dewatering drinking water sludge’s in...

14th European Biosolids and Organic Resources Conference and Exhibition 1

Organised by Aqua Enviro Technology Transfer

www.aquaenviro.co.uk

www.european-biosolids.com

DEWATERING DRINKING WATER SLUDGE’S IN REED BED SYSTEMS

Nielsen, S1. and Cooper D.J.,2

1Orbicon A/S, Denmark, 2ARM Ltd, UK

Author Emails [email protected]; [email protected]

Abstract

Sludge drying reed beds have been used for dewatering and mineralisation of sewage sludge’s since

the late 1980’s, however there has been no experience of treating drinking water sludge’s in reed

bed systems.

Water Treatment Works (WTW) treat water to a potable standard, part of this treatment includes

dosing with Iron Sulphate to help dirt particles to coagulate, producing a ferric sludge. Historically

the sludge wastes have been pumped to the sludge lagoons at Hanningfield Reservoir. However,

these are now nearing the end of their serviceable life and a new sludge handling process is

required.

As part of a trial with Essex and Suffolk Water, ARM Ltd and Orbicon A/S set up a series of six trial

beds, each 20m2 at Hanningfield Reservoir in Essex to examine the dewatering processes of the

sludge produced from the water treatment process. These were monitored from 2008 – 2009. The

purpose of the test is to clarify whether the sludge is suitable for treatment in a sludge reed bed

system, the dimensions criteria and the quality of reject water from the sludge reed bed system.

The operation of the test system has been very positive especially because of the very intensive

loading program for some of the basins on such a young system. It is possible to drain and treat

ferric sludge (approximately 300,000 mg Fe/kg dry solid) in a reed bed system.

The system has a good draining efficiency and the filtrate water from the sludge loading is out of the

system the next day. The sludge residue surface is drying as indicated by evidence of desiccation

cracks shortly after the water drains out of the system. In spite of the different loading programs,

volume reduction is very high at over 99%. The ferric sludge is dewatered to approximately 30-40 %

dry solid and desiccates in the trial beds. It is possible to get the vegetation to grow in ferric sludge,

where the pH was measured to 7, 7. It has not been necessary to use fertilizer.

The proposed use of reed bed systems not only reduces the capital and operating cost, but also

provides the site with an environmental friendly operation area. The overall reduction of the sludge

volume occurs without the use of chemicals. The process involves only a very low level of energy

consumption for pumping the sludge and reject water.

Key words

Sludge treatment, reed bed System, drinking water sludge, load, dewatering, ferric sludge

Introduction

It has been identified by the asset management strategy for Essex and Suffolk Water that there is a

need to provide a new process for sludge handling at the Hanningfield Water Treatment Works.

mailto:[email protected]

mailto:[email protected]





Hanningfield WTW (Figure 1) is supplied with raw water from Hanningfield Reservoir. The existing

treatment works comprises pre-ozonation, clarification, filtration, main ozonation, pesticide

removal, chlorine disinfection prior to the pumping of treated water. Sludge waste is generated

primarily from the de-sludging of the Pulsator clarifiers producing a mineral sludge with seasonally

fluctuating levels of algae and suspended solids.

Figure 1 Hanningfield Water Works and the water works sludge lagoon

The final part of the solids removal process is the rapid gravity filters and the granular activated

carbon contactors used during backwashing. Historically these sludge wastes have been pumped to

the sludge lagoons at Hanningfield Reservoir. However, these are now nearing the end of their

serviceable life and a new sludge handling process is required. The two options being considered are

a mechanical centrifugal solution and sludge treatment reed beds. The proposed use of reed bed

systems not only reduces the capital and operating cost, but also provides the site with an

environmentally friendly operation area.

The purpose of the trial is to clarify:

Whether the sludge from Hanningfield Water Treatment Works is suitable for further

treatment in a sludge reed bed system.

The dimensions required (capacity, operations, loads, area, number of basins, etc.) for a full

scale plant at Hanningfield Water Treatment Works.

The quality of reject water from a sludge reed bed system treating sludge from Hanningfield

Water Treatment Works.





Historically, sludge reed beds (Figure 2) have been used for the dewatering and mineralization of

sewage sludge in Europe since 1986. The European Union Water Framework Directive calls for

cleaner discharges from our waste water treatment facilities. Improved treatments to achieve higher

quality effluents results in the production of more sludges. In countries like Denmark, France,

Germany and Sweden, sludge treatment in Reed Bed Systems are a common and a well-proven

method. Long-term sludge reduction takes place in reed-planted basins, partly due to dewatering

(draining, evapotranspiration) and partly due to mineralisation of the organic solids in the sludge.

