
sensitive to the availability of recharge during the rainy season. Two drought years in succession
can pose a serious problem. Under such conditions the low permeability of hard rock aquifers is
a redeeming feature, because it makes small quantities of water available, at least for drinking
purposes, in the dug wells or bore wells in the central valley portion of a sub-basin. If the hard
rocks have a very high permeability, the groundwater body will move quickly towards the main
river basin, thereby leaving sub-basins high and dry. The low permeability in the range of 0.05 to
1.0 m/d thus helps in retarding the outflow and regulating the availability of water in individual
wells. More farmers are able to dig or drill their wells and irrigate small plots of land without
causing harmful mutual interference.
15.3 Groundwater development
Development of a natural resource like groundwater towards its optimum utilisation will benefit
mankind. In the highly populated but economically backward areas in hard rock terrain, many
governments in the developing countries have taken up schemes to encourage small farmers to
dig/drill wells for irrigation. This is especially true for the semi-arid regions where surface water
resources are meagre. For example, in peninsular India, hard rocks such as granite, gneiss, schist,
quartzite and basalts (Deccan traps) occupy about 1.15 million km2 area, out of which about 40%
is in the semi-arid zone, receiving less than 750 mm rainfall per year. Over 3.5 million dug wells
and bore wells are being used in the semi-arid region for irrigating small farm plots and pro-
viding domestic water supply.
Development of groundwater resources for irrigation and domestic use is thus a key
factor in the economy of vast stretches of semi-arid, hard rock areas. The basic need of millions
of farmers in such areas is to obtain an assured irrigation supply for at least one crop per year
and to have a protected, perennial drinking water supply within a reasonable walking distance.
The hard-rock hydrogeologists in many developing countries have to meet this challenge to
impart social and economic stability to the rural population, which would otherwise migrate to
the neighbouring cities. Exploration and assessment of groundwater, finding suitable sites for
locating dug wells and bore wells and planning for long term sustainability of the wells, are the
main tasks under these circumstances.
For promoting large-scale irrigation development from groundwater resources, institu-
tional finance has to be made available to the farmers at concessional rates of interest, for
digging or drilling wells. Before selecting an area covering several sub-basins, for financing
irrigation wells, the primary requirement is the assessment of the available groundwater. In
the absence of reliable data on rainfall, evapotranspiration, flood discharge of a stream and
pumpage from existing wells in the sub-basin drained by the stream, a fair estimate of ground-
water resource available for new development can be made by assessing the dry season flow
and underflow of the stream. This base flow is supported by the effluent drainage from ground-
water in the sub-basin and a part of it can be tapped by the proposed new wells. Once the
technical feasibility is ascertained, the financing institutions or banks need to estimate the econ-
omic viability of an average individual well, based on the average yield of the well in different
cropping seasons, supporting a suitable cropping pattern. The total incremental income from
such a cropping pattern should enable the farmer to repay the loan with interest, over an
average period of 7 to 9 years. Some banks also make provision for insurance of failed wells.
Groundwater development in a sub-basin results in increased pumpage due to the new

385
Groundwater studies
wells. This lowers the water table and hence reduces the effluent drainage from the sub-basin.
Such development in several sub-basins draining into a main river basin reduces the surface
flow and the underflow of the river, thereby affecting some of the surface water schemes
dependent on the river flow. In order to minimise such an interference, it is advisable to apply
water conservation and recharge augmentation techniques in the sub-basins, simultaneously
with the groundwater utilisation programmes.
15.4 Types of wells
It is common to use dug wells and dug-cum-bored wells to obtain shallow, phreatic ground-
water in low permeability hard rock aquifers. A dug well offers more area than a tubewell for
inflow of groundwater into it and also provides sizeable storage for accumulating water. The
volume of rock excavated while constructing a dug well is often much more than the REV
(Representative Elemental Volume) of the fractured rock aquifer and there are reasonable
chances of tapping water-bearing vertical and horizontal fractures and fissures during the
excavation, especially if the site for the well is carefully selected. Otherwise, these fractures may
be dry or may bear water for only a few months of the year.
A typical dug well has a diameter of 3 to 8 metres and depth of 5 to 15 metres. It usually
penetrates through the weathered mantle and goes a few metres into the underlying fractured
rock. A retaining wall of stone or brick masonry is required to support the upper portion of a
dug well excavated in soil, sub-soil and highly weathered rock overlying hard rock. The
masonry wall is provided with weep holes, especially near its base to allow inflow of phreatic
water into the well. In areas where hard rock is covered by laterite or a weathered zone rich in
lime nodules, a retaining wall may not be necessary to support the excavation. Some of the
deeper dug wells are dug through the weathered mantle and the fissured rock, into the massive
hard rock below, so as to provide adequate storage for water flowing into the well during a
period of 8 to 10 hours. This allows the farmer to conveniently pump the stored water, once in
the morning and once in the evening, by using a low-cost centrifugal pump set. It is customary
to provide two or three platforms at different levels inside the well, so that the pump set can be
installed according to the water levels in different seasons. The foot valve of the pump is kept in
a small pit excavated in the well bottom, so as to facilitate emptying of the well. On shallow
wells, animal-drawn devices or human-powered water wheels can also be employed to lift
water for irrigation of small plots.
Horizontal bores can be drilled radially outward from a dug well, at various levels below
the water table, so as to increase well yield. Horizontal bores of 25 mm to 75 mm diameter and
of lengths up to 10 to 15 metres are drilled in laterite, alluvial cover and weathered rock
overlying the fractured hard rock, manually or with a drilling machine. In very soft strata, such
bores have to be supported by inserting slotted casing pipes. Such bores increase the effective
diameter of the dug well, without a proportional increase in the cost. Horizontal boring in
fractured hard rock is done to tap more water from the neighbouring vertical fissures and
fractures. Such bores usually extend radially outward from the wet patches or mouths of springs
issuing into the dug well from the fractured strata.
Vertical bores of 100 mm to 150 mm diameter and depths up to 50 to 60 metres are drilled
in the bottom of the dug wells to tap semi-confined water in the deeper horizontal or sub-
horizontal fractures, joints or flow junctions. Water from such bores rises over the bottom of the
dug well and accumulates there if the bottom of the dug well is at least 5 metres below the water
table. A centrifugal pump installed on the dug well can then pump this water out. If water from
the vertical bore does not rise into the well, a separate pump is required. A submersible pump
can be installed if the yield from the bore is above 4 to 5 m3 per hour, and the water can be

386
directly taken for irrigation, rather than stored in the dug well. Submersible pumps being
expensive, low water level guards are installed to stop the pump operating when the water
level approaches the top of the pump. This prevents dry running of the pump. The pump can be
restarted by using a timer switch on the control panel or by using a high water level sensor in
the bore. If the yield is small, a Jet or Ejecto pump is used to pump out water from the bore. The
water is stored in the dug well until a sizable quantity has accumulated in 8 to 10 hours and
then pumped out in 2 to 3 hours by the centrifugal pump installed on the dug well. In this case,
a casing pipe of about 3 to 4 metres length is taken and half of it is inserted into the bore, with
the remaining half sticking out over the well bottom. Cement concrete is put around this pipe so
that the water accumulating in the dug well does not flow back into the bore.
Figures 15.1(a) and (b) illustrate dug wells with horizontal and vertical bores respectively,
in horizontally and vertically fractured strata. In the first figure, the horizontal bores in the dug
well are successful because they intersect a vertical fissure, however the vertical bore has failed.
In the second figure the vertical bore in the dug well is successful but horizontal bores have
failed. Figure 15.1(c) shows two wells on either sides of a stream in metamorphic strata of alter-
nating bands of low permeability schist and more permeable fractured quartzite. Well A, located
on the down dip side from the stream, gives some water from the phreatic aquifer in weathered
rock and also has a successful vertical bore in its bottom. Well B is a relative failure.
It will be clear from the figures that in areas where vertical fractures, fissures and joints
occur in the vicinity of a dug well, horizontal bores will be more successful, while in areas where
horizontal or sub-horizontal joints, bedding planes, cleavages, flow junctions or fractures occur
below the bottom of dug wells, vertical bores will be more successful.
The cost of a dug well of about 6 m diameter and of 15 m depth in India and other Asian
countries like Nepal, Pakistan and Sri Lanka, would be around US$3,000. In African countries
like Kenya or Zimbabwe, the cost would be almost double due to scarcity of trained labourers.
In the soft strata near the ground surface, the excavation is cheaper but a masonry wall has to be
built to support the strata. In the underlying hard fractured rock, the excavation is expensive
because dynamite blasting is often necessary. However, this portion of the dug well does not
need support. When blasting in the fractured rock for well excavation below the water table
level, it is advisable to blast only one or two holes at a time, instead of simultaneously blasting
all the 40 or 50 holes drilled in the bottom of the excavation. If all the holes are simultaneously
blasted, the fracture network often receives a heavy shock and the yield of water into the dug
well may be reduced. The cost of drilling horizontal bores is around US$2 to 3 per metre; in
India vertical bores cost around US$4 to 5 per metre.
Other types of dug wells include large diameter (10 to 20 metres) shallow dug wells of
4 to 5 metres depth, excavated in coastal areas for skimming fresh water floating over saline
water and rectangular trench wells of 1 to 2 metres width and 10 to 50 metres length excavated
at right angles to the direction of groundwater flow. Such a trench well dug across the bed of a
stream or river, at a carefully selected location, taps a large quantity of the underflow, even after
the stream dries up. It is customary to back-fill the excavation, creating an inverted filter, with
boulders in the bottom, gravel in the middle and coarse sand in the upper portion. The porous
and permeable trench thus created across the river bed is connected to a small, circular dug well
or jack well, on the river or stream bank.
The construction of a dug well or a dug-cum-bore well is expensive and time consuming.
Down-the-hole-hammer rigs have provided an alternative that enables a surface bore well of
150 mm diameter and up to 100 metres depth to be drilled in one day. The cost in India is
around US$600. The upper portion of the bore well in soft mantle and weathered rock has to be
supported with steel or PVC casing pipe. Unless this casing pipe has slots near its lower end to
allow inflow of water from the phreatic aquifer, the bore well yields all of its supply from the
network of fractures and fissures in the hard rock. However, this network is hydraulically

387
Groundwater studies
Figure 15.1 The well in (a) is mainly filled by flow from the horizontal bores, in (b) the well is mainly filled
by flow from the vertical bore and (c) shows two wells of which well A is successful and B
is a relative failure
connected to the phreatic aquifer, and pumpage from the bore well results in elastic compression
of the network plus dewatering of the phreatic aquifer.
Experience in Afro-Asian countries indicates that most of the bore wells in fractured rock
terrain meet their supply within the first 50 metres. In some cases the supply increases by
drilling deeper, up to 100 metres. Very few bores, less than one percent in India, tap water-
bearing fissures or fractures below 100 metres depth. However, some of these deep bores do
come into the category of high-yielding bores.
In peninsular India, basement complex and basalt (Deccan trap) cover an area of around
1.15 million km2. Most of the bore wells drilled here fall in the category of low-yielding bores,
yielding up to 2 m3 per hour. In basaltic terrain there is some correlation between shallow water
table and good yield of the bore wells but such a correlation is not observed in granitic terrain.

388
Revitalisation of low-yielding bores is a topic of great interest to hydrogeologists and farmers. If

the low yield is due to some fissures becoming clogged with rock dust during the drilling
operation, the yield gradually improves after using the bore for a few years. Immediately after
the rainy season, when the water table is high, the bore well is pumped to create a large draw-
down. The clogged fissures get flushed when water rushes into the bore under a greater head
difference than normally encountered. In some cases, working a rubber rimmed plunger up and
down like a piston, below the water table in the bore well, also gives beneficial results. Blasting
up to 2 kg of dynamite charge inside a low-yielding bore well is the last resort. The charge may
be shrouded in coarse sand. The bore well is filled with water up to the brim and the top of
casing pipe is plugged with a cotton ball. Blasting helps in some cases where the fissures, that
have opened up or are newly formed by blasting connect the bore well to more permeable
network of fissures in the vicinity. In some other cases the bore well collapses and is rendered
useless. In the remaining cases, blasting has no effect on the yield.
Hydrofracturing of low-yielding bore wells is a relatively new technique and requires
expensive equipment for injecting large volumes of water under high pressure into the bore, so
as to jack up the existing low permeability fractures met within the bore well. Some of these
fractures become extended and connect with the network of more permeable fractures if there is
such a network in the vicinity. Hydrofracturing is commonly used in United States of America,
Australia and South Africa. In South Africa, flow rates of injection water in a successful oper-
ation are more than 15 l/s at a pressure of 80 bars. (Less et al., 1993). It has been observed that
hydrofracturing is more successful at sites selected on the basis of lineament mapping and
geophysical exploration. At such sites the initial low yield is due to the unfortunate fact that the
bore well has missed the main fracture network by a few metres, hence during hydrofracturing
its connection to the main network is established.
15.5 Drinking water supply
Villages and communities situated away from any surface water source depend on locally
available groundwater to meet their drinking water needs. Economically it is not possible for
them to bring water in pipeline from distant sources. In remote hilly areas in semi-arid hard
rock terrain, any small supplies of groundwater available from dug wells or bore wells are of
vital importance for survival, especially in summer months, when yields as low as 1 m3/d are
considered as very useful. Figure 15.2 shows the situation of a village located not far from a
steep incised plateau. The water table in the fractured and weathered rock overlying the hard
rock is high after the rainy season and the villagers get their drinking water supply from the
spring at the cliff or from the dug well. Towards summer, the water table gradually declines and
the spring dries or its yield is reduced to a wet patch, just supporting some local vegetation at
the mouth of the spring. The thickness of the saturated aquifer is negligible at the spring, but
away from the cliff face the saturated thickness gradually increases up to a metre or so and the
dug well is able to yield a small quantity of drinking water during summer. Due to the small
hydraulic gradient and the low permeability, the groundwater body does not get completely
drained and assumes a quasi-static stage towards the summer season. In some villages, only
the spring water is used as long as the spring runs, and the dug wells are brought into use only
after the spring dries at the beginning of the summer. The cone of depression thus starts devel-
oping around the dug well in summer, not earlier.
In some villages in hard rock terrain, bore wells do not yield perennial water supply, due
to the absence of deeper fissures and fractures. Dug wells of 6 to 8 metres depth also go dry at
the start of the summer, due to the small thickness of the weathered zone, rugged terrain and
limited rainfall. It is expensive to provide drinking water to such villages in the summer season

389
Groundwater studies
Figure 15.2 Effect of natural drainage on a village well
by tankers, especially in hilly terrain. In Maharashtra state in western India, experiments in

creating an artificial aquifer by blasting around the existing dug well, have been successfully
carried out. In these experiments about 80 to 120 bore holes of 100 mm diameter and 6 to
8 metres depth were drilled around the dug well up to about 60 metres distance and blasted
simultaneously, in order to create a jacket of fractured rock aquifer around the dug well. Such a
dug well could then provide a small quantity of drinking water, even in summer.
However, water from dug wells easily becomes polluted due to percolation from cattle
sheds, septic tanks, manure pits, garbage dumps and fertilisers used on farms. Dug wells
located a few hundred metres away and on the up-gradient side of a village have a better
quality of water. Biological contamination and high nitrate values appear in some cases. Villages
accessible to drilling rigs are therefore increasingly being supplied with drinking water from
bore wells. The commonly used norm in Asian countries is: one bore well of 150 mm diameter
and up to 60 to 100 metres depth, fitted with a hand pump, for providing water supply to a
population of 250 or less. Such a bore well is often the least cost solution for providing pro-
tected drinking water supply to rural communities, as the cost works out at only around
US$0.48 per person per year, even after allowing for interest, depreciation and maintenance at
16% on the initial capital cost of about US$ 750 (Limaye, 1995). The quality of water is usually
good if the bore well has a casing pipe up to about 6 metres depth and if the surroundings are
kept clean. In some granitic areas the problem of fluorosis is present but small-scale water treat-
ment plants have now been developed.
The experience from several Afro-Asian countries, in programmes undertaken by Govern-
ment agencies and by NGOs and voluntary organisations, for providing drinking water bore
wells in villages, indicates that the following factors are of crucial importance for successful
implementation:
(i) selection of a drilling site should be done in the presence of village leaders, after
careful hydrogeological and, if necessary, geophysical resistivity survey;
(ii) participation of the local rural community should be encouraged at all stages of the
programme, such as providing labourers during the preliminary survey, preparing
an access road for the drilling rig, providing food for the drilling crew, assisting in
the construction of a concrete slab around the bore well and looking after proper
operation and maintenance of the pump.

