Bioresource Technology 99 (2008) 1–6

The effects of digestion temperature and temperatureshock on the biogas yields from the mesophilic anaerobic

digestion of swine manure

K.J. Chae a, Am Jang a, S.K. Yim b, In S. Kim a,*

a Department of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST),

1 Oryong-dong, Buk-gu, Gwangju 500-712, South Koreab R&D Institute, Kolon Engineering and Construction, 199-5, Jeondae-ri, Pogok-eup, Chein-gu, Yongin-si, Kyunggi-do 449-815, South Korea

Received 5 July 2004; received in revised form 14 September 2006; accepted 25 November 2006Available online 16 February 2007

Abstract

In order to obtain basic design criteria for anaerobic digesters of swine manure, the effects of different digesting temperatures, tem-perature shocks and feed loads, on the biogas yields and methane content were evaluated. The digester temperatures were set at 25, 30and 35 �C, with four feed loads of 5%, 10%, 20% and 40% (feed volume/digester volume). At a temperature of 30 �C, the methane yieldwas reduced by only 3% compared to 35 �C, while a 17.4% reduction was observed when the digestion was performed at 25 �C. Ultimatemethane yields of 327, 389 and 403 mL CH4/g VSadded were obtained at 25, 30 and 35 �C, respectively; with moderate feed loads from 5%to 20% (V/V). From the elemental analysis of swine manure, the theoretical biogas and methane yields at standard temperature and pres-sure were 1.12 L biogas/g VSdestroyed and 0.724 L CH4/g VSdestroyed, respectively. Also, the methane content increased with increasingdigestion temperatures, but only to a small degree. Temperature shocks from 35 to 30 �C and again from 30 to 32 �C led to a decreasein the biogas production rate, but it rapidly resumed the value of the control reactor. In addition, no lasting damage was observed for thedigestion performance, once it had recovered.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Anaerobic digestion; Biogas; Methane; Swine manure

1. Introduction

Anaerobic digestion is becoming more and more attrac-tive for the treatment of high strength organic wastes suchas swine manure, since it produces renewable energy, meth-ane, and valuable digested residues, liquid fertilizer and soilconditioner (Angelidaki and Ahring, 1994, 2000; Bonmatiet al., 2001; Hansen et al., 1999). In spite of these advanta-ges, an anaerobic digester for treating swine manure hasnot been attractive in South Korea. Only few farm-scaleor centralized full-scale digesters operating with a mixtureof mainly animal manure supplemented with household

0960-8524/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.biortech.2006.11.063

* Corresponding author. Tel.: +82 62 970 2436; fax: +82 62 970 2434.E-mail address: iskim@gist.ac.kr (I.S. Kim).

organic wastes have been built in South Korea. This isthe result from improper process design, frequent opera-tion failures and insufficient land to utilize digested manureas liquid fertilizer (Chae et al., 2004). However, an eco-nomic evaluation of these digesters reveals that the plantscan be economically viable. In addition, a strong demandfor renewable energy generation and new and forthcomingenvironmental legislation, such as the climate change levy,has gradually increased the interest in anaerobic digestiontechnology (van Lier et al., 2001).

The optimum digester temperature setting, consideringboth the potential biogas yield and heat requirement, isone of the most critical factors for the economically viabledigester operation in temperate countries, like SouthKorea, since most annual temperatures are below the mes-ophilic condition. Therefore, the cost for digester heating

2 K.J. Chae et al. / Bioresource Technology 99 (2008) 1–6

comprises a large portion of the whole operating costs.Many previous researches have suggested that the gas pro-duction during anaerobic digestion is related to tempera-tures. However, different results have been reported(Angelidaki and Ahring, 1994; Hansen et al., 1998;Hashimoto et al., 1981; Hobson et al., 1980; Pfeffer,1974). Hobson et al. (1980) suggested that the gas produc-tion was linearly correlated with temperature from 25 to44 �C, which was between 0.26 and 0.42 CH4 m3/kg totalsolids (TS). Hashimoto et al. (1981) showed that tempera-ture had no effect on the ultimate methane yield of beef cat-tle manure between 30 and 60 �C, although there wasobviously more rapid degradation at higher temperatures.Other researchers suggested that an increase in the temper-ature resulted in a reduction of the biogas yield, due to theincreased inhibition of free ammonia (NH3) whichincreases with increasing temperature (Angelidaki andAhring, 1994; Hansen et al., 1999). According to the stud-ies investigating the effect of temperature on the biogasyield in swine manure anaerobic digestion, much of theprevious researches mainly focused on the thermophiliccondition, with concern regarding the increased ammoniainhibition with increasing temperature. The ammonia con-centration is especially of major concern in agriculturaldigesters when protein rich co-substrates are digested withslaughterhouse waste or waste from collective kitchens.However, thermophilic digestion is not economical due tothe large energy input and operational difficulties in SouthKorea.

