Process Biochemistry 40 (2005) 12251232

Effect of soluble organic, particulate organic, and hydraulic shock loads on anaerobic sequencing batch reactors treating slaughterhouse wastewater at 20 C
L. Masse , D.I. Mass
Agriculture and Agri-Food Canada, P.O. Box 90, 2000 Route 108, East Lennoxville, Que., Canada J1M 1Z3 Received 1 September 2003; accepted 18 April 2004

Abstract Anaerobic sequencing batch reactors (ASBRs) treating slaughterhouse wastewater at 20 C were subjected to soluble organic, particulate organic and hydraulic shock loads. The normal organic loading rate (OLR) was increased by a factor ranging from 1.5 to 3.6. During the soluble organic shock load, efuent soluble chemical oxygen demand (SCOD), volatile fatty acid (VFA), and suspended solids (SS) concentrations increased from averages of 200, 58 and 249 mg/l, respectively, in control reactors to maxima of 5626, 1642 and 10 723 mg/l, respectively, in overloaded reactors. However, concentrations were rapidly reduced to pre-shock load levels shortly after normal operating conditions were resumed. The particulate and hydraulic shock loads had small and temporary effects on efuent SCOD and VFA concentrations, but high SS concentrations were observed in the efuent of most overloaded reactors. However, solids loss during efuent discharge did not signicantly affect long-term performance of the ASBRs, probably because the sludge bed contained high initial volatile SS (VSS) concentration (21.9 3.2 g/l), and also a fraction of the VSS removed from the reactors during efuent discharge was composed of undegraded colloidal solids fed with the substrate, as opposed to bacterial ocs. Nevertheless, maintaining high VSS concentrations in the sludge bed will probably prevent long-term negative effect of biomass loss during sudden shock loads. 2004 Published by Elsevier Ltd.
Keywords: Anaerobic sequencing batch reactor; Overload; Shock load; Slaughterhouse wastewater

1. Introduction Anaerobic digestion in high-rate reactors represents an attractive alternative for wastewater treatment at slaughterhouse plants. Slaughterhouse wastewater contains high concentrations of biodegradable organics, mostly fats and proteins, and adequate nutrient concentrations for bacterial growth. It is well buffered and exits the plant at a relatively warm temperature, typically ranging from 20 to 30 C [1]. The anaerobic sequencing batch reactor (ASBR) is particularly well suited for wastewater treatment at slaughterhouses, because it entails low capital and operating costs as well as minimum daily maintenance. It can be fed on an intermittent basis, as the wastewater is produced, thus eliminating the need for an equalizing tank or recycling line. The ASBR process is efcient with efuent contain

Corresponding author. Fax: +1-819-565-9171. E-mail address: massel@agr.gc.ca (L. Masse).

ing high suspended solids (SS) content, such as slaughterhouse wastewater, because it allows prolonged contact between the biomass and the organics. Laboratory ASBRs (42 l) treating hog slaughterhouse wastewater could sustain steady state organic loading rates (OLRs) of 4.93, 2.94 and 2.75 g/l/day at operating temperature of 20, 25 and 30 C, respectively [2]. At all temperatures, the soluble chemical oxygen demand and total chemical oxygen demands (SCOD) and (TCOD) were consistently reduced by more than 92%. At the slaughterhouse, the ASBRs could be subjected to sudden shock load conditions. A blood spill could translate into a large soluble overload. The COD content of blood, one of the major dissolved pollutants of slaughterhouse wastewater, was evaluated at 375 g/l [3]. The reactors could also be fed an excess of particulate organics, which would require extended hydraulic retention time (HRT) to process, while a temporary increase in wastewater volume or the sudden break down of one of the reactors may impose an hydraulic overload on the biological process.