390
(iii) a facility which is available totally free often gets neglected. The villagers should
therefore, be required to contribute an amount between 5% to 20% of the cost of
bore well and pump. If this is not possible for poor communities, they should at
least volunteer their labour;
(vi) the village council should collect a monthly fee per family for maintenance of the
pumps;
(v) women must be involved in the management, through proper representation in the
Village council. Also, arrangements must be made to train some of the young men
and women in the village in hand-pump maintenance.
15.6 Exploration
Exploration for locating sites for well digging and drilling in hard rock terrain is vitally impor-
tant for successful completion of irrigation or drinking water supply projects, because hard rock
aquifers are not extensive and their properties vary in short distances. Basic exploration is done
by collecting topographical and geological maps, aereal photographs and satellite images, if
available, and by conducting a hydrogeological survey during which the following data and
information are collected:
(i) Inventory of existing wells. Their depth, diameter and yield and the type of strata
encountered. Elevation of the water table. Area irrigated by each well. Type of
pump and pumping schedule. Seasonal fluctuation of the water table.
(ii) Rainfall and drainage patterns.
(iii) Lithological units encountered in wells. Strike and dip in sedimentary strata. The
thickness of soft overburden and its relation to topography. The extent of fissuring
in hard rock and its relation to topography.
(iv) The sandy or rocky nature of the stream or river bed. Whether the stream is
seasonal or perennial. The prospects of attracting influent seepage from the stream
to a pumping well on the bank.
(v) Shifting and meandering of river. Erosional or depositional features on the river
bank. Evidence, if any, of rejuvenation.
(vi) Locations and discharge of natural springs, if any, in the area.
(vii) Locations of surface water reservoirs, if any, in the area. Possibility of receiving
recharge during the dry season from surface water reservoirs and/or the irrigation
canals leading off from the reservoir. The possibility of obtaining recharge from
deep percolation below the root zone in the irrigated area.
(viii) The occurrence of dykes, pegmatite veins, etc. in the area and their nature as
groundwater conduits or barriers. Whether there are any good wells upstream from
dykes. The direction of any preferred fracture orientation in the area as observed
from rock exposures and strata encountered in dug wells.
(ix) Correlation, if any, between the lineaments observed in air photos or satellite
images and the locations of successful wells in the area or patches of dense natural
vegetation in an otherwise sparsely vegetated landscape.
(x) Variations, if any, in the quality of groundwater along its general flow direction.
(xi) Whether there are any erratically successful or erratically failed wells, which do not
fit into the conceptual model of groundwater occurrence in the area. Such wells
indicate discontinuity and lateral variation in the aquifer.
Such observations and information are useful in delineating promising zones for groundwater
development in a sub-basin. Geophysical resistivity or electromagnetic surveys can then be
carried out in these zones for selection of suitable well sites. Given that there are lateral varia-

391
Groundwater studies
tions in the strata, Wenner profiling is more useful than Wenner or Schlumberger sounding.
Profiling is carried out with electrode spacings between 20 to 50 metres, to locate a fractured,
low resistivity zone in the hard rock covered by a soft mantle. To find out the fracture
orientation, azimuthal resistivity survey can be carried out over the low resistivity zone. In such
a survey, resistivity readings are taken around one central spot, with the same electrode
configuration but in different directions.
Even within a fractured zone, the intensity of fracturing, interconnections, apertures,
infilling matter and recharge from the overlying phreatic aquifer determine the quantity of water
available in a dug well or bore well. Tension fracture zones usually have a higher storage
capacity, while shear fracture zones could be tight or permeable, depending on their stage of
development (UNESCO, 1984).
15.7 Recharge augmentation
With the increase in pumpage due to new wells in a sub-basin, the water table falls and the
effluent drainage from the sub-basin is reduced. In the hard rock areas, where both the total
storage of groundwater and the average residence time are small, the system is much more
sensitive to variations in pumpage and recharge than a similar system in alluvial or carbonate
aquifer areas. As mentioned earlier, it is advisable to start soil and water conservation and
recharge augmentation activities concomitantly with groundwater development schemes. Some
of these activities, such as hill slope trenching, contour bunding, afforestation, gully plugging,
are useful in increasing the infiltration to a groundwater body during the rainy season. But the
geometrical factors of the sub-basin and the thickness and storativity of weathered rock and
fractured rock set a limit to the recharge that can be accepted in the rainy season.
Many developing countries have a well defined rainy season in the year which is fol-
lowed by a prolonged dry period. In such a climate, a sizable portion of the recharge received in
the rainy season may leave the sub-basins by way of effluent stream flow and groundwater out-
flow, within the first few months of the dry season. Water scarcity may thus occur towards the
later months of the dry period, which in the monsoon climate is the summer season. It is there-
fore necessary to undertake activities which would retard the groundwater outflow and which
would lead to recharge to the groundwater body during the earlier months of the dry season.
Construction of underground impermeable bunds (low dikes) across stream beds is a
useful technique in semi-arid regions for retarding the groundwater outflow. Construction of a
percolation tank by putting an earthen bund with side waste weirs across a stream is also very
useful, because during the dry season the water stored behind the bund during the rainy season
gradually percolates to recharge the groundwater body. A typical percolation tank has a bund
of 8 to 10 metres height and a catchment area of about 20 to 50 km2. Ideally, the water stored
in the tank should percolate within first 3 to 4 months of the dry season so that the shallow
water body is not exposed to excessive evaporation rates in summer months. In western India,
thousands of percolation tanks have been constructed in semi- arid regions to augment recharge
after the rainy season is over. In this drought-prone area, construction of a percolation tank is
also a relief measure preferred by the government authorities, because it provides employment
to about 2,000 people for 3 to 4 months in a drought year. The percolation tank becomes oper-
ative from the next years rainy season.
The storage efficiency of a percolation tank is the ratio of the volume of water stored in
the tank at the end of rainy season to the volume of runoff water available from the catchment.
The percolation efficiency is the ratio of the volume of water percolated to the volume of water
stored. The overall efficiency is the product of the above two efficiencies and is around 40 to 70%
for percolation tanks constructed at technically suitable locations (Limaye and Limaye, 1985).

392
Occurrence of exposed hard rock in the tank bed impedes percolation. Similarly, silting in the
tank bed over the years reduces both storage and percolation efficiencies. Regular desilting is
therefore necessary. A percolation tank and a few smaller stream bunds and underground bunds
is an ideal combination for augmenting recharge in a sub-basin.
An additional benefit of percolation tanks in coastal areas is to improve the quality of
water in drinking water wells excavated on the downstream side of the tank. In granitic areas,
where a high fluoride content in groundwater is a problem, dug wells near the percolation tanks
show much less fluoride and are preferred for obtaining drinking water supply.
Recently, a novel experiment in recharge augmentation was implemented in semi-arid,
basaltic terrain in western India, by a voluntary organisation. In this experiment, surface water
flowing in effluent streams during the early part of the dry season was lifted by the farmers,
using their pump sets and was delivered into several dug wells in each stream basin. About
100,000 dug wells were thus charged with water that would have left the area as surface flow in
a few weeks after the rains. The residence time was prolonged to more than a few months. The
beneficial effects of this experiment were felt by the farmers during the summer season. The
wells which used to dry up in summer started to yield small supplies of drinking water for the
family and the cattle. Wells which used to give a small supply in summer started supporting
irrigation of summer vegetables on small plots of land.
15.8 Sustainability and pumpage control
Sustainable development is achieved when the quality and quantity of water available from the
wells remains unaffected over many years. In hard rock areas, groundwater or the resource itself
is modest in quantity, erratic in occurrence and sensitive to changes in pumpage and recharge. It
is therefore not easy to ensure sustainability of all the wells over decades because new wells are
constructed each year and the pumpage in a sub-basin increases year by year. Some of the sub-
basins become over-developed and the yields from the wells decline due to mutual interference
and general depletion leading to a permanent decline of the water table. Under such conditions
the drinking water supply wells need to be protected first. This can be achieved either by pre-
venting the construction of new wells within a specified distance from a drinking water supply
well or by giving the government authorities a right to acquire any private well in the vicinity,
after paying due compensation, to provide water supply to the village.
However, digging and drilling of new wells cannot be stopped in a sub-basin to protect
the yields of the existing, private irrigation wells, because this would amount to of a scarce
resource being captured and others being permanently denied an equitable share of the
resource. Recharge augmentation efforts should receive priority in over-developed sub-basins, in
co-operation with the farmers, their cooperative societies and financing institutions. It is also
desirable to incorporate voluntary agencies and NGOs in these efforts and through them to
advise the farmers to reduce wastage of water. If these efforts are not enough, pumpage control
may have to be imposed.
Pumpage control is a negative way of management but when it has to be imposed, it
should only be through mutual monitoring by farmers or by local council at the village level.
The concept of sustainability in such a case may consider a period of only about 7 to 9 years in
which a farmer usually recovers his investment in constructing the well. The owners of high-
yielding wells should be the first to cut down on their pumpage and pump only an equitable
share, so that other farmers may also dig/drill new wells. This is more likely to be successful
through persuasion and social pressure rather than through any rigid legislation. In complex,
anisotropic and discontinuous hard rock aquifers, any rigid legislation is technically unsound
and may merely lead to endless and futile court battles.

393
Groundwater studies
LESS, C.; ANDERSEN, N., 1994. Hydrofracture; State of the Art in South Africa. Applied Hydro-
geology, Vol. 2, pp. 5963.
LIMAYE, S. D., 1995. Groundwater Management in Semi-Arid, Hard Rock Terrain in India, (Keynote
Address). Proc. 2nd South Asia Geological Congress, Geological Society of Sri Lanka,
Colombo.
, 2000. Some Aspects of Sustainable Groundwater Development. Abstract, Special Symposium
B-7, Proc. 31st International Geological Congress, Rio de Janeiro, Brazil.
LIMAYE, S. D.; LIMAYE, D. G., 1985. Groundwater Development in Deccan Traps of Western India.
In: Proc. International Workshop on Rural Hydrogeology and Hydraulics in Fissured Basement
Zones. Dept. of Earth Sciences, University of Roorkee.
UNESCO, 1984. Groundwater in Hard Rocks. Studies and Reports in Hydrogeology 33, UNESCO,
Paris.

394
16 H y d ro g e o l o g y o f v o l c a n i c ro c k s
16.1 Introduction
Volcanic rocks constitute thick and widespread formations in many areas of the world such as
Eastern Siberia, the Ural mountains, the Deccan in India, the Paran basin in eastern South
America, the Permian basin of Siberia or several areas in central western United States. In large
regions of such areas, as well as in many archipelagos and islands, such as the Canaries, Azores,
Iceland, Hawaii and Runion, volcanic rocks are the dominant or only significant geological
formation around and most of the available groundwater resources are held in them. In other
cases, volcanics are simply one of the local geological formations, below, interlayered or topping
other geological formations, as in Sicily, or in many cases in Mexico, Central America and
the Andes. Their hydrogeological role as aquifers, aquitards or even aquicludes depend on the
relative water-bearing and hydraulic properties and location of the volcanic rocks with respect
to the other formations.
Various books deal with volcanic rocks (e.g. MacDougall, 1988; Middlemost, 1985; Fisher
and Schminke, 1985; Araa and Lpez-Ruiz, 1974; Araa and Ortz, 1984; Francis, 1993) and how
they are related to global tectonics (Smellie, 1994). More general descriptions can be found in
standard geology books (e.g. Duff, 1993; Aubouin et al., 1981).
Specialised literature exists on hydrogeology of volcanic rocks is scarce, although the topic
is explicitly considered in some widely used textbooks. Chapters on volcanic rock hydrogeology
can be found in Davis and De Wiest (1966) and Custodio and Llamas (1983). Details on coastal
and island volcanic aquifers can be found in Falkland and Custodio (1991). A specialised report
(Custodio, 1978) largely summarises and enlarges the contents of the UNESCO sponsored
International Symposium on Volcanic Rocks and Islands Hydrogeology, which took place in
1974 in Arrecife de Lanzarote (published by CEDEX, 1987), and comments on the experience up
to that time in the Canary Islands, including the UNDP-Government of Spain project on
groundwater resources in the Canaries.
16.2 Volcanic rocks and formations
Volcanic rocks cover a wide range of chemical and mineralogical compositions. Generally,
volcanic rocks are basic, with a low silica content (basalt and basalt-like rocks), some are rich in
plagioclase (tholeiites) and others may be rich in olivine (alkaline basalts). A high silica content
is found in acidic rocks, such as rhyolites, and andesites and phonolites have an intermediate
silica content.

