Asian Institute of Technology

Bangkok, Thailand –
Urban Env. Engineering &
Managament Program

Use of Reed Beds for Faecal

Sludge Dewatering

Udo Heinss* and Thammarat Koottatep**

A Synopsis of Reviewed Literature

and
Interim Results of Pilot Investigations with Septage Treatment
in Bangkok, Thailand

1998

* EAWAG, Swiss Federal Institute for Environmental

Science & Technology
SANDEC, Dept. for Water and Sanitation in Developing Countries

** AIT, Asian Institute of Technology

School of Environment, resources and Development
Environmental Engineering Program
 Acronyms and Abbreviations

AIT Asian Institute of Technology, Bangkok

BMA Bangkok Metropolitan Administration
EAWAG Swiss Federal Institute for Environmental Science &
Technology, Zurich
EEP Environmental Engineering Program (AIT)
SANDEC Dept. of Water & Sanitation in Developing Countries
(EAWAG)

AG-WSP Attached-Growth Waste Stabilisation Ponds

CW Constructed Wetlands
FS Faecal Sludge
HRT Hydraulic retention time
SLR Solids loading rate
STP Sewage Treatment Plants

BOD Biochemical Oxygen Demand

COD Chemical Oxygen Demand
SS Suspended Solids
TKN Total Kjeldahl Nitrogen
TS Total Solids
 Contents

Preface i

1 Rationale and Introduction 1

Part A
2 Findings from the Literature.
2.1 Comparative Performance of Planted vs. Unplanted Beds 2
2.2 Hygienic Quality of the Percolating Water 2
2.3 Need for Bottom Ventilation 3
2.4 Reeds 3
2.5 Close-Up View 3

3 Conclusions 5

4 Recommended Design for a Pilot Reed Bed 6

Part B

5 Interim Results of Pilot Cattail Bed (Constructed Wetlands)

Operations at AIT, Bangkok

5.1 Introduction 7
5.2 Operations 7
5.3 Evaluation 10

Table 1 Summary of Reed Bed Performance

Data for Sludge Dewatering/Drying 14

Table 2 Reviewed Publications 15

Preface i

Preface

Most urban dwellers in developing countries use on-site excreta disposal systems,
such as public and family latrines, aqua privies and septic tanks. Contrary to
wastewater collection and treatment, improvements in faecal sludge management
have received little attention until very recently. To date, in the majority of cases,
faecal sludges collected from on-site disposal systems are still disposed of untreated,
mainly for lack of affordable treatment options. Therefore, SANDEC has embarked
on practice-oriented R+D of faecal sludge (FS) treatment options, with the objective
to develop guidelines for the design and operation of sustainable treatment
technologies based mainly on practical field research.

So far, SANDEC mainly examined the use of pond systems for FS treatment through
collaborative field research carried out with Ghana's Water Research Institute.
Results of these investigations and recommendations for practitioners have been
published in "Solids Separation and Pond Systems for the Treatment of Faecal
Sludges in the Tropics" by Heinss, Larmie and Strauss (SANDEC report no.05/98).
Another focus project was on the use of septage for land reclamation of Lahar
(volcano ash) stricken areas. In a new collaborative field research project with AIT
(Asian Institute of Technology), reed beds will be tested for their ability to treat
septage.
Rationale and Introduction 1

1. Rationale and Introduction

Sludge drying beds may be classified as a solids-liquid separation process,

producing a dewatered or dried solids product and a liquid (percolate) requiring
further treatment prior to discharge or use. Mostly unplanted drying beds have
been used to date to handle sludges from sewage treatment plants (STP).
However, in Europe and in the United States, planted sludge drying beds are
finding increasing use. Reeds normally are the plants-of-choice for such beds that
are therefore also designated “reed beds”. They require substantially less operating
care than unplanted beds as they need to be emptied every few years, only.
SANDEC hypothesizes that reed beds might prove a feasible option for treating
faecal sludge.