Sludge from wastewater treatment plants is pumped onto the basin surface. The dewatering phase

results in the dry solids content of the sludge remaining on the basin surface as sludge residue,

whereas the majority of its water content continues to flow vertically through the sludge residue.

The water content is further reduced through evapotranspiration.

Figure 2 Sludge reed bed system

A sludge treatment reed bed system utilizes the forces of nature to reduce and treat sludge. The

only appreciable power consumption is by the pumps used to transport sludge and reject water. This

means that the reed bed system uses much less power than other systems. Transport costs will be

reduced substantially, while the volume of sludge can be reduced to approximately 1.5-2.5 % of its

original volume.

The sizing and design of reed bed systems depends on the sludge production (TDS per annum),

sludge type, quality and regional climate. The treatment period is approximately 8 - 12 years and the

operation of the system may be divided into a number of phases related to different periods in the

lifetime of a system. Each phase consists of commissioning, full operation, emptying and re-

establishment of the system. When the system is setup, it requires only a weekly control-visit to the

site of about one to two hours, and there is no contact with the sludge, which gives a better working

environment.

The plan is to empty the Sludge Reed Bed System over a period, with 2-3 basins selected for

emptying per year. Capacity during the emptying period is maintained despite the reduction in

number of the basins the emptying phase. After a period of years, the basins are emptied, and its

contents can be recycled and spread on farm land.

After a basin has been emptied you can start the load for another operations period. Maintaining full

capacity during emptying is possible provided that the basins are re-established after emptying with





sufficient regeneration of vegetation, and provided that the loading rate is adapted to vegetation

growth. So re-planting basins has not been necessary.

Sludge treatment in reed bed systems is a thoroughly tested method with a number of proven

advantages. The overall reduction of the sludge volume occurs without the use of chemicals. The

process involves only a very low level of energy consumption for pumping the sludge and reject

water resulting in a minimum of CO2-emissions.

The control system (SCADA) can be setup at the main working site and the daily surveillance can be

done from there. There is no noise from the system as there is from many other types of treatment

systems. The system works effectively to reduce pathogenic bacteria like Salmonella, Enterococci

and E. Coli, thus making it a lot safer to be on site.

Sludge treatment in Reed Bed Systems uses no chemicals in the dewatering process. This means a

considerable improvement in the working environment along with a reduction of the chemical

residue in the treated waste water passing into the environment. The content of substances in

sludge that are foreign to the environment can be reduced. After the treatment of sludge, recycling

options are good, particularly in agriculture. The sludge quality is cleaner and more adaptable in the

natural cycle than mechanically dewatered sludge.

Methods and results

The Hanningfield test system (Figure 3) was built at Hanningfield Water Treatment Works with 6

basins each of 20m2 with a design comparable to a full-scale system with reeds, ventilation, sludge

input; reject water systems as well as filters and drains. The reeds were planted in February 2008.

The development during 2008 was good. The reeds grew well and began to cover the whole surface

and achieved a height of approximately 1 meter. It hasn’t been necessary to replant as the reeds

have thrived in the test system. In 2009 the reed growth continued and the surface of the basins was

approximately 100 % covered with reeds. In July the reeds had achieved a height over 2 meters

(Figure 4). It has not been necessary to add fertilizer.





Figure 3 Basin no. 1 – 29.07.2009 (Hanningfield Sludge Treatment System)

Sludge quality

The data (Figure 4) shows that there is a large variation in the total solids being loaded into the

basins. Therefore it is very important to know the % of dry solid in each load in order to calculate the

load (kg ds/m2/year) as precisely as possible. The contents of solids are lower in 2009 and it has been

necessary to adjust the loading program to achieve the demanded area loads. The sludge quality

seems to have good sedimentation ability (Figure 5) and settles well within a few minutes.

0

0,1

0,2

0,3

0,4

0,5

0,6

01.03.08 01.06.08 01.09.08 02.12.08 04.03.09 04.06.09

%

Total SolidsAverage2008: 0,1862009: 0,163

Figure 4 Total solid (%) – Sludge

2008-2009

average





Figure 5 Sedimentation of the sludge. A: Sludge sample (28.07.2009). B: Sedimentation period.

Load

The basins have been loaded for approximately 7 and 4 months in 2008 and 2009, respectively.

During the test period the load has been intensified. Each load representing a volume of 3m3, has

been loaded over a period of approximately 1 hour. The accumulated volume has resulted in a total

loaded volume of 69 - 180 m3 and 45 – 150 m3 in 2008 and 2009 (as of 21.07.2009), respectively

(Figure 6).

