Antonie van Leeuwenhoek 69: 1-14, 1996. @ 1996KluwerAcademicPublishers. Printedin the Netherlands. 1 Limitations of thermophilic anaerobic wastewater treatment and the consequences for process design Jules B. v a n L i e r Department of Environmental Technology, Wageningen Agricultural University, Bomenweg 2, 6703 HD Wageningen, the Netherlands Key words." anaerobic, biogas production rate, high-rate, inhibition, plug-flow, thermophilic, volatile fatty acids, wastewater treatment Abstract Thermophilic anaerobic digestion offers an attractive alternative for the treatment of medium- and high-strength wastewaters. However, literature reports reveal that thermophilic wastewater treatment systems are often more sensitive to environmental changes than the well-defined high-rate reactors at the mesophilic temperature range. Also, in many cases a poorer effluent quality is experienced while the carry over of suspended solids in the effluent is relatively high. In this paper recent achievements are discussed regarding the process stability of thermophilic anaerobic wastewater treatment systems. Laboratory experiments reveal a relatively low sensitivity to temperature changes if high-rate reactors with immobilized biomass are used. Other results show that if a staged process is applied, thermophilic reactors can be operated for prolonged periods of time under extreme loading conditions (80-100 kg chemical oxygen demand.m -3.day-1),,while the concentrations of volatile fatty acids in the effluent remain at a low level. Introduction Thermophilic anaerobic treatment of waste and wastewater is often found to be less stable than mesophilic treatment (e.g. Buhr & Andrews 1977; Rudd et al. 1985; Soto et al. 1992) restraining most industries and constructors to implement this new technology. Literature reports mention several drawbacks of thermophilic digesters, such as a high susceptibility for i) temperature increases (Varel et al. 1977; Schraa 1983; Zinder et al. 1984); ii) feed interruptions (Wiegant 1986); and iii) shock loadings (Duff & Kennedy 1982; Seif et al. 1992; Soto et al. 1992). In comparison to mesophilic sludge digesters, thermophilic digesters are generally characterized by relatively high concentrations of volatile fatty acids (VFA) in the effluent, indicating a lower degree of process stability as reviewed by Wiegant (1986). In contrast, in other studies very low effluent VFA concentrations were found, in both, thermophilic sludge digesters (Aoki & Kawase 1991; Ghosh et al. 1980) and thermophilic wastewater treatment plants (Schraa & Jewell 1984; Cail & Barford 1985; Ohtsuki et al. 1992; Van Lier et al. 1992, 1994; Wiegant & De Man 1986). In addition, Ahring (1994) recently made a comparison between similar mesophilic and thermophilic full scale digesters and concluded that the VFA concentrations were very similar. Also, Pavan et al. (1994) found a very high stability of thermophilic digesters treating the organic fraction of municipal solid waste under sub-optimal conditions. Yet, it remains unclear to what extent the frequently cited process instabilities are intrinsic disadvantages of thermophilic digestion, and if so, what then must be the measures to adapt the current technology in order to use thermophilic treatment in the most optimal way. In this paper attention will be paid to the state of art and perspectives of thermophilic wastewater treatment in high-rate reactors with immobilized biomass. Thermophilic anaerobic wastewater treatment in high-rate reactors Research on thermophilic wastewater treatment in high-rate reactors is mainly performed using model compounds like VFA and carbohydrates, and with industrial wastewaters which are discharged at high temperatures, e.g. effluents from pulp and paper industries and wastewaters from food processing industries such as alcohol distilleries and canneries. A summary of this research is presented in Table 1. The listed substrates were tested under various loading conditions, however, Table 1 shows only the most optimal values. Going through the literature results it emerges that extreme loading rates are generally accompanied with decreased removal efficiencies, particularly in the studies performed with industrial (non-synthetic) wastewater (e.g. Rintala & Lepist6 1992; Wiegant et al. 1985). However, due to the variety of wastewaters studied and the different processes applied, it is very difficult to conclude whether or not high effluent VFA concentrations are inevitable during thermophilic wastewater treatment. Roughly, the results reveal that a high degree of VFA removal can only be achieved in moderately low loaded thermophilic high-rate reactors (e.g. Ahring et al. 1993; Schraa & Jewell 1984; Van Lier etal. 1992). In addition to high effluent VFA concentrations, perturbed reactors are sometimes characterized by excessive wash-out of active thermophilic biomass as well (Duff & Kennedy 1982). Problems with biomass retention were also observed by Soto et al. (1992) who found that immobilisation of anaerobic biomass in upflow anaerobic filters is more difficult to achieve under thermophilic conditions than under mesophilic conditions. The main difference between the two filters was that under mesophilic conditions the biomass could be entrapped in between the support material, while in the thermophilic filter the void volume was virtually free of biomass. The latter was attributed to the formation of dispersed sludge with poor settling abilities (Soto etal. 1992). Growth of a similar type of thermophilic sludge was also observed by others (Uemura & Harada 1993; Van Lier etal. 1992; Wiegant & Lettinga 1985). The formation of dispersed sludge might partly be attributed to the higher degree of sludge mineralization under thermophilic conditions (Soto etal. 1992), which, consequently, results in a lower amount of extracellular polymers. Schmidt & Ahring (1994) recently indeed found that the content of extracellular polysaccharides and protein in thermophilic granular sludge was much lower than in mesophilic granules. A low content of these polymers may hinder the formation of dense and firm granules. It should be noted that the formation of dispersed biomass was particularly found when VFA mixtures were used as the sole substrate, while addition of sucrose or glucose to the influent resulted in the formation of thermophilic granular sludge with a high specific activity and good settling abilities (Wiegant & Lettinga 1985; Uemura & Harada 1993; Van Lier etal. 1992, 1994). Vanderhaegen et al. (1992) previously discussed the important role of acidifiers and/or their excretion products, such as extracellular polysaccharides, for the cultivation of sludge granules in the mesophilic temperature range. Due to the higher mineralization rate, the requirement for extracellular polymers might even be higher under thermophilic conditions. Another reason explaining dispersed growth, might be the morphological appearance of the predominant methane bacteria at high temperatures. In contrast to mesophilic conditions, long filamentous bacteria are less frequently observed, while often short rod-shaped or even coccoid methane bacteria dominate (Uemura & Harada 1993; Van Lier et al. 1992; Schmidt and Ahring 1993). A low surface to volume ratio of thermophilic methanogens may negatively influence the adherence capacity of the cells, assuming that under thermophilic conditions the cell-hydrophobicity also plays an important role in the initial attachment, similar to mesophilic conditions (Grotenhuis et al. 1992). Finally, due to the tow liquid viscosity and the eventual high biogas production rate under thermophilic conditions, the cells which are only weakly bound, are more easily rinsed from the