APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1985, p. 1472-1477 0099-2240/85/061472-06$02.

00/0 Copyright 1985, American Society for Microbiology

Vol. 49, No. 6

Methanogenesis in an Upflow Anaerobic Sludge Blanket Reactor at pH 6 on an Acetate-Propionate Mixture

ERIK TEN BRUMMELER,' LOOK W. HULSHOFF POL,1* JAN DOLFING,2t GATZE LETTINGA,' AND ALEXANDER J. B. ZEHNDER2 Department of Water Pollution Control' and Department of Microbiology,2 Agricultural University, 6703 BC Wageningen,

The Netherlands
Received 13 November 1984/Accepted 13 March 1985

High-rate anaerobic digestion can be applied in upflow anaerobic sludge blanket reactors for the treatment of various wastewaters. In upflow anaerobic sludge blanket reactors, sludge retention time is increased by a natural immobilization mechanism (viz. the formation of a granular type of sludge). When this sludge is cultivated on acid-containing wastewater, the granules mainly consist of an acetoclastic methanogen resembling Methanothrix soehngenii. This organism grows either in rods or in long filaments. Attempts to cultivate a stable sludge consisting predominantly of Methanosarcina sp. on an acetate-propionate mixture as substrate by lowering the pH from 7.5 during the start-up to approximately 6 failed. After 140 days of continuous operation of the reactor a filamentous organism resembling Methanothrix soehngenii prevailed in the sludge. The specific methanogenic activity of this sludge on acetate-propionate was optimal at pH 6.6 to 6.8 and 7.0 to 7.2, respectively.

In recent years anaerobic wastewater treatment an increasing interest. This is mainly a result of the energy balance of anaerobic treatment processes

has met positive and the

development of inexpensive, high-rate treatment systems. Various reactor types for anaerobic wastewater treatment have been developed. They are the anaerobic filter, the fluidized and expanded bed reactors, the upflow anaerobic sludge blanket (UASB) reactor, and the fixed film reactors which can be operated either as a down- or an upflow reactor. Results of studies of these reactor types have been presented by Henze and Harremoes (5), Speece (18), and van den Berg (19). The UASB process, as described by Lettinga and co-workers (12, 14), so far has been the most frequently applied treatment process. At present, full-scale plants are successfully treating a variety of wastewaters like wastewaters from sugar beets, potato processing, and corn and potato starch (17). In conventional, completely mixed anaerobic digestors the sludge retention time is controlled by the hydraulic loading rate and the organic matter (volatile suspended solids [VSS]) content of the primary and secondary sludge. In high-rate anaerobic wastewater treatment systems, however, sludge rnust be immobilized by some mechanism, causing the sludge retention time to be almost independent of the flow rate in the system. In UASB reactors imnmobilization is achieved by a natural mechanism, i.e., the formation of highly settleable aggregates of microorganisms. In specific cases this phenomenon can be indicated as granulation, viz. when the aggregates are growing in the form of the distinct granules (7, 8, 9). These granules frequently are high in VSS content (up to 90%), show a high specific activity (2.2 kg of CH4 chemical oxygen demand [COD] VSS-1 day-' at 30C), and are 1 to 5 mm in size (8). The predominant organism in the granules, which develop in UASB reactors in which a volatile fatty acids-containing wastewater is being treated, appears to be a bacterium resembling Methanothrix soehn* Corresponding author. t Present address: Department of Microbiology and Public Health,

Michigan State University, East Lansing, MI 48824.