395
Groundwater studies
Molten rock, called magma, includes variable quantities of dissolved gases (mainly CO2)
and water vapour. Magmas may be fully melted matter or a mix of melt and crystals. Basic
magmas are hot (about 1100C), present a relatively low viscosity and typically have a low
dissolved gas content. Acidic magmas, which are not as hot (900C), are much more viscous
and often have a higher dissolved gas content. Magmas may vary from molten earth mantle
material, more or less chemically differentiated due to partial crystallisation and separation of
the denser constituents before being ejected, to melted earth crust that has been carried down to
great depths in plate subduction zones. The latter are acidic, and the former are dominantly
basic. Variable differentiation and mixing of rising deep magmas with melted crust produce a
wide range of magma and volcanic rock compositions, with conspicuous changes in space and
along time.
The deep melt is lighter than the surrounding rocks and slowly ascends by buoyancy.
Most often it accumulates in large magma chambers, some of them relatively shallow (a few km
deep), from which the magma flows out through tension cracks (fissures). Generally only a
fraction of the magma contained in the chamber flows out, the rest remains in place. After
cooling, it may leave a coarse-grained rock mass or it may be intruded through enlarged fissures
into the surrounding rock as dykes, sills, or cone sheets. The surrounding rock, which may or
may not be of volcanic origin, may be deeply altered (metamorphosed) by heat and/or hot gases
and water, and sometimes intensively intruded by dykes. There are alteration products that
correspond to a wide range of underground conditions, from hot to cold ones.
The magma may escape from the chamber through fissures but the flow often con-
centrates in short sections when approaching the land surface and pours out at discrete points as
volcanoes. It seems that the large outflows that formed the large volcanic plateaus, such as the
Deccan in India or the Paran in eastern South America, were intense fissural emissions, not
devoid of point outflows. Important recent volcanic episodes show volcanoes along lines repre-
senting the main regional fissures, as in the seventeenth and eighteenth century major effusions
of Lanzarote, Canary islands. In other cases, cylindrical volcanic vents can be seen after erosion
has exposed them, as near Stuttgart, Germany. But interpretations differ from researcher to
researcher. Magmas poured out in rift zones and into transform faults of sliding crustal plates
are dominantly basaltic, whilst those poured out in plate subduction areas are dominantly of
the acidic type. Basaltic magmas form the extensive plateaus of Deccan and Paran, and others
in Siberia, South Africa and Central Europe, as well as the large oceanic ridges, whose top
emerge as islands such as Iceland, Azores, Ascension, Hawaii or Tahiti. Other dominantly
basaltic islands such as the Canaries and Madeira seem to correspond with transform faults.
Acidic magmas form large masses along the western part of North, Central and South America,
and around the Mediterranean sea, and also form long island arcs such as the Kuriles, Nippon
and Philippine islands, and dominate in Java and Sumatra (Indonesia). The upwelling of mantle
magma plumes, sustained by buoyancy, persists through geological times and leaves a trace on
top of the plates. The age of the volcanic deposits increases as one moves away from the present
point of eruption.
Submarine volcanism is more extensive than continental volcanism, but is of little hydro-
geological interest since the formations are mostly impermeable due to intense alteration. They
appear on continental land masses as relatively old, compact materials after having been
tectonically uplifted, such as some important volcanic deposits dating from the Permian and the
Keuper.
The build up of a volcanic oceanic island is rather complex (Carracedo et al., 1997). First a
submarine edifice rises from the ocean bottom and comprises extruded and intruded materials.
Later on, when the submarine edifice attains the sea surface eruptions are explosive and produce
loose, fragmented materials which are easily transported downslope by shallow water erosion. If
volcanism persists or the materials are rised, finally an island is formed, with an erosive surface

396
Hydrogeology of volcanic rocks
and the possible outcrop of the resistant inner core of intrusive bodies and densely dyke injected
volcanics. Afterwards subareal volcanics pile up and form the island. Vertical movements may
be relatively important in islands due to regional tectonics (Watts et al., 1997) and the bulging
due to magma chamber emplacement. Thus, relatively young submarine volcanic rocks and
intrusive bodies may crop out, as in some of the Canary islands volcanic complex formations.
The sea floor on which these islands lie is about 3,000 to 4,000 m below sea level; Miocene sub-
marine volcanics may be found 1,000 m above present sea level, while, due to these important
vertical tectonic movements subaerial volcanic rocks have been found at a depth of 2,000 m
below sea level.
The magma that pours out through the volcanic vents suffers a loss of pressure that
allows dissolved gases and water vapour to exsolve. The molten, more or less degassed rock
forms the lava, which spreads away from the outflow area. Gases may exsolve quietly or
violently in the form of bursts, in which part or much of the magma is fragmented and thrown
into the air, forming pyroclasts (tephra) varying in size from coarse blocks to fine dust (ashes or
cinder). Pyroclasts may be deposited as cooled fragments which remain loose, or as hot frag-
ments which may be welded to some extent. The block, gravel and sand size pyroclasts (lapilli
or picon) pile up around the outflow area, forming a volcano, which is a symmetrical or
deformed cone, depending on wind strength and land slope. The volcano may include lavas and
intruded material. It is not rare for lavas to outflow at a lower altitude instead of from the crater-
like summit of the volcano. Large gas outbursts and explosions produce large quantities of ash
which the wind may transport far away. Volcanism in the Andes has supplied large quantities of
volcanic dust that is now included in the sediments that cover large areas in Argentina and
Uruguay, and plays a significant role in groundwater chemistry. The finest particles produced in
large explosions may reach the stratosphere and extend over a large region and even the entire
earth. The change in atmospheric dust content and the gases released during large volcanic
events may effect the earths climate for some time (Robock, 2000). The sudden outpouring of
gas-rich magmas, especially of the acidic and intermediate type, in the response to a rapid loss
of pressure may produce a fluidised layer of dust and fragments suspended in the very hot gas
which is able to move easily down slope until it settles down and cools off. This forms the ash
flow (tuff) deposits or ignimbrites that differ from the ash fall formations that are formed in a
cool environment. After cooling, ignimbrites look like lava, especially when alteration and
weathering obscure their texture. Large explosive events may also produce extensive and thick
formations of volcanic conglomerates. The so called Roque Nublo formation once covered most
of the island of Gran Canaria and filled previous relief.
In the following sections subaerial (continental and island) volcanism will be considered,
focussing on the relatively young formations of Quaternary, Tertiary and in some cases Mesozoic
age, because of their hydrogeological importance.
From the hydrogeological point of view, the most important volcanic formations are lavas,
pyroclastics, ignimbrites and dykes. Dykes are more numerous deeper and closer to the magma
chambers. When considering lava and ignimbrite flows, it is important to distinguish between
the dense, slowly cooling inner part and the highly vacuolar and brecciated top and bottom
(scoria) of a flow. Some lava flows are almost devoid of top scoria (pahoehoe type) but others
are formed almost entirely by scoria and blocks (aa type). This depends on viscosity, cooling
speed and gas content.
Fluid lavas spread laterally, forming discrete lava flows on sloping ground, and when the
land is flat, collect into large lava fields. In this way hundreds of square metres of lava sheets,
commonly 1 to 10 m thick, accumulate. In Lanzarote (Canary islands) in the early seventeenth
century, a 6-year-long series of eruptions covered more than 200 km2 and deposited up to 300 m
of volcanics in some areas. On the same island subaerial volcanics may be up to 2,500 m thick.
Viscous lava tends to pile up around the outflow point. In some cases only a plug of semi-solid

397
Groundwater studies
hot rock has extruded, accompanied by sudden outbursts of ash flows. In the Paran basin,
major basaltic outflows covered thousands of km2 of land with very extensive pyroclast-free lava
sheets that are in some cases up to 1,000 m thick.
Near the volcanoes the ratio of lavas to pyroclasts is highly variable, from high in some
basaltic shield volcanoes, such as those in Hawaii, to very low in explosive eruptions, mainly of
the acidic type, but also of the basic eruptions when groundwater increased the gas content of
upflowing magma (phreatovolcanoes). Explosions not only form large quantities of pyroclasts
but also take broken parts (xenoliths) of deep volcanic or non volcanic formations to the surface.
Volcanoes commonly have steep sides, close to the geotechnical slope equilibrium. Thus
erosion, earthquakes, uplifting due to deep magma emplacement and volcanic explosions may
easily start large landslides that leave deep scars (Elworth et al., 1999). This is well known in
Hawaii and in the Canaries. The pressure and friction of the sliding mass may form a layer of
finely crushed and altered material, sometimes rich in bentonite, called mortalon in the
Canaries. Such layers are poorly permeable and difficult to drill, and may be regional hydro-
geological boundaries, although they are still poorly known. Large mudflows (lahars) can form
when the hot gases and pyroclasts cause the fast melting of snow on top of high volcanoes
(Nevado de Ruiz, Colombia; Mount Saint Helene, NW United States of America) or when the
explosions are accompanied by intense rain storms and lake water displacement, as in Java. The
sudden loss of weight after a major landslide may reactivate volcanism from still hot deep
magma chambers, as happened on some of the Canary islands (Tenerife, Hierro).
Dykes often present one or two main orientations. They tend to be vertical but actually
their thickness and slope accommodate the host rock conditions. The horizontal length of the
dykes varies from a few tens of metres to several km and the width varies from some cm to a
few metres. Close intrusive bodies and around long-lasting active calderas swarms of dykes are
common, and their volume may be larger there that of case rock. Some magma may be intruded
horizontally, interlayered between existing formations, forming a sill or horizontal dyke, a few
to some tens of metres thick and up to many km2 of surface area.
Over a long time (up to several million years) the slow cooling of magma chambers up to
many km3 in volume, releases large quantities of hot gases, mostly CO2 and water vapour,
which play an important role in groundwater chemistry and in modifying rock properties by
thermal metamorphism.
There is an important distinction between volcanic formations close to the eruption
points, where pyroclasts as well as other features that will be described below are found, and
volcanic formations far from these eruption points. In the former case coarse pyroclasts and
other features linked to the eruption area predominate. In the latter case the volcanic deposits
are mostly represented by the accumulation of gently sloping flows with occasional sediments
and weathered layers sandwiched in between. The most important of these formations are
basalts, such as those in the Paran and in the Deccan. These formations are also called traps,
derived from the Scandinavian word for stairs, referring to the appearance of these horizontal
layers after erosion.
Volcanism is a constructive (positive) agent, but erosion is a negative agent that tends to
destroy what has been formed. Volcanic formations are easily erodible. Steep slopes around
eruption areas are prone to fast destruction by gravity and running water, as well as by sea
abrasion in coastal areas. This is especially conspicuous on islands, in which building episodes
(eruptions) alternate with destructive (erosive) ones during calm stages. The oldest volcanic land
forms are more erosion-resistant than the more recent ones, because of the greater dyke density
and higher degree of consolidation (Fig. 16.1). The valleys and depressions of old landscapes
carved into relatively poorly permeable volcanics may later be fully or partly covered by new
outpourings and fallout (Fig. 16.2). Thus rock hydraulic properties change in both vertical and
horizontal directions. Figure 16.2 shows a large variety of structures that produce contrasting

398
Figure 16.1 Structure of volcanic formations affecting groundwater as exemplified by the Canary Islands,
Spain (Custodio, 1978)
The upper figure represents Fuerteventura which is a high, deeply dissected massif, with poor
permeability, densely intruded by dykes (Araa and Carracedo, 1979). It resists erosion due to the
massiveness of the old, mostly submarine volcanic complex. The remaining part of the island is formed
by permeable, recent volcanics extending seaward and filling the spaces between the dykes.
The lower figure represents Tenerife island in which the remnants of old (Miocene) deeply dissected
volcanic structures outcrop at the ends of three main fissured areas. At the confluence point young
volcanic deposits have covered part of the old formations and spread towards the sea, filling the
intermediate depressions. A similar structure can be found in Mauritius (Jawaheer et Proag, 1978) in
which a deeply eroded central volcano with a caldera was later filled by younger volcanics; the present
island is a plateau of recent lavas from which the old volcanics outcrop forming radial ridges.

399
Groundwater studies
Figure 16.2 Simplified cross section through island volcanic formations
A and B represent two cross sections from La Palma island (Canary Islands, Spain, from the Taburiente
caldera in the centre, towards the north (modified after the studies for the La Palma island water plan).
Resting on very poorly permeable, densely dyke intruded submarine and subaerial old volcanics (1),
there is a basaltic cover (2) flowing from the flanks of the of the Taburiente caldera, which is less
permeable than the younger basaltic flows (3) that fully cover it or leave only small windows. The
water table is inside formation (2) except near the top, where it is in formation (3). In case A, a small
part of the recharge is discharged through springs inside the caldera, the remainder going under-
ground towards the coast, if not tapped by wells or galleries. In case B, a large part of the recharge
outflows through large midslope springs. The water table slope decreases towards the coast due to
radial flow pattern, and when it is in formation (3). C shows the effect of large landslides (4) which
produce a low permeable breccia (5), which in Tenerife (Canary Islands) is locally called mortaln
(modified after the studies for the Tenerife island water plan). The water table attains high altitudes in
the low permeable, dyke-intruded and altered old materials (core), and groundwater mostly flows
through the thin overlying layer of younger volcanics.

400
hydraulic properties, including the interbedding of sediments and palaeosoils. They have to be
taken into account when dealing with groundwater occurrence and flow. Uplifting allows
erosion to exhume deep structures and even the underlying magmatic chambers and submarine
volcanic formations.
Large violent eruptions produce a fast reduction of pressure in the magma chamber and
may cause the roof of volcanic or sedimentary formations to collapse, producing a caldera
bounded by ring faults and crossed by radial faults. Afterwards sediments and volcanics from
subsequent activity tend to fill up the caldera. Figure 16.3 shows two examples.
Volcanism is associated with land elevation disturbances related to the emplacement of
magma (uplifting) and to the sinking after the eruptions, as well as to the general tectonics of the
area. Furthermore there are intrusions of dykes, landslides, explosions, cover collapse after
sudden outbursts and to relief internal pressure, differential compaction of collapsible porous
materials due to overburden weight, stress relief due to erosion and retraction due to cooling.
All this produces fissures, joints, cracks and discontinuities.
16.3 Hydrogeological properties of volcanic formations
Volcanic formations are highly variable and so are their hydrogeological properties, which cover
the full range of possible values found in natural formations. The scale of observation is
important. Heterogeneity is great at small-scale (say up to metres) but is smoothed at large-scale
(hundreds of metres) in large and extended formations.
Total porosity of volcanic formations is generally high, due to voids created by exsolved
gases and to the frequent scoriaceous and brecciated parts, as well as to their often clastic nature.
Values above 40 percent are not rare in lapilli, pumice and ash deposits. But compact lavas and
dykes have less than 5 percent total porosity. In some rocks most of this porosity is due to
closed, isolated voids, through which water cannot readily flow unless the rock becomes
weathered. The same can be said for closed-end pores. The result is that the porosity available
for groundwater flow is often much less than total porosity, especially in fresh (young) vol-
canics. Specific yield (drainable porosity) is small for fine-grained pyrocIastic formations, ashes
and weathered rocks. See Table 16.1 for some indicative values.
Permeability varies considerably, from almost zero (less than 0.0001 m/day) to more than
1,000 m/day. Massive lavas and welded ash flows (ignimbrites) present very low permeability,
while fresh, highly brecciated lavas and loose lapilli are very permeable. Table 16.2 shows some
representative values, but they have to be taken as indicative, since large variations can be
expected from site to site and inside a given formation; sometimes secondary permeability is
dominant.
The distribution of permeability and porosity is related to the characteristics of the rocks.
This is important when explaining the hydraulic properties and when choosing sites for drilling
wells to exploit groundwater resources.
Lava flows generally present a massive core with scoria and breccia layers at top and
bottom that are intrinsically more porous and permeable. Some lava flows are almost devoid of
top layers while others are dominantly brecciated from top to bottom. Some lavas are highly
vacuolar, which produces high porosity and often a large fraction of scoria and breccia, while
others are compact. A lava flow may lie directly on top of another lava flow that outpoured
shortly before, or rest on older materials covered by a weathered zone, even soil. In the latter
case the heat from the new lava flow thermally alters (bakes) the underlying materials, which
acquire a reddish colour (rubefaction). If a more or less developed soil existed, an often reddish,
brick-like layer of variable thickness is formed, commonly known as red layer or almagre.
Sometimes it may be traced for many km when exposed in cliffs that cut across the formation.