Part A contains the literature review. Experience gained with the use of reed beds for
dewatering/drying sewage treatment plant or excess activated sludge in mostly
temperate climate is summarised. The purpose of the review was to make the
information available to persons dealing with faecal or STP sludge management.
Further to this, SANDEC needed hints as to the possible design of planted drying
beds to carry out pilot or demonstration schemes for FS dewatering/drying. The
findings from the literature review are discussed in Chpt. 2 below. Table A1 in the
Annex is a synopsis of the reported design and operating parameters. Fig. 1 shows
the schematic design of a ventilated reed bed. Table A2 of the Annex provides the
list of reviewed publications.

In Part B, interim results of AIT/SANDEC collaborative field research conducted in

Bangkok, Thailand, are presented. Three pilot reed beds (“constructed wetlands”)
treating septage have been in operation and subjected to intensive monitoring since
early 1997.
Findings from the Literature 2

Part A
2. Findings from the Literature
2.1 Comparative Performance of Planted vs. Unplanted Beds

Three experiments compare the performance of planted and unplanted drying

beds (4,7,8). The results obtained show the advantages of reed beds as over the
conventional sludge drying beds.

Drying

- The average reported solids content in planted beds amounts to 40%

compared to 32% in unplanted beds.

- According to the authors, the lower sludge layer in a planted bed is

aerobic and brown in colour whereas the sludge in unplanted beds is
anaerobic and black.

BOD and COD

- BOD and COD concentrations in the percolating water of reed beds are
35-55% and 50-60%, respectively, lower than in unplanted beds.

Nitrification

- The ammonium in the percolating water of reed beds is 70-80% lower

than in the liquid of conventional beds. Nitrate levels in the effluent of reed
beds are, in turn, 3-10 times higher than in other beds. It can therefore be
concluded that nitrification in reed beds is greater than in conventional
unplanted beds.

Ventilated vs. Non-Ventilated Beds

- The difference in quality between the percolating waters of planted and

unplanted beds becomes smaller when the beds are equipped with a
ventilation system inducing an air current along the bottom of the bed (7)
(see Fig.1). As a result, oxidising reactions also tend to occur in unplanted
beds.

2.2 Hygienic Quality of the Percolating Water

- Hygienic quality improvement was also investigated (8,9). Concentrations

of enterobacteria (mainly E. coli) were reduced during percolation through
the reed beds by 2-3 orders of magnitude. According to one author, an
unrestricted agricultural use of the sludge from reed beds is possible after
14 months of storage and drying of the sludge (8).
Findings from the Literature 3

2.3 Need for Bottom Ventilation

- Loading rates for anaerobically stabilised sludges should be kept small in

reed beds without bottom ventilation (4). Planted reeds died in non-
ventilated beds when loaded with activated sludge from an oxidation ditch
(7). However, in a ventilated bed the reeds survived and the average
drying rate was significantly higher than in a non-ventilated bed. Reeds
grew very well when aerobic stabilised sludge was applied in a non-
ventilated bed (8). However, if the beds were loaded with a fresh excess
sludge or digested sludge, the stock of reed plants decreased.

2.4 Reeds

- All reed beds were planted with Phragmites. In beds where also other
Helophytes were planted, Phragmites was the only plant that developed a
permanent plant stock and supplanted all other plants.

- Newly planted reeds reportedly formed a dense standing crop several

months later (9).

- As regards cutting, different opinions prevail: from no cutting to cutting

every year or after several years.

2.5 Close-Up View

- It is assumed that reeds influence the sludge dewatering and

mineralisation process in different ways. Bacteria may thrive on the
extensive root system as attached biofilm. The hollow roots possibly allow
transport and supply of O2 to attached aerobic bacteria. Since the root
system is continuously growing, the sludge/soil layer remains permeable
even through the increasing sludge layer.

- Compared to unplanted drying beds, higher TS contents but lower

percolating water were obtained in planted beds (4,8). This can be
attributed to the high suction pressure exerted by the root system and the
leaves which evapotranspirate high quantities of water (15-25 mm/d (1,
2)).

3. Conclusions
Use of reed beds for faecal sludge dewatering in tropical countries may constitute a
promising alternative to conventional sludge dewatering beds. Compared to
conventional drying bed, the thereby resulting advantages are the following.

