low temperature treatment of domestic sewage in upflow anaerobic sludge blanket and anaerobic hybrid...

PII: SO273-1223(99)00100-6

Wat. Sci. Tech. Vol. 39, No. 5, PP. 177-185, 1999 Q 1999 IAWQ

Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved

0273-1223199 $19.00 + 0.00

LOW TEMPERATURE TREATMENT OF DOMESTIC SEWAGE IN UPFLOW ANAEROBIC SLUDGE BLANKET AND ANAEROBIC HYBRID REACTORS

Tarek A. Elmitwalli, Marcel H. Zandvoort, Grietje Zeeman, Harry Bruning and Gatze Lettinga

Department of Environmental Technology, Wageningen Agricultural University, Bomenweg 2, 6703 HD Wageningen, The Netherlands

ABSTRACT

The treatment of sewage at a temperature of 13°C was investigated in three reactors (each 3.84 litm) a UASB and two anaerobic hybrid (AH) reactors with small sludge granules with an average diameter of 0.73 mm. The media used in the AH reactors were vertical polyurethane foam sheets. The reactors were operated at a HBT of 8 h. The use of small sludge granules and operating the reactors at low upflow velocity (1.8 m/d) improved suspended COD removal effkiencies for the UASB reactor. Moreover, the use of sheets in the AH reactors significantly increased suspended COD removal efftciencies as compared to the UASB and reached to 87% for pre-settled sewage treatment. The treatment of pre-settled sewage instead of raw sewage in AH reactors significantly increased colloidal and dissolved COD removal efficiencies with 13% and 12% respectively and colloidal COD removal efftciency for the UASB reactor with 13%. At ‘steady state’ for pre- settled sewage treatment, the AH reactors removed 64% of the total COD which is significantly higher by 4% than the UASB reactor. Therefore, the anaerobic treatment of domestic sewage at low temperature can be improved by treating pre-settled sewage in shallow AH reactors containing small sludge granules. D 1999 IAWQ Published by Elsevier Science Ltd. All rights reserved

KEYWORDS

Anaerobic treatment; domestic sewage; granular sludge; low temperature; polyurethane foam.

INTRODUCTION

Anaerobic digestion is an attractive process for the treatment of domestic sewage as little energy is needed for the process itself and moreover removed organic matter is converted to biogas which can be used for the production of energy. As hardly any nutrients are removed during digestion, the effluent of the treated sewage followed by pathogens removal is a valuable product for irrigation and fertilization to close the water and nutrient cycle. Therefore, anaerobic digestion of domestic wastewater is an attractive treatment system for community on-site (Lettinga et al. 1993; Lettinga, 1996). Although the anaerobic treatment of domestic sewage has been applied on a large scale in several tropical countries (Hulshoff Pol ef al., 1997), the process is so far not applied on a full scale in countries with lower temperatures, mainly as results of lower removal effkiencies. Also, at low temperature, a longer HRT (>12 h) is needed (Inamori et al., 1983; Man et al., 1986; Lettinga, 1996) and the accumulated SS increases with decreasing temperature (Inamori et al., 1983; Genung et al., 1985; Sanz and Fdz-Polanco, 1990).

177

~ Pergamon

PH: S0273-1223(99)00100-6

Wal. Sci. Tech. Vol. 39, No.5, pp. 177-185,1999

© 19991AWQPublished by Elsevier Science Ltd

Printed in Great Britain. All rights reserved0273-1223/99 $19.00 + 0.00

LOW TEMPERATURE TREATMENT OFDOMESTIC SEWAGE IN UPFLOWANAEROBIC SLUDGE BLANKET ANDANAEROBIC HYBRID REACTORS

Tarek A. Elmitwalli, Marcel H. Zandvoort,Grietje Zeeman, Harry Bruning and Gatze Lettinga

Department ofEnvironmental Technology, WageningenAgricultural University,Bomenweg 2, 6703 HD Wageningen, The Netherlands

ABSTRACT

The treatment of sewage at a temperature of 13°C was investigated in three reactors (each 3.84 litre) a UASBand two anaerobic hybrid (AH) reactors with small sludge granules with an average diameter of 0.73 mm.The media used in the AH reactors were vertical polyurethane foam sheets. The reactors were operated at aHRT of 8 h. The use of small sludge granules and operating the reactors at low upflow velocity (1.8 mid)improved suspended COD removal efficiencies for the UASB reactor. Moreover, the use of sheets in the AHreactors significantly increased suspended COD removal efficiencies as compared to the UASB and reachedto 87% for pre-settled sewage treatment. The treatment of pre-settled sewage instead of raw sewage in AHreactors significantly increased colloidal and dissolved COD removal efficiencies with 13% and 12%respectively and colloidal COD removal efficiency for the UASB reactor with 13%. At 'steady state' for presettled sewage treatment, the AH reactors removed 64% of the total COD which is significantly higher by4% than the UASB reactor. Therefore, the anaerobic treatment of domestic sewage at low temperature can beimproved by treating pre-settled sewage in shallow AH reactors containing small sludge granules. © 1999IAWQ Published by Elsevier Science Ltd. All rights reserved

KEYWORDS

Anaerobic treatment; domelltic sewage; granular sludge; low temperature; polyurethane foam.

INTRODUCTION

Anaerobic digestion is an attractive process for the treatment of domestic sewage as little energy is neededfor the process itself and moreover removed organic matter is converted to biogas which can be used for theproduction of energy. As hardly any nutrients are removed during digestion, the effluent of the treatedsewage followed by pathogens removal is a valuable product for irrigation and fertilization to close thewater and nutrient cycle. Therefore, anaerobic digestion of domestic wastewater is an attractive treatmentsystem for community on-site (Lettinga et ai. 1993; Lettinga, 1996). Although the anaerobic treatment ofdomestic sewage has been applied on a large scale in several tropical countries (Hulshoff Pol et al., 1997),the process is so far not applied on a full scale in countries with lower temperatures, mainly as results oflower removal efficiencies. Also, at low temperature, a longer HRT (>12 h) is needed (Inamori et ai., 1983;Man et ai., 1986; Lettinga, 1996) and the accumulated SS increases with decreasing temperature (Inamori etai., 1983; Genung et ai., 1985; Sanz and Fdz-Polanco, 1990).

177

178 T. A. ELMITWALLI

The removal and degradation of colloidal particles which represent 20-30% of the total COD (COD,) for domestic sewage appear to become limiting in UASB reactors at lower temperatures (Mergaert et al., 1992). Wang (1994) mentioned that the removal and conversion of colloidal particles was the rate limiting step in a two step UASB+EGSB system treating raw sewage at a HRT of 3+2 h at low temperatures. He found average removal efficiencies for colloidal COD (COD& of 40% and 49% at temperatures of 17 and 12°C respectively. Yoda et al. (1985) also reported that the colloidal particles in the influent were difficult to remove and represented up to 60-70% of the effluent of an anaerobic fluidized bed reactor. Sayed and Fergala (1995) considered that the entrapment mechanism involved in removing solids is not sufficient to remove colloidal particles through the sludge bed with high porosity and under upflow and gasification. Non-published results of a fed batch recirculation experiment with raw and pre-settled sewage indicate that the presence of SS in the influent negatively affects the removal of COD,,,.

The UASB reactor inoculated with granular sludge showed a better performance than that inoculated with flocculant sludge (Lettinga et al., 1983). Anaerobic treatment of sewage with UASB and EGSB reactors using granular sludge has been studied extensively by Lettinga et a/.(1983), Man et al. (1988), Last et a1.(1992) and Wang (1994) with large sludge granules. The recent researches in anaerobic granular sludge showed that the small sludge granules have a higher methanogenic activity than large sludge granules (Grotenhuis et al. 1991, Alphenaar, 1994, Rebac et al., 1997). The lower activity of large sludge granules may be due to the substrate transport limitations increase with the diameter of the granules (Alphenaar, 1994). The use of small sludge granules can also improve the filtration and entrapment mechanisms in the sludge bed which will lead to improving COD,, and COD,,! removal.

