Bioresource Technology 154 (2014) 215–221

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Comparison of high-solids to liquid anaerobic co-digestion of food wasteand green waste

0960-8524/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.biortech.2013.12.054

⇑ Corresponding author. Tel.: +86 571 8898 2179; fax: +86 571 8898 2191.E-mail address: kcsheng@zju.edu.cn (K. Sheng).

Xiang Chen a, Wei Yan a, Kuichuan Sheng a,⇑, Mehri Sanati b

a College of Biosystems Engineering and Food Science, Zhejiang University, Hangzhou 310058, Chinab Faculty of Engineering, Department of Design Sciences, Lund University, P.O. Box 118, SE-22100 Lund, Sweden

h i g h l i g h t s

� High-solids and liquid co-digestion of food waste (FW) and green waste (GW).� Optimal biogas production was achieved at FW:GW mixing ratio of 40:60.� Methane yields at 15–20% total solids (TS) were higher than that at 5–10% TS.� Organic overloading at high TS content (25%) caused inhibition of methanogenesis.� Volumetric productivity at 15–25% TS was 3.8- to 4.6-fold higher than that at 5% TS.

a r t i c l e i n f o

Article history:Received 17 October 2013Received in revised form 10 December 2013Accepted 12 December 2013Available online 22 December 2013

Keywords:High-solids anaerobic digestionCo-digestionFood wasteGreen wasteBiogas

a b s t r a c t

Co-digestion of food waste and green waste was conducted with six feedstock mixing ratios to evaluatebiogas production. Increasing the food waste percentage in the feedstock resulted in an increased meth-ane yield, while shorter retention time was achieved by increasing the green waste percentage. Foodwaste/green waste ratio of 40:60 was determined as preferred ratio for optimal biogas production. About90% of methane yield was obtained after 24.5 days of digestion, with total methane yield of 272.1 mL/gVS. Based the preferred ratio, effect of total solids (TS) content on co-digestion of food waste and greenwaste was evaluated over a TS range of 5–25%. Results showed that methane yields from high-solidsanaerobic digestion (15–20% TS) were higher than the output of liquid anaerobic digestion (5–10% TS),while methanogenesis was inhibited by further increasing the TS content to 25%. The inhibition maybe caused by organic overloading and excess ammonia.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Bioenergy recovery and pollution control through anaerobicdigestion (AD) of organic wastes is a promising greenhouse gasmitigation option and considered to be a sustainable waste treat-ment practice (Pantaleo et al., 2013; Rajagopal et al., 2013). Sincemethane rich biogas is the main end product of AD, methaneproduction must be improved to maximize revenues from energygeneration and hence, to make digestion facilities more profitable(Fdez-Güelfo et al., 2012). Driven by a complex and diverse com-munity of microbial organisms, the performance of AD is affectedby a variety of operational factors, such as temperature, pre-treat-ment of substrates, and digester mixing. The total solids (TS)content in association with the organic loading rate is also one ofthe key factors that affect the performance, cost and stability ofAD systems (Alvarez and Liden, 2008; Wu et al., 2009). It has been

reported that the TS content affects the following parameters:rheology and viscosity of the digester contents, fluid dynamics,clogging, and solid sedimentation that can directly influence theoverall mass transfer rates within the digesters (Karthikeyan andVisvanathan, 2013).

Since the TS content is an important parameter, two main typesof AD processes have been developed: liquid and high-solids AD.Liquid AD (L-AD) systems typically operate with 0.5–15% TS, whilehigh-solids AD (HS-AD) refers to a process that generally operatesat 15–40% TS (Shi et al., 2013). It has been claimed that HS-AD isadvantageous over L-AD for a number of reasons including highervolumetric loading capacity, reduced energy input for heating andmixing, and greater ease in handling the compost-like digestate (Liet al., 2011). However, both HS-AD and L-AD have their ownadvantages and disadvantages with respect to methane productionmaximization and process optimization. Even though the HS-ADprocess is reported to tolerate high organic loadings, low opera-tional stability still hinders wide application of HS-AD technology(Schievano et al., 2010). HS-AD may be particularly sensitive to

216 X. Chen et al. / Bioresource Technology 154 (2014) 215–221

the inhibition caused by overproduction of volatile fatty acids(VFAs) and ammonia, due to organic overloading. However, sofar, information is lacking concerning the quantitative thresholdof the TS content below which methane production from HS-ADis higher or comparable to the output of L-AD.

