BG 50 - 2013 Water Treatment For Closed Heating and Cooling Systems

Licensed copy from CIS: southbank, Southbank University, 06/03/2018, Uncontrolled Copy.
A single copy of this document is licensed to
southbank
On
06/03/2018
This is an uncontrolled copy. Ensure use of the

most current version of the document by searching
the Construction Information Service.
56GF=5;i]XY lll#Whg^V#Xd#j`
KUhYfHfYUhaYbhZcf7`cgYX
<YUh]b[UbX7cc`]b[GmghYag
:YîZYWnGZ\^cVaY7gdlc
5^c]bhjYbhifYk]h\"""
7<*%$'%&(
1
ACKNOWLEDGEMENTS
The technical information in this publication was provided by members of

a steering group representing BSRIA, the British Association for Chemical
Specialities, the Water Management Society, the Commissioning Specialists
Association, the Institute of Corrosion and the United Kingdom Water
Treatment Association.
The steering group was chaired by Alan Edwards representing the Water
Management Society, and the editor for the publication was Reginald Brown
of BSRIA. Design and production was carried out by Joanna Smith of BSRIA.
BSRIA wishes to thank all the members of the steering group for their
contributions:
David Bleicher BSRIA

Reginald Brown BSRIA
Jason Bruce Commissioning Specialists Association
Tony Collins Institute of Corrosion
Jill Cooper Water Management Society
Daniel Davies United Kingdom Water Treatment Association
Liz Day Commissioning Specialists Association
Alan Edwards Water Management Society
Mike Hunter Water Management Society
Mike Iddon Water Management Society
John Lane Water Management Society
Stewart McGillivray Institute of Corrosion
Phil Munn Institute of Corrosion
Pamela Simpson Institute of Corrosion
John Smith British Association for Chemical Specialities
Jonathan Usher Water Management Society
Geoff Walker British Association for Chemical Specialities
Stuart Wilton Water Management Society
Every opportunity was taken to incorporate the views of the steering group,
however final editorial control of the publication rested with BSRIA.
BSRIA acknowledges with thanks the organisations that provided additional

images for use in this publication:
Midland Corrosion Services Ltd.

Chesterfield WT Consultants Ltd.
B & V Water Treatment
The guidance given in this publication is correct to the best of BSRIA’s knowledge. However BSRIA
cannot guarantee that it is free of errors. Material in this publication does not constitute any warranty,
endorsement or guarantee by BSRIA. Risk associated with the use of material from this publication is
assumed entirely by the user.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means electronic or mechanical including photocopying, recording or
otherwise without prior written permission of the publisher.
© BSRIA October 2013 ISBN 978-0-86022-724-3 Printed by Charlesworth Press
WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

© BSRIA BG 50/2013
PREFACE
The treatment of water in modern closed heating and cooling systems

is essential for the avoidance of microbiological fouling (biofouling),
corrosion and scale. These problems can result in energy wastage, poor
system performance, and the need for early replacement of plant and
components. The consequences of inappropriate or non-existent water
treatment can sometimes be disastrous.
This guide is intended for use by design engineers, installing contractors

and the maintenance staff responsible for looking after the completed
systems. It provides an introduction to current theory and practice
of water treatment in closed systems including minimising the risk of
corrosion through system design features and proactive monitoring.
In particular the guide will help facilities managers and others to engage in
constructive discussion with water treatment professionals and choose the
most appropriate water treatment programme for their systems.
The common causes of water quality and corrosion problems are

explained and their implications for closed heating and cooling systems
are described. Consideration is given to design, system operation, routine
control, treatment and monitoring of water.
The guide partially replaces BSRIA AG 2/93 Water treatment for building
services systems in respect of closed system applications. The guidance is
consistent with BSRIA BG 29/2012 Pre-commission cleaning of pipework
systems, BS 8552:2012 Sampling and monitoring of water from building
services closed systems. Code of practice and the European Biocidal products
Regulation (528/2012, commonly known as BPR).

© BSRIA BG 50/2013
CONTENTS
1 INTRODUCTION 1
2 SYSTEM DESIGN AND OPERATION 3

2.1 Choice of materials 3
2.2 Flow rates 5
2.3 Minimising dissolved oxygen 6

2.4 Design for water treatment 8
2.5 Pre-commission flushing and cleaning 13
2.6 Remedial flushing and cleaning 14
3 CORROSION PROCESSES 21
3.1 Chemical corrosion processes 21
3.2 Microbially influenced corrosion (MIC) 31
4 CHEMICAL WATER TREATMENT 35

4.1 Corrosion inhibition 35
4.2 Bacteria and biofouling inhibition 44
4.3 Scale inhibition 49
5 NON-CHEMICAL WATER TREATMENT 50

5.1 Filtration 50
5.2 Deaeration 53
5.3 Sacrificial anodes 56
5.4 Other non-chemical treatments 56
6 MANAGEMENT OF THE WATER TREATMENT

PROGRAMME 59
6.1 Overview 59
6.2 Water sampling and analysis 60
6.3 Sampling for water chemistry 61
6.4 Sampling for microbiology 63
6.5 Water quality analysis and guidelines 69
6.6 Corrosion Monitoring 74
7 REGULATIONS 78
7.1 Health and safety 78
7.2 Other regulations 79

© BSRIA BG 50/2013
APPENDICES
APPENDIX: A USE OF TEMPORARY PUMPS 81

APPENDIX: B CASE STUDIES 83
APPENDIX: C BIOCIDES 94
APPENDIX: D QUANTITATIVE MEASUREMENT OF BIOFILM 98
APPENDIX: E ANTI-FREEZE 99
GLOSSARY 101
REFERENCES AND BIBLIOGRAPHY 103

© BSRIA BG 50/2013
INTRODUCTION 1
1 INTRODUCTION
The scope of this guide is water treatment for the control of corrosion
and maintenance of water quality in closed heating and cooling systems in
buildings other than individual dwellings. This includes low temperature
hot water heating systems, heat pump and chilled water systems and
condenser water circuits. It is also relevant to community heating and
cooling networks operating at temperatures below 110°C.
The scope does not include the specific requirements for water treatment
for medium and high temperature heating systems, for which reference
should be made to BS 2486[3].
This scope does not include water treatment for open systems such as
cooling towers and steam boilers or for domestic hot and cold water
services.
The objectives of a water treatment programme in closed heating and

cooling systems are to:
s maintain the system in a clean condition
s assist in maintaining system efficiency
s prolong system life
through the control of corrosion and fouling.
This is aided by the following elements:

1. System design
a. Specification of appropriate materials
b. Avoidance of dead legs and areas of stagnation
c. Operating characteristics that minimise the risk of air ingress and
problems associated with low flow, such as sedimentation and
biofouling
Further information on system design is provided in Section 2.
2. Installation, testing & pre-commission cleaning
Further information on these topics is provided in Section 3 and BSRIA
BG 29[1].
3. Application of a correct and appropriate water treatment programme
including:
a. Control of corrosion by:
i. Application of corrosion inhibitors
ii. Control of scale, biofouling & other deposits
b. Control of scale by:
i. Minimising system losses & make-up
ii. Use of scale inhibitors
iii. Pre-treatment of make-up water when necessary
WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS 1

© BSRIA BG 50/2013
1 INTRODUCTION
c. Control of biofouling/biofilms by:

i. Avoidance of low flow
ii. Use of appropriate biocides
iii. Control of nutrient ingress
iv. Pre-treatment of make-up water when necessary
v. Control of scale, corrosion and sedimentation
d. Control of sedimentation by:
i. Avoidance of low flow

ii. Use of dispersants
iii. Use of side-stream filtration
iv. Pre-treatment of make-up water when necessary
v. Control of scale, precipitation, biofouling & corrosion
Further information on these topics is provided in Sections 4 and 5.
4. Management of the programme including:
a. Knowledge of the system including:
i. Materials used
ii. Operational characteristics (temperature, flow rates, etc.)
iii. Volume
b. Programme design including
i. Product selection
ii. Sampling & testing regime
iii. Guidelines and action limits etc.
c. Implementation of programme
d. Monitoring & control
e. Training & competence of staff involved
f. Definition of reporting lines, roles & responsibilities
g. How to handle/report problems/areas of concern
Further information on these topics is provided in Section 6.
Section 7 briefly summarises the main legislation that is relevant to water

treatment activities.
Various technical issues are discussed in detail in the appendices and there
is a glossary and bibliography at the end of this guide.
2 WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

© BSRIA BG 50/2013
SYSTEM DESIGN AND OPERATION 2
2 SYSTEM DESIGN AND OPERATION
There are many factors in the design of closed heating and cooling systems
that can influence corrosion. It is impossible to completely eliminate the
corrosion of ordinary metals in such systems, but it is possible to control
the corrosion rate to tolerable levels.
Issues that should be considered by the system designer include:

s choice of materials.
s flow rates and control strategy.
s minimising dissolved oxygen.
s filling, flushing and cleaning prior to use.
s chemical and physical water treatment.
s monitoring and maintenance.
The benefits of good design can be wasted if installation, commissioning

or maintenance practices are poor or ineffective. Advice on provision of
facilities for flushing and cleaning of systems prior to commissioning can
be found in Section 2 of BSRIA BG 29/2012 Pre-commission cleaning of
pipework systems[1].
2.1 CHOICE OF A wide range of well-understood materials are used in closed water
MATERIALS
systems. These include low carbon steel, stainless steel, galvanised steel, cast
iron, aluminium, copper and copper alloys, elastomers and plastics.
Metals are used for distribution pipework and equipment including heat
exchangers, pumps, valves, radiators and other heat emitters. Elastomers
are used as seals including “O-rings” in valves and fittings. Plastics are
used for pipework (PVC, ABS, HDPE etc.) and individual components of
fittings and equipment (acetal, nylon, PTFE etc.).
Steel, copper alloys and aluminium are suitable for closed heating and
cooling systems as the corrosion rate should be limited by the low
availability of dissolved oxygen in the water and the appropriate use of
corrosion inhibitors. However, there are other factors that may influence
corrosion. For example:
s Mild steel, stainless steels and copper may be prone to microbially
influenced corrosion (MIC).
s Heating and cooling systems contain different metals that when
joined together may induce galvanic corrosion in the presence of
dissolved oxygen.
s Brasses may be subject to dezincification.
s Aluminium and stainless steels are prone to pitting corrosion in high
chloride environments.

© BSRIA BG 50/2013
Although non-metallic components such as seals, diaphragms and linings

are not subject to oxygen corrosion they may be subject to chemical attack
causing swelling, softening, cracking etc. This is important as degradation
over time may result in leaks or restriction of flow or allow oxygen to
enter the system.
All materials, be they metallic or non-metallic, should have industry

approval for use in heating and cooling systems and be compatible with
current water treatment chemistry. Table 1 summarises the corrosion

issues of materials used in closed water systems. The corrosion processes
affecting different metals in closed water systems are discussed in Section
3.1 and BS EN 14868[10].
Table 1: Corrosion properties of frequently used materials in closed water systems

Material Where used Corrosion resistance Other issues
Aluminium Some boiler heat exchangers Good overall corrosion resistance Exposure to high pH causes
and radiators in oxygenated waters of neutral or rapid loss of metal and
slightly alkaline pH. Should not be formation of aluminium
exposed to pH greater than 8.5 hydroxide sludge. Susceptible
to pitting corrosion at
moderate chloride or sulfate
concentrations
Copper and copper alloys Copper tube, brass valves and Good overall corrosion resistance in Copper ions entering the water
fittings water of neutral or moderately alkaline can result in pitting corrosion of
pH. steel and aluminium. Brass can
be subject to stress corrosion
In oxygenated water copper is subject cracking when exposed to
to attack from turbulent erosion, flux ammonia.
residues and under-deposit corrosion.
Mild steel and cast iron Steel pipe, boiler heat Low concentrations of dissolved Formation of insoluble iron
exchangers, circulating pumps oxygen result in uniform corrosion and oxides as suspended solids
the production of magnetite sludge. increases wear in pumps and
High concentrations of dissolved the risk of under-deposit
oxygen result in pitting attack under corrosion in low flow areas
tubercles. where sedimentation occurs.
Galvanised steel Some piping systems Internally galvanised pipes and fittings Formation of zinc hydroxide as
should not be used in heating systems suspended solids.
(see Section 3.1.2)
Stainless steel (SS) Plate heat exchangers, Very good resistance to general
pump casings, minor parts. corrosion but may be susceptible to
Occasionally pipework. pitting, crevice corrosion and stress
corrosion cracking at high chloride or
sulfate concentrations.
Plastic Plastic pipe including underfloor Resistant to corrosion but may be Oxygen permeation through
heating. Minor parts. subject to physical degradation for plastic pipe. Pressure resistance
example by sunlight. decreases with temperature.
Rubber Flexible hose liners (EPDM), Resistant to corrosion but may be Amenable to the formation of
O-rings and seals. subject to gradual chemical and biofilm.
physical degradation leading to loss of
flexibility and cracking.

© BSRIA BG 50/2013
2.2 FLOW RATES 2.2.1 Flow velocity

Flow rates (litres per second) are set by the system designer to satisfy the
requirement for efficient heat transfer in heat exchangers and coils and to
achieve the required distribution of heating and cooling in the building.
However it is the flow velocity rather than flow rate that influences the
risk of corrosion.
In oxygenated waters, the uniform corrosion rate will generally increase

with flow velocity but this is less important than what happens at very low
or very high flow velocity.
If the water velocity is too low then settling of solids may occur. This is
mainly an issue in small bore pipework and terminal units, such as fan coils
and radiators, at the extremities of the building. Accumulation of deposits
increases the likelihood of under-deposit corrosion as well as biofouling
and MIC.
If the velocity is too high then there is a risk of turbulent erosion.

However, for this to occur there also needs to be significant oxygenation
of the system water. Turbulent erosion is common in poorly balanced
domestic water systems but rarely encountered in heating and cooling
systems at normal design flow rates. Turbulent erosion is not the same as
cavitation, which is a pressure-related phenomenon that does affect closed
systems.
2.2.2 Circulation
It is essential to avoid prolonged stagnation in heating and cooling systems.
Stagnation will allow the sedimentation of suspended solids and promote
the formation of biofilms, increasing the risk of microbially influenced
corrosion. Also, most inhibitors work more effectively under flowing
rather than stagnant water conditions.
The system control strategy should therefore ensure that circulation pumps
and control valves are periodically operated to reduce the risk of stagnation
and ensure full distribution of all water treatment chemicals.
The minimum recommended period of circulation for heating and cooling

circuits is one hour per day. Ideally this should be programmed into the
building management system, where one exists. It is not necessary to
operate boilers and chillers during this period.
Water sampling should not be carried out in no-flow conditions, as this

could influence the results obtained.

© BSRIA BG 50/2013
2.3 MINIMISING O2 to OH−) is the primary reaction that drives anodic dissolution and
DISSOLVED OXYGEN
therefore corrosion. Control of dissolved oxygen is therefore a key aspect
of controlling corrosion.
The most important way of controlling corrosion in use is to ensure that

the systems are tight and do not allow ingress of oxygen, either through
excessive additions of make-up water or inadequate pressurisation.
Systems should be designed with adequate numbers of air vents at high

points of the system and drain valves at low points. Air pockets persisting
in low flow areas of horizontal pipework and where there is inadequate
venting can lead to waterline pitting of partially flooded pipes. There
should also be a sufficient number of isolating valves at strategic locations
to allow inspection and/or replacement of system components without
having to drain large parts of the system. This reduces the exposure of
the pipework to atmospheric air when drained and fresh water containing
oxygen when refilled.
BS EN 14868[10] describes the different ways that oxygen can enter a

system and the effect on the likelihood of corrosion of different metals and
alloys used in these systems.
In traditional closed heating systems constructed using mild steel pipe,

oxygen that enters the system dissolved in the fill water is evolved as a
gas during heating (and vented) or consumed by chemical reaction with
the large area of steel surface. The loss of metal from general corrosion of
the steel pipe due to the dissolved oxygen is not significant. If the system
contains relatively little steel or other corrodible material compared to
the system volume then the dissolved oxygen concentration can remain
high indefinitely, resulting in rapid corrosion of the remaining steel
unless adequately protected by the water treatment programme. Steel
radiators are particular at risk of failure due to pitting corrosion in these
circumstances.
In closed cooling systems, oxygen concentrations may remain high after

filling as there is no thermal evolution of the dissolved oxygen from the
fill water. However, the corrosion rate is also slower at the lower operating
temperature and there may be no radiators or other thin steel components
to worry about.
The risk of oxygen corrosion can be minimised by:

s good standards of design, construction and commissioning to limit
air ingress and persistence.
s inclusion of physical devices to remove oxygen e.g. vacuum
deaerators (see section 5.2).
s adding chemical corrosion inhibitors to passivate the metal surface
or otherwise interfere with the corrosion reactions (see Section 4.1).

© BSRIA BG 50/2013
The key recommendations to designers are:

s Design pipework systems in accordance with the recommendations
of CIBSE Code W[2].
s Make sure that air can easily be vented from all parts of the system.
s Ensure that a positive pressure exists at all times under all operating
conditions in all parts of the system.
s Ensure that there is sufficient expansion capacity.
s Ensure that plastic pipe that is prone to oxygen diffusion (such as

polyethylene and polybutylene) is specified with an effective oxygen
barrier.
2.3.1 Oxygen barriers in plastic pipe

One route by which oxygen can enter a closed heating or cooling system
is through the use of non-barrier polyethylene (PE-X) or polybutylene
(PB) plastic pipe. Oxygen can diffuse through the wall of a non-barrier
plastic pipe (irrespective of the thickness and hydrostatic pressure) because
the partial pressure of oxygen is much higher outside the pipe than inside.
Barrier plastic pipework (for PE-X and PB) became available in the late
1980s. This barrier is either a plastic film on the inner surface of the
pipe that reduces oxygen diffusion or a concentric aluminium layer in
the middle of the pipe wall (composite pipe) that prevents it. In order to
maintain a continuous barrier throughout the system it is important that
the pipe is used with the fittings recommended by the manufacturer.
The use of non-barrier plastic pipework for heating systems is still allowed
in BS 7291[5]. If it is used then it is crucial to maintain the correct level of
a corrosion inhibitor to control corrosion.
2.3.2 Removal of air and dissolved gases

All closed systems should be designed so that displaced air can easily be
removed from the system during filling.
The fill water also contains dissolved air (mostly oxygen and nitrogen)
that will be evolved from solution if the water is heated. Automatic air
vents allow gases evolved during the initial operation of heating systems, or
created by subsequent chemical or biological processes, to be vented from
high points of the system. Methods to accelerate deaeration of system
water are described in Section 5.2.
2.3.3 Pressurisation of systems

Closed systems must be appropriately pressurised to operate at optimum
performance and minimise the risk of corrosion. Insufficient pressurisation
risks air being drawn into the system through seals, gaskets and vents.
Over-pressurisation risks water loss through pressure relief valves as the
system heats up.
Maintenance of adequate pressure will also reduce the risk of cavitation. If

the pressure for that temperature of the water passing through a valve drops
below the saturated vapour pressure for that temperature, microscopic

© BSRIA BG 50/2013
steam bubbles can be formed. This is known as cavitation and these

bubbles can cause considerable damage particularly to brass valve seats as
they implode on the metal surface.
However, increasing the pressure more than necessary could be

counterproductive. Increased pressure will increase the solubility of gases
such as oxygen and carbon dioxide and if these are retained in the system
rather than being vented they will increase the corrosion rate.
2.3.4 Expansion vessels

Water expands by some 3% when heated from 15°C to 82ºC (the design
temperature for conventional LTHW systems). In closed systems this
expansion is accommodated by an expansion vessel or expansion system.
The design process should:

s calculate the pressurisation unit and pressure relief valve settings
to maintain a positive pressure throughout the system under all
operating conditions.
s calculate the required expansion volume from the total system
volume and operating temperature range.
s size the expansion vessel, calculate the bladder pressure and locate it
in the system according to the methodology in BS 7074-2[4].
These steps should ensure that the maximum design pressure of the
heating system will not be exceeded during heating up and negative
pressure will not develop in any part of the system during cooling down.
Actual system pressure should be checked during commissioning and the
pressurisation settings adjusted upwards if necessary.
2.4 DESIGN FOR Most closed heating and cooling systems rely on chemical water treatment
WATER TREATMENT
for the control of corrosion. This section describes the system features
necessary to implement a chemical water treatment programme. The
chemicals that may be employed are discussed in Section 4.
2.4.1 Fill water pre-treatment

Closed heating and cooling systems are usually filled with mains water,
though borehole water is sometimes used. Mains water complies
with drinking water quality standards and should therefore be free of
contamination from dirt or debris. However, there are other characteristics
that may render the water undesirable for use within the system. Pre-
treatment should be considered for any fill water that does not comply
with the recommendations of BG 29[1]. In particular pre-treatment may
be required for:
s very hard water (particularly if the intention is to use a phosphate
inhibitor programme).
s water containing a high concentration of dissolved solids
(particularly chlorides and sulfates).
s water with undesirable microbiological activity, for example with
high concentrations of pseudomonads or sulfate reducing bacteria
(SRB).

© BSRIA BG 50/2013
Pre-treatment systems available can be broadly classified as:

s Water conditioning:
- Water softening
- Filtration
- Automated chemical dosing (proportional to the metered
volume or controlled by redox potential)
s Microbiological control:
- Ultraviolet disinfection
- Reverse osmosis (provides a barrier to bacteria)
- Automated biocide dosing
Water softening
Where the available water is particularly hard, a softening plant may be
temporarily installed to fill the system with reduced hardness water. It is
not recommended to install a permanent softening plant solely for fill
water as that will be underused and may become a source of bacterial
contamination.
The fill water should not be fully soft (complete removal of calcium) as
such waters are non-scale forming and, in theory, can be more corrosive
than non-softened waters, particularly for copper. This risk can be offset
by appropriate corrosion inhibitors e.g. azoles for copper protection.
Base-exchange softening replaces scale forming calcium and magnesium

ions with sodium ions. If base-exchange softened water is used to fill the
system, it is important to check that all inhibitors are suitable for use with
the softened water.
Deionisation (utilising resin beds) is similar to base-exchange softening but

removes both the scale forming and corrosive ions from the water and thus
reduces the total dissolved solids (TDS) and conductivity. Deionisation of
fill and make-up water is more expensive than base-exchange softening
and generally restricted to high temperature and/or high heat-transfer
heating systems. Deionised water is rarely used to fill a closed heating
and cooling system but could be helpful where there is a high sulfate
or chloride concentration in the raw water. Again, the water treatment
package should be tailored to the use of deionised water.
Softened and deionised water have reduced buffering capacity so chemical

pH buffers may need to be added as part of the water treatment package to
maintain the pH at the required level.
Softening is not normally required for make-up water for closed systems
as there should be minimal make-up during normal operation and the
consequent addition of scale formers should be insignificant compared to
the volume of the system. Some water treatment practitioners do however
favour the use of blended softened water for both filling and topping
up where it is already available on site (usually for hot and cold water
services).