401
Groundwater studies
Figure 16.3 Schematic internal structure of a volcanic massif conditioning the permeability distribution
A. Caldera after the collapse of the magmatic chamber, following a violent episode in which
differentiated magma is outpoured. 1. Pre-volcanic formations (old volcanics or other formations);
2. volcanic formations from the magma chamber, generally of basic composition; 3. magma chamber
containing scarcely differentiated magma from a deeper chamber; 4. differentiated (more acidic)
magma in the top of the chamber, responsible for the violent volcanic episodes whose products;
5. partly fill the collapsed caldera; 6. extra-caldera volcanics thrown out during the violent episodes;
they may include lahars (mudflows produced by water) and debris from large landslides; 7. filling
of the caldera with erosion material and products of reactivated volcanism; 8. final stages of filling,
which may include water borne sediments from the sides. The figure may represent, with due
changes, the Caadas del Teide (Tenerife, Spain), the Crater Lake (Oregon, United States of America),
or Mount Etna (Sicily, Italy), which lies on top of folded and thrusted pre-volcanic formations.
B. System of various volcanoes, inspired by the Gran Sarcouy (Auvergne, France), but also
representative of the Gomera necks (Canary Islands, Spain). 1. basement, which may or may not be
volcanic; 2. volcanic landforms dating from before the violent episode; 3. new volcanics ejected when
the area collapsed, thus forming a caldera; 4. ejection of viscous, more acidic poorly permeable lavas
that fill the caldera and produce a tight plug; 5. post-eruption sediments or volcanics.

402
Table 16.1 Common porosity values of volcanic formations. The ranges are only indicative. Values are
derived from a wide literature review and the experience in the Canaries
Total Drainable
porosity porosity
Material (%) (%) Comments
Basalt flows 0.8 20 0.1 8 Dense to highly vacuolar

Basaltic interflows 20 50 5 15 Breccia at the top and bottom
of a lava flow
Basaltic formations 5 40 28 Increases with content of scoria
and pyroclasts
Basalt sheets (traps) 4 10 <1 2 Several flows, no pyroclasts,
moderately old
Core basalts 25 < 0.1 4 Thermally altered and dyke-
intruded
Conglomerates 2 25 14 Mostly basaltic
Loose pyroclasts 25 50 5 l0 Fresh lapilli and blocks
Ash fall 25 40 15 Relatively fresh
Phonolites 24 16 Dense flows
Phonolitic ignimbrites 20 60 0.5 8 Dense to poorly welded tuff
Pumices 50 85 < 0.1 1 Non-connected pores, unfractured
Obsidians ~ 0.5 < 0.1 Not a formation
Rhyolites 0.1 30 0.5 5 Dense to vacuolar
Rhyolitic ignimbrites 15 70 0.5 l0 Dense to poorly welted tuff
Rhyolitic interflows 30 70 15 Breccia at the bottom
of a lava flow
Lahars (mudflows) 20 35 < 0.5 Heterometric, variably
consolidated
Volcanic soils 40 60 <1 5 Variable
Note: Total porosity and drainable porosity decrease with age.
Its continuity depends on the shape of the previous land surface and soil cover, and the extent of
the baking lava flow. There are often changes of thickness and permeability, and discontinuities,
and it may be crossed by joints and dykes.
The inner part of ignimbrites (ash flows) tends to be compact and dense, like lavas, while
the outer parts become more porous and loosely consolidated towards the top and bottom, often
grading into breccias. Here some primary permeability may be found while the remaining rock
is almost impervious.
Both lavas and ignimbrites develop joints due to shrinking during cooling. The joints run
from top to bottom forming vertical faces in a hexagonal prism-like distribution, spaced ten to
hundreds of centimetres. There are usually also some horizontal joints. Generally joints are
deformed (curved joints, irregular patterns). These joints are primary tight.
When lavas, hot pyroclasts and volcanic tuff engulf tree trunks from pre-eruption forests,
what remains after the wood has been charred are tube-like cylindrical moulds.
In gently sloping land, when a lava flow stops moving because its surface has consoli-
dated, but the inner part is still fluid, sudden outbursts of the molten inner part may happen after
internal pressure increases. This forms a downflow extension of the lava flow, leaving behind an
elongated cavity below the hardened crust. Most often the roof collapses and the cavity becomes
filled with rubble and, later on, by materials from new lava flows, but sometimes a more or less

403
Groundwater studies
Table 16.2 Common ranges of permeability for water at normal temperatures (10 to 25C) and values
of transmissivity. The ranges are only indicative. Values are derived from a wide literature review
and experience in the Canary Islands
Perme- Reported trans-

ability missivity values
Material (m/day) (m2/day) Comments
Basalt flows 10 5 10 2 100 Dense, unfractured to young,

highly brecciated
Basaltic formations 0.01 20 2 100 Several flows with pyroclasts,
aa to pahoehoe types
Core basalts 0.001 0.05 0.1 10 Thermally altered, dyke
intrusion
Basaltic traps 0.001 10 1 50 Several flows, no pyroclasts,
moderately old
Recent basaltic
formations 1 10 3 100 10 5 Unaltered
Conglomerates 0.01 0.5 3 50 Mostly basaltic
Loose pyroclasts 0.1 50 10 500 Young
Ash fall 0.01 0.1 0.5 5 Relatively fresh
Andesites 0.1 1 2 200
Phonolites 0.1 20 20 1500 Effect of major fissures
Phonolitic ignimbrites 10 6 0.01 0.1 10 Welded to fractured
Trachy-syenites 0.01 0.l 15
Rhyolites 0.01 0.1 0.1 10
Rhyolitic ignimbrites 10 4 0.01 0.02 0.4
Alluvium and terraces 1 10 2 200 Poorly sorted, derived from
volcanics
Lahars (mudflows) 0.01 10 1 100
Note: Permeability decreases with increasing age and degree of thermal alteration. Old volcanics are generally poorly
pervious, but after exhumation they develop fractures. Reported transmissivity values may not represent the whole
formation but the part penetrated by the well.
long, complex cave remains for a while. All these are short-lived features in recent lava flows,
easily collapsible or fillable, and generally above the saturated zone. In areas close to sea level or
near large lakes, or when recharge is very intense, saturated local permeability may be very
high. In Lanzarote island (the Canaries) advantage has been taken of these circumstances in lava
flows extending beyond the former coast line; wells have been drilled to get large flows of sea
water free from sand and algae for feeding desalination plants.
Young pyroclasts are often loose and very permeable, but in other cases they form more or
less cemented volcanic agglomerates and tuff, very porous and prone to collapse under weight.
Permeability may vary from great to almost impervious. Other volcanic structures like dykes,
sills and necks, have low porosity and very small permeability.
It is very important to note that what has been said so far refers to primary volcanic rock
characteristics, e.g. the characteristics of the rock when formed. Subsequently there may be
important changes that are able to transform these properties. The two main causes of change
are internal adjustments and rock alteration, which includes the ageing effect. These secondary
properties may dominate the hydraulic behaviour of the volcanic formation.

404
The designation internal adjustments includes a series of poorly studied circumstances

which generally produce an increase in the large-scale permeability, with little change in total
porosity. They include small differential movements due to compaction of collapsible formations
under increasing load and after alteration, the bulging or shrinking in response to deep magma
emplacement or emptying and pressure relief. They are also the result of the tectonic response to
deep faults, large landslides, the stress produced by dyke intrusion, the stress relief after erosion
and so on. The result is that joints are subjected to small relative displacements which widen
them, and new joints are created. This makes massive lavas and ignimbrite flows permeable, the
more frequent the dislocations, the more permeable they become. Fissure widths of some mm
are not uncommon, especially near volcanic centres (they are rarer further away) and in some
cases they grow centimetres wide. It seems that subvertical fissures dominate. Fissures remain
open unless thermal and low temperature weathering line and fill them with minerals. The
effect of these fissures is less important in breccias, loose pyroclasts and similar formations,
but dominant in harder rocks. The role of dykes deserves special attention. In Hawaii, because
of their long linear extent, dykes are often considered impervious walls against groundwater
flowing through the more pervious volcanics in between. This is true when the dyke is parallel
to the ridge, not fissured due to internal adjustments generally it is stiff and fragile and the
country rock is permeable, as is the case in Hawaii, where dykes penetrate through relatively
fresh volcanics in the very high rainfall highlands (Fig. 16.4). Circumstances may change dra-
matically with depth or in older volcanics, since then the country rock may have a lower macro-
scopic permeability than the fractured dyke and, what is more important, the contact between
the case rock and the dyke may develop into a fissured zone through which groundwater may
flow readily.
The preferential orientation of dykes makes the permeability anisotropic. When perpen-
dicular to the ridge they may facilitate the drainage. But it must be clear that each dyke behaves
differently from the others according to the complex local circumstances, and consequently an
individual behaviour cannot be generalised, as the mean characteristics may differ greatly from
what is found in a particular well or water gallery. This can be seen in the numerous, long water
galleries in the Canary Islands, where the most permeable, recent volcanics are unsaturated due
to the relatively low rainfall.
Volcanic formations are easily altered, both at the surface and at depth. Their glassy
nature, great specific surface and geochemistry explain this. In wet environments thick soils and
weathered rock layers may develop under moderate erosion rates. But more important are
internal changes. Around magma chambers hot fluids and gases (mainly CO2) metamorphose
the surrounding rocks into a poorly permeable and dense mass; frequently criss-crossed by a
dense network of dykes near the magma chamber. The action extends outwards with progres-
sively decreasing intensity, but still with large geochemical changes produced by CO2-rich, slow-
flowing groundwater. The formation of low temperature silicates may often produce a dramatic
reduction of permeability. Fissures tend to be filled with new minerals. This is why around the
volcanic centres not only are dykes frequent, but the rock forms a core with a very low per-
meability. The slow release of CO2 from the magma chamber and from originally closed pores of
the volcanic rock is a long lasting process.
The above processes explain why old volcanics are much less permeable than young ones.
This is the ageing effect. There is no absolute scale of ageing effects, since the result depends on
a series of geochemical factors, including groundwater, but also on the compactness of the rock.
Formations such as breccias, lapilli, ash fall deposits and loosely-cemented ash flows are easily
altered to a low permeability mass in only a hundred thousand to a few million years, whilst
massive lavas and dense ignimbrites are much more resistant to changes at low temperature,
conserving some of their primary and secondary properties for a much longer time, up to some
tens of million of years. Thus, Pleistocene volcanic tuff and Pliocene block and scoria lava flows

405
Groundwater studies
Figure 16.4 Two modes of interpreting groundwater in the volcanic island of Oahu (Hawaii Islands,
United States of America)
The upper one (from Hunt et al., 1988) shows the assumed compartmental behaviour of the water table
in the densely dyke (dike) intruded area, while away from that area a continuous, gently sloping water
table is depicted; this basal aquifer does not extend below the dyke impounded area. There are
unshown springs in the dyke-impounded water area and all recharge is assumed to be transferred to
the basal area, where discharge is along the near coast, either directly or through the cap rock, on top
of the marine salt water wedge. Only occasional perched springs are depicted.
The lower figure (from Takasaki and Mink, 1985) shows a more continuous water table and explains
in more detail the cap rock of sediments along the coast and offshore, although the salt water wedge
is not shown. The sudden water table fall to the left represents the basal water at the other side of the
island, without details on how it is connected to the dyke-intruded zone. These two figures do not
represent the same cross section of the island. In the first one the dyke-intruded area is far from the
coastal area shown at the right hand side of the figure, and in the second one the cap rock covering it
is depicted.

406
(a-a type) may become quite a poorly permeable mass whilst Cretaceous massive lavas may
conserve a relative good permeability in distorted joints and near the top and bottom of the
flows (pahoe-hoe flows). Volcanics several million years old are generally quite impervious, and
generally behave as aquitards if not as aquicludes. The effect is even more conspicuous when
they occur with other sediments such as sandstones, as happens in the Jurassic-Cretaceous
Paran basin (Serra Geral basalts juxtaposed to the Botucat sandstones), or in the Quaternary
patchy volcanic formations in Olot (Catalonia, Spain) in which confined fluvial gravels
constitute the main aquifer, and altered ash fall and lapilli act as aquitards (Fig. 16.5); the lava
flows are only considered good local aquifers when no gravels are present.
Figure 16.5 Cross section along the Fluvi river valley, Catalonia, Spain (modified from Fayas and Domnech,
1974)
It is shown how successive lava flows from the Pleistocene volcanoes of the Olot area have dammed
the valley until the river was able to breach the barrier. The lavas and pyroclasts confine very
permeable river terrace gravels. Behind the dam a lake formed, which progressively filled up with
clays and silts, which are much less permeable than the volcanics. Large springs appeared when the
present river valley cut through the saturated permeable, mainly volcanic, formations. The deep
confined river terraces are the more permeable formations and are exploited by means of deep wells.
Buried alluvium and terrace deposits may play an important role in volcanic islands as well, as in
Runion (Join et Coudray, 1992, 1993), or in Telde, eastern Gran Canaria island (Cabrera et al., 1992).
16.4 Groundwater flow in volcanic formations
A volcanic formation can be considered as a combination of porous layers and blocks criss-
crossed by fissures. When weathering is intense, thick altered zones (regolith) are formed if
erosion is not too intense and removes them.
Recharge water originates as soil moisture and tends to move vertically downwards
towards the water table. The actual detailed movement depends on the combination of layers,
blocks and fissures. In the unsaturated medium most of the flow is probably through the layers

407
Groundwater studies
Figure 16.6 Schematic, highly idealised cross sections through volcanic formations
The upper one corresponds to an area close to the eruption centres (inspired by what is seen in cliffs in
Gomera and Hierro, Canary Islands, Spain), and the lower one to a piling up of basaltic flows far from
the eruption centres (inspired by descriptions of the Deccan traps, India, or of the Serra Geral basalts,
Brazil). It shows the virtual water table (it can be determined only from permeable features, correcting
for possible vertical groundwater head gradients) and how it changes due to the effect of water
abstraction from a well. The arrows show the local groundwater flow (from Custodio, 1989).

