Layman’s Report

0 INTRODUCTION The use of Waste Stabilization Ponds (WSPs) to treat wastewater has increased greatly since the end of the seventies and is frequently chosen by small communities. There are now some 2,500 to 3,000 WSP systems in France for an average of 600 p.e. representing 20 % of sewage treatment plants (Racault et al., 2005). In France, WSPs are usually designed by 3 ponds (facultative + 2 maturation ponds) for a total area of 11 m2.p.e.-1. For small communities, in the case of fragile receiving bodies, the effluent quality is not always sufficient according to the limits set by some regulations (i.e. French Quality Level D4: <125 mg/L COD; < 25 mg/L BOD5). In effect, the classical WSP configuration doesn’t permit to achieve these limits in France (see Table 1, Racault et al., 1995). As a result, a complementary treatment is required. Several technologies have been proposed for upgrading pond quality effluent (Middlebrooks, 1995; Sezerino et al, 2003; Johnson and Mara, 2005). It would be desirable that these post-treatments also ensure that the global treatment system will maintain the primordial advantages of the WSPs (easy operation and maintenance, environmental integration).

COD SS KN NH4+ -N PT

Mean 198 69 25 17 10 Separated networks SD 109 5 17 15 8 Mean 141 54 19 12 7,5 Combined sewers SD 69 41 11 9 6

Table 1: Mean outlet water concentrations measured on different French WSPs (mg.l-1).

Moreover, many WSPs are fed by combined sewers or partially separated sewers which can induce great hydraulic overloads. They usually are by-passed at the inlet of the plant direct to the receiving bodies without any treatment. Additionally, it is also to consider that the small amount of sludge produced have to be managed. Ponds produce about 2 cm/year of wet anaerobic sludge (110L/Person Equivalent (PE)/year) (Racault and Boutin, 2005); which must be removed periodically to keep the pond performing properly. Taking into account the loading conditions of WSPs in France, it seems necessary to remove the sludge approximately every 10 years. Desludging represents a major factor of maintenance operational costs in WSPs (Cavalcanti et al., 2002; Racault and Boutin, 2005). In order to reduce these costs the association of a Sedimentation Tank (placed upstream the pond) and Sludge Dewatering Reed Beds can work as a solution. In this context, a new treatment line using an association of WSP and ISFs plus a sludge line was purposely built in Aurignac as a full scale demonstration facility. with the following objectives :

• To Associate Pond and ISF to respect the French D4 quality level and complete nitrification.

SS due to the presence of algae, is the main parameter responsible of the non-respect of the D4 water quality level (WEF, 1992). The implementation of Intermittent Sand Filters (ISFs) as polishing treatment after WSP can offer a solution to the previous concern due to the potentialities of these technologies for retaining algae (Neder et al., 2002; Kayser et al. 2002), removing organic matter and nitrifying the ISF influent. ISFs system are nevertheless sensitive to clogging phenomena and thus have to be designed and manage

correctly to ensure biomass regulation. A first step done on pilot experiment (Boutin et al., 2002) allowed to go on this way by experimenting the process in full scale.

• To treat storm-waters drained by combined sewers

The aim is to use the hydraulic capacity of a pond system to store temporally hydraulic overloads in the pond using a supplementary freeboard of 50 cm allowing the storage of a part of the storm waters drained by the combined sewerage network.

• To allow an easier sludge treatment and management.

This point is studied by the construction of a prototype of sludge drying reed beds to treat the sludge pumped daily in a primary settle tank.

The final aim of this study is to propose a wastewater treatment train for rural districts (with combined sewerage network) easy to operate.

1 PRESENTATION DU SITE EXPERIMENTAL The WWTP is designed for 300 p.e. in a first step with the aim in the future to extend the plant capacity to 1200 p.e.. The sewerage system is of combined type.

Figure 1 : Layout of the Aurignac WWTP

The plant has been in operation since July 2003, presented on Figure 1, consists of :

• Settling Tank Wastewaters pass trough a first step of automatic screening. Then, they go to settling tank (ST). Primary sludge is extracted each day from the bottom of the ST (1.5 m3 of useful volume) placed upstream the pond and is pumped towards four Sludge Dewatering Reed Beds

• Pond system The facultative pond, with a surface of 2100 m2, has a supplementary freeboard of 50 cm allowing the storage of a part of the storm waters drained by the combined sewerage network. Therefore, the water heights can fluctuate between 80 and 130 cm (there is a by-pass at 130cm).