© BSRIA BG 50/2013
Filtration and reverse osmosis

In principle filtration of mains water should not be necessary but strainers
or coarse filters are routinely used as protection against large debris and
scale for pumps and control valves. Filters are discussed in Section 6.1.
Mains water is not sterile but contains a wide range of bacteria that are
harmless to people but may become implicated in corrosion processes.
Low pressure membrane filtration is increasingly being used to provide a
physical barrier to bacteria and fungi entering the system from the water
supply. This is claimed to reduce the need for biocides for subsequent
microbiological control.
Certain types of high pressure reverse osmosis (RO) membrane will

also reduce the concentration of scale forming ions as an alternative to
softening or deionisation. This may be useful on borehole supplies of hard
water. RO softened water may need the addition of pH buffers to maintain
pH control as for other softened water.
Ultraviolet (UV) disinfection

Ultraviolet treatment of the fill water is also used as a precautionary
measure for water supplies with high bacteria counts. Ultraviolet
treatment does not provide an absolute barrier to bacteria but can
significantly reduce the number of bacteria entering the system.
Conventional UV treatment cells are not recommended for treating the

makeup water for closed systems as makeup should be infrequent and
insignificant in volume.
Automated proportional dosing

Automated proportional dosing of water treatment chemicals and biocides
may be used for the initial fill water and as part of pre-commission
cleaning.
Automated proportional dosing is not recommended for treating

makeup water for closed systems as the makeup should be infrequent
and insignificant in volume. Maintenance additions of water treatment
chemicals and biocides should be added via dosing pots or dosing pumps
under manual control.
2.4.2 Solids removal and filtration

The design of pipework systems to enable flushing and cleaning and the
methodology of pre-commission cleaning is discussed in Section 2.5 and
BG 29[1].
Strainers
Strainers should be fitted in accordance with design codes to protect
sensitive equipment and heat exchangers. Strainers may also be fitted
where secondary circuits are fed from the primary circuit. The appropriate
strainer size and media must be selected according to the design flow rates
and maximum allowable pressure drops. Importantly, strainers must be
isolable and easily accessed for inspection and routine maintenance.

© BSRIA BG 50/2013
Side stream filtration

Side stream filtration can be used to help maintain a low concentration
of suspended solids in the circulating water following pre-commission
cleaning and reduce the risk of solids deposition and under-deposit
corrosion in areas of low flow. In addition to the longer term benefits of
early installation of side stream filtration, it can also help to maintain water
quality during the commissioning period.
In side stream filtration a small proportion of the circulating flow is passed

through the filter on a continuous basis. If however the filter becomes
blocked then the risk of corrosion can be increased due to bacterial
colonisation. Filters should therefore be fitted with pressure differential
indicators, frequently checked and properly maintained.
Various filtration media and bespoke filter units can remove suspended
solids and some microorganisms. Particular guidance on filtration
technologies can be found in Section 5.1.
2.4.3 Chemical water treatment

The chemical dosing facility should be sized relative to the system volume
to allow sufficient amounts of treatment chemicals to be added efficiently.
For small systems a dosing pot is used for maintenance treatments. This
is a flow-through container, plumbed in parallel with part of the circuit.
It is fitted with valves so that it can be isolated, drained and filled with
chemical at atmospheric pressure, then reconnected to the circuit (see
Figure 1).
Figure 1: Dosing pot
Tundish
Safety check valve
Vent
Inlet
Outlet
Serial
number
plate
Drain

© BSRIA BG 50/2013
It is important that the dosing pot should remain open to the system
with continuous through-flow so that it does not become a dead leg and
encourage the growth of bacteria. One manufacturer now combines the
function of dosing pot and strainer.
Larger volumes of chemicals (for the initial dosing of small systems or

maintenance dosing of larger systems) can be pumped into the system at
pressure using chemical dosing pumps.
Designers should consider not only the dosing arrangement but

arrangements for the safe storage of chemicals near to the dosing point.
2.4.4 Sampling points and monitoring equipment

There should be sufficient sample points located throughout the system
to allow it to be properly monitored and managed. For most systems
the drain points will provide the majority of potential sample points (see
Section 2.4.5). Pressure test points (Binder points) can also be used for
certain types of samples as indicated in Table 2.
Table 2: Selection of sampling points (from BS 8552[6])

Type of sample Sampling Point
Full bore drain Reduced bore drain Pressure test point
point point
Settled solids Yes No No
Suspended solids & Yes Yes No

“total” metals
Dissolved solids Yes Yes Yes
Microbiology Yes Yes Yes
Dissolved oxygen No See BS 8552, A.2 See BS 8552, A.2
The system designer should also consider the installation of permanent

monitoring equipment such as corrosion coupon racks and on-line
sensors (see Section 6.6). It is preferable to design and install monitoring
equipment at the same time as pipework rather when the system is
in operation. This will also provide a full history of the corrosion
environment. Electrochemical sensors may need to be temporarily
removed during flushing and cleaning activities to avoid damage.
In addition to corrosion coupons, easily demountable sections of pipework

can be incorporated into the system for inspection purposes e.g. flanged
spool pieces in short isolable sections.
2.4.5 Maintenance & inspection

Isolation valves and drain points should be provided to enable maintenance
and inspection of the following:
s Heat exchangers (boiler heat exchangers, plate heat exchangers,
calorifier coils etc.)
s Terminal units (AHU heating/cooling coils, fan coils, chilled beams,
radiant panels etc.)
s Control valves

© BSRIA BG 50/2013
s Pump sets
s Strainers and filters
The majority of samples needed to assess water quality in the system can
be taken from suitable drain points (see Section 6.3.3).
Filters and plate heat exchangers should be fitted with differential pressure
gauges to detect signs of blockage or fouling.
Ideally, pipework systems and equipment should not be drained for

internal inspection (other than for corrosion coupon access) as this will
allow wet surfaces to be exposed to oxygen and dilution of inhibitor by
replacement water, both of which will increase the risk of corrosion.
2.4.6 External corrosion

Pipework is also at risk of external corrosion due to environmental factors.
Leaky joints, condensation and moisture leaching through insulation
materials can cause localised corrosion and eventually lead to perforation
or catastrophic failure.
All steel pipework should be protected against corrosion by suitable

coatings. Chilled water pipework insulation should be vapour sealed to
avoid the accumulation of condensation on the pipe surface. Externally
installed heating and chilled pipework must be thoroughly protected
against the ingress of rainwater.
Although external corrosion is not a water treatment issue there are some
failures where it may not be immediately obvious whether the problem
lies inside or outside the pipe. Stress corrosion cracking (SCC) particularly
affects brass fittings and valves on chilled water pipework when exposed
to ammonia or other aggressive chemical species. These can be leached
from certain types of fibre and rigid foam insulation when moistened by
condensation on the outer surface of the pipe. Where there is any doubt
as to the risk of SCC on chilled water pipework, it is recommended
to terminate and seal the suspect insulation before it reaches the brass
component. Neoprene insulation can then be used to individually insulate
the brass component but should extend to at least 300 mm either side of it.
2.5 PRE-COMMISSION The installation of pipework systems results in surfaces that may be
FLUSHING AND
CLEANING
oxidised or contaminated with dirt and debris. The objective of pre-
commission cleaning is to remove these contaminants and bacteria that
may have colonised the system after filling and pressure testing. If this
is not done then the system is at increased risk of operational problems
during commissioning and corrosion problems thereafter.
The process of pre-commission cleaning, and the measures by which it

judged to be successful, are explained at length in BG 29[1] so there is no
need to do so here. Nevertheless it is useful to reiterate the guidelines
for water quality at practical completion as these represent the minimum
requirements for system water quality at the beginning of the routine
water treatment regime.

© BSRIA BG 50/2013
It is important to stress that while Table 3 is applicable to many closed

heating and cooling systems some of the parameters (particularly the
guidelines for soluble metals) are dependent on the specific water
treatment products being employed. It is the responsibility of the water
treatment provider to declare the acceptable range of each parameter in
relation to the correct application of their products.
Table 3: Practical completion guidelines extracted from BG 29[1]

Parameter Sample location Guideline for practical completion

(unless otherwise agreed by the specifier)
Suspended solids (mg/l) Circulating pump Less than 30 mg/l at system pump during circulation
(circulate for at least one hour prior to sampling)
Settled solids (mg/l) Representative points Less than 60 mg/l in pipework at extremes of system. Less than 45 mg/l if
(as defined in BG 29) previously sampled within 3 weeks
Less than 90 mg/l at terminal units not previously sampled. Less than 45
mg/l if previously sampled (or flushed) within 3 weeks
Any individual results of greater than 90 mg/l should be investigated. The
point should be re-sampled after 2 weeks (or 2 weeks after remedial action)
in which case the result should be less than 45 mg/l
Soluble iron (mg/l) Representative points Less than 3 mg/l

Subject to advice from water treatment chemical supplier
Total iron (mg/l) Representative points Less than 15 mg/l
Subject to advice from water treatment chemical supplier
Total copper (mg/l) Representative points Less than 1 mg/l or as recommended by the specialist
TVC at 22°C Representative points <100,000 cfu/ml and no increasing trend
Pseudomonads Representative points <10,000 cfu/100 ml at 30oC and no increasing trend
Sulfate reducing bacteria Representative points Absent
Nitrite/nitrate reducing Circulating pump For information only

bacteria
Note that the guidelines for chemical parameters should be applied to individual samples whereas the guidelines for bacteria are set
relative to the geometric mean result for samples taken at the same time (the numbers and location of samples being defined in BS
8552[6]). Also, it may not be possible to achieve these guidelines for remedial flushing of an existing system.
2.6 REMEDIAL 2.6.1 Why remedial flushing and cleaning may be necessary
FLUSHING AND If water quality is being properly managed then the system may never
CLEANING
need to be re-flushed. However there are circumstances where the system
condition and water quality have deteriorated to the extent that no
amount of chemical additions will correct the evident problems.
Some scenarios that can cause the water quality to deteriorate are:
s Persistent ingress of oxygen into the systems resulting in corrosion
of components
s Stagnant/low flow water conditions, resulting in microbial growth
or poor surface passivation
s Reduced system operation and/or building dilapidation
s Poor management or failure of the water treatment regime
s Inappropriate water treatment chemicals
s External contamination
s High rates of leakage and makeup resulting in loss of inhibitors
© BSRIA BG 50/2013
Tell-tale signs that the system may need to be flushed include:

s Poor heating or cooling performance in localised areas of the
building
s Blockage of terminal unit strainers
s Large amounts of solids appearing from dirt pockets and drain valves
s Unstable commissioned flows
s Excessive wear of pump seals
s Build-up of solids in radiators and terminal units

s Blockage of plate heat exchangers
s Perforation of pipework and radiators
s Heat exchanger failures
s Rapid depletion of inhibitors
s Rapid re-colonisation with high levels of bacteria after biocide
treatment
s Persistently high levels of suspended solids
Dynamic flushing can be carried out using the mains water supply and
back flushing small circuits and terminal devices (fan coils, AHU, etc.) to
drain. In some circumstances flushing can be considerably aided by the
prior use of mild dispersants and polymers to aid the suspension and hence
removal of settled solids.
However, in cases where the system has suffered significant degradation it

will be necessary to use chemical treatments to restore the surface of the
pipe and other wetted surfaces to a condition where control of corrosion
with a conventional water treatment regime can be re-established.
2.6.2 Planning remedial flushing and cleaning

Detailed guidance on flushing and cleaning, including the treatment of
new additions to the existing system, can be found in BG 29[1]. This
includes the sampling and analysis procedures that should be used to verify
the success of the clean. Note that the guidelines in BG 29 for water
quality after the clean are appropriate to new systems and may not be
achievable for remedial work in older systems.
For the avoidance of doubt, the cleaning contractor and client should
review the guidelines in BG 29 and agree realistic criteria by which to
assess the success of the clean prior to commencing the work. In addition
to water quality guidelines modified from BG 29, those criteria may
include a relative improvement against current system water quality and the
physical condition of representative pipework surfaces before and after the
clean.
Remedial works should be planned and executed by personnel who

are experienced in pre-commission cleaning activities as the use of
inappropriate chemical treatments and techniques could be further
detrimental to the system.

© BSRIA BG 50/2013
The following issues should be considered when planning remedial

flushing and cleaning works:
Planning considerations
s Condition of the installation - how bad is the problem and what is
the minimum intervention that will correct it?
s Costs of cleaning compared to partial or complete replacement of
the installation.
s Tenant restrictions – practical restrictions on what can be done and

when
s External restrictions – are there limitations on the water supply?
s Risk to sensitive equipment – pumps, valves and sensors should be
checked for compatibility with the proposed works or temporarily
removed
s The pros and cons of off-line cleaning against on-line cleaning. It is
not usually possible to clean a badly fouled system on-line.
s Programming of the works to minimise disruption to the occupier.
It is usually possible to carry out an effective programme of sectional
cleaning at night or during weekends.
s The choice of cleaning chemical(s) to be used. The efficacy of the
cleaner in removing corrosion products from pipework and other
components and the compatibility with the alloys to be cleaned may
need to be demonstrated beforehand in a laboratory.
s Availability of drainage
s Foul drain discharge permits for flushing and cleaning activity
Considerations for the flushing contractor

s Age of the system
s System service history/water treatment log book
s Availability of up-to-date plans and schematics
s Availability of O&M manuals
s Operational history of the system including any previous remedial
works
s Records of original pre-commission cleaning works (if available)
s Records of original installation and commissioning tests (if available)
s Records of installation materials
s Nature of the problem: bacteria, suspended solids, sediment, type of
corrosion etc.
s Compatibility of cleaning chemicals with system materials and
condition. The use of aggressive chemicals should be avoided where
there is a history of leaks; alkaline cleaners should not be used in
systems with aluminium boilers or radiators etc.
s Health and safety precautions for the works
s Storage of materials and equipment
s Method statement and risk assessment

© BSRIA BG 50/2013
Considerations for the building operator

s Programming of works to minimise disruption (usually out of
normal working hours)
s Temporary provision of alternative heating/cooling where required
s Effects on normal building operation and equipment when using
mains cold water for flushing should the works be of a large scale
s Protective measures for furnishings and equipment
s Re-instatement and/or re-commissioning checks on completion

of the works to verify that the Landlord’s and other Tenant’s systems
have not been affected
s Health and safety precautions for building occupiers (segregation of
work areas)
s Security requirements during the works
s Interaction with the building FM/maintenance team: access,
operation of controls, safety etc.
s Risk management (measures to mitigate leakage etc.)
s Insurances
2.6.3 Methodology for remedial flushing

Depending on the size of the system and scope of the remedial works,
and particularly for badly fouled systems, it may be necessary to tackle the
system in sections. In this case the bulk recirculating water quality needs
to be improved and maintained by means of filtration while allowing
sections (usually individual floors) to be temporarily isolated (possibly
overnight or during a weekend), cleaned and put back into use without
risk of re-contamination. Flushing out of hours reduces the risk that there
will be insufficient water pressure available for the building occupier.
The contractor undertaking the works must provide a suitable and

appropriate site-specific method statement. This should describe the
cleaning process and the criteria for assessing its success or otherwise, to
ensure that a good standard of water quality is achieved for integration
back into the bulk water system.
Temporary pumps can be used for flushing and cleaning most of the
isolated sections but system pumps may be used for the main plant areas.
Specific provisions for the safe use of temporary pumps are discussed in
Appendix A.
The cleaning specialist must be aware of the existing system chemistry and
products in use. The replacement water treatment chemicals (inhibitors
and biocides) must be fully compatible with those in use and be approved
by the water treatment specialist as being compatible with their water
treatment regime and the system materials.

© BSRIA BG 50/2013
Preparations for remedial flushing should include:

s Surveying the system
s Checking the locations and ratings of pressure relief valves and/or
pressure reducers
s Checking the location and settings of all control and regulating
valves. These may be opened or closed as required during the
flushing process.
s Locating foul drains of an adequate size as close to the works as
possible
s Locating a good water supply with adequate pressure and as close to
the works as possible.
s Checking mains or boosted water supply pressures.
s Ensuring that the water connection and pumps can provide an
adequate flow rate without losing effective pressure
s Locating or cutting in to pipework for flushing connections and
drains
s Providing power supplies if temporary flushing pumps are to be
utilised
s Confirming temporary pump size and power requirements if these
are to be utilised
s Agreeing the site-specific methodology and water quality results
required to re-integrate the cleaned pipework with the operational
system
s Agreeing the risk assessment and precautionary measures to be
implemented
Generally, if the building is in operation and has a good level of occupation

the mains water should be of an adequate quality for flushing, though it
may be prudent to sample the mains cold water prior to commencement.
Whatever cleaning methodology is implemented, ingress of air should be

avoided. Installing deaerators, preferably of the vacuum type, to rapidly
reduce dissolved oxygen concentrations following any form of system
cleaning is recommended.
2.6.4 Cleaning on-line

As an alternative to a balanced flush, solids and colloidal matter in
suspension can be removed by use of filtration over an extended time scale
using the on-line method. Mild cleaning agents and polyelectrolytes will
also help to dislodge and suspend loose deposits for removal by filtration.
Static debris and deposits will not be removed by filtration.
Various type and grades of filter medium are used depending on particle
sizes to be removed.
In side-stream filtration, a small proportion of the system flow is passed

through the filter. The filter may have a dedicated pump set, which does
not influence the hydraulic balance of the main system, whether operating
or not. Side-stream filtration will take longer than full flow filtration to

© BSRIA BG 50/2013
reduce the solids concentration in the circulating water but has numerous
advantages:
s the filter housing is relatively small and inexpensive
s the connecting pipework is also small and easy to retrofit
s blockage of the filter will not impact on the main system
performance
s the filter can be left in situ after the flushing process for long term
maintenance operation but must be regularly checked
Filter manufacturers can provide advice and recommendations for the

removal of whatever contaminant is targeted. Generally, the smaller the
particle size the larger the filter required to sustain the required flow rate at
a reasonable pressure drop. Filtration equipment with reusable media must
have an automatic or manual back flush and drain facility and precautions
must be taken to ensure that the media does not become microbiologically
contaminated.
Filtration equipment is discussed in more detail in Section 5.1.
2.6.5 Chemical cleaning

The application of any chemical cleaner should be in accordance
with the manufacturer’s guidance regarding safe working, the effective
concentration range, the means to verify this, and the contact time. The use
of aggressive chemical cleaners on existing systems should be avoided if
possible.
Non-acidic cleaners are a blend of chelating agents, surfactants and

dispersants that ‘digest’ the rust but leave the base metal unaffected and
assist in flushing of particulate matter. They are typically applied at
concentrations of 1% to 5% of the bulk water content dependant on the
level of deposits to be removed. The contact time could be anything from
12 hours to three weeks, depending on the characteristics of the cleaner
and state of the system.
Certain mild acid products can be utilised where there are deposits of
certain compositions that cannot be removed with neutral formulations.
The system water should be tested for iron and pH levels during the
chemical cleaning process to determine the efficacy of the product and the
exhaustion rate so as to determine the need for further chemical additions.
At the end of the cleaning process, the cleaning chemicals are usually
flushed from the system which is then re-treated with inhibitors and
biocides.
2.6.6 Post-works commissioning

Decommissioned equipment and control or balancing valves must be re-
commissioned after the clean to ensure that the system is returned to the
operational condition. Normally it is sufficient to return control valves
to their original settings but since the remedial works may have affected
system resistance it is advisable to carry out spot checks to assess whether a
full re-commissioning may be needed.

© BSRIA BG 50/2013
Depending on the scope of the works it may be desirable to temporarily

increase the frequency of sampling until it is demonstrated that the water
quality within the system is stable. If that is not the case then the remedial
flush may need to be repeated or alternative measures considered.
2.6.7 Validation of the remedial works

A full record of the remedial works carried out and results obtained
(including the chemicals used, concentrations and contact times, flushing
flow rates and final dosage of inhibitors and biocides) should be kept and
issued to the building operator.
Sampling and analysis of water quality on completion of the remedial

works should be commensurate with the objectives and scope of works
and agreed prior to commencement.
Large scale works may require fully validated documentation as described

in BG 29[1] and BS 8552[6] and confirmatory sampling with independent
analysis seven days after completion of the clean. Small scale works may
require only a brief description of what was done and sufficient analysis to
demonstrate that it had the desired effect.
Whether the analysis and sampling is carried out by the cleaning or

water treatment specialist using test kits or an independent authority
using an accredited laboratory is a contractual matter, not a fundamental
requirement of the process.

© BSRIA BG 50/2013
CORROSION PROCESSES 3
3 CORROSION PROCESSES
3.1 CHEMICAL BS EN ISO 8044[11] formally defines corrosion as “physicochemical

CORROSION
PROCESSES
interaction between a metal and its environment that results in changes in
the properties of the metal, and which may lead to significant impairment
of the function of the metal, the environment, or the technical system, of
which these form a part”.
In this guide corrosion is more simply defined as the loss of metal from a
component due to the electrochemical reaction with its environment. In
closed systems this includes loss of metal from pipes and components due
to surface oxidation as well as pitting, dezincification and stress corrosion.
3.1.1 Basic theory

Corrosion of a metal in water is essentially an electrochemical process
similar to that which occurs in a battery. In order for any electrochemical
cell to operate all of the following are required:
s Anode
s Cathode
s Electrolyte
s Conducting circuit
In the corrosion cell:

s the anode is the area where the oxidation of the metal takes place,
releasing metal ions into solution and generating electrons - this is
where the primary corrosion damage occurs
s the cathode is the area where electrons are dissipated from the metal
surface, usually by the reduction of dissolved oxygen in the water to
hydroxide ions
s the electrolyte is water with dissolved salts
s the conducting circuit is provided by the underlying metal
The driving force for the corrosion process is the energy released by the
oxidation of the metal at the anode.
In general corrosion of a single material, anodic and cathodic areas are not
fixed but evolve all over the surface of the material. In pitting corrosion of
a single material, the anodic reaction becomes fixed in a particular location
surrounded by cathodic sites (Figure 2). In this case M++ is the metal ion
released by the pitting process. Metal cations may react with the hydroxyl
ion formed at the cathode, ultimately to form an insoluble material, which
may hinder further corrosion, for example iron oxides in the case of steel
corrosion.

© BSRIA BG 50/2013
Figure 2: Basic mechanism of pitting corrosion
AQUEOUS
ELECTROLYTE
O2 2OH− M++
H2O
CATHODE
2e¯ ANODE
Conductive path
METAL
The metal oxidation at the anode(s) must be balanced by the chemical

reduction taking place at the cathode(s) i.e. the rate of the anodic
reaction(s) must equal the rate of the cathodic reaction(s). Therefore
corrosion rates can be reduced by inhibiting either the anodic reaction
(for example by coating the surface) or the cathodic reaction (for example
by limiting the amount of dissolved oxygen in the water or increasing the
pH).
Anodic and cathodic activity can initiate on any heterogeneous metal

surface, with the possible exception of inert metals such as gold and
platinum.
Factors which affect the rate of corrosion include:

s Dissolved oxygen
s Temperature
s Galvanic potential difference
s pH
s Carbon dioxide
s Total dissolved solids
s Chlorides
s Sulfates
s Bacteria
s Flow velocity
s Surface condition
s Stress
In instances where the anode is small and localised the corrosion may be
locally intense resulting in pitting and eventual perforation. These factors
are explained below.

© BSRIA BG 50/2013
Dissolved Oxygen
The concentration of dissolved oxygen is by far the most important factor
controlling the rate of corrosion of metals in closed systems.
The rate of corrosion of active metals in near neutral (neither acid nor
alkaline) solutions is proportional to the concentration of dissolved oxygen,
and minimising this provides one method of controlling the corrosion
rate. However, a reduced oxygen concentration may not necessarily be an
advantage for metals protected by a passive film, e.g. stainless steel, since the
ability to maintain the film will be governed by the availability of sufficient
cathodic current.
Steel in contact with water containing low concentrations of dissolved

oxygen and no inhibitors will corrode with the formation of black iron
oxide (magnetite) sludge. At higher concentrations of dissolved oxygen
red iron oxide (haematite) may be formed with pitting of the steel surface.
In reality, the reaction pathways are rather complex and will result in a
mixture of iron oxides in different mineral forms.
Temperature
Changes in temperature can modify the mechanism and rate of corrosion
reactions.
Generally, the corrosion rate increases with increase in temperature. For

example, the rate of corrosion of mild steel in dilute chloride solutions
is approximately doubled by a 30°C rise in temperature, while that of
admiralty brass and cupronickels doubles for a 20°C rise in temperature.
In a few cases an increase in temperature can bring about a reversal of

polarity at bimetallic contacts for example in an iron/zinc couple the iron
is protected by sacrificial corrosion of the zinc at ambient temperatures but
begins to corrode preferentially as the temperature is raised above 60°C.
Cyclic temperature variations may encourage the onset of stress corrosion

cracking in susceptible alloys.
There are also secondary temperature effects. Increasing the water

temperature reduces the solubility of oxygen so temporary heating above
the normal operating temperature can sometimes be used as a method
of accelerating deaeration. Increasing the temperature can also result in
thermal disinfection of bacteria.
Galvanic potential difference

The term “galvanic corrosion” is normally reserved for corrosion induced
by the presence of two metals of different electrode potential that are
electrically connected in the presence of an electrolyte. The rate of this
corrosion is strongly influenced by the electrode potential difference
between these areas.
The galvanic series is an arrangement of the metals in the order of their

electrode potential, the most noble being at the top of the list and the most
active at the bottom. An example of this is given in table 4.