There are some studies related to the effect of the TS content onthe performance of AD process. Forster-Carneiro et al. (2008) ana-lyzed the AD process of food waste with three different TS levels.The results showed that reactors at 20% TS achieved a higher meth-ane production compared to 25% and 30% TS. In a study conductedby Wu et al. (2009), no significant differences were observed in themethane production ranging from 351 to 381 mL/g VSfeedstock,applied to four TS contents of 1%, 2%, 5% and 10%. Recently, Brownet al. (2012) evaluated several lignocellulosic feedstocks (switch-grass, corn stover, wheat straw, yard waste, leaves, and maple)for biogas production under L-AD (5% TS) and HS-AD (18–19%TS). The study found no significant difference in methane yieldbetween L-AD and HS-AD. These studies investigated the influenceof TS control levels on AD, but the TS contents studied were withina narrow range; studies on a wider range of TS contents affectingperformance of anaerobic reactors under both L-AD and HS-ADare limited.

Food waste and green waste are available year round andaccount for a significant portion of municipal solid waste (MSW)(Brown and Li, 2013). The use of food waste and green wastemay improve the overall economic benefits of AD process due tothe low or zero cost associated with collecting these two feed-stocks (Brown et al., 2012; Brown and Li, 2013). However, due tothe high biodegradability and relatively low carbon to nitrogen(C/N) ratio, mono-digestion of food waste may encounter variouspotential inhibitors, including fast VFAs production from starchand free ammonia from protein (Brown and Li, 2013; Xu and Li,2012). Mono-digestion of lignocellulosic green waste also faceschallenges, including its poor nutrient content, slow start-up andlong retention time (Pohl et al., 2013). Better methane productionperformance is expected in co-digestion systems. Co-digestion is awell-accepted process that enhances organic matter degradationand biogas production by synergistic and complementary effects,which improve the balance of nutrients and dilute inhibitorycompounds (Kim and Oh, 2011; Wan et al., 2011). Consequently,anaerobic co-digestion may be a promising solution for centralizedtreatment of food waste and green waste.

The objective of this study was to investigate the effect of TScontrol levels on anaerobic co-digestion of food waste and greenwaste. Anaerobic batch tests were conducted under L-AD andHS-AD, with TS contents ranging from 5% to 25%. A preferred foodwaste to green waste (FW/GW) mixing ratio was also needed tooptimize biogas production for co-digestion. Hence, the effect ofFW/GW mixing ratio on the performance of co-digestion was as-sessed first. Then, a comparison of high-solids to liquid anaerobicco-digestion of food waste and green waste was evaluated basedon the pre-determined preferred FW/GW ratio.

2. Methods

2.1. Feedstock and inoculum

The food waste was collected from one student canteen in Zhe-jiang University, Hangzhou, China. Impurities contained in the foodwaste, such as bones, eggshell, wastepaper and plastics wereremoved manually after sampling. Then the food waste wasground up using a blender (CPEL-23, Shanghai Guosheng, China).The ground food waste slurry was sealed in plastic bags and storedin a freezer at �20 �C. The food waste was thawed overnight underambient conditions before usage.

Green waste was collected on the campus of Zhejiang Univer-sity and mainly contained grass clippings and fallen leaves. Thegreen waste was air-dried at room temperature for 48 h, and thenground with a grinder (DYQ-188, Ruian Huanqiu, China). Then theground green waste was screened through a 5-mm sieve, andstored at 4 �C until used.