• Infiltration Sand Filters 6 independent ISFs of a unit surface of 50 m2 are filled with 3 different media (river sand, river sand planted with Phragmites australis, and crushed sand).

Figure 2 : Lits d’infiltration percolation-

The granulometry of the sands has been chosen according to the up-to-date recommendations in France (Liénard et al., 2001). Filters are composed, from the top to the bottom, of the sand (see Table 2) followed by a 5 cm transition layer of 6/15 mm gravel and then a 40 cm drain layer of 30/60 mm gravel.

Characteristics River sand Crushed sand d10 mm 0.25 0.19 UC 4.7 9.3

% fines % 2.1 4 Porosity % 0.42 0.44

Table 2: particle size characteristics of the sand.

Moreover, for each type of media two filter depths are used: 25 and 65cm. This in the aim to determine the optimal depth of sand that have to be used. Table 3 precise the appellations used in this document for each type of filter’s configuration. Type of filter appellation River sand, 25 cm R25 River sand, 65 cm R65 Crushed sand, 25 cm C25 Crushed sand, 65 cm C65 River sand with Macrophytes, 25 cm M25 River sand with Macrophytes, 65 cm M65

Table 3: appellations of ISFs

Filters are fed by pumps (55m3.h-1) by distribution systems under pressure. Batch volume id managed by an ultrasonic probes into the feeding tank. The volume of each batch is regulated to produce a water level of batch between 2.5 and 5 cm in relation to the experimental needs. ISFs are alternately fed: the two filters of the same media are fed at the same time during 3.5 days and then rest for 1 week.

• Sludge treatment. Primary sludge is settled and extracted each day from the bottom of the ST (1.5 m3 of useful volume) placed upstream the pond and is pumped towards four Sludge Dewatering Reed Beds (SDRBs). This system allow:

- A dewatering step of the sludge by mechanical filtration. Percolates return to the pond.

- A mineralization step of the sludge with years leading to a reduce sludge production (65 % of reduction is expected).

A ten years sludge storage period is expected. Each bed has a surface of 17 m2; designed for a load of 40 kg of solids m-2.y-1 and

planted with Phragmites australis. The filters are alternately fed; with a feeding/rest

period of 7/21 days. After percolating through the SDRBs the treated lixiviate is recycled to the settler. They are composed, from the top to the bottom, of 30 cm of 2/6 cm of gravel followed by a 10 cm transition layer of 6/15 mm gravel and then a 15 cm drain layer of 30/60 mm gravel.

Figure 3: SDRBs

2 SAMPLING AND METHODS The instrumentation implemented permitted to follow up continuously the hydraulic behavior of the whole WWTP (flows, pond water level, infiltration rates of filters) as well as the climatic conditions (temperature, humidity, solar radiation, rainfall) and the temperatures inside the pond and filters. All these data, which could be locally or remotely accessed, were recorded continuously on an electronic central station of data acquisition minute by minute. Physicochemical and microbiological parameters were evaluated in each component of the plant: (a) WWTP inlet, (b) ST outlet (c) SDRBs inlet and outlet, (d) WSP outlet and (e) ISFs outlet. Analysis of 24-hours composite-samples (40 campaigns) and grab samples (collected weekly) were performed from September 2003 to October 2005. COD, SS, KN, NH4-N, NO3-N, PO4-P and TP were analyzed according to French standard methods (AFNOR, 2005). The bacteriological indicators (fecal coliforms (FC) and E. coli) were evaluated according to the protocol for membrane filter procedures of Standard methods (APHA, 1998). The quantitative and qualitative algal biomass evolution has been also studied in the pond by optical microscopy and Chlorophyll-a analysis (AFNOR method NT 90-117). Tracer experiments with NaCl have also been done on the filters to study the evolution of their hydrodynamic.