© BSRIA BG 50/2013
The further apart the metals are on this list, the greater will be the
reactivity between them and therefore the greater the speed at which the
anodic site will corrode.
Table 4: Galvanic series of some commercial metals and alloys
The actual ordering of metals in the series may vary according
to the chemical environment but the following can be taken as
indicative for a closed heating or cooling system.
More noble or cathodic

Silver Solder (more corrosion resistant)
Stainless Steel (passive)

Nickel-Copper Alloys
Bronzes
Copper
Brasses
Lead-Tin Alloys
Lead
Tin
Soft solder
Stainless Steel 316
Stainless Steel 430
Stainless Steel 410
Cast Iron
Low-carbon Steel
Aluminium Alloys
Aluminium
Zinc
Magnesium More active or anodic
(less corrosion resistant)
In Figure 3, “Metal 1” is the more cathodic metal and “Metal 2” is the

more anodic metal. Therefore “Metal 2” will be preferentially corroded.
Figure 3: Basic mechanism of galvanic corrosion
AQUEOUS
ELECTROLYTE
M++
O2 2OH -
H2O
CATHODE 2e - ANODE
Conductive path
METAL 1 METAL 2

© BSRIA BG 50/2013
Closed pipework systems invariably include several different metals and

are potentially susceptible to galvanic corrosion. However this is rarely a
serious concern where:
s The anodic area is large compared to the cathodic area
s Oxygen concentrations are low
s The correct inhibitors are deployed
The ratio of the area of anode to cathode is particularly important.

The oxidation of metal at the anode must be balanced by the chemical

reduction taking place at the cathode. If the anode is large relative to the
cathode then the corrosion will be constrained by reaction at the cathode
and any small loss of metal is spread over the relatively large area of the
anode. If the cathode is large relative to the anode, corrosion will be
constrained by reaction at the anode but loss of metal will be concentrated
in a small area and could lead to a corrosion related failure.
In many heating and cooling systems copper is the most electropositive

metal present in significant quantity and would be the cathode of most
galvanic couples. However, provided the dissolved oxygen concentration
is low, the cathodic reduction on copper is slow and consequently any
galvanic effect should be insignificant. Nevertheless, one should always
avoid the direct coupling of copper or brass to aluminium or galvanised
steel as the electrochemical potential difference is much greater and pitting
of the aluminium or zinc coating of the galvanised pipe is likely to occur.
Where there are undesirable combinations of materials that cannot be

avoided then it may be possible to electrically disconnect the anode and
cathode by using insulating materials between vulnerable parts of the
system. This should only be done with expert advice.
However, in some situations the galvanic couple is not created by the

design choice of materials but by the electrochemical process that causes a
more noble metal to plate out from solution, particularly copper onto steel.
This forms a cathodic area that can induce aggressive pitting corrosion on
the surrounding steel surface.
pH
The pH (hydrogen ion concentration) of system water significantly affects
the rate and degree to which metals corrode and the fate of the metal ions
released into solution. Most metal will corrode more rapidly in acidic
solutions (less than pH 7). Water treatment programmes for steel pipework
systems are typically designed to operate between pH 8 and pH 11.
Aluminium is corroded at high pH so water treatment programmes for

systems containing aluminium are limited to a pH of 8.5 or less, subject to
advice from the aluminium component manufacturers.
Carbon dioxide
Dissolved carbon dioxide contributes to corrosion by reducing the pH of
the water and increasing its capacity to corrode metals.

© BSRIA BG 50/2013
Total dissolved solids

A high concentration of total dissolved solids in water increases its
conductivity and therefore the risk of certain corrosion processes.
However, it is the composition of the TDS in terms of the concentrations
of individual ions that may be of more concern.
Chlorides
Chlorides are potentially aggressive as they interfere with the development
of protective films and also allow passive films to be broken down more
readily. The danger of having partial passivation is that the corrosive attack
is concentrated at active sites leading to pitting corrosion.
Chloride at the typical levels found in natural waters should not be an

issue for copper and brass in closed systems, since oxygen levels are very
low. If however chloride levels are unusually high (>100 mg/l) the dose
rate and concentration of passivating inhibitors, particularly nitrite, may
need to be increased (see Section 4.1.2).
For other materials, including those that have a naturally formed protective
oxide film such as aluminium and stainless steel alloys, certain grades may
have reduced susceptibility to chloride attack and should be chosen where
this is potentially an issue (see Section 3.1.2).
Sulfates
High concentrations of sulfates may also increase corrosion rates but are
not as aggressive as chlorides at the same concentration. However, sulfate
ions can be metabolised by various sulfate reducing bacteria (SRBs)
causing pitting corrosion in anaerobic conditions under biofilm and
deposits (see Section 3.2).
Bacteria
There are many other bacteria (aerobic and anaerobic) that can result in
corrosion, either directly or indirectly, as discussed in Section 3.2.
Flow velocity
The general corrosion rate in a pipe gradually increases with flow velocity.
At a high flow velocity, particularly in the presence of entrained gases
or solids, turbulent erosion may occur. This causes uneven attack on
pipework surfaces after flow discontinuities. Turbulent erosion mainly
affects the softer metals such as copper and brass, particularly in domestic
water systems, but is relatively uncommon in closed systems.
In closed systems it is a low flow velocity or stagnation that is more of

a problem. Under these conditions suspended solids can settle out on
pipework, increasing the likelihood of under deposit attack and MIC
leading to rapid pitting corrosion.
Surface condition
Dirty surfaces, such as those contaminated with grease or scale are more
susceptible to localised corrosion than clean surfaces. One explanation
is that isolated areas of contamination on an otherwise clean surface can
initiate differential oxygen corrosion. Dirty surfaces also encourage the
growth of biofilm and MIC.
© BSRIA BG 50/2013
Stress
Stresses in certain metals and alloys can cause stress-corrosion cracking
(SCC) when they are exposed to specific corrosive environments. The
factors required to produce this effect are a susceptible alloy/environment
combination, tensile stress (either applied or residual from the metal
forming processes) and exposure to an aggressive species. For example
ammonia causes SCC of brass and chloride causes SCC of stainless steel. A
valve that has suffered from SCC is shown in figure 4.
Figure 4: SCC of ball valve
Other factors
Other factors that increase heterogeneity and therefore encourage the
initiation of corrosion cells and should be avoided where possible include:
s Discontinuities in the natural or applied protective surface coating
(including inhibitor films)
s Local differences in oxygen concentration caused by accumulation
of deposits on surfaces and in crevices
s Differences in metallurgy due to welding
s The proximity of dissimilar metals

© BSRIA BG 50/2013
3.1.2 Corrosion effects on various metals

Aluminium and aluminium alloys
Cast aluminium alloys are used extensively in boiler heat exchangers while
wrought aluminium alloys are sometimes used for radiators. Although
aluminium is a very reactive metal, it has a high resistance to corrosion
because of the tenacious and inert oxide film which forms on the surface.
This oxide layer is generally stable in the pH range of 6.5 to 8.5 but starts
to break down or dissolve in more acid or alkaline conditions. In any water
treatment programme, it is important that account must be made of any

aluminium in the system, so that the pH does not exceed 8.5. This is lower
than that typically used for steel systems without aluminium.
The behaviour of aluminium in water systems depends on carbonate

hardness and concentrations of chlorides and copper. Higher chloride
(and sulfate) content encourages general corrosion and pitting whilst
higher carbonate hardness reduces the number of pits formed. If oxygen is
present, dissolved copper in the water may have a profound effect on pitting
corrosion; the amount of copper necessary to initiate pitting varies with
hardness and chloride content, but can be as low as 0.02 ppm.
Contact with copper and copper alloys, and to a lesser extent, iron and iron
alloys, will induce galvanic corrosion in all environments where dissolved
oxygen is present. Aluminium is intentionally used a sacrificial anode in
some situations to protect more vulnerable materials.
Cast iron
Typical applications of cast iron include boiler sections (heat exchangers),
pump housings, large valves and pipe fittings (malleable iron). Corrosion
rates for cast iron in closed heating and cooling systems are similar to those
for mild steel and the same programmes of water treatment apply.
Copper and copper alloys

Copper and its alloys comprise a versatile range of materials which are
used in a wide variety of building services applications. Copper is used
extensively for pipework, hot water cylinders and heat exchanger coils due
to its excellent ductility and thermal conductivity. A wide range of copper
alloys including brasses and bronzes are available and mainly used for valves,
pipe fittings and pump components.
Copper metal – Copper tube is widely used in closed heating and

cooling systems and acute failures due to internal corrosion are rare.
Where general surface corrosion occurs due to oxygen ingress, the
release of copper ions into the water may increase the rate of corrosion
and pitting on other more active metals in the system such as mild
steel. The risks are greatest in the early life of the system when oxygen
concentrations are high and metal surfaces not yet fully protected by the
water treatment programme.
Brasses - The common brasses consist of a range of copper alloys

containing from 10% to 50% zinc with other minor alloying elements.
Brasses can suffer from two particular forms of corrosion leading to

© BSRIA BG 50/2013
catastrophic failure of components. These are dezincification and stress

corrosion cracking.
Dezincification is where zinc is preferentially removed from the alloy

leaving a porous mass of copper having little strength. Where leakage or
breakage of a brass fitting has occurred, dezincification may be suspected
if the defective area has a dull, coppery appearance. Dezincification is
well known in domestic water systems but rarely encountered in closed
systems as it requires a high oxygen environment. The risk can be

reduced by specifying components manufactured from dezincification
resistant (DZR) brasses.
Stress corrosion cracking (SCC) occurs where brass components are

exposed to both mechanical stress and certain corrosive chemical
species, particularly ammonia and related compounds. The stress can be
residual from the manufacturing process or imposed during installation
for example brass nuts of compression fittings are inevitably subject to
high hoop stresses as a result of tightening. SCC is a particular problem
in chilled water systems where corrosive species can be leached from
insulation materials as condensate forms on the outer surface of the pipe
and fitting. SCC is rarely initiated from inside the fitting as the aggressive
chemical species are unlikely to be present in sufficient concentration in
the system water to cause SCC. However, there are examples, including
the failure of valves, where the crack appears to originate from a surface
in contact with system water. The risk of SCC (and dezincification)
can be eliminated by specifying bronze components but these are more
expensive than brass.
In common with copper, brasses are relatively soft and can suffer attack
by erosion, impingement or cavitation. Cavitation in heating and
cooling systems is most frequently encountered on valve seats where
the valve has been operated in an almost closed position. This risk is
exacerbated in water with high suspended solids or debris. An example is
shown in figure 5.
Figure 5: Control valve damaged by cavitation

© BSRIA BG 50/2013
Bronzes (gun metals, phosphor bronzes, and tin bronzes) - The

term ‘bronze’ should refer to copper-tin alloys, but in practice a variety
of copper alloys are termed bronzes irrespective of whether they contain
tin. This wide variation in composition makes corrosion behaviour of a
component difficult to predict unless the specific composition is known.
Gun metals are probably the best all round choice for general valves and
fittings in water-based heating systems.
Aluminium bronzes - These materials generally have a high resistance

to corrosion, impingement attack and cavitation erosion. They are
however surprisingly prone to pitting corrosion in natural waters and can
suffer selective corrosion analogous to dezincification. They are used for
tube plates, back boilers and pump impellers.
Cupro-nickel alloys - Alloys containing copper and nickel have

excellent resistance to corrosion in many environments, including those
containing chlorides. They are more stable than brasses under flow
conditions, less susceptible to stress corrosion and are used in shell and
tube heat exchanger applications.
Stainless steel
The term “stainless steel” covers a wide of alloys incorporating
iron, chromium and nickel. These can be grouped by metallurgical
characteristics as martensitic, ferritic and austenitic. Austenitic grades have
the highest corrosion resistance and ductility and are most frequently used
in closed systems. Examples of the use of stainless steel in closed systems
include valve stems, pump shafts and impellers, strainers, parts of fittings,
expansion joints, plate heat exchangers and buffer vessels. Stainless steel
is occasionally used for pipe sections and associated fittings but less so in
closed systems than domestic water systems.
Stainless steels are potentially subject to pitting, corrosion and stress

corrosion, particularly in the presence of high concentrations of chlorides.
The stainless steel grades used in components used for closed systems
are unlikely to be adversely affected at the concentrations normally
encountered in closed systems filled from mains water. Stainless steel
pipework can however be affected by external corrosion if protective
coatings fail, particularly on buried pipe.
Steel (low carbon and low alloy)

Low carbon and low alloy steel, generically known as mild steel, is used in
a wide variety of applications due to its ease of fabrication and mechanical
properties. Applications include pipework, boiler heat exchangers
and radiators. Mild steel will readily corrode when exposed to water
containing dissolved oxygen unless protected in some way.
At dissolved oxygen concentrations above 1 mg/l in untreated water,

the corrosion products may form bulky mounds on the surface called
tubercles. These overlie areas where localised attack is occurring and can
lead to pipe wall perforation or reduce the carrying capacity of pipes
due to the reduction in cross sectional area. At lower concentrations of
dissolved oxygen, iron oxide corrosion products can form a magnetite

© BSRIA BG 50/2013
sludge that restricts flow and may lead to blockages or increased risk of
MIC. This is commonly found in the base of steel radiators and the cast
iron heat exchangers of older boilers.
Steel corrosion can occur even in the absence of oxygen, if the water
contains sulfate reducing bacteria (see Section 3.2.1).
Zinc
Zinc is used as a protective coating on some steel pipes and fittings,
mainly for cold water services. Internally coated pipes and fittings are
NOT suitable for hot water or heating systems as the corrosion rate in
oxygenated water increases dramatically above 55°C and a reversal of
potential occurs above 60°C. The zinc then becomes cathodic to steel,
stimulating attack which can result in localised corrosion (pitting) and
subsequent perforation.
3.2 MICROBIALLY Bacteria are single celled microorganisms that often aggregate into
INFLUENCED colonies. The typical cell size is 1 μm to 10 μm. The population of
CORROSION (MIC)
microorganisms can be diverse and variable depending on the degree
of sensitivity to the oxidation-reduction potential of their environment.
Bacteria can be classed into the following categories depending on how
much oxygen they require for respiration:
s Aerobes (grow in the presence of oxygen), e.g. Pseudomonas,
Aeromonas, Flavobacter.
s Facultative anaerobes (able to grow with or without oxygen) e.g.
Klebsiella, E.coli.
s Anaerobes (grow in the absence of oxygen) e.g. Sulfate reducing
bacteria (SRB).
Each type of bacteria therefore requires different conditions for survival

and growth. If conditions are right for growth they can multiply rapidly,
with numbers doubling in as little as 20 minutes. One bacterial cell can
potentially become 200 million in less than 10 hours!
All natural sources of water (including tap water) contain many different
types of bacteria, some of which may multiply and lead to problems
within a pipework system if they encounter suitable conditions for growth.
Bacteria can circulate freely in the system water as “planktonic” bacteria or
adhere to the pipe walls as “sessile” bacteria.
Closed systems, if left filled and untreated, can quickly develop a biofilm
layer on pipe surfaces. A biofilm is a microbial mass of aquatic bacteria
and other microorganisms together with trapped particles and dead
bacteria cells, bound together by an exopolymer. This is an extracellular
polysaccharide exudate produced by slime forming bacteria such as
pseudomonads, constituting up to 90% of the biofilm. It protects the
sessile bacteria from hostile conditions including exposure to biocides. A
consortium of bacteria is therefore able to develop within the biofilm on
the pipe wall where both aerobic and anaerobic species can flourish. This is
the starting point for microbially influenced corrosion.

© BSRIA BG 50/2013
During the late 1990s, some bacteria (particularly Pseudomonas) were

linked to a series of damaging contamination problems within closed
heating and cooling systems. The symptoms included blockages at
valves and strainers, sludge formation, and severe gassing affecting flow
measurements and commissioning results. There is no evidence to confirm
that Pseudomonas was directly responsible for these problems – other
bacteria may have been to blame. However, it has been found that where
concentrations of Pseudomonas-like bacteria (pseudomonads) in system
water are high, the risk of these problems is increased. Pseudomonad

concentrations are therefore increasingly used as one indicator of the
biological quality of system water.
Other bacteria of concern are nitrate/nitrite reducing bacteria (NRB) and

nitrite oxidising bacteria. These bacteria can cause rapid loss of nitrite-
based inhibitor from the system and so increase the risk of electrolytic
corrosion. In addition, some of these bacteria can produce ammonia that
can increase corrosion of copper and brass.
There is a natural order in which bacteria develop within closed heating

and cooling systems. First the aerobic bacteria colonise surfaces and
multiply in oxygenated sections of the system. In stagnant areas of the
system these species quickly reduce the available oxygen allowing growth
of anaerobic bacteria such as SRB and NRB.
There are three ways in which bacteria are able to cause or influence
corrosion:
a. directly, by obscuring the surface, resulting in a differential aeration
cell
b. directly, by producing metabolic by-products which are in themselves
aggressive towards metal surfaces
c. indirectly, by acting on the treatment chemicals in a way that renders
them less effective
These are described in more detail in the following sections.
3.2.1 Direct MIC

Direct MIC occurs when sessile bacteria are in direct contact with
the metal surface, usually under the biofilm. Different bacteria can be
implicated in various forms of corrosion in different systems but specific
examples for closed heating and cooling systems include:
Oxygen depletion corrosion: If there is a substantial biofilm, a zone

of oxygen depletion can occur at the interface between the biofilm and
the metallic surface. The different oxygen potentials between the inside
and outside of the biofilm cause the inside to become the anode and the
outside to become the cathode. This then allows electrolytic corrosion to
take place. This is illustrated in figure 6.

© BSRIA BG 50/2013
Figure 6: Differential oxygen cell formed under biofilm
O2
O2
Aerated Water
O2
O2
O2 O2 O2
BIOFILM
O2 O2
O2 O2 O2 O2
O2 O2 Aerobic O2
O2 O2 O2
BIO
O2
FILM
O2 O2
FILM
O2 O2 O2 O2 O2
BIO
O2 Anaerobic O2 O2
O2 O2
O2 O2
O2
OH– OH– O2 OH– OH–
M+ M+
M+
Cathode e– e– Cathode
–
e e–
Anode METAL
Corrosion of stainless steel: Stainless steel is corrosion resistant due

to the presence of a film of a stable oxide on the surface. However in the
low oxygen conditions under a biofilm, this oxide layer breaks down thus
removing that protection and allowing corrosion to occur.
Corrosion due to SRB: The absence of oxygen under the biofilm allows
the growth of anaerobic organisms such as SRB which are one of the
most frequent causes of MIC. SRB reduce sulfate from the fill water to
hydrogen sulfide which reacts with the metal to produce metal sulfides
as corrosion products. The involvement of SRB in corrosion failures
is inferred from the evolution of hydrogen sulfide when the corrosion
deposit is acidified. Figure 7 shows two examples of SRB attack on pipe
surfaces under the microscope (at low magnification).
Figure 7: Examples of SRB pitting corrosion
It is possible to reduce the risk of SRB activity by limiting essential

bacterial nutrients such as phosphorus, nitrogen, and sulfate. Reverse
osmosis or deionised fill water will have fewer problems with SRB. Also,
any practices that minimise biofilm such as flushing, disinfection and the
removal of dead legs, will reduce the growth of SRB.

© BSRIA BG 50/2013
3.2.2 Indirect microbial influence

Examples of the indirect influence of bacteria include:
Nitrate/nitrite reducing bacteria: Nitrite can be metabolised by

certain bacteria (for example Pseudomonas denitrificans and Bacillus)
and rapidly depleted from a closed system. As the concentration of the
nitrite falls it becomes less effective and may actually increase the rate of
corrosion. The degradation products may also be detrimental for example
nitrate/nitrite reducing bacteria (NRB) may reduce nitrates, nitrites and

organic nitrogen compounds to ammonia, which is aggressive to copper
and copper alloys.
Interference with the action of treatment chemicals: In general

terms chemical corrosion inhibitors are able to reduce corrosion by
coating the exposed metal surfaces. Biofilm prevents the inhibitor from
reaching the metal surface and therefore allows corrosion to take place as
discussed in section 3.2.1.
Degradation of glycols: Bacteria can survive in water containing up to

20% glycol and are able to use the glycol as a nutrient (a source of organic
carbon). It is important to maintain the glycol concentration above
20% to prevent proliferation of those bacteria unless there are additional
biocides present. The glycol concentration should be checked and
corrected after any water replacement that could have resulted in dilution.
The physical characteristics of glycols and measurement of concentration
are discussed in Appendix E.