The anaerobic sludge taken from the bottom settlement of amesophilic anaerobic digester in Hangzhou, China was used asinoculum. The digester was a 300 m3 tank fed with livestock man-ure. Before sampling, the digester stirring was stopped for 1 day.The sludge was kept in air-tight buckets under ambient conditions(about 25 �C) after sampling.

2.2. Batch anaerobic digestion system

Each batch AD system consisted of a 500-mL digestion glassbottle, a 2-L gas collection glass bottle and a 500-mL liquid collec-tion beaker. The digestion bottle was loaded with feedstock andinoculum. Once biogas was produced in the digestion bottle, itwas automatically distributed into the gas collection bottle whichwas filled with diluted hydrochloric acid solution (pH < 3), andthen an equivalent volume of acid solution to the produced biogaswas displaced to the liquid collection beaker. Thus, the biogas pro-duction volume could be measured periodically by means of thewater displacement method.

2.3. Experimental design and set-up

Two sets of experiments were carried out in the batch AD sys-tem. The first set of experiments studied the effect of FW/GW mix-ing ratios on biogas production via anaerobic co-digestion. Sixfeedstock mixing ratios (FW/GW: 100:0, 80:20, 60:40, 40:60,20:80, and 0:100, based on VS) were studied. Based on the initialTS contents of the food waste, green waste and inoculum, a suffi-cient amount of deionized water was added in each condition toadjust the TS content of the mixture inside the batch system to15%.

After completing the first set of experiments, a preferred FW/GW mixing ratio for optimal biogas production was determined:40:60. Based on this ratio, the second set of experiments investi-gated the effect of TS content on co-digestion of food waste andgreen waste. Food waste and green waste (40% food waste and60% green waste, based on VS) were digested at five TS levels:5%, 10%, 15%, 20% and 25%. Based on the initial TS contents of thefeedstock and inoculum, a sufficient amount of deionized waterwas added in each digestion test to adjust the corresponding TScontent. For the digestion tests with higher TS contents (i.e. 20%and 25%), the inoculum sludge was centrifuged (Centrifuge5810R, Eppendorf, Germany) at 3000 rpm for 30 min. After remov-ing the decanted liquid from the solid, the solid portion wascollected, and then its TS and volatile solids (VS) contents weremeasured again for the digestion tests.

In all the digestion tests, the feedstock and inoculum wereloaded into the batch system at a feedstock/inoculum ratio of 1.0(5.0 g VS of feedstock and 5.0 g VS of inoculum were added). Thefeedstock/inoculum ratio was calculated based on the amount offeedstock to the amount of inoculum on a VS basis. Blank trialscontaining inoculum only were performed to correct for the biogasproduced from the inoculum. All the tests were carried out induplicate. After adding the feedstock and inoculum, the anaerobicreactor was tightly closed with a rubber stopper and a screw cap,and then flashed with argon gas for 5 min. Thereafter, the ADsystems were incubated at 37 ± 1 �C.

60

70

80

90

(g V

S·d)

)

100% FW + 0% GW80% FW + 20% GW60% FW + 40% GW40% FW + 60% GW

(a)

X. Chen et al. / Bioresource Technology 154 (2014) 215–221 217

2.4. Analytical methods and data analysis

The TS and VS contents of food waste, green waste, inoculumsludge and digestate were determined according to the StandardMethods (APHA, 1998). Total carbon and nitrogen contents weremeasured by an elemental analyzer (EA 1112, CarloErba, Italy).To determine pH, NH4–N, total VFAs, and alkalinity (total inorganiccarbon) before and after each test, samples were prepared by mix-ing a 5-g sample with 50 mL of deionized water, and subsequentlythe dilution was centrifuged at 10000 rpm for 15 min. The super-natant was then filtered through qualitative filter paper, and thefiltrate was analyzed. The pH was determined by a pH meter(PHS-3D, Shanghai Jinghong, China). The NH4–N concentrationwas measured by spectrophotometry according to the StandardMethods (APHA, 1998). Total VFAs and alkalinity were determinedthrough a two-step titration method (Voß et al., 2009). Cellulose,hemicellulose, and lignin contents in the green waste were ana-lyzed according to the procedure described by Van Soest et al.(1991).