The performance of an anaerobic hybrid (AH) reactor depends on the contact of the wastewater with both the biomass in the sludge bed and attached and suspended biomass in the anaerobic filter (AF) at the top of the reactor. Moreover, the AF layer helps in retaining biomass inside the reactor and gas/solids separation (Tilche and Vieira, 1991). Huysman et al. (1983) mentioned that reticulated polyurethane foam (RPF) appeared an excellent colonization matrix for AF. The RPF has a high specific surface area which can reach up to 2400 mZ/m3 and high porosity of 97% (Huysman et al., 1983). The RF’F is able to retain over 15 gVS/l in the attached form (Tilche and Vieira, 1991). Table 1 summarized recent results for treatment of sewage at low temperatures (~20’C) with UASB, EGSB, AF and AH reactors.

Table 1. Summary of recent results for treatment of sewage at low temperatures (120°C) with UASB, EGSB, AF and AH reactors

Reactor Sludge Sewage T HRT Vu, Influent Removal (%) Reference

UASB We

F R CC)

20 (h) 18

ww 0.11

COD,(mg/l) 550

COD, 55-75

COD,(TSS) Lettinga et aZ., 1981

UASB F R 20 8

UASB F R 8-20 8

UASB G R 7-8 9-14

AF F R 10 6

UASB G R 20 4

UASB F R 12-18 18

UASB G R 12-20 7-8

UASB R 20 6

EGSB G S >I3 l-2

UASB+ F,Gb R 17 3,2b

UASB+ F,Gb R 12 3,2b

UASB+ F R 20 9,9b

UASB G R 13 8

AH G R 13 8

UASB G S 13 8

AH G S 13 8

0.40

0.40

0.14

0.70

0.70

0.23

6

1,6b

1,6b

0.2,0.Zb

0.075 0.075

0.075

0.07s

500 75"

400 30-50

467-700 57

529 53.7 424 60 465 65

190-1180 30-75

1076 64

391 16-34

697 69

507 51

536= 80

456 65

456 66

344 59

344 61

72

(69) (73)

(Z)

79

67

(95) 88

92

79

87

Grin et al., 1983 Grin et al., 1983

Man et al., 1986

Derycke and Verstraete, Vieira and Souza, 1986 Monroy et al., 1988 Man et al., 1988 Mergaert et al., 1992 Last and Lettinga, 1992 Wang, 1994 Wang,1994

Tang el al., 1995 this study this study this study this study

F: flocculant sludge; G: granular sludge; R: raw sewage; S: pre-settled sewage. a: COD removal based on total influent and filtered effluent; b: the second reactor; c: the wastewater was raw sewage and brewery wastewater


The re~oval and degradation of colloidal particles which represent 20-30% of the total COD (CODt) for

domestIc sewage appear to become limiting in UASB reactors at lower temperatures (Mergaert et ai., 1992).Wang (1994) mentioned that the removal and conversion of colloidal particles was the rate limiting step in atwo step UASB+EGSB system treating raw sewage at a HRT of 3+2 h at low temperatures. He foundaverage removal efficiencies for colloidal COD (CODeol) of 40% and 49% at temperatures of 17 and 12°Crespectively. Yoda et ai. (1985) also reported that the colloidal particles in the influent were difficult toremove and represented up to 60-70% of the effluent of an anaerobic fluidized bed reactor. Sayed andFergala (1995) considered that the entrapment mechanism involved in removing solids is not sufficient toremove colloidal particles through the sludge bed with high porosity and under upflow and gasification.Non-published results of a fed batch recirculation experiment with raw and pre-settled sewage indicate thatthe presence of SS in the influent negatively affects the removal of CODeol.

The UASB reactor inoculated with granular sludge showed a better performance than that inoculated withflocculant sludge (Lettinga et ai., 1983). Anaerobic treatment of sewage with UASB and EGSB reactorsusing granular sludge has been studied extensively by Lettinga et ai.(l983), Man et ai. (1988), Last etal.(l992) and Wang (1994) with large sludge granules. The recent researches in anaerobic granular slUdgeshowed that the small sludge granules have a higher methanogenic activity than large sludge granules(Grotenhuis et ai. 1991, Alphenaar, 1994, Rebac et ai., 1997). The lower activity of large sludge granulesmay be due to the substrate transport limitations increase with the diameter of the granules (Alphenaar,1994). The use ofsmall sludge granules can also improve the filtration and entrapment mechanisms in thesludge bed which will lead to improving CODss and CODcol removal.

The performance of an anaerobic hybrid (AH) reactor depends on the contact of the wastewater with boththe biomass in the sludge bed and attached and suspended biomass in the anaerobic filter (AF) at the top ofthe reactor. Moreover, the AF layer helps in retaining biomass inside the reactor and gas/solids separation(Tilche and Vieira, 1991). Huysman et ai. (1983) mentioned that reticulated polyurethane foam (RPF)appeared an excellent colonization matrix for AF. The RPF has a high specific surface area which can reachup to 2400 m2/m3 and high porosity of97% (Huysman et ai., 1983). The RPF is able to retain over 15 gVS/lin the attached form (Tilche and Vieira, 1991). Table I summarized recent results for treatment of sewage atlow temperatures (,;;20°C) with UASB, EGSB, AF and AH reactors.

Table 1. Summary of recent results for treatment of sewage at low temperatures (~20°C) with UASB,EGSB, AF and AH reactors

Reactor Sludge Sewage T HRT Vu• Influent Remova1(%) Reference

type (ue) (h) (mIh) COD,(mgll) COD, COD,,{TSS)

VASB F R 20 18 0.11 550 55-75 Lettinga et ai., 1981VASB F R 20 8 0.40 500 75' Grin et ai., 1983

VASB F R 8-20 8 0.40 400 30-50 Grin et ai., 1983

VASB G R 7-8 9-14 0.14 467-700 57 72 Man et ai., 1986

AF F R 10 6 0.70 529 53.7 Derycke and Verstraete,

VASB G R 20 4 0.70 424 60 (69) Vieira and Souza, 1986

VASB F R 12-18 18 465 65 (73) Monroy et ai., 1988

VASB G R 12-20 7-8 0.23 190-1180 30-75 60 Man et ai., 1988

VASB R 20 6 1076 64 (88) Mergaert et ai., 1992

EGSB G S >13 1-2 6 391 16-34 Last and Lettinga, 1992

VASB+ F,Gb R 17 3,2b 1,6& 697 69 79 Wang, 1994

VASB+ F,G& R 12 3,2b 1,6b 507 51 67 Wang, 1994

VASB+ F R 20 9,9& 0.2,0.2b 536< 80 (95) Tang et ai., 1995

VASB G R 13 8 0.075 456 65 88 this study

AH G R 13 8 0.075 456 66 92 this study

VASB G S 13 8 0.075 344 59 79 this study

AH G S 13 8 0.075 344 61 87 this study

F: flocculant sludge; G: granular sludge; R: raw sewage; S: pre-settled sewage.a: COD removal based on total influent and filtered effluent; b: the second reactor; c: the wastewater wasraw sewage and brewery wastewater

Low temperature treatment of domestic sewage 179

In the present research, the anaerobic treatment of sewage using UASB and AH reactors with small sludge granules has been investigated at a temperature of 13“C. The RPF sheets were used as packing material (PM) which were oriented vertically with two different spacings. Also, the effect of influent COD, concentration on the removal of COD,, has been studied by operating the reactors with raw sewage and pre-settled sewage with settling time of 14 h.