408
and blocks, and not through the fissures, except when recharge is intense, during and shortly
after heavy rains, when surface water infiltrates or below perched saturated layers, both near the
surface or in depth. The formation of perched saturated zones is not rare when recharge rate is
high, and some very poorly permeable layers are found, such as extensive red beds. In sloping
areas, seeps and springs appear when such a layer intersects the land surface. The outflow is
generally only a fraction of local recharge, the rest following the more or less vertical path
toward the water table. Part of the flow of the perched springs and seeps may be locally
evaporated or transpired by the vegetation growing at the site. The remaining flow often
reinfiltrates downstream if permeable formations are found, but in other situations ravines with
perennial flow can be found.
The thickness of the unsaturated zone may vary from shallow (high rainfall, flat areas,
poorly permeable materials, or a combination of them) to very deep, sometimes several hundred
metres (low rainfall, sloping areas, highly permeable materials).
Water flow in the saturated zone tends to follow preferential paths such as little- altered
brecciated zones and open joints, but most groundwater is stored in the bulk of the volcanic
formations as slowly flowing water (Fig. 16.6). As a whole, the volcanic formations approach a
double-porosity/double-permeability system, but less clear-cut than in other hard rocks such as
granite or limestone. The model of storage in blocks and flow in fissures is a better description
than that of a granular-like aquifer, but the actual behaviour is somewhere between the two. The
behaviour changes from place to place according to the dominant type of volcanic formation,
such as lava flows, pyroclasts, ignimbrites or altered volcanics with intruded dykes. Ground-
water flow in the Deccan basalt traps is mostly through vertical fissures and horizontal interflow
layers, and storage is in the blocks and these interflow layers (Kulkarni et al., 2000).
The actual groundwater flow system in lavas and ignimbrites far from eruptive centres is
best described by the fissure and block model.
Large lava fields may confine older aquifers e.g. in the Paran basin and the Cretaceous
trap basalts (Serra Geral, Alto Paran and Arapei formations). The latter are up to several
hundred metres thick; there are poor aquifers in the 50 to 100 m upper fractured layer below the
clayey soils and the rest is an aquitard or even an aquiclude that confines the underlying
Jurassic Guaran sandstone aquifer, a main regional aquifer that is exploited by deep wells,
locally free flowing and producing warm water, that extends over Brasil, Uruguay, Argentina
and Paraguay.
The groundwater flow description becomes more uncertain when approaching the
eruptive centres as well as in volcanic islands. In these islands young volcanics may form a
relatively permeable cover on top of a poorly permeable, densely dyke-intruded core in which
groundwater flows slowly. The water table may attain high elevations, up to the gradational
transition to the cover of young volcanics. In some of the Canaries the water table may be more
than 1,000 m above mean sea level, with slopes greater than 0.1, which indicates the existence of
a poorly permeable core of older, more altered volcanics. Permanent springs appear in areas
where the young volcanic cover thins out and is no longer able to transmit all the groundwater
flowing inside it (Fig. 16.7).
The existence of such low permeability core of old, even submarine volcanics, or physi-
cally and chemically highly metamorphosed ones is not always clear since it may not outcrop or
it is difficult to recognise due to slope debris and dense vegetation cover, as in humid tropical
volcanic islands. However, geophysical surveys and thermal and hydrological characteristics
may show its existence, as in La Fournaise, southern Runion island (Violette et al., 1997;
Robineau et al., 1997; Courteaud, 1996), and also in Hierro, Canary Islands, after the recent
drilling and excavation of a deep, long water gallery.
When erosion cuts through the core, numerous very small springs may form, many of
them just sustaining vegetation. The landscape is very different from areas covered by young

409
Groundwater studies
Figure 16.7 Schematic cross section through volcanic islands of the high type
The figure shows the formation of a water-saturated high dome due to the presence of a poorly
permeable core. This corresponds to the situation of Gran Canaria and Gomera (Canary Islands) and
Madeira, and to some extent can be applied to Tenerife and La Palma (Canary Islands). The core may
not exist above sea level in recent islands and then this mounding is not produced.
volcanics in which the vegetation is reduced to what the local soil cover and rainfall permits,
except around the sparse, relatively large springs. Some of these springs form permanent
streams, although water may be partly hidden in the torrential bed deposits. When the young
cover is thick enough and when there are no large permanent springs the entire recharge is
discharged into the valley bottom or into the sea.
Groundwater flow may be modified by major features, such as the sedimentary for-
mations (marls and limestones) sustaining Mount Etna, Sicily (Italy); the terrestrial and marine
deposits (cap rock) confining permeable basalts in some coastal areas in Oahu, Hawaii, or
acting as a main drain in eastern Gran Canaria; the river terraces and lake deposits in Olot, Cata-
lonia, Spain; the sandstone formations below the trap basalts in the Paran basin, etc. Inside vol-
canic formations densely dyke-intruded zones may act as anisotropic ridges, with open fissures
parallel to the dominant dyke orientation and very low regional permeability at right angles to

410
this. Major landslides may produce fine-grained breccias at the slipping surface, which may be a
major internal discontinuity, as in the case of the mortaln in Tenerife (Canary Islands).
Major discharges (springs, nacientes) tend to concentrate in breccias and enlarged joints.
Poorly consolidated local features surrounded by more competent formations tend to result in
caves being formed in the cliffs, since loose material is easily removed, producing what have
sometimes been erroneously called karst-like features. These features do not behave like karst,
nor are they produced the same way. The same holds for springs outflowing from tree log
moulds and volcanic caves in high rainfall, high water table areas. These are occasional features
which allow the concentration of otherwise diffuse groundwater discharge, but lack the organ-
isation of karst. The same can be said of the deep vertical holes sometimes found near young
volcanoes. They are vertical moulds made from thermally hardened rock or vertical lava pipes
after the magma inside was suddenly expelled by a burst of gas. They are in no way like a karst
sinkhole.
Groundwater flow in volcanic formations has been successfully described at regional scale
by means of anisotropic, continuous media equations. Adjusted parameters are not true storage
and permeability values (or transmissivity in two-dimensional horizontal situations) but do not
differ greatly from values obtained by means of pumping tests. Double continuum models com-
bining a high permeability, low storativity continuum with a high storativity, low permeability
continuum linked by means of a block fissure water exchange function seem a better approach
to groundwater flow simulation. Up to now these models have seldom been applied to volcanic
formations except for studying poorly permeable formations considered as possible repositories
for nuclear waste. Many of them correspond to Yucca Mountain, in western United States of
America (Hinds et al., 1999; Sonnenthal and Bodvarsson, 1999), where advanced tests method
have been devised (Huang et al., 1999).
The validity of Darcys law to describe macroscopic flow in volcanic formations is a recur-
rent topic. Detractors often rely on speculation and poorly interpreted local behaviours. Exper-
ience in the Canaries and elsewhere show that macroscopic behaviour can be linearly described
by groundwater heads and permeability-like coefficients. Measurements obtained from several
piezometric boreholes in water galleries in Lanzarote (Fig. 16.8) and Gran Canaria show that the
water head line along the gallery is macroscopically continuous and has a constant slope, in
spite of the different lava flows, pyroclastic layers, dykes and red beds appearing along the test
tunnel.
Pumping tests to determine hydraulic properties are generally interpreted in the same
way as the tests carried out in granular materials. There is little experience of tests carried out in
well bounded conditions. In such cases the common models (Theis, Hantush) are able to
reasonably reproduce drawdown-distance-time behaviour, yielding transmissivity and storage
coefficient values which are averaged properties of a relatively large volume of rock. It seems
that the consideration of anisotropic effects (Papadopulos-Cooper model) improves the fit, but
multi-piezometer tests are rare. In other cases the Neuman model of delayed response of the
water table seems to improve the fit to real data. Often only drawdown at the pumping well is
measured. Where the wells are large diameter ones sometimes with enlargements, galleries
and horizontal drains that partially penetrate a variable formation aquifer, well storage may
dominate the test. In such circumstances it is difficult to obtain aquifer parameters, and the good
fits sometimes obtained may owe more to the fitting effort than to the interpretation model.
16.5 Hydrogeochemistry and mass transport in volcanic formations
Volcanic formations do not usually contain significant quantities of water-soluble material, such
as alkaline halides or sulphates, because these compounds are easily altered (hydrolysed) if

411
Groundwater studies
Figure 16.8 Experimental water gallery in the Famara massif, Northern Lanzarote, Canary Islands, Spain
The massif consists of more than 700 m of late Tertiary sheet lava flows, with some extensive red layers,
buried cinder cones and a moderately dense network of dykes, oriented approximately NS. Several
vertical and horizontal bores were drilled along the gallery, and fitted with pressure meters in the
section below the original water table. The heads that were recorded correspond to the points in the
upper figure. These points lie on a nearly perfect straight line. The small deviation can be explained by
vertical head gradients created by the water flowing into the gallery. Although the gallery is relatively
short, it penetrates several dykes and cinder cones and affects several lava flows and red layers.
The lower figure shows the Famara massif water table contours in 197072 (Custodio, 1978, 1989).
They show the mound-forming effect inside the poorly permeable volcanics of the massif, even for the
small recharge shown below. The water table flattens in the more recent, even sub-historical lava flows
of the eastern side. The massif has some very small, perched springs (seeps), supported by the red layers
when these are locally thick and continuous. This water has the same chemical characteristics as water
from the saturated part of the gallery and shows similar aridity effects on its isotope composition.

412
water is pH buffered. Dissolved CO2 plays this buffering role. Low temperature alteration
(weathering) frees alkaline and earth-alkaline metals in the rock as soluble ions, together with
some silica, and bicarbonate ions. The rock minerals are transformed into clay minerals which
include oxides of other metals such as iron. Most iron in fresh volcanic rocks is generally in the
reduced state. In oxygen-free acidic water it dissolves as Fe++, but usually it is oxidised by
oxygen dissolved in the water and precipitated as Fe3 oxides and oxyhydroxides. Aluminium is
practically immobile and remains in place as part of clay minerals, or as aluminium hydroxide
(gibbsite) if there is intense leaching.
Most of the Cl and SO42 in groundwater comes from rainfall and dry atmospheric
deposition (fallout). But some volcanic rocks, especially the submarine ones, may contain
enough sulphide to contribute significant quantities of SO42 upon oxidation.
In some cases there is a lithological origin of Cl and SO42, as an Fuerteventura island
(Canary island), probably related with submarine volcanic formations (Herrera and Custodio,
2000) and sometimes the joint consideration of Cl and Br is an useful tool to know the origin of
salinity and the possible contribution of volcanic activity (Davis et al., 1998; Custodio and
Herrera, 2000).
Though their overlying soils contain oxidisable organic matter, volcanic formations do
not contain significant quantities of oxygen-consuming substances other than reduced iron and
manganese. The groundwater is rarely under strong reducing conditions and most often it is
under oxidant conditions. Buried organic matter from soils and vegetation covered by lava flows
may create strong reducing conditions, even producing methane, but they are rare and local.
The cationic composition of groundwater is related to the type of volcanic rock
(Fig. 16.9). Considering the ratio of alkaline to earth-alkaline ions, the lower values correspond
to basic (low silica) volcanic rocks such as basalts, and the high values to acidic (high silica)
rocks such as rhyolites. The results are more realistic if the contribution of atmospheric cations is
first subtracted. This can be done by using the Cl content as a guide.
Groundwater salinity in volcanic rocks is the result of atmospheric deposition of salt ,
climatic conditions and rock weathering. There are no major differences compared with other
easily weatherable rocks under similar circumstances (Custodio et al., 1997). Since chloride is not
significantly contributed by the rock in areas with no significant human influence (agriculture,
urbanisation, etc.) the simplified long-term balance (without considering storage terms) is:
P . CP = R . CR + S . C S
in which:
C = mean concentration of the conservative solute of interest in the water balance

component shown by the subscript [M L 3]
P = mean local rainfall = ET + R + S [L T1]
ET = mean actual evapotranspiration
R = mean recharge to the aquifer
S = surface runoff
CP includes the chloride contributed by atmospheric fallout and may be obtained from an
open, rainfall collector protected against evaporation. If D is total chloride contribution per unit
land surface area and unit time [M L2 T1], CP = D/P. Typically D may vary from 25 g m2 a1
near the coast to 0.2 1 g m2 a1 in continental situations. Then, since no other sources are present
and chloride does not precipitate, except in very extreme climatic conditions:
CR = (P/R) CP (S/R) CS

413
Groundwater studies
Figure 16.9 Relationship between rock and water composition
The figure shows the relationship between mobile cation rock composition and water composition and
major monovalent and divalent cation composition of groundwater from different volcanic rock types
(lower figure), as deduced from studies in Gran Canaria (CEDEX, 1987; Custodio, 1989). The more basic
(closer to basalt) the rock is, the higher is the ratio of earth-alkaline to alkaline ions. The anion content
is to some extent influenced by arid climatic conditions and by transformation of volcanic CO2 into
bicarbonate ion. The r represents concentration in meq/L.
In permeable soils the second term of the right-hand side of the equation is small, and then:
CR (P/R) CP
(P/R) is the climatic-pedological concentration factor, which means that when converted
into recharge rainfall salinity increases moderately in humid climates, but greatly in dry
climates. Rainfall salinity depends on distance from the sea along the dominant wind direction
(Custodio et al., 1997). The same is true for Br and also for SO42 if the contribution by rock
sulphide oxidation is negligible and there is no net sulphate precipitation in the upper soil. Near
populated areas and far from the coast SO42 increases relative to Cl due to anthropogenic
sources and recycling.
For other ions the balance has to be changed into:
P . CP + W = R . CR + S . C S + F
in which:
W = contribution from weathering [M L2 T1]
F = separation by net precipitation, ion exchange and/or incorporation
into weathering products [M L2 T1]