The biogas composition (CH4, CO2, H2, and N2) from the head-space of the AD system was analyzed using a gas chromatograph(GC 2014, Shimadzu, Japan) equipped with a thermal conductivitydetector. The temperatures of the column oven, injector port anddetector were 100, 120, and 120 �C, respectively. Argon at a flowrate of 30 mL/min was used as a carrier gas. The biogas and meth-ane yields at the end of each test were calculated by dividing thecumulative gas yields by the mass of VS in the feedstock loadedinto the reactors at start-up. The volumetric methane productivityexpressed in Vmethane/Vwork was calculated as the volume of meth-ane production (Vmethane) per unit working volume of the reactor(Vwork).

Analysis of variance (ANOVA) was performed using MicrosoftExcel 2007 software to determine statistical significance with athreshold p-value of 0.05.

0

10

20

30

40

50

0 5 10 15 20 25 30 35 40 45 50

Dai

ly b

ioga

s yi

eld

(mL/

Digestion time (Days)

20% FW + 80% GW0% FW + 100% GW

350

400

450

mL/

g VS

) (b)

3. Results and discussion

3.1. Characteristics of feedstocks and inoculum

Characteristics of the food waste, green waste and inoculumsludge are shown in Table 1. TS contents of the food waste andgreen waste were 26.9% and 86.8%, respectively on a wet weightbasis, both of which were above 25%. This indicates that thesetwo materials with high solids contents were suitable for HS-AD.The anaerobic sludge had a TS content of 13.7%. Centrifugation ofthe sludge was adopted in this study to ensure that the TS of feed-stock and sludge mixture in the reactors were around 20% orabove. It has been reported that up to 50% of inoculum is requiredin HS-AD systems for a rapid start-up (Brown and Li, 2013; Martinet al., 2003). Consequently, a highly active and concentrated

Table 1Characteristics of feedstocks and anaerobic sludge.

Parameters Food waste Green waste Anaerobic sludge

TS (%, w.b.) 26.9 ± 0.3 86.8 ± 0.3 13.6 ± 0.2VS (%, w.b.) 25.2 ± 0.3 74.3 ± 0.9 6.4 ± 0.1VS/TS (%) 93.6 ± 0.5 85.7 ± 1.2 47.1 ± 0.2Total carbon (%, d.b.) 46.3 ± 0.7 45.3 ± 0.3 29.4 ± 0.3Total nitrogen (%, d.b.) 2.1 ± 0.2 1.1 ± 0.1 2.6 ± 0.5C/N 22.0 ± 1.1 41.2 ± 1.3 11.3 ± 0.9pH 4.51 ± 0.01 ND 7.46 ± 0.01Cellulose (%, d.b.) ND 32.1 ± 0.9 NDHemicellulose (%, d.b.) ND 23.7 ± 0.7 NDLignin (%, d.b.) ND 14.1 ± 0.7 ND

Note: w.b., wet base; d.b., dry base; ND, not determined.

inoculum source was critical to speed up the HS-AD process(Brown and Li, 2013; Forster-Carneiro et al., 2008). The effluentof L-AD after dewatering and drying, or the digestate of HS-ADcan be utilized in industrial HS-AD reactors. The cellulose, hemicel-lulose, and lignin contents of green waste were 32.1%, 23.7%, and14.1%, respectively.

