Potential of anaerobic digestion of complex waste(water)

G. Zeeman and W. Sanders Section Environmental Technology, Wageningen University, PO Box 8129, 6700 EV Wageningen, The Netherlands. (E-mail: [email protected]) Abstract Although they differ greatly in origin complex waste(water)s mainly consist of proteins, lipids, carbohydrates and sometimes lignin in addition. Hydrolysis is the first and generally rate-limiting step in the process of anaerobic digestion of particulate organic substrates. Hydrolysis of particulate polymers can be described by Surface Based Kinetics, but for use in practice the empirical first order relation is advised. Unlike the hydrolysis of protein and carbohydrate, lipid hydrolysis is hardly occurring in the absence of methanogenesis. The latter is probably a physical rather than a biological process and affects the choice for either a one- or a two-step (phase) anaerobic reactor. In the chain of collection and transport, complex wastes often become complex wastewaters simply because of dilution. Dilution not only changes the reactor technology to be applied but also complicates the post-treatment and possibilities for resource recovery. Combining concentrated with diluted waste streams will almost always end up in much more complicated treatment technologies. Keywords Anaerobic digestion; carbohydrates; complex waste(water); domestic sewage; faeces; hydrolysis; lipids; particulate; polymers; primary sludge; proteins; sanitation concepts; suspended solids; urine; waste activated sludge

Introduction

Complex wastewaters and wastes can be both defined as substrates containing a large fraction of suspended solids. The main difference between the two, is determined by the total solids (TS) and therewith the total COD concentration. The latter results in a different approach in treatment technology to be applied. This paper will focus on both the resemblance and the difference in the treatment processes and the biological and physical aspects involved. Moreover post-treatment and or reuse necessities or options will be discussed. The paper shows that considering the whole sanitation chain, instead of only considering end of pipe technology can facilitate the final treatment. In such an approach it becomes clear that many wastewaters originally were produced as concentrated wastes. The dilution complexes the subsequent treatment and often prevents reuse of energy and nutrients. Among the main components of complex waste(water), viz., carbohydrates, lipids and proteins, carbohydrates are known to be easily and rapidly converted via hydrolysis to simple sugars and subsequently fermented to volatile fatty acids (VFA) (Cohen, 1982, Miron et al., 2000). Protein is hydrolysed to amino acids and further degraded to VFA either through anaerobic oxidation linked to hydrogen production or via fermentation according to the Stickland reaction (McInerney, 1988). The former is dependent on the presence of hydrogen-scavengers while the latter is independent of the methanogenic activity in the reactor (Nagase and Matsuo, 1982). Among the lipids, triglycerides are hydrolysed to long chain fatty acids (LCFA) and further oxidised via b oxidation to acetate or propionate. Accumulation of hydrogen inhibits the b oxidation (Novak and Carlson, 1970) since it is thermodynamically unfavourable under standard conditions. b oxidation only occurs when the hydrogen partial pressure is kept low by the presence of hydrogen-scavengers. Several authors report that the presence of methanogenic activity also enhances hydrolysis of lipids (Palenzuela-Rollon,1999, Miron et al., 2000, Sanders, 2001). This paper will discuss the hydrolysis mechanism and kinetics for particulate polymers and will hypothesise on the

Water Science and Technology Vol 44 No 8 pp 115–122 © IWA Publishing 2001

Potential of anaerobic digestion of complex waste(water)

115

mechanism of the observed difference in lipid hydrolysis at methanogenic and acidogenic conditions. Recommendations for the operational design of both complex wastes and wastewaters are presented. Composition and concentration G. Zeeman and W. Sanders

Complex waste(water)s which can be (pre)treated anaerobically are amongst others; slaughterhouse waste(water) (Sayed 1987, Batstone, 2000), fish processing waste(water) (Palenzuela-Rollon, 1999), faeces plus urine (Zeeman et al., 2000), domestic sewage (Elmitwalli, 2000), primary sludge (Miron et al., 2000), dairy (waste)water (Zeeman et al., 1997), waste activated sludge (Zeeman et al., 1997), animal manure (Zeeman, 1991) and the organic fraction of municipal solid waste (Ten Brummeler, 1993). Although they differ greatly in origin these complex waste(waters) all consist of proteins, lipids, carbohydrates and sometimes lignin in addition. The proteins and carbohydrates in waste activated sludge do mainly persist in the aerobic biomass cells (Zeeman et al., 1997). Hydrolysis

Hydrolysis is the first step in the process of anaerobic digestion of particulate organic substrates. During hydrolysis, polymeric compounds are converted into soluble monomeric or dimeric substrates. Many authors describe the hydrolysis with first order kinetics based on biodegradable substrate at constant pH and temperature (Pavlostathis and Giraldo-Gomez, 1991): dXdeg r . dt

= - kh . Xdeg r .