MATERIALS AND METHODS

Figure 1 shows a schematic diagram of the experimental set-up. The experiment consisted of three reactors (each 3.84 l), one UASB reactor and two AH reactors. The three reactors have the same shape except for the gas/solids separator which is replaced by PM in the AH reactors. Each reactor made of plexiglass had a square cross section with 8 cm rib length. The height of each reactor was 70 cm with wastewater height of 60 cm. The PM used were RPF sheets (type Filteren TM10 from Recticel, Buren, The Netherlands) which were oriented vertically in the AH reactors. Each sheet had nobs at one side while the other side was flat. The characteristics of the used RPF sheets are presented in Table 2.

-----------------------

0

Figure 1. Schematic diagram of the experiment. 1, influent; 2, peristaltic pump; 3, UASB reactor; 4, AH reactor with 3 sheets; 5, AH reactor with 2 (raw sewage) or 4 sheets (pre-settled sewage); 6, effluent; 7, biogas; 8,

temperature controlled room at 13%.

Table 2. The characteristics of the RPF sheets used in the experiment

Parameter unit Value

Sheet width cm 8

Total sheet thickness mm 20

Base thickness mm 10

Nob thickness mm 10

Specific surface area m2/m3 500

Density kg/m3 19-22

Number of pores pore/inch 7-15

Pore size mm 2.5

The two AH reactors varied from the number of RPF sheets. The first AH reactor contained 3 sheets corresponding to a spacing of 0.5 cm between the sheets. The second AH reactor contained 2 sheets corresponding to a spacing of 1.3 cm between the sheets. After 57 days of operation, the ‘Zsheets’ reactor was modified to a ‘Csheets’ reactor without spacing between the sheets. The HRT for each reactor was 8 h. For each reactor, the C!& gas was collected by using a gas bag. The volume of CH4 was determined by pumping the biogas through a NaOH solution (3%) and wet gas-meter. Dissolved CH4 in the effluent was calculated according to Henry’s law and added to the produced C& gas.


In the present research, the anaerobic treatment of sewage using UASB and AH reactors with small sludgegranules has been investigated at a temperature of 13°C. The RPF sheets were used as packing material (PM)which were oriented vertically with two different spacings. Also, the effect of influent COD.. concentration onthe removal of CODcol has been studied by operating the reactors with raw sewage and pre-settled sewage withsettling time of 14 h.

MATERIALS AND METHODS

Experimental set-up

Figure 1 shows a schematic diagram of the experimental set-up. The experiment consisted of three reactors(each 3.841), one UASB reactor and two AH reactors. The three reactors have the same shape except for thegas/solids separator which is replaced by PM in the AH reactors. Each reactor made of plexiglass had asquare cross section with 8 cm rib length. The height of each reactor was 70 cm with wastewater height of60 cm. The PM used were RPF sheets (type Filteren TMIO from Recticel, Buren, The Netherlands) whichwere oriented vertically in the AH reactors. Each sheet had nobs at one side while the other side was flat.The characteristics ofthe used RPF sheets are presented in Table 2.

-8--------------------

-----------------------o

Figure 1. Schematic diagram of the experiment. I, influent; 2, peristaltic pump; 3, UASB reactor; 4, AH reactorwith 3 sheets; 5, AH reactor with 2 (raw sewage) or 4 sheets (pre-settled sewage); 6, effluent; 7, biogas; 8,

temperature controlled room at B·C.

Table 2. The characteristics of the RPF sheets used in the experiment

Parameter Unit Value

Sheet width cm 8

Total sheet thickness mm 20

Base thickness mm 10

Nob thickness mm 10

Specific surface area m2/m3 500

Density kg/m3 19-22

Number ofpores pore/inch 7-15

Pore size mm 2.5

The two AH reactors varied from the number of RPF sheets. The first AH reactor contained 3 sheetscorresponding to a spacing of 0.5 cm between the sheets. The second AH re~ctor co?tained 7sheetscorresponding to a spacing of 1.3 cm between the sheets. After 57 days of operation, the 2-sheets reactorwas modified to a '4-sheets' reactor without spacing between the sheets. The HRT for each reactor ~as 8 h.For each reactor, the C~ gas was collected by using a gas bag. The vOI.ume of CH4 ~as determmed bypumping the biogas through a NaOH solution (3%) and wet gas-meter. Dissolved CH4 m the effiuent wascalculated according to Henry's law and added to the produced C~ gas.


The sewage used in the experiment was from the village Bennekom, The Netherlands. The sewage is collected in a combined sewer system and is continuously pumped to the experimental hall. The experiment was operated for the first 57 days with raw sewage and after that with pre-settled sewage from a 26 m3 settler with 2 hours settling time followed by a 35 litre settler in the refrigerator for 12 hours. The experiment with pre-settled sewage lasted for 93 days and was subsequently operated for 7 days with 2 hours pre-settled sewage. Table 3 shows COD fractions for sewage used in the experiment

Table 3. COD fractions for the sewage used in the experiment. Standard deviations are presented between two brackets

Parameter

Raw sewage

Pre-settled sewage (14 h settling time)

Pre-settled sewage (2 h settling time)

Period COD (mg/l)

(days) COD, COD,, C0Dco1 CObis

57 456(129) 229(81) 114(47) 112(34)

93 339(70) 82(34) 136(27) 124(36)

7 403(63) 181(64) 108(27) 114(25)

Sampling of the influent was made by using an additional head in the influent peristaltic pump. Composite samples for 24, 48 and 72 hours of the influent and the effluent of each reactor were collected in containers stored in the refrigerator at 4°C.

The reactors were inoculated with granular sludge from a full scale UASB reactor treating alcohol wastewater from Nedalco, Alcohol producing industry, Bergen op Zoom, The Netherlands. The seed sludge granules had an average diameter of 0.73 mm. The total amount of granular sludge added to each reactor was approximately 29 gVSSl1.

COD was analyzed using the micro-method as described by Jirka and Carter (1975). Raw samples were used for COD,, 4.4 pm folded paper filtered (Schleicher & Schuell 5951/z) samples for CODr and 0.45 pm membrane filtered (Schleicher Schuell ME 25) samples for dissolved COD (CODdi,). The COD,, and COD,,, were calculated by the differences between CODt and CODf, CODr and CODdi, respectively. Volatile fatty acids (VFA) were measured from membrane filtered samples. VFA were determined by gas chromatography. The chromatograph (Hewlett Packard 5890A, Palo Alto, USA) was equipped with a 2m x 2 mm (inner diameter) glass column, packed with Supelco port (loo-120 mesh) coated with 10% Fluorad FC 43 1. Operating conditions were: column, 13O“C; injection port, 2OO’C; flame ionization detector, 28O’C. Nz saturated with formic acid at 20°C was used as a carrier gas (30mFmin). TS, VS, SS and VSS were measured according to the Dutch Standard Normalized Methods (1969). All measurements were determined in duplicate.

Statistical Comparison between the performances of the reactors were done according to Snedecor and Co&ran (1980). The performance of each reactor was assumed to be independent with different varhCe.


The sewa.ge used i~ the experiment was. from the village Bennekom, The Netherlands. The sewage iscollected 10 a combmed sewer system. and IS continuously pumped to the experimental hall. The experimentwas oper~ted for the first .57 d~ys WIth raw sewage and after that with pre-settled sewage from a 26 m3

settle~ WIth 2. hours settlIng time followed by a 35 litre settler in the refrigerator for 12 hours. Theexpenment WIth pre-settled sewage lasted for 93 days and was subsequently operated for 7 days with 2hours pre-settled sewage. Table 3 shows COD fractions for sewage used in the experiment

Table 3. COD fractions for the sewage used in the experiment. Standard deviations are presented betweentwo brackets

Parameter Period COD (mg/I)

(days) CODt COD" CODeol CODdis

Raw sewage 57 456(129) 229(81) 114(47) 112(34)

Pre-settled sewage (14 h settling time) 93 339(70) 82(34) 136(27) 124(36)

Pre-settled sewage (2 h settling time) 7 403(63) 181(64) 108(27) 114(25)

SampliOK

Sampling of the influent was made by using an additional head in the influent peristaltic pump. Compositesamples for 24, 48 and 72 hours of the influent and the effluent of each reactor were collected in containersstored in the refrigerator at 4°C.