© BSRIA BG 50/2013
CHEMICAL WATER TREATMENT 4
4 CHEMICAL WATER TREATMENT
4.1 CORROSION The purpose of water treatment in a closed heating or cooling system is to
INHIBITION
ensure that the system maintains its efficiency and achieves or even extends
its life expectancy. In practice the focus tends to be on protecting the
metal components of the system from corrosion processes that may lead to
the formation of suspended solids and pitting, though the adverse effects of
poor water quality on other materials should not be ignored.
4.1.1 Corrosion inhibitor treatment programme

The inhibitor programme/products should tolerate and be effective under
the following conditions:
s Make-up water characteristics within the ranges specified by the
programme/product provider
s Dissolved oxygen concentrations within the range specified by the
programme/product provider
s Total dissolved solids and suspended solids within the range specified
by the programme/product provider
s pH appropriate to the materials of construction
s System operating temperature specified by the designer
The inhibitor programme/products should:

s exhibit good wetting properties. They should not support or
promote foaming conditions in areas of high turbulence.
s provide a protective layer on clean metal surfaces that is effective
both in flowing water during system operation and stagnation
during extended plant shut-downs.
s ideally provide both cathodic as well as anodic inhibition, thereby
giving some protection to metal surfaces under debris.
s provide protection to all vulnerable metals in the system.
Inhibitors work by a number of different mechanisms, but those used in

closed systems all result in microscopically thin molecular films that are
invisible to the naked eye.
The inhibitor programme should be effective under conditions of

slight dilution or over-concentration and protect all the metals present
in the system. The intended application (systems and materials) and
effective concentration of each inhibitor should be documented by the
programme/product provider.
The addition of each inhibitor required to reach a target concentration

should be calculated using an accurate estimate of the system volume. The
actual concentration of inhibitor should always be checked after addition
and circulation of the inhibitor.
Note that some inhibitors, such as nitrite, can promote pitting corrosion
when present below a critical concentration. It is very important to

© BSRIA BG 50/2013
maintain inhibitor concentrations above the manufacturer’s recommended

“safe” concentration when using these.
The water treatment programme should specify sufficient reserves of

inhibitors so that precise remedial additions are not required for minor
losses of system water and inhibitors should only require occasional
topping up. Such reserves should allow for any degradation or adsorption
of inhibitor on system surfaces and debris.
The water treatment chemicals should not be aggressive to any materials

found in the system including solders, jointing compounds, rubber and
plastic components. Where necessary, the treatment programme should be
compatible with antifreeze components such as glycols.
The water treatment programme should have sufficient buffering capacity

to maintain the system pH within the specified range in the presence of
cleaning residues, antifreeze, degradation products of the water treatment
chemicals and bacteria metabolites.
Chemical spot tests and performance tests should be available so that the
efficiency of the water treatment can be measured in the field.
The water treatment programme should not promote the establishment

or growth of fungi or bacteria. Refer to Section 4.2 for the application of
biocides.
Where appropriate, the water treatment programme should inhibit scaling

or precipitation of hardness salts or other impurities. Refer to Section 4.3
for the application of scale inhibition.
4.1.2 Properties of corrosion inhibitors

Inhibition may result from:
s the inhibitor molecule being chemisorbed on the metal surface so
forming a protective film either by itself or in combination with
metallic ions.
s the formation of passive oxide films.
s the inhibitor reacting with corrosion promoting species present in
the system water.
Corrosion inhibitors can be generally classified into anodic, cathodic or

organic film-forming.
Anodic inhibitors build a very thin and generally invisible protective

film initially at the anode which may eventually spread all over the metal
surface. Whilst the degree of protection offered by many anodic inhibitors
is high, care must be taken to ensure that they do not fall below their
critical concentration as this can cause pitting attack on the metal surface.
Cathodic inhibitors are generally less effective than the anodic type
and may form visible films on the cathode surface. Whilst there is no
critical lower concentration, excessively high concentrations could cause
precipitation, increasing the risk of under-deposit corrosion.
© BSRIA BG 50/2013
Cath-anodic corrosion inhibitors have properties of both anodic and

cathodic inhibitors, though there are few examples in common use. More
usually the correct balance of protection is achieved by blending different
anodic and cathodic inhibitors to achieve the required performance.
Organic film forming inhibitors usually have a mixed mode of action

- they inhibit both anodic and cathodic reactions. However, they are
generally used to prevent corrosion under acidic conditions rather than in
neutral oxygenated waters.
In general terms, if the mean corrosion rate is reduced to an acceptable

rate for all the metals present and there is no sign of pitting attack then
effective corrosion protection has been achieved.
The common corrosion and scale inhibitors used in closed systems

are listed in Table 5 and the subsequent text. The dose rates and target
concentrations should be specified by the programme/product provider
and documented by the water treatment specialist in the system water
treatment log book.
Table 5: Common corrosion and scale inhibitors

Component Inhibitor function
Nitrite Corrosion inhibitor for ferrous metals
Nitrate Corrosion inhibitor for aluminium
Molybdate Corrosion inhibitor
Azoles Corrosion inhibitors for copper and copper alloys
Phosphate Corrosion inhibitor for steel
Polyphosphate Scale and corrosion inhibitor (reverts to

orthophosphate)
Phosphonates Scale and corrosion inhibitor
Silicates Corrosion inhibitor for steel, copper alloys and

aluminium
Tannins Film forming corrosion inhibitor and oxygen
scavenger
Benzoate Anodic inhibitor
Triethanolamine, monoethanolamine, Organic film formers and corrosion inhibitors

alkylcarboxylates and substituted triazines
Phosphono- and phosphino- carboxylic acids Cathodic inhibitor and scale dispersant
Diethylhydroxylamine Oxygen scavenger
Borate pH buffer, biocide, corrosion inhibitor
Nitrite – The nitrite anion (NO2-) is an oxidizing anodic inhibitor for

mild steel that has been in use for many years. Unlike molybdate, nitrite
does not require dissolved oxygen to function as a passivating inhibitor.
The optimum dose depends on the concentration of aggressive ions such
as chloride and sulfate and system temperature. As with other anodic
inhibitors, if the concentration falls below a critical level then severe

© BSRIA BG 50/2013
pitting can ensue. Too little nitrite can be worse than none at all. Table
6 suggests the critical concentrations of chloride and sulfate for pitting of
steel in nitrite inhibitor based on information from Corrosion and Corrosion
Control[18].
Table 6: Critical concentrations for pitting of steel in nitrite inhibitor
Nitrite inhibitor Chloride Sulfate
NO2- mg/l Cl- mg/l SO42- mg/l
33 127 14
67 278 37
333 1207 304
Give that nitrite concentrations will normally be held above 300 mg/l and
mains water chloride and sulfate concentrations should be well below 100
mg/l this issue is unlikely to be a major concern in most systems filled
from the mains. However chloride and sulfate can be much higher in
borehole water. Table 8 of BG 29[1] suggests that water containing up to
250 mg/l sulfate or chloride should be acceptable for filling the system.
Also bear mind that chloride and sulfate can accumulate in the system in
the longer term as the breakdown product of some biocides.
Nitrite may be used alone or synergised with alkaline buffering agents

and other inhibitors such as molybdate. Azoles are added when protection
of copper or copper alloys is required. If aluminium is present, a specific
aluminium inhibitor such as silicate or nitrate will also be needed.
Nitrite is compatible with glycol antifreeze and commonly used non-

oxidizing biocides. Nitrite should not be mixed with organic amines,
amides or other organic inhibitors that dissociate under heat transfer or
catalysed conditions to form nitrosamines or other suspected carcinogens.
Nitrite is readily degraded by various microorganisms, generally described

and nitrite reducing (denitrifying) or nitrite oxidising (nitrifying) bacteria.
Nitrifying bacteria oxidize nitrite to nitrate, while denitrifying bacteria
reduce nitrite to nitrogen gas or ammonia. In addition to the loss of
protection resulting from the reduction in nitrite concentration, the
degradation products may also be detrimental. For example nitrate is also
a nutrient for other bacteria (though it also has an inhibitor function),
whereas ammonia is aggressive to copper and copper alloys. Diligent
use of a non-oxidizing biocide regime is imperative when using a nitrite
program. Oxidizing biocides should not be used with nitrite treatments,
since they will oxidize nitrite ions to nitrate ions.
Nitrate – The nitrate anion (NO3-) may be included with other inhibitors
for the protection and aluminium components and solder. In particular
the addition of nitrate can overcome the detrimental effect of nitrite on
aluminium.
Nitrate functions by controlling the growth of the aluminium hydroxide

film and also as a pitting inhibitor by the preferential reaction of nitrate
with the active aluminium surface.

© BSRIA BG 50/2013
If copper or copper alloys are present, an azole is recommended, since

neither nitrite nor nitrate provides protection against corrosion of these
materials.
Whilst nitrate is not as problematical as nitrite in terms of critical

concentration, care must be taken with the biocide regime.
Molybdate – The molybdate anion (MoO42 - ) is widely used as an anodic

inhibitor. Molybdate (added to the system as sodium molybdate) requires

some dissolved oxygen to function effectively though the concentrations
typically found in closed heating and cooling systems are usually sufficient
2-
for this. If there is insufficient oxygen for molybdate inhibition the system
is unlikely to be at immediate risk of oxygen corrosion.
Molybdate is most often synergised with other inhibitors such as nitrite.

In this combination, nitrite takes the place of oxygen in forming the
protective oxide film. The molybdate also helps to retard the growth of
pits should the concentration of nitrite fall. The benefits of molybdate/
nitrite programmes include enhanced corrosion control, a reduced
propensity for pitting relative to nitrite alone, and both low carbon
steel and aluminium corrosion control. Molybdate/nitrite programmes
are compatible with commonly used non-oxidizing biocides as well as
ethylene and propylene glycols.
The absorbance of molybdate onto the outer hydrated iron oxide

layer means that when molybdate is fed to a poorly maintained system
containing old corrosion products, molybdate will be consumed.
Therefore, prior to using a molybdate based treatment in a fouled system,
the system should be cleaned to remove existing corrosion products.
Azoles - Aromatic azoles such as tolyltriazole (TT), benzotriazole

(BT), mercaptobenzotriazole (MBT) and other substituted azoles are
used in closed systems for controlling copper or copper alloy corrosion.
Substituted azoles have also been utilised for controlling aluminium
corrosion. The azole requirement increases with increasing concentration
of chloride and sulfate ions.
In addition to controlling copper or copper alloy corrosion, azoles react

with cuprous ions in solution to prevent them from plating out as metallic
copper on low carbon steel or aluminium surfaces which can lead to
aggressive local pitting.
Orthophosphates – Orthophosphates (PO43-) are used as either the

mono, di or tri metal salts depending upon the pH of the system. They
are anodic inhibitors for low carbon steels but are rarely used alone as
they form insoluble precipitates with water hardness salts, and do not
function in the absence of oxygen. Orthophosphates should be used
with care especially with aluminium heat exchangers as they can induce
pitting attack under high heat flux conditions. Several European countries
prohibit the discharge of phosphates or orthophosphates into the sewerage
system.

© BSRIA BG 50/2013
Polyphosphates - Polyphosphates are complex linear condensed or

polymerised phosphates with variable chain length. They are used as the
metal salts of hexametaphosphate, tripolyphosphate and pyrophosphate, all
of which act as cathodic inhibitors for steel when combined with calcium.
Threshold concentrations of these chemicals will not only provide
corrosion protection to the system but in the range pH 7.0 to 8.0 will also
precipitate scale formers. Polyphosphates hydrolyse to orthophosphate
under strongly acid or alkaline conditions or under the influence of high
heat flux.
Phosphonates - Phosphonates are organic compounds that are

structurally related to orthophosphates. They are chelating agents and
bind strongly to metal surfaces. Several are now being used as cathodic
corrosion inhibitors for low carbon steels and as deposit (calcium
carbonate) control agents.
Phosphonates work in a similar manner to polyphosphates but are more

resistant to hydrolysis and thermal breakdown under high heat transfer
conditions. They usefully synergise with other inhibitors and can stabilise
formulation concentrates. Overdosing of phosphonate formulations,
especially in soft water areas, can induce cupro-solvency. Failure to control
this can cause problems when attempting to protect mixed metal systems
especially where aluminium is present.
Phosphonates degrade to orthophosphate and organic acids in the

presence of oxidising species. As these are potential nutrients for bacteria
the implication of changing an existing programme to or from one with
phosphonates and interaction with other chemicals that may be present
should be carefully considered.
Silicates - Silicates are considered to be anodic filming inhibitors. They

have a variable composition of nNa2O: mSiO2. Generally, a ratio of m/n
between 2.5 and 3.0 is effective. With silicate treatments the protective
film develops slowly and may take weeks to form. The film is believed
to consist of silica gel along with ferric hydroxide precipitates. Silicates
provide protection for low carbon steel, copper, copper alloys and
aluminium.
Silicates are not used as stand-alone corrosion inhibitors for low carbon
steel protection but may be added to the water treatment programme for
control of aluminium corrosion and/or to augment the performance of
other inhibitors.
An advantage of silicates is that they do not act as a nutrient for

microbes and are an option where microbiological control is difficult. A
disadvantage is the tendency for silica deposition to occur if the product is
over-dosed, used with hard water at high temperature or where pH is not
properly controlled. For example, if silica is used for aluminium corrosion
control and the pH is elevated then aluminium silicate deposition may
occur, which can plug nozzles and other small orifices.

© BSRIA BG 50/2013
The efficiency of silicate anodic protection is less than that for

polyphosphates. Concentrations of the order of 100 ppm are required
for silicate to be effective. The silicates are stable only at relatively high
pH and readily form insoluble salts with hardness formers (calcium and
magnesium salts). These precipitates can be highly abrasive and can also
form thermally insulating layers particularly in boiler heat exchangers.
Silicates are, however, useful in neutralising the effects of acidic soft waters
to minimise the dezincification of brass and cupro-solvency.
Tannins – Tannins (and lignins) are mixtures of materials derived

from wood bark and have been widely used for many years as oxygen
scavengers, crystal modifiers, sludge dispersants and cathodic film formers.
They are not generally used in isolation but synergised with other
additives.
Careful control of pH may be necessary to avoid precipitation of tannins

leading to rapid depletion of the active reserve. Unlike many inorganic
anions, tannins do not promote localised attack when present at below
critical concentrations. Excessive concentrations of tannins have however
been observed to form thick films on heat exchanger surfaces under
conditions of high heat flux and these films can become detached as large
platelets (suspended solids) which can settle in areas of low flow and cause
pump and valve problems.
Benzoate – The benzoate anion may be loosely described as an anodic

inhibitor but is rarely used in isolation. In heating and cooling systems
benzoate (as the sodium salt of benzoic acid) is usually combined with
nitrite. Benzoate does not promote localised corrosion when used below
critical concentration. Benzoate is readily biodegradable and requires
regular top-up and close control of biocides to maintain the required
concentration.
Organic film formers – These are organic molecules that react with
metal surfaces in contact with the water in which they are dissolved to
form a protective barrier.
Triethanolamine and monoethanolamine form barrier films that prevent

oxygen from reaching the metal surface, thus inhibiting the reaction at the
cathode. They act as anodic inhibitors for aluminium by blocking pores in
the surface oxide film.
Alkylcarboxylates are absorbed onto metal/metal oxide surfaces, possibly

through hydrogen bonding. The hydrophobic hydrocarbon tail provides a
barrier that prevents water and hence dissolved oxygen from reaching the
metal surface.
Substituted triazines form a film over anodic sites. There is a wide

range of possible compounds with different organic substitution groups.
Substituted triazines do not provide protection for aluminium, copper or
copper alloys so other inhibitors are added for systems containing these
e.g. silicates for aluminium.

© BSRIA BG 50/2013
Combinations of organic film formers and vapour corrosion inhibitors

(VCIs) are now being used in closed heating and cooling systems
particularly for low conductivity systems. There is however concern
that the addition of organic carbon to the system could promote
microbiological growth.
Phosphono- and phosphino- carboxylic acids – Both of these

copolymers act as cathodic inhibitors and scale inhibitor/dispersants.
Oxygen scavengers – Sodium sulfite and other oxygen scavengers

including diethylhydroxylamine (DEHA) have historically been used in
high temperature heating systems with steel pipework on the basis that
removal of dissolved oxygen will prevent oxygen corrosion.
Sulfites are still used for steam boiler plant and high temperature heating
systems but rarely in low temperature systems as they could promote the
growth of SRB by providing a source of oxidised sulfur and anaerobic
conditions. This is not an issue in high temperature systems as the bacteria
are eliminated by thermal disinfection.
DEHA is used for controlling ferrous metal corrosion in both high and
low temperature systems, particularly those that require the water to
maintain a low conductivity. In addition to being an oxygen scavenger and
pH buffer, DEHA acts as a passivator, converting haematite to magnetite.
Systems relying on oxygen scavengers require frequent (even daily)

monitoring, especially where there is oxygen ingress either directly or via
makeup water.
Borate - Sodium borate is a pH buffer, biocide and corrosion inhibitor. It

is compatible with both molybdate and nitrite in high pH water treatment
programmes. Its use is likely to be curtailed as a consequence of the
Biocidal Products Regulation (see Appendix C).
4.1.3 Practical considerations in choosing a water treatment

programme
There is no one treatment programme that eliminates corrosion and
deposition in every application, and there is usually more than one
possible option available for each application. A practical water treatment
programme is likely to combine several chemicals with different functions,
often as proprietary formulations, to provide overall protection for the
system.
It is up to the provider to:

1. Obtain relevant information on the closed system to be treated
2. Evaluate which treatment options will achieve the corrosion control
and water quality objectives, taking account of customer concerns and
references, regulatory issues and product selection constraints
3. Follow up the treatment programme with proper monitoring of
chemical, microbiological and physical parameters, ideally including
monitoring of corrosion rates.

© BSRIA BG 50/2013
Where multi-component formulations are used, the scope of analysis

should be sufficient to detect whether the treatment programme is
becoming unbalanced and to allow remedial dosing to restore the balance.
If depletion of threshold components goes unnoticed the system could be
vulnerable to rapid localised attack.
For example, when using nitrite/molybdate inhibitor package it is usual to

test the nitrite first, as this is the component more likely to disappear. If
the nitrite concentration is low then the molybdate should also be tested
so that the remedial dosing can be adjusted to bring both components up
to the recommended concentration. In some situations it will only be
necessary to add nitrite.
Water treatment chemicals that will be exposed to metal surfaces with

high rates of heat transfer should have been tested for possible effects of
thermal degradation and catalytic oxidation.
Some water treatment chemicals encourage microbiological growth, either

directly or indirectly through their degradation products. In such cases a
complementary biocide should normally be used.
All biocides (see Section 4.2.2) must be compatible with the components
of the corrosion inhibition programme so that neither reduces the
effectiveness of the other.
Finally it is essential that the treatment formulation concentrate has a

long shelf life. Liquid formulation concentrates should not re-crystallise
or undergo phase separation on freeze-thaw testing. The activity of the
ingredients should not be depleted with time and the concentrate should
not be susceptible to biodegradation in storage.
The water treatment specialist should be aware of the possibility of

feedwater pre-treatment to control dissolved gas concentrations, water
hardness, dissolved solids and pH.

© BSRIA BG 50/2013
4.2 BACTERIA AND In closed systems the primary concern is biofouling, though fungal growth
BIOFOULING
INHIBITION
can sometimes be significant. The bacteria that cause most problems in
closed heating and cooling systems are:
s Sulfate reducing bacteria (SRB)
s Nitrite reducing bacteria
s Pseudomonads
It is likely that over 90% of the microorganisms present in a badly fouled

closed system exist in biofilms attached to surfaces within that system.
Bacteria in the biofilm may be up to 3000 times more resistant to biocides

than bacteria present within system water. This illustrates how difficult it is
to eliminate microorganisms when they are present in established biofilms.
Biofilms cause a number of operating problems including:

s microbially influenced corrosion (MIC).
s reduced efficacy of corrosion inhibitors due to their inability to
access pipework surfaces.
s reduction of heat transfer in heat exchangers.
s in extreme cases, reduced water flow due to partial or complete
blockage of valves, strainers etc.
Biofouling and corrosion of metal pipe surfaces occurs in several stages.

Firstly one or more slime forming bacteria attach to the pipe surface and
begin to multiply and exude exopolymer allowing other bacteria to attach
to the surface and multiply. The accumulation of bacteria and exopolymer
quickly forms a complex three dimensional structure. This interferes
with the action of inhibitors and allows the formation of an anoxic zone
adjacent to the metal surface. This results in a differential aeration cell
where areas under the biofilm are anodic relative to areas beyond the
biofilm with greater oxygen availability, and the potential for localised
corrosion. Partial lifting of a biofilm a can also result in differential
aeration.
The anoxic conditions under the biofilm also provide ideal conditions
for the growth of anaerobic bacteria, such as the SRB that accelerate
corrosion and pit formation. SRB accelerate corrosion by reducing sulfate
to hydrogen sulfide that reacts to form insoluble sulfides and cathodic
depolarisation (removal of hydrogen from the cathode). Other bacteria
can also create acidic environments (with sulfuric acid, hydrochloric acid
etc.) under the biofilm, resulting in corrosion.
Biofilms can propagate through detachment of clumps of cells, or by

a type of “seeding dispersal” that releases individual cells. Either type
of detachment allows bacteria to attach to a surface or to a biofilm
downstream of the original community.
The ability of biofilm to trap particulate matter increases operational

problems and cleaning challenges. Overall the presence of a biofilm
increases the risk of damaging corrosion and equipment failures.

© BSRIA BG 50/2013
4.2.1 Factors promoting growth of bacteria and MIC

pH: Most microorganisms grow best at pH 6.5 to 7.5. Few thrive below
pH 4 or above pH 9 though there are reported cases of microbes surviving
at less than pH 1 or above pH 11. In general, bacteria tend to predominate
above pH 6 or above whereas fungi and yeasts can flourish in acidic
conditions, below pH 6.
Temperature: Most spoilage microorganisms tend to grow best at

temperatures of 15 - 40°C. Although the usual design temperatures for

heating and chilled water systems may not appear to favour bacteria
growth, some areas within the system may provide ideal temperatures for
bacteria growth, particularly under part load operating conditions or when
the system is inactive.
Nutrients: It is unfortunate that some of the chemicals that are added

to the system as corrosion inhibitors, or their degradation products, also
provide essential nutrients for certain bacteria. Even biocides may become
nutrients as they degrade. While this situation cannot be avoided, the
water treatment specialist should be aware of the issue and where necessary
take additional steps to counteract the problem.
Biofilms: Biofilms protect bacteria from the action of biocides and

provide a range of environmental conditions suitable for the growth of
different bacteria as well as being directly implicated in corrosion processes.
Stagnation: Closed water systems should be designed to minimise the

risk of biofouling. In practice this means avoiding areas of stagnation
including dead-legs. Dead-legs are lengths of pipework that are nominally
more than three times the pipework diameter in length and have no flow.
These should be eliminated or modified to provide water circulation.
Dead-legs allow undisturbed microbial proliferation and biofilms that can
affect water quality throughout the system. All closed water systems should
be designed to provide routine circulation through all wetted areas.
Dirt and debris: Dirt and debris will absorb and reduce the effectiveness
of biocides and other water treatment chemicals and provide areas of local
stagnation. Dirt and debris particles may also become incorporated in the
biofilm and may become a source of nutrients for bacteria.
Materials: The selection and installation of pipes, fittings and

equipment should avoid use of materials or design features that support
microbiological growth. Once colonised, an unsuitable material may
continue to seed the system with bacteria even after shock dosing with
biocide.

© BSRIA BG 50/2013
Table 7 suggests an assessment of microbial control applicable to a typical

system based on TVC results. The numerical levels should however be
adjusted by the water treatment specialist to take account of the specific
requirements of the system and also the incubation temperature used for
TVC determination. Also, an assessment of microbial control should not
be based on a solely on single result or set of results but on the trend of
those results.
Table 7: Assessment of microbial control
TVC in circulating water Microbial control

(cfu/ml)
>106 Biocide failing to provide protection
4 6
10 - 10 Unsatisfactory biocide control
102 - 104 Biocide control satisfactory
<102 Excellent biocide control
4.2.2 Selection of biocides

It is much easier and more cost effective to maintain microbiological
control within a closed heating and cooling system than to clean up a
badly fouled system containing a heavy build-up of biofilm.
Microbiological control of a system is achieved by:

s ensuring that the system remains free of debris and settled solids
s including an appropriate biocide in the pre-commission cleaning
process for new systems (or system flush for older systems)
s ensuring that an appropriate biocide is present in the system at the
correct concentration following pre-commission cleaning. This can
be classed as a maintenance biocide.
s on-going microbiological monitoring of the system.
s shock dosing of an appropriate biocide if levels or bacteria or
biofilm become unacceptable.
It is difficult if not impossible to clean up a microbiologically fouled

system with only one dose of biocide, as biocides are never 100% effective
at killing or dispersing slimes. Where shock dosing of a particular biocide
has not proved effective it may be necessary to carry out a laboratory
biocide evaluation in order to select an alternative biocide for use in the
system.
When selecting a biocide for use in a system, whether as a biocidal flush,

maintenance dose biocide or a shock dose biocide, a number of factors
need to be considered:
s Contact time: This is only appropriate for biocides used during
a biocidal flushing procedure. A biocide must be selected that
works optimally within the timescale of the actual biocidal flushing
process.
s Concentration: The biocide supplier will provide recommended
dose rates for their products. The system volume needs to be
known to ensure accurate dosing of the correct amount of biocide.

© BSRIA BG 50/2013
s Temperature: Most biocides have an optimum temperature range for

use, above which rapid degradation will occur.
s pH: Biocides have an optimum pH range for their use, with an
upper limit. More alkaline conditions may cause rapid degradation
of the biocide.
s Compatibility with the other chemicals present within the system,
particularly the corrosion inhibitors and glycols.
s Broad spectrum of activity against different bacteria or particular
effectiveness against target organisms.

s Implications of discharges to the environment
s Cost effectiveness
Although a broad spectrum biocide is ideal for flushing and maintenance

dosing, certain biocides are more effective against certain bacteria. If there
are specific problems with particular bacteria it is important to choose a
biocide and dosing regimen that has been shown to be effective against
these. This information can generally be obtained from the company
supplying the biocide.
Different dose rates are required for a maintenance dose biocide (to
prevent microbiological growth and therefore produce a biostatic effect)
and a shock dose biocide (to kill high microbiological numbers present
and bring the system under control and therefore produce a biocidal
effect). The activity of antimicrobial agents is often quantified as a
minimum concentration that is required to inhibit the growth of the target
organisms (minimum inhibitory concentration) or as a concentration that
leaves no detectable survivors after a specified contact time (minimum
bactericidal concentration, generally taken as the minimum concentration
required to achieve 99.9 percent reduction in bacteria).
The volume and frequency of biocide applications are dependent on

system volume, temperature, pH and biocide half-life under those
conditions and will be specific for each site. These parameters need to be
considered prior to the implementation of any biocide treatment regime.
The water treatment specialist must be aware of the possibility of the

build-up of immunity of organisms to biocides. This may require control
agents to be periodically changed even when they might still appear to be
effective (see Section 4.2.3).
Low temperature hot water (LTHW) heating systems that run

continuously with flow temperatures greater than 60°C should have
greatly reduced susceptibility to microbiological growth though all
systems have cooler sections and dead legs. Also, many heating systems are
switched off during the summer months and provide ideal temperatures
for bacteria to grow during these ‘down times’. It is recommended that
these are dosed with biocides prior to seasonal shut downs in order to
prevent growth of bacteria during this time. It is strongly recommended
that both heating and cooling systems continue to be circulated (for at
least an hour a day) during their off-season to avoid stagnation and provide
continuing protection from biocides and inhibitors (see Section 2.2.2).