Therefore, in this study the anaerobic digestion of swinemanure was performed at different temperatures withinmesophilic ranges (25–35 �C) and at different feed loads(5–40%) to obtain basic design criteria for economicalanaerobic digesters in South Korea. In order to evaluatethe exact theoretical gas yield and methane content, an ele-mental analysis of swine manure was performed. In addi-tion, the responses of anaerobic bacteria to temperatureshocks were investigated.

2. Methods

2.1. Feed characteristics

The raw swine manure for a series of batch experimentswas collected from the swine feedlots (slurry type) in theNational Institute of Livestock Research in South Korea,delivered to the laboratory and stored at 4 �C. The compo-sition of feed was as follows (mg/L): total solids (TS)23885, volatile solids (VS) 16310, chemical oxygen demand(COD) 45350, NHþ4 –N 1910 and PO3�

4 –P 110. The pollu-tant concentrations of the feed manure were relativelylow, due to the large amounts of washing water used as aresult of the concerning of foot and mouth disease, a highlyinfectious viral disease of cloven hoofed animals, duringthis experiment. The swine manure was filtered using a28-mesh sieve to remove large particles prior to feedinginto the reactors.

2.2. Anaerobic master culture reactor

In order to minimize possible test variation and to sup-ply the adapted inoculums for batch experiments, threeanaerobic master culture reactors (MCR, 5 L each) wereoperated at 25, 30 and 35 �C, respectively. These MCRswere initiated with digested swine manure from a meso-philic anaerobic digester as a seed inoculum, and thenoperated on a semi-continuous basis with raw swine man-ure for four weeks until a steady state reached. A steadystate was assumed when the daily biogas yield, pH andCOD values were shown to be constant. After confidencein the steady state of MCRs had been obtained, theenriched inoculums were transferred into batch reactorsunder anaerobic conditions.

2.3. Batch experiments

The purpose of this study was to obtain basic design cri-teria; mainly biogas yields at different temperatures, appro-priate feed loads and digesting temperatures for amesophilic swine manure digester in South Korea. In addi-tion, the response to temperature shocks was also exam-ined. For this study, the batch reactors (1.2 L glass bottlewith a working volume of 1.0 L) were filled with 300 g ofinoculums transferred from anaerobic MCRs operating atthe same three temperatures (25, 30 and 35 �C) that wereused in the batch experiments, pH buffer solution, and freshswine manure as 5%, 10%, 20% and 40% (feed volume/digester volume) feed loads. Reactors were flushed withN2:CO2–80:20 for 3 min to minimize air contamination,tightly closed with butyl rubber stoppers and plastic caps,and subsequently incubated at 25, 30 and 35 �C for 20 days,respectively. Details of the batch test conditions were sum-marized in Table 1. In order to distinguish the amount ofgas produced by the inoculum itself, all batch experimentswere performed with control reactor which contains thesame composition (inoculum + buffer solution) only with-out fresh manure, and that amount was subtracted fromthe test reactor. The pH of the inoculums from the anaero-bic MCRs ranged from 7.2 to 7.4, and the TS and VS con-tents were 15060 and 8700 mg/L, respectively. The cellcomposition of the inoculum for the batch experimentswas C10.10H19.13O2.35NS0.14. In order to prevent a pH drop6000 mg/L NaHCO3 was added, with a pH of 7.2 at thebeginning of the experiments. A magnetic stirrer was intro-duced for vigorous mixing. Each experiment was performedin duplicate to determine experimental errors. The configu-ration of the batch experiments is shown in Fig. 1. A statis-tical analysis was performed using Microsoft excel 2003program, and the data of repeated experiments were givenas the mean ± standard deviation (SD).