Seed sludKe

The reactors were inoculated with granular sludge from a full scale UASB reactor treating alcoholwastewater from Nedalco, Alcohol producing industry, Bergen op Zoom, The Netherlands. The seed sludgegranules had an average diameter of 0.73 mm. The total amount of granular sludge added to each reactorwas approximately 29 gVSS/1.

Analysis

COD was analyzed using the micro-method as described by Jirka and Carter (1975). Raw samples were usedfor CODt. 4.4 ).110 folded paper filtered (Schleicher & Schuell 5951/2) samples for CODr and 0.45 ).1mmembrane filtered (Schleicher Schuell ME 25) samples for dissolved COD (CODdis). The COD" andCODeol were calculated by the differences between CODI and CODr, CODr and CODdis respectively.Volatile fatty acids (VFA) were measured from membrane filtered samples. VFA were detennined by gaschromatography. The chromatograph (Hewlett Packard 5890A, Palo Alto, USA) was equipped with a 2m x2 mm (inner diameter) glass column, packed with Supelco port (100-120 mesh) coated with 10% FluoradFC 431. Operating conditions were: column, 130°C; injection port, 200°C; flame ionization detector, 280°C.N

2saturated with fonnic acid at 20°C was used as a carrier gas (30mllmin). TS, VS, SS and VSS were

measured according to the Dutch Standard Normalized Methods (1969). All measurements were detennined

in duplicate.

Statjstjcal analysis

Statistical Comparison between the perfonnances of the reactors were done according to Snedecor andCochran (1980). The perfonnance of each reactor was assumed to be independent with different variance.

Low temperature treatment of domestic sewage

RESULTS AND DISCUSSION 181

Treatment of raw sewage The results of the performance of the reactors are shown in Table 4. In the start-up period, removal efficiencies of CODt improved with time. In the third week (from day 15 to day 23), CODt removal efficiencies reached to 72%, 72% and 70% for UASB, ‘3-sheets-AH’ and ‘2-sheets-AH’ reactor respectively.

After the start-up period (33 days), a part of the sludge bed in the UASB reactor floated to the gas solid separator and returned again with time. The floating and returning of the sludge bed occurred continuously until termination of the experiment with raw sewage. From visual observations, aggregation of individual granules followed by entrapment of biogas in the aggregates can be considered the main cause for sludge bed flotation. The presence of vertical sheets in the AH reactors prevented the flotation of the sludge bed and only channels and gas pockets were formed in the sludge bed. The channels and gas pockets disappeared with time and returned again in another part of the sludge bed. The CODt removal efficiencies of the UASB reactor and AH reactors were practically the same. As the performance of the ‘2-sheets-AH’ reactor was similar to the ‘3-sheets-AH’ reactor, another 2 sheets were added.

Treatment ofpre-settled sewage The operational period with pre-settled sewage is divided into three periods, acclimatization period (from day 1 to day 64), ‘steady state’ period (from day 64 to day 93) and SS shock load period (from day 93 to day 101). Table 4 summarized the results of treating pre-settled sewage in each reactor.

Table 4. The average removal efficiencies for COD fractions for the UASD and AH reactors treating raw and pre-settled sewage. Standard deviations are shown between two brackets

Period Removal efficiency (%)

UASB reactor ‘3-sheets-AH’ reactor ‘2 or 4-sheets-AH’ reactor

CODt COD, COD,, CODdi, COD, COD,, COD,1 CODdi, COD, COD,, COD-1 CODdi,

Raw sewage treatment:- start-up

(Z) (E) (E) (Z) (E) 864) (Z) (ii) (Z) (“:, (ii) (i”9)

Pm-settled sewage treatment:- Whole

(‘93 ,:“, (Z) (E) (T:) (ii) (Z) (E) r; (“9”, ::) (Z) Acclimatization 58

(10) (E) (E) (E) (:$ (Z) (E) (T:) (E) (Z) (E) (E)

Steady state (“5”, (E) t: (“9”, li (“9’, (64: (if) ri (“5: (“4’, (2)

SS shock load (Z)

During the first two weeks of operation with pre-settled sewage, flotation of the sludge bed in the UASB and formation of channels and gas pockets in the sludge bed of the AH reactors still occurred although the influent COD,, decreased from 229 mg/l during the raw sewage period to 102 mg/l during the first two weeks of pre-settled sewage treatment. The sludge bed flotation and formation of channels and gas pockets gradually decreased with time and definitively stopped after two weeks. So, the sludge bed flotation in the UASB and formation of channels and gas pockets in the sludge bed of the AH reactors can be attributed to high COD,, concentration of raw sewage. During the acclimatization period, the effluent CODt, COD,,1 and CODdi, concentration decreased with operation time.

After an acclimatization period of 64 days, removal efficiencies and effluent concentrations became more stable and standard deviation decreased. The reactors were considered to be in ‘steady state’ in this period. The criteria for ‘steady state’ set by Noyola et al. (1988) and Polprasert et al. (1992) were met during this period. Noyola et al. (1988) assumed ‘steady state’ of an anaerobic reactor treating domestic wastewater to

Low temperature treatment of domestic sewage

RESULTS AND DISCUSSION

Reactor perfonnance

181

Treatment ofraw sewage

The results of the performance of the reactors are shown in Table 4. In the start-up period, removaleffic~enc~es of COD! improved with time. In the third week (from day 15 to day 23), COD! removalefficIencIes reached to 72%, 72% and 70% for UASB, '3-sheets-AH' and '2-sheets-AH' reactor respectively.After the start-up period (33 days), a part of the sludge bed in the UASB reactor floated to the gas solidseparator and returned again with time. The floating and returning of the sludge bed occurred continuouslyuntil termination of the experiment with raw sewage. From visual observations, aggregation of individualgranules followed by entrapment of biogas in the aggregates can be considered the main cause for sludgebed flotation. The presence of vertical sheets in theAH reactors prevented the flotation of the sludge bedand only channels and gas pockets were formed in the sludge bed. The channels and gas pockets disappearedwith time and returned again in another part of the sludge bed. The COD! removal efficiencies of the UASBreactor and AH reactors were practically the same. As the performance of the '2-sheets-AH' reactor wassimilar to the '3-sheets-AH' reactor, another 2 sheets were added.

Treatment ofpre-settled sewageThe operational period with pre-settled sewage is divided into three periods, acclimatization period (fromday I to day 64), 'steady state' period (from day 64 to day 93) and SS shock load period (from day 93 to day101). Table 4 summarized the results oftreating pre-settled sewage in each reactor.