3 RESULTS

3.1 Wastewater treatment

3.1.1 Loads received by the plant.

Designed for 300 p.e. (45 m3.d-1 ; 36 kg COD.d-1), the station received in dry periods 40% of its nominal organic load and 66 % of its nominal hydraulic load. Inlet pollutant concentrations measured during dry weather flow, slightly diluted cf. Table 4), are characteristic of water resulting from combined sewerage networks (Racault et al., 1995). The relatively frequent pluviometry in the area of Aurignac (approximately 50% of time) induced important hydraulic overloads. The average hydraulic load received by the plant during the study is 53 m3.d-1 (70 m3.d-1 by mean during rainy periods). The load can varies considerably in relation to the magnitude of the storm event. The maximum hydraulic load measured was of 500 m3.d-

1 (11 times the dry weather flow). Therefore, pollutant concentrations during rainy periods are

lower due to the dilution effect except on solids because of accumulation of solid into the network during dry periods that are released during rainy periods.

COD BOD5 SS KN TP mean (mg.l-1) 515 188 183 65 8.9 Dry periods SD 76 26 56 15 1.9 mean (mg.l-1) 364 173 38.2 6.3 Rainy periods SD 143 154 12 2.4

Table 4: characteristics of the WW inlet.

Organic loads evolved during the study from 30/40 % of the nominal load to 40/50% at the end thanks to the progressive connections of users to the sewer network.

3.1.2 Settling tank

As it was expected, the sedimentation tank upstream the pond removed at least 50 % of the SS and about 35 % of the COD of the raw wastewater. On the contrary, the removal of dissolved contaminants (10 %) and bacterial indicators was not significant. High flows measured during storm event do not appear to penalize too much the efficiency of the settling tank. Sand and gravel drained during storm event are easily removed by the settling tank despite of bad upward flow-rate in that case.

3.1.3 The pond

Two points have been studied in relation to the pond performance. One focused on pollutant removal efficiency and the other one focused on the hydraulic overload acceptance.

3.1.3.1 Hydraulic flow acceptance Pond water level fluctuations between Sept. 2004 and July 2005 are presented in Figure 4.

Pond water level

700

800

900

1000

1100

1200

1300

1400

a-04 s-04 o-04 n-04 d-04 j-05 f-05 m-05 a-05 m-05 j-05 j-05

date

leve

l (m

m)

Pond overflow detection Measure simulated pond water level

Figure 4: Pond water level fluctuation from September 2004 to July 2005.

On this period storm water represents 42 % of the inlet flow. Water volume by-passed by the pond maximum level represents 14 % of the inlet flow. This indicates that a sizeable part of the inlet water has not been treated by the plant. Nevertheless it has to be mentioned that pond overflow water that has been discharged into the receiving body corresponds to period with

non optimized filter feeding (low hydraulic load, non loaded …) due technical or experimental problems. Taken a stable hydraulic load of 75 cm.d-1 on the filter, only 0.4 % of the inlet flow would have been discharged by the pond overflow and 1.4 % taken into account a nominal dry-weather flow (45 m3.d-1) at the inlet of the plant. The freeboard onto the pond appears to be a relevant solution to treat storm waters of combined systems.

3.1.3.2 Pollutant removal efficiency The average concentrations and removal rates (settling tank + pond) are presented in Table 5. The quality of the pond effluent is representative of WSPs in France (Racault et al, 1995). The mean concentrations reached by the lagoon didn’t respect the French quality level D4 (<125 mg/L COD, < 25 mg/L BOD5) for discharge in sensitive receiving bodies.

Dry periods Rainy periods Average outlet

concentration Removal efficiency

Concentration moyenne

Rendement moyen

COD (mg.l-1) 148 89 % 144 53 % dCOD (mg.l-1) 93 65 % 98 23 % BOD5 (mg.l-1)L) 58 69 % n.d. n.d. SS (mg.l-1) 48 76 % 45 74 % KN (mg.l-1) 21 70 % 20 45 % NH4

-N (mg.l-1) 13 72 % 9.6 56 % TP (mg.l-1) 3.6 62 % 3.6 41 % FC (Ulog) 2.7e4 2.5 Ulog 7.9e4 2.0 Ulog E.coli (Ulog) 2.8e4 2.7 Ulog 5.1e4 2.1 Ulog

n.d. : not determined

Table 5: Average concentration of pollutants in the WSPs effluent and removal efficiency

The quality of the effluent was not affected by the rainy episodes due to the buffer capacity of the pond; except for the bacterial indicators. It is to note that the water level in the lagoon (of 0.8 m to 1.3 m) could change thus allowing the system to withstand with these rainy episodes. On the contrary, fluctuations in the effluent quality according to the season were observed (Figure 5).