(1)

with: Xdegr = concentration degradable substrate (kg/m3), t = time (days), kh = first order hydrolysis constant (1/day). The first order relation is however an empirical relation and even when the reactor conditions and substrate type are kept constant, different kh values can be found due to changes in the particle size distribution of the substrate (Hills and Nakano, 1984; Chyi and Dague, 1994). To gain more insight into the hydrolysis process some authors have tried to develop and/or verify a deterministic model for the anaerobic hydrolysis (Hills and Nakano, 1984, Hobson, 1987, Vavilin et al., 1996). In this model, it is assumed that the substrate particles are completely covered with bacteria that secrete an excess of hydrolytic exo-enzymes during digestion. The hydrolysis rate is therefore constant per unit area available and the hydrolysis constant, Ksbk, is not affected by the particle size of the substrate. The model will further be referred to as the Surface Based Kinetics (SBK) model (formula 2). dM = - K sbk * A dt

(2)

with: M = mass of substrate (kg), t = time (days) Ksbk = surface based hydrolysis constant (kg/m2 day), A = surface available for hydrolysis (m2).

116

Sanders et al. (2000) present a mathematical description of the surface related hydrolysis kinetics for spherical particles in a batch digestion and a verification of this model with

Lipids

The results of different investigations into digestion of lipid containing waste(water), like primary sludge, fish wastewater and palm oil in CSTR and batch experiments, show that unlike the hydrolysis of protein and carbohydrate, lipid hydrolysis is hardly occurring in the absence of methanogenesis (Miron et al., 2000, Palenzuela-Rollon, 1999, Sanders et al., 2001). Sanders et al. (2001) illustrate that the lower volumetric lipid hydrolysis rate during batch digestion under acidogenic conditions could not be ascribed to a low pH, accumulation of LCFA or accumulation of di- or mono-glyceride. The results only indicate a (possible non-causal) relation between the hydrolysis rate and the hydrogen concentration. Calculations indicate that the amount of surface available for the hydrolysis is an important factor determining the difference in the volumetric hydrolysis rates between acidogenic and methanogenic lipid digestion. The larger available surface for the hydrolysis under methanogenic conditions as compared to acidogenic conditions could be caused by the emulsifying effect of the gas production. The latter has important consequences for the design of anaerobic reactors treating complex waste(water) containing a considerable fraction of lipid-COD. Reactor design aspects are discussed below.

G. Zeeman and W. Sanders

particulate starch as a substrate. CH4 production and particle size distribution (PSD) were determined in time during batch digestion of two starch fractions with different PSD and fitted with the model. The theoretical PSD, calculated based on CH4 production, shows good similarity with the experimental PSD, proving that the SBK model is very well capable to describe the anaerobic hydrolysis of particulate substrates and that the amount of substrate surface available for hydrolysis is the essential factor determining the hydrolysis rate. Hydrolysis of particulate polymers can be described by Surface Based Kinetics, but for use in practice the determination of the available surface is so complicated that the empirical first order relation is advised.

Reactor systems With or without biomass retention

Anaerobic treatment systems can in general be divided into systems with and without sludge/biomass retention. When considering complex waste(water), the systems without sludge retention are applied for more concentrated waste streams or slurries, e.g. faeces plus urine, primary sludge, waste activated sludge or animal manure. The CSTR (completely stirred tank reactor) is the most generally applied system for slurry digestion. The main characteristic of a CSTR system is that its SRT (sludge retention time) is equal to its HRT (hydraulic retention time). In general, mesophilic CSTR systems can be applied when the slurry is so concentrated that it will provide enough biogas to produce the energy for heating the system. The higher the concentration, the more surplus energy is produced for other applications. At low temperature climates, like in Western Europe, it is forbidden to apply fertilisers on the field during the winter period. When the digested slurry is applied on the fields as a fertiliser, storage of 3–5 months will be necessary to overcome this winter period. In that situation, combined storage and digestion in a fed-batch or accumulation system (AC) at ambient temperatures, is a feasible alternative for a CSTR system (Zeeman, 1991; Zeeman and Lettinga, 1999). For complex wastewater’s like, fish-, slaughterhouse-, dairy-wastewater and domestic sewage, anaerobic systems with sludge retention can be applied, characterised by a much longer SRT in comparison with the HRT. The UASB reactor is the most frequently applied system with biomass retention. For the application of UASB systems for the treatment of wastewater with a high fraction of suspended solids, both physical and biological processes will determine the final removal efficiency and conversion of organic compounds to