Table 4. The average removal efficiencies for COD fractions for the UASa and AH reactors treating rawand pre-settled sewage. Standard deviations are shown between two brackets

Period Removal efficiency (%)

UASB reactor '3-sheets-AH' reactor '2 or 4-sheets-AH' reactor

COD, COD" CODcol CaDdis COD, COD" CODcol CaDdis COD, COD" CODcol CaDdis

Raw sewage treatment:-

Start-up 67 90 48 39 67 94 37 40 63 91 31 39(18) (24) (24) (30) (16) (6) (39) (25) (16) (5) (22) (29)

Pre-settled sewage treatment:-

Whole 59 79 59 45 61 87 57 47 62 88 57 48(9) (16) (14) (13) (11) (10) (14) (16) (9) (9) (14) (14)

Acclimatization 58 78 61 42 58 86 55 44 59 89 55 46

(10) (18) (13) (14) (12) (10) (14) (17) (10) (10) (14) (15)

Steady state 60 79 60 49 64 82 65 52 64 85 65 51

(5) (12) (8) (9) (4) (9) (4) (11) (4) (5) (4) (10)

SS shock load 60 86 36 40 65 93 43 40 66 90 43 49

(9) (8) (28) (15) (7) (6) (26) (21) (8) (5) (25) (12)

During the first two weeks of operation with pre-settled sewage, flotation of the sludge bed in the UASB andformation of channels and gas pockets in the sludge bed of the AH reactors still occurred although theinfluent CODss decreased from 229 mg/l during the raw sewage period to 102 mg/l during the first twoweeks of pre-settled sewage treatment. The sludge bed flotation and formation of channels and gas pocketsgradually decreased with time and definitively stopped after two weeks. So, the sludge bed flotation in theUASB and formation of channels and gas pockets in the sludge bed of the AH reactors can be attributed tohigh CODss concentration of raw sewage. During the acclimatization period, the effluent CODt, CODcol andCODdis concentration decreased with operation time.

After an acclimatization period of 64 days, removal efficiencies and effluent concentrations became morestable and standard deviation decreased. The reactors were considered to be in 'steady state' in this period.The criteria for 'steady state' set by Noyola et ai. (1988) and Polprasert et ai. (1992) were met during thisperiod. Noyola et ai. (1988) assumed 'steady state' of an anaerobic reactor treating domestic wastewater to


be achieved after an operation time equal to 10 times the new HRT with a minimum of 2 weeks. Polprasert et al. (1992) considered ‘steady state’ when the eftluent COD concentrations at the anaerobic treatment of slaughterhouse wastewater were constant or varied within 10%.

The removal efficiency of COD, was the same for the three reactors during the acclimatization period. At ‘steady state’, the AH reactors removed 64% of the CODt which is significantly higher by 4% than the UASB reactor (level 2.5% for ‘3-sheets-AH’ reactor, level 5% for ‘4-sheets-AH’ reactor) mainly caused by significantly better removal of COD,,1 (level 0.5% for ‘3-sheets-AH’ reactor, level 2.5% for ‘4-sheets-AH reactor). Although, the pre-settled sewage COD,, concentration was low, the three reactors showed better removal efficiencies for COD,, than that found in previous researches at the same temperature (see Table 1) probably due to operating the reactors at low upflow velocity and the use of small sludge granules. The use of sheets in the AH reactors significantly increased COD,, removal efficiency as compared to the UASB reactor for the whole period of pre-settled sewage treatment.

No significant differences were found for CODdi, removal efftciency between the AH reactors and the UASB reactor. Although the reactors were operated at low upflow velocity (0.075 m/h), it seemed that there was a sufficient contact between dissolved substrate and biomass. The average CODdi, removal efficiencies during ‘steady state’ for the UASB,‘S-sheets-AH’ and ‘4-sheets-AH’ reactor represented respectively 91%, 96% and 94% of the maximum removal efficiency of 54% achieved by Last and Lettinga (1992) at batch recirculation of pre-settled sewage at upflow velocity of 6 mih. VFA were completely removed in all reactors indicating that methanogenesis was not rate limiting. The use of small sludge granules, which have less substrate transport limitations (Alphenaar, 1994), may be the reason for high removal efficiency for CODdi, and complete removal of VFA.

The shock load of COD,, did not affect the overall removal efticiency of the three reactors while the effluent COD concentration increased. The effluent concentration of COD,,, increased for all reactors and removal efficiency decreased considerably. The decrease in removal efficiency of COD,,, might be caused by

production of colloidal from extra SS in the influent.

No significant difference was shown between the performance of the ‘3-sheets-AH’ and ‘4-sheets-AH’ reactor except for CODdi, which was significantly better at level 10% for the ‘4-sheets-AH’ as compared to the ‘3- sheets-AH’ reactor on applying the SS shock load.

The results of methane production of the ‘3-sheets-AH’ reactor are summarized in Table 5. The average methane production during raw sewage treatment was similar to that found by Lettinga et al. (1983) (0.21 m3 CH4/kg COD removed), by Kobayashi et al. (1983) (0.16 m3 CH&g COD removed) or by Noyola et al. (1989) (0.17 m3 CH4 (STP)/kg COD removed). Moreover, the methane production increased with pre-settled sewage treatment and the conversion of removed COD to methane was higher than the theoretical value (0.35 m3 CH4 (STP)Acg COD removed) during ‘steady state’. Although, the reactors met the ‘steady state’ based on the stability of the effluent COD concentration as mentioned by Noyola et al. (1988) and Polprasert et al (1992), the high methane production showed that a real steady state was not achieved. The extra methane was presumably produced from the hydrolysis of SS accumulated in the sludge bed during the raw sewage treatment.

Table 5. Methane production during the treatment of raw sewage and pre-settled sewage. standard deviations are shown between two brackets

Sewage Period Methane production Conversion of removed

m’ CH4 (STP)/kg m’ CH, (STP)ikg COD COD to CH, (%) COD added removed

Raw fromday38to57 0.17(0.03) 0.25(0.04) 72(12)

Pre-settled Acclimatization 0.18(0.04) 0.30(0.06) 86(17)

‘steady state’ 0.25(0.04) 0.39(0.07) 112(20)


be achieved after an operation time equal to 10 times the new HRT with a minimum of 2 weeks. Polprasertet al. (1992) considered 'steady state' when the effiuent COD concentrations at the anaerobic treatment ofslaughterhouse wastewater were constant or varied within 10%.

The removal efficiency of CODt was the same for the three reactors during the acclimatization period. At'steady state', the AH reactors removed 64% of the CODt which is significantly higher by 4% than the DASBreactor (level 2.5% for '3-sheets-AH' reactor, level 5% for '4-sheets-AH' reactor) mainly caused bysignificantly better removal of CODeol (level 0.5% for '3-sheets-AH' reactor, level 2.5% for '4-sheets-AH'reactor). Although, the pre-settled sewage CODss concentration was low, the three reactors showed betterremoval efficiencies for COD" than that found in previous researches at the same temperature (see Table 1)probably due to operating the reactors at low upflow velocity and the use of small sludge granules. The useof sheets in the AH reactors significantly increased COD" removal efficiency as compared to the DASBreactor for the whole period ofpre-settled sewage treatment.

No significant differences were found for CaDdis removal efficiency between the AH reactors and theDASB reactor. Although the reactors were operated at low upflow velocity (0.075 mIh), it seemed that therewas a sufficient contact between dissolved substrate and biomass. The average CaDdis removal efficienciesduring 'steady state' for the DASB,'3-sheets-AH' and '4-sheets-AH' reactor represented respectively 91 %,96% and 94% of the maximum removal efficiency of 54% achieved by Last and Lettinga (1992) at batchrecirculation of pre-settled sewage at upflow velocity of 6 mIh. VFA were completely removed in allreactors indicating that methanogenesis was not rate limiting. The use of small sludge granules, which haveless substrate transport limitations (Alphenaar, 1994), may be the reason for high removal efficiency forCaDdis and complete removal ofVFA.

The shock load of COD" did not affect the overall removal efficiency of the three reactors while the effiuentCOD concentration increased. The effiuent concentration of CODeol increased for all reactors and removalefficiency decreased considerably. The decrease in removal efficiency of CODeo! might be caused byproduction of colloidal from extra SS in the influent.

No significant difference was shown between the performance of the '3-sheets-AH' and '4-sheets-AH' reactorexcept for CaDdis which was significantly better at level 10% for the '4-sheets-AH' as compared to the '3sheets-AH' reactor on applying the SS shock load.