0032-9592/$ see front matter 2004 Published by Elsevier Ltd. doi:10.1016/j.procbio.2004.04.012

1226

L. Masse, D.I. Mass e / Process Biochemistry 40 (2005) 12251232

Borja et al. [4] subjected a mesophilic anaerobic lter reactor (AFR) treating slaughterhouse wastewater to temporary organic and hydraulic shock loads representing 22.5 times the steady state load. Efuent volatile fatty acid (VFA) and COD concentrations increased during the overload period, but concentrations returned to normal values within 12 h of steady state operation. The efuent contained high protein and VFA concentrations, and the authors suggested that hydrolysis and acidication were the limiting processes. The AFRs are particularly robust to soluble shock loads, because biomass immobilization on support reduces the risk of biomass washout. With the ASBR process, on the other hand, biomass retention depends on an adequate settling of the microorganisms before efuent removal. Under shock load conditions, high biogas production rate could create turbulence during the settling phase, thereby preventing complete biomass sedimentation. Most studies on anaerobic reactor stability were conducted with mesophilic or thermophilic systems. Psychrophilic reactors are considered more stable than thermophilic systems [5], but biomass loss could be detrimental at lower operating temperatures, because of slow bacterial growth. The objective of this study was to evaluate the performance and stability of ASBRs treating hog slaughterhouse wastewater at 20 C under sudden and temporary organic and hydraulic shock load conditions.

to remove hair and large particles, transferred to 10 l jugs, and stored at 2 C. The wastewater was allowed to reach room temperature before reactor feeding. Wastewater characteristics during the experimental period are presented in Table 1. During the soluble shock load experiment, wastewater SCOD was increased by adding 1.1 l of pure uncoagulated blood to 11.1 l of slaughterhouse wastewater. The uncoagulated blood had a COD of 140 6 g/l (the coagulated fraction had been removed). During the particulate organic shock load, the suspended solids concentration was increased by removing about 75% of the supernatant of settled slaughterhouse wastewater before feeding the experimental ASBRs. Wastewater SCOD was similar to that in control reactors, but the particulate COD (PCOD) was substantially increased (Table 1). During the hydraulic shock load, all reactors were fed the same wastewater, but the HRT of the experimental ASBRs was decreased from 2 to 1 day. 2.2. Experimental set-up A schematic of the four 42 l ASBRs used for the experiment was presented in Mass and Masse [2]. Before the initiation of this study, the ASBRs were fed hog slaughterhouse wastewater for 18 months at 20 C. At the start of the experiment, the sludge was removed from all ASBRs, mixed in Nalgene bottles under anaerobic conditions, and redistributed to the bioreactors. The four ASBRs were then operated at 20 C under normal conditions, dened as an OLR of 2.60 0.36 g/l/day and an HRT of 2 days (Table 2). Throughout the experiment, the OLR was based on the amount of TCOD fed per volume of sludge present at the start of a cycle per total cycle time. The ASBRs were batch-fed over a 1 h period. During the react phase, ASBR content was mixed for 1 min every 30 min

2. Materials and method 2.1. Substrate Raw wastewater was collected once a month at a hog slaughterhouse in St-Valrien, Qubec. In the laboratory, the wastewater was mixed, screened through a 1 mm sieve

Table 1 Characteristics of the slaughterhouse wastewater fed to control ASBRs (normal operating conditions) and during each shock load experiment Parameters (mg/l except pH) COD total Soluble particulate SS total Volatile VFA total TKN NH4 Total protein Soluble protein Total fat Alkalinity pH Normal operating conditionsa 7083 1211 3623 752 3460 840 1592 321 1418 298 467 108 547 174 2257 1531 168 78 59 492 465 46 Soluble organic shock load 19367 2821 15252 3420 4385 1347 1845 281 1709 262 793 129 1887 318 9806 8885 379 NA NA 229 87 1060 943 52 Particulate organic shock load 13595 3057 3041 902 10554 3724 6883 3459 5560 2496 435 103 NA NA NA NA 317 4 667b 6.88b Hydraulic shock load 7924 584 4189 263 3742 621 1769 217 1614 194 508 10 525 121 2527 2107 227 61 50 121 129 80

903 226 6.85 0.44

958 59 6.89 21

NA: not analysed a Average over the 7 months experimental period. b Only one sample analysed.