3.2. Effect of food waste to green waste mixing ratios on biogasproduction

The daily and accumulative biogas yields during the co-diges-tion of food waste and green waste at different mixing ratios areshown in Fig. 1. The biogas production processes ran for about50 days until no more biogas production was observed. For allthe digestion tests, biogas production started immediately fromthe first day, and peak daily biogas production rates were observedafter 1.5 days of digestion. The highest biogas production rate wasobtained at an FW/GW mixing ratio of 100:0, with a peak dailybiogas production rate of 80.6 mL/g VS/d, which was 1.5-fold(p < 0.05) higher than that of the digestion system with an FW/GW mixing ratio of 0:100. However, biogas production ratedropped immediately after the peak for the digestion system withan FW/GW mixing ratio of 100:0, and no biogas was produced fromday 6.5 to 12.5, indicating that an apparent severe inhibitionoccurred. The inhibition was probably caused by the higherdigestibility of food waste compared to the lignocellulosic greenwaste, leading to overproduction of VFAs that inhibited the

0

50

100

150

200

250

300

0 5 10 15 20 25 30 35 40 45 50

Cum

ulat

ive

biog

as y

ield

(

Digestion time (Days)

100% FW + 0% GW80% FW + 20% GW60% FW + 40% GW40% FW + 60% GW20% FW + 80% GW0% FW + 100% GW

Fig. 1. (a) Daily and (b) accumulative biogas yields during co-digestion of foodwaste and green waste at different mixing ratios (note: FW = food waste,GW = green waste).

218 X. Chen et al. / Bioresource Technology 154 (2014) 215–221

methanogenesis process (Brown and Li, 2013). Then after about10 days of self-recovery, the system started again to produce bio-gas, and a higher daily biogas production occurred from day 30to 45. Digestion systems with FW/GW mixing ratios of 80:20 and60:40 also had similar biogas production processes.

For the digestion systems with FW/GW mixing ratios of 40:60,20:80 and 0:100, the biogas production rates could be separatedinto two phases: an initial rapid production for the first 12 daysfollowed by a slower rate over the rest of the digestion test. Withincreasing the amount of green waste from 60% to 100% in thefeedstock, biogas production increased until day 32.5, 24.5, and16.5, respectively, and subsequently biogas was produced at a neg-ligible level until the end of experiments (day 50.5). It was foundthat higher daily biogas production occurred later in the digestionprocess when the food waste was the major component in thefeedstock (60–100% food waste). The retention time for mono-digestion of food waste (50.5 days) was two times longer than thatfor mono-digestion of green waste (16.5 days). It seemed theaddition of green waste can result in shorter retention time.

From Fig. 1b, it can also be seen that at the end of the digestionprocess, the total biogas yields were 409.8, 389.4, 388.8, 390.2,324.7, and 270.9 mL/g VS for the digestion system with FW/GWmixing ratios of 100:0, 80:20, 60:40, 40:60, 20:80, and 0:100,respectively. Approximately 44.0%, 43.7%, 66.5%, 90.3%, 96.7%,and 96.9% of the total biogas yields were obtained after the first24.5 days of digestion, respectively, for FW/GW mixing ratios of100:0, 80:20, 60:40, 40:60, 20:80, and 0:100. During the periodof 20–32.5 days, the highest accumulative biogas yield was ob-served for the FW/GW mixing ratio of 40:60. Consequently,increasing the amount of food waste in the feedstock led to anincrease in the biogas yield, while higher biogas productionefficiency was achieved by increasing the amount of green waste.

The average methane contents of biogas produced from co-digestion of food waste and green waste at different mixing ratiosare shown in Fig. 2. Statistical analysis shows that the mixing ratioshad significant effects (p < 0.05) on methane contents. The highestmethane content of 79.7% was observed in the digestion systemwith 100% food waste, which was comparable to a study of theHS-AD (with 20% TS) of the organic fraction of MSW where themethane content remained practically constant at 80% in the per-iod 15–40 days (Fernández et al., 2010). With the addition of greenwaste, methane content of the biogas started to decrease. Thehigher the composition ratio of green waste, the lower was themethane content in the digestion system. The system with 100%green waste had the lowest methane content of 60.7%. The highermethane content at higher composition ratios of food waste was

0

50

100

150

200

250

300

350

400

0

10

20

30

40

50

60

70

80

90

100

100:0 80:20 60:40 40:60 20:80 0:100

Met

hane

yie

ld (m

L/g

VS)