414
W is limited by oxygen availability (which is transported with recharge water and slowly dif-
fused from the atmosphere) and especially by carbon dioxide supply. In very simplified terms,
each molecule of CO2 contributed is transformed into HCO3 + H+ and each H+ is exchanged (in
mass balance terms) by an atom equivalent quantity of cations. The actual behaviour is some-
what more complex, depending on the final pH and redox reactions.
CO2 availability in the soil is limited by the organic matter oxidation rate. It is low in arid
and cold climates and relatively high in warm, humid conditions with abundant vegetation. In
this case weathering (rock factor) dominates groundwater chemistry. The resulting water con-
tains the rock dominant cation and bicarbonate, is rich in dissolved silica and is relatively
diluted due to the high recharge. In dry climates the climatic-pedological factor may remain
dominant and in areas under marine influence recharge water generally is of the sodium
chloride type, but local circumstances play an important role.
As in other rock formations, the chemical composition of recharge water is related to the
soil rock which is being weathered, with little change afterwards, except for groundwater paths
with very long residence time, but in volcanic formations volcanic CO2 has to be taken into
account. The CO2 is slowly released from the cooling magma chambers and diffuses through the
volcanic rock mass, even if this is water saturated, towards the external atmosphere. This
includes CO2 in the closed pores of volcanic rocks that is released after weathering or increasing
fracturing due to overburden pressure and tectonics.
CO2 release from volcanic effluent is intense at the initial cooling stages and decreases
with time. It may produce bicarbonate-rich water if enough contact time with the rock is
allowed. In the areas affected by outgasing, chloride-poor but very bicarbonate-rich (up to 2 g/l)
water may form, especially in the volcanic cores where the poor permeability ensures that there
is enough contact time. Acidic and intermediate volcanics and alkaline basalts may thus contain
water rich in sodium bicarbonate. Less alkaline water, rich in magnesium and calcium bicar-
bonate and generally with a high pH, is found in plagioclase basalts. Calcium carbonate may
precipitate, filling voids and cracks.
Released CO2 from active or recent volcanism, coming from degassing magma, generally
diffuses slowly through the rocks and groundwater. In some case it may accumulate in deep
lakes (Delmelle and Bernard, 1994) and even create episodically sudden releases. It may also
accumulate in confined aquifers that when tapped by a borehole may burst out during some
time. A spectacular case has attracted the attention of newspapers during the summer of year
2000 in the now extinct Quaternary volcanic area of Campo de Calatrava (Ciudad Real, Central
Spain).
Chemical denudation rate depend on recharge by rainfall and dissolution capacity, mostly
due to CO2 (Louvat and Algre, 1997). Actual conditions vary from very slow in arid areas
(rainfall may be less than 0.1 m/a, as in some parts of the Canary and Cape Verde islands), up to
more than 10 m/a in some tropical, humid areas, as in some places in Central America, Runion,
Tahiti and Hawaii.
The stable isotope composition of groundwater is not affected by volcanic rocks except
under hot rock conditions or intense gas flow. In a hot environment the 18 O in the water is
readily exchanged with that in the rock, and the water becomes heavier (less negative 18 O), but
the 2H (deuterium) value does not change. The exchange is less marked than with carbonates.
If enough CO2 bubbles through the water the exchange makes the water lighter in 18O, but the
2H is unchanged if there is no evaporation. Evaporation makes the water heavier, both in 18 O
and 2H, with a 18 O2H slope less than 8, depending on temperature. The use of 18 O and 2H
to study and characterise groundwater flow does not present special difficulties. Classical
interpretation rules and models can be applied in most situations. They are very useful tools in
areas with large topographical variations, as in many volcanic islands and formations
(Fig. 16.10). The same can be said for radioactive 3H (tritium), but in this case its transient

415
Groundwater studies
Figure 16.10 Water isotopes applied to volcanic areas
The figure shows the 18O vs. D(2H) plot of rainfall data across Gran Canaria (dots) and the results of
sampling small seeps, shaft wells and deep bored wells in the southeastern part of the island (Amurga
phonolitic massif), which is an arid area. There is some recharge from occasional storm water which
follows the World Mean Meteoric line (m = 8; d = +10), but part of it evaporates from the bare rock and
the soil surface due to the sparse vegetation cover. The lower figure shows the 18 O vs. altitude relation-
ship for Gomera (Canary islands). Most water samples come from springs, from near sea level up to
1,300 m. The isotopic slope is about 0.225/100 m. The lower point of each plot is the sampling altitude
and the upper one the mean altitude of the assumed recharge area. Some springs and galleries receive
water from a higher altitude, which means significant lateral transfer of water through the more
permeable upper volcanic formations resting on older volcanics.

416
concentration introduces additional mass transport problems. Waters of different origin and
variable residence time can be identified (Falkland and Custodio, 1991; Pennisi et al., 2000).
Since volcanic rocks are devoid of carbonate minerals except for some secondary
deposits in veins and pores of old volcanics the dissolved inorganic carbon (DIC) comes
exclusively from CO2. In areas not affected by volcanic carbon dioxide, DIC isotopic composition
is derived directly from CO2 released by decaying vegetation. Thus, if vegetation is characterised
by 13Cv , groundwater has a composition between 13 Cv (closed system to CO2) and 13Cv +
(open system to CO2 when dissolved inorganic carbon is dominantly HCO3; pH between 6.8
and 7.8). is the temperature-dependent fractionation factor between CO2 gas and dissolved
HCO3. Thus, for Calving-cycle plants and temperatures around 18C, the dissolved inorganic
carbon 13C varies between 25 and 17 relative to the PDB standard. Carbonate precipi-
tation in arid areas may change these results.
The same is true when the dominant DIC source is CO2 from magma degassing, but
in that case 13Cv is close to 8. For a system that does not lose CO2 (closed system to CO2)
13CD 8 but when groundwater is almost stagnant and CO2 diffuses through it toward
outer regions the behaviour of the CO2 is close to an open system and 13CD 0 at about 18oC,
when HCO3 dominates.
A series of intermediate situations between the two origins of CO2 is possible. This makes
it difficult to interpret radioactive 14 C content to obtain groundwater age. The carbon hydro-
geochemical behaviour in the volcanic system under study needs to be considered carefully. If
volcanic rock circumstances are not duly taken into account, classical correction procedures,
mostly devised for carbonate rich formations, may fail to give reasonable answers. Water devoid
of significant quantities of radiocarbon are not necessarily old but may merely reflect a
dominant volcanic CO2 contribution. This pose additional problems to interpretation (Rose and
Davisson, 1996). Correlation between 13 C and total dissolved carbon and between 13 C and
14 C content may help in obtaining correct interpretations.
From the few applications so far it is clear that mass transport of solutes through volcanic
formations cannot be accurately described by the common simple continuum mass transport
equation. A double continuum approach seems to improve results, but there is little experience
of using this method. The solute is diffusely exchanged between the porous blocks and the low
storativity fissures. Chemical or temperature changes are dispersed and delayed, with sharp
fronts and long tails. Thus, conservative solutes such as tritium are affected and groundwater
turnover time calculation by means of environmental tritium is more complex than in granular
aquifers under similar geometrical circumstances.
Case studies dealing with the transport of contaminants in volcanic rocks are scarce, espe-
cially for reactive contaminants and those which may decay. It seems that absorption isotherms
and ion-exchange rules can be applied to some extent, but actual properties are poorly known.
Studies on pesticide transport in Oahu island, Hawaii (United States of America) show that
transport in soils can be described conventionally (Oki and Giambelluca, 1989), when there are
thick soils (Loague, 1994). In other cases soils are thin or absent. In this case little knowledge is
available.
16.6 Groundwater quality issues in volcanic formations
Groundwater quality issues in volcanic rocks are not essentially different from other rocks, but
there are some aspects that deserve consideration:
the easy weathering of the rock means that before being released to the atmosphere
soil CO2 is used to increase total dissolved solids in a larger proportion than in other
rocks under similar conditions;

417
Groundwater studies
the contribution of volcanic CO2 in areas with relatively recent volcanism or magma
emplacement episodes may produce water rich in dissolved solids, rich in sodium
and/or magnesium bicarbonate, sometimes sparkling at atmospheric pressure, and
sometimes with rather high dissolved Fe++. These waters are unsuitable for domestic
supply, may be highly corrosive to metal and fibro-cement pipes, or highly incrusting,
depending on circumstances, and most unsuitable for irrigation.
is not rare for fluor ions, F, to attain concentrations well above the upper limit
it
permitted for drinking water, especially in acidic, soft water.
Occasionally, problems are caused by high concentrations in boron or in some heavy semi-
metals, such as vanadium and arsenic but most problems of related to the effects of late stages of
volcanism on some springs. Problems of high V and As concentrations in pore waters in the
semi-arid plains (Pampa) of Argentina are linked to glassy volcanic dust from the Andean
volcanoes that becomes incorporated into the aeolian-fluvial deposits; a process still active. In
Mexico, arsenic is a problem in some rhyolite aquifers. It is well known that large metallic
mineralizations are associated to important ancient volcanic events, mainly in subduction areas.
This is well characterised in the Pyrite belt and Ossa-Morena areas of the Iberian Peninsula. This
is the subject of classical research and mapping studies. The erosion, transport and subsequent
sedimentation, adsorption or precipitation of some components may be the origin of noxious
concentration of some elements in groundwater, although in many cases the geochemical
conditions are complex and poorly known.
16.7 Groundwater exploitation in volcanic formations
Groundwater exploitation from volcanic formations relies on the same methods commonly used
in other formations, especially in hard rocks, although there are specific issues.
In lava formations far from the eruption centres, springs outflowing at valley bottoms are
tapped. Often these valleys correspond to tectonic features in which not only erosion acts prefer-
entially but permeability is also locally greater. The most frequent form of groundwater abstrac-
tion is by means of wells. Large-diameter dug wells, 2 to 8 m in diameter and a few tens of
metres deep are common in the Deccan traps. The large storage capacity of water inside the well
allows for relatively high pumping rates during short periods, though the yield of the aquifer is
often low.
Currently, drilled wells are common. They are 0.1 to 0.6 m in diameter and up to several
hundred metres deep. Drilling with a pneumatic down-the-hole hammer rig is currently the
most common drilling method, although not without problems of soft sections being unstable or
caving. Cable tool percussion is also used, but penetrating thick massive and abrasive lava flows
is a slow and costly process. Almost any drilling procedure for hard rock has been used with
variable success, depending on local circumstances and drillers experience. Wells are cased
when soft sections cave in, or to protect the pump. Torch or machine slotted tube is generally
sufficient and preferred to more elaborate screen sections.
The groundwater yielding formations are mainly the breccia and scoria between lava
flows, especially the less altered and thicker flows. In some cases it is possible to assign water-
yielding properties to identifiable individual lava flows. Data for the Deccan traps show that
permeability decreases with depth.
In the Paran basin basalt outcrops the situation is not so clear, but water-yielding proper-
ties decrease with depth, too. In this case, large fissures are also a main source of groundwater,
but they are not easily intersected by wells, since these fissures tend to be vertical. There is a
greater probability of encountering fissures close to tectonic features, which in some cases show
up as lineaments in aerial pictures and as deep valleys on the ground.

418
In volcanics at or near eruption areas matters are more complex due to the variable
internal structure of the formations. They may present a wide range of groundwater-yielding
properties. As these areas are generally of pronounced relief, relatively high altitude springs
may appear, as in many volcanic islands receiving relatively good rainfall recharge. The direct
tapping of the streams that originate from them allow water to be transported by gravity to
urban areas or to irrigated fields at lower elevations, and even to generate some electricity if
there is enough flow and head difference, as in the SE of Runion island. Water that is not used
is lost downstream or to the sea. In water-scarce areas, reservoirs have been built to store excess
water for periods of high water demands. Surface water storage reservoirs may be built by
damming the valleys when these are in poorly permeable volcanic material, as in Gran Canaria;
otherwise, artificial impermeable linings must be constructed.
In some situations spring tapping is improved in order to collect much of the otherwise
dispersed outflow in one point. This is done by digging horizontal water galleries at or below
the spring outflow point. After excavation the water flow often increases dramatically for some
time, due to the drainage of the local groundwater storage. When this is depleted, the final
steady flow may be considered to be too small, requiring the gallery to be extended. This
process has produced in Tenerife (Canary Islands) hundreds of water galleries up to 6 km long
and up to 1 km below ground level at the front. Some of them have reached the high-
temperature, sodium bicarbonate rich waters of the island core. Water galleries are efficient
groundwater abstraction works if they penetrate perpendicular to dykes and major vertical
fissures, but the final steady flow depends on the fraction of the recharge they intercept. This
may be quite small, even for a long gallery, as most of the yield is produced at the head section
when much of gallerys length may become unsaturated zone due to depletion and the effect of
other water galleries at the same or at a lower altitude. Bores drilled horizontally from the
gallery sides help to increase the yield temporarily, in the same way as in large diameter wells.
Transient periods may last from weeks to many years. As in the case of springs, water that is not
used is lost downstream. This is of special concern if this water is one time storage water.
Locally bulk-heads have been installed deep inside the galleries. A few examples can be found
in Mount Etna (Sicily), in Tenerife and in Hierro (Canary Islands). Bulkheads are difficult to
construct, due to the high pressure that may develop behind them. A massive dyke may be a
good place to anchor them. Water may penetrate through most of the active length of the gallery
but most often it appears in short sections separated by long, poorly yielding and dry sections.
Some sections have to be lined to avoid caving in.
Wells are the most commonly used groundwater abstraction works, although water has to
be pumped out, expending energy and using costly machinery. This was a major problem in the
past when steam or oil driven motors were needed to move piston pumps to push up the water.
Currently, submersible pumps and electrical energy have simplified many of the old problems.
In remote areas electrical energy is locally generated by means of oil-driven motors. The appro-
priate solutions depend on the area (Kulkarni et al., 1997).
Drilled wells using the methods commented on before are now common, but
groundwater-yielding sections are often separated by long sections of non-yielding rock. This
means that deep wells have to be drilled, with the risk of bypassing good quality, short turnover
groundwater below the water table and instead tapping more mineralised, deep groundwater,
or mixing the two at the pump inlet. Drilling deep below the water table is a problem for air-
driven bottom hammers, since air pressure has to overcome the pressure created by the water
column in the bore. High pressure air compressors or supplementary airlift is needed. Caving-in
of loose material is a major problem, producing serious failures. Casing is a method to avoid
problems of caving-in but makes drilling slower and more expensive. Vertically drilled wells are
not efficient groundwater abstraction works in areas where vertical fissures are important water-
yielding features that help in draining large volumes of rock.

419
Groundwater studies
In the Canary Islands, and especially in Gran Canaria, the solution to accommodate the
old bulky pumping machinery was to construct large diameter wells (about 3 m), excavated like
a mine shaft with lining and reinforcement of soft sections. At the same time the large diameter
of the well allowed horizontal water galleries to be excavated from the well bottom in order to
intersect vertical groundwater-yielding internal features, thus greatly increasing the well yield.
More recently the dangerous, expensive and risky gallery excavation method has been replaced
by horizontally drilled bores (catas) from the well bottom, using especially adapted, electrically
powered drilling machinery. The catas are often 50 to 200 m long, but some are longer than
500 m to increase the yield of the expensive excavated wells, which are often up to 500 m deep.
The drilling machine has to be placed below the static water table and the well shaft has to be
continuously dewatered by pumping.
Groundwater production from the well drains water stored in the rock and the discharge
decreases as this storage is depleted. When well yield becomes too small to meet needs, the well
has to be deepened and new catas drilled. This is an expensive process that need frequent or
continuous work inside the well. Moreover, more energy is needed to pump out the water.
Whether redrilling is advisable depends on the price of water in the market, in combination with
water quantity issues and administrative regulations from the Water Authority. Existing wells
are maintained and deepened, but new wells are rarely constructed, due to the large investment
required to reach the deep saturated zone. Currently these large diameter shafts are extended
down by means of small diameter, mechanical drilling, substituting lateral drainage by deep
penetration, but this currently cheaper method is hydraulically less efficient in draining large
rock volumes.
In strongly sloping ground, side galleries are constructed to take out the water at a lower
elevation, thus saving energy. Something similar is done in long water galleries, from which
outflowing water has to be pumped prior to distribution. By using a closed pipe, instead of an
open canal to tap the water at the gallery bottom where most of it is produced, saves part of the
hydraulic head above the gallery mouth due to the slope of the gallery (about 0.02).
16.8 Groundwater balance in volcanic formations
The groundwater balance in volcanic formations follows the same principles as in any
other aquifer system (Custodio et al., 1997), and will not be described here, except for some
specific comments.
Groundwater recharge by infiltrating rainfall and snowmelt is a major term which is
difficult to calculate with accuracy. This is not new. In many volcanic areas, soil is very thin and
permeable, and vegetation is shallow rooted. Consequently, in wet climates a large part of rain-
fall becomes recharge. Values greater than 50% of local rainfall are possible. But in arid climates,
even in sparsely vegetated areas, most of the rainfall returns to the atmosphere, partly by direct
soil evaporation. This is reflected in the isotopic composition of the water. Studies carried out in
southeast Gran Canaria (Canary Islands), in a dry area receiving between 100 and 200 mm of
rainfall annually, showed a mean recharge of 2 to 10 mm/year with a salinity of up to 5 g/l total
dissolved solids due to the rainfall taking up airborne marine salts (Custodio, 1992).Recharge is
zero in most of the years, except the wettest ones. As a consequence, a thick unsaturated zone
and the thick saturated zone underneath results in old radiocarbon ages for pumped water, even
though this water comes from a water table aquifer receiving some recharge.
Many volcanic areas may develop fertile soils on which a dense vegetation cover can
develop. This cover often uses rainfall water very efficiently in moderate precipitation areas,
thus dramatically reducing recharge. Some of this vegetation may satisfy part of its water needs
from capturing air moisture in foggy areas.