117

methane gas. Removal of suspended solids occurs by physical processes such as settling, adsorption and entrapment. The subsequent hydrolysis and methanogenesis of the removed solids depends on the process temperature and the prevailing SRT. Zeeman and Lettinga (1999) present a model for the calculation of the minimal HRT when a certain SRT is a prerequisite (formula 3). G. Zeeman and W. Sanders

HRT = (C*SS/X)*R*(1–H)*SRT,

(3)

where HRT = hydraulic retention time, C = CODinfluent concentration (g/l); SRT = sludge retention time (days); R = fraction of the CODSS removed; H = fraction hydrolysed of removed solids; X = sludge concentration in the reactor (g COD/l); SS = CODSS/CODinfluent and no distinction is made between the fraction of CODSS that is removed but not hydrolysed and the biomass yield. Long SRTs are needed to provide a sufficient amount of hydrolysis and methanogenesis, especially at low temperature conditions. Miron et al. (2000) show that an SRT of at least 10 days is necessary to provide methanogenesis in the anaerobic treatment of primary sludge at a process temperature of 25°C while a SRT of 15 days is necessary for sufficient hydrolysis and acidification of lipids. For temperatures as low as 15°C, an SRT of at least 75 days has to be provided to achieve methanogenic conditions (Zeeman et al., 2000). According to Wiegant (2001) the hydraulic design criteria, superficial biogas velocity or the solids retention time (SRT) determines the design of the reactor compartment of a UASB reactor. For domestic sewage treatment at temperatures of 18–25°C and minimum SRTs of 45 and 31 days respectively, it is calculated that at influent TSS concentrations < 375 and 500 mg/l, respectively, the hydraulic design criteria, rather than the criteria pertaining to the solids retention time, are governing the design of UASB reactors. At high TSS concentrations and/or low temperature conditions the SRT will become the design criterion. In arid climates, with low water consumption, as in Jordan, influent sewage concentrations are as high as 1,500 mg COD/l, with a suspended fraction of 75%. Moreover temperatures can in winter decrease to ca. 15°C. Under such conditions, SRT rather than hydraulic conditions will determine the design of the reactor. Also for fish and slaughterhouse wastewaters with high concentrations of SS the latter becomes the case even at higher temperature conditions (Palenzuela-Rollon, 1999). One or two step (phase) systems

118

When wastewater is treated in an up-flow reactor the particles become part of the sludge fraction whereas the dissolved components remain part of the water fraction. Therefore to achieve sufficient hydrolysis of the particles the SRT is the most important design parameter. For the hydrolysis of the dissolved polymers the hydraulic retention time is the most important design parameter (Sanders, 2001). High concentrations of suspended solids in the influent make high volumetric loading rates in one-stage anaerobic reactors virtually impossible, especially when low temperatures prevail (Elmitwalli, 2000) or lipids are involved (Palenzuela-Rollon, 1999). Lipids can cause severe problems during the digestion. Namely, accumulation of free LCFA and possible LCFA toxicity, due to the high lipid-water interface of the lipid emulsions (Batstone, 2000). Secondly, formation of scum layers and wash out of sludge, due to the low specific density of the lipids can occur. Palenzuella-Rollon (1999) advises to remove lipids from the wastewater prior to anaerobic treatment to achieve a better process stability. The application of a highly loaded first step UASB system for this purpose was explored. The system is similar to the upflow anaerobic solid removal (UASR) system as used by Zeeman et al. (1997) in the pre-treatment of waste activated sludge, domestic sewage and dairy wastewater, a similarly lipid-rich wastewater.