Met

hane

con

tent

(%)

Food waste to green waste mixing ratio

Methane contentMethane yield

Fig. 2. Average methane contents and total methane yields produced from co-digestion of food waste and green waste at different mixing ratios.

probably caused by the high protein content contained in the foodwaste. Compared with carbohydrate-rich feedstocks, such as ligno-cellulosic green waste, methane content in the biogas producedduring the degradation of protein-rich materials was higher (Wei-land, 2010). A study by Liu et al. (2009) also found higher methanecontents in the biogas were achieved from digestion of food wastethan from green waste and the mixture (50% food waste and 50%green waste, based on VS).

The total methane yield of the digestion systems can also beseen in Fig. 2. Methane yield comparisons closely resemble thoseof the methane content comparisons at the same FW/GW mixingratios. By decreasing the food waste percentages in the feedstockfrom 100% to 0%, the methane yield decreased by 49.6% (from326.4 to 164.6 mL/g VS). The methane yield from digestion of greenwaste was lower than the yield obtained from food waste. Sincefood waste and green waste were co-digested at different mixingratios, the synergistic effect of co-digestion could be estimated asan additional methane yield for co-substrates over the weightedaverage of the individual substrate’s experimental methane yield(EMY) (Labatut et al., 2011). If the differential (EMY – WeightedEMY) was positive and greater than the standard deviation (SD)of EMY, the synergistic effect could be confirmed. The weightedEMY of co-substrates was calculated according to the followingformula:

Weighted EMY ¼ EMYFW � PFW þ EMYGW � PGW ð1Þ

where, Weighted EMY is the weighted average of experimentalmethane yield for co-substrates (mL/g VS); EMYFW and EMYGW

are the experimental methane yields for food waste and greenwaste, respectively (mL/g VS); PFW and PGW are the percentage offood waste and green waste in the co-substrates, respectively ona VS basis.

As seen in Table 2, synergistic effects were found in all the co-digestion systems, since the positive differentials in methaneyields were all greater than their SD. In the digestion system withan FW/GW mixing ratio of 40:60 in particular, the EMY was 18.7%higher than the weighted EMY. The synergism observed in the co-digestion may arise from the adjustment of the C/N ratios in thereactors from 14.4 (food waste alone) and 16.9 (green waste alone)to 14.9–16.4, since the C/N ratios generally fell in the optimumrange of 15–30 (Weiland, 2010). Co-digestion can actually improvethe methane production of the AD process (Ward et al., 2008).

In summary, the first set of experiments showed that the FW/GW mixing ratios had significant effects (p < 0.05) on anaerobicco-digestion of food waste and green waste. In AD systems operat-ing at a TS content of 15%, increasing the percentage of food wastein the feedstock led to an increase in the methane yield, whileshorter retention time was achieved by increasing the amount ofgreen waste. Considering balanced methane production capacityand efficiency, an FW/GW mixing ratio of 40:60 was regarded asa preferable mixing ratio. With this ratio, 90% of the methane yieldwas obtained after 24.5 days of digestion, and the total methaneyield was determined to be 272.1 mL/g VS.

3.3. Comparison of liquid to high-solids anaerobic co-digestion of foodwaste and green waste

3.3.1. Biogas productionA preferable FW/GW mixing ratio of 40:60 was determined in

the first set of experiments to further study the effect of the TScontrol level on the co-digestion of food waste and green waste.Co-substrates were digested with five different TS contents: 5%,10%, 15%, 20% and 25%, and the corresponding initial VS loadingswere 15.4, 30.7, 46.1, 61.5, and 76.9 g VS/L, respectively. The dailyand accumulative biogas yields during the co-digestion of foodwaste and green waste at different TS contents are shown in Fig. 3.

Table 2Synergistic effect evaluation of co-digestion of food waste and green waste.