L. Masse, D.I. Mass e / Process Biochemistry 40 (2005) 12251232 Table 2 Experimental conditions during soluble, particulate and hydraulic shock loads in ASBRs treating slaughterhouse wastewater at 20 C Type of shock load Soluble S-I 26.9 0.8 2 1 1 1 3 2 3 5 Experiment Initial VSS in sludge beda (g/l) Number of consecutive cycles Operating conditionb Sludge mixing and redistribution Normal Shock load (4.22 g/l/day) Normal Shock load (7.93 g/l/day) Normal Sludge mixing and redistribution

1227

S-II

23.3 0.6

Normal Shock load (8.95, 7.55 and 9.32 g/l/day) Normal Sludge mixing and redistribution

Particulate P-I 16.1 0.4 2 1 2 2 1 3 2 2 4 3 4 4 2 4 4

Normal Shock load (3.92 g/l/day) Normal Sludge mixing and redistribution Normal Shock load (6.25 g/l/day) Normal Sludge mixing and redistribution Normal Shock load (4.14 and 5.76 g/l/day) Normal Sludge mixing and redistribution Normal Shock load (5.29 g/l/day) Normal Sludge mixing and redistribution Normal Shock load (5.40 g/l/day) Normal

P-II

20.8 0.6

P-III

16.0 0.6

Hydraulic H-I 22.7 1.2

H-II

20.1 0.6

a b

VSS concentration in sludge bed after the sludge has been mixed and redistributed to the four ASBRs. During normal operating conditions, the ORL and HRT were 2.60 0.36 g/l/day and 2 days, respectively.

by biogas recirculation with dual-head air pumps. At the end of cycles, the efuent was removed from the bioreactors over a 1 h period, leaving14 l of sludge in the ASBR to start the next cycle. During each experiment, two ASBRs received target shock loads while two control reactors were fed at normal OLRs (Table 2). After each shock load, all reactors were operated under normal conditions for a few cycles to monitor long-term effects of the shock load. Between each experiment, the sludge was removed from all ASBRs, mixed under anaerobic conditions, and redistributed to the bioreactors (Table 2). This insured that the four ASBRs had similar volatile SS (VSS) concentration and biomass activity rate at the beginning of an experiment. 2.3. Sampling and data analysis Biogas production was monitored twice daily with wet cup gas meters. Biogas composition (methane, and carbon dioxide) was determined in samples collected 24 h after feeding.

Approximately 250 ml of slaughterhouse wastewater was collected before each ASBR cycle. Reactor efuent was thoroughly mixed in a large vessel, before a 250 ml sample was collected. During some ASBR cycles, mixed-liquor samples (100 ml) were collected from each replicate reactor at various times during the react phase. Sludge samples (100 ml) were collected from the 14 l sludge bed before and after each shock load experiment. Sample analyses included SCOD, TCOD, SS, VSS, NH4 , total Kjeldahl nitrogen (TKN), fat, VFA (acetic, propionic, butyric, valeric, isovaleric and caproic acids), pH and alkalinity. All analytical methods have been described in Mass and Masse [2]. Particulate COD (PCOD) was calculated by subtracting SCOD from TCOD values. Protein concentration was calculated by multiplying the difference between the TKN and NH4 N values by 6.25 [6]. During the particulate shock load experiment, the amount of SS transformed into methane during each cycle was calculated as follows: SSCH4 = PCODCH4 /CODSS (1)

1228

L. Masse, D.I. Mass e / Process Biochemistry 40 (2005) 12251232

Experiment SI 10 Experiment SII

CH4 (l/g TCODfed)

where SSCH4 is the amount of SS transformed into methane (g); CODSS is the COD equivalent of the SS in the inuent (g/g), calculated by dividing PCOD by SS concentration in the inuent sample; and PCODCH4 is the PCOD transformed into methane (g). The PCODCH4 was calculated by subtracting methane production attributed to SCOD degradation from total methane production, as follows: PCODCH4 = CODCH4 [SCODin SCODeff ] (2)

OLR (g/l/d)

(a)

(b)
0.4 0.2 0.0 6000 4000 2000 0

where CODCH4 is total methane production during the cycle in g of COD, using the conversion factor of 0.38 l of methane per g of COD at 20 C; SCODin and SCODeff are SCOD in inuent and efuent, respectively (g). Signicant differences (P < 0.05) between control and experimental reactors during each cycle were estimated using the GLM subroutine in SPSS 9.0 for Windows. During the hydraulic shock load experiment, the experimental ASBRs were fed and emptied daily, while the control reactors were maintained on a 2-day cycle. For statistical analysis purposes, efuent characteristic was compared only on days when efuent was removed from both control and experimental ASBRs, while results for all cycles are presented on the graphs.