system. Nevertheless, immobilization of anaerobic biomass by formation of granular sludge is a common phenomenon in thermophilic methanogenic upflow reactors (Ohtsuki et al. 1992; Ohtsuki etal. 1994; Schmidt & Ahring 1993; Souza et al. 1992; Uemura & Harada 1993 1995; Van Lier et al. 1994; Wiegant & Lettinga 1985; Wiegant & De Man 1986). Generally, thermophilic granular sludge exerts a very high specific activity. The high activity of the immobilized biomass is attributed to the high energy consumption for maintenance of the thermophiles (Wiegant 1986). As already mentioned above, the presence of a fraction non-acidified substrate apparently is indispensable for the cultivation of thermophilic granules in high-rate reactors. In contrast, other authors observed granulation of methanogenic biomass with a VFA mixture as the sole substrate both under mesophilic (Hulshoff Pol 1989) and thermophilic conditions (Ahring 1991; Wiegant & De Man 1986). In addition to the substrate, other factors such Table 1. Thermophilic anaerobic wastewater treatment in 'high rate' reactors. Substrate Temp. ~ process a working volume 1 sucrose 55 AAFEB sucrose-VFA 55 USSB sucrose-VFA VFA Ac/Ac-Bu 55 55 55 sucrose-Ac sucrose-Ac coffee waste coffee waste Ref. Influent g COD.1 - I loading rate g COD.I-I.day - i HRT h COD removal % 2.0 g 2-16 30-40 0.5-5 70 1 5.1 a 9-10 80-100 2-2.5 90 2 UASB UASB UASB 5.75 5.75 5.75 8 14.7 3-4.5 35 104 120-150 5.7 3.2 0.6 84.5 78 84-93 3 3 4 55 65 55 53 UASB UASB UASB CSTR-FB c 5.5 5.5 10 e 5 + 0.45 3 3 4-8 23 87 41 6-10 7 0.8 1.8 32 78 96 77 70 20 5 5 6 7 coffee waste corn steep liq. 55 53 UAF AFR 1.t 4 1.1 4 5.5 4 11 24 12 60 90 8 9 beer brewing process fish and sea food non-fat dry milk 55 55 56 UASB (two stage) UAF UAF 1400 0,92 16.8 20 10-26 25 30-40 / 12 50 < 24 53 12 80-90 74 67-78 10 11 12 wood hydrolysate bean blanching 55 55 UFF DSFF 0.45 a 1.254 22.5 10 10 6.6-17.4 54 13-31 84.4 84-91 13 14 meat waste 57-60 FB 4 2.5 4.6 13 49-59 15 boiled soya bean boiled soya bean ice cream distillery vinasseb vinasse b 54 54 55 42 55 55 UFF/AF UFF/AF UAF UAF UASB UASB lab pilot 1.1 500 5.75 75000 32-616 32-66 10 45-50 diluted 31.5 47 40 36 389 100 25-30 20 n,r. 6.6 29 2.5 11 90 90 85 40-50 60 72 16 17 17 18 19 20 TMP-white water 55 UASB 0.25 2-3.5 3 14-22 max.80 6 1 65-75 60 21 TMP-white water pharma.-glucose 70 55 UASB UFF 0.25 35 d 3 2.5-7 13 0.5-1.5 7 113 60 51-58 21 22 Abbreviations: Temp., temperature; HRT, hydraulic retention time; Ref., reference; liq., liquor; TMP, thermo-mechanical pulping; pharma., pharmaceutical; n.r., not reported. Ref.: 1 = (Schraa & Jewell 1984); 2 = (Van Lier et al. 1994); 3 = (Wiegant & Lettinga 1985); 4 = (Wiegant & De Man 1986); 5 = (Uemnra & Harada 1993); 6 = (Lanting et al. 1989); 7 = (Kida et al. 1992); 8 = (Fernandez & Forster 1993); 9 = (Yang et al. 1992); 10 = (Ohtsuki et al. 1994); 11 = (Lema et al. 1988); 12 = (Harris & Dague 1993); 13 = (Good et al. 1982); 14 = (Kennedy & Van de Berg 1982); 15 = (Rudd et al. 1985); 16 = (Kawase et al. 1989); 17 = (Ugurlu & Forster 1992); 18 = (Braun & Huss 1982); 19 = (Wiegant et al. 1985); 20 = (Souza et al. 1992); 21 = (Rintala & Lepist5 1992); 22 = (Self et al. 1992). a process: AAFEB, anaerobic attached film expanded bed; USSB, upflow staged sludge bed; UASB, upflow anaerobic sludge bed; CSTR, continuous stirred tank; FB, fluidized bed; UAE upflow anaerobic filter; AFR, anaerobic fixed-bed reactor; UFE upflow fixed film; DSFF, downflow stationary fixed film, b effluents from alcohol distillery. c batch digestion of coffee waste, flow is kept by recirculation, COD reduction based on CH4 yield. a total volume. e results from 6 m 3 UASB pilot study were slightly worse. f a load of 50 kg COD.m - 3 . d a y - ~ was reached after replacement of 30% of the COD load with sugar. g loading rate in kg VS.m-3.day -1 . as i n o c u l u m source, might determine whether or not at e x t r e m e g a s p r o d u c t i o n r a t e s ( W i e g a n t 1 9 8 6 ) . T h i s indicates results reveal that the transport of the pro- sludge granules are formed. granular duced biogas from the reactor medium can become crit- sludge, an excessive carry-over of biomass may occur ical w h e n c o n v e n t i o n a l u p f l o w a n a e r o b i c s l u d g e b e d Despite the presence of thermophilic Table 2. Temperatureoptimaand growthkineticpmametersof severalacetate utilizingmethanogeniccultures. Acetate utilizingmethanogens Topt. (~ Tmax (~ ~max (h-l) Ks (Ac) (mg COD.I-a) Ref. Methanosarcina barkeri Methanosarcina thermophila MethanosarcinaCALS-1 Methanosarcina MP Methanosarcina MSTA-1 MethanosarcinaCHTI55 Methanothrixsoehngenii Methanothrix concilii Methanosaetasp.PT TAM Methanothrixsp. CALS-1 Methanothrix thermoacetophila Acetate oxidizingco-culture 35-40 50 55-58 55 55 57 37 35-40 55 60 60 65 60 n.r.a 55-60b 60 60 65 63 45-50 b 40-45 b 65-70b 70 65-70 b 70 n.r. 0.023 0.058 0.058 n.r. 0.053 0.085 0.0085 0.029 0.020 0.012 0.028 n.r. 0.019 320 288 n.r. n.r. 685 614 45 77 n.r. 51 < 64 n.r. n.r. 1 2,3 4 5 6 7 8 9 10 11 12 13 14 Ref.: 1 = (Smith & Mah 1978);2 = (Ziuder& Mah 1979);3 = (Zinderet al. 1985);4 = (Zinderet al. 1984);5 = (Ollivieret al. 1984); 6 = (Clarens& Moletta 1990); 7 = (Touzelet al. 1985); 8 = (Huseret al. 1982); 9 = (Patel 1984); 10 = (Kamagata& Mikami 1991); 11 = (Ahring& Westermann1985); 12 = (Zinderet al. 1987): 13 = (Nozhevnikova&Chudina1984); 14 = (Zinder& Koch 1984). a n.r. = not reported. b no growth observedat highest temperatureof givenrange. (UASB) reactors are used for thermophilic wastewater treatment. Obviously, if dispersed sludge with a poor settling ability is cultivated, such critical biogas load will already manifest at relatively low organic loading rates (Van Lier et al. 1992). Also, due to the lower liquid viscosity at high temperatures, sludge particles are easily lifted from the sludge bed and subsequently rinsed to the settling area. Biological and physical limitations of thermophilic wastewater treatment Temperature susceptibility The presumed susceptibility for temperature fluctuations seems to be one of the major concerns for most industries to abandon thermophilic wastewater treatment. However, it should be noted that most of the early studies on thermophilic wastewater treatment were performed using completely mixed systems which are characterized by a high sensitivity at any temperature range (Zinder et al. 1984; Zinder 1986). A 5 ~ temperature increase in a 58 ~ stirred tank reactor (CSTR) led to complete reactor failure with a concomitant shift in the methanogenic population (Zin- der et al. 1984). Recently we described the occurrence of various temperature optima for acetate degradation in the temperature range of 45-65 ~ (Van Lier et al. 1993a). In these experiments thermophitic sludge was cultivated in batch reactors at 46, 55 and 64 ~ with mesophilic granular sludge (MGS) as inoculum and acetate as the sole substrate. After an 'adaptationperiod' of 1.5-2 months the maximum acetate degradation rate was assessed at various temperatures. For the sludge cultivated at 46, 55 and 64 oC, temperature optima were found at 50, 57 and 65 ~ respectively. The activity of the (sub)thermophilic sludges did not drop immediately to zero if the temperature of the medium exceeded the optimum temperature. This was most pronounced for the sludge cultivated at 46 o C. Probablyl methanogens with a higher optimum temperature than the cultivation temperature were present in the sludge as well. The occurrence of various temperature optima for methanogenesis in the (sub)thermophilic temperature range, agrees with the optimum growth temperatures of thermophilic acetate-utilizing methanogens described in the literature (Table 2). For comparison, Table 2 includes some mesophilic acetate