FW/GW mixing ratioa C/N EMY SD of EMY Weighted EMY Differential (EMY – Weighted EMY) Increasing rate of methane yield (%)b Synergistic effect

100:0 14.4 326.4 30.7 326.4 – – –80:20 14.9 305.9 7.2 294.0 11.9 4.0 Synergistic60:40 15.3 284.6 10.0 261.7 22.9 8.8 Synergistic40:60 15.8 272.1 26.6 229.3 42.8 18.7 Synergistic20:80 16.4 211.2 9.1 196.9 14.3 7.2 Synergistic0:100 16.9 164.6 4.8 164.6 – – –

FW: food waste; GW: green waste; C/N: C/N ratio in the mixture of food waste, green waste and anaerobic sludge; EMY: experimental methane yield (mL/g VS); SD: standarddeviation.

a Based on volatile solids (VS).b The increasing rate of methane yield was calculated based on the comparison between EMY and weight EMY.

0

10

20

30

40

50

60

0 5 10 15 20 25 30 35 40 45 50

Dai

ly b

ioga

s yi

eld

(mL/

(g V

S·d)

)

Digestion time (Days)

5% TS10% TS15% TS20% TS25% TS

(a)

0

50

100

150

200

250

300

350

400

450

0 5 10 15 20 25 30 35 40 45 50

Cum

ulat

ive

biog

as y

ield

(mL/

g VS

)

Digestion time (Days)

5% TS10% TS15% TS20% TS25% TS

(b)

Fig. 3. (a) Daily and (b) accumulative biogas yields during co-digestion of foodwaste and green waste at different TS control levels.

0

50

100

150

200

250

300

350

0

10

20

30

40

50

60

70

80

90

100

5 10 15 20 25

Met

hane

yie

ld (m

L/g

VS)

Met

hane

con

tent

(%)

TS control level (%)

Methane contentMethane yield

Fig. 4. Average methane contents and total methane yields produced from co-digestion of food waste and green waste at different TS control levels.

X. Chen et al. / Bioresource Technology 154 (2014) 215–221 219

Similar to the first set of experiments, biogas production startedimmediately from the first day for all the digestion tests, indicatingfast acclimation of the microorganisms to the co-substrates. Thedaily biogas production rates in the digestion systems at TS levelsof 5%, 10%, 15%, 20% and 25% reached their peak values of 42.5,40.2, 55.3, 60.5 and 39.0 mL/g VS/d, respectively, after 1.5 days ofdigestion. For the digestion systems at TS contents of 5%, 10%and 15%, the biogas production rate dropped immediately afterthe first 4.5 days of digestion, and then resumed to reach anotherpeak at day 10.5. This was followed by a slower rate over the restof the digestion test. For the L-AD systems (5% and 10% TS), biogasproduction increased until day 24.5, and then remained almostconstant until the end of experiments (day 50.5). Similar to diges-tion systems with higher percentages of food waste in the first setof experiments, there was a suspension phase during the biogas

production process for the HS-AD systems (20% and 25% TS). Afterthe biogas production peak on day 1.5, the rate declined rapidly,and the digesters had zero biogas production from day 4.5 to 8.5.Then, after about 6 days of self-recovery, biogas production re-sumed in the reactors, and a higher daily biogas production wasobserved from day 24.5 to 45.

Fig. 3b shows that the total biogas yields from the digestion sys-tems operating at TS of 5%, 10%, 15%, 20% and 25% were 312.8,335.7, 390.2, 348.4, and 269.0 mL/g VS, respectively. After16.5 days of digestion, about 82.8%, 82.7%, 73.0%, 46.1%, and50.2% of the total biogas yield was achieved, and after 24.5 daysof digestion, about 95.8%, 96.1%, 90.3%, 64.3%, and 63.1% of the to-tal biogas yield was achieved, respectively, for TS of 5%, 10%, 15%,20% and 25%. It can be seen that the retention time of L-AD systemswith 5% and 10% TS was about 25 days for complete digestion,while a 2.0-fold increased retention time was found for the HS-AD systems with 20% and 25% TS. This could be explained thatthe mass transportation in HS-AD was much slower than that inL-AD (Li et al., 2011).