SCOD (mg/l)

(c)

12000

(d)

SS (mg/l) VFA (mg/l)

8000 4000 0 1500 1000 500 0 0 Control 10 20 time (d) 30 40 50 Soluble shockload

(e)

3. Results and discussion 3.1. Performance of control ASBRs The control ASBRs produced 0.292 0.050 l CH4 /g TCODfed . Based on a maximum theoretical yield of 0.38 l CH4 /g TCOD degraded at 20 C, an average of 76% of the TCOD fed to the ASBRs was transformed into methane during cycles. Efuent TCOD, SCOD and SS concentrations averaged 561 85, 210 38 and 234 70 mg/l, respectively, representing a reduction of 93.9 1.8, 91.8 1.5 and 84.1 6.1% in the three parameters, respectively. Biomass synthesis and solids accumulation in the bioreactor probably accounted for most of the difference between TCOD reduction and transformation into methane. 3.2. Soluble organic shock load Two soluble shock load experiments (S-I and S-II) were conducted (Table 2, Fig. 1). During experiment S-I, the ASBRs received two individual shock loads separated by one cycle at normal operating conditions (Table 2). The rst soluble shock load (4.22 g/l/day on day 8) had minimal effect on efuent quality, but specic methane yield decreased signicantly to 0.247 l/g TCODfed (Fig. 1be). The second shock load (7.93 g/l/day on day 12) caused a signicant (P < 0.05) increase in efuent SCOD, SS and VFA concentrations, while specic methane yield decreased to 0.211 l/g TCODfed . However, efuent quality and specic methane yield returned to pre-shock load levels within one cycle at normal OLR. The high initial VSS concentration

Fig. 1. OLR, methane yield, and efuent characteristics during soluble organic shock loads in ASBRs treating slaughterhouse wastewater at 20 C. The empty () and full () square marks indicate shock load and normal cycles, respectively.

in the reactors (Table 2) may have mitigated potential adverse effects of biomass loss during the second shock load cycle. During experiment S-II, the ASBRs received three consecutive soluble shock loads (days 33, 35 and 37) at OLRs of 8.95, 7.55 and 9.32 g/l/day, respectively (Fig. 1a). Efuent SCOD and VFA concentrations increased to maximum values of 5626 and 1642 mg/l, respectively, at the end of the third shock load cycle (Fig. 1c and e). Although residual SCOD and VFA concentrations decreased rapidly when normal operating conditions were resumed, they remained signicantly (P < 0.05) higher in the efuent of experimental than control ASBRs for four and two cycles, respectively, after the last shock load. Specic methane yield decreased signicantly during the three shock load cycles, but tended to be higher in the experimental than control ASBRs when normal operating conditions were resumed (Fig. 1b). Increased methane yield was probably due to the digestion of residual COD in the experimental ASBRs. Efuent SS concentration was signicantly higher in experimental than control ASBRs during shock loads 1 and 3, and was similar

L. Masse, D.I. Mass e / Process Biochemistry 40 (2005) 12251232

1229

in all ASBRs in subsequent cycle (Fig. 1d). VSS concentration in the sludge bed decreased from an initial value of 23.3 to 21.2 g/l at the end of experiment S-II. The rapid recovery of the ASBRs following the soluble shock loads may be partly attributed to an adapted and active biomass that efciently degraded the substrate throughout the experiment. Additionally, a fraction of the VSS removed during efuent discharge was probably composed of undegraded colloidal solids from the substrate, as opposed to bacterial ocs. Finally, the operational mode of the ASBR also favours rapid recovery of the system following soluble shock loads. At the end of a cycle, a settled efuent high in soluble organics is removed from the bioreactor, and is replaced with a wastewater sometimes lower in organic substrate. For example, at the end of the three consecutive shock loads of experiment S-II, 12 l of efuent having a SCOD of 5626 mg/l was removed from the overloaded reactors and replaced with 18 l of fresh slaughterhouse wastewater with a SCOD of 2603 mg/l. The resulting mixed-liquor SCOD after feeding was 3926 mg/l, which represented a 30% SCOD reduction in the reactor. Fig. 2 presents acetic, propionic, and isovaleric acid concentrations in the mixed-liquor of overloaded and control reactors during the three consecutive shock loads. In control reactors, the acetic acid concentration was reduced by 80% within the initial 24 h of treatment, and by 84% at the end of the 2 days cycles. In overloaded reactors, the acetic acid concentration increased in the initial 6 h, suggesting rapid acidication of the soluble organics. Although concentration decreased during the remaining 18 h of treatment, acetic acid tended to accumulate in the reactor with each new shock load cycle. After three shock load cycles, the acetic acid concentration averaged 502 mg/l in the overloaded reactors, while the raw wastewater contained 345 mg/l.
Shock load cycle 1 800 600 400 200 0 800 600 400 200 0 800 600 400 200 0 31 (a) Acetic acid