050

100150200250

26/09/200304/01/200413/04/200422/07/2004 30/10/200407/02/200518/05/200526/08/2005

mg/

L

COD mg/L

SS mg/LdCOD mg/L

D4 level

Figure 5 : Evolution of COD, dCOD and SS content in the WSPs final effluent

The photosynthetic activity, higher in summer, induces elevated algal production compared to winter (Figure 6). As a consequence, COD and SS evolve in the same way (Figure 5). The high concentrations of dCOD are probably resulting from the excretion of carbohydrates by the algae; as was observed for the carbohydrates analyses performed during the study. On the other hand, a higher reduction of nitrogen forms and bacterial indicators in warm periods was observed (Figure 7). In summer, ammonia concentrations remained very often under 10 mg/L (Figure 4). The data suggest that ammonia removal also has a seasonal behavior: the improved ammonia elimination in summer was linked with the higher water

temperatures and chlorophyll-a concentrations. The removal of FC ad E. coli was also higher in summer (> 3 Ulog) when the concentrations of chlorophyll-a were important. High ambient temperature, solar radiation and pH due to the growth of algae have been reported to encourage pathogen inactivation and die-off (Davies-Colley et al., 1999; Alcalde et al. 2003).

0300600900

12001500

06/10/2003 14/01/2004 23/04/2004 01/08/2004 09/11/2004 17/02/2005 28/05/2005 05/09/2005

µg/L Chlorophyll a

/LScenedesmus

Chlorella EuglenaEuglena

Figure 6: Evolution of Chlorophyll a in the pond and the predominant algae genera

05

10152025

26/09/2003 04/01/2004 13/04/2004 22/07/2004 30/10/2004 07/02/2005 18/05/2005 26/08/2005

mg/

L

NH4

Figure 7 : Evolution of N-NH4 content in the WSPs final effluent

Filters appear to be necessary to reach the D4 quality level not only for filtration of solids (and so reduction of COD), but for dissolved COD as well. Moreover, to assure a complete nitrification filters also appear to be an essential treatment step (Figure 7). Their utility would be more important for an inlet nominal organic load.

3.1.4 Filters

3.1.4.1 Feeding operation Different organic and hydraulic loads were applied all through the study, and different volume of batches as well. Four periods can be pointed out : - Hydraulic load of 20 cm.d-1 on the filters in operation and batches of 5 cm. - Hydraulic load of 40 cm.d-1 on the filters in operation and batches of 5 cm. - Hydraulic load of 80 cm.d-1 on the filters in operation and batches of 5 cm. - Hydraulic load of 80 cm.d-1 on the filters in operation and batches of 2.5 cm. Periods of low load were necessary to overcome a clogging problem that appeared in autumn 2003 for all the filters. This clogging was due to simultaneous effect :

- Presence of Scenedesmus (Figure 6) : large algae with high clogging effect. - Unstable flow at the outlet of the pond (unadapted valve) - Mistakes done on the rotation of the feeding and resting periods of the filters.

As a consequences several corrections have been done (permanent or temporal) to resolves the clogging and start again on good base:

- Rake the algal deposit on the unplanted filters. - Practice a perfect rotation of the filter - Lower the hydraulic load until they recover good infiltration rates (June 2004)

- Ensure a stable pond outlet flow.

7.1.1.1 Pollutant removal All filters achieved the quality fixed by the French regulations for discharge in sensitive areas with effluent concentrations lower than 125 mg/L for COD and 25 mg/L for BOD5 (see Table 6). However, it must be noted that the effluent quality from 25 cm depth and crushed sand filters presents COD and BOD5 concentrations close to the required discharge limits. Filters of 65 cm depth applied with security the D4 quality level with outlet level always below 90 mg.l-1 of COD for example. In relation to this goal, filters appears to be a good solution to treat pond effluent. Moreover, ISFs also nitrifies the pond effluent, with KN concentrations <7 mg/L and NH4-N concentrations ≤3mg/L for almost all the filters.