© BSRIA BG 50/2013
Biocides can be divided into two main groups: oxidising biocides and non-
oxidising biocides.
Oxidising biocides
Commonly used oxidising biocides include chlorine, bromine, chlorine
dioxide and hydrogen peroxide. Oxidising biocides kill bacteria but will
also react with any organic matter present in the system. Also, at the
concentration required for effective antimicrobial activity most oxidising
biocides will be corrosive. The use of oxidising biocides in closed heating
and cooling systems is therefore not recommended.
Non-oxidising biocides
The following non-oxidising biocides are currently still permitted under
the Biocidal Products Regulation (BPR) for use in closed systems:
s Isothiazolinones
s DBNPA
s Bronopol
s Gluteraldehyde
s THPS
s Triazines
Details of the properties and application of these biocides are given in

Appendix C.
4.2.3 Microbial resistance

Bacteria can become resistant to biocides. However it is important to note
that the term “resistance” is often used loosely and somewhat inaccurately
in the context of closed systems. Unlike antibiotic resistance, an increase in
the minimum inhibitory concentrations of a biocide does not necessarily
mean that a biocide will have no effect. It may simply be that an
increased concentration of the biocide is required to achieve a sustained
improvement as most of the bacteria in the system are present within the
biofilm.
It is also possible that a particular biocide has limited impact in the system
due to degradation by high pH or temperature. After the initial dosing and
significant kill of bacteria, the biocide will be rapidly depleted to below its
effective concentration. Bacteria will then be released into the water from
the biofilm resulting in a sharp increase in the measured numbers. In this
case a biocide which has better pH and/or temperature stability should be
selected.
However, genuine resistance does occur. When a biocide is no longer

controlling bacteria at the recommended dose, alternating the treatment
with different biocide might help to rid of the system of any resistant or
harder to kill strains. The second biocide could be applied as shot dose.
The system dynamics need to be fully understood to decide on the most

effective treatment. Options might include a quick acting biocide followed
by a more persistent maintenance biocide, a change of biocide or a shorter
time between dosing the existing biocide. Generally it is recommended

© BSRIA BG 50/2013
to carry out a biocide screening test to determine the most effective

combination and concentration of biocides.
4.3 SCALE INHIBITION Calcium scale (limescale) is not usually considered an acute problem in
heating and cooling systems with minimal fresh water additions. However,
over the lifetime of a system, repeated draining and refilling of a system
can result in scale deposition that will significantly reduce heat transfer
efficiency, particularly in boilers. Most commercially available corrosion

inhibitors therefore contain chemicals for scale control.
In hard water areas, base-exchange softening of the fill water can be

beneficial. If this is carried out it is important that the water is not over-
softened and the chemical corrosion inhibitors are compatible with
softened water. Water softeners should not be left in-situ after filling. This
is discussed in Section 2.4.1.

© BSRIA BG 50/2013
5 NON-CHEMICAL WATER TREATMENT
5.1 FILTRATION 5.1.1 Strainers

All systems are fitted with strainers to protect pumps and control elements.
Strainers are intended to capture large debris that might cause blockage
or damage for example foreign objects, metal fragments, jointing tape and
large corrosion flakes. None of these should be present in a previously
cleaned system. Strainers do not retain fine particles of metal oxides, scale
and precipitates that contribute to the suspended solids result in a normal
system.
The element in an in-line “Y” pattern strainer will be either:

s a perforated screen (available hole sizes 0.8 – 3.2 mm)
s a mesh screen (available hole sizes down to 0.08 mm)
The typical strainer is supplied with a 0.8 mm stainless steel mesh though
different sizes can be specified. It is counterproductive to specify a smaller
mesh than is actually needed as this will increase the pressure drop and the
risk of blockage of the strainer itself if not regularly inspected. A Y pattern
strainer is shown in figure 9.
Figure 9: Y pattern strainer
On large pipework systems or where the concentration of large solids

is such that frequent cleaning of the filter may be required (e.g. open
condenser water systems), duplex basket filters are used. These would not
usually be required for routine operation of closed systems but may be
employed during flushing activities.
5.1.2 Dirt separators

Dirt separators are essentially in-line pots containing a mesh to create
a stagnation region that encourages sedimentation of particles from the
system water (Figure 10). The sedimented particles can be periodically
drained from the base of the dirt separator, which may be removed for
cleaning or mesh replacement.

© BSRIA BG 50/2013
NON-CHEMICAL WATER TREATMENT 5
Figure 10: Dirt separator

Drain
valve
Some manufacturers produce devices that combine dirt removal with

deaeration (see Section 5.2.1).
5.1.3 Magnetic filtration

Magnetic filters are similar to dirt separators but incorporate permanent
magnets to trap ferromagnetic material. They are widely fitted when
replacing boilers in domestic heating systems, mainly to protect the new
boiler heat exchanger and pump seals from magnetite that has been
previously formed in the old boiler and radiators. In principle, magnetic
filtration should be not be advantageous on a properly maintained system
with low suspended solids.
5.1.4 Hydro-cyclones
The principle of the hydro-cyclone is that the centripetal forces generated
in a rapidly spinning water column cause particles to move to a region
where they can be separated from the main water flow, collected and
drained off. The typical form is an inverted cone where the water enters
tangentially at the top. Clean water is discharged on the vortex axis at the
top, and particles collected via a drain at the bottom (figure 11).
The size range of particles removed by the hydro-cyclone depends on the

rotational velocity of the water column and the dimensions of the cyclone.
Increasing the efficiency and reducing the minimum size of particles
collected will increase the pressure drop and hence the pumping energy
absorbed by the cyclone.
Although cyclones could be sized for the total system flow (particularly
in flushing applications), they are most often installed as side stream
devices, usually with a separate pump. Although not as effective as media
filters, there is no cost of consumables apart from the energy requirement
associated with the pumping.

© BSRIA BG 50/2013
Figure 11: hydro-cyclone

Collector
Drain
valve
5.1.5 Disposable media filters

These are typically cartridge or bag filters, usually installed as a side stream
across the system circulating pump, accepting flow from the line on the
pressure (downstream) side of the pump and returning to the suction
(upstream) side of the pump. These filters remove mobile particles down
to 5 μm or less.
Side stream units are normally sized such that the whole water system
volume passes through the filter at least once over a 24 hour period. As
they are on a side-stream, clogging up the filter does not cause back-
pressure problems on the main system. Cartridge filters are periodically
removed and replaced with a new cartridge, the need for replacement
being determined by pressure drop or visual inspection.
The pore size of the filter media can be varied and reduced over time to
“polish” the system water.
5.1.6 Reverse osmosis

Reverse osmosis is a form of pressure driven filtration through micro-
porous membranes. It can be used to improve the quality of mains water
and borehole water supplies. It is not applicable to water already within
the heating or cooling system. Different types of membrane can be used
for the rejection of microscopic dirt particles and bacteria or for softening
i.e. the rejection of scale forming species (see figure 12).

© BSRIA BG 50/2013
One difference between reverse osmosis and conventional filtration is that

the rejected species are continuously carried away to waste, the wastewater
flow being up to 20% of the input flow. Therefore the systems are
relatively low maintenance, though the removal of chlorine and addition of
small amounts of chemicals to reduce scale formation may be required to
prolong the life of the membrane.
Figure 12: Reverse osmosis spiral-wound membrane element
Perforated central tube olution

Feed s
Anti-telescoping
device
Feed channel spacer
Membrane
a te
Perme
ntrate
Conce
Permeate collection material
Membrane
Feed channel spacer
Outer wrap
The sole application of reverse osmosis for closed systems is the pre-
treatment of fill water, particularly to provide a physical barrier against the
ingress of bacteria.
5.2 DEAERATION Cold tap water used to fill systems is often more than 95% saturated with
dissolved gases including oxygen. The equilibrium concentration of
oxygen depends on temperature, pressure and other dissolved components
in the water but could be as high as 10 mg/l. In order to manage the
corrosion risk, the oxygen concentration should ideally be reduced to less
than 1 mg/l, and as quickly as possible.
The solubility of oxygen in water decreases with increasing temperature

so normal operation of a heating system will start to reduce the oxygen
concentration compared to the fill water. Passive deaerators speed up
the process of reaching an equilibrium concentration of oxygen at the
maximum temperature to which the water is heated. Note that the
maximum temperature reached at the surface of a boiler heat exchanger
will generally be higher than the system flow temperature.
Once the system has achieved an equilibrium concentration of oxygen by

thermal evolution of dissolved air then further reduction in the oxygen
concentration is achieved mainly through chemical reaction with metal
surfaces i.e. corrosion processes.

© BSRIA BG 50/2013
Chilled water systems may never see a temperature higher than the
ambient in the building, so the dissolved oxygen can only be removed by
corrosion, microbiological processes, chemical oxygen scavengers (rarely
used) or the application of technologies such as vacuum deaeration.
Vacuum deaeration can also be used for heating system water. This and
other possibilities are described below.
5.2.1 Passive deaeration

All systems are should be fitted with manual and/or automatic air vents at
the top of risers and at high points where air (and other gases generated in
the system) will naturally collect when released from solution.
Passive deaerators encourage the release of air from solution by providing

a matrix that stimulates the formation of small bubbles, particularly after a
temperature or pressure change. Those bubbles then coalesce and can be
vented. Unlike normal vents, the system flow goes through the deaerator.
The deaerator can also include a dirt collection (sedimentation) function.
This is illustrated in figure 13.
Figure 13: Air vent and combined deaerator with dirt trap
Automatic
air vent
Stagnation
volume
Automatic
Branch air vent
Riser
Deaerator
and dirt trap
Drain
valve
5.2.2 Vacuum deaeration

Vacuum deaeration (also known as vacuum degassing) is the process of
exposing the system water (usually automatically in batches) to greatly
reduced pressure to encourage deaeration. Figure 14 shows the principle
of a batch deaerator where water is sprayed into a deaeration cylinder
under partial vacuum. This causes dissolved gases to be released from
the water and expelled through the vacuum pump. When the deaerated
water reaches level L2, the vacuum is then released and the deaerated
water pumped back into the system to complete the cycle. Since the
pressure reduction is created by the deaerator vacuum pump rather than
system water flow, the unit can be located at any convenient point in the
system. An additional expansion vessel may be installed to compensate

© BSRIA BG 50/2013
for the volume of water temporarily removed from the system during the
deaeration cycle.
Vacuum deaeration is sometimes integrated with the system pressurisation

set so that fill water goes through the deaeration cycle before it enters the
system and a single pump can be used for pressurisation and reinjection.
However, many vacuum deaerators are installed on a temporary standalone
basis, to provide rapid reduction of the oxygen concentration during
construction and commissioning of the system, and then removed.

Figure 14: Principle of vacuum deaerator
Vacuum
relief valve Vacuum
pump
L2
Deaeration
vessel
L1
Reinjection
Inlet pump
valve
Non-return
valve
Main Branch
5.2.3 Anodic oxygen removal

In passive anodic oxygen removal, system water is circulated at low
velocity through a vessel containing a highly reactive anode material (such
as magnesium or zinc) so that dissolved oxygen can be consumed by
electrolytic action between this and a cathodic material.
In active anodic oxygen removal, otherwise known as the ‘impressed

current method’, oxygen is removed by imposing a direct current on the
anode. The current, and hence the rate of oxygen removal, is controlled
in proportion to the concentration of dissolved oxygen measured by a
suitable detector located downstream of the reaction vessel.
In both these methods, the oxygen reacts with the anode, initially to form
hydroxyl ions which tend to elevate the pH of the system water. Also,
subsequent reactions of the metal ions may result in increased suspended
solids and scale.

© BSRIA BG 50/2013
5.3 SACRIFICIAL The use of sacrificial anodes is an extension of cathodic protection

ANODES
measures, whereby a vulnerable structure is coupled with a more active
metal, such as zinc or magnesium and occasionally aluminium. A galvanic
cell is produced in which the active metal becomes the anode and provides
a flux of electrons to this structure. The structure then becomes the
cathode, and is protected, while the anode is gradually destroyed, hence the
term sacrificial anode.
The most common applications of sacrificial anodes in building services

are tap water storage tanks, direct fired water heaters and stainless steel hot
water cylinders, particularly in areas of extremely hard water where the
oxygen and conductivity levels are high.
Sacrificial anodes are rarely used within closed systems. The principle can
however be adopted where lengths of pipework are to be protected from
external corrosion. Due to the low open circuit potential differences
between steel and for example a magnesium strip, one anode can only
protect a limited length of pipe line. However, this low potential
difference (voltage) can be preferable to a higher impressed voltage as the
danger of over-protecting some portions of the installation is less.
Sacrificial anodes require regular replacement and must be included in

the planned maintenance regime. It should also be borne in mind that
excessive or inappropriate cathodic protection can in fact accelerate
corrosion processes.
5.4 OTHER NON- Closed heating and cooling systems are predominantly treated with
CHEMICAL
TREATMENTS
chemical inhibitors and biocides. However, in recent years a number of
well-authenticated non-chemical methods of scale inhibition, corrosion
inhibition and bacterial inhibition have been reported, mainly in cooling
towers, but also occasionally in closed systems.
The authors of this text would strongly advise the owner/operator of any
system considering using a non-chemical system to seek the following
evidence before relying on innovative solutions:
s Proof of current usage:
- Is there authentic case study data relevant to closed systems?
- Are there references from users, independent of the supplier?
- Are site visits to existing installations possible?
- Is the supplier open about installations that may not have
performed to expectations and the reasons why?
s Effect of installation of the proposed unit on the existing treatment:
- If the unit purports only to control bacteria is it compatible with
the current corrosion inhibitor programme?
s Measures that will need to be put in place before installation for
comparative performance evaluations including:
- Corrosion rate (various metals)
- Total bacterial counts
- Biofilm monitoring
© BSRIA BG 50/2013
The following text describes some approaches that have been tried:
Magnetic water treatment: Magnetic water treatment has been

promoted since the 1930’s. The main application is to inhibit scale
deposition in domestic water systems. This is not relevant to closed
systems. Magnetic filters may be relevant to closed systems as discussed in
Section 5.1.3.
Electrostatic water treatment: These units generate a strong electric

field in a flow-through cell. This is claimed to disrupt bacteria growth as
well as reducing scale formation. The units are principally marketed for
cooling tower applications.
pH management units: These consist of vessels packed with reactive

mineral media, usually a blend of calcite and dolomite (calcium and
magnesium carbonates). The circulating water is passed through the unit
on a side stream basis. These units are claimed to be ‘self-regulating’ in that
while the initially-produced effluent from the bed is increased in hardness,
alkalinity and pH, this ‘surplus’ hardness is precipitated in the form of a
fine suspension which is then filtered out by re-passage through the media.
The claimed result is to produce water that is “naturally balanced” having
low hardness and high pH, which is non-scaling and non-corrosive.
Hydrodynamic cavitation: This is a heavily engineered system that

circulates a proportion of the system water through a ‘cavitation chamber’.
This imposes stresses on the water such that when it is super-saturated
with calcium bicarbonate, partial decomposition occurs and calcium
carbonate is precipitated (in the form of amorphous ‘aragonite’) with
the liberation of carbon dioxide. The precipitated calcium carbonate is
continuously removed via a downstream filtration system.
In addition to the shock-precipitation of calcium carbonate, the stresses

in the cavitation chamber also destroy bacteria (including legionellae). A
number of these systems are in use in the USA and Europe (including the
UK). Most are in evaporative cooling tower systems, although there are
reports of successful usage for bacterial control in closed systems.
Ultraviolet irradiation: Ultraviolet (UV) irradiation is undoubtedly

effective in disinfecting tap water and therefore useful in removing bacteria
from fill water. It is seldom used for treating water already inside closed
systems due to attenuation by suspended solids and other constituents
of the water including dissolved iron and inhibitors. Also, UV only kills
bacteria at the point of exposure. There is no residual effect in the water
so the biofilm in the rest of the system is unaffected.
Catalysed UV: A relatively recent entry to the water-treatment market

(for bacterial inhibition) has been the combination of UV with titanium
dioxide. Irradiation of a titanium dioxide coated surface by UV light of a
specific wavelength stimulates the production of hydroxyl radicals which
are effective against a range of bacteria, fungi etc. This is more effective
than UV alone and is claimed to prevent the build-up of biofilms in
circulating systems. The technology is currently confined to the treatment

© BSRIA BG 50/2013
of fill-waters and cooling towers but may have a future role in closed
systems.
Ultrasonic cavitation: These units generate high frequency ultrasonic

waves in a side stream treatment chamber, creating shear stresses that are
lethal to bacteria. The units are principally marketed for cooling tower
applications but may be used in other recirculating water applications.
Although ultrasonic treatment operates principally within the treatment
chamber, it is claimed that the interference with microbial populations has

an indirect effect on biofilm.
Pulsed power units: These units consist of a solenoid coil around a

reaction chamber though which this system water is passed on a side-
stream basis. The water passing though the chamber receives a series
of high frequency electric pulses and a varying DC field. The units are
claimed to control scale formation by the controlled precipitation of
calcium carbonate in the amorphous (aragonite) form. They are also
claimed to inhibit bacteria by a process termed ‘electroporation’ where
the high frequency pulsing action damages the membranes of planktonic
bacteria. The units are principally marketed for cooling tower applications.

© BSRIA BG 50/2013
MANAGEMENT OF THE WATER TREATMENT PROGRAMME 6
6 MANAGEMENT OF THE WATER TREATMENT PROGRAMME
6.1 OVERVIEW 6.1.1 Roles and responsibilities

It is essential that any water treatment programme is implemented within
a clear and documented framework of roles and responsibilities. In many
cases, corrosion failures in the early life of a closed heating or cooling
system have been traced back to management failures, particularly in the
period between pre-commission cleaning and handover. This period is

now addressed by specific guidance in BG 29[1]. However, loss of control
at any stage in the life of the system can rapidly lead to failures.
There are four main actors associated with the water treatment
programme:
s Client: the organisation that commissions the cleaning or water
treatment activity
s Cleaning specialist: a person or organisation appointed to carry out
pre-commission cleaning or remedial cleaning of a pipework system
and related activities
s Water treatment specialist: a person or organisation appointed
to maintain the water quality and manage the corrosion, scaling,
sedimentation & biofouling risk within a pipework system; and to
supply the water treatment products.
s System operators: Staff who control the system on a daily basis and
who may be responsible for interim monitoring tasks and in some
cases interim dosing of chemicals.
6.1.2 Objectives of monitoring

It is essential that regular monitoring is carried out to evaluate the
effectiveness of the water treatment programme. That may include
sampling and analysis of system water, direct monitoring of corrosion
rates and inspection of corrosion coupons. It should not be assumed
that because a treatment programme has performed satisfactorily in one
application it will be equally successful in another. All systems are different
both in design and operational history.
The objective of monitoring a closed heating or cooling system is to

provide information about the current condition of that system and/or the
water within it.
Monitoring should be carried out using appropriate methodologies and

equipment to ensure that the results obtained are accurate, meaningful
and comparable. Samples should be taken at a sufficient number of
representative locations to provide an overall view of the system condition
and to allow for assessment of the significance of the results.
Water analysis alone cannot give the full picture of the condition
of the system and, since it is only done periodically, rarely picks up
sudden changes in the system condition e.g. the loss of inhibitor due to
unscheduled drain-down or leakage. However, there are opportunities for
the continuous monitoring of certain system parameters as explained in
Section 6.6.4.
© BSRIA BG 50/2013
6.1.3 Frequency of sampling

Sampling and analysis of system water is an essential part of a treatment
regime and should include both the circulating water and fill/make up
water supplied to the system. Sampling and analysis may be carried out,
at various times, by the cleaning specialist, water treatment specialist or
system operators, provided they have been trained to do so.
The programme of monitoring appropriate to each system should be

defined, documented and implemented by the water treatment specialist.

The frequency and extent of this routine monitoring will depend on the
operating characteristics, susceptibility to disturbance, and operational
importance of the system and may vary through the life of the building.
The typical frequency of routine monitoring for heating and cooling

systems in non-domestic buildings in mid-life is one to three months.
More frequent monitoring may be desirable during the first 6 months of
operation of a new system and immediately after any significant system
changes such as replacement or addition of plant or terminal units.
Detailed guidance is provided in BS 8552[6].
6.2 WATER Sampling and analysis of the water circulating through the main plant
SAMPLING AND
ANALYSIS
is the primary method of assessing the efficacy of the water treatment
programme i.e. control of scale, suspended solids, corrosion and biofouling.
Other methods such as corrosion and biofouling coupons may be used
in a supporting role. Sampling, analysis, interpretation of results and the
recommendation of corrective actions should be carried out by persons
with suitable training, competence and experience.
BS 8552[6] recommends that additional samples are taken from remote

locations in the building to help demonstrate that the water treatment
regime is effective throughout the system. The number of such samples
depends on the volume and complexity of the system.
If these remote location samples are taken from low flow areas including
terminal units then they may contain higher concentrations of solids and
bacteria than at the main plant. This effect should be considered in the
interpretation of results. Where the results are deemed unacceptable then
further investigation and specific remedial actions may be required for
example remedial flushing or operational changes.
Routine monitoring can be carried out utilising site test kits. These
are available for determination of most chemical and microbiological
parameters of importance to water treatment. The benefit of test kits for
routine monitoring of chemical parameters is that they provide rapid on
the spot results that can be acted upon without the delays associated with
analysis by external laboratories. However it is essential that the selected
test kit provides a suitable range and accuracy for the analysis and the
operator is aware of the limitations of the test method including possible
interferences.