The shock loads had little effect on butyric acid in the reactor, but propionic and isovaleric acid concentrations increased during each shock load cycle and tended to accumulate in the reactors (Fig. 2b and c). In efuent, isovaleric acid concentration remained signicantly (P < 0.05) higher in overloaded than control reactors for three cycles following the last shock load. Propionic and isovaleric acids were found to be the most sensitive indicators of reactor instability during start up of ASBRs treating slaughterhouse wastewater [7]. Soluble protein concentration averaged 8885 mg/l in the concentrated wastewater and 1188 mg/l in the control substrate. At the end of shock load cycles, soluble protein concentration was similar in the efuent of control (80 48 mg/l) and overloaded (103 47 mg/l) ASBRs, suggesting that the microorganisms responsible for hydrolysis and acidication of the soluble protein were not signicantly affected by the shock loads. The VFAs, on the other hand, accounted for 4050% of efuent SCOD in the shock load reactors. VFA reduction and methanization would thus seem to have been the rate limiting process during the soluble organic shock loads. Ammonianitrogen averaged 418 45 and 1263 214 mg/l in the efuent from control and overloaded reactors, respectively. Additional ammonianitrogen was probably produced during the mineralization of the high soluble protein substrate in the shock load reactors. A sudden increase in ammonianitrogen could be detrimental to the biomass, but efuent removal at the end of a cycle and the input of fresh substrate with a NH4 N concentration of less than 200 mg/l would help diluting the high concentrations at the start of the next cycle. Efuent alkalinity also increased signicantly (P < 0.05) from an average of 2151 mg/l in control to 4600 mg/l in
Shock load cycle 3

Shock load cycle 2

mg/l

(b)

Propionic acid

mg/l

mg/l

(c)

Isovaleric acid

32

33

34

Control reactors

35 36 37 time (d) Experimental reactors

38

Fig. 2. Concentration of acetic, propionic and isovaleric acid in the mixed-liquor during the three consecutive shock load cycles of experiment SII in ASBRs treating slaughterhouse wastewater at 20 C. The arrows indicate ASBR feeding.

1230

L. Masse, D.I. Mass e / Process Biochemistry 40 (2005) 12251232

Experiment PI 8

Experiment PII

Experiment PIII

OLR (g/l/d)

6 4 2 0
0.5

(a)

a release of CO2 , which may explain the lower methane content in overloaded ASBRs. Biogas quality returned to pre-shock load levels as soon as normal operating conditions were resumed. 3.3. Particulate organic shock load

(b)

0.4 0.3 0.2 0.1 0.0

400

(c)

300 200 100 0 3000

(d)

2250 1500 750 0 0 10 20 time (d) Particulate shock load Control 30 40

Fig. 3. OLR, methane yield, and efuent characteristics during particulate organic shock loads in ASBRs treating slaughterhouse wastewater at 20 C. The empty () and full () square marks indicate shock load and normal cycles, respectively.

overloaded ASBRs, probably due to the formation of ammonium bicarbonates through the reaction of ammonia with the carbonate system during protein mineralization [8]. Efuent pH was also slightly higher in overloaded (7.4) than control (7.2) reactors during the three consecutive shock loads. In biogas samples collected 24 h after feeding, methane content was signicantly (P < 0.05) higher in control (78.4 1.2%) than overloaded (72.9 2.8%) ASBRs. Higher acidication rate in the overloaded reactor probably caused