2.4. Analytical methods

A complete analytical characterization was carried outaccording to the Standard Methods (APHA, 1995). The

Table 1Experimental set-up for batch tests

Feed load Materials added (mL) Function Test temperature (�C)

Inoculum Feed swine manure Buffer solution Distilled water Total volumea

Control 300 0 200 500 1000 Blank 255% 300 50 200 450 1000 Test unit 3010% 300 100 200 400 1000 Test unit 3520%b 300 200 200 300 1000 Test unit Respectively40% 300 400 200 100 1000 Test unit

a Physical reactor volume is 1.2 L with 0.2 L head space.b One extra set was operated for the batch experiment of the temperature shock investigation.

1

3

2

45

6

7

8

Inoculum (300g)

Swine manure(4 levels of feed loads)

Buffer solution Rectorvolume

(1L)

Head space (0.2L)

Fig. 1. Schematic diagram of batch experiments (1) reactor, (2) waterbath, (3) magnetic stirrer, (4) thermometer, (5) temperature controller, (6)counting cell base of respirometer, (7) computer for data acquisition, (8)gas bag.

K.J. Chae et al. / Bioresource Technology 99 (2008) 1–6 3

volume of biogas produced was automatically measuredusing a Challenge ANR-100 anaerobic respirometer (Chal-lenge Environmental Systems, Inc., Arkansas, USA). Themethane production rates were estimated by analysis ofthe methane contents in the biogas using gas chromatogra-phy (GC, Hewlett Packard 5890 series II) with a flame ion-ization detector. In order to calculate the theoretical biogasyield of swine manure, an elemental analysis was con-ducted using an EA 1110 elemental analyser (CE Instru-ments, Milano, Italy).

3. Results and discussion

3.1. Theoretical biogas yields

The exact expectation of the producible biogas amountand its methane content is one of the most importantaspects in the design of an anaerobic digester. The chemicalcomposition of a feedstock determines the potential biogasyields, as well as the gas composition. Therefore, the meth-ane content obtainable from a given swine manure can beestimated if the chemical composition of the feed is known.The result of elemental analysis, conducted in triplicate,showed the chemical composition of feed used in this studyto be C14.25H28.80O4.43NS0.03. In order to calculate the bio-gas production in the anaerobic digestion, an empiricalequation (Eq. (1)) was used in this study. According to

our calculation, the theoretical biogas and methane yieldsof the feed swine manure at standard temperature and pres-sure (STP, i.e. 0 �C, 760 mmHg) were 1.12 L Biogas/gVSdestroyed and 0.72 L CH4/g VSdestroyed, respectively. Thetheoretical methane content was estimated to be 65.0%,and agreed closely with the batch experiment results.

CnHaObNcSd þ ½A�H2O

! ½B�CO2 þ ½C�CH4 þ cNH3 þ dH2S ð1Þ

where

[A] = (n � a/4 � b/2 + 3c/4 + d/2)[B] = (n/2 � a/8 + b/4 + 3c/8 + d/4)[C] = (n/2 + a/8 � b/4 � 3c/8 � d/4)

CH4 ðSTP;mL=g VSÞ

¼ CH4 ðT �C;mL=g VSÞ � 273

273þ T� 760� P

760ð2Þ

where

T = temperature (�C),P = vapor pressure (mmHg)

3.2. Digesting temperature effect on gas yields and methane

contents

All gas production data at the different temperatures arerepresented as the dry gas production at STP to eliminatethe effects of water vapor pressure (Eq. (2)). The resultsfrom the batch experiments are summarized in Table 2.The influence of the digestion temperatures and feed loadson the methane yields is shown in Fig. 2. Regardless of thedigestion temperature, increasing the feed load from 5% to40% decreased the biogas yield, as would be expected dueto the incomplete degradation for 20 days. In Fig. 2, eventhough the methane yields were different according totested temperatures, the slope of the graph does not clearlyvary within the range of feed loads from 5 to 20% at alltested temperatures. Meanwhile, with a 40% feed load,the methane yields were significantly lower (approximately54% reduction) than with the lower feed loads regardless of

Table 2Summary of methane yields with varying temperatures and feed loads

Temperature(�C)

Feed loadsa

(v/v %)Methane yields CH4 yield of theoretical

value (%)bCODi

c (mg/L) SCODfd (mg/L)