1472

genii, an acetoclactic methanogen described previously by Huser (10). In pure culture Methanothrix soehngenii grows in two different shapes: rod-like units of 2 to 10 cells and long filaments of several hundred cells (4). Granules grown in full-scale UASB reactors mainly consist of Methanothrixlike organisms growing as rods. A minority of the granules contains filamentous organisms. During the start-up of a UASB reactor a selective washout of dispersed growing sludge is considered to be essential in the granulation process (7, 9). As a result of washout, only bacteria which grow in the heavier particles remain in the reactor. Occasionally, the washout of biomass may be fairly high, leading to undesirable prolonged start-up periods. Certain conditions, such as long periods of underloading, lead to the formation of anaerobic bulking sludge (7) in which little formation of granules occurs. An important role in this bulking is played by Methanothrix sp. because it grows in long filaments. It was thought previously that the formation of anaerobic bulking sludge, apart from specific start-up procedures, can be overcome by creating conditions that favor the development of Methanosarcina sp., another acetoclastic methanogen. Methanosarcina sp. readily forms agglomerates in pure culture (2, 15, 23). These agglomerates settle very well, are spherical in shape, and are up to 2 to 3 mm in size. Consequently, by stimulating the growth of Methanosarcina sp., i.e., making it the predominant organism in the anaerobic sludge of a UASB reactor, the danger of a massive washout can be reduced during the start-up period. In Table 1 the most important kinetic and physiological characteristics of Methanothrix sp. and Methanosarcina sp. obtained in pure culture are summarized. In digesting sludge Methanothrix sp. predominates under conditions of low acetate concentrations. However, at high acetate concentration Methanosarcina sp. prevails. This phenomenon can be explained on the basis of a considerable difference of the Ks for acetate of these organisms. The two organisms also differ in their optimal pH growth on acetate. According to Huser (Ph.D. thesis, Swiss Federal Institute of Technology, Zurich, Switzerland, 1981), Methanothrix soehngenii has its optimum at pH 7.8 and shows no activity below pH

VOL. 49, 1985

METHANOGENESIS AT pH 6

1473

TABLE 1. Characteristics of the acetoclastic methanogens Methanosarcina sp. and Methanothrix soehngenii grown in pure culture on acetate
Organism (no. tested)

(h-1) (temp)a
0.0032 (33C)

K, (mM)

Y (g g_ I)b

OptimuM Optimum temp (TC) pH

7.8
37

Habitat Other substrates Morphology OtesusatsorhogHbit

Methanothrix soehngenii (10)

0.72 5.0

2.1

Rods (2 to 10 cells),

Methanosarcina sp. (23)

a

0.02-0.03

1.4

6-8

40-45

Methanol, H2, methylamines

filaments (100 to 300 cells) Packets of cocci growing in clumps

Digestion sludge

Acetate-rich anaerobic environments

ILmax, Maximum specific growth rate. Y, Growth yield. ity, as determined in a batch-fed experiment, amounted to 0.12 kg of CH4 COD kg of VSS-1 day-'. Averaged over the total reactor volume 10 g of VSS/liter was supplied to the reactors. The start-up procedure applied was identical to that described previously by de Zeeuw and Lettinga (4). The space loading rate (in terms of kg of COD m-3 day-') was increased for 75% once the COD reduction of the system exceeded 85% of the influent concentration (2,850 mg of COD liter-'). To prevent channeling of the medium in the sludge bed during the initial stages of start-up, the reactors were mechanically stirred (30 rpm) every 30 min for 5 s. Once the methane production exceeded 1 liter per reactor volume a day, stirring was terminated because beyond that point the natural mnixing of the system appeared to be sufficient, according to our observations. One reactor, which acted as a reference, was operated at pH 7.5, whereas in the second reactor the pH was kept at approximately 6. The pH of the feed was adjusted to values of respiratioh of 6.5 and 4.5 to 6 with 6 N HCL and 6 N NaOH, respectively. The pH in the reactors was measured several timnes a day. This was done by taking a sample from the reactor fluid just above the sludge bed. The pH of the sample was measured in a small vessel containing the pH electrode (Knick mV meter; Berlin, Federal Republic of Germany). In this way the release of CO2 from the oversaturated fluid could be prevented. The experiments were performed in a temperature-controlled room at 30C (+iC). VFA analyses. Samples (one sample per day per reactor) were taken fromn the settler compartment of the reactors and