Three particulate shock load experiments (P-I to P-III) were conducted (Table 2, Fig. 3). The ASBRs received single particulate shock loads of 3.92 and 6.25 g/l/day (days 2 and 16) during experiments P-I and P-II, respectively, and two consecutive particulate shock loads of 4.14 and 5.76 g/l/day, respectively (days 33 and 35) during experiment P-III. During all three experiments, SCOD concentration (Fig. 3c), pH, alkalinity, VFA and ammonianitrogen concentrations were similar in the efuent from overloaded and control reactors. Average SS concentration in efuent increased during all shock load cycles, and average concentration remained higher in overloaded than control reactors for one cycle at normal operating conditions (Fig. 3d). However, for two of the particulate shock load cycles, high SS concentration was observed in only one of the two overloaded reactors. In these reactors, methanization of undigested particles trapped in the sludge bed during the settling period probably caused the release of methane bubbles which temporarily resuspended the solids just before efuent removal. The efuent from overloaded reactors containing low SS concentrations may have been removed prior to such event. Since SCOD concentration was low in all reactors, methanization of the soluble organics cannot account for the turbulence during the settling period. Solids mass balance in the reactors during experiments P-I to P-III is presented in Table 3. Calculated SS accumulation in the sludge bed zone of the control reactors during each experiment ranged from 19 to 38 g, corresponding to a net solid accumulation of 0.0660.174 g SS/g CODremoved . After the two rst particulate shock loads (P-I and P-II), SS accumulation was lower in overloaded than control reactors, due to the large SS loss during efuent removal (Table 3). During experiment P-III, however, large SS accumulation was observed in overloaded reactors. Consequently, SS concentration was higher in the sludge bed zone of overloaded

Table 3 Suspended solids mass balance in ASBRs under particulate shock loads and control ASBRs Experiment SS input with substrate (g) Shock load P-I P-II P-III
a b c d

SS (mg/l)

SCOD (mg/l)

l CH4/g TCODfed

SS loss in efuent (g) Control 90 89 89 Shock load 38 52 49 Control 11 11 14

SS transformed into methane (g)a Shock load 73 70 73 Control 60 55 37

Calculated SS accumulation in the ASBRs (g)b Shock load 1.1 13 117 Control 19 23 38

Observed SS accumulation in the ASBRs (g)c Shock load 13 NDd 92 Control 13 NDd 31

113 135 240

Calculated as outlined in Section 2. SSinput (SSlossinefuent + SSmethane ). Calculated from SS in sludge bed at the beginning and the end of each particulate shock load experiment. Not determined.

L. Masse, D.I. Mass e / Process Biochemistry 40 (2005) 12251232

1231

(29.0 1.0 g/l) than control (24.9 1.1 g/l) reactors at the end of the experimental period, despite high SS losses from the overloaded ASBRs. Accumulated SS may have been mainly undegraded solids from the substrate. Biogas methane content was similar in overloaded and control reactors. Specic methane yield in overloaded reactors ranged from 0.227 to 0.121 l CH4 /g TCODfed (Fig. 3b), and was signicantly (P < 0.05) lower than in control reactors. However, specic methane yield tended to be higher in overloaded than control reactors in the two cycles following the shock loads, probably due to the digestion of accumulated organic particles. 3.4. Hydraulic organic shock load Hydraulic shock loads were created by decreasing the HRT of overloaded reactors from 2 to 1 day for four consecutive cycles, thereby increasing the OLR to 5.29 and 5.40 g/l/day in experiments H-I and H-II, respectively (Table 2). Efuent SCOD ranged from 274 to 773 mg/l for the overloaded reactors, and was generally signicantly (P < 0.05) higher than for the control ASBRs (Fig. 4c). Efuent VFA concentration ranged from 37 to 338 mg/l in overloaded reactors, and was also signicantly (P < 0.05) higher

Experiment SI

Experiment SII

OLR (g/l/d)