Bj0ll'as productjon

The results of methane production of the '3-sheets-AH' reactor are summarized in Table 5. The averagemethane production during raw sewage treatment was similar to that found by Lettinga et al. (1983) (0.21m3 ClLJkg COD removed), by Kobayashi et at. (1983) (0.16 m3 CH4/kg COD removed) or by Noyola et al.(1989) (0.17 m3 CH4 (STP)/kg COD removed). Moreover, the methane production increased with pre-settledsewage treatment and the conversion of removed COD to methane was higher than the theoretical value(0.35 m3 CH4 (STP)/kg COD removed) during 'steady state'. Although, the reactors met the 'steady state'based on the stability of the effiuent COD concentration as mentioned by Noyola et al. (1988) and Polprasertet al (1992), the high methane production showed that a real steady state was not achieved. The extramethane was presumably produced from the hydrolysis of SS accumulated in the sludge bed during the rawsewage treatment.

Table 5. Methane production during the treatment of raw sewage and pre-settled sewage.standard deviations are shown between two brackets

Sewage

RawPre-settled

Period

from day 38 to 57Acclimatization

'steady state'

Methane productionm3 CH4 (STP)/kg m3 C~ (STP)/kg COD

COD added removed0.17(0.03) 0.25(0.04)

0.1 8(0.04) 0.30(0.06)

0.25(0.04) 0.39(0.07)

Conversion ofremovedCODtoC~(%)

72(12)

86(17)

112(20)


The sludge was wasted two times mainly from the top of the sludge bed for maintaining the sludge bed almost at the same level (Table 6). Miron (1997) mentioned that sludge residence time (SRT) plays a crucial rule in anaerobic treatment and SRT is an efficient parameter in controlling the process. Calculating the SRT from total biomass in the reactor and wasted biomass gave a high SRT (~500 d) for all reactors. The wasted biomass mainly contained flocculant sludge while the biomass in the reactors was mainly granular sludge. Separation between SRT of flocculant sludge and granular sludge is more realistic. However, accurate separating and measuring of flocculant sludge and granular sludge was difficult especially in the AH reactors which contained attached biomass to the media.

Table 6. Wasted sludge during the treatment of raw and pre-settled sewage

Time UASB reactor ‘3-sheets-AH’ ‘2 or 4 sheets-AH’ reactor reactor (gVSS/d)

(g VSS/d) (g VSS/d)

At the end of raw sewage treatment 0.21 0.2 1 0.22 After 41 days of pre-settled sewage treatment 0.16 0.15 0.17

The treatment of raw sewage in the UASB reactor with small sludge granules under lower upflow velocity and temperature seemed to be not practical due to the sludge bed flotation. The high influent COD, was the reason for sludge bed flotation. The presence of RPF sheets in the AH reactor prevented sludge bed flotation and only channels and gas pockets were formed in the sludge bed of the AH reactors.

The treatment of pre-settled sewage instead of raw sewage improved COD,,, removal efficiency of the UASB reactor by 13% (from 46% to 59%) and this difference was significant at level 2.5%. The ‘3-sheets- AI-I’ reactor removal efficiency for COD c0l as well as CODdis improved significantly (level 10%) by 13% (from 44% to 57%) and 12% (from 36% to 48%) respectively. The reduced COD,, levels in the influent of the pre-settled sewage are quite likely the cause of the improved COD,, removal. The latter indicates that part of the colloids in the effluent at the anaerobic treatment of raw sewage is produced from SS in the influent.

Wang (1994) showed that a high loaded UASB (HRT=3 hours) resulted in winter (12°C) in a COD,, removal efftciency of 49% and a mean effluent of 110 mg CODJl. As a result of the high upflow velocity in the subsequent EGSB reactor the total removal efficiency of COD,1 amounts to 49%. Combination of a high loaded UASB with a hybrid system could improve both the COD,, and the COD,,, removal, finally resulting in a COD, removal efficiency of ca 80% at 13’C at a HRT of 11 hours for the two systems together. The proposed two phase system could be attractive for raw sewage treatment at lower temperatures especially for community on-site. Optimization of the process with respect to HRT and .upflow velocity should be provided.

CONCLUSIONS

The use of small sludge granules and operating the reactors at low upflow velocity improved suspended COD removal efficiencies for the UASB reactor. Moreover, the use of sheets in the AH reactors significantly increased suspended COD removal efficiencies as compared to the UASB and reached to 87% for pre-settled SL vage treatment.

The treatment of pre-settled sewage instead of raw sewage in AH reactors significantly increased colloidal and dissolved COD removal efficiencies with 13% and 12% respectively and colloidal COD removal efftciency for the UASB reactor with 13%.

Wasted sludl'e


The sludge was wasted two times mainly from the top of the sludge bed for maintaining the sludge bedalmo.st at the s~e level (Table 6). Miron (1997) mentioned that sludge residence time (SRT) plays a crucialrule m anaerobIc treatment and SRT is an efficient parameter in controlling the process. Calculating theSRT from total biomass in the reactor and wasted biomass gave a high SRT (>500 d) for all reactors. Thewasted biomass. mainly contained flocculant sludge while the biomass in the reactors was mainly granularsludge. SeparatIon between SRT of flocculant sludge and granular sludge is more realistic. However,accurate separating and measuring of flocculant sludge and granular sludge was difficult especially in theAH reactors which contained attached biomass to the media.

Table 6. Wasted sludge during the treatment of raw and pre-settled sewage

Time

At the end of raw sewage treatment

After 41 days ofpre-settled sewage treatment

UASB reactor

(g YSS/d)

0.21

0.16

'3-sheets-AH'reactor

(g YSS/d)

0.21

0.15

'2 or 4 sheets-AH' reactor(gYSS/d)

0.22

0.17

Comparison between the treatment of raw sewal'e and pre-settled sewal'e

The treatment of raw sewage in the UASB reactor with small sludge granules under lower upflow velocityand temperature seemed to be not practical due to the sludge bed flotation. The high influent COD.. was thereason for sludge bed flotation. The presence of RPF sheets in the AH reactor prevented sludge bed flotationand only channels and gas pockets were formed in the sludge bed of the AH reactors.

The treatment of pre-settled sewage instead of raw sewage improved CODeol removal efficiency of theUASB reactor by 13% (from 46% to 59%) and this difference was significant at level 2.5%. The '3-sheetsAH' reactor removal efficiency for CODeol as well as CODdis improved significantly (level 10%) by 13%(from 44% to 57%) and 12% (from 36% to 48%) respectively. The reduced COD" levels in the influent ofthe pre-settled sewage are quite likely the cause of the improved CODeol removal. The latter indicates thatpart of the colloids in the effluent at the anaerobic treatment of raw sewage is produced from SS in theinfluent.

Wang (1994) showed that a high loaded UASB (HRT=3 hours) resulted in winter (l2°C) in a COD"removal efficiency of49% and a mean effluent of 110 mg CODsJl. As a result ofthe high upflow velocity inthe subsequent EGSB reactor the total removal efficiency of CODeol amounts to 49%. Combination of a highloaded UASB with a hybrid system could improve both the COD" and the CODeol removal, finally resultingin a COOt removal efficiency of ca 80% at 13°C at a HRT of 11 hours for the two systems together. Theproposed two phase system could be attractive for raw sewage treatment at lowe~ temperatures especially forcommunity on-site. Optimization of the process with respect to HRT and upflow velocity should be

provided.

CONCLUSIONS

The use of small sludge granules and operating the reactors at low upflow velocity improved suspendedCOD removal efficiencies for the UASB reactor. Moreover, the use of sheets in the AH reactorssignificantly increased suspended COD removal efficiencies as compared to the UASB and reached to 87%for pre-settled Sl vage treatment.

The treatment of pre-settled sewage instead of raw sewage in AH reactors significantly increased colloidaland dissolved COD removal efficiencies with 13% and 12% respectively and colloidal COD removalefficiency for the UASB reactor with 13%.