M65 R65 C65 M25 R25 C25 mg/L % mg/L % mg/L % mg/L % mg/L % mg/L %

COD 57.7 (15) 62 59.3

(18) 57 76.4 (19) 49 79.0

(16) 44 79.8 (18) 42 96.9

(17) 35

dCOD 45.0 (19) 52 49.1

(21) 47 67.1 (18) 27 53.8

(16) 42 55.3 (17) 35 69.2

(15) 23

BOD56.2

(2.5) 89 7.9 (3.5) 86 17.0

(4.1) 70 13.5 (4.2) 76 13.6

(5.7) 76 18.1 (6.2) 68

SS 9.8 (6.3) 78 11.5

(5.9) 75 19.7 (11) 69 17.1

(8.6) 63 17.8 (8.2) 69 26.3

(10) 52

KN 4.9 (4.1) 78 4.3

(3.9) 79 6.7 (3.9) 70 6.9

(5.4) 69 6.5 (4.1) 70 8.7

(5.5) 63

NH4-N 1.7 (3.7) 92 1.6

(2.6) 92 4.0 (3.1) 73 3.0

(4.3) 82 2.7 (4.2) 83 4.3

(4.5) 71

NO3-N 10.4 (6.3) * 14.3

(11) * 11.5 (6.4) * 11.1

(4.8) * 14.3 (8.8) * 11.3

(7.8) *

TP 1.9 (1.2) 52 3.0

(1.2) 35 3.3 (0.5) 10 2.8

(1.1) 27 3.0 (0.8) 9 3.4

(0.3) 2

Table 6: ISFs efficiency (physicochemical parameters): average outlet pollutant concentration (standard deviation) and % removal

The TP ISFs outlet concentrations ranged from 1,9 to 3,4 mg/L. Retention of phosphorus is comparable with the one from systems with fixed-biomass on fine media in the absence of specific material for phosphorus removal. In one year the removals diminished drastically for all filters (i.e. from 80% to 20 % for planted-filters). During the whole study the average FC concentrations in the outlet of all filters was higher than 1000 CFU/100 mL. Microbiological decontamination, influenced by the hydraulic retention time, is relatively poor (around 1.5 U.log for the 65 cm filters). Infiltration rates are too fast to comply all the time reuse exigencies. However, it has been observed that the effluent quality of the 65 cm beds complies with the WHO guidelines for agricultural reuse and can be used for unrestricted irrigation in warm periods. The increase of the hydraulic load induce a reduction of removal efficiency as it can be seen on Figure 8 and Figure 9. Nevertheless, until 80 cm.d-1, outlet concentrations comply with the D4 quality level.

01020

30405060708090

100

C25 R25 M25 C65 R65 M65

Rem

oval

rate

(%)

COD SS NK

0

102030405060708090

100

C25 R25 M25 C65 R65 M65

Rem

oval

rate

(%)

COD SS KN

Figure 8 : Removal rate for hydraulic loads of 40 cm.d-1 Figure 9 : Removal rate for hydraulic loads of 80 cm.d-1

7.1.1.2 Type of media.

Filters with crushed sand performed worse than the filters with river sand for organic matter (COD, dCOD, BOD5) and SS removal. The use of crushed sand is not to discard but requests a higher security margin when dimensioning (depth of the sand). The presence of reeds had notable effects in COD, dCOD and SS removal for HLs <50 cm.d-1. However, planted and non-planted river sand filters presented similar reductions for higher HLs.

7.1.1.3 Operation strategy. The good performances achieved in this WWTP could be obtained by keeping strictly the alternate feeding of the filters. One of the main problems when working with algae influents is the risk of clogging the filters. The respect of the periods of feeding and rest is of major importance for the hydraulic reliability of the filters. The alternation of periods of rest and feeding regulates the growth of biomass and the formation of surface deposits, thus minimizing the risks of clogging. At the time, and respecting the operation mode of 3-4 days feeding and 7 days of rest for each filter (3 filters in parallel), no problems of clogging appeared with hydraulic and organic loads of 80 cm/day and 180 g/COD/m2 respectively. On the other hand, continuous feeding (> 8 days spreading) or too short periods of rest (3-4 days), leads to a weak mineralization of the surface deposits composed of algae. These conditions can diminish infiltration rates depending on the algae population (i.e. Scenedesmus) thus inducing the clogging of filters, particularly in winter. If surface clogging appears, the removal of this surface layer is necessary to restore the infiltration capacity. However, for planted filters the manual elimination of this layer is not possible. In that case specific hydraulic loads have to be applied until the clogging disappears by itself. When working with planted ISFs at the beginning of winter every year the faded aerial part of the reeds was cut and removed from the beds. Non-planted ISFs didn’t need this maintenance activity. On the contrary, the maintenance operations during the study were more frequent and complicated due to the continuous growing of weeds in the beds. Generally, the use of planted-ISFs is recommended.