utilizing methanogens as well. Regarding the differences in optimum growth temperatures, methanogens with the highest growth rate at a fixed process tempera- ture will be predominant in the cultivated sludge. This phenomenon may also explain the occurrence of complete process deterioration in thermophilic digesters for sewage sludge and man~;e, when the temperature is increased with only a few degrees celsius (Garber et al. 1975; Ahring 1995). Recovery of the process in such case, obviously depends on the growth of new bacterial mass, indicating a shift in the microbial population similar to the results of Zinder et al. (1984). Very striking are the results of previous experiments in which the temperature response curve was assessed from thermophilic sludge cultivated in UASB reactors at 46, 55 and 64 ~ (Van Lier et al. 1993a). In contrast to the sludge cultivated in batch reactors, a single temperature optimum for acetate conversion was found at 60-65 ~ irrespective of the cultivation temperature in the range 46-64 ~ Recently, Uemura & Harada (1993) found a similar temperature response of thermophilic sludges which were cultivated in two different UASB reactors at 55 ~ and 65 ~ respectively. For both sludges a maximum acetate conversion rate was found at 65 ~ Most likely, other selection criteria than the specific growth rate are of importance during the development of thermophilic methanogenic consortia in sludge bed reactors (Van Lier et al. 1992, 1993a). Kinetic and adherence properties of thermophilic methanogens rather than optimal growth temperatures may determine the bacterial composition of a sludge bed. From Table 2 follows that all acetate-utilizing methanogens with an optimum growth temperature between 60 and 65 oC belong to the genus Methanothrix ("Methanosaeta"; Patel & Sprott ' 1990) (Table 2). Interestingly, granular anaerobic sludge from UASB reactors is mostly dominated by Methanothrix-type bacteria, both in the mesophilic (De Zeeuw 1984; Grotenhuis et al. 1991) and the thermophilic range (Uemura and Harada, 1993 1995; Van Lier et al. 1992, 1993a; Wiegant and De Man, 1986). In addition to the frequently observed Methanothrix species, granular sludge may consist also of syntrophic acetate-oxidizing consortia or bacteria belonging to the genus Methanosarcina. This will be discussed in more detail in the next sections. The above results indicate that if the maximum specific growth rate is the predominant selection criterion for the methanogenic consortium, a high sensitivity to temperature changes can be expected. This will be the case during the start-up of a high-rate reactor but also during the operation of batch reactors and/or convential completely mixed systems under high loading conditions. Therefore, high-rate systems with a high solids retention time are preferred over the CSTR-type systems for the application of thermophilic wastewater treatment. Effect of biomass immobilization on the temperature susceptibility High-rate reactors are characterized by the formation of biofilms and/or granular aggregates. While the activity of anaerobic microorganisms is affected strongly by temperature (Heitzer et al. 1991), the effect of temperature on substrate diffusivity is only marginal (Perry & Green 1984). Therefore, immobilization of bacteria in biofilms and/or granules, may enhance the thermostability of the process due to the fact that the maximum conversion rate will be determined by diffusion limitation of the substrate (Smith 1981 ; Lens et al. 1993; Pavlostathis & Giraldo-Gomez 1991). Obviously, such effect will manifest particularly when the methanogenic sludge is characterized by a high specific activity. Most likely, mass transfer limitation plays a major role in the thermostability of thermophilic high-rate reactors with immobilized biomass. Figure 1 shows the temperature response of the maximum acetate utilization rate of thermophilic granular sludge cultivated on a VFA-sucrose mixture in an upflow staged sludge bed (USSB) reactor at 55 ~ (Van Lier et al. 1996). Surprisingly, the maximum rate of approximately 2.5 g chemical oxygen demand (COD).g -1 volatile suspended solids (VSS).day- 1 remained unaffected by temperature in the range between 50 ~ to 65 ~ Crushing these thermophilic granules led to a 2 to 3-fold increase in the methanogenic activity at 65 ~ This was accompanied by a high temperature dependency. Apparently, biomass immobilization brings about a dual effect. On one hand it decreases the specific substrate conversion rate due to mass transfer limitation. On the other hand, it enhances the thermostability of the overall process by providing an extra biomass buffer for a sudden drop in temperature. One may speculate that such mechanism is of big importance regarding any change in environmental conditions. Maximum temperature The maximum applicable temperature in anaerobic digestion is not clear yet. Previous research revealed that acetate conversion even proceeds at 75 ~ (Van Lier et al. 1991). To investigate the feasibility of extreme thermophilic wastewater treatment, an UASB 6 Acetate conversion rate (g COD/g VSS. day) 6 [ crushed gran 4 ~" 9 , t v ", ,9 50 70 temp. (~ Fig. 1. Maximum acetate conversion rate at various temperatures of intact- and crushed thermophilic granular sludge cultivated at 55 ~ in an USSB reactor (Van Lier et al. 1996). ;ubstrate conversion rate '.g COD/g VSS. day) 1.2 75~ - sludg e //~ / J 9 Acetate 1 9 Propionate J \ I J 0.8 0.4 temp. (~ Fig. 2. Maximum substrate conversion rate at various temperatures of extreme therophilic sludge cultivated at 75~ on a VFA mixture (triplicate). Assays performed according to Van Lier et al. (1996). reactor was started up at 75 ~ with MGS as inoculure and a VFA mixture as the substrate (Van Lier et al. 1993c). Interestingly, the optimum temperature for acetate degradation of the cultivated 75 ~ sludge differed distinctly from the optimum temperature found for the sludge cultivated under moderate thermophilic conditions i.e. 45-65 ~ (Fig. 2; Van Lier et al. 1993a). Apparently, there is a significant shift in the methanogenic population if the tempera- ture is increased beyond 65/70 ~ Regarding propionate and butyrate such a shift is less clear. Temperature response experiments revealed that the specific propionate degrading activity of the 75 ~ was below 0.03 g COD.g - t VSS.day -1 for the entire temperature range between 45 ~ and 90 ~ No clear optimum was found. The specific butyrate degrading activity showed an optimum of 1.2 g COD.g-1 VSS.day -1 at 65 ~ while little if any activity was found at the cultivation temperature of 75 ~ (Fig. 2). The UASB process at 75 ~ appeared to be limited by severe wash out of fine, active methanogenic biomass, which resulted in a rapid decrease of the sludge bed volume (Van Lier et al. 1993c). One may expect an even higher degree of mineralization and thus, a lower content of extracellular polymers under extreme thermophilic conditions. As mentioned above, the UASB reactors at 75 ~ were seeded with MGS. The relatively high abundance of extreme thermophilic organisms in the mesophilic biomass is not yet understood. Recently, Rintala et al. (1993) described similar startup experiments at 70 ~ using the same seed material. Surprisingly, the cultivated 70 ~ showed no clear temperature optimum for acetate conversion in the range 55-70 ~ Moreover, the methane production rate dropped distinctly at 75 ~ Apparently, 7075 ~ is the critical temperature border beyond which the 'moderate' acetate-utilizing methanogens cannot survive. Acetotrophic methanogenesis at 75 ~ is not yet described in the literature. In general, acetate is split in a so-called aceticlastic reaction in which the methyl-group of the acetate molecule is