(L/g VSadded) (L/g CODadded)

25 5 0.317 ± 0.017 0.114 ± 0.020 43.8 3310 ± 127 668 ± 3110 0.352 ± 0.017 0.127 ± 0.010 49.3 5620 ± 35 1667 ± 6120 0.312 ± 0.024 0.112 ± 0.024 43.1 10240 ± 85 2483 ± 1840 0.122 ± 0.031 0.044 ± 0.013 16.9 19000 ± 283 7900 ± 113

30 5 0.397 ± 0.010 0.143 ± 0.018 54.8 3310 ± 42 796 ± 8210 0.388 ± 0.025 0.139 ± 0.020 53.6 5620 ± 35 964 ± 2020 0.383 ± 0.018 0.138 ± 0.011 52.9 10240 ± 127 1184 ± 11940 0.170 ± 0.014 0.061 ± 0.007 23.4 19480 ± 141 6000 ± 191

35 5 0.437 ± 0.017 0.163 ± 0.010 60.4 3225 ± 86 908 ± 2510 0.421 ± 0.016 0.157 ± 0.013 58.2 5450 ± 99 1005 ± 3320 0.319 ± 0.014 0.119 ± 0.014 44.1 9900 ± 573 1510 ± 2840 0.228 ± 0.018 0.085 ± 0.010 31.5 18800 ± 184 5580 ± 113

a Feed swine manure% of total reactor volume.b Chemical formula of feed swine manure was C14.25H28.80O4.43NS0.03.c Initial COD concentrations in the reactors.d Final soluble COD concentrations in the reactors after 20 days digestion, values are given as means and standard deviation of two replicates.

4 K.J. Chae et al. / Bioresource Technology 99 (2008) 1–6

digestion temperatures due to incomplete degradationwithin 20 days. Consequently, the upper limit of applicablefeed load of swine manure seems to be up to 20% in meso-philic digestion ranging from 25 to 35 �C.

According to the results, the total gas production wasgreatest at 35 �C, and a temperature increase from 25 to35 �C resulted in an increased biogas yield (17.4% high at35 �C relative to 25 �C). Ultimate methane yields of 317,397 and 437 mL CH4/g VSadded were obtained at 25, 30and 35 �C, respectively, when swine manure was fed at5%. These values corresponded to 44%, 55% and 61% ofthe theoretical yield, which means VS reductions in therange of 44–61%. These were relatively high compared tothe methane yields normally achieved, i.e. 220–350 mLCH4/g VS from swine manure (Bonmati et al., 2001; Han-sen et al., 1998). Bonmati et al. (2001) reported a 347 mLCH4/g VS from the anaerobic digestion of swine manure

Swine manure load (v/v, %)0 10 20 30 40 50

Met

hane

yie

ld (

CH

4 L/

g V

Sad

ded)

0.0

0.1

0.2

0.3

0.4

0.5

25 °C

°C°C30

35

Fig. 2. Methane yield from swine manure as functions of temperature andfeed load (digestion time = 20 days). The data are expressed as themean ± SD (n = 2).

for 80 days at 35 �C when investigating the thermal pre-treatment effect on the gas production potential. Hansenet al. (1998) presented that the maximum methane poten-tial of swine manure was 300 ± 20 mL CH4/g VS, and alsoa relatively low methane yield of 188 mL CH4/g VS at37 �C due to the inhibition caused by a high ammonianitrogen concentration of 6000 mg/L. The relatively highmethane yield obtained from this experiment might haveresulted from the high carbon and hydrogen contents inthe feed, as shown in the results of elemental analysis,C14.25H28.80O4.43NS0.03.

The relative biogas yield (% of gas production at 35 �C)at different temperatures in the digestion of swine manureis shown in Fig. 3. The biogas yield was influenced by tem-perature in the range of 25–35 �C, but was not linear withinthe tested range. The difference in the methane yield wasnot obvious between 35 and 30 �C; approximately 97% of

Temperature (ºC)

25 30 35

Rel

ativ

e bi

ogas

yie

ld

0.0

0.2

0.4

0.6

0.8

1.0VS base

COD base

Fig. 3. Relative biogas yields (% of gas production at 35 �C, means of fourfeed loads) for various digester operating temperatures. The data areexpressed as the mean ± SD (n = 4).