7.2 Sludge drying reed beds prototype

7.2.1 Load applied

Different mode of feeding have been tested during the study. Three periods can be pointed out:

- One feeding of 3 mn per day (263 days) - Two feeding of 3 mn per day (122 days) - Three feeding of 2 mn per day (175 days)

About 1 m3.d-1 of sludge was applied onto the beds for the two last modes of feeding.

The sludge SS average content was 5.3 g/L with high variations (from 4 to 22g/L). A tendency can be observed between the concentration of the inlet wastewater and the sludge extracted from the settling tank (see Figure 10).

y = 28,002xR2 = 0,4643

0

5000

10000

15000

20000

25000

0 100 200 300 400 500 600

Concentration des eaux usées brutes (mg/L)

Con

cent

ratio

n de

s bo

ues

prim

aire

s (m

g/l)

Figure 10 : SS concentration of the inlet wastewater and the extracted sludge.

The average load received by the beds is 28 kg DM.m-2.y-1, which represent 70 % of the designed load. Nevertheless, mode of feeding influence the extracted sludge concentration. The most concentrated sludge was observed for the feeding mode of 3 feeding of 2 mn per day with a organic load of 35 kg DM.m-2.y-1 (87 % of the designed load).

7.2.2 Removal rates

The percolate quality and the removal rates of the SDRBs are presented in Table 7. The removal of particulate contaminants was very high; almost all the solids extracted from the settler were retained on the top of the SDRBs creating, by the same way, a filtration layer strongly biologically active.

COD dCOD SS KN NH4-N TP Removal rate (%) 92 48 97 87 62 77

Concentration (mg.l-1) 301 (91) 187 (51) 90 (45) 27 (7) 15 (6) 6.7 (9.3) Table 7: Removal and outlet concentrations (standard deviation) of the SDRBs.

Drainage and infiltration capacity of the SDRBs reveal that the hydraulic behavior of the beds is correct during one week of feeding. In March (critic period for the bed because of the absence of reeds), it is only after 9 days of feeding that beds drainage becomes detrimental for oxygen renewal. Thus a feeding/rest period of 1 week/ 3 week appears to be well adapted for primary sludge at the beginning of the plant functioning.

7.2.3 Sludge evolution

At the end of the experiment, only 5 cm of sludge are accumulated on the filters. The dry matter content of the accumulated sludge varied from 20 to 50 % depending on the season and the distance from the feeding point. This height of sludge is smaller than the one expected. A high mineralization rate can be mentioned to explain this phenomena in relation to the mass of sludge accumulated over the two years of the study (600 kg of DM). Nevertheless, this has to be moderated because of the limited VSS measured on the sludge (72 % of VSS for the primary sludge and 50 % of VSS for the stored sludge) Analyses of the sludge (Table 8) confirm that the treated sludge complies with pertinent regulations to be disposed of as agricultural amendment.

Aurignac sludge concentrations (mg/kgDM)

Normative values for spreading (mg/kgDM)

NF U 44-095 Cd 1,1 10 Cr 39 1000 Cu 205 1000 Hg 0.35 10 Ni 22 200 Pb 72 800 Zn 891 3000

Cr + Cu + Ni + Zn 1157 4000

Table 8: concentrations in trace element of the sludge accumulated onto the filter 4.

This concentrations of trace elements comply with the French legislation for agricultural spreading. This way of reutilization have to be confirmed by others analyses of trace organic compounds (PCB, Fluoranten, Benzo(b)Fluoranten, Benzo(a)pyren).

8 ENVIRONNEMENTAL/ÉCONOMIQUE BENEFITS Environmental benefits induced by this process can be resumed in the following points :

A total needed surface minimized in relation to classical ponds systems and thus a diminished constraint for some communities.

Improvement of water quality discharge into the receiving bodies with low level of oxydable compounds.

An high capacity of storm water treatment (SS drained during such events). Mass of pollutant, in some case non-negligible, being generally discharge directly in the receiving bodies in classical plant configuration.

An ecological in situ sludge treatment allowing a diminution of the sludge volume production. Sludge are moreover well mineralized and aerobic. The withdrawal of the sludge is easier and allow to not stop the water treatment line.

9 TRANSFERABILITY AND APPLICATION Results learnt on the Aurignac demonstration plant give major elements to design new WWTP and extend existing ponds systems. In term of sizing new WWTPs, two configuration can be pointed out (see Figure 11).