reduced to CH4, while the carboxyl-group is oxidized to HCO3(e.g. Zehnder et al. 1980, 1982). So far, only two genera of methanogenic bacteria, i.e. Methanosarcina and Methanothrix, are capable of catabolizing acetate to CH4, both under mesophilic and thermophilic conditions (Zinder 1990). In addition to the above aceticlastic conversion, acetate can also be degraded via a two-step reaction in which acetate is first oxidized to H2/CO2, followed by a subsequent conversion to CH4 (Zinder & Koch 1984; Weber et al. 1984). Although the latter reaction is thermodynamically less favourable, recent findings reveal that the two-step reaction might become important under high temperature conditions and particularly at low acetate concentrations (Ahring 1995). Petersen & Ahring (1991) demonstrated that syntrophic acetate oxidation might contribute to up to 14% of total acetotrophic methanogenesis in a thermophilic (60 oC) digester. Moreover, results of Ahring (1995) show that when the acetate concentration drops to below the threshold level for the dominating aceticlastic methanogen in a 55 ~ digester, the two step reaction becomes the predominant mechanisms for acetate conversion. Uemura & Harada (1993, 1995) and Ahring et al. (1993) recently showed that a relatively large fraction of granular sludge grown in 55 oCUASB reactors also consists of such syntrophic consortia. An even more dominant role of non-aceticlastic acetate conversion is expected at temperatures higher than 65 ~ (Ahring et al. 1995; Uemura & Harada 1993, 1995), which is beyond the temperature range of the aceticlastic methanogens known thus far (Table 2; Zinder 1990). The addition of acetate to methanogenic sludge cultivated at 70-75 oC in UASB reactors, resulted in a rapid build-up of H2 in the head-space of closed serum vials (Rintalaet al. 1993; Van Lier et al. 1993c). Moreover, recent investigations with the 75 oC-sludge, using 13C-labelled acetate as the substrate, showed that approximately 95% of the total acetate conversion was performed by the syntrophic acetate oxidation reaction (Van Lier et al., unpublished results). The predominance of the latter reaction in extreme thermophilic (70 ~ UASB reactors was recently confirmed by Ahring et al. (1995). Furthermore, syntrophic oxidation of acetate was shown to be the principal pathway to acetate conversion in Islandic hot springs of 70 ~ (Ahring 1992). These results indicate that under extreme thermophilic conditions hydrogenconsuming methanogens are mainly responsible for the final COD removal from waste and wastewater (Fig. 3). The possible biotechnological implications of these findings are yet not understood. Regarding the possible role of hydrogen in the anaerobic conversion of organic matter, Westermann (1994), recently made a reversed observation under psychrophilic conditions. He observed a decreased contribution of hydrogenotrophic methanogens in swamp slurries with decreasing temperatures from the mesophilic (37 ~ to the psychrophilic (2 ~ range. Combining these results with the observations made under thermophilic conditions, one can speculate on a biological rule that the importance of H2 in methanogenic ecosystems increases with increasing temperatures from below 0 ~ to higher than 100 ~ Although methanogenesis from acetate even proceeds under extreme thermophilic conditions, it is recommended to keep the process temperature of a thermophilic wastewater treatment plant below 60 oC. This temperature limit is determined by the propionateoxidizing bacteria rather than by the acetate converting organisms. Recent studies show that propionate con- Complex Polymers (proteins, polysaccharides, etc) Mono and Oligomers (sugars, amino acids, peptides) Acetate, Propionate, Butyrate, etc. (long-chain fatty acids) "f H2+ C O 2 ]J I OH4, CO2 Fig. 3. Hypothetical carbon flow in the anaerobic degradation of organic matter under extreme thermophilic conditions, assuming that ac6ticlastic methanogenesis does not contribute to the ultimate COD removal. version might become critical at temperatures higher than 60 ~ (Rintala & Lepist6 1992; Van Lier et al. 1992, 1993a). Also in thermophilic manure digesters, propionate degradation, together with hydrolysis of solid matter, probably determines the maximum applicable temperature (Ahring 1995). In addition, in previous experiments we found only one single optimum temperature for propionate conversion at about 5560 ~ despite the fact that the thermophilic sludge was cultivated in UASB reactors at 46, 55 or 64 ~ (Van Lier et al. 1993a). These results are, however, in contrast with the recent results of Uemura and Harada (1995) who found a similar propionate-utilizing methanogenic activity of thermophilic (55 ~ granular sludge in the temperature range of 55-65 ~ Because in the latter study the maximum propionate utilization rate was calculated from the methane production rate, unknown temperature effects might have masked the actual effect on the propionate conversion. Effluent VFA concentrations Acetate conversion Single stage thermophilic high-rate reactors which are operated under high loading conditions are often characterized by relatively high VFA concentrations in the effluent. The high concentration of residual VFA can only partly be explained by the low substrate affinity of some thermophilic organisms. Various thermophilic methanogenic species have a considerable lower affinity for the substrate than their mesophilic homologues, e.g. Methanosarcina sp. (Clarens & Moletta 1990; Touzel et al. 1985) and Methanobacterium sp. (Sch6nheit et al. 1980). However, the substrate affinity of thermophilic Methanothrix spp. is similar to that of the mesophilic homologue (Ahring & Westermann 1985; Zinder et al. 1987). As already mentioned, Methanothrix sp. are the predominant acetotrophic methanogens in granular sludge from thermophilic upflow reactors (Wiegant & De Man 1986; Van Lier et al. 1992; Uemura & Harada 1993, 1995). However, high effluent acetate concentrations (> 700 mg COD.l- 1) may result in the development of sludge granules predominated by Methanosarcina spp. (Wiegant & De Man 1986). The latter type of granules are also characterized by a high specific activity, but they appear to be smaller and weaker, and are more easily rinsed from the system than the Methanothrix dominated granules (Wiegant & De Man 1986). Interestingly, Ahring (1991 ) recently reported the development of thermophilic sludge granules in which Methanosarcina spp. were the only acetotrophic methanogens, despite the fact that low effluent acetate concentrations were applied during cultivation in UASB reactors. The occurrence of these 'Methanosarcina-granules' might be attributed to the kind of inoculum used and the rather low loading rates applied. The UASB reactors from the latter study were seeded with sludge from a CSTR-type large scale biogas plant, treating manure and industrial waste at 53-55 ~ (Ahring 1991; Ahring et al. 1995). Because Methanosarcina spp. predominated in the seed sludge (Ahring 1991), an eventual predominance of thermophilic Methanothrix spp. will be very difficult, if possible at all. Moreover, the reactors were operated at relatively low loading rates of about 10 g COD.l- 1.day- 1 (Ahring et al. 1993), which prevents exposure of the 'Methanosarcina-granules' to a high biogas turbulence and a high upward liquid velocity. According to Wiegant & De Man (1986), Methanosarcina dominated sludge will particularly ~ 1001 "~ ~ 6o ~9 ~ 60. -~ ,~ ~ V = Vm / (1 + Km/S + (S/Ki) ^ n) i~ ~ ~J, e~ - - ~ " - - Thermophilic-1 ..... 9..... Thermophilic-2 - - e - - - . Mesophilic ~, ~ 100% = (g/g.day} 1.86 0.92 0.43 . 