420
Recent volcanic formations, still devoid of significant vegetation, highly porous and with
a very irregular land surface (badlands or malpais) seem prone to transform most of the rain-
fall into recharge. This is true for high rainfall areas (in which weathering proceeds fast) but not
necessarily for arid areas. Air may flow through the upper few metres of the ground due to
wind effects and temperature gradients, taking away most of the rain-water held in the large
specific surface lapilli, scoria and breccia.
Although often poorly permeable, volcanic formations may conserve connected porosity,
some permeability and open fissures down to large depths. This has been shown in studies for
possible nuclear waste repositories in North America and also in the long water galleries of
Tenerife (Canary Islands). In the latter there is some groundwater permeability in the first
thousand metres and probably deeper. This has to be taken into account when putting
boundaries to a groundwater balance domain.
Low permeability, the possibility of preferential drainage through fissures intersected by
groundwater abstraction works and regionalised systems due to the large active thickness
means that any hydraulic disturbance is followed by a long transient period which can be
measured in terms of
L2 S/T L2 Sv/k
in which:
L = linear dimension of the system [L]
S = storage coefficient [dimensionless]
Sv = specific storage coefficient [L1]
T = transmissivity [L2T1]
k = permeability [LT1]
In core situations, characteristic times in the order of one year are common values for a
deep well or gallery, due to the effect of the low, local Sv value (elastic behaviour), but for the
whole formation they may attain hundreds of years for a long-term drainable porosity of 0.02 to
0.06. This mean that storage terms may play an important role in the groundwater balance of the
system. This is something often ignored, especially in volcanic islands, where the groundwater
outflow to the sea is not measurable and mostly unnoticed. Thus, even when groundwater
abstraction is less than recharge, continuous large drawdowns may be observed -larger around
individual wells and water galleries- since the steady state can only be attained by reducing the
outflow to the sea or a main river valley. When groundwater abstraction is far from the coast or
the valley bottom, this is only possible with a large total drawdown.
Groundwater management issues are complex ones and they depend on local circum-
stances and rules. When water is scarce due to aridity and/or high agricultural demand, prices
are high. These rules are set, although they do not get necessarily an efficient use of water and a
rational development of groundwater (Cabrera et al., 1997). It is not rare that private ground-
water markets develop.
16.9 Groundwater monitoring in volcanic formations
Groundwater monitoring in volcanic formations covers three main topics:

water levels,
water physico-chemical and quality characteristics, and
water abstraction.

421
Groundwater studies
In thick, poorly permeable formations, the three-dimensional nature of groundwater flow

and chemical characteristics has to be taken into account. Water table slopes may be greater than
0.01 and sometimes exceed 0.1 (often under land surface control). In these cases, vertical flow
components are important. This is clearly shown by the drawdown of the static water level in
wells in highlands when they are deepened. Some authors describe groundwater flow in the
massif as a series of numerous perched aquifers on top of a deep regional level corresponding to
what has been called a basal aquifer. This is generally incorrect and the very permeable deep
layer cannot exist except if it lies on top of other formations such as fractured limestones. The
conceptual model is a thick saturated poorly permeable formation with large vertical head
gradients. This comes from the classic description of aquifers in Oahu, Hawaii, in which
pervious recent basalts in the coastal plain have a gently sloping water table (the basal aquifer).
This aquifer is at the foot of the much less permeable, densely dyke-intruded island core, with
high altitude perennial springs. This is a common situation in many other volcanic islands (e.g.
Tahiti, Runion, Martinique), but it does not mean that the basal (coastal) aquifer groundwater
levels continues inland below a thick series of perched aquifers.
Groundwater monitoring of pumping wells may give distorted results, due to local
storage depletion. Moreover, in large-diameter wells the water levels measured when the pump
is idle may show poor water level recovery, due to the large well capacity. Thus, the monitoring
network has to be carefully designed and some data may need corrections. A deep open bore-
hole shows a water head which is a composite value of the point heads along its length. Inside
water galleries the locally influenced water table position can be seen when it is penetrated, but
since there are large, poorly permeable sections and the ambient air is almost water vapour
saturated, often the exact position cannot be located. Furthermore, the gallery is continuously
yielding water and thus the intersection with the water table is locally distorted.
Groundwater level monitoring is best done by means of drilled observation boreholes, but
this is expensive, especially when the unsaturated zone is thick. The ideal situation is when
various separated piezometers (different bores or various isolated tubes in a large diameter
borehole) are installed at different depths at a site, each with a short open section, grouting the
rest. Since permeable features may be separated by long, poorly permeable sections, the placing
of the open sections has to be decided after drilling and testing the borehole. Open sections in
very poorly permeable rock may need weeks to show the true water level, and water head
changes are greatly dampened and delayed. In such circumstances isolated water pressure
transducers may be a solution for monitoring.
Groundwater quality monitoring can be carried out by periodically sampling wells and
springs, generally once or twice a year. Consideration has to be given to what represents the
water sample. When the well is not pumped, the water in it comes from the highest water head
level or is dominated by water falling from possible perched aquifers. During pumping a mix-
ture of the different yielding levels is produced, which varies with time and with pumping
discharge, and in large-diameter wells it is influenced by water in storage. It is not easy to obtain
good water samples which are representative of particular situations. Pumped water is a mix-
ture of many different origins and residence times. Seasonal changes are sometimes observed,
but they are more the result of variable mixtures due to seasonally affected groundwater levels
than of a true seasonal change in water quality. To study agricultural pollution, the upper part of
the water table aquifer must be sampled.
In coastal aquifers the existence of sea water and the mixing zone has to be considered.
Salinity stratification can be studied by logging the electrical conductivity (EC) of the water
along long open boreholes, but it is not rare for small vertical flows along the borehole to modify
salinity distribution and the mixing zone thickness. Water temperature logging is very useful in
identifying anomalous situations. Young volcanics far from the eruption centres and near the

422
coast are easier to monitor, since mean permeability may be more homogeneous, and vertical
flow negligible.
Monitoring groundwater abstraction seems an easy task in principle but it is very difficult
in practice. Often, wells are operated discontinuously, pump discharge varies with water table
elevation and the efficiency changes with time. Generally there is neither a flow meter nor an
hour meter, or if they have been installed they do not work properly or at all due to lack of
maintenance. Excess water from springs and galleries is seldom gauged; the same applies to
water in streams. Field survey and interpretation of aerial and satellite pictures is a gook com-
bination, but direct involvement of groundwater users through their representative associations
is of paramount importance.
Recharge is seldom measured due to its complexity and the restricted value of local
results. It is preferentially calculated from meteorological data and some soil observations and
calibrated against water level changes or spring discharge if possible. This is a difficult task in
volcanic formations with thin or discontinuous soil cover.
16.10 Geothermal effects in volcanic formations
Geothermal energy will not be dealt with here. Since volcanic formations may contain magma
intrusions such as magma chambers, sills or buried necks, the vertical thermal gradient may be
increased and convection of hot fluids may take place, affecting groundwater resources. The
effect of CO2 release has already been discussed.
The interaction of water with hot rock affects the equilibrium between water and silicate.
This decreases the Na/K ratio and produces a 18O shift to more positive values. If there is some
boiling the shift also effects deuterium. Condensed steam may produce small seeps that are
characterised by lighter isotope composition and a larger deuterium excess.
Groundwater involved in geothermal convection is dominantly of meteoric origin, but on
small islands seawater is involved as well (Henry et al., 1996).
ARAA, V; LPEZ-RUIZ, J., 1974. Volcanismo: dinmica y petrologa de sus productos. Ediciones Istmo,
Madrid, 481 pp.
ARAA, V.; ORTZ, R., 1984. Volcanologa. Ed. Rueda. Madrid, 510 pp.
AUBOUIN, J.; BROUSSE, R.; LEHMAN, J. P., 1981. Prcis de gologie: ptrologie. Bordas, Paris. Translated,
Tratado de geologa: petrologa. Ed. Omega, Barcelona, 602 pp.
CABRERA, M. C.; CUSTODIO, E.; ROQUE, F., 1997. Las aguas subterrneas y la planificacin hidrolgica en
las Islas Canarias. Asoc. Intern. Hidrogelogos, Grupo Espaol, 266 pp.
CABRERA, M. C.; NUEZ, J. A.; CUSTODIO, E., 1992. Contribucin al conocimiento geolgico del
subsuelo de Telde (Gran Canaria, Islas Canarias). Actas del III Congreso Geolgico de Espaa/
VII Congreso Latinoamericano de Geologa, Salamanca, Vol. 2, pp. 25660.
CARRACEDO, J. C.; DAY, S. J.; GUILLOU, H.; RODRIGUEZ-BADIOLA, E.; CANAS, J. S.; PREZ-TORRADO,
F. J., 1997. Origen y evolucin del volcanismo de las Islas Canarias. In: Ciencia y Cultura en
Canarias. Ed. Museo de la Ciencia y del Cosmos, Canarias, pp. 6789.
CEDEX, 1987. Simposio internacional sobre hidrologa de terrenos volcnicos. Arrecife de Lanzarote,
1974. CEDEX. Madrid, 1296 pp.
COURTEAUD, M., 1996. tude des structures gologiques et hydrogologiques du massif de La
Fournaise par la mthode audiomagnetotellurique. Doc. Thesis. Universit de la Runion.
Saint Denis, 205 pp.

423
Groundwater studies
CUSTODIO, E., 1978. Geohidrogeologa de terrenos e islas volcnicas. Instituto de Hidrologa/Centro de

Estudios Hidrogrficos, Pub. No. 128, Madrid, 303 pp.
, 1989. Groundwater Characteristics and Problems in Volcanic Rock Terrain. In: Isotope
Techniques in the Study of the Hydrology of Fractured and Fissured Rocks. STI/PUB/790, Intern.
, 1992. Coastal Aquifer Salinisation as a Consequence of Aridity: The Case or Amurga
Phonolitic Massif, Gran Canaria Island. Study and Modelling of Salt Water Intrusion.
CIMNE-UPC, Barcelona, pp. 8198.
, 2000. Utilizacin de la relacin Cl/Br como trazador hidrogeoqumico en hidrologa
subterrnea. Bol. Geolog. y Minero, Vol. 11, Madrid, pp. 4968.
CUSTODIO, E.; LLAMAS, M. R., 1983. Hidrologa subterrnea (2 vol). Ed. Omega, Barcelona, 2350 pp.
CUSTODIO, E.; LLAMAS, M. R.; SAMPER, J., 1997. La evaluacin de la recarga a los acuferos en la plani-
ficacin hidrolgica. Asoc. Intern. Hidrogelogos, Grupo Espaol-Inst. Tecnol. Geominero
de Espaa. Madrid. 455 pp.
DAVIS, S. N.; DE WIEST, R. J. M., 1966. Hydrogeology. Wiley, 463 pp.
DAVIS, S. N.; WHITTEMORE, D. O.; FABRYKA-MARTIN, J., 1998. Uses of Chloride/Bromide Ratios in
Studies of Potable Water. Ground Water, Vol. 36, pp. 33850.
DELMELLE, P.; BERNARD, A., 1994. Geochemistry, Mineralogy and Chemical Modelling of the Acid
Crater Lake of Kawah Ijen Volcano, Indonesia. Geochim. et Cosmochim. Acta, Vol. 58,
pp. 244560.
DUFF, P. MCL. D., 1993. Holmes Principles of Physical Geology. Chapman, 791 pp.
ELSWORTH, D.; CARRACEDO, J. C.; DAY, S. J., (eds.), 1999. Deformation and Flank Instability of
Oceanic Island Volcanoes: A Comparison of Hawaii with Atlantic Island Volcanoes. J.
Volcanology and Geothermal Research, Vol. 94, 340 pp.
FALKLAND, A.; CUSTODIO, E., 1991. Guide on the Hydrology of Small Islands. Studies and Reports in
Hydrology, No. 49, UNESCO, Paris, 435 pp.
FAYAS, J. A.; DOMNECH, J., 1974. Morfologa volcnica de Olot y su inters hidrolgico. Documentos
de Investigacin Hidrolgica, Centro de Estudios, Investigacin y Aplicaciones del Agua,
Barcelona.
FISHER, R. V.; SCHMINCKE, H.-U., 1984. Pyroclastic Rocks. Springer Verlag, 472 pp.
FRANCIS, P., 1993. Volcanoes: A Planetary Perspective. Oxford Univ./Clarendon Press, 435 pp.
HENRY, P.; GUY, C.; CATTIN, R.; DUDOIGNON, P.; SORNEIN, J. F.; CARISTAN, Y., 1996. A Convective
Model of Water Flow in Mururoa Basalts. Geochim et. Cosmochim. Acta, Vol. 60,
pp. 2087109.
HERRERA, CH.; CUSTODIO, E., 2000. Saline Water in Central Fuerteventura Island, Canary Islands,
Spain. XVI Salt Water Intrusion Meeting, Miedzyzdroje, Poland. (In press.)
HINDS, J. J.; GE, SH.; FRIDRICH, CH. J., 1999. Numerical Modelling of Perched Water under Yucca
Mountain, Nevada. Ground Water, Vol. 37, pp. 498504.
HUANG, K.; TZANG, Y. W.; BODVARSSON, G. S., 1999. Simultaneous Inversion of Air-Injection Tests
in Fractured Unsaturated Tuff at Yucca Mountain. Water Resources Research, Vol. 35,
pp. 237586.
HUNT ET AL., 1988. The Geology of North America. The Geological Society of America, Vol. 30,
pp. 25562.
JAWAHEER, K.; PROAG, V., 1988. Groundwater Aquifers of Mauritius. Hydrogologie, Vol. 2,
pp. 12534.
JOIN, J. L.; COUDRAY, J., 1992. Exemple dun coulement souterrain chenalis dans les brches
volcaniques peu permables: la ravine Saint-Gilles, le de la Runion. Hydrogologie, Vol. 3,
pp. 16372.
, 1993. Charactrisation gostructurale des mergences et typologie des nappes daltitude en
milieu volcanique insulaire (le de la Runion). Geodinamica Acta, Vol. 6, Paris, pp. 24354.