Unlike a conventional clarifier, the influent wastewater in this reactor passes through the sludge bed where the SS can be entrapped or adsorbed, and partially hydrolysed and acidified depending on the temperature and solid retention time (SRT). The influent flow constantly flushes the sludge bed, hence preventing possible accumulation of intermediate products, such as amino acids (Doi, 1972; Glenn, 1976), which could otherwise be inhibitory to hydrolysis or acidification. Elmitwalli (2000) recommends the application of an Anaerobic Filter (AF) for the removal of SS prior to a methanogenic Anaerobic Hybrid reactor for the treatment of domestic sewage at low temperature conditions (13°C). The AF demonstrates much higher removal efficiency (80%) for CODss as compared to the high loaded UASB (Zeeman et al., 1997). The total system achieves total COD removal efficiencies of ca 70% at 13°C and a HRT of 4+8 hours, which is similar to those at tropical conditions (Wiegant, 2001). The sludge produced in the first step of such a two-step system is hardly stabilised. The prevailing short SRT in the first step only provides some protein and carbohydrate hydrolysis while the absence of methanogenic activity limits the lipid hydrolysis. Application of sludge digestion at elevated temperatures can achieve stabilisation of the produced sludge and recovery of CH4 gas. While lipid-containing wastewater is advisable to treat in a two step system, concentrated slurries and waste with high lipid concentration should preferably treated in a one-stage slurry digester, for two reasons. (1) Lipids will not be hydrolysed in absence of methanogenic activity. (2) The possible decrease of the lipid-water interface in the first stage of a two-stage sludge digester can even result in a longer required SRT in the second stage. Nevertheless, several authors recommend applying a two-phase instead of a one-phase system in order to promote the hydrolysis step (Ghosh et al., 1995; Perot et al., 1989). Apart from the negative effect on the lipid hydrolysis it is however also shown that hydrolysis and acidification of proteins and carbohydrates are not promoted by acidogenic conditions (Miron et al., 2000). Acidogenic conditions even negatively effect protein hydrolysis as a result of the low pH (Palenzuela-Rollon, 1999). Therefore the application of a two-phase system will not improve efficiencies, when hydrolysis is recognised as the rate-limiting step. Concepts for collection, transport and treatment

Comparing the anaerobic treatment of complex wastewaters and wastes it becomes clear that for both substrates, lipids, proteins and carbohydrates are the main components to be converted. Often complex wastes become complex wastewaters simply because of dilution somewhere in the chain of collection and transport. Not only the anaerobic reactor technology to be applied will change as a result of the degree of dilution but also the post treatment needed and/or the possibilities for reuse will change tremendously. The latter is very obvious in the chain of collection, transport and treatment and reuse of domestic waste(water). The substrates, faeces plus urine, domestic sewage, primary sludge and waste activated sludge, are very much related in origin. The applied collection, transport and treatment have however resulted in totally different substrates with respect to either composition or concentration, with large consequences for the treatment technologies to be applied. The latter is illustrated in Figure 1. The scheme in Figure 1 illustrates the interactions between the 4 types of waste(water). While faeces plus urine are diluted with flushing-, shower and bath-, kitchen- and laundry water, producing domestic sewage, primary sludge is in its turn produced by settling domestic sewage. These subsequent dilutions and concentrations result in the production of pre-settled sewage which is polluted with nitrogen, phosphorus, potassium and moreover pathogens. As the latter stream is very much diluted, recovery and reuse of the present resources becomes in general very complicated, except for reuse in tropical agriculture combined with fertilisation. Though anaerobic digestion of domestic sewage is applied on a

119

39 litres toilet flushing

52 litres shower/ bath

42 litres Laundry, kitchen, others

Rain


133+ litres domestic sewage

1.2 l i t r e s faeces plus urine

P r esettled sewage

Primary sludge

oxygen

Waste activated sludge

Figure 1 Interactions between the 4 types of domestic waste(water)s, faeces plus urine, domestic sewage, primary sludge and waste activated sludge. Water consumption figures are mean values for the Netherlands (Zeeman et al., 2001)