40- ~ 200 0:2 014 016 0.8 Acetic acid (g HAc-COD/I) Fig. 4. Methanogenic activity versus acetic acid concentration of thermophilic granular sludge grown in compartments1 and 5 of an USSB reactor (Fig. 5) which was fed with a 1:1 socrose-VFAmixture (based on COD). Acetate concentrations in the compartments were ( I ) 1500and ( 9 150 mg acetate-COD.l-l, respectively.Figure depictsthe activityof the seed materialas well. Solid lines were computed using Haldane's equation for substrate inhibition (Morvai et al. 1992). The pH during the assay ranged between 6.8-7.3 and between 6.3-6.7 for the thermophilicand mesophilic sludges, respectively. deteriorate under conditions of high 'shear stress'. No information about the strength of the 'Methanosarcinagranules' was given. As mentioned above granulation of thermophilic sludge results in a lower overall activity due to mass transfer limitation. A concomitant effect of biomass immobilization is a decrease in the overall substrate affinity. For thermophilic granular sludge, apparent K,~ values as high as 1800 mg COD.I- 1were measured for acetate conversion at 55 oC, while for crushed granules, apparent K,~ amounted 850 mg acetate-COD.l- ] at 55 ~ (Van Lier et al. 1996). Due to the extremely high apparent Kin, effluent acetate concentrations may easily rise under high loading conditions. Inhibition effects In addition to the lower substrate affinity also toxicity effects may influence the VFA-pool in the reactor. Experiments performed in our laboratory revealed that thermophilic acetate conversion was inhibited severely at relatively low acetate concentrations (Van Lier et al. 1996). Figure 4 shows the acetate utilizing activity of two thermophilic granular sludges, cultivated at either 1500 or 150 mg acetate-COD.l -] , as well as the activity of the mesophilic inoculum. All calculations were based on the concentration of the undissociated acetic acid which is thought to determine the actual degree of inhibition (Fukuzaki et al. 1990; Morvai et al. 1992). Surprisingly, both the thermophilic sludges showed a similar sensitivity, despite the big difference in acetate concentration at which the sludges were cultivated. The sensitivity of the seed sludge was much lower. Other experiments revealed that propionate degradation was severely inhibited by addition of 3.2 g acetate-COD.l-1 to the influent of an UASB reactor which was fed with 3.9 g propionate-COD.l-1 as the sole substrate (Van Lier et al. 1993b). The inhibitory effect remained, even when the acetate concentration in the effluent was below detection level. Recovery of propionate oxidation occurred only when the acetate was omitted from the influent medium. Wiegant (1986) proposed a two step digestion process whenever propionate is a major compound and hydrogen production is inevitable during the thermophilic treatment of wastewater. In the second reactor of this system propionate would be removed. The overall better performance was attributed to a decreased hydrogen partial pressure in the second reactor which, for thermodynamic reasons, stimulates propionate oxidation. The above results show that in addition to a low hydrogen concentration, a low level of acetate seems to be indispensable for an effective degradation ofpropionate. The possible occurrence of toxicity effects is another reason why industries are not interested to implement thermophilic anaerobic treatment so far. In comparison to mesophilic sludge, thermophilic sludge indeed was found to be more sensitive to various organic compounds as well as to heavy metals (Disley et al. 1992; Macleod & Forster 1988). Wiegant et al. (1985), found a clear correlation between the strength of vinasse wastewater (i.e. effluents from an alcohol distillery) and the VFA concentration in the effluent of a thermophilic UASB reactor. Organic loading rates of 25-40 kg COD.m -3 reactor.day-1 resulted in total effluent VFA concentrations of 3,000-8,000 mg COD.I-1. Apparently, thermophilic UASB reactors are rapidly limited by toxic compounds in the bulk of the solution. However, by applying a more appropriate reactor design certain toxicity effects can be minimized, particularly those which are caused by the substrate and/or the intermediates produced, and probably also toxicity effects caused by volatile gaseous compounds. This will be discussed in more detail in the next section. Staged reactor concept In principle, thermophilic process conditions may lead to much higher loading potentials of anaerobic reactors than mesophilic conditions. However, from the above literature review and experimental results it becomes clear that high-rate thermophilic wastewater treatment can become limited by various biological and physical factors. In single stage sludge bed reactors biomass retention might become critical when COD removal rates exceed 50-60 kg.m -3 reactor.day -1 for prolonged periods of time. Sludge separation will even be more critical at high temperatures because of the lower liquid viscosity and the sometimes occurrence of less stable aggregates under thermophilic conditions. Taking into account the high sensitivity to relatively low concentrations of intermediate compounds such as hydrogen and acetate, it will be clear that low effluent VFA concentrations are virtually impossible to realize in single stage thermophilic high-rate reactors. However, in applying a two-stage process with two methanogenic reactors in series, significant improvements are made (Kida et al. 1992; Lanting et al. 1989; Ohtsuki et al. 1994; Wiegant et al. 1986). Kaiser et al. (1993) successfully operated the second reactor under mesophilic conditions. Apparently, a staged reactor system provides a higher treatment efficiency. In recent experiments we clearly demonstrated that when a compartmentalized system is applied, a very high degree of VFA removal can be achieved (Van Lier et al. 1994). The upflow staged sludge bed (USSB) reactor in which the various stages of the degradation process are separated, consists of 5 compartments (Fig. 5). Each compartment is equipped with its own gas-solids separator. This set-up enabled us to perform detailed analysis about the staged degradation and the various factors affecting this pattern. The start-up of the staged reactor system proceeded rapidly. Within 2-3 months the organic loading rate of a reactor treating a sucrose-VFA mixture could be increased up to 100 kg COD.m-3.day -1 (Fig. 6). The COD removal rate exceeded 90%. Long term operation at these high loading rates did not lead to a significant wash-out of the thermophilic biomass despite the application of an extreme biogas loading rate of about 50 m3.m -3 reactor.day -l. The main achievements of the plug-flow reactor are: i) very low effluent VFA concentrations under extreme loading conditions, i.e. 10120 mg acetate COD.l-1 and 200-500 mg propionate COD.l-I; ii) high degree of sludge retention due to very low turbulence in the final reactor compartment; 10 Sucrose, VFA (g COD/I) Biogas Effluent Sucrose Acetate Propionate Butyrate 9 9 9 0 Influent W udge I 1 2 3 4 5 Effluent Compartment Fig. 7. Degradation pattern of the partial acidified substrate in the various compartments of the USSB reactor at a loading rate of 75 kg COD.m-3.day -1 (Van Lier et al. 1994). Fig. 5. Upflow staged sludge bed (USSB) reactor. COD Removal (g COD/I. day) /,,..'l 100% Removal ] I / 80 / I ~ VFA .Zli~ T M -~ 0 i i 20 . i= o J i 40 * i 60 i l 80 I t i 100 Organic Loading Rate (g COD/I. day) Fig. 6. COD removalrate versus applied organicloading rate. USSB reactors were fed with VFA or Sucrose-VFA and were inoculated with mesophilic granular sludge (Van Lier et al. 1994). and iii) stable reactor performance due to the segregated development of methanogenic sludge along the reactor height. The advantage of using compartmentalized reactors was clearly demonstrated