(a)
5 4 3 2 1

than in the efuent from the control reactors (Fig. 4e). The efuent contained high SS concentrations, ranging from 407 to 5501 mg/l (Fig. 4d). Consequently, VSS concentration in the sludge bed of overloaded reactors decreased from 22.7 1.2 to 18.6 1.2 g/l during experiment H-I, and from 20.1 0.6 to 17.2 0.7 g/l during experiment H-II. However, VSS loss did not seem to affect subsequent operation of the ASBRs since efuent quality returned to pre-shock load levels within one cycle at regular operating conditions after both hydraulic shock load experiments. Again, the high initial VSS concentration in the ASBRs may have allowed some biomass loss without signicant decrease in activity. Specic methane yield averaged 0.201 0.016 and 0.281 0.281 l/g TCODfed in overloaded and control reactors, respectively, corresponding to the transformation of 53 and 74%, respectively, of input TCOD into methane gas (Fig. 4b). Difference was partly due to a lower SCOD reduction in overloaded (88 4%) than control (94 2%) ASBRs. However, the lower methane yield from reactors under hydraulic shock load may be mainly due to lower hydrolysis of particulate COD during the short HRT. A methane yield of 0.400 l/g in the rst cycle at normal operating conditions following the second hydraulic shock load (H-II) probably indicated the methanization of accumulated solids. However, an increase in methane yield was not observed after the rst hydraulic shock load (H-II), suggesting that a large part of the undegraded SS from the substrate were removed during efuent discharge. Colloidal solids from the substrate would not settle as well as bacterial ocs. There was no shock load effect on biogas methane content.

l CH4/g TCODfed

(b)
0.4 0.3 0.2 0.1 800 600 400 200 0

4. Conclusion Soluble organic, particulate organic and hydraulic shock loads, which increased the normal OLR by a factor ranging from 1.5 to 3.6, were given to ASBRs treating slaughterhouse wastewater at 20 C. During the soluble organic shock load, SCOD and VFA concentrations tended to accumulate in the reactors, but were rapidly reduced to pre-shock load levels when normal operating conditions were resumed. Increase in mixed-liquor alkalinity prevented a large decrease in pH due to an increase in VFA production during shock loads. Particulate organic and hydraulic shock loads had small and temporary effects on SCOD and VFA concentrations. However, during most shock loads, high SS concentrations were observed in the efuent. Solids loss did not signicantly affect long-term performance of the ASBRs, probably because the sludge bed had high initial VSS concentration (average of 21.9 3.2 g/l). Moreover, a fraction of the VSS lost in the efuent was probably undegraded colloidal solids from the slaughterhouse wastewater as opposed to biomass. Nevertheless, maintaining a high VSS concentration in the sludge bed will probably prevent long-term negative effect of biomass loss during sudden shock loads.

SCOD (mg/l)

(c)

SS (mg/l)

4500 3000 1500 0 400 300 200 100 0 0

(d)

VFA (mg/l)

(e)

10 time (d) Control

20

30 Hydraulic shock load

4 0

Fig. 4. OLR, methane yield, and efuent characteristics during hydraulic organic shock loads in ASBRs treating slaughterhouse wastewater at 20 C. The empty () and full () square marks indicate shock load and normal cycles, respectively.

1232

L. Masse, D.I. Mass e / Process Biochemistry 40 (2005) 12251232 transient changes in process parameters. Biotechnol Appl Biochem 1994;20:37183. Mass, DI. Psychrophilic anaerobic digestion of swine manure slurry in intermittently fed sequencing batch reactors. Ph.D. Thesis. Ottawa, ON: University of Ottawa; 1995. AOAC. In: William, S, Baker, D, editors. Ofcial methods of analysis of the association of ofcial analytical chemists. Arlington, VA: AOAC; 1984. Mass D, Masse L, Verville A, Bilodeau S. The start-up of anaerobic sequencing batch reactors at 20 and 25 C for the treatment of slaughterhouse wastewater. J Chem Technol Biotechnol 2001;76:393400. Fannin, KF. Start-up, operation, stability, and control. In: Chynoweth DP, Isaacson R, editors. Anaerobic digestion of biomass. London; 1987. p. 17196.

References
[1] Mass DI, Masse L. Characterization of wastewater from hog slaughterhouse in Eastern Canada and evaluation of their in-plant wastewater treatment systems. Can Agric Eng 2000;42:139 46. [2] Mass DI, Masse L. The effect of temperature on slaughterhouse wastewater treatment in anaerobic sequencing batch reactors. Biores Technol 2000;76:918. [3] Tritt WP, Schuchardt F. Materials ow and possibilities of treating liquid and solid wastes form slaughterhouse in Germany. Biores Technol 1992;41:23545. [4] Borja R, Banks CJ, Wang Z. Stability and performance of an anaerobic downow lter treating slaughterhouse wastewater under [5]

[6]

[7]

[8]