6.8. Methanosarcina sp. forms methane in a much wider pH range, namely 5 to 8 (22). UASB reactors are generally operated at pH 6.9 to 7.5. When considered in combination with the low acetate concentrations generally pursued in UASB reactors, the reason for the predominance of Methanothrix sp. in the sludge becomes obvious. In view of what has been mentioned above, it is expected that little, if any, growth of Methanothrix-like organisms will occur when a UASB reactor is started up at pH 6; instead, Methanosarcina sp. will prevail. The purpose of this was to prevent the formation of anaerobic bulking sludge. Anaerobic wastewater treatment can proceed well at a pH range of 6.6 to 7.6 (16). Reactor failure is often caused by decreases in pH (10). This is especially true for propionate degradation. The oxidation of this acid is strongly inhibited for a long period after a pH shock, although adaption is possible (11). Presumably, the main reason for the inhibition at low pHs is the high toxicity of the undissociated fatty acids which become abundant below pH 6 (3). If a methanogenic population could develop at pH 6, it would undoubtedly represent a significant and beneficial feature of anaerobic treatment because this would extend the application of the process to fairly acidic wastewaters without the need for a supply of alkali. The purpose of this study was (i) to investigate whether a reactor start-up at a low pH could be accomplished and whether it would be feasible; and (ii) to confirm that induced Methanosarcina sp. prevail in the cultivated sludge under a reactor pH of 6 and that this is accompanied by an improved settleability of the sludge as compared with the start-up at normal pH (7.5).
MATERIALS AND METHODS Medium, seed sludgea and experimental conditions. Two identical UASB reactors (Fig. 1), each with a volume of 2.5 liters, were used in the experiments. The cylindrical reactors were made of plexiglass, and each had a diameter of 12 cm. The lower part of the reactors was conically shaped to enhance contact of the substrate with the biomass. In the upper part of the reactors a three-phase separator was situated to capture evolved gas and to allow settling. They were fed with a medium consisting of the following (in milligrams per liter): acetic acid (1.250), propionic acid (1.000), H3BO3 (0.5), FeCl3 3H20 (20), ZnCl2 (0.5), MnCl2 * 4H20 (5), NH4Mo7 * 4H20 (0.5), CoCl2 * 6H20 (0.5), NiCl (0.5), EDTA (5), NaSeO3 (1), AlCl3 (0.5), resazurin (5), HCl (36% solution, 10-2 mi/liter of medium), yeast extract (100), NH4Cl (60), (NH4)2SO4 (15), KH2PO4 (15). The seed sludge was obtained from the municipal sewage sludge digester in Ede, The Netherlands. The volatile suspended solids (VSS) content of the sludge was 62% and the maximum specific activ-

e: -Effluent
Gas

FIG. 1. UASB reactors (2.5 liter) applied in the start-up

experiments.

1474

ten BRUMMELER ET AL.

APPL. ENVIRON. MICROBIOL.

time (days)

VFA (mg COD/ l)

14001

made anaerobic by flushing with O2-free nitrogen gas, and they were subsequently filled with 40 ml of an anaerobic buffer solution. The nitrogen gas was made free from residual 02 by passing it over hot copper coils. The buffer was made anaerobic by boiling and then by cooling to room temperature under continuous gassing with O2-free nitrogen. The buffer solution (pH values as indicated) was a 0.2 M KH2PO4-K2HPO4 buffer containing 0.5 g of NH4Cl per liter of demineralized water. Sludge was anaerobically distributed over the vials in portions of 100 to 1,000 mg. Substrate was added to the mixed liquor from concentrated stock solutions to reach final concentrations of 20 to 50 mM. The vials were closed with serum bottle caps and incubated in a shaking water bath at 30C. The sludge was stored at 4C and was reacclimatized by incubating it overnight in the presence of small amounts of substrate. After an overnight reacclimatization, the head space of the vials was flushed with O2-free nitrogen before the methane production rate was determined. At the end of the activity tests, the pH was checked, and the amount of sludge present in the vials was determined by weighing the pellet after the mixed liquor was centrifuged and the supernatant was discarded. The pellet was dried to constant weight overnight at 105C. For methane analysis samples were taken from the head space with a gas pressure lock
syringe.

time (days)

FIG. 2. (a) Course of gas production rate (VG), space loading rate (VL), and pH during the start-up experiment with the UASB reactor operated at pH 7.5. (b) Levels of acetate (0) and propionate (0) expressed as milligrams of COD per liter in the effluent of the UASB reactor operated at pH 7.5.