120

large scale in tropical countries (Wiegant, 2001), aerobic treatment is so far the most frequently applied treatment technique. The latter results in the production of large amounts of waste activated sludge that is very much resistant to further anaerobic stabilisation, due to the limited dying and lysis of aerobic cells under anaerobic conditions. The sludge is often too much polluted with heavy metals to be reused in agriculture due to mixing with industrial wastewater. Otterpohl et al. (1999) and Zeeman et al. (2000) showed the potential to separate different waste(water) streams in order to treat them in the most sustainable way with energy production and recovery and the reuse of nutrients as the main objectives. The three concepts, proposed by Zeeman et al. (2000), are presented in Figure 2. Anaerobic treatment of domestic sewage or separately collected toilet waste(water) in combination with kitchen and food waste (swill) plays a key role in closing the water and nutrient cycle in these sanitation concepts. The separate collection of toilet waste(water) is very important since the total wastewater contains all the nutrients and pathogens, which were originally produced in a small volume as faeces and urine. After anaerobic digestion of the total wastewater (concept 1), therefore, a more complex post-treatment system is needed to subsequently remove these compounds from a large water volume in comparison with separate treatment of faeces plus urine and grey water (concept 3). So “dilution” not only affects the anaerobic treatment but also the post-treatment. The post-treatment to be applied will depend on the subsequent use of the treated water. The reactor volume for the digestion of faeces plus urine as proposed in concept 3 will mainly depend on the amount of water used for flushing the toilets. The conventional flushing toilet system does not comply with the demands of the production of concentrated slurry. Even 4 litre flushing toilets still produce a too diluted slurry for efficient slurry digestion (Zeeman et al., 2001). Interesting alternatives are the vacuum toilets that are already available on the market and use flushing water quantities as low as 0.7–1.2 litres. Another possibility for reducing the volume of faeces plus urine produced daily without affecting the gross methane potential is separation of the

Conclusions

• Complex waste(water)s mainly consist of proteins, lipids, carbohydrates and sometimes lignin. • Hydrolysis is the first and generally rate-limiting step in the process of anaerobic digestion of particulate organic substrates. • Hydrolysis of particulate polymers is a surface related process and can be described with the Surface Based Kinetics model. • Unlike the hydrolysis of protein and carbohydrate, lipid hydrolysis is hardly occurring in the absence of methanogenesis, affecting the choice for one- or two-step (phase) anaerobic digestion of complex waste(water).


urine, as mentioned before. Urine contains the main fraction of the nutrients, and could be used as a resource for agriculture. Dilution of concentrated waste is not only a concern in public sanitation. Also in animal husbandry different methods of manure collection lead to large differences in manure concentrations, with similar consequences for treatment as sketched above. Also in agricultural industry like the slaughterhouse and fish industries, the processing, collection of waste(water), and treatment and reuse of water and resource should be looked upon in an integrated way. Each step at the beginning of the chain will have its effect on the end of the chain. Combining concentrated with diluted waste streams will almost always end up in much more complicated treatment technologies.

Concept 1: Collection and treatment of total sewage sewage

UASB sludge

Swill

BIOGAS

p o s t-treatment for pathogens

Irrigation, f e r t i l i s a t i on

Soil conditioning, fertilisation

CSTR or AC system

Concept 2: Separate collection and treatment of black and grey water Black water

UASB sludge

S will

Grey water

BIOGAS

p o s ttreatment for pathogens

Irrigation, fertilisatio

Soil conditioning, f tili ti

CSTR or AC system Filtration system

Concept 3: Separate collection and treatment of faeces plus urine and grey water Faeces plus urine and Swill

CSTR or AC system

Grey water

Filtration

BIOGAS

Soil conditioning, fertilisation

Irrigation

Figure 2 Different concepts for the collection and treatment of domestic waste(water), with energy production and recovery and the reuse of nutrients as the main objectives. Total sewage = faeces + urine with water flushing + shower/bath water + laundry water + kitchen water; blackwater = faeces + urine with flushing water

121

• Dilution of complex wastes to wastewaters not only changes the anaerobic reactor technology to be applied but also complicates the post-treatment and possibilities for resource recovery. References G. Zeeman and W. Sanders 122