• Simplicity of operation:
Findings from the Literature 4

Compared to unplanted beds, far less effort is required for sludge removal in
reed beds as planted beds may be loaded over several years before
desludging becomes necessary.

• Nitrification:

Faecal sludge treatment in facultative ponds is often difficult as high ammonia

concentrations may prevent the growth of algae. Nitrification in reed beds
may therefore improve treatability of the drained effluent in facultative ponds.

• Limited clogging

Clogging of the filter layer is likely to be minimal due to the continuous growth
of rhizomes (rootstock).

Similar to sludge drying beds, the main disadvantage of reed beds is their large
land requirement. This will limit the use of this treatment option to areas where
sufficient land is available or where the treatment strategy consists in opting for
decentralized systems.
Beside the operational advantages, the application of constructed wetlands, as the
reed beds are also called, leads to beneficial effects such as wetland conservation
and restoration of wildlife habitat.
Pilot Reed Bed Design 5

4. Recommended Design for a Pilot Reed Bed

Based on the data in Table 1 for temperate climates, experimental or pilot reed
beds treating faecal sludges in tropical zones could possibly be designed as
follows:
2
- Size: 20-40 m , multiple units; 1 - 1.5 m freeboard
- Bottom ventilation to allow passive aeration
- Interval of faecal sludge application: weekly
- Loading rate: 100 kg TS/m2y, initially; provision may be made to allow
a stepwise increase of the loading rate to 150 kg TS/m2.yr or above if
the reed growth and the drying process are not negatively influenced.

Since reeds do not appear to grow well under highly anaerobic conditions, a
ventilation system allowing for natural bottom aeration is, therefore, necessary,
particularly when drying highly anaerobic sludges (Fig. 1). Passive aeration is
achieved with the use of hollow blocks at the filter bottom and a vent pipe of a large
diameter ( 20 cm). The pipe should extend at least 1 m above the reeds at their
highest stand; i.e. prior to bed emptying.
It should be noted that uncertainties still exist regarding long-term performance and
optimum operational patterns. Extended monitoring of septage-loaded constructed
wetlands shall answer essential questions: Will percolation rate in the soil filter
remain constant over several years of sludge loading? How will the plants react to
high organic loading and a changing water regime induced by alternating septage
loading and drying?

Wind action

Vent pipe Reeds

Freeboard
Faecal sludge

Compost*
Sand
Filter bottom

Air
Drainage *for initial planting

Fig. 1: Functional sketch of a reed bed with natural bottom ventilation

Results of Pilot Cattail Bed Operations 6

Part B
5. Interim Results of Treating Septage in Pilot Cattail Beds
(Constructed Wetlands) at AIT, Bangkok
5.1 Introduction
Experience to date in using constructed wetlands or planted soil filters for sludge
dewatering and stabilisation is limited almost exclusively to treating excess
activated sludge. The majority of these sludges are either aerobically stabilised or
anaerobically digested. Performance data shown in Table 1 above were
determined in systems situated in temperate climates. Therefore, the three pilot
sludge drying beds installed at the AIT campus and treating faecal sludge
(septage1) hauled from Bangkok city represent a new approach in using wetlands.

5.2 Operations
The three pilot reed beds have got a size of 5 x 5 m each. Fig. 2 shows
schematically the structure of the beds installed at AIT. The support and filtering
media installed in the concrete embankment lined units are composed as follows
(from bottom to top):
- Hollow blocks underdrain each 20x40x20 cm
- 40 cm of large gravel ∅ 25-50 mm
- 15 cm of small gravel ∅ 10-25 mm
- 10 cm fine sand ∅ eff. 0.3-0.75 mm
Ventilation pipes (∅ 20 cm) which are connected to the underdrain system secure root
zone aeration. Fig. 3 shows schematically how the reed beds are operated,
including feeding arrangements and post treatment in attached-growth waste
stabilisation ponds.
The wetlands have been loaded with septage for 9 months. 1,650 m3 of septage
with an average TS content of 1-2 %, TCOD 14,000 mg/l, and TKN of 1,240 mg/l)
have been treated and dewatered to a TS level of around 45 % during this period.
Fig. 4 shows the water balance in the constructed wetlands as calculated from
cumulative septage loading, water stored in the dewatered septage and percolate
flow measurements.
The pilot reed beds were operated at the following regime:
- Septage loading at once-per-week and twice-per-week intervals
- Septage loading rates:
Equivalent to 80 – 160 kg TS/m2year, initially
More recently set at 250 kg TS/m2year
1
Faecal sludge is a general expression for sludges of variable consistency collected from so called
on-site sanitation systems; viz. latrines, non-sewered public toilets, septic tanks, and aqua privies.
Septage as one kind of faecal sludge, consists of the contents of septic tanks and usually
comprises settled and floating solids as well as the interstitial liquid.
Results of Pilot Cattail Bed Operations 7