Lagune de maturation

6 m6 m22//EqHEqH

Filtres plantés roulés ou concassésh=40 cm (ou 25cm)

1 m1 m22//EqHEqH

2,5 m2,5 m22//EqHEqH

Lagune facultative

Lagune facultative

Filtres plantésroulés ou concassésh=40 cm

6 m6 m22//EqHEqH

1m1m22//EqHEqH

Décanteur

Figure 11 : designs for new plant realization.

In the case where it is not chosen to use primary settling tank with daily extraction, it is recommended to use two ponds in series (one facultative and a maturation pond) to ensure an acceptable organic load on the filter. This configurations allow to reduce the classical required surface for ponds systems (11 m2.p.e.-1 in France) with a better outlet water quality. This design can be used for all continental climate countries and also for Mediterranean climate countries.

The acceptance of storm waters of combined systems is also a point which can be generalized for all European countries taken into account the hydrogram of each communities to well design the pond. In term of pond systems arriving at or exceeding their nominal load, this study gives necessary elements to design a final adapted filtration step to improve the water outlet quality. In the same way, the SDRBs prototype has showed the perfect adequacy of the settling tank and the SDRBs to treat the primary sludge and facilitate its management. Also, it has to be mentioned that such a process is transposable to other countries of the European community.

10 REFERENCES AFNOR, (2005), Recueil Normes & Réglementation Environnement. Qualité de l’eau Vol 1 (552p) et Vol 2 (502p). Alcalde, L. G. Oron, L. Gillerman, M. Salgot and Y. Manor. (2003). Removal of fecal coliforms, somatic coliphages and F-specific bacteriophages in a stabilization pond and reservoir system in arid regions. Water Supply, 3(4), 177-184. APHA (1998). Standard Methods for the examination of water and wastewater. 19th ed. American Public Health association, Washington DC. Boutin, C., A. Liénard, N. Billote and J.P. Naberac. (2002). Association of waste stabilization ponds and intermittent sand filters: the pilot results and the demonstration plant of Aurignac. Proceedings of the 8th International Conference on Wetland system for water pollution control, Arusha (Tanzania), 543-555. Cavalcanti P., A. van Haandel and G. Lettinga. (2002). Sludge accumulation in polishing ponds treating anaerobically digested wastewater. Water Science and Technology, 45(1), 75–81 Davis-Colley, R. J., Donnison, A. M., Speed, D., Ross, C.M. and Nagels, J. W. (1999). Inactivation of faecal indicator microorganisms in waste stabilization ponds: interactions of environmental factors with sunlight. Water Research, 33(5), 1220-1230. Johnson M. and D.D. Mara (2005). Aerated rock filters for enhanced nitrogen and faecal coliform removal from facultative waste stabilization pond effluents. Water Science and Technology, 51(12), 99-102. Kayser, K., Kunst S., Fehr G., Voermanek H., (2002). Nitrification in reed beds-capacity and potential control methods. Water Science and Technology, 46(6-7), 363–370. Liénard, A., Guellaf H and Boutin C. (2001) Choice of the sand for sand filters used for secondary treatment of wastewater. Water Science and Technology, 44(2-3), 189-196. Middlebrooks, E. J. (1995). Upgrading pond effluents: an overview. Water Science and Technology, 31(12), 353-368. Neder, K.D., G.A. Carneiro, T.R. Queiroz and M.A.A. de Souza. (2002). Selection of natural treatment processes for algae removal from stabilisation ponds effluents in Brasilia, using multicriterion method. Water Science and Technology, 46(4-5), 347-354. Racault Y., Boutin C. (2005). Waste Stabilisation Ponds in France. State of the art and recent trends. Water Science and Technology, 51(12), 1-9. Racault, Y. C. Boutin and A. Seguin. (1995). Waste stabilization ponds in France: a report on fifteen years experience. Water Science and Technology, 31(12), 91-101. Sezerino, P.H., V. Reginatto, M. A. Santos, K. Kayser, L. S. Philippi and H. M. Soares. (2003).Nutrient removal from piggery effluent using vertical flow constructed wetlands in southern Brazil. Water Science and Technology, 48(2), 129-135. WEF (1992). Design of Municipal Wastewater Treatment Plants, Vol. 2, Alexandria, VA.