analyzed for acetate and propionate with a gas chromatograph (model 417; Becker; flame ionization detector [FID] 200C chromosorb column [200 by 0.2 cm ]; carrier gas, N2) equipped with a computing integrator (Spectra Physics 4100). Peak areas were measured and compared with a standard volatile fatty acid (VFA) mixture (precision, 3%). Methane measurement. (i) Volume. The total volumetric methane production was measured every 24 h, after the biogas (CH4-CO2 mixture) was put through a 3 N NaOH solution to scrub the C02, with a wet gas meter (Dordrecht Meterfabrieken, The Netherlands). (ii) Concentration. The methane concentration was determined with a gas chromatograph (Packard-Becker 406) equipped with a thermal conductivity detector and a molecular sieve column, operated at 50C. The carrier gas was argon and was used at a flow rate of 20 ml min-'. Activity tests. Serum vials with a volume of 130 ml were

RESULTS Reactor run at pH 7.5. Figures 2A and B show the concentrations of acetate and propionate in the effluent solution, the course of the rate of gas production, and the reactor pH in relation to the space loading rate that was applied. The conversion of propionate was poor until day 50. At day 50 the pH of the medium was increased from 5.8 to 6.5 to prevent the occurrence of pH shocks after the loading increments. Granules of 0.5 to 1 mm were detected in the sludge bed after 80 days of operation. The predominant organism in the washed out sludge was similar to Methanothrix soehngenii. The mean VSS content of the reactor decreased from 10 g liter-' (day 0) to 1.1 g liter-1 at day 80 and then gradually increased to 1.8 g liter-' at day 150. After the experiments were terminated, the sludge was characterized by the measurement of its specific activity and scanning electron microscopy. The characteristics of the sludge are given in Table 2. The scanning electron micrographs are shown in Fig. 3 and Fig. 4. Reactor run at pH 6. The results of the reactor run at pH 6 are shown in Fig. 5A and B. The reactor feed was interrupted 7 days after the start of the experiment, because no gas could be detected. However, during the feed interruption a slow but distinct increase in gas production occurred. This was noticed by observing the evolution of small gas bubbles. Therefore, at day 17, it was decided to resume the feeding of the reactor. During the experiment it appeared to be difficult to maintain the pH constant at 6 because of the

TABLE 2. Characteristics of sludges obtained in UASB -reactors run at pH 7.5 and after 150 days of operation
Reactor pH Biomass concn (g of VSS

liter-')

CH4of dayo-)

Sp act (g of
g

Yield (g of VSS g of

Predominant

COD-')a
0.031 0.034

morphologyb
Filaments Filaments

organism

Avg granule
diameter (mm)

7.5 1.8 1.8 6 3.5 1.3 a Calculated from the total converted COD during the experiment. b Organisms resembling Methanothrix soehngenii.

1.5 1.0

VOL. 49, 1985

VFA (mg COD!))

METHANOGENESIS AT pH 6

1475

1000-8
VL ,V (9COD.
1

100

day1)

2 00

pH

FIG. 3. Scanning electron micrograph of a part of a granule cultivated in a UASB reactor operated at pH 7.5. Bar, 10.0 ,um.
460

low pH of the influent and the absence of a sufficient buffering capacity. This resulted in pH fluctuations and a pH drop at day 120, which caused a strong inhibition of the process. Moreover, the flow rate was doubled, despite the fact that only 85 to 90% of the acetate was converted and that propionate conversion was negligible. After day 50 a distinct attachment of biomass to the reactor wall became visible. This process of attachment and subsequent growth of the attached biomass continued until at the termination of the experiment at day 140, roughly 50% of the biomass was attached to the reactor wall and 50% was present in the sludge bed. The attached biomass was grey in color, and white, dispersed aggregates (ca. 0.5 mm) could be distinguished. The propionate conversion started after day 50, which coincided with an increase in wall growth. The sludge washed out from the reactor during the period from days 0 to 50 consisted mainly of Methanosarcina-like organisms, whereas after day 50 filamentous organisms predominated in the biomass that was present in the effluent. Granulation of the sludge became apparent 80 days after the start of the experiment. The VSS content of the reactor decreased from 10 g liter -1 at day 0 to 2.0 g liter-' at day 80, but then it gradually increased to 3.5 g liter-' at day 140. The main characteristics of the sludge after termination of the experiments are shown in Table 2. It appeared that the biomass on