© BSRIA BG 50/2013
If the cleaning or water treatment specialist employed is not the product

supplier they should consult with the product supplier as to which analyses
are to be used to quantify the levels of active inhibitors in the system and
control limits to be maintained. These should be clearly documented in
the site/system water treatment logbook.
All results should be recorded in the site/system water treatment log book.
Where results exceed action limits then predefined remedial actions should
be carried out and recorded or the matter reported to the responsible

person for action.
6.3 SAMPLING The detailed scope of sampling and analysis should be defined and
FOR WATER
CHEMISTRY
documented by the water treatment specialist in the sampling and analysis
plan. This scope should include key components of the water treatment
programme and various indicators of water quality (see Section 6.5).
Detailed guidance on sampling, sample handling and reference analytical

methods is given in BS 8552[6], but an overview is given below.
6.3.1 Sample bottles

Where the sample is destined for laboratory analysis it is preferable to
use single use sample bottles provided by the laboratory. This will ensure
that the correct bottle is used and sufficient volume is available for the
proposed analysis.
Where the sample is destined for immediate on-site analysis then a clean
high density polyethylene bottle with a wide neck is suitable for collecting
the sample prior to chemical analyses.
6.3.2 Sample labelling and paperwork

All sample bottles should be suitably labelled using an adhesive label or
permanent marker. The information provided should, as a minimum,
allow the sample to be clearly identified, state the analysis required and
allow the sample to be referenced to the sampling worksheet (sometimes
referred to as the chain of custody form).
Where the sample is destined for laboratory analysis, pre-numbered

labels and a corresponding chain of custody form can be supplied by the
laboratory. The level of information provided with the sample should, as
a minimum, unambiguously identify the precise origin of the sample, the
condition of the sample when taken and the required analysis.
The information required on the chain of custody form should include:

s Date and time of sampling
s Client/site address
s Origin of the sample (system and sample point)
s Type of water being sampled (chilled, heating, condenser water etc.)
s Identity (name & organisation) of the sampler
s Analysis required

© BSRIA BG 50/2013
When the sample arrives at the laboratory the chain of custody form
should be signed by the person receiving the sample, with time and date.
The laboratory will allocate a unique reference number to the sample (if
not already printed on the label) that will allow it to be tracked during
analysis and reporting.
BS 8552[6] suggests additional information to be recorded including:

s Nominal water treatment package (nominal concentrations of
inhibitors, biocide, anti-freeze etc.)

s Temperature
s Visual appearance (clarity, colour, presence of settled solids)
s Odour (ammoniacal, sulfide etc.)
s Storage conditions (whether refrigerated)
s Other observations (anything unusual about the sample or the way
it was collected)
The same information should be recorded for samples analysed on site.
6.3.3 Sample collection

Not all potential sampling points are suitable for all analyses, particularly
for settled or suspended solids.
Table 8 summarises which sample points are suitable for different samples.
[6]
Table 8: Selection of sampling points (from BS 8552 )
Type of sample Sampling Point
Full bore drain Reduced bore drain Pressure test point
point point
Settled solids Yes No No
Suspended solids & Yes Yes No

“total” metals
Dissolved solids Yes Yes Yes
Microbiology Yes Yes Yes
Dissolved oxygen No See BS 8552, A.2 See BS 8552, A.2
6.3.4 Storage and transportation

Samples should be stored and transported generally in accordance with BS
EN ISO 5667-3[13].
Samples should be analysed as soon as possible. Samples for laboratory

analysis should arrive such that analysis can be commenced within 24
hours of the sample being taken. The impact of the time between sampling
and analysis should be minimised by:
s filling sample containers to the brim;
s storing samples in the dark; and
s refrigerating samples to between 2°C and 8°C, except where
otherwise indicated in the analytical method.

© BSRIA BG 50/2013
It is sometimes argued that samples for chemical analysis do not need to

be refrigerated and there are pros and cons to doing so. Refrigeration will
reduce the activity of certain bacteria that might influence the chemistry
(particularly nitrite concentrations) but it may also increase the rate of
precipitation of dissolved solids from solution.
6.4 SAMPLING FOR Sampling is an important part of the analytical process and the manner
MICROBIOLOGY
in which the sample is collected and subsequently stored and transported

prior to analysis can greatly affect the final results. Results from poorly
handled samples can then be misleading and result in inappropriate actions
or treatments being undertaken.
A sample should be representative of the water within a system. Ideally

the water will be exactly the same when it is tested in the laboratory as
when it was in the system. In reality the sample will start to change as
soon as it is put in a sample bottle. In order to minimise these changes
the sample should be collected, stored and transported in an appropriate
manner. Detailed guidance on sampling, sample handling and reference
analytical methods is given in BS 8552[6], but an overview is given below.
6.4.1 Sample bottles

Sample bottles for microbiological analysis should be made of plastic or
glass, pre-sterilised and free from any toxic substances. Where the sample
is destined for laboratory analysis it is preferable to use single-use sample
bottles provided by the laboratory. This will ensure that the correct
bottle is used, the required preservative is present and sufficient volume is
available for the proposed analysis
Sample bottles provided by laboratories are usually pre-loaded with a small

quantity of sodium thiosulfate solution, equivalent to a final concentration
of approximately 18 mg/l in the full bottle. Sodium thiosulfate neutralises
any residual oxidising biocide that may be present in the sample as that
could change the result. Pre-loaded bottles should only be used for the
microbiological sampling. If the bottle is damaged in any way, or the seal
is broken, the bottle should not be used.
Ideally a separate bottle (which should be clean but not necessarily

sterile) should be used for chemical analysis (see Section 6.3). If there
is no sodium thiosulfate in the microbiological sample bottle and there
is sufficient sample, both the microbiological and chemical analysis
may be undertaken from that bottle. However in this situation, the
microbiological analysis must be undertaken first in order to minimise
the risk of contaminating the sample. Conversely microbiological analysis
should NOT be undertaken from a chemical sample bottle as it may not
have been sterile before use.
6.4.2 Sample labelling and paperwork

Refer to section 6.3.2.

© BSRIA BG 50/2013
6.4.3 Sample collection

Samples of system water for microbiological analysis can be taken from
dedicated sampling points, drain valves or binder points. BS 8552[6]
provides guidance on the selection of sampling points for other samples
(refer to Table 8).
The following procedure should be used for drain valves:

1. Disinfect the sampling point prior to taking the microbiological
sample by spraying with propan-2-ol (minimum 70% solution). This

should be allowed to evaporate before the sample is taken. If a binder
point is used, the binder probe should be sterilised internally and
externally before use. Samples should not be taken through a hose or
non-sterile conduit.
2. Fully open the sample valve, discharge sufficient volume of water to
flush the dead leg into a waste container (for disposal) and close the
sample valve.
3. Re-open the sample valve and discharge the required volume of water
slowly into the sample bottle. The sample bottle should be held in one
hand and the cap removed and held in the other – do not put the cap
down and take care not to touch the neck of the bottle with fingers.
The bottle should be inserted into a steady flow of water from the
outlet, without touching the neck of the bottle sample point. The
bottle should be filled but not overflowed.
If using a binder point and tube, the needle and tube should be disinfected
before use by flushing the inside, and swabbing the outside, with propan-2-
ol solution. There is not usually any system dead leg but the binder probe
should be inserted and the sample tube briefly flushed to remove any
residual disinfectant before collecting the sample as described above in
step 3.
Note: It has been common practice when taking water samples for
microbiological analysis from open systems to leave a small airspace in the
sample bottle. However, given that the sample is being extracted from
a low oxygen environment in a closed system, BS 8558[7] states that the
sample bottle should be filled to the brim. This is to avoid increasing the
dissolved oxygen concentration during handling and is consistent with
BS EN ISO 5667-3[13].
The chemical samples bottles should be filled in the same way.
When taking both chemical and microbiological samples from the same
sample point, the microbiological sample should be taken first.
6.4.4 Storage and transportation

Samples should be stored and transported generally in accordance with
BS EN ISO 5667-3[13].
Samples should be analysed as soon as possible. Samples for laboratory

analysis should arrive such that the analysis can be commenced within

© BSRIA BG 50/2013
24 hours of the sample being taken. The impact of the time between
sampling and analysis should be minimized by:
1. Filling sample containers to the brim (as described in Section 6.4.3)
2. Refrigerating samples to between 2°C and 8°C (except where
otherwise indicated in the analytical method). Cool boxes with ice
packs can be used for transportation.
3. Storing samples in the dark
6.4.5 Laboratory analysis

The majority of methods used for the analysis of water from closed
systems are based on, but not necessarily identical to, those used for the
analysis of bacteria in drinking waters. These include quantitative methods
that give an estimate of bacteria numbers and semi – quantitative methods
that give an indication of bacteria numbers. The analysis may be designed
to encourage growth of the general population of bacteria (e.g. total viable
count) or to highlight specific bacteria, either by using a selective growth
media or by causing the bacteria of interest to stand out from the other
bacteria in some way (e.g. colour changes).
In order to obtain a bacterial plate count the organisms are grown on

a suitable medium until the colonies formed reach a size that can be
counted (see figure 15).
Figure 15: Plate count
There are several reasons why the final result will always be an estimate:
1. The sample may not be representative of the number of planktonic
bacteria in the system. Occasionally a fragment of biofilm containing
hundreds of bacteria will end up in the sample and may distort the
result.
2. The number of bacteria in the sample could increase or decrease
during transportation and while waiting for the analysis to begin,
depending on how the sample is handled.

© BSRIA BG 50/2013
3. Growth media do not provide 100% recovery of bacteria in the sample

– it could be as low as 60%. However, this effect should be consistent
between samples.
4. It is assumed that one colony comes from one bacterium but because
bacteria sometimes clump together, a single visible colony may in fact
be derived from a number of individual bacteria. The result is therefore
expressed as a colony forming unit per standard volume rather than a
definitive number of bacteria.
Table 9 lists the main microbiological tests and the time required to obtain
a result. These tests are further explained in the following sections.
Table 9: Comparison of the main microbiological tests

Test Time to result Results
Pseudomonads 2 days (plus 2 further days if cfu/100ml

identification is required)
Total viable count 3 days cfu/ml
Nitrite reducing bacteria 5 days Presence/absence
Semi-quantitative sulfate 6 days (or 21 day reference Presence/absence

reducing bacteria method)
Quantitative SRB 1 day cfu/ml
6.4.6 Pseudomonads
Pseudomonads are “pseudomonas–like” bacteria. At present there is no
standard method specifically for the analysis of pseudomonads in closed
systems. BS 8552[6] states that a suitable reference method is BS EN ISO
13720[14], using a pseudomonas selective agar base with CFC selective
supplement for “pseudomonas species”. The incubation period is 48 hours
at a temperature of 30°C.
In practice most laboratories will use an adapted drinking water method

with a proprietary selective agar. The principle of this method is for 1
ml of sample to be filtered through a membrane filter with a pore size of
0.45 μm. The pore size of the filter membrane is small enough to trap all
the bacteria from the sample on its surface. The drinking water method
requires 100 ml to be filtered but closed systems often have much higher
numbers of pseudomonads so there may be to many bacteria to count if
100 ml is used.
The membrane and attached bacteria is then transferred to a selective

media designed to encourage the growth of the pseudomonads and inhibit
the growth of other bacteria that are not of interest. The membrane and
media are then incubated at 30°C for 48 hours
After incubation the membrane is examined for blue-green, greenish-

brown, yellow-green or colourless colonies, or colonies which exhibit
fluorescence under exposure to ultraviolet light. Any such colonies are
counted and recorded. These colonies are considered to be “pseudomonas

© BSRIA BG 50/2013
species” or pseudomonads. On occasion the analysis may be taken a step

further to allow for the identification of the predominant bacteria types.
Note that the results are usually expressed as cfu per 100 ml for
consistency with guidelines in BG 29[1]. Results provided “per ml” should
therefore be multiplied by 100 for comparison with guidelines.
6.4.7 Total viable count (TVC)

This analysis is considered to provide an indication of overall water quality
with respect to bacteria. However it is recognised that this method will

measure only a proportion of the bacteria present as only aerobic bacteria
that grow on the nutrient media and at the incubation temperature of
the method will be recorded. The TVC result is nevertheless useful for
monitoring the trend in water quality, particularly to identify loss of
control of bacteria.
BS 8552[6] states that a suitable reference method is BS EN ISO 6222[12]

(see also BS EN ISO 19458[15]).
For the laboratory analysis, 1 ml of sample is spread on the surface of a

nutrient rich media. Alternatively a 1 ml volume of sample is added to
a Petri dish and mixed with molten nutrient rich media and the media
allowed to set.
The next step is to incubate the plate for the specified incubation
temperature and time: 22°C for 72 hours, 30°C for 72 hours or 37°C
for 48 hours depending on the application and objectives of the analysis.
Following incubation, the plates are counted and the result expressed as
cfu/ml.
Note that since the result of the TVC test is sensitive to the incubation
temperature, results obtained at different incubation temperatures cannot
be directly compared or included in the same trend analysis.
TVC results can also be obtained using test kits, in which case the
manufacturer’s instructions should be followed. Note that the results from
laboratory analysis and test kits are likely to differ (though they should be
self-consistent) and should not be combined for the purpose of assessing
trends.
6.4.8 Nitrate/Nitrite reducing bacteria (NRB)

The usual method for determination of NRB is a semi-quantitative
proprietary test kit. For one such test, 2 ml of sample is added to a tube of
indicator media. This is then incubated at 30°C for five days, with the tube
being examined every day and any changes recorded.
If NRB are present the media turns pink and bubbles of ammonia gas
are produced. The time it takes for these reactions to take place gives an
indication of the number of bacteria present. A positive reaction after
1 day of incubation indicates heavy contamination (> 106 cfu/ml) and
a positive reaction after 5 days indicates low level contamination (10 -
100 cfu/ml). However, for practical purposes this should probably be
considered as a presence/absence test.
© BSRIA BG 50/2013
The shortcoming of this method is that the generation of the pink colour
is due to an increase in the alkalinity of the media due to the production
of ammonia by the NRB. If the sample is alkaline in the first place, the
media will change to pink regardless to the presence of NRB. Conversely
if the sample is acidic the media will turn yellow, and even if NRB are
present they may not be able to generate enough ammonia to change
the colour of the indicator and therefore a false negative result may be
reported.
There are also dip slide versions of this test.
6.4.9 Sulfate reducing bacteria (SRB)

BS 8552[6] recommends the 21 day SRB test for reference purposes but
shorter tests may provide sufficient confidence for routine management of
the water system provided sufficient numbers of samples are collected and
analysed (see BS 8552[6] Annex C).
Test kit methods rely on the fact that SRB reduce sulfates to produce
hydrogen sulfide gas, which in turn results in a blackening of the culture
media.
In the first method, 2 ml of sample is added to the surface of the

indicator gel in a tube. The tube is then incubated at 30°C for five days
and checked daily for a blackening of the media. If a black line appears
on the surface it indicates a positive reaction. The number of bacteria
present is then estimated by how quickly the blackening moves down the
tube. For example if 100% of the tube is black within 2 days there are
>100,000 cfu/ml, but if there is only 20% blackening after 6 days there are
approximately 100 cfu/ml.
The second method is a quantitative method based on that developed

for drinking water analysis. A volume of the sample (100 ml) is filtered
through a membrane filter with a pore size of 0.45 μm. The pore size of
the membrane is small enough to trap all the bacteria from the sample on
its surface. The membrane is then transferred, face-upwards to a selective
nutrient medium and incubated for 24 hours at 37°C under anaerobic
(without oxygen) conditions. Each viable organism responsive to the
media should multiply to give a visible colony.
Colonies of sulfate reducing clostridia are recognised by specific growth

characteristics and interaction with the selective media. Black colonies are
counted and recorded as sulfate reducing bacteria.
Caution needs to be applied to the interpretation of results because of

the specificity of growth media to SRB species. For example, in many
affected systems the predominant SRB is desulfovibrio desulfuricans rather
than clostridium and this may respond differently in different tests. A
study of SRB methods and confidence limits based on experimental data
is discussed in BS 8552[6] Annex C.

© BSRIA BG 50/2013
6.4.10 Other microorganisms

Usually only the organisms mentioned in Table 10 are included the routine
analysis of closed systems. However, there are numerous other species of
bacteria and fungi that can be found within closed heating and cooling
systems. Where a system has a particular problem or routine control
measures appear not to be working, additional analysis may be required,
especially a check for the presence of fungi in cooling systems. Under
certain conditions fungi can proliferate and severely affect the efficiency of
the system.
6.5 WATER QUALITY The possible range of analysis likely to be carried out on site on samples
ANALYSIS AND
GUIDELINES
from closed systems together with guideline control limits, where
applicable, are detailed in Table 10 below. Notes are provided below
the table. The essential analysis and specific control limits for the system
and water treatment programme in use must be specified by the water
treatment specialist.
Table 10: Comparison of the main microbiological tests

Group Parameter Guideline Note
Physical Visual appearance Clear with no visible suspended solids 1

condition (suspended/settled solids) or gassing
Odour No strong sulphurous or 2
ammoniacal smell
pH Within specified range 3
Conductivity and total Within specified range 4

dissolved solids (TDS)
Suspended solids Limit as specified for the circulating 5
water. No increasing trend
Settled solids (as defined in Limits as specified for pipework at 6
BG 29
[1] [6]
& BS 8552 ) the extremes of the systems and for
terminal units
Water Nitrite (where used) Within specified range 7
treatment
status Nitrate Within specified range 8
Molybdate (where used) Within specified range 9
Boron Within specified range 10
Phosphonate Within specified range 11
Organic inhibitors See note 12
Silicate Within specified range 13
Orthophosphate Within specified range 14
Hardness Total alkalinity For information 15
Calcium hardness For information 16
Total hardness For information 17
Cont...

© BSRIA BG 50/2013
Table 10 cont: Comparison of the main microbiological tests

Group Parameter Guideline Note
Corrosion Total iron Limit as specified 18

products
Dissolved iron Limit as specified 19
Copper Limit as specified 20

No increasing trend
Zinc Limit as specified 21
No increasing trend
Aluminium Limit as specified 22
No increasing trend
Other Sulfate >90% of fill water sulfate 23
No decreasing trend
Chloride Limit as specified 24
Ammonia Limit as specified 25
Dissolved oxygen For information 26
Microbiology Total viable count (TVC) Limit as specified 27

No increasing trend
Pseudomonads Limit as specified 28
No increasing trend
Sulfate reducing bacteria Not detected 29
(SRB)
Nitrite reducing bacteria For information 30
(NRB)
Notes to Table 10:

1. Visual appearance.This is the simplest test but also the most subjective and
difficult to report consistently. Some practitioners recommend that the sample
should be photographed routinely or if there is anything at all unusual. It is
also possible that the sample can significantly change between sampling and
arrival at the laboratory e.g. due to precipitation of dissolved iron on exposure
to atmospheric oxygen. Cloudiness that clears quickly on standing is probably
due to dissolved gases such as air, nitrogen or carbon dioxide released when the
sample is depressurised and the reasons for this should be investigated.
2. Odour. A strong foul or sulfide smell is suggestive of stagnant anaerobic
conditions and significant SRB activity. This should not be present in a
normal system. A strong ammoniacal smell is suggestive of NRB activity
though a slight ammoniacal odour can also be associated with some water
treatment chemicals. The microbiological status can be checked with the
relevant bacteria test.
3. pH. All closed system water treatment is intended to produce neutral
to alkaline conditions i.e. above pH 7. The maximum allowable pH is
dependent on the materials present e.g. up to pH 8.5 for systems containing
aluminium but up to pH 11 for all steel systems. The pH will usually
decline if there is a loss of water treatment. An increase in pH over the upper

© BSRIA BG 50/2013
limit of the range may indicate that the wrong water treatment chemical or
wrong dosage has been added to the system.
4. Conductivity and TDS. The relationship between conductivity and TDS
varies according to the chemistry of the system so any estimate of TDS based
on conductivity reading should be treated with caution. Ideally the specified
range should be expressed as conductivity as this will be the easiest to measure
on site. A sudden drop in conductivity could be indicative of water loss from
the system. For TDS determination the sample should be evaporated at
105°C. The laboratory should be made aware of the presence of glycol as this
can interfere with the results.
5. Suspended solids. This is the key indicator of water quality. Suspended
solids should normally be less than 30 mg/l in the circulating water and a
well-controlled system may consistently achieve less than 10 mg/l. A high
suspended solids result may indicate poor control of corrosion rates and should
be considered together with the results for corrodible metals.
6. Settled solids (as defined in BG 29[1] and BS 8552[6]).The result will
indicate the degree of sedimentation of suspended solids at the sample location
since the last discharge of water from that sample point (for sampling or
maintenance). BG 29 suggests that the limit for settled solids in pipework
at the extremes of the system should be no more than double the suspended
solids limit (i.e. 60 mg/l) and less than three times for terminal units (i.e.90
mg/l). If the sample point has been used within the last 3 weeks and the
system is well controlled there should be no significant increase over the
suspended solids result. In that case the limit for either should be less than
150% of the suspended solids limit (i.e.45 mg/l).
7. Nitrite is the principal corrosion inhibitor in many water treatment packages
and it is essential to maintain the concentration in the range recommended by
the water treatment specialist. A low concentration of nitrite could actually
increase the corrosion rate. A rapid loss of nitrite may indicate the presence of
nitrite reducing bacteria. This should be confirmed with the NRB test.
8. Nitrate may be included as a component of the water treatment package (to
synergise with nitrite) but can also be formed by oxidation of the nitrite.
9. Molybdate is a common inhibitor in many water treatment packages.
Molybdate will gradually be lost through precipitation and over-dosing can
result in an increase in suspended solids.
10. Boron can be present in the water treatment package as borate.
11. Phosphonate may be included in the water treatment package as a scale
inhibitor. Loss of concentration is not critical.
12. Organic corrosion and scale inhibitors can be difficult and/or expensive to
monitor. The water treatment specialist will provide advice on what is feasible
and the associated guidelines.
13. Silicate may be present at trace levels in the water supply and as a film
forming inhibitor in the water treatment package for high pH water treatment
programmes.
14. Orthophosphate may be included in the water treatment package as a
corrosion inhibitor.

© BSRIA BG 50/2013
15. Total alkalinity (also known as methyl orange alkalinity from the use of
methyl orange indicator in the analytical titration) is used together with pH to
calculate the carbonate concentration in the sample.
16. Calcium hardness is a measured of the scale forming potential of the water.
17. Total hardness is the sum calcium and magnesium hardness of the water
expressed as mg/l CaCO3.
18. Total iron. The result is used to comment on the source of suspended and
settled solids.
19. Dissolved iron is used as a primary indicator of corrosion control in systems
containing steel. A limit of 3 mg/l or less may be set for typical nitrite/
molybdate regimes (see BG 29[1])but may be much higher for some proprietary
products. Care must be taken in the interpretation of dissolved iron results in
high pH systems, and where there is only a small amount of steel, as a low
result may not mean that the system is safe.
20. Copper (dissolved or total). New systems with fresh copper surfaces are
vulnerable to cupro-solvency and high concentrations of dissolved copper could
result in pitting corrosion of steel and aluminium.
21. Zinc (dissolved or total). Zinc plated surfaces should not be used in heating
systems as the plating will be rapidly corroded. If zinc is found in other
circumstances it could be indicative of dezincification.
22. Aluminium (dissolved or total) is an indicator of corrosion control in systems
with aluminium components.
23. Sulfate is present in the fill/make-up water. The concentration of sulfate
in the system water should remain above 90% of that for the fill water. A
downward trend in sulfate could indicate the presence of active SRB.
24. Chloride is present from the fill water and the degradation of biocides. A high
concentration of chloride will increase the risk of pitting and stress corrosion.
There is no generally accepted “safe” level of chloride.
25. Ammonia can be produced by the action of bacteria on nitrite, nitrate and
organic water treatment components. High concentrations of ammonia are
detrimental to copper and copper alloys. Some practitioners recommend
remedial action if ammonia concentrations exceed 30 to 40 mg/l. However,
there is no published evidence to justify this recommendation in terms of an
increased risk of SCC.
26. Dissolved oxygen is required for most of the metallic corrosion processes
in closed systems. Cold tap water is often nearly saturated with dissolved
oxygen. Since corrosion processes will tend to reduce the concentration of
dissolved oxygen in the system water, the trend of results should be decreasing.
A persistently high concentration (>2 mg/l) in circulating water may be
indicative of oxygen continuing to enter the system e.g. though inadequate
system pressurisation (see section 2.3) or through plastic pipe. Very low
concentrations of oxygen relative to the circulating water in samples taken from
terminal units may indicate stagnation and/or local corrosion.
27. Total viable count (TVC) is a measure of the concentration of aerobic bacteria
in the water.These bacteria may not be implicated in corrosion but high levels

© BSRIA BG 50/2013
of TVC are indicative of poor microbiological control.The typical control

range is 1,000 to 10,000 cfu/ml (see Section 4.2.1 and Table 7) and
no increasing trend.This is based on the geometric mean of the samples as
explained in BS 8558[7].
28. Pseudomonads are slime forming bacteria that promote the formation of
biofilms. They are obligate anaerobes which means they can act as either
aerobic or anaerobic depending on conditions within the system.When
anaerobic they act as nitrate/nitrite reducers
There is no defined relationship between the concentration of pseudomonads

in the water and the much higher concentration in the biofilm where they do
the damage (see Section 3.2). High concentrations of pseudomonads in system
water are however taken as indicative of poor microbiological control.
BG 29[1] states that the limit for pseudomonads in system water at handover
should be less than 10,000 cfu/100ml (100 cfu/ml) and no increasing
trend. This is based on the geometric mean of the samples as explained in BS
8558[7]. Opinions vary about the significance of high pseudomonad counts
in older systems with some practitioners arguing that the limits should be the
same as for TVC at 30°C and that 100 cfu/ml would certainly be no cause
for concern.Various data from actual sites suggest that pseudomonas counts
are typically around 10% of TVC counts in a “dirty” system (one operating
around the TVC limit) but only 1% of TVC in a clean system, hence the
criteria in BG 29.The water treatment specialist should decide on a suitable
control limit for pseudomonads based on his knowledge of the system and
operational requirements.
29. Sulfate reducing bacteria (SRB) are anaerobic bacteria that are a significant
cause of under-deposit pitting corrosion in pipework systems. As with other
bacteria, there is no relationship between the concentration of bacteria in
the water and in the biofilm. There are also difficulties in the sampling
and analysis as SRB may become dormant when exposed to oxygen in the
sample bottle and there are at least three different methods of analysis that
may respond differently to different SRB (see Section 6.4.9). A positive
result should be treated as an indication of the presence of SRB but a “non-
detectable” result does not mean the absence of SRB.
30. Nitrite reducing bacteria (NRB) reduce nitrite (or nitrate) inhibitor to
ammonia or nitrogen gas. NRB analysis is generally carried out to investigate
the reasons for rapid loss of nitrite. In most situations the major concern is
the loss of inhibitor, though a high concentration of ammonia is corrosive to
copper alloys. NRB include some pseudomonads when exposed to anaerobic
conditions.
Conclusions about the state of the system and any decisions about
remedial work should take account of all the available evidence. Often
the first step in dealing with water that falls outside the acceptable range
for a particular parameter should be to repeat the sampling to confirm the
result.