under the extreme loading conditions applied. A characteristic sequence in the degradation of the sucrose-VFA mixture was found. In the first compartment sucrose was converted, followed by the conversion of butyrate and acetate in the next compartments. As usual, propionate was the most difficult intermediate to degrade, but in the last compartments also this fatty acid was degraded almost completely (Fig. 7). Sudden changes in the influent substrate concentration hardly affected the VFA concentration in the effluent (Van Lier et al. 1994). The withdrawal of the produced biogas along the reactor height appeared to be very beneficial for the overall treatment process. As mentioned above, the reactor performance was not limited by a wash-out of thermophilic biomass due to the optimal settling conditions at the top part of the reactor. Also, the hydrogen, which is mainly produced during acidification of organic matter, is removed effectively from the lower compartments. For thermodynamic reasons, high hydrogen concentrations may affect the acetogenic reactions in successive compartments. Such 'stripping effect' also might be of importance if sulphate-rich wastewater is treated. Due to acid formation the pH in the lower compartments will be relatively low while most of the sulphate will 11 be reduced here. Consequently, a large fraction of the produced sulphide will be stripped from the solution. Stripping of sulphide will minimize eventual sulphide inhibition effects on acetogenic and methanogenic conversions in the higher compartments. Moreover, in the higher compartments merely sulphide-free biogas will be produced. If protein-rich wastewater is treated and the first compartment is operated under moderately high pH conditions, i.e. pH 7.5-8, then such stripping effect also might be beneficial to remove NH3 from the solution. It is known that thermophilic methanogenesis is very sensitive to relatively low concentrations of free NH3 (Angelidaki & Ahring 1994; Zeeman et al. 1985). The staged degradation of organic compounds results in a segregated development of methanogenic sludge along the reactor height (Van Lier et al., 1996). In the first compartment of the reactor treating the sucrose-VFA mixture, the sludge was whitish and fluffy, likely consisting of acidifying biomass. Sludge granules grown in the top part of this reactor possessed the highest acetogenic and methanogenic activity, and the highest granule strength as well. Apparently, the sludge in each compartment will differ depending on the specific environmental conditions prevailing there and the remaining compounds to be degraded. Since mixing of the sludge over the total system is preventod, from compartment to compartment a specific sludge will develop. Long term experiments revealed that the conversion of the sucrose-VFA mixture into methane gradually deteriorated at prolonged operation under high-rate conditions. The sludge bed in the higher compartments became also whitish and fluffy, similar to the sludge cultivated in the first compartment. Likely, the acidifying sludge gradually replaced the methanogenic granules in the system. In these experiments, the excess methanogenic sludge produced was collected from the top part of the reactor once the sludge bed had reached the effluent overflow. Obviously, a stable reactor performance on the long term can only be achieved by preserving the sludge segregation in the reactor. Otherwise the acetogenic and methanogenic activity of the sludge in successive compartments will drop to low levels. In using an upflow reactor, sludge withdrawal from one extraction point, preferentially at the bottom of the reactor, very likely will suffice to ensure stable reactor performance. It should be mentioned that the above reactor set-up was particularly developed for high-rate thermophilic anaerobic wastewater treatment. Conventional reactor systems can be used without any problems if ther- mophilic treatment is applied for other reasons than aiming at the highest possible loading rate, e.g. for dissolving specific compounds and/or 'pasteurization' of the digested residue (Aitken & Mullennix 1992; Bendixen 1994). C o n c l u s i o n s The effect of temperature fluctuations on the process stability of thermophilic wastewater treatment systems is most severe if continuous stirred tank reactors and/or batch reactors are used. These type of systems may be characterized by a very narrow temperature range for methanogenesis. While a temperature decrease immediately affects the conversion capacity of a high loaded CSTR system, an increase in the process temperature may result in complete reactor failure. In contrast, the thermostability in sludge retention systems is much higher which can be attributed to: - the presence of one single temperature optimum at about 60 ~ irrespective of a lower cultivation temperature; and - the formation of immobilized biomass which leads to a certain degree of substrate diffusion limitation, and consequently, to the creation of a 'biomass buffer' which can be drawn on when the temperature suddenly drops. The maximum applicable temperature in thermophilic wastewater treatment systems tentatively can be set at 60 ~ Although methanogenesis from acetate is possible at temperatures up to 75 ~ other conversion steps, such as propionate oxidation, hardly proceed under extreme thermophilic conditions. Thermophilic wastewater treatment reactors are often limited by high concentrations of VFA in the effluent which can be attributed to: higher intrinsic half saturation constants of thermophilic bacteria; increased half saturation constants due to biomass immobilization; - toxicity effects which may limit the conversion rate of specific (intermediate) compounds. As a result, high organic loading rates are often accompanied with decreased removal efficiencies, particularly when industrial, non-synthetic, wastewater is fed to the reactor. A more stable reactor performance under extreme loading conditions is achieved when a staged reactor is applied, where the various steps in the anaerobic conversion of organic matter are separated. In such system the produced biogas is evenly withdrawn - - 12 o v e r t h e r e a c t o r h e i g h t , w h i c h r e s u l t s in a n e n h a n c e d s l u d g e r e t e n t i o n capacity. Acknowledgements This work was financially supported by the Ministry of VROM, Novem-grant 51120/1510, The Netherlands a n d P a q u e s B.V. B a l k , T h e N e t h e r l a n d s . T h e a u t h o r w i s h e s to t h a n k Dr. J o s 6 L u i s S a n z f o r his h e l p w i t h t h e 75 o C - s l u d g e e x p e r i m e n t s . References Ahring BK (1991) Methanogenesis during thermophilic anaerobic digestion with focus on acetate. In: Verachtert H & Verstraete W (Eds) Proceedings of the International Symposium on Environmental Biotechnology, part 1, Oostende (Belgium), Koninklijke Vlaamse Ingenieurs-verenigingVZW, Antwerp, Belgium (pp 275-283) Ahring BK (1992) Turn-over of acetate in hot springs at 70 ~ In: Thermophiles: Science and Technology, International Conference, Reykjav~, Iceland, 23rd-26th August 1992, IceTec Publ. p 130 Ahring BK (1994) Status on science and application of thermophilic anaerobic digestion. Wat. Sci. Tech. 30(12): 241-249 Ahring BK (1995) Methanogenesis in thermophilic bioreactors. Antonie van Leeuwenhoek 67:91-102 Ahring BK, Rintala J, Nozhevnikova AN & Mathrani IM (1995) Metabolism of acetate in thermophilic (55 oC) and extreme thermophilic (70 ~ UASB granules. In: Proceedings of International Meeting on: Anaerobic Processes for Bioenergy and Environment, Copenhagen, 25-27 January, 1995 Ahring BK, Schmidt JE, Winther-Nielsen M, Macario AJL & Conway de Macario E (1993) Effect of medium composition and sludge removal on the production, composition and architecture of thermophilic (55 o C) acetate-utilizing granules from an upflow anaerobic sludge blanket reactor. Appl. Environ. Microbiol. 