K.J. Chae et al. / Bioresource Technology 99 (2008) 1–6 5

the methane produced at 35 �C was still produced at 30 �C.In contrast, the digestion proceeding at a temperature of25 �C showed only 82.6% of that at 35 �C. These resultswere in agreement with previous results that showed animprovement in the biogas yields with increasing tempera-ture (Hobson et al., 1980).

There was a faster degradation at the higher tempera-tures, as shown in Fig. 4. The degradation of swine manureat 25 �C took almost twice as long as at 35 �C. Therequired times to complete the digestion of a 10% feed loadwere 7, 12, 15 days at 35, 30 and 25 �C, respectively. Thesewere fairly short due to the removal of the large particlesby the 28-mesh sieve prior to feeding into reactors. There-fore, approximately 15–20 days seems to be the minimumfor optimal digestion of swine manure for a digester inwhich somewhat larger particles can be fed occasionally.

The biogas composition differed according to digestiontemperature, with methane contents in the biogas of65.3%, 64.0% and 62.0% at 35, 30 and 25 �C, respectively,but these differences were statistically not significant. Theobtained methane contents were similar to the 65% theoret-ical value calculated from Eq. (2).

3.3. Optimum digesting temperature on the basis of net

energy recovery

Within the mesophilic range, the predominant operatingtemperature in the full-scale anaerobic digesters operatingin South Korea is 35 �C, with the common belief that theoptimum gas production occurs at 35 �C. Running a diges-ter at 25 instead of 35 �C results in a remarkable saving ofenergy, especially at low retention times, if the producedgas is fairly comparable between both temperatures. How-ever, with digestion at 25 �C the methane yield is only82.6% of that produced at 35 �C. It led to a lesser calcu-lated net energy recovery of the digester of 25 comparedto 35 �C, even with a lower energy demand for heatingthe digester than at 35 �C. Consequently, the net energy

Reaction Time (days)0 5 10 15 20 25 30

Met

hane

Yie

ld (

CH

4 L/

g V

Sad

ded)

0.0

0.1

0.2

0.3

0.4

0.5

25 °C

°C°C30

35

Fig. 4. Methane production versus time at 25, 30 and 35 �C (10% swinemanure load). The data are expressed as the mean ± SD (n = 2).

balance between the energy demands to heat the digesterand improved energy production from the increased meth-ane yield as the temperature is increased must be simulta-neously considered when deciding the optimum operatingtemperatures.

However, this net energy balance is also directly depen-dent upon the feed VS concentration, as the absolute meth-ane potential depends on the VS amount. If the input VSconcentration is higher than about 45000 mg/L, a temper-ature of 35 �C is more economical than 30 �C since theimproved methane yield at 35 �C can overcome the addi-tional heating energy demands. However, at a VS concen-tration of less than 45000 mg/L, digestion at 30 �C is morefavorable than at 35 �C. Therefore, with respect to the netenergy recovery, the optimum temperature in South Koreamight be between 30 and 35 �C, as the usual VS content isapproximately 40000 mg/L.

Furthermore, temperature has a strong effect on theconcentration of free ammonia (NH3), the real inhibitorrather than ammonium ðNHþ4 Þ (Hashimoto, 1986). Freeammonia concentration increases with increasing tempera-ture, by influencing the equilibrium. Therefore, careful con-sideration is required when increasing the digestiontemperature for the purpose of enhancing the methaneyield due to the simultaneous increase in the ammoniainhibition.

3.4. Response of digestion performance on temperature

shocks

The effect of temperature shocks on the digestion perfor-mance of swine manure is shown in Fig. 5. In a digesternormally operated at 35 �C, the temperature was suddenlylowered to 30 �C, maintained for 170 h, and then raisedagain to 32 �C. Temperature shock, from 35 to 30 �C, ledto a decrease in the biogas production rate compared tothe control. However, it resumed to the value of the controlreactor in about 40 h. Once recovered, no difference in the

Time (hr)0 100 200 300 400 500

Bio

gas

prod

uctio

n ra

te (

mL/

hr/L

)

0

2

4

6

8

10

12

14

ControlTemperature Shock

°C35

°C32

°C30

Fig. 5. Influence of temperature shock on the digestion of swine manureat a 20% (v/v) feed load.