At ‘steady state’ conditions of pre-settled sewage treatment, the AH reactors removed 64% of the total COD which was significantly higher by 4% than the UASB reactor.

ACKNOWLEDGEMENTS

We acknowledge the Egyptian Ministry of Higher Education for the scholarship given for the first author. We are grateful to Recticel, Buren, The Netherlands for providing the polyurethane foam sheet. We are also grateful to: R.E. Roersma, B. Willemsen, D. van Doom, J. van der Laan, H. Danker and Ilse Bennehey for technical support.

REFERENCES

Alphe==, A. (1994). Anaerobic gramJar sludge: characterization and factors affecting its functioning, ph. D. thesis, Wageningen Agricultural University, The Netherlands.

Derycke, D. and Verstraete, W. (1986). Anaerobic treatment of domestic wastewater in a lab and pilot scale polyurethane carrier reMor. In Proc. Anaerobic Treatment a Grown-up Technology, Amsterdam, The Netherlands, 437-450.

Dutch Stand& Normalized Methods (1969). The Netherlands Normalisation Institute, Del& The Netherlands, Genung, R. K., Donaldson. T. L. and Reed, G. D. (1985). Pilot-scale development of anaerobic filter technology for municipal

waskwater treatment. In Proc. Seminar/WorkFhop: Anaerobic Treatment of Sewage, Amherst, Mass., 127-160 Grin P. C., Roersma R. and Lettinga G. (1983). anaerobic treatment of raw sewage at lower temperatures. In proc. European

Symp. Anaerobic Wastewater Treatment, Noordwijkerhouf The Netherlands, 335-347. Grotenhuis, J. T. C., Kissel, J. C., Plugge, C. M., Stams, A. J. M. and Zehnder, A. J. B. (1991). Role of substrate concentration in

particle size distribution of metbanogenic granular sludge in UASB reactors. Wat. Res., 25, 21-27. Hulshdf pal, L., Euler, H., Eitner, A. and Grohganz, T. B. W. (1997). State of tbe art sector review. Anaerobic Treds. WQI

July/August, 31-33. Huysman, P., Meenen, P. van, Assche, P. van and Verstraete, W. (1983). Factors affecting tbe colonization of non porous and

porous packing materials in model upflow methane reactors. Biotech. Lett., S(9), 643-648. Inamori, Y., Iketani, M. and Sudo, R. (1983). Effect of temperature on the performance of anaerobic/aerobic submerged filter

system for domestic sewage treatment. J. Sew. Works Association. 20(10), 10-17. Jirka, A. and Catter (1975). Micro semi-automated analysis of surface and waste waters for chemical oxygen demand. Analytical

Chemistry, 41, 1397-1401. Kobayashi, H. A., Stenstrom, M. K. and Mah, R. A. (1983). Treatment of low strength domestic wastewater using the anaerobic

filter. Wut. Res. 17,903-909. Last, A. R. M. van der and Lettinga, G. (1992). Anaerobic treatment domestic sewage under moderate climatic (Dutch) conditions

using upflow reactors at increased superticial velocities. Wat. Sci. Tech., 25(7), 167-178. Lettinga, G., Roersma, R., Grin, P., Zeeuw, W. de, Hulshoff Pol, L., Velsen, L. van, Hombma, S. and Zeeman, G. (1981).

Anaerobic treatment of sewage and low strength wastewater. In Proc. 2nd. fnt. Symp. on Annerobic Digestion, Travemunde, 271-291

Lettinga, G., Roersma, R. and Grin, P. (1983). Anaerobic treatment of raw domestic sewage at ambient temperatures using a granular bed UASB reactor. Biotechnol. Bioeng., 25. 1701-1723.

Lettinga, G., Man, A. W. A. de, Last, A. R. M. van der, Wiegant, W., Knippenberg, K. van, Frijns, J. and Buuren, J. C. L. van (1993). Anaerobic treatment ofdomestic sewage and wastewater. Wet. Sci. Tech., 27(9), 67-73.

Lettinga, G. (1996). Sustainable integrated biological wastewater treatment. Wat. Sci. Tech., 33(3), 85-98. Man, A. W. A. de, Grin, P. C., Roesma, R., Grolle, K. C. F. and Lettinga, G. (1986). Anaerobic treatment of sewage at low

temperatures. In Proc, Anaerobic Treatment a Grown-up Technology, Amsterdam, The Netherlands, 451-466. Man, A.W.A. de, Rijs, G. B. J., Lettinga, G. and Starkenburg, W. (1988). Anaerobic treatment of sewage using a granular sludge

bed UASB reactor. In Proc. 5th. Int. Symp. on Anaerobic Digestion. Bologna, Idly, 735-738. Mergaert, K., Vanderhaegen, B. and Verstraete, W. (1992). Application and trends of anaerobic pre-treatment of municipal

wastewater. Wat. Res., 26, 1025-1033. Miron, Y. (1997). Anaerobic treatment of concentrated sewage in a two step UASR-UASB system: The effect of SRT, sludge

characteristics and t&low velociry on the pe&rmance of the UASR, Report 97-5 1, Wageningen Agriculm1 University, The Netherlands.

Momoy, O., Noyola, A., Raminez, F. and Guiot, J. P. (1988). Anaerobic digestion of water hyacinth as a highly efficient treatment process for developing countries. In Proc. 5th. fnt. Symp. on Anaerobic Digestion. Bologna, Italy, 747-757.

Noyola, A., Capdeville, B. and Roqnes (1988). Anaerobic treatment of domestic wastewater with a rotating-stationary fixed-film reactor. Wat. Res., 22(12), 1585-1592.

Polprasert, C., Kemmadamnrong, P. and Tran, F. T. (1992). Anaerobic baffle reactor (ABR) process for treating slaughterhouse wastewater. Environm. Tech., 13, 857-865.

R&c, s., Lier, J. B. van, Janssen, M. G. J., Dekkers, F., Swinkels, K. T. M and Lettinga, G. (1997), High-rate anaerobic

treatment of mlthg wastewater in a pilot-scale EGSB system under psychrophilic conditions. J. Chem. Tech.

Biotechnol. 68, 135-146. Sanz, I, and F&-p0 anco, F. (1990). Low temperature treatment of municipal sewage in anaerobic fluidtied bed reactors. #‘at. f

Res., 24,463-469.

184 T. A. ELMITWALL!

At 'steady state' conditions ofpre-settled sewage treatment, the AH reactors removed 64% of the total CODwhich was significantly higher by 4% than the DASB reactor.

ACKNOWLEDGEMENTS

We acknowledge the Egyptian Ministry of Higher Education for the scholarship given for the first author.We are grateful to Recticel, Buren, The Netherlands for providing the polyurethane foam sheet. We are alsograteful to: R.E. Roersma, B. Willemsen, D. van Doorn, J. van der Laan, H. Donker and IIse Bennehey fortechnical support.

REFERENCES

Alphenaar, A. (1994). Anaerobic granular sludge: characterization and factors affecting its functioning. Ph. D. thesis,Wageningen Agricultural University, The Netherlands.

Derycke, D. and Verstraete, W. (1986). Anaerobic treatment of domestic wastewater in a lab and pilot scale polyurethane carrierreactor. In Proc. Anaerobic Treatment a Grown-up Technology, Amsterdam, The Netherlands, 437-450.