- TS concentrations of more than 50 % were obtained in the dried septage

layer. It was however observed that at this TS concentration cattails
wilting occurred presumably because of insufficient soil moisture.
Sustainable plant growth is ensured at TS concentrations of 25-35%. This
corresponds to a septage loading rate of 250 kg TS/m2year a. Percolate
ponding for 2-6 days is being tested to support conditions for healthy
cattail growth.

20-cm stainless
steel ventilation 7-cm thick
pipe ferro-cement

Outlet
pipe

0.25-cm
plastic sheet 1.2-cm mesh
20-cm concrete
blocks
10-cm reinforced
concrete slab
20-cm PVC
2-m wood pile drainage pipe
3-m3 effluent
receiving tank

Fig. 2: Pilot Reed Bed Installed at AIT (schematic)

Evapotrans-
piration 22
%

Dried
Sludge
Percolate 1%
77%

Fig. 3: Water Balance in the Pilot Constructed Wetlands Used for the
treatment of Septage in Bangkok, Thailand (Initial phase)
Results of Pilot Cattail Bed Operations 8

BMA
To constructed
wetland units
Mixer-equipped
feeding tank
Screening
tank

Storage tank

CW-1

From screening ERT*-1

& storage units

To AGWSP units
& polishing ponds
CW-2
ERT*-2

CW-3

ERT*-3

From constructed
wetland units
Percolate
feeding tank

AGWSP-1
AGWSP-2

Percolate
feeding tank
Polishing pond
HRF**
To canal
AGWSP-3
AGWSP-4
* ERT = Effluent receiving tank
** HRF = Horizontal-flow rock filter

Fig. 4 Functional sketch for unit operations of the

constructed wetlands and AGWSP systems
Results of Pilot Cattail Bed Operations 9

5.3 Evaluation

Results generated during the first nine months of operations (1997-98), tend to
indicate that the process is feasible in principle, for treating septage. However,
long-term operation and monitoring over 2-4 years are required to determine
design and operating criteria, which will allow for a stable process. Moreover,
faecal sludges other than septage would have to be tested too, to determine the
feasibility of the process for varying types of FS. It is expected that sludges, which
exhibit a fair degree of biochemical stabilisation, such as pond sludges, may lend
themselves well for reed bed dewatering. Treatment of fresh, public toilet type
sludges may constitute a greater challenge, though.

The important question is whether it will be possible to operate the beds over years
without removing the sludge and without clogging of the filter layer.

Encouraging results with respect to removal efficiencies in the percolating liquid

and with respect to sludge dewatering have been obtained to date. Periodic reed
wilting has occurred at the same time. Further investigations are required to
understand the reasons for it and to develop sustainable measures to achieve
stable plant growth. One measure, which has been tried to date, and which
appears to improve plant growth is the effluent ponding to a level just below the
stored sludge layer. At the current regime, the effluent is then allowed to flow off
every 2-6 days, the main criterion being to avoid anaerobicity in the root zone.

Table 1 shows the average values of percolate quality achieved to date. Post-
treatment of the percolate will be required for most discharge situations in spite of
the fact that high removal efficiencies are attained in the reed beds.