424
KILKARNI, H.; LALWANI, A.; DEOLANKAR, S. B., 1997. Selection of Appropriate Pumping Systems
for Bore Wells in the Deccan Basalt of India. Hydrogeology Journal, Vol. 5, pp. 75 81.
KULKARNI, H.; DEOLANKAR, S. B.; LALWANI, J. B.; PAWAS, S., 2000. Hydrogeological Framework of
the Deccan Basalt Groundwater Systems, West-Central India. Hydrogeology Journal, Vol. 8,
pp. 36878.
LOAGUE, K., 1994. Regional Scale Ground-Water Vulnerability Estimates: Impact of Reducing
Data Uncertainties for Assessments in Hawaii. Ground Water, Vol. 32, pp. 60516.
LOUVAT, P.; ALLGRE, C. J., 1997. Present Denudation Rates on the Island of Runion Determined
by River Geochemistry: Basalt Weathering and Mass Budget between Chemical and
Mechanical Erosion. Geochim. et Cosmochim. Acta, Vol. 61, pp. 364569.
MACDOUGALL, J. D., 1988. Continental Flood Basalts. Kluwer, 341 pp.
MIDDLEMOST, E. A. K., 1985. Magmas and Magmatic Rocks: An Introduction to Igneous Petrology.
Longman, 266 pp.
OKI, D. S.; GIAMBELLUCA, TH. W., 1989. Groundwater Contamination by Nematicides: Influence of
Recharge Timing under Pineaple Crop. Water Resources Bull., Vol. 25, pp. 28594.
PENNISI, M.; LEEMAN, N. P.; TONANINI, S.; PENNISI, A.; NABELEK, P., 2000. Boron, Sr, O, and H
Isotope Geochemistry of Groundwater from Mt. Etna (Sicily): Hydrologic Implications.
Geochim. et Cosmochim. Acta, Vol. 64, pp. 96174.
ROBINEAU, B.; RITZ, M.; COURTEAUD, M.; DESCLOITRES, M., 1997. Electromagnetic Investigation of
Aquifers in the Gran Brul Coastal Area of Piton de la Fournaise Volcano, Runion Island.
Ground Water, Vol. 35, pp. 58592.
ROBOCK, A., 2000. Volcanic Eruptions and Climate. Reviews of Geophysics, Vol. 38, pp. 191219.
ROSE, T. P.; DAVISSON, M. L., 1996. Radiocarbon in Hydrologic Systems Containing Dissolved
Magmatic Carbon Dioxide. Science, Vol. 273, pp. 136770.
SMELLIE, J. L., 1994. Volcanism Associated with Extension at Consuming Plate Margins. Geolog. Soc.
London, 292 pp.
SONNENTHAL, E. L.; BODVARSSON, G. S., 1999. Constraints on the Hydrology of the Unsaturated
Zone at Yucca Mountain, NV, from Three-Dimensional Models of Chloride and Strontium
Geochemistry. J. Contam. Hydrol., Vol. 38, pp. 10756.
TAKASAKI, K. J.; MINK, J. F., 1985. Evaluation of Major Dike-Impounded Ground-Water Reservoirs,
Island of Oahu. U.S. Geological Survey, Water-Supply Paper 2217, Washington D.C., 77 pp.
VIOLETTE, S.; LEDOUX, E.; GOBLET, P.; CARBONNEL, J. P., 1997. Hydrologic and Thermal Modelling of
an Active Volcano: The Piton de la Fournaise, Runion. J. of Hydrology, Vol. 191, pp. 3763.
WATTS, A. G.; PEIRCE, C.; COLLIER, J.; DALWOOD, R.; CANALES, J. P.; HEUSTOCK, T. J., 1997. A Seismic
Study of Lithospheric Flexure in the Vicinity of Tenerife, Canary Islands. Earth and
Planetary Sci. Letters, Vol. 196, pp. 43147.

425
kruseman-app.qxd 27/11/2003 12:09 Page 427
APPENDICES
Appendix A-1 Conversion factors for physical data
To convert To Multiply by
Grams Ounces (avoirdupois) 0.03527

Ounces (Avoirdupois) Grams 28.35
Tons per acre Metric tons per hectare 2.2417
Grains per gallon Milligrams per litre 17.12

Milligrams per litre Grains per gallon 0.05841
Milligrams per litre Tons per acre-foot 0.001360
Gallons (Imperial) Gallons (U.S.) 1.2009

Gallons (U.S.) Litres 4.55
Litres Gallons (U.S.) 0.219
Cubic-feet/second (cfs) Gallons (U.S.) per minute 448.8
Cubic-feet/second/day* Acre-feet 1.983471
Cubic-feet/second/day Gallons (U.S.) per day 646,317
Acre-feet Gallons (U.S.) 325,851
Acre-feet Cubic feet 43,560
Acre-feet Cubic metres 1,232
Cubic feet Cubic metres 0.028317
Cubic feet Gallons (U.S.) 7.481
* 1 sec-ft day = 1cfs for 24 hr.

427
Groundwater studies
Appendix A-2 Conversion factors for hydrochemical data
milligrams/litre (mg/l) * F1 = milliequivalents/litre (meq/l);

milligrams/litre * F2 = millimoles/litre (mmol/l).
Element and Element and

reported species F1 F2 reported species F1 F2
Aluminium (Al+++) 0.11119 0.03715 Lead (Pb) 0.00483

Ammonium (NH4+) 0.05544 0.05544 Lithium (Li+) 0.14411 0.14411
Barium (Ba++) 0.01456 0.00728 Magnesium (Mg++) 0.08226 0.04113
Beryllium (Be++) 0.22192 0.11096 Manganese (Mn++) 0.03640 0.01820
Bicarbonate (HCO3) 0.01639 0.01639 Molybdenum (Mo) 0.01042
Boron (B) 0.09250 Nickel (Ni) 0.01703
Bromide (Br ) 0.01251 0.01261 Nitrate (NO3) 0.01613 0.01613
Cadmium (Cd++) 0.01779 0.00890 Nitrite (NO2) 0.02174 0.02174
Calcium (Ca++) 0.04990 0.02495 Phosphate (PO4 ) 0.03159 0.01053
Carbonate (CO3 ) 0.03333 0.01666 Phosphate (HPO4 ) 0.02084 . 0.01042
Chloride (Cl) 0.02821 0.02821 Phosphate (H2PO4) 0.01031 0.01031
Chromium (Cr) 0.01923 Potassium (K+) 0.02667 0.02557
Cobalt (Co++) 0.03394 0.01697 Rubidium (Rb+) 0.01170 0.01170
Copper (Cu++) 0.03148 0.01574 Silica (SiO2) 0.01664
Fluoride (F) 0.05264 0.05264 Silver (Ag) 0.00927
Germanium (Ge) 0.01378 Sodium (Na+) 0.04350 0.04350
Gallium (Ga) 0.01434 Strontium (Sr++) 0.02283 0.01141
Gold (Au) 0.00511 Sulphate (SO4 ) 0.02082 0.01041
Hydrogen (H+) 0.99209 0.99209 Sulphide (S ) 0.06238 0.03119
Hydroxide (OH) 0.05880 0.05880 Titanium (Ti) 0.02088
Iodide (I) 0.00788 0.00788 Uranium (U) 0.00420
Iron (Fe++) 0.03581 0.01791 Zinc (Zn++) 0.03060 0.01530
Iron (Fe+++) 0.05372 0.01791

428
Appendices
Appendix A-3 List of authors
Aggarwal, P. (Chapter 9)
Isotope Hydrology Section of the International Atomic Energy Agency (IAEA)
Wagrammer strasse 5, P.O. Box 100, A-1400 Vienna, Austria
Tel: +43 1 2600-21736; Fax: +43 1 26007
Araguas-Araguas, L. (Chapter 9)
Tel: +43 1 2600-21736; Fax: +43 1 26007
Becht, Robert (Chapter 5)

International Institute for Aerospace Surveys and Earth Sciences (ITC)
P.O. Box 6, 7300 AA Enschede, The Netherlands
Tel. +31 53 4874 444, Fax:+31 53 4874 400, e-mail: [email protected]
Custodio, Emilio (Chapter 16)

Department of Ground Engineering and Mapping,
International Centre for Groundwater Hydrology,
Universitat Politcnica de Catalunya,
Gran Capitn s/n, Ed. D-2, 08034 Barcelona, Spain
Tel: +93 401 6866; Fax: +93 401 7251
Froehlich, S. (Chapter 9)
Tel: +43 1 2600-21736; Fax: +43 1 26007
Griffioen, Jasper (Chapter 2 and Chapter 8)

Netherlands Institute of Applied Geoscience-TNO,
P.O. Box 6012, 2600 JA Delft, The Netherlands
Tel: +31 15 269 7164, Fax: +31 15 256 4800; e-mail: [email protected]
Garcia-Agudo, E. (Chapter 9)
Tel: +43 1 2600-21736; Fax: +43 1 26007
Hodgson, Frank D. I. (Chapter 4)

Institute for Groundwater Studies, University of the OF S
P.O. Box 339, Bloemfontein, 9300 South Africa
Fax: +27 51 401 9111
Kovalevsky, Vlademir S. (Chapter 11 and Chapter 13A)

Institute of Water Problems
3 Gubkina str., 117971 Moscow, Russia
Fax: +7 095 135 5415; e-mail: [email protected]

429
Groundwater studies
Kruseman, Gideon P. (Chapter 1 and Chapter 8)

Netherlands Institute of Applied Geoscience-TNO
P.O. Box 6012, 2600 JA Delft, The Netherlands
Tel: +31 30 234 0695; Fax: +31 15 2564800; e-mail: g.p.kruseman@ planet.nl
Kulkarni, K. M. (Chapter 7)
Bhabha Atomic Research Centre
Mumbai, India
LaMoreaux, Philip E. (Chapter 14)

P.O. Box 2310, Tuscaloosa, AL 35403, United States of America;
Tel: +1 205 752 5543; Fax: +1 205 391 3534; e-mail: [email protected]
Limaye, S. D. (Chapter 15)

Director, Groundwater Institute
2050, Sadashiv Peth, Poona 411030, India
E-mail: [email protected]
Margat, Jean (Chapter 10 and Chapter 12)

BRGM
P.O. Box 6009, Avenue C. Guillemin, 45060 Orlans Cedex, France
Tel: +33 2 38 64 32 72; Fax: +33 2 38 64 39 48
Rebouas, Aldo C. da (Chapter 7)

Centre for Ground Water Research, University of So Paulo, Brazil
Tel: +55 11 268 2862; Fax: +55 11 869 0483; e-mail: [email protected]
Rushton, Ken R. (Chapter 1 and Chapter 11)

Dept. of Civil Engineering, University of Birmingham
Edgbaston, Birmingham B15 2TT, United Kingdom
Sporry, Robert (Chapter 6)

International Institute for Aerospace Surveys and Earth Sciences (ITC)
P.O. Box 6, 7300 AA Enschede, The Netherlands
Tel: +31 53 4874 444; Fax:31 53 4874 400; e-mail: [email protected]
Van Lanen, Henny A. J. (Chapter 3)

Department of Environmental Sciences, Agricultural University
Nieuwe Kanaal 11, 6709 PA Wageningen, The Netherlands
Tel: 31 317 482 778; Fax: 31 317 484 885; e-mail: [email protected]
Vrba, J. (Chapter 13B)

Korandova 32, 14700 Prague 4, Czech Republic
Tel/Fax: 420 2 472 7447; e-mail: [email protected]
Yuan Daoxian (Chapter 14)

The Institute of Karst Geology, CAGS
No. 40 Qixing Road, Guilin-Guangxi, China
Tel +86 773 581 3179; Fax: +86 773 581 3708
e-mail: [email protected]

430

Groundwater Studies UNESCO

Uploaded by

Copyright:

Available Formats

Groundwater Studies UNESCO

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Groundwater Studies UNESCO

Uploaded by

Copyright:

Available Formats

krus-couv-avec-dos.

qxd 27/11/2003 11:35 Page 1

IHP-VI, SERIES ON GROUNDWATER NO. 3

IHP-VI, SERIES ON GROUNDWATER NO. 3

The designations employed and the presentation

Published in 2004 by the United Nations

Layout and typesetting by Marina Rubio 93200 Saint-Denis

1 Occurrence of groundwater, regime and dynamics 19

2.4.2 Characterisation of groundwater quality 51

3 Groundwater networks and observation methods 67

4 Processing and presentation of data 95

5 Remote sensing techniques for groundwater prospection 121

6 Geophysical techniques in groundwater investigations 133

7 Well drilling and design methods 185

7.5 Well drilling techniques 187

8 Determining hydrodynamic and contaminant transfer

9 Nuclear techniques in groundwater investigations 239

10 Hydrogeological mapping 271

11 Assessment of groundwater resources and groundwater

11.2.4 Important questions when identifying resources 301

12 Groundwater management 319

13 (A) The influence of changes in hydrogeological conditions

14 Hydrogeology of carbonate rocks 363

15 Hydrogeology of hard rocks 383

15.2 Occurrence of groundwater 384

16 Hydrogeology of volcanic rocks 395

Fig. 3.3 Hydrological variables to be monitored 72

Fig. 6.6 Principle of the Slingram EM technique 143

Fig. 8.7 Time drawdown and residual drawdown 228

1.1 Hydrological cycle

Figure1.1a The hydrological cycle (After De Wiest,1965)

Occurrence of groundwater, regime and dynamics

1.2 Groundwater flow

1.2.1 Groundwater flow processes and continuity

1.2.2 Groundwater head

Occurrence of groundwater, regime and dynamics

1.2.3 Darcys Law, hydraulic conductivity and transmissivity

1.2.4 Inflows and outflows

1.2.5 Storage coefficients and time dependency

1.3 Groundwater composition

1.3.1 Physical and chemical properties

Occurrence of groundwater, regime and dynamics

1.3.2 Risk of groundwater pollution

may pollute groundwater are grouped according to their physico-chemical characteristics in

(iii) Dissolved organics

Occurrence of groundwater, regime and dynamics

(iv) Non-aqueous phase liquids

(v) Colloids and radionuclides

Contamination associated with nuclear activities deserves special attention. Radionuclides

(vi) Bacteria and viruses

Occurrence of groundwater, regime and dynamics

1.4 Groundwater assessment and exploration

1.4.1 Aquifer and groundwater systems

(i) Aquifer system

The aquifer system comprises:

The transmissivity for vertical flow through an aquitard is:

Tv = Kv/M [T 1]; (1.6a)

c = M/K v [T] (1.6b)

Occurrence of groundwater, regime and dynamics

(ii) Groundwater system