-40

time (days)

FIG. 5. (a) Course of the gas production (VG), space loading rate (VL) and pH during the start up experiment with the UASB-reactor operated at pH 6. (b) Levels of acetate (0) and propionate (0) expressed as milligrams of COD per liter of effluent of the UASB reactor operated at pH 6.

the reactor wall was very loosely attached, so it was impossible to analyze it independently of the biomass of the sludge bed. The predominant organism in the cultivated sludge was a filamentous bacterium (Fig. 6). Only a few granules consisted of Methanosarcina-like organisms (Fig. 7). To assess the effect of pH on the acetate and propionate conversion rates of the sludge, the specific activity for acetate and propionate degradation was measured at several

FIG. 4. Scanning electron micrograph of the Methanothrix-like organisms of Fig. 3 shown in more detail. Bar, 1.0 p.m.

L--

FIG. 6. Scanning electron micrograph showing the filamentous

nature of the granules obtained in a UASB-reactor operated at pH 6.

Bar, 10.0 p.m.

1476

ten BRUMMELER ET AL.

APPL. ENVIRON. MICROBIOL.

FIG. 7. Scanning electron micrograph of a part of a Methanosarcina granule cultivated in the UASB reactor operated at pH 6. Bar, 10.0 Jim.

pHs in batch experiments. The results are shown in Fig. 8 and 9. For acetate degradation an optimum was found at pH 6.6 to 6.8. The sludge continued to show a slight but distinct activity at pH 5. For propionate degradation one distinct optimum was found. Activity measurements at pH 6.6 indicated that the sludge might have a second optimum in this range (Fig. 9).
DISCUSSION The predominance of a filamentous bacterium in the sludge cultivated at pH 6 indicates that Methanosarcina sp. cannot compete successfully for acetate with this organism. Consequently, the washout during the start-up period, as was intended originally, cannot be minimized significantly, in comparison with a start-up period at pH 7.5. The difference that was found, i.e., the difference in the lowest calculated VSS content during the experiments of 0.9 g liter-', in favor of the reactor operated at pH 6, can be explained by an improved attachment of biomass in the

reactor operated at pH 6. This phenomenon was not observed in the reactor run at pH 7.5. During the period from days 0 to 50 there was no substantial acetoclastic activity. As a result the acetate concentration prevailing in the reactor was high (1,000 to 1,200 mg of COD liter-'), and consequently preferential growth of Methanosarcina sp. occurred. Once the acetate degradation proceeded satisfactorily, the kinetic circumstances gradually seemed to change in favor of the filamentous organisms. Indeed, beyond day 50 a gradual shift occurred in the bacterial composition of the sludge, viz. the filamentous organisms became predominant instead of Methanosarcina sp. Apart from the kinetic advantage of the filamentous organisms, another mechanism favored the population shift. Methanosarcina sp. tend to form relatively small aggregates of less than 0.5 mm, in comparison with the sludge aggregates, in which the filamentous organisms dominate (1.0 to 1.5 mm). These small Methanosarcina sp. aggregates wash out more readily from the reactor, at comparable hydraulic retention times, than the conglomerates formed by the filamentous organisms. The Methanothrix-like organisms show a strong tendency to attach to either the reactor wall or the inert particles that originate in the inoculum. This phenomenon leads to improved retention of this type of sludge, and consequently Methanosarcina sp. is increasingly outcompeted. The low methanogenic activity from days 0 to 50 of the experiment at pH 6 indicates that both Methanosarcina sp. and the filamentous organism were present in the inoculum in relatively low numbers. Two possible explanations for the development of the filamentous organisms at this pH level can be given: (i) the bacterium was already present in the inoculum in very low numbers and represents a Methanothrix strain with a lower pH optimum on acetate; and (ii) Methanothrix soehngenii adapted to the lower pH. van den Berg et al. (20) have described an acetate-utilizing methanogenic culture with a similar optimal pH range for the specific activity at pH 6.6 to 6.9. The predominant organism in this culture also appeared to be a filamentous bacterium (filaments of 100 to 200 ,um), which is possibly the same organisms as described in our experiments. The results of the experiment at pH 6 show that propio: activity

specitic activity (gmol CH, .gVSS 1min')