Batstone, D. (2000). High rate anaerobic treatment of complex wastewater, PhD. Thesis, The university of Queensland. Chyi, Y.T. and Dague, R.R. (1994). Effects of particulate size in anaerobic acidogenesis using cellulose as a sole carbon source, Water Env. Res., 66(5), 670–678. Cohen, A. (1982). Optimisation of anaerobic digestion of soluble carbohydrates containing wastewater by phase separation, PhD thesis, The University of Amsterdam, Amsterdam. Doi, R.H. (1972). Role of proteases in sporulation. Curr. Top. on Cell Regul. 6, 1–20. Elmitwalli, T.A. (2000). Anaerobic treatment of domestic sewage at low temperature, PhD. Thesis, Wageningen Agricultural University. Ghosh, S., Buoy, K., Dressel, L., Miller, T., Wilcox, G. and Loos, D. (1995). Pilot and full scale two-phase anaerobic digester of municipal sludge. Water Environ. Res., 67(2), 206–214. Glenn, A. R. (1976). Production of extracellular proteins by bacteria. Microbiol. 30, 41–62. Hills, D.J. and Nakano, K. (1984). Effects of particle size on anaerobic digestion of tomato solids waste, Agr. Wastes, 285–295. Hobson, P.N. (1987). A model of some aspects of microbial degradation of particulate substrate, J.Ferment. Technol., 65(4), 431–439. McInerney, M.J. (1988). Anaerobic hydrolysis of fats and proteins, In: Zehnder, A.J.B (ed.): Biology of Anaerobic Micro-organisms, John Wiley & sons, New York, 373–415. Miron, Y., Zeeman, G., van Lier, J.B. and Lettinga, G. (2000). The role of sludge retention time in the hydrolysis and acidification of lipids, carbohydrates and proteins during digestion of primary sludge in CSTR systems, Wat. Res., 34(5), 1705–1713. Nagase, M. and Matsuo, T. (1982). Interactions between amino acid-degrading bacteria and methanogenic bacteria in anaerobic digestion, Biotech. Bioeng. 24, 2227–2239. Novak, J.T. and Carlson, D.A. (1970). The kinetics of anaerobic long chain fatty acids degradation, JWPCF 42(2), 1932–1943. Otterpohl, R., Oldenburg, M. and Zimmermann, J. (1999). Source control in urban sanitation and waste management: Ten options with resource management for different social and geographical conditions, Wat. Sci. Tech., 39(5), 153. Palenzuela-Rollon, A. (1999). Anaerobic digestion of fish wastewater with special emphasis on hydrolysis of suspended solids. PhD thesis Agricultural University, Wageningen. Pavlostathis, S.G. and Giraldo-Gomez, E., (1991). Kinetics of anaerobic treatment, Wat. Sci. Tech., 24(8), 35–59. Perot, C. and Amar, D. (1989). Optimisation of sludge anaerobic digestion by separation of the hydrolysisacidification and methanogenesis, Enviro. Tech. Letters, 10, 633–644. Sanders, W.T.M., Geerink, M., Zeeman, G. and Lettinga, G. (2000). Anaerobic hydrolysis kinetics of particulate substrates, Wat. Sci. Tech., 41(3), 17–24. Sanders, W.T.M. (2001). Anaerobic hydrolysis during digestion of complex substrates. PhD thesis Agricultural University, Wageningen. Sayed, S.K.I. (1987). Anaerobic treatment of slaughterhouse wastewater using the UASB process. PhD thesis, Agricultural University, Wageningen. Ten Brummeler, E. (1993). Dry anaerobic digestion of the organic fraction of municipal solid waste, PhD thesis, Agricultural University, Wageningen. Vavilin, V.A., Rytov, S.V., Lokshina, L.Ya. (1996). A description of the hydrolysis kinetics in anaerobic degradation of particulate organic matter, Bio. Res. Tech., 56, 229–237. Wiegant, W.M. (2001). Experiences and potentials of anaerobic wastewater treatment in tropical regions. Wat. Sci. Tech., 44(8), 107–113. Zeeman, G., Kujawa-Roeleveld, K. and Lettinga, G. (2001). Anaerobic treatment systems for high strength wastewater. In: Decentralised Sanitation and Reuse, eds. Lens, P. Zeeman, G. and Lettinga, G. IWA Publishing, London, pp. 218–234. Zeeman, G., Sanders, W. and Lettinga, G. (2000). Feasibility of the on-site treatment of sewage and swill in large buildings, Wat. Sci. Tech., 41(1), 9–16. Zeeman, G. and Lettinga, G. (1999). The role of anaerobic digestion of domestic sewage in closing the water and nutrient cycle at community level, Wat. Sci. Tech., 39(5), 187–194. Zeeman, G., Sanders, W.T.M., Wang, K.Y. and Lettinga, G. (1997). Anaerobic treatment of complex wastewater and waste activated sludge – application of an upflow anaerobic solids removal reactor (UASR) for the removal and pre-hydrolysis of suspended COD, Wat. Sci. Tech., 35, 121–128. Zeeman, G. (1991). Mesophilic and phychrophilic digestion of liquid manure, PhD, thesis, Wageningen University.