The dry solid content in the sludge has varied during the test. The average values were 0.18 % and

0.16 % in 2008 and 2009, respectively (Figure 5). The accumulated load has resulted in a total load of

110 - 300 kg and 70 - 240 kg in 2008 and 2009 (21.07.2009), respectively. During the test-loading

period in 2008, each of the basins have been loaded for 1-3 days (loading days) with 1-4 batches

daily, followed by a rest period of approx. 10-30 days (resting days). In 2009 the basins have and will

be loaded for 2-5 days with 2-4 batches daily, followed by a rest period of approx. 30-55 days.

0

50

100

150

200

250

300

01.03.08 01.06.08 01.09.08 02.12.08 04.03.09 04.06.09 04.09.09

m3

Loading basin 1

Figure 6 Load of basin no 1 (m3)

A

B A B





In 2008 the area load had been applied for a whole year the area load would have varied between

10 to 26 Kg ds/m2/year. In 2009 the area load was varied between an area load 25 – 50 kg

ds/m2/year for a whole year.

-0,005

0

0,005

0,01

0,015

0,02

0,025

0

0,5

1

1,5

2

2,5

3

3,5

4

4,5

06-07-09 00:00 06-07-09 12:00 07-07-09 00:00

l/se

c/m

2

m3

/hBasin no 1

Inlet

Outlet

Figure 7 Four loads and the resulting dewatering profile.

Dewatering (l/sec/m2)

The majority of the water in a sludge load continues to flow vertically through the sludge residue

and filter. The dewatering (liters/second/m2) is good (Figure 7 and 8). The dewatering efficiency has

a positive development in 2009.

The basins are now more mature with a good crop of reeds. The decrease in the dewatering profile

has not been so pronounced and the variation in the dewatering efficiency from the basins has been

reduced. The load has increased in 2009 for all the basins. The maximum loads are 2-4 loads per day

per basins (approx. 6-12 m3/day).

0

0,005

0,01

0,015

0,02

0,025

0,03

0,035

0,04

0,045

0,05

3000 3100 3200 3300 3400 3500 3600 3700 3800

Reject (l/s/m2)BASIN 6

Figure 8 Reject water from the basin no. 6 (Liters/second/m2)





Generally the dewatering profile is a peak with a maximum over 0.015 l/sec/m2. In some basins the

maximum dewatering speed is over 0.020 – 0.025 l/sec/m2 (Figure 8).

Dewatering (m3)

The times for dewatering of 6-12 m3 are approximately 15 hours and over 90 % of the load is

dewatered in that period (Figure 9). Other positive results are that the dewatering profiles generally

are very narrow, short, steep (up and down) and reach the bottom line with a short period of time,

even for the basins which had been loaded with 12 m3 each day for 4 days.

0

20

40

60

80

100

120

140

1 2 3 4 5 6

%

Basin no.

Reject - % of load Average system

Figure 9 Volume of reject water - percentage of load (basin no 1-6)

Dewatering – Volume reduction

The dewatering phase results in the dry solids content of the sludge remaining on the basin surface

as sludge residue. The water content is further reduced through evapotranspiration. The sludge

residue shows evidence of desiccation very quickly (Figure 10). In spite of the different loading

programs, volume reduction is very high over 99 %.

In 2009 the reduction is slightly higher than in 2008 with an increase in the loads (Table 1), and with

more mature basins. The dry solid (0.16 -0.20) in the sludge has been concentrated approximately

200 times to a dry solid estimate in the sludge residue of approximately 28 - 42 % (Table 1).

Table 1 Reduction in sludge volume – 01.03.2008-31.01.2009

Bassin

no.

Sludgevol.

m3

Sludge dry solids (%) Reduction

%

Sludge

residue

vol.

m3

Sludge residue dry

solids (%)

Min. Max. Avg. Min. Max. Avg.

1 180 0.16 0.2 0.186 99.33 1.2 23.9 29.9 27.8

2 159 0.16 0.2 0.186 99.55 0.7 35.6 44.4 41.3

3 120 0.16 0.2 0.186 99.48 0.6 30.8 38.5 35.8

4 81 0.16 0.2 0.186 99.56 0.4 36.4 45.5 42.3

5 105 0.16 0.2 0.186 99.43 0.6 28.1 35.1 32.6

6 87 0.16 0.2 0.186 99.43 0.5 28.1 35.1 32.6





Figure 10 Dewatering efficiency and change in the sludge surface in 14 days (September/October

2008)

Reject Water - Quality

The dewatering phase results in the dry solids content of the sludge remaining on the basin surface

as sludge residue. Samples of Water Works sludge and samples of reject water treated in the reed

bed system show the efficiency of the filtration through the reed beds (Figure 11).