41

(Amol CH,, gVSS mnin '1)

30
0

2-

pH FIG. 8. Specific activity on acetate in relation to the pH of the sludge cultivated at pH 6.

FIG. 9. Specific activity on propionate in relation to the pH of the sludge cultivated at pH 6.

VOL. 49, 1985

METHANOGENESIS AT pH 6

1477

nate degradation is possible in a UASB reactor operated at relative low pHs. This may be a result of the formation of microenvironments, in which higher pHs might exist within the granules or the biofilm attached to the reactor wall. According to the investigations of Arvin and Kristensen (1), higher pH values prevail in denitrifying biofilms compared with the pH in bulk solutions. The maximum difference measured amounted to 0.5 to 2 pH units. Arvin and Kristensen have assumed that this phenomenon is a result of lower diffusion coefficients of H+ and HCO3 ions inside the biofilm. In the case of methanogenic biofilms the existence of higher pH values inside the biofilm (granules) is fairly likely because VFAs are being degraded here as a result of the high bioactivity which necessarily causes a rise in the pH of the entrapped solution. According to Dolfing (J. Dolfing, Ph.D. thesis, Agricultural University, Wageningen, The Netherlands, in press), on the other hand, a high pH gap is not likely, because mass transfer resistance is very limited in the granules. Another possible explanation for the propionate degradation at pH 6 may be the existence of a second group of propionate utilizers, as suggested by Heyes and Hall (6), which is faster growing and less sensitive to pH shocks in comparison with the propionate degraders normally found in anaerobic digesters. From our experiments it can be concluded that high rate of anaerobic digestion in a UASB reactor in which an acetatepropionate mixture is treated is possible at pH 6. In a period of 4 months of continuous operation at pH 6, a space load of almost 10 kg COD m-3 day -1 could be reached. The start-up period of a pH 6 reactor can possibly be shortened by applying a more proper seed material, i.e., a sludge adapted to low pH conditions. As reported previously by Williams (21), methanogenic activity in some acid peatlands is found to be optimal at pH 6. For the treatment in UASB reactors of acid wastewater, complete neutralization of the influent is not a prerequisite. This fact has economic implications because fewer chemicals for neutralization are needed. The importance and the effect of different pHs in microenvironments and bulk solutions need to be investigated further.
ACKNOWLEDGMENT This work was supported by a grant from the Dutch Ministry of Housing, Physical Planning and the Environment.
LITERATURE CITED 1. Arvin, E., and G. H. Kristensen. 1982. Effect of denitrification on the pH in biofilms. Water Sci. Technol. 14:838-848. 2. Bochem, H. P., S. M. Schoberth, B. Sprey, and P. Wengler. 1982. Thermophilic biomethanation of acetic acid: morphology and ultrastructure of a granular consortium. Can. J. Microbiol. 28:500-510. 3. Buhr, H. O., and J. F. Andrews. 1977. The thermophilic anaerobic digestion process. Water Res. 11:129-143. 4. de Zeeuw, W., and G. Lettinga. 1983. Start-up of UASB-reactors, p. 348-369. In W. J. van den Brink (ed.), Proceedings of the European Symposium on Anaerobic Wastewater Treatment. TNO Corporate Communication Department, the Hague, The Netherlands. 5. Henze, M., and P. Harremoes. 1983. Anaerobic treatment of