© BSRIA BG 50/2013
6.6 CORROSION Analysis of system water quality provides useful information about the
MONITORING
concentration and performance of corrosion inhibitors and biocides but
only gives a partial picture of the overall condition of the system and the
extent to which corrosion is taking place. Several additional techniques
and methods of corrosion monitoring can be applied to provide a more
complete picture of the health of the system.
The key ingredients of a corrosion monitoring programme are:

1. System knowledge
2. Site testing of critical performance criteria
3. Documentation of observations and recommendations
System knowledge: Before taking on responsibility for monitoring

and treating a system, the water treatment specialist should review the
operational and design characteristics of the system. They should be aware
of:
s Materials present
s Working temperature and pressure
s Operating volume (from previous testing)
s History of the system (including any pre-commission cleaning
activity, any known corrosion problems and remedial works)
s History of the water treatment (programmes and products used)
Although some of this information may be available in the operation and

maintenance (O&M) manuals for the building it should be copied to the
water treatment logbook for easy reference.
Documentation: The water treatment specialist should document

the system parameters, site testing results, and any observations and
recommendations provided. Verbal communication of critical observations
and recommendations must be confirmed in writing. It is also a good
practice to document (as explicitly as possible) the consequences of not
heeding these observations and recommendations.
Site testing techniques are described in the following sections.
6.6.1 Corrosion coupons

Corrosion coupons are used to assess the rate of corrosion of metals in
cooling and heating systems. Standard coupons of most common metals
and alloys are commercially available, usually with a bead-blast finish.
These come in vapour phase inhibitor impregnated bags to prevent any
corrosion before use. They may be supplied pre-weighed but, if weight
loss determinations are to be performed by a laboratory other than the
coupon supplier, they should be weighed immediately before use to a
resolution of 0.1 mg.
The coupon materials should correspond to the at-risk materials in the

system for example mild steel, copper, brass or aluminium. Note however
that extruded aluminium coupons may behave differently to the cast Al-Si

© BSRIA BG 50/2013
alloys typically used to make heat exchangers. At least two coupons from
each metal of interest should be used to obtain mean corrosion rates.
Coupons can be assembled into bundles and placed in coupon chambers
or attached to coupon racks for insertion in the system.
In making coupon bundles, the coupons are normally electrically insulated

from each other by the use of PTFE spacers, with PTFE sleeves over
the brass bolts used to hold the bundle together. However, coupons of
different metals (e.g. steel and copper) may be bolted together to produce
a galvanic couple if required. It is important that clean gloves are used
when handling the coupons at any time to prevent grease and other
contaminants affecting the results.
In order to obtain accurate corrosion rates, coupons should be left in the

system for at least one month. To obtain trends in corrosion rates, coupons
need to be placed in the system and taken out at different intervals. The
mean corrosion rates for each metal averaged over the exposure period can
then be determined.
On removal of coupons from the system, they should be immediately

rinsed in demineralised water, followed by methanol and then dried
in warm air. Coupons should then be stored in a desiccator or bags
containing silica gel for transport to the laboratory for inspection and
weighing.
Exposed coupons should be assessed for any pitting corrosion by visual

inspection under a low power microscope. The general condition and any
salient features should be photographed for record purposes. Note that it
is not unusual for some crevice corrosion to be observed at the mounting
points but this is not of concern.
Corrosion rates are estimated by carrying out weight loss determinations

on each coupon in turn. Several cleaning stages are carried out on each
coupon using inhibited acids in order to obtain a corrected weight loss, as
specified in ASTM G-1[16].
6.6.2 Corrosion probes

Corrosion probes allow continuous monitoring of corrosion rates. The
most common probes for closed systems are based on the principle of
linear polarisation resistance (LPR). Two or three electrode versions are
available and whilst mild steel electrodes are the most common, other
probe materials can be obtained.
The LPR method is based on the fact that polarising electrodes by less
than 20 mV produces a linear current response. The slope of the graph
of the polarisation voltage plotted against polarisation current is the
polarisation resistance, which is inversely proportional to corrosion rate.
A more accurate determination of corrosion rates can be obtained by the
instrument, using default “Tafel” slopes in the Stern-Geary equation. A
combination of the LPR technique and corrosion coupons will provide
a fairly accurate assessment of the corrosion rates of various metals in the
system and trends in corrosion rates over time.

© BSRIA BG 50/2013
If pitting corrosion is taking place, this will result in an imbalance between

the two electrodes and give rise to a fluctuating potential difference
between them. Some devices (e.g. Corrater®) can exploit this effect to
give an indication of pitting in addition to the general corrosion rate.
LPR meters and probes can be installed on site to give intermittent

readings of corrosion rates. However, they can also be used with data
loggers to store information, which makes on-line monitoring possible.
This enables alarms to be set if corrosion rates exceed pre-set values due to
sudden changes in water quality.
The user of LPR probes needs to be aware of a number of things:

s Low conductivity water samples can cause corrosion rates to be
underestimated, as the solution resistance is added to the polarisation
resistance. However, in typical heating and cooling systems the
conductivity is usually more than high enough for these errors to be
insignificant.
s LPR probes should be aligned perpendicular to the flow so that one
electrode is not shielding the other, which could affect the readings.
s LPR probes located in stagnant areas will give results that are
unrepresentative of system conditions. This point is common to all
methods of measuring corrosion.
Electrical resistance corrosion probes are not recommended for use in

closed systems as they only provide useful information at much higher
rates of corrosion than would normally be encountered.
6.6.3 Biofilm monitoring

There are a number of possible approaches to biofilm monitoring in closed
system but unlike corrosion monitoring, there are no standard or widely
used methods. Qualitative judgements on the presence of biofilm are
often made on the surface condition of corrosion coupon racks.
Quantitative methods rely on inserting a surface into the system for a

period of weeks or months, then removing that surface and recovering the
bacteria for analysis. Alternatively the surface may be assessed directly with
microscopy or a bioluminescence technique.
A method for biofilm monitoring in closed systems using glass beads is

described in Appendix D.
6.6.4 Continuous monitoring of key system parameters

Corrosion in closed heating and cooling systems is governed primarily by
the concentration of dissolved oxygen, the presence of microorganisms and
how these factors are mediated by the use of water treatment, i.e. corrosion
inhibitors and biocides.
Periodic water analysis can measure the concentration of inhibitors and

pick up signs of metal corrosion. LPR probes and corrosion coupons can
give an accurate assessment of corrosion rates. However, to understand
how the whole system is performing and pre-empt damage due to loss of
control, continuous monitoring can be useful.

© BSRIA BG 50/2013
For instance, it is no use adding more chemicals to the system to tackle a

leak. Many inhibitors cannot cope with gross aeration of system waters
due to fresh make-up water, or negative pressures at the top of the system
pulling in air. Water treatment should not be used to address problems that
could and should be remedied by simple engineering changes.
There are a number of robust sensors available that can be used for in-line
monitoring of heating and cooling systems (Table 11). They will need to
be calibrated before use but once installed they should not drift or give
noisy signals.
Table 11: Sensors for in-line monitoring of closed heating and cooling systems
Parameter Comment
Temperature The temperature difference between flow and return

sensors can be used to verify that heating or cooling
water is circulating as required. Sufficient temperature
sensors may already be installed as part of the building
management system
Pressure A pressure sensor should be installed at the top of
the system, at the minimum pressure point, to verify
that positive pressure is maintained under all operating
conditions to prevent air entering the system through
vents etc.
Pressurisation system water meter Persistent make-up may be indicative of a leak in the
system and adds dissolved oxygen. Planned events such
as partial drain down/refill for maintenance reasons
should be followed by water sampling and analysis to
check inhibitor concentration.
Conductivity A sudden change in conductivity indicates that something
has happened to the system and, if the reasons are not
obvious, should trigger further identification to identify
the cause.
LPR probe LPR technology (e.g. Corrater®) can be used for

monitoring general corrosion rates and as a pitting
indicator for steel and other metals (see section 6.6.2).
Galvanic current The galvanic current flowing between steel and copper
gives a good indication of the level of corrosion
protection in system. Without inhibitors present, this
responds quickly to changes in dissolved oxygen.
Dissolved oxygen Oxygen sensors are relatively expensive and therefore

permanently installed sensors are not widely used but
may be justified for large systems. Both optical and
electrolytic sensors are available. In-line sensors need to
have a stainless steel body.
Unfortunately this list of robust in-line sensors does not include pH and
ion selective electrodes as these need frequent calibration and cannot
withstand high temperatures. pH and ion selective electrodes can be
incorporated in automated sampling and analysis systems but these are
generally too expensive for most closed heating and cooling systems.
Continuous monitoring of the available sensors via the building

management system allows alarms to be set and the water treatment
specialist to be automatically informed of loss of control. In addition,
the logged data should be available to help the water treatment specialist
diagnose the cause of the problem.

© BSRIA BG 50/2013
7 REGULATIONS
7 REGULATIONS
7.1 HEALTH AND The UK approach to health and safety is generally implemented through
SAFETY
the Health and Safety etc. Act 1974 (as amended) and the regulations
made under the act. Further information on these regulations and
practical guidance on safe working practice can be found at www.hse.gov.
uk. The following regulations (as amended) are particularly relevant to
water treatment operatives:
Management of Health and Safety at Work Regulations 1999

The Regulations require employers to assess risks to health and safety and
record findings. Having done so they must implement measures to control
risks, appoint competent people, set up emergency procedures and provide
information and training for employees and anyone else who needs to
know.
Employees have to use equipment in accordance with training and have

the duty to report dangerous situations to their employers. A particular
issue for water treatment personnel is that they often work alone and out
of sight of other building occupants and therefore need to be particularly
conscious of the safe working practice when it comes to handling
potentially hazardous materials.
Personal Protective Equipment at Work Regulations 1992

Personal protective equipment (PPE) is defined as all equipment designed
to be worn or held to protect against a hazard at work. PPE should be
relied on as a last resort but PPE should be suitable for both the risk and
the user. The employer has a duty to provide and maintain PPE and ensure
it is properly used. Virtually all PPE should carry the CE mark.
These regulations do not apply to processes covered by other regulations,

which means they do not cover hearing protection or respiratory
protective equipment.
Control of Substances Hazardous To Health Regulations 2002

The aim of the COSHH Regulations is to protect workers’ health from
hazardous substances used at work. The employer has to assess the risks
from such substances, and for certain substances he has to monitor the
employee’s exposure to them. The regulations define what constitutes
suitable personal protective equipment.
Hazardous substances include both chemical and biological agents.
Reporting of Injuries, Diseases and Dangerous Occurrences

Regulations 1995
RIDDOR covers the notification and reporting of injuries and dangerous
occurrences. The specific injuries and incidents which legally must be
reported are specified in the Regulations. However, any injuries sustained
by a water treatment operative, however minor, must be reported both to
their employer and the person responsible for health and safety on the site.

© BSRIA BG 50/2013
REGULATIONS 7
Confined Spaces Regulations 1997

These regulations require employers to avoid entry to confined spaces
by finding alternative methods of work, but if entry to a confined space
is unavoidable, a safe method of work must be in place together with
adequate emergency arrangements which will also safeguard rescuers.
Although people tend to think of confined spaces as tunnels and tanks,

some plant rooms can be classed as confined spaces within the meaning of
the regulations.
Work at Height Regulations 2005

These regulations impose health and safety requirements with respect
to work at height. They cover organising and planning work at height,
competency and supervision, risk avoidance, selection of work equipment
and inspection regimes. Duties are placed on employers, the self-employed
and contractors.
Step ladders are often used when taking samples in site. The safe use of
step ladders for sampling in ceiling voids always requires two persons.
Electricity at Work Regulations 1989

These regulations impose health and safety requirements with respect to
electricity at work. They impose duties on employers, and requirements as
to systems, work activities and protective equipment.
It is a well-known saying that electricity and water do not mix. Work

involving temporary pumps needs to be properly checked and safety tested
by competent persons before use.
The Carriage of Dangerous Goods and Use of Transportable Pressure

Equipment Regulations 2009
The regulations governing what chemicals may be carried in what
quantities, the type of packaging and how the vehicle must be marked
are very complex. Suffice to say that the best source of guidance on the
practical requirements will normally be the manufacturer or supplier of the
chemical. Under no circumstances should water treatment chemicals that
are marked as hazardous be transported in unmarked vehicles.
7.2 OTHER Various regulations apply to water systems in general including the
REGULATIONS installation and operation of closed heating and cooling systems. The
applicable legislation for England is shown below. Different legislation
may apply in other parts of the UK but the practical requirements are
similar.
Water Supply (Water Fittings) Regulations 1999

The Water Regulations impose requirements for backflow prevention that
are applicable when filling or flushing a heating or cooling system.

© BSRIA BG 50/2013
7 REGULATIONS
Water Industry Act 1991

The Water Industry Act provides for discharges of “trade effluent” to the
sewer to require a permit from the sewerage undertaker. Although this
principally affects the permitting of pre-commission cleaning activity,
any drain down of a heating or cooling system at any stage in the life
of the building could require a permit. There is no minimum volume.
The application for the permit will require details of the volume and
constituents of the discharge and may take several weeks to obtain. Each
sewerage undertaker has different requirements for information and it is

advisable to consult them as early as possible if a discharge is planned.

© BSRIA BG 50/2013
USE OF TEMPORARY PUMPS A
APPENDIX A USE OF TEMPORARY PUMPS
All temporary pumps should be:

s fitted with pressure control.
s located in bunds to retain water in the event of a leak.
s connected to the building system with hoses and components
that are tested and rated to at least 1.5 times the maximum pump
pressure.
Figure 16 (from BG 29[1], with commentary) illustrates the main features of

a temporary pump installation. In addition:
s Sampling and dosing connections should be available (not shown in
Figure 16).
s The water inlet and outlet connections to the system should be of
adequate size (ideally line size).
s Hoses should be of sufficiently large diameter to achieve the
required flow rates given the available pressure at the pumps.
s Hoses should be checked for damage and sterilised before use.
s The pump should be connected to a suitably rated electrical supply
via an isolator and appropriate protective devices.
s A safety barrier should be erected around the work area.
s A notice board detailing chemical COSHH information should be
displayed.

© BSRIA BG 50/2013
A USE OF TEMPORARY PUMPS
[1]
Figure 16: Temporary pump arrangement (from BG 29 )
2
IV
1
IV
8
3 4 T
IV STR NRV IV
OP DRV
7
IV
IV
1. Flexible pipe between the temporary pump and system is acceptable for short periods. However, if the
temporary pump is to be left in place for more than 1 month then connections should be hard piped. Flexible
hoses used for flushing purposes should be checked for damage before use and sterilised to avoid the risk of
bacterial contamination.
2. All flexible hose connections should be screwed. Quick release couplings or jubilee clips could present a hazard if
they work loose or are released under pressure.
3. A regulating valve with flow measurement device is required on the return connection from the system to
permit measurement and regulation of the flushing velocity. For this application, a variable orifice regulating valve
offers greater flexibility since it can be used to measure a wider range of flow rates.
4. A strainer on the pump inlet is required to protect the pump against circulating debris.
5. The entire temporary pump set should be mounted in a tank or bund so that drips or spillages do not cause
damage to under-floor services, structure or finishes.
6. A water supply will be required to replenish dirty water drained from the system. The smaller the supply available
the longer it will take to clean the system. To facilitate the process, flushing connections should be provided in the
cold water riser. A double check valve or a break tank with type A air gap will be required to prevent possible
back contamination of the boosted supply in compliance with the Water Regulations.
7. A drain connection will be required to discharge dirty water. A suitable drain point should be installed in the
main foul drainage system at a convenient location for this purpose. An effluent permit will be required and this
may need to include all chemicals proposed for use and those present in the system. Where a trapped waste
connection is provided solely for the purpose of facilitating the flushing process, this should be removed and
capped after completion.
8. A thermometer on the outlet from the pump will provide a check that circulating temperatures are below levels
at which bacteria could multiply.

© BSRIA BG 50/2013
CASE STUDIES B
APPENDIX B CASE STUDIES
Case study 1 – MIC in radiators
Building: Large government office building.
Manifestation: Blocking of radiators with sludge and pinhole leaks at the

bottom of radiators under sludge within a few years of construction.
Figure 17 shows a transverse section through the partially blocked lower

waterway of a radiator with severe pitting corrosion.
Figure 17: Sludge and pitting corrosion in lower waterway of radiator
Cause: The LTHW system contained a large amount of non-barrier

plastic pipework which allowed oxygenation of system water. Iron
oxide corrosion debris had accumulated in the bottom of radiators and
growth of sulfate reducing bacteria (SRB) under sludge led to microbially
influenced corrosion (MIC) and localised pitting.
Solution: The system was too badly affected for any on-line remediation
to be effective and the radiators were all replaced.

© BSRIA BG 50/2013
B CASE STUDIES
Case study 2 – Sporadic pitting in new radiators and towel rails
Building: Residential apartments on a community heating scheme
Manifestation: Pinhole leaks in radiators and towel rails within 6-12

months of completion.
Figure 18: Pinhole corrosion of radiator (outside surface)
Paint blister
Figure 19: Inside surface of radiator
Corrosion site
Dislodged
tubercle
Note: the corrosion tubercle has been dislodged during the cutting of the sample
Figure 20: Inside surface of towel rail
Corrosion
deposit
Note: The corrosion deposit is contaminated with swarf from cutting

the sample. The perforation occurs through the tube under the
corrosion deposit, near to the junction between the tube and header.

© BSRIA BG 50/2013
CASE STUDIES B
Cause: Progressive commissioning and occupancy together with a plastic

pipe distribution system led to persistently high oxygen levels. This rapid
pitting is a form of differential oxygen corrosion that can occur in isolated
locations anywhere on the inner surface but usually adjacent to a seam
or crevice between the opposing faces of the radiator. Note that the
surrounding surface has no corrosion as the pitting area becomes anodic.
This is different to longer term radiator corrosion that is associated with
the build-up of sludge and debris in the lower waterway.
Solution: In this case the developer changed from plastic to steel pipe
to reduce oxygen ingress. Other similar projects use heat interface units
with plate heat exchangers so that any problems with water quality in the
distribution system do not spread into the completed apartments.
Case study 3 – Pinhole leaks in thin wall carbon steel tube
Building: Refurbished office building
Manifestation: Pinhole leaks in several areas of the new chilled water

distribution system became evident within a few weeks of occupation.
Figure 21: Pinhole in thin wall carbon steel tube
Figure 22: Internal corrosion deposits obscuring the leak site

© BSRIA BG 50/2013
B CASE STUDIES
Figure 23: Corrosion and pitting in heat affected zone (cleaned sample)
Cause: Examination of the affected pipe revealed pitting corrosion in

the heat affected zone of the seam weld. There were known problems
in achieving effective pre-commission cleaning and subsequent water
treatment due to high levels of suspended solids in the pre-existing systems
but also a suspicion of copper deposition around the perforation sites.
Unfortunately no detailed surface analysis was undertaken in this case.
Solution: The affected areas were re-cleaned and pipe replaced as necessary.
Case study 4 – Rapid pitting corrosion of thin wall carbon steel tube
Building: New educational building
Manifestation: Pinhole leaks, poor water quality and sludge in the heating
systems soon after pre-commission cleaning.
Figure 24: Severe pitting on thin wall carbon steel tube (cleaned sample)
Cause: Examination of the pipe samples taken from a number of different

areas of the building found some instances of severe pitting under a thick
layer of brown/black corrosion deposit. The causes of the problem included
deficiencies in pre-commission cleaning, ingress of oxygen and poor
management of water quality immediately after commissioning.
Solution: Large parts of the system were recommended for replacement

and the remainder for re-cleaning.

© BSRIA BG 50/2013
CASE STUDIES B
Case study 5 - Widespread corrosion in heating and chilled water

systems
Building: Large business park
Manifestation: Blockage of pipework together with failure of boilers and

condenser coils
Cause: Insufficient expansion vessel capacity and poor maintenance of

system pressures leading to large water losses, loss of inhibitors and gross
aeration.
Figure 25 shows the tube bundle of a large shell and tube condenser.
Several tubes had failed due to corrosion and there was considerable
corrosion debris collected in the bottom of the shell. Figure 26 shows
sections of 20 mm steel pipe with oxygen pitting corrosion.
Figure 25: Heat exchanger shell and tubes
Figure 26: Oxygen corrosion in steel pipe
Solution: Remedial flushing of pipework was possible in some areas but

extensive replacement of pipework and system components was necessary.
Larger expansion vessels were installed to avoid water losses and the system
pressure increased. A proper maintenance regime was introduced to
monitor system conditions and water quality.

© BSRIA BG 50/2013
B CASE STUDIES
Case study 6 - Inappropriate water treatment
Building: College
Manifestation: Blockage of radiators due to large amounts of alumina

debris.
Cause: A strongly alkaline corrosion inhibitor which resulted in corrosion

of the aluminium boilers and circulation and deposition of alumina

corrosion products as shown in Figure 27 and Figure 28.
Figure 27: Oxygen corrosion in steel pipe
Figure 28: Debris viewed under microscope (x8)
Solution: Remedial flushing of the radiators and system pipework. The

system was re-treated with a more appropriate inhibitor and biocide and
carefully monitored.

© BSRIA BG 50/2013
CASE STUDIES B
Case study 7 – SCC of brass fitting under insulation
Building: Office chilled water system
Manifestation: Leakage of brass valves under thermal insulation. Figure 29

shows longitudinal stress corrosion cracks running through a brass fitting
and nut.
Figure 29: SCC in brass nuts

Cause: Poorly fitted phenolic foam insulation allowing condensation on

pipe. High hoop stresses on fittings coupled with exposure to ammonia
leached from the insulation resulted in stress corrosion cracking of fitting
from outside.
Solution: The affected valves were replaced and the insulation improved
to prevent condensation on the pipework and leaching of aggressive
species.

© BSRIA BG 50/2013
B CASE STUDIES
Case study 8 - SCC in a stainless steel water jacket
Building: Food Factory
Manifestation: Leaks in the stainless steel hot water jacket surrounding

a pipe delivering liquid chocolate. The leaks were due to stress corrosion
cracks initiating from pits in stainless steel. Figure 30 shows a crack at
approximately 20 times magnification.
Figure 30: Stress corrosion cracking in stainless steel (x20)
Cause: Incorrect grade of austenitic stainless steel (304 instead of 316)

susceptible to pitting and stress corrosion in high chloride water. Note
that although this is not a conventional building heating and cooling
application the same problem could occur in any closed system with
stainless steel heat exchangers or pipework.
Solution: RO water installed to eliminate chloride from water. The RO

water was then passed through a dolomitic limestone filter to reintroduce
some hardness.