59: 2538-2545 Ahring BK & Westermann, P (1985) Methanogenesis from acetate: physiology of a thermophilic acetate-utilizing methanogenic bacterium. FEMS Microbiol. Lett. 28:15-19 Aitken MD & Mullennix RW (1992) Another look at thermophilic anaerobic digestion of waste water sludge. Water Environ. Res. 64:915-919 Angelidaki I & Ahring BK (1994) Anaerobic thermophilic digestion of manure at different ammonia loads: effect of temperature. Wat. Res. 28:727-731 Aoki N & Kawase M (1991) Development of high-performance thermophilic two-phase digestion process. Wat. Sci. Technol. 23 (7-9): 1147-1156 Bendixen HJ (1994) Safeguards against pathogens in Danish biogas plants. Wat. Sci. Tech. 30(12): 171-180 Braun R & Huss S (1982) Anaerobic digestion of distillery effluents. Process Biochem. 17-4:25-27 Buhr HO & Andrews JF (1977) The thermophilic anaerobic digestion process. War. Res. 11:129 143 Cail R & Barford JP (1985) Thermophilic semi-continuous anaerobic digestion of palm-oil mill effluent. Agr. Wastes 13:295-304 Clarens M & Moletta R (1990) Kinetic studies of acetate fermentation by Methanosarcina sp. MSTA-1. Appl. Microbiol. Biotechnol. 33, 239-244 De Zeeuw W (1984) Acclimatization of anaerobic sludge for UASB reactor start-up. Ph.D. thesis, Department of Environmental Technology, Agricultural University, Wageningen, The Netherlands Disley RS, Walmsley MJ & Forster CF (1992) Inhibition of gas production by thermophilic anaerobic sludges: The effect of organic compounds. Environ. Technol. 13:1153-1159 Duff SJB & Kennedy KJ (1982) Effect of hydraulic and organic overloading on thermophilic down flow stationary fixed film (DSFF) reactor. Biotechnol. Lett. 4:815-820 Fernandez N & Forster CF (1993) A study of the operation of mesophilic and thermophilie anaerobic filters treating a synthetic coffee waste. Biores. Technol. 45:223-227 Fukuzaki S, Nishio N, Shobayashi M & Nagai S (1990) hthibition of the fermentation of propionate to methane by hydrogen, acetate, and propionate. Appl. Environ. Microbiol. 56:719-723 Garber WE Ohara GT, Colbaugh JE & Raksit SK (1975) Thermophilic digestion at the Hyperion treatment plant. J. Wat. Poll. Contr. Fed. 47: 950-961. Ghosh S, Klass DL Christopher RW & Edwards VH (1980) Thermophilic biogasification of biomass. In: Proc. 7 th Energy Technol. Conf. and Expo., March 24-26, Washington DC Good P, Moundry R & Fluff P (1982) Use of fixed film and CSTR reactor for anaerobic treatment of stillage of wood hydrolysate. Biotechnol. Lett. 4:595-600 Grotenhuis JTC, Smit M, Plugge CM, Xu Y, Lammeren AAM van, 'Stares AJM & Zehnder AJB (1991) Bacteriological composition and structure of granular sludge adapted to different substrates. Appl. Environ. Microbiol. 57:1942-1949 Grotenhuis JTC, Stams AJM & Zehnder AJB (1992) Hydrophobicity and electrophoretic mobility of anaerobic isolates from methanogenic granular sludge. Appl. Env. Microbiol. 58: 10541056 Harris WL & Dague RR (1993) Comparative performance of anaerobic filters at mesophilic and thermophilic temperatures. Water Environ. Res. 65:764-771 Heitzer A, Kohler H-PE, Reichert P & Hamer G ( 1991 ) Utility of phenomenological modeis for describing temperature dependence of bacterial growth. Appl. Environ. Microbiol. 57:2656-2665 Hulshoff Pol LW (1989) The phenomenon of granulation of anaerobic sludge. Ph.D. Thesis, Department of Environmental Technology, Agricultural University, Wageningen, The Netherlands Huser BA, Wuhrmann K & Zehnder AJB (1982) Methanothrix soehngenii gen. nov. sp. nov., a new acetotrophic non-hydrogen oxidizing methanobacterium. Arch. Mierobiol. 132:1--9 Kaiser SK, Harris WL & Dague RR (1993) Initial studies on the temperature phased anaerobic biofilter process. In: Proe. of 66th Annual Conference & Exposition of Water Environment Federation, October 3-7 1993, Anaheim, California USA, pp 319-329 Kamagata Y & Mikami E (1991) Isolation and characterization of a novel thermophilic Methanosaeta strain. Int. J. Syst. Bacteriol. 41:191-196 Kawase M, Nomura T & Majima T (1989) An anaerobic fixed bed reactor with a porous ceramic carrier. Wat. Sci. Technol. 21-4/5: 77-86 Kennedy KJ & Van den Berg L (1982) Thermophilic downflow stationary fixed film reactors for methane production from bean bleaching waste. Biotechnol. Lett. 4:171-176 Kida K, lkbal & Sonoda Y (1992) Treatment of coffee waste by slurry-state anaerobic digestion. J. Ferm. Bioeng. 73:390-395 Lanting J, Jordan JA, Schone MT, Kull A, Carey WW & Kitney BL (1989) Thermophilic anaerobic digestion of coffee wastewater. 13 In: Dalton CS & Wukasch RF (Eds) Proceedings of 43rd Industrial Waste Conference, May 1988, Lafayette, Indiana, USA. (pp 513-524) Lewis Publishers, Chelsea Michigan Lema JM, Soto M, Mdndez R & Bl~izquez R (1988) Comparison of mesophilic and thermophilic filters treating very high saline wastewaters, In: Poster-Proceedings of 5th International Symposium on Anaerobic Digestion, Bologna Italy, May 22-26, (pp 547-549) Lens PM, De Beer D, Cronenberg CH, Houwen FE Ottengraf SPP & Verstraete W (1993) Heterogenous distribution of microbial activity in methanogenic aggregates: pH and glucose microprofiles. Appl. Environ. Microbiol. 59:3803-3815 Macleod C & Forster CF (1988) Heavy metal inhibition of gas production by thermophilic anaerobic sludges. Microbiol. 54: 31-40 Morvai L, Mihaltz P & Hollo J (1992) Comparison of the kinetics of acetate biomethanation by raw and granular sludges. Appl. Microbiol. Biotechnol. 36:561-567 Nozhevnikova AN & Chudina VI (1984) Morphology of the thermophilic acetate bacterium Methanothrix thermoacetophila sp. nov.. Microbiology 53:618-624 Ohtsuki T, Watanabe M & Miyaji Y (1992) Start-up of thermophilic UASB (upflow anaerobic sludge blanket) reactors using microcarrier and mesophilic granular sludge. Wat. Sci. Tech. 26(3/4): 877-886 Ohtsuki T, Tominaga S, Morita T & Yoda M (1994) Thermophilic UASB system start-up and management-change in sludge characteristics in the start-up procedure using mesophilic granular sludge. In: Paper Pre-prints of the Seventh International Symposium on Anaerobic Digestion, Cape Town, January 23-27, 1994, South Africa (pp 348-357) Ollivier B, Lombardo A & Garcia JL (1984) Isolation and characterization of a new thermophilic Methanosarcina strain (strain MP). Ann. Microbiol. (Inst. Pasteur) 135B: 187-198 Patel GB (1984) Characterization and nutritional properties of Methanothrix concilii sp.nov., a mesophilic aceticlastic methanogen. Can. J. Microbiol. 30:1383- 1396 Patel GB & Sprott GD (1990) Methanosaeta concilii gen. nov., sp. nov. ("Methanothrix concilii") and Methanosaeta thermoacetophila nom. rev., comb. nov.. Int. J. Syst. Bacteriol. 40:79-82 Pavan P, Musacco A, Cecchi E Bassetti A, & Mata-Alvarez J (1994) Thermophilic semi-dry anaerobic digestion process of the organic fraction of municipal solid waste during transient conditions. Environ. Technol. 15:1173-1182 Pavlostathis SG & Giraldo-Gomez E (1991) Kinetics of anaerobic treatment: a critical review. Critical Rev. in Environ. Control. 21 (5,6): 411-490 Petersen SP & Ahring BK (1991) Acetate oxidation in a thermophilic anaerobic sewage-sludge digestor: the importance of non-aceticlastic methanogenesis from acetate. FEMS Microbiol. Ecol. 86:149-158 Perry RH & Green DW (eds) (1984) Perry's chemical engineers' handbook, section 14, 6th edition, McGraw-Hill Publishing Co., New York, USA Rintala JA & Lepist6 SS (1992) Anaerobic treatment of thermomechanical pulping whitewater at 35-70 ~ Wat. Res. 26: 1297t305 Rintala JA, Lepist0 SS & Ahring BK (1993) Acetate degradation at 70 ~ in upflow anaerobic sludge blanket reactors and temperature response of granules grown at 70 ~ Appl. Environ. Microbiol. 