waste water in fixed film reactors-a literature review. Water Sci. Technol. 15:1-101. 6. Heyes, R. H., and R. J. Hall. 1983. Kinetics of two subgroups of propionate-using organisms in anaerobic digestion. Appl. Environ. Microbiol. 46:710-715. 7. HulshoffPol, L. W., W. J. de Zeeuw, C. T. M. Veizeboer, and G. Lettinga. 1983. Granulation in UASB-reactors. Water Sci. Technol. 15:291-305. 8. Hulshoff Pol, L. W., J. Dolfing, W. de Zeeuw, and G. Lettinga. 1982. Cultivation of well adapted pelltized methanogenic sludge. Biotechnol. Lett. 5:329-332. 9. HulshoffPol, L. W., J. Dolfing, K. van Straten, W. J. de Zeeuw, and G. Lettinga. 1984. Pelletization of anaerobic sludge in upflow anaerobic sludge bed reactors on sucrose containing substrates, p. 636-642. In M. J. Klug and C. A. Reddy (ed.), Current perspectives in microbial ecology. American Society for Microbiology, Washington, D.C. 10. Huser, B. A., K. Wuhrmann, and A. J. B. Zehnder. 1982. Methanothrix soengenii gen. nov. sp. nov. a new acetotropic non-hydrogen oxidizing methanobacterium. Arch. Microbiol. 132:1-9. 11. Kaspar, H. F., and K. Wuhrmann. 1978. Product inhibition in sludge digestion. Microbiol. Ecol. 4:241-248. 12. Lettinga, G., S. W. Hobma, L. W. HulshoffPol, W. de Zeeuw, P. de Jong, P. Grin, and R. Roersma. 1982. Design operation and economy of anaerobic treatment. Water Sci. Technol. 15: 177-197. 13. Lettinga, G., A. T. van der Geest, S. Hobma, and J. van der Laan. 1979. Anaerobic treatment of methanolic wastes. Water Res. 13:725-737. 14. Lettinga, G., A. F. M. van Velsen, S. W. Hobma, W. de Zeeuw, and A. KlapwUk. 1980. Use of the upflow sludge blanket (USB) reactor concept for biological wastewater treatment, especially for anaerobic treatment. Biotechnol. Bioeng. 22:699-734. 15. Mah, R. A., M. R. Smith, and L. Baresi. 1978. Studies on a acetate-fermenting strain of Methanosarcina. Appl. Environ. Microbiol. 35:1174-1184. 16. McCarty, P. L. 1964. Anaerobic waste treatment fundamentals. Part two, Environmental requirements and control, p. 123-126. Public works, October 1964. 17. Pette, K. C., and A. I. Versprille. 1982. Application of the UASB-concept for wastewater treatment, p. 121-137. In D. E. Hughes (ed.), Anaerobic digestion 1981. Elsevier/NorthHolland Biomedical Press, Amsterdam. 18. Speece, R. E. 1983. Anaerobic biotechnology for industrial wastewater treatment. Environ. Sci. Technol. 17:416-427. 19. van den Berg, L. 1983. Comparison of advanced anaerobic reactors, p. 71-91. In R. L. Wenthworth (ed.), Proceedings of the Third International Symposium on Anaerobic Digestion. Cambridge, Mass. 20. van den Berg, L., G. B. Patel, D. G. Clark, and C. P. Lentz. 1976. Factors affecting rate of methane formation from acetic acid by enriched methanogenic cultures. Can. J. Microbiol. 22:1312-1319. 21. Williams, R. T., and R. L. Crawford. 1984. Methane production in Minnesota Peatlands. Appl. Environ. Microbiol. 47:12661271. 22. Zehnder, A. J. B., B. A. Huser, T. D. Brock, and K. Wuhrmann. 1980. Charactization of an acetate-decarboxylating, non-hydrogen-oxidizing methane bacterium. Arch. Microbiol. 124:1-11. 23. Zehnder, A. J. B., K. Ingvorsen, and T. Marti. 1982. Microbiology of methane bacteria, p. 45-68. In D. E. Hughes (ed.), Anaerobic digestion 1981. Elsevier/North-Holland Biomedical Press, Amsterdam. 24. Zhilina, T. N. 1976. Biotypes of Methanosarcina. Mikrobiologya 45:481-489.