© BSRIA BG 50/2013
CASE STUDIES B
Case study 9 - Erosion corrosion of brass valve seat
Building: Large four story office block
Manifestation: Brass control valves to fan coil units ‘letting by’ and
therefore giving poor temperature control and wasting of energy for
the air conditioning system. Figure 31 shows an example at 8 times
magnification.
Figure 31: Erosion of brass valve seat (x8)
Cause: Erosion corrosion with possible cavitation due to low static

pressures at top of building. Also high concentrations of ammonia in the
system as a result of microbial degradation of the nitrite/nitrate
inhibitor.
Solution: The affected valves were replaced. The system was flushed out
and inhibited with non-nitrite inhibitor to avoid ammonia formation.
The system static pressure was increased to reduce the risk of cavitation at
the top of the building. In these situations the control strategy and pump
settings should also be reviewed with the aim of reducing the differential
pressure across the control valve.

© BSRIA BG 50/2013
B CASE STUDIES
Case study 10 - Poor water quality leading to blockages
Building: LTHW system in large office block.
Manifestation: Loss of flow due to fan coil unit strainers repeatedly

becoming blocked.
Figure 32: Blocked fan coil strainer
Cause: Loss of control of water quality and corrosion leading to high

levels of circulating solids and pseudomonad bacteria.
Solution: The main plant and risers were cleaned using system pumps,
followed by side stream filtration to maintain physical cleanliness of
recirculating water. Each floor was then individually cleaned, off-line,
using a temporary pump (one floor per weekend). Full building and
system operation was maintained through all working hours.

© BSRIA BG 50/2013
CASE STUDIES B
Case study 11 – Copper sulfide scale
Building: Large office building with fan coil heating and cooling
Manifestation: Persistent suspended solids in the form of black flakes

following pre-commission cleaning.
Figure 33: Copper sulfide scale
Cause: Examination of copper tube from the system found a uniform

layer of black scale that was friable when dried. Analysis of the scale using
x-ray diffraction found this to be almost pure copper (I) sulfide (Cu2S).
Some pitting was found underneath the scale layers but this was considered
unlikely to result in perforation. The origin of the sulphide was never fully
explained but believed to be due to microbiological action in the system
prior to pre-commission cleaning.
Figure 34: Copper sulfide scale
Solution: A pilot scale cleaning rig was set up to assess cleaning strategies
for the distribution pipework as the scale was resistant to normal cleaning
chemicals. The system was eventually cleaned using ammonium citrate
followed by hydrogen peroxide to dissolve the scale. The fan coils were
replaced as a precautionary measure as the tube in these is much thinner
(0.3 mm) than distribution pipework.

© BSRIA BG 50/2013
C BIOCIDES
APPENDIX C BIOCIDES
One of the main driving forces behind the current and future availability
of biocides in the water treatment industry is the European Biocidal
Products Regulation (528/2012, commonly known as BPR) which came
into force on the 1st September 2013, repealing the European Biocidal
Products Directive (98/8/EC, commonly known as BPD).
For a summary of the implementation of the BPR in the UK see

www.hse.gov.uk/biocides/
The characteristics of some of the non-oxidising biocides that are

currently still permitted for use in closed heating and cooling systems are
summarised below.
Isothiazolinones
These biocidal products consist of a 3:1 ratio mixture of 5-chloro-2-
methyl-2H-isothiazol-3-one plus 2-methyl-2H-isothiazol-3-one (CMIT/
MIT) (abbreviated name).
O O
N CH3 N CH3
S S
Cl
CMIT MIT
The characteristics of isothiazolinones are:

s Broad-spectrum bactericide. Effective against the majority of
bacteria likely to be found in closed heating and cooling systems.
s Effective over the range pH 6.5 to 8.5.
s The dose rate of 1.5% active isothiazolinone is typically 200 ppm
for maintenance dose rate and 400 ppm for shock dose.
s The contact time for a shock dose is typically 5 to 6 hours.
s Stable up to 50°C

© BSRIA BG 50/2013
BIOCIDES C
DBNPA
2-2-Dibromo-3-nitrilopropionamide. Biocide with an active halogen
group.
O
N
NH2
Br Br
The characteristics of DBNPA are:

s Broad spectrum bactericide.
s Effective over the range pH 6.0 to 8.0. DBNPA hydrolyses rapidly
above pH 8.0.
s Fast acting. The contact time for a shock dose may be less than 1
hour.
s The dose rate of a 20% active solution is typically 100 ppm for a
maintenance dose and 200 ppm for a shock dose.
s Stable up to 50°C.
Bronopol
2-Bromo-2-nitropropane-1,3-diol.
O¯
N+=O
Br
HO OH
The characteristics of bronopol are

s Broad spectrum micro-biocide, slimicide and aerobic/anaerobic
bactericide.
s Effective over the range of pH 6.0 to 8.0 although the breakdown
products are also micro-biocidal under alkaline conditions
s The dose rate for 20% active material is typically 100 ppm for
maintenance dose and 200 ppm for a shock dose.

© BSRIA BG 50/2013
C BIOCIDES
Gluteraldehyde
CH2(CH2CHO)2
O O
The characteristics of glutaraldehyde are:

s Broad spectrum bactericide.
s Good for difficult and persistent organisms including SRB and
biofilms
s Effective over a wide pH range (typically pH 6.5 to 9.0)
s Fast-acting. Usually 3 to 4 hours contact time but perhaps 4 to 6
hours with difficult slimes.
s The dose rate for a 45% active material is typically 100 to 125 mg/l
though a heavily slimed cooling system may need 200 - 300 mg/l as
a shock dose for the initial clean-up.
s Stable up to 60°C
THPS
Tetrakis (hydroxymethyl) phosphonium sulfate
(P(CH2OH)4+)2 SO4
HO OH
OH O HO
P+ P+
O– S O–
HO OH
O
OH HO
The characteristics of THPS are:

s Broad spectrum bactericide.Very effective against SRB.
s Acts by disrupting the structure of bacterial cell membranes leading
to cell lysis.
s Fast-acting. 3 to 4 hours contact time for optimum activity.
s Effective over a very wide pH range (typically pH 3 to 10).
s The dose rate for 25% active material is typically 60 to 80 mg/l.
s Note: Overdosing with THPS has been linked to the formation of
copper sulfide scale in copper tube.

© BSRIA BG 50/2013
BIOCIDES C
Triazines
A number of triazine based biocides can be used within closed heating and
cooling systems, the predominant one being 1,3,5-tris(hydroxy-ethyl)-s-
hexahydrotriazine (HHT).
HO OH
N N
OH
The characteristics of HHT are:

s Broad spectrum bactericide. Effective against SRB and biofilms
s Relatively slow acting with optimal biocidal activity after 24 hours
contact.
s Effective over a wide pH range (optimum pH 8.0 to 11.0).

© BSRIA BG 50/2013
D QUANTITATIVE MEASUREMENT OF BIOFILM
APPENDIX D QUANTITATIVE MEASUREMENT OF BIOFILM
The principle of this method is the insertion of a small canister of glass

beads of known surface area into the system water flow, possibly at the
same location as the corrosion monitoring coupon. Biofilm forms on the
surface of the beads. At the end of the monitoring period the canister is
removed and a number of the glass beads are agitated in a known volume
of water to release the biofilm. A sample of this water is then analysed for
TVC in the normal way.
In the version developed by the General Electric Company for cooling

towers (Figure 35) a 100 mm diameter transparent Perspex® tube is
installed vertically and connected between a flow and return branch. The
flow through the tube is controlled to 10-15 litres per minute. The idea of
the transparent tube is to allow a visible indication of the accumulation of
gross biofilm or debris but it must be protected from sunlight. 20 of the
3 mm diameter beads are removed for each biofilm determination.
Figure 35: Biofilm monitor
Return
Perspex tube with

screwed ends
containing glass
beads
Flow meter
Control
valve
Flow

© BSRIA BG 50/2013
ANTI-FREEZE E
APPENDIX E ANTI-FREEZE
The most common anti-freeze components for building services chilled

water applications are ethylene glycol and propylene glycol. Propylene
glycol is nowadays preferred to ethylene glycol as it is less hazardous and
the impact of discharges to the environment and on sewage treatment
processes is less.
Glycol solutions are more corrosive than pure water and must be used in
combination with corrosion inhibitors. Also, the breakdown products are
acidic so pH should be monitored.
Glycols also provide a carbon source for bacteria growth and can be
broken down by bacteria, though high concentrations (>15% by volume)
are biocidal.
The basic characteristics of glycol solutions are shown in Table 12 and

figures 36 and 37. Note that the freezing points will be influenced
by other water treatment components. Also, unlike pure water, glycol
solutions become slushy and more difficult to pump before they reach
their freezing point.
Glycol concentration can be verified by density measurements or by

optical properties. There are simple low-cost devices available for both
these approaches.
Glycols interfere with gravimetric suspended solids determination so the

presence of glycols in water samples should be notified to the laboratory
prior to their analysis.
Table 12: Characteristics of glycols
Ethylene glycol Propylene glycol
% by mass % by volume Freezing Density % by mass % by volume Freezing Density
point °C @ 20°C point °C @ 20°C
kg/m3 kg/m3
0 0 0 998.2 0 0 0 998.2
5 4.4 -1.4 1005.9 5 4.8 -1.6 1003.3
10 8.9 -3.2 1013.3 10 9.6 -3.3 1008.1
[17]
Source data: ASHRAE Handbook. Fundamentals
15 13.6 -5.4 1021.6 15 14.5 -5.1 1013.4

20 18.1 -7.8 1029.7 20 19.4 -7.1 1019.0
21 19.2 -8.4 1031.1 21 20.4 -7.6 1019.5
22 20.1 -8.9 1032.7 22 21.4 -8 1020.5
23 21 -9.5 1034.2 23 22.4 -8.6 1021.5
24 22 -10.2 1035.8 24 23.4 -9.1 1022.5
25 22.9 -10.7 1037.4 25 24.4 -9.6 1023.6
26 23.9 -11.4 1039.0 26 25.3 -10.2 1024.6
27 24.8 -12 1040.5 27 26.4 -10.8 1025.6
28 25.8 -12.7 1042.1 28 27.4 -11.4 1026.6
29 26.7 -13.3 1043.7 29 28.4 -12 1027.6
30 27.7 -14.1 1045.3 30 29.4 -12.7 1028.4

© BSRIA BG 50/2013
E ANTI-FREEZE
Figure 36: Freezing point of ethylene glycol solutions
-2
-4
Freezing point °C
-6
-8
-10
-12
-14
0 5 10 15 20 25 30
% by volume
Figure 37: Freezing point of propylene glycol solutions
-2
-4
Freezing point °C
-6
-8
-10
-12
-14
0 5 10 15 20 25 30
% by volume

© BSRIA BG 50/2013
GLOSSARY
GLOSSARY
ABS Acrylonitrile Butadiene Styrene – a thermoplastic

Aerobic Requires oxygen to support life
Alkalinity A measure of the concentration of carbonate and bicarbonate ions
in solution.
Anaerobic Does not require oxygen to support life.
Anion Ion possessing a positive charge.

Anode The positive pole in a chemical reaction cell e.g. in a battery. Metals
always corrode at the anodes
Bactericide A biocide effective against bacteria
Bacterium Microscopic usually unicellular organism
Back flushing The process of removing debris from a terminal unit by discharging
system water from a convenient drain point such that the direction
of flow through the terminal unit is opposite to the normal
operating flow
Base exchange softener An ion exchange system that replaces calcium and magnesium ions
with sodium ions to reduce the hardness of the water
Binder point Self-sealing pressure test point from which fluid can be discharged
by insertion of a matching needle probe
Biocide Chemical agent which either kills or prevents the multiplication of
organisms in water systems
Biofilm A mass of bacteria and other microorganisms and particles
embedded in an exopolymer attached to a surface.
Biofouling Fouling of surfaces caused by bacteria and biofilms.
Cathode The negative pole in a chemical reaction cell e.g. in a battery.
Metals tend to be passive at the cathode.
Cation Ion possessing negative charge
Cfu Colony forming unit
Cavitation Formation and sudden collapse of vapour bubbles in a liquid.
Cavitation can occur in pumps causing erosion of the pump
impeller.
Chelating agent Organic compound causing a substance that would normally be
insoluble to be held in solution or to pass into solution
Closed system System where the water is circulated without loss or addition
Conductivity Measure of the ability of a solution to carry an electric current.
Usually expressed in microsiemens per centimetre (μS/cm)
Corrosion Loss of metal from a component due to the electrochemical
reaction with its environment. The term is also used to describe
the chemical degradation of non-metallic materials.
Corrosion coupons Small strips of metal placed in a water system to enable the rate of
corrosion in that system to be assessed.
Corrosion inhibitor Compound that suppresses corrosion. Inhibitors may work by
coating the corrodible surface or interfering with the chemical
reaction that leads to corrosion.
Crevice corrosion Localised corrosion occurring in a crevice or crack in a metal
surface.
Deaerator A means of removing dissolved gases including oxygen from system
water.
Dead leg A length of pipe normally closed at one end
Dezincification Corrosion of brass (an alloy of copper and zinc) where the zinc is
preferentially removed to leave a weak porous matrix of copper
Dip slide Disposable plastic slide coated with microbiological nutrients for
the assessment of bacteria concentrations in water

© BSRIA BG 50/2013
GLOSSARY
Erosion Loss of metal by physical removal though impingement with water

alone or gases and particles suspended within it
Hardness A measure of the calcium and magnesium content of the water
Heterogeneous metal A surface that is not completely homogeneous i.e. there are
surface variations in surface condition
HDPE High Density Polyethylene – a thermoplastic
LTHW Low Temperature Hot Water in the context of heating systems
Method statement Detailed description of the on-site work activity, including
preparatory works and access provisions
MIC Microbially Influenced Corrosion - corrosion that is initiated or

accelerated by microbial activity, particularly the formation of
biofilms
Microorganism Microscopic life forms that can thrive in water system including
algae, fungi and bacteria
NRB Nitrate/Nitrite Reducing Bacteria
pH A measure of the whether the water is alkaline or acidic based on
the logarithm of the hydrogen ion concentration. Water at pH 7
is neutral. Water at greater than pH 7 is alkaline and water at less
than pH7 is acidic. For most metals corrosion rates will increase as
the pH falls below 7.
Pitting Locally severe corrosion forming deep pits in the metal surface.
This can lead to perforation of pipes and radiators.
Practical completion The point of issuing a certificate of practical completion of
construction works for the relevant system by the responsible
person appointed by or on behalf of the client, after which
responsibility for that system passes to the client
Pseudomonads Pseudomonas – like bacteria. Prior to the advent of DNA
classification methods, many bacteria were classified as
Pseudomonas species on the basis of appearance and growth
characteristics. These are now recognised as distinct species but
are still relevant to pipework corrosion.
PTFE Polytetrafluoroethylene – a thermoplastic
PVC Polyvinyl Chloride – a fluoropolymer plastic
Sampling and analysis plan A detailed description of the objectives of sampling, numbers and
locations of samples, the analyses to be carried out for each sample
and guidelines for interpretation of the results
Scale A dense, adherent deposition of mineral particles usually composed
of insoluble salts of calcium, magnesium and iron. Scale is formed
when the water is heated or when the pH or alkalinity of the water
is changed. The risk of scale forming is greatest with hard water and
smallest with soft water.
SCC Stress Corrosion Cracking - a specific form of corrosion that leads
to cracking and mechanical failure of stressed brass and stainless
steel components when exposed to certain chemical species
SRB Sulfate Reducing Bacteria
Suspended solids Insoluble dispersed solids that may be separated from the water by
filtration. The concentration of suspended solids is determined by
passing a measured quantity of water through a pre-weighed filter
and re-weighing the filter after drying.
TDS Total Dissolved Solids - solids that are dissolved in the water but
may be measured by weighing the residue remaining after the
water is evaporated
TVC Total Viable Count - a measure of the number of aerobic bacteria
present in the sample. The result is expressed as cfu/ml.
Tubercle A wart-like mound of corrosion deposit on the surface of a pipe.
Pitting corrosion may be occurring under the tubercle potentially
leading to perforation of the pipe.

© BSRIA BG 50/2013
REFERENCES AND BIBLIOGRAPHY
REFERENCES
1. BSRIA BG 29/2012 Pre-commission cleaning of pipework systems

2. CIBSE Commissioning Code W
3. BS 2486:1997 Recommendations for treatment of water for steam boilers and water heaters
4. BS 7074-2:1989 Application, selection and installation of expansion vessels and ancillary equipment for sealed
water systems. Code of practice for low and medium temperature hot water heating systems
5. BS 7291-1:2010 Thermoplastics pipe and fitting systems for hot and cold water for domestic purposes and
heating installations in buildings. General requirements
6. BS 8552:2012 Sampling and monitoring of water from building services closed systems. Code of practice
7. BS 8558:2011 Guide to the design, installation, testing and maintenance of services supplying water for
domestic use within buildings and their curtilages. Complementary guidance to BS EN 806
8. BS EN 12502-1:2004 Protection of metallic materials against corrosion. Guidance on the assessment of
corrosion likelihood in water distribution and storage systems. General
9. BS EN 12502-2:2004 Protection of metallic materials against corrosion. Guidance on the assessment of
corrosion likelihood in water distribution and storage systems. Influencing factors for copper and copper alloys
10. BS EN 14868:2005 Protection of metallic materials against corrosion. Guidance on the assessment of corrosion
likelihood in closed water circulation systems
11. BS EN ISO 8044:2000 Corrosion of metals and alloys. Basic terms and definitions
12. BS EN ISO 6222:1999, BS 6068-4.5:1999 Water quality. Enumeration of culturable micro-organisms.
Colony count by inoculation in a nutrient agar culture medium
13. BS EN ISO 5667-3:2012 Water quality. Sampling. Preservation and handling of water samples
14. BS EN ISO 13720:2010 Meat and meat products. Enumeration of presumptive Pseudomonas spp.
15. BS EN ISO 19458:2006 Water quality. Sampling for microbiological analysis
16. ASTM G1 - 03(2011) Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens
17. ASHRAE Handbook. Fundamentals (SI) edition. 2009
18. Corrosion and Corrosion Control: An Introduction to Corrosion Science and Engineering, Fourth Edition
(2008). R Winston Revie, Herbert H Uhlig

© BSRIA BG 50/2013
REFERENCES AND BIBLIOGRAPHY
BIBLIOGRAPHY
CIBSE Guide B Heating, ventilation, air conditioning and refrigeration.
CIBSE Guide M Maintenance engineering and management.
Fundamentals of Corrosion: Mechanisms, Causes and Preventative Methods (2010). Philip A Schweitzer.
Fundamentals of Hydraulic Engineering Systems, Fourth edition (2009). Robert J. Houghtalen, Ned H.C.
Hwang, A.Osman Akan.
Industrial Biofouling: Detection, Prevention and Control (2000). James Walker, Susanne Surman, Jana Jass.
The corrosion of copper and its alloys: A practical guide for Engineers (2009). Roger Francis
Microbiologically influenced corrosion (2007). Brenda J. Little, Jason S. Lee.
Heat and mass transfer in building services design (1998). Keith Moss
NACE SP0169-2007 Control of external corrosion on underground or submerged metallic piping systems.

© BSRIA BG 50/2013
7HG>6ⳮ i]ZWjâi
Zck^gdcbZciZmeZgih
6GF=5[]jYgmciWcbZ]XYbWY]bXYg][b UXXYXjUìY]b
aUbiZUWhifY WcadYh]h]jYUXjUbhU[Y]baUf_Yh]b[
dfcZ]hUV`YWcbghfiWh]cb UbXYZZ]W]YbhVi]`X]b[g
IZhi^c\ IgdjWaZh]ddi^c\
BdYZaa^c\ >c[dgbVi^dc
GZhZVgX] IgV^c^c\
8dchjaiVcXn EjWa^XVi^dch
K\UhYjYfmcifVi]`X]b[ >chigjbZci]^gZ! BVg`ZigZhZVgX]VcY

gYfj]WYgfYei]fYaYbh hVaZhVcYXVa^WgVi^dc ^ciZaa^\ZcXZ
WcbhUWh6GF=5.
I/ ))%&()))+*+%%
;/ ))%&()))+*+'+ AYaVYfg\]d]gh\YZcibXUh]cbcZ6GF=5»g
:/ Whg^V5Whg^V#Xd#j`
L/ lll#Whg^V#Xd#j` YldYfh]gYUbX]bXYdYbXYbWY
DaY7gVX`cZaaAVcZLZhi!
7gVX`cZaa!7Zg`h]^gZ!
G<&',6=!J@
D[[^XZh^c7gVX`cZaa!7Z^_^c\!9jc[Zgba^cZ!
@jVaVAjbejg!DiiVlV!Hi=ZaZch!Hijii\Vgi
VcYIdjadjhZ#6hhdX^ViZh^c6gbV\]

BG 50 - 2013 Water Treatment For Closed Heating and Cooling Systems

Uploaded by

Copyright:

Available Formats

BG 50 - 2013 Water Treatment For Closed Heating and Cooling Systems

Uploaded by

Document Information

Original Title

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

BG 50 - 2013 Water Treatment For Closed Heating and Cooling Systems

Uploaded by

Copyright:

Available Formats

Licensed copy from CIS: southbank, Southbank University, 06/03/2018, Uncontrolled Copy.

A single copy of this document is licensed to

This is an uncontrolled copy. Ensure use of the

The technical information in this publication was provided by members of

David Bleicher BSRIA

BSRIA acknowledges with thanks the organisations that provided additional

Midland Corrosion Services Ltd.

WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

The treatment of water in modern closed heating and cooling systems

treatment can sometimes be disastrous.

This guide is intended for use by design engineers, installing contractors

The common causes of water quality and corrosion problems are

WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

2 SYSTEM DESIGN AND OPERATION 3

2.3 Minimising dissolved oxygen 6

4 CHEMICAL WATER TREATMENT 35

5 NON-CHEMICAL WATER TREATMENT 50

6 MANAGEMENT OF THE WATER TREATMENT

WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

APPENDIX: A USE OF TEMPORARY PUMPS 81

REFERENCES AND BIBLIOGRAPHY 103

WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

cooling networks operating at temperatures below 110°C.

The objectives of a water treatment programme in closed heating and

This is aided by the following elements:

WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS 1

c. Control of biofouling/bioﬁlms by:

i. Avoidance of low ﬂow

Section 7 brieﬂy summarises the main legislation that is relevant to water

2 WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

2 SYSTEM DESIGN AND OPERATION

Issues that should be considered by the system designer include:

The beneﬁts of good design can be wasted if installation, commissioning

WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS 3

Although non-metallic components such as seals, diaphragms and linings

All materials, be they metallic or non-metallic, should have industry

current water treatment chemistry. Table 1 summarises the corrosion

Table 1: Corrosion properties of frequently used materials in closed water systems

4 WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

2.2 FLOW RATES 2.2.1 Flow velocity

In oxygenated waters, the uniform corrosion rate will generally increase

If the velocity is too high then there is a risk of turbulent erosion.

The minimum recommended period of circulation for heating and cooling

Water sampling should not be carried out in no-ﬂow conditions, as this

WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS 5

The most important way of controlling corrosion in use is to ensure that

Systems should be designed with adequate numbers of air vents at high

BS EN 14868[10] describes the different ways that oxygen can enter a

In traditional closed heating systems constructed using mild steel pipe,

In closed cooling systems, oxygen concentrations may remain high after

The risk of oxygen corrosion can be minimised by:

6 WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS

The key recommendations to designers are:

s Ensure that plastic pipe that is prone to oxygen diffusion (such as

2.3.1 Oxygen barriers in plastic pipe

2.3.2 Removal of air and dissolved gases

2.3.3 Pressurisation of systems

Maintenance of adequate pressure will also reduce the risk of cavitation. If

WATER TREATMENT FOR CLOSED HEATING AND COOLING SYSTEMS 7