59:1742-1746 Rudd T, Hicks SJ & Lester JN (1985) Comparison of the treatment of a synthetic meat waste by mesophilic and thermophilie anaerobic fluidized bed reactors. Environ. Technol. Lett. 6:209-224 Schmidt JE & Ahfing BK (1993) Effects of hydrogen and formate on the degradation of propionate and butyrate in thermophilic granules from an Upflow Anaerobic Sludge Blanket reactor. Appl. Environ, Microbiol. 59:2546-255 l Schmidt JE & Ahring BK (1994) Extracellular polymers in granular sludge from different upflow anaerobic sludge blanket (UASB) reactors. Appl, Microbiol. Biotechnol. 42:457-462 Sch6nheit P, Moll J & Thauer RK (1980) Growth parameters (Ks, #raaz, Ys) of Methanobacterium thermoautotrophicum. Arch. Microbiol. 127:59-65 Schraa G (1983) Conversion of soluble organic matter with the thermophific anaerobic attached film expanded bed process. PhD thesis, Comell Univ., Ithaca, NY, USA Schraa G & Jewell WJ (1984) High rate conversions of soluble organics with the thermophilic anaerobic attached expanded bed. J. Wat. Poll. Contr. Fed. 56:226-232 Seif HAA, Joshi SG & Gupta SK (1992) Effect of organic load and reactor height on the performance of anaerobic mesophilic and thermophilic fixed film reactors in the treatment of pharmaceutical wastewater, Environ. Technol. 13:1161-1168 Smith JM (1981 ) Chemical Engineering Kinetics. McGraw-Hill, Inc. USA Smith MR & Mah RA (1978) Growth and methanogenesis by Methanosarcina strain 227 on acetate and methanol. Appl. Environ. Microbiol. 36:870-879 Soto M, Mdndez R & Lema JM (1992) Characterization and comparison of biomass from mesophilic and thermophilic fixed bed anaerobic digesters. Wat. Sci. Technol. 25-7:203-212 Sohza ME, Fuzaro G & Polegato AR (1992) Thermophilic anaerobic digestion of vinasse in pilot plant UASB reactor. Wat. Sci. Teehnol. 25(7): 213-222 Touzel JP, Petroff D & Albagnac G (1985) Isolation and characterization of a new thermophitic Methanosarcina, the strain CHTI 55. Syst. Appl. Microbiol. 6: 66-7l Uemura S & Harada H (1993) Microbial characteristics of methanogenic sludge consortia developed in thermophilic UASB reactors. Appl. Mierobiol. Biotechnol. 39:654-660 Uemura S & Harada H (1995) Inorganic composition and microbial characteristics of methanogenic granular sludge grown in a thermophilic upflow anaerobic sludge blanket reactor. Appl. Microbiol. Biotechnol. 43:358-364 Ugurlu A & Forster CF (1992) The impact of shock loadings on the performance of thermophilic anaerobic filters with porous and non-porous packings. Biores. Technol. 39:23-30 Vanderhaegen B, Ysebaert E, Favere K, Van Wambeke M, Peeters T, Panic V, Vandenlangenbergh & Verstraete W (1992) Acidogenesis in relation to in-reactor granule yield. Wat. Sci. Technol. 25: 75-81 Van Lier JB, Boersma F, Debets MMWH & Lettinga G (1994) High-Rate thermophilic anaerobic wastewater treatment in compartmentalized upflow reactors. Wat. Sci. Technol. 30(12): 251261 Van Lier JB, Hulsbeek J, Stares AJM & Lettinga G (1993a) Temperature susceptibility of thermophilic methanogenic sludge: implications for reactor start-up and operation. Biores. Technol. 43: 227-235 Van Lier JB, Groeneveld N & Lettinga G (1996) Characteristics and development of thermophilic methanogenic sludge in compartmentalized upflow reactors. Biotechnol. Bioeng. (Accepted) Van Lier JB, Grolle KCF & Lettinga G (1991) Anaerobic digestion at 75 oC. In: Poster-Abstracts of the Sixth International Symposium on Anaerobic Digestion, Sao Paulo, Brazil, May 12-16, p 163 Van Lier JB, Grolle KCF, Frijters CTMJ, Stams AJM & Lettinga G (1993b) Effects of acetate, propionate and butyrate on the ther- 14 mophilic anaerobic degradation of propionate by methanogenic sludge and defined cultures. Appl. Environ. Microbiol. 59:10031011 Van Lier JB, Grolle KCF, Stams AJM, Conway de Macario E & Lettinga G (1992) Start-up of a thermophilic Upflow Anaerobic Sludge Bed (UASB) reactor with mesophilic granular sludge. Appl. Microbiol. Biotechnol. 37: 130-135 Van Lier JB, Macario AJL, Conway de Macado E & Lettinga G (1993c) Permanent increase of the process temperature of mesophilic Upflow Anaerobic Sludge Bed (UASB) reactors to 46, 55, 64 and 75 ~ In: Dalton CS & Wukasch RF (Eds), Proceedings of 47th Industrial Waste Conference, May 1992, Lafayette Indiana, USA (pp 445-459)Lewis Publishers, Chelsea Michigan, USA Van Lier JB, Sanz Martin JL & Lettinga G (1996) Effect of temperature on the anaerobic thermophilic conversion of volatile fatty acids by dispersed and granular sludge. War. Res. 30:199-207 Varel VH, Isaacson HR & Bryant MP (1977) Thermophilic methane production from cattle waste. Appl. Environ. Microbiol. 33:298307 Weber H, Kulbe KD, Chmiel H & Trfsch W (1984) Microbial acetate conversion to methane: kinetics, yields and pathways in a twostep digestion process. Appl. Microbiol. Biotechnol. 19:224-228 Westermann P (1994) The effect of incubation temperature on steady-state concentrations of hydrogen and volatile fatty acids during anaerobic degradation in slurries from wetland sediments. FEMS Microbiol. Ecol. 13:295-302 Wiegant WM (1986) Thermophilic anaerobic digestion for waste and wastewater treatment. PhD. thesis, Department of Environmental Technology, Agricultural University, Wageningen, The Netherlands Wiegant WM, Claassen JA & Lettinga G ( 1985) Thermophilic anaerobic digestion of high strength wastewaters. Biotechnol. Bioeng. 27:1374-1381 Wiegant WM & De Man AWA (1986) Granulation of biomass in thermophilic anaerobic sludge blanket reactors treating acidified wastewaters. Biotechnol. Bioeng. 28:718-727 Wiegant WM, Hennink M & Lettinga G (1986) Separation of the propionate degradation to improve the efficiency of thermophilic anaerobic treatment of acidified wastewaters. Wat. Res. 4: 517524 Wiegant WM & Lettinga G ( 1985) Thermophilic anaerobic digestion of sugars in upflow anaerobic sludge blanket reactors. Biotechnol. Bioeng. 27:1603-1607 Yang M, Ishihara T, Okada M, Nagai S & Snnahara H (1992) Stability and performance of thermophilic anaerobic fixed-bed reactor packed with a saddle-shaped slag biocarrier. Environ. Technol. 13:671--678 Zeeman G, Wiegant WM, Koster-Treffers ME & Lettinga G (1985) The influence of the total ammonia concentration on the thermophilic digestion of cow manure. Agr. Wastes 14:19-35 Zehnder AJB, Huser BA, Brock TD & Wuhrmann K (1980) Characterization of an acetate-decarboxylating non-hydrogen oxidizing methane bacterium. Arch. Microbiol. 124:1-11 Zehnder AJB, Ingvorsen K & Marti T (1982) In: Hughes DE et al. (Eds), Anaerobic Digestion 1981 (pp 45-68). Elsevier Biomedical Press, Amsterdam Zinder SH (1986) Thermophilic waste treatment systems. In: Brock TD (Ed) Thermophiles: general, molecular and applied biology (pp 257-277). Wiley-lnterscience, New York Zinder SH (1990) Conversion of acetic acid to methane by thermophiles. FEMS Microbiol. Rev. 75:125-138 Zinder SH, Anguish T & Cardwell SC (1984) Effects of temperature on methanogenesis in a thermophilic (58 oC) anaerobic digester. Appl. Environ. Microbiol. 47:808-813 Zinder SH, Anguish T & Lobo AL (1987) Isolation and characterization of a thermophilic acetotrophic strain of Methanothrix. Arch. Microbiol. 146:315-322 Zinder SH & Koch M (1984) Non-aceticlastic methanogenesis from acetate: acetate oxidation by a thermophilic syntrophic coculture, Arch. Microbiol. 54:263-272 Zinder SH & Mah RA (1979) Isolation and characterization of a thermophilic strain of Methanosarcina unable to use H2-CO2 for methanogenesis. Appl. Environ. Microbiol. 38:996-1008 Zinder SH, Sowers KR & Ferry JG (1985) Methanosarcina thermophila sp. nov., a thermophilic, acetotrophic, methaneproducing bacterium. Int. J. Syst. Bact. 35:522-523