Dutch Standard Normalized Methods (1969). The Netherlands Nonnalisation Institute, Delft, The Netherlands.Genung, R. K., Donaldson, T. 1. and Reed, G. D. (1985). Pilot-scale development of anaerobic filter technology for municipal

wastewater treatment. In Proc. Seminar/Workshop: Anaerobic Treatment ofSewage, Amberst, Mass., 127-160Grin P. C., Roersma R. and Lettinga G. (1983). anaerobic treatment of raw sewage at lower temperatures. In Proc. European

Symp. Anaerobic Wastewater Treatment, Noordwijkerhout, The Netherlands, 335-347.Grotenhuis,1. T. C., Kissel, J. c., Plugge, C. M., Stams, A. J. M. and Zehnder, A. J. B. (1991). Role of substrate concentration in

particle size distribution ofmethanogenic granular sludge in UASB reactors. Wat. Res., 25, 21-27.Hulshoff Pol, L., Euler, H., Eitner, A. and Grohganz, T. B. W. (1997). State of the art sector review. Anaerobic Trends. WQI

July/August, 31-33.Huysman, P., Meenen, P. van, Assche, P. van and Verstraete, W. (1983). Factors affecting the colonization of non porous and

porous packing materials in model upflow methane reactors. Biotech. Lett., 5(9), 643-648.Inamori, Y., Iketani, M. and Sudo, R. (1983). Effect of temperature on the perfonnance of anaerobic/aerobic submerged filter

system for domestic sewage treatment. J. Sew. Works Association. 20(10),10-17.Jirka, A. and Carter (1975). Micro semi-automated analysis of surface and waste waters for chemical oxygen demand. Analytical

Chemistry, 47, 1397-1401.Kobayashi, H. A., Stenstrom, M. K. and Mah, R. A. (1983). Treatment of low strength domestic wastewater using the anaerobic

filter. Wat. Res. 17,903-909.Last, A. R. M. van der and Lettinga, G. (1992). Anaerobic treatment domestic sewage under moderate climatic (Dutch) conditions

using upflow reactors at increased superficial velocities. Wat. Sci. Tech., 25(7),167-178.Lettinga, G., Roersma, R., Grin, P., Zeeuw, W. de, Hulshoff Pol, L., Velsen, L. van, lfombma, S. and Zeeman, G. (1981).

Anaerobic treatment of sewage and low strength wastewater. In Proc. 2nd. Int. Symp. on Anaerobic Digestion,Travemunde,271-291

Lettinga, G., Roersma, R. and Grin, P. (1983). Anaerobic treatment of raw domestic sewage at ambient temperatures using agranular bed UASB reactor. Biotechnol. Bioeng., 25.1701-1723.

Lettinga, G., Man, A. W. A. de, Last, A. R. M. van der, Wiegant, W., Knippenberg, K. van, Frijns, 1. and Buuren, J. C. 1. van(1993). Anaerobic treatment ofdomestic sewage and wastewater. Wat. Sci. Tech., 27(9), 67-73.

Lettinga, G. (1996). Sustainable integrated biological wastewater treatment. Wat. Sci. Tech., 33(3), 85-98.Man, A. W. A. de, Grin, P. C., Roesma, R., Grolle, K. C. F. and Lettinga, G. (1986). Anaerobic treatment of sewage at low

temperatures. In Proc. Anaerobic Treatment a Grown-up Technology, Amsterdam, The Netherlands, 451-466.Man, A.W.A. de, Rijs, G. B. J., Lettinga, G. and Starkenburg, W. (1988). Anaerobic treatment of sewage using a granular sludge

bed UASB reactor. In Proc. 5th. Int. Symp. on Anaerobic Digestion. Bologna, Idy, 735-738.Mergaert, K., Vanderhaegen, B. and Verstraete, W. (1992). Application and trends of anaerobic pre-treatment of municipal

wastewater. Wat. Res., 26,1025-1033.Miron, Y. (1997). Anaerobic treatment of concentrated sewage in a two step UASR-UASB system: The effect of SRT, sludge

characteristics and upflow velocity on the performance ofthe UASR, Report 97-51, Wageningen Agricultural University,

The Netherlands.Monroy, 0., Noyola, A., Rarninez, F. and Guiot, J. P. (1988). Anaerobic digestion of water hyacinth as a highly efficient treatment

process for developing countries. In Proc. 5th. Int. Symp. on Anaerobic Digestion. Bologna, Italy, 747-757.Noyola, A., Capdeville, B. and Roques (1988). Anaerobic treatment of domestic wastewater with a rotating-stationary fixed-film

reactor. Wat. Res., 22(12), 1585-1592.Polprasert, C., Kemrnadanmrong, P. and Tran, F. T. (1992). Anaerobic baffle reactor (ABR) process for treating slaughterhouse

wastewater. Environm. Tech., 13, 857-865.Rebac, S., Lier, J. B. van, Janssen, M. G. J., Dekkers, F., Swinkels, K. T. M and Lettinga, G. (1997), High-rate anaerobic

treatment of malting wastewater in a pilot-scale EGSB system under psychrophilic conditions. J. Chem. Tech.

Biotechnol. 68, 135-146.Sanz, I. and Fdz.Potanco, F. (1990). Low temperature treatment of municipal sewage in anaerobic fluidized bed reactors. Wat.

Res., 24, 463-469.


Sayed, S. K. I. and Fergala, M. A. A. (1995). Two-stage UASB concept for treatment of domestic sewage including sludge stabilization process. War. Sci. Tech., 32( 1 1), 55-63.

Snedecor, G. W. and Cochran, W. G. (1980). Statistical methods. The Iowa state university press, Ames, Iowa, USA. Tang, N. N., Torres, C. L. and Speece, R. E. (1995). Treatment of low strength domestic wastewater by using upflow anaerobic

sludge blanket process. In Proc. 50th. Purdue Industrial Waste Co@, 437-448. Tilche, A. and Vieira, S. M. M. (1991). Discussion report on reactor design of anaerobic filters and sludge bed reactors. Wat. Sci.

Tech., 24(8), 193-206. Vieira, S. M. M. and Souza, M. E. (1986). Development of the technology for the use of the UASB reactor in domestic sewage

treatment. Wat. Sci. Tech., 18(12), 109-121. Wang, K. (1994). Integrated Anaerobic and Aerobic Treatment ofSewage. Ph. D. thesis, Wageningen Agricultural University,

The Netherlands. Yoda, M., Hattori, M. and Miyaji, Y. (1985). Treatment of municipal wastewater by anaerobic fluidized bed: behaviour of organic

suspended solids in anaerobic treatment of sewage. In Proc. Seminar/Workshop: Anaerobic Treatment of Sewage, Amherst, Mass., 161-197.


Sayed, S. K. I. and Fergala, M. A. A. (1995). Two-stage UASB concept for treatment of domestic sewage including sludgestabilization process. Wat. Sci. Tech., 32(11), 55-63.

Snedecor, G. W. and Cochran, W. G. (1980). Statistical methods. The Iowa state university press, Ames, Iowa, USA.Tang, N. N., Torres, C. L. and Speece, R. E. (1995). Treatment of low strength domestic wastewater by using upflow anaerobic

sludge blanket process. In Proc. 50th. Purdue Industrial Waste Conf, 437-448.Tilche, A. and Vieira, S. M. M. (1991). Discussion report on reactor design of anaerobic filters and sludge bed reactors. Wat. Sci.

Tech., 24(8), 193-206.Vieira, S. M. M. and Souza, M. E. (1986). Development of the technology for the use of the UASB reactor in domestic sewage

treatment. Wat. Sci. Tech., 18(12),109-121.Wang, K. (1994). Integrated Anaerobic and Aerobic Treatment ofSewage. Ph. D. thesis, Wageningen Agricultural University,

The Netherlands.Yoda, M., Hattori, M. and Miyaji, Y. (1985). Treatment of municipal wastewater by anaerobic fluidized bed: behaviour of organic

suspended solids in anaerobic treatment of sewage. In Proc. Seminar/Workshop: Anaerobic Treatment of Sewage,Amherst, Mass., 161-197.

low temperature treatment of domestic sewage in upflow anaerobic sludge blanket and anaerobic hybrid...

Documents