6 K.J. Chae et al. / Bioresource Technology 99 (2008) 1–6

biogas yield was observed between the control and temper-ature shock reactor. The 2nd temperature shock, from 30to 32 �C, again showed a decrease in the biogas yield, butreturned to its former level more rapidly. Obviously, inorder to maintain a digester at its optimum performance,constantly keeping the temperature at the optimum levelis very important, as the methane forming bacteria are verysensitive to temperature changes. According to the results,however, no lasting damage was seen in the digestion per-formance, even with fairly large changes in temperature.The temperature shock reactor showed only 7.2% less ulti-mate gas production compared to the control. This resultsuggests that an effective adaptation of the mesophilic bac-teria to different temperatures is possible as long as thechange occurs smoothly.

4. Conclusions

The digestion temperature has an influence on theultimate methane yield, as well as the methane content.In the mesophilic temperature range, 25–35 �C, the higherthe temperature, the better the methane yield. However,the yield did not linearly increase with increasing tempera-ture. The yields at 30 and 35 �C were similar, but were quitehigh compared to that at 25 �C, by more than 13–17%.However, this result does not mean the higher temperaturethe more optimal, due to the larger energy requirement athigher digesting temperatures. Therefore, careful consider-ation of the net energy balance between the increased heat-ing energy demands and improved additional methaneproduction at higher operating temperatures must besimultaneously taken into account when deciding the eco-nomical digesting temperature. Furthermore, this netenergy balance is inevitably dependent upon the VS con-tent of the feed, as methane comes from the degradationof VS. Temperature shocks led to a reduction in the biogasproduction rate compared to that of the control, but recov-

ered rapidly. Once recovered, no clear difference in the bio-gas yield was observed between the control and thetemperature shock reactor. This result indicates that,although methanogens are quite sensitive to temperaturethey have considerable ability to adapt to temperaturechanges.

References

Angelidaki, I., Ahring, B.K., 1994. Anaerobic thermophilic digestion ofmanure at different ammonia loads: Effect of temperature. Water Res.28 (3), 727–731.

Angelidaki, I., Ahring, B.K., 2000. Methods for increasing the biogaspotential from the recalcitrant organic matter contained in manure.Water Sci. Technol. 41 (3), 189–194.

APHA, 1995. Standard methods for the examination of water andwastewater, 19th ed. American public health association, Washington,DC.

Bonmati, A., Flotats, X., Mateu, L., Campos, E., 2001. Study of thermalhydrolysis as a pretreatment to mesophilic anaerobic digestion of pigslurry. Water Sci. Technol. 44 (4), 109–116.

Chae, K.J., Yim, S.K., Choi, K.H., Kim, S.K., Park, W.K., 2004.Integrated biological and electro-chemical treatment of swine manure.Water Sci. Technol. 49 (5–6), 427–434.

Hansen, H.H., Angelidaki, I., Ahring, B.K., 1998. Anaerobic digestion ofswine manure: Inhibition by ammonia. Water Res. 32 (1), 5–12.

Hansen, H.H., Angelidaki, I., Ahring, B.K., 1999. Improving thermophilicanaerobic digestion of swine manure. Water Res. 33 (8), 1805–1810.

Hashimoto, A.G., 1986. Ammonia inhibition of methanogenesis fromcattle wastes. Agr. Wastes 17, 241–261.

Hashimoto, A.G., Varel, V.H., Chen, Y.R., 1981. Ultimate methane yieldfrom beef cattle waste: effects of temperature, reaction constituents,antibiotics and manure. Agr. Wastes 3, 241–256.

Hobson, P.N., Bousfield, S., Summers, R., Mills, P.J., 1980. Anaerobicdigestion of piggery and poultry wastes. In: Stafford, B.E., Wheatley,B.I., Hughes, D.E. (Eds.), Anaerobic digestion. Applied SciencePublishers, London, pp. 237–253.

Pfeffer, J.T., 1974. Temperature effects on anaerobic fermentation ofdomestic refuse. Biotechnol. Bioeng. 16, 771–787.

van Lier, J.B., Tilche, A., Ahring, B.K., Macarie, H., Moletta, R.,Dohanyos, M., Hulshoff Pol, L.W., Lens, P., Werstraete, W., 2001.New perspectives in anaerobic digestion. Water Sci. Technol. 43 (1), 1–18.