Samples of sludge and reject water taken in the hours after the sludge load. Sludge treatment in the

reed bed system showing the high efficiency of the filtration. The quality of the reject water during

the hours after the loading has a very low content of suspended solid after one load. More than one

load on the same basin on the same day results in turbidity variations in the reject water (Figure 11).

Figure 11 Sludge samples (A) and samples of reject water (1, 2, 3, 4,) during a period with first 3 of 4

loads





0

50

100

150

200

250

300

350

400

19-05-09 03-06-09 18-06-09 03-07-09 18-07-09

NTU

Turbidity - Total

Figure 12 Reject water – Turbidity (NTU)

In May 2009 a turbidity meter was installed on the outlet pipe. The data recording started in May

(Figure 12) before it was calibrated in May and early June. In the initial period before the meter was

calibrated it shows incorrect values.

Generally since June the turbidity is lower than 5 NTU (Figure 13). During the loading (eg. 4 loads in

one day) there is an increase in the turbidity, up to approximately 20 NTU. A few results reach 50

NTU. The trend shows that the main volume of water has a turbidity level below 5 NTU even in the

loading periods (Figure 12).

Sludge residue

Sludge residue levels are measured approximately every 14 days. There has been a reduction in

depth from 12 to 6 cm over a period of approximately 15 weeks which indicates the effect

dewatering in the resting period (Figure 13).

The results have shown that the system has good dewatering properties resulting in good dry solid

%. The amount of nitrogen and phosphorus in the sludge residue appears at present, to be sufficient

to support reed growth (Table 2).





-2

0

2

4

6

8

10

12

14

01-06-2008 01-10-2008 01-02-2009 01-06-2009 01-10-2009

cm

Basin 1Average depth gauge 1-3

Scale pole

Figure 13 Basin no 1 - Sludge residue height

Table 2 Sludge residue quality in the lagoon and in the basins (29.07.2009). The numbers in the

bracket are days since last load

Conclusions

In the first phases (March 2008 – January 2009) of the test the results have shown that:

1. The operation of the test system has been very positive especially because of the very

intensive loading program for some of the basins on such a young system.

2. The reeds are approximately 2.0 meters high and they are beginning to cover the whole

surface.

Parameters Unit Basin 1 Basin 2 Basin 3 Basin 4 Basin 6 Lagoon Site 1 Lagoon Site 2

Dry solid % 39 (21) 40 (18) 44 (28) 42 (4) 31 (6) 28 28

Loss on

ignition

% of DS 24 25 26 25 28 22 14

Total

nitrogen

mg/kg DS 4,800 7,200 7,900 8,600 7,600 5,300 3,900

Total

phosphor

mg/kg DS 11,000 10,000 9,700 9,600 8,100 9,100 58,000

Aluminium mg/kg DS 1,700 1,400 1,400 1,300 2,200 4,100 2,700

Calcium mg/kg DS 47,000 47,000 46,000 43,000 50,000 160,000 200,000

Iron mg/kg DS 370,000 390,000 390,000 370,000 370,000 230,000 130,000

FeS2 (pyrite) mg/kg DS 610 830 920 180 140 1,900 250

Grease + Oil mg/kg DS <60 81 140 110 160 190 110





3. The filters have a good draining efficiency and the water from the loads is out of the system

the next day. The sludge residue is dewatered to approximately 30-40 % dry solid.

4. Shortly after the water is out of the system the sludge residue surface is cracking up very

well.

5. The dewatering is good.

6. Samples of the reject water have shown that the filter has a good filtration capability.

7. A full scale system will need 16 basins and the loading can be 30 kg dry solid/m2/year.

References

Nielsen, S. (2003) Sludge Drying Reed beds. Water Sci. and Technology.48, No 5 pp. 101 – 109.

Nielsen, S. (2003) Paper: 16 years of experience with sludge treatment in Reed beds system

Conference: Achievements and Prospects of Phythoremediation in Europe. COST Action 837. 15 - 18

October 2003 Vienna.

Nielsen, S.(2005) Sludge Reed beds facilities: Operation and problems. Water Sci. and Technology, 51

No 9 pp. 99-107.

Christensen, L.B. (1991) Acidification and Ochre formation in Fenlands. BSSS and IPS Symposium,

Cambridge, England April 1991.

Acknowledgements

The authors do wish to thank the different partners involved in this study: Essex & Suffolk Water,

Hanningfield Treatment Works, and Northumbrian Water Limited.

dewatering drinking water sludge’s in...

Documents