Table 1 Reed Bed Performance With Respect to Percolate Quality

Raw septage Percolate

mg/l mg/l % Removal

SS mg/l 10, - 20,000 700 > 90

CODtot mg/l 14,000 1,500 95

CODfilt. mg/l 470 200 57
TKN mg/l 1,200 150 90
 Table 2 Reviewed Publications

Reviewed Publications
Ref. Title Reference

1 Klärschlammvererdung mit Hilfe von Helophyten Reinhofer M./Berghold H. (1994).

(in German) (Sludge Humification Using Helophytes) Korrespondenz Abwasser, 8/94, pp.1302-1305.
2 Klärschlammvererdung in bepflanzten Mori K. (1996).
Schlammtrockenbeeten (in German) MORITEC Internal document.
(Sludge Humification in planted drying beds)

3 Sludge Dewatering and Mineralisation in Reed Bed Systems Nielsen S. M. (1990). In Proceedings - Constructed
Wetlands in Water Pollution, Pergamon Press.

4 Use of Phragmites in Sewage Sludge Treatment Hofmann K. (1990). In Proceedings - Constructed

Wetlands in Water Pollution, Pergamon Press.

5 Use of Reed Beds for Dewatering Sludge in the USA Kim B.J./ Cardenas R. (1990). In Proceedings - Constructed
Wetlands in Water Pollution, Pergamon Press.

6 Klärschlammvererdung auf zwei ostfriesischen Inseln Ingenieurgesellschaft für Umweltplanung (INFU) (1995).
(in German) (Sludge humification on two East Frisian Internal document.
Islands)

7 A Study of Activated Sludge Dewatering in Experimental Liénard A. / Duchène Ph./ Gorini D. (1994). In: Proceedings,
Reed-Planted or Unplanted Sludge Drying Beds 4th IAWQ Int. Specialist Conference on Wetland Systems for
Water Pollution Control, Guangzhou, P.R. China, 6-10 Nov..

8 Der Einsatz von Pflanzen zur Klärschlamm-Entwässerung Zacher B. / Strauch D. (1987).

(in German) (Using plants for sludge dewatering) Korrespondenz Abwasser 9/87, pp. 922-930.

9 Klärschlammvererdung in Schilfbecken Ergebn. und Scholl W. /Wurster H. /Thalmann A. (1985).

Erkenntn. eines praxisbez. Pilotprojektes (in German) Korrespondenz Abwasser 5/85, pp. 386-395.
(Sludge humification using reed beds: results of a pilot
project)

10
 Summary of Reed Bed Performance Data
Table 1 Summary of Reed Bed Performance Data for Sludge Dewatering/Drying

Ref. no. Size of No. of Pop’n Comparison Bottom Kind of TS in Sludge Applic. TS loading TS
(see one bed drying Equiv. w. unplanted ventilation ? treated Loaded vol. Applied interval Rate Attained
Table 2) (m^2) beds bed ? sludge Sludge, g/l m3/m2·year Kg TS/m2,yr %

1 80 2 10,000 no no aerobically 2.8 4.4 weekly 120 33.5

stabilised 4-7cm
2 - - - no no not specified - 5.2 - 75 45

3 100 4 - no yes aerob stab. 0.7 1.9 - - 20

anaerob dig. 3.5 - -
4 60 4 - yes no aerob. stab. - - fort- 20-30 40
nightly SS
5 37-279 - no no anaerobically 1 to 10 0.41-0.82 - 13-74
(survey 1 to 10 digested 40-50
results) - no no aerob. stab. 0.5 to 5 0.73-3.38 - 29-106

6 - 50,000 no no ? -8 - - - 40-50

7 20 3 exp. yes yes aerob. stab. 0.3 13-15 daily 55 15

beds

8 46 4 yes no aerob stab. 1.5 1.5 weekly 25 -

9 66 2 no yes excess 8 0.45 weekly 100 40

activated sludge

Remarks: 1 Water balance evaporation/drainage 50:50; average

evapotranspiration: 11mm/d
2 40 % org. N elimination in the drainage; average
Ref. No evapotranspiration: 12-20 mm/day
5 Survey of 28 beds in the TS - Total Solids
US
6 Average evapotranspiration: 5 mm/d SS - Suspended solids

11