The average methane contents of biogas produced fromco-digestion of food waste and green waste at different TS controllevels are shown in Fig. 4. The statistical analysis shows that the TScontrol level had significant effects (p < 0.05) on methane contents.Methane contents varied from 65.5% to 70.9% at the TS level rangeof 5–25%. The total methane yields were therefore calculated to be215.9, 224.0, 272.1, 246.9 and 176.3 mL/g VS, respectively at TSlevels of 5%, 10%, 15%, 20% and 25% (shown in Fig. 4). The highermethane yields at TS levels of 15% and 20% indicated that methaneproduction from HS-AD could be higher or comparable to theoutput of L-AD (Li et al., 2011; Luning et al., 2003). However, thecontinual increase of TS content from 20% to 25% caused a 35.2%decrease in the methane yield, with respect to the yield at 15%

220 X. Chen et al. / Bioresource Technology 154 (2014) 215–221

TS. Additionally, the methane yield at 25% TS was 18.3% lower thanthat at 5% TS. These results were not completely consistent with aprevious study conducted by Brown et al. (2012). They evaluatedthe methane production from yard waste under L-AD (5% TS) andHS-AD (18–19% TS), and found no significant difference in methaneyield between L-AD and HS-AD. However, the HS-AD of yard wasteat higher TS contents (above 20% TS) was not investigated byBrown et al. (2012). In this study, methane yields obtained fromHS-AD at TS levels of 15–20% were comparable to the yields ofL-AD, while the methane yield started to decrease with an increas-ing TS level to 25%.

The volumetric methane productivities of the digestion systemsin both L-AD and HS-AD are presented in Fig. 5. It can be seen thatvolumetric productivity comparisons generally resemble those ofthe methane yield comparisons at the same TS levels. L-AD at 5%TS showed the lowest volumetric productivity (0.9 Lmethane/Lwork).Increasing the TS level from 5% to 10% increased the volumetricproductivity 2.2-fold (p < 0.05). HS-AD systems at TS levels of15%, 20%, and 25% had similar volumetric productivity (3.6–4.3Lmethane/Lwork), which showed increases of 278–357% comparedto that of L-AD at 5% TS. The higher volumetric productivity ofHS-AD confirmed a main advantage over L-AD due to highervolumetric loading capacity and smaller reactor volume (Brownet al., 2012; Guendouz et al., 2008).

3.3.2. VS reductionThe VS reduction of the co-substrates is also shown in Fig. 5. It

can be seen that the VS reduction values were highly correlated

0

10

20

30

40

50

60

70

0

1

2

3

4

5

6

5 10 15 20 25

VS re

duct

ion

(%)

Volu

met

ric p

rodu

ctiv

ity (L

met

hane

/Lw

ork)

TS control level (%)

Volumetric productivityVS reduction

Fig. 5. Volumetric methane productivity and VS reduction at different TS controllevels.

Table 3Variation of pH, VFA/alkalinity ratio, and TAN concentration during co-digestion of food w

Food/Green mixing ratioa TS level (%) pH

Initial Final

100:0 15 7.66 ± 0.05 8.36 ± 0.0680:20 15 7.76 ± 0.00 8.39 ± 0.0660:40 15 7.81 ± 0.03 8.34 ± 0.0440:60 15 7.81 ± 0.03 8.37 ± 0.0820:80 15 7.88 ± 0.01 8.36 ± 0.000:100 15 7.94 ± 0.04 8.37 ± 0.0040:60 5 7.30 ± 0.01 7.39 ± 0.0840:60 10 8.03 ± 0.01 8.25 ± 0.0140:60 15 7.81 ± 0.03 8.37 ± 0.0840:60 20 7.89 ± 0.04 8.70 ± 0.1140:60 25 7.95 ± 0.04 8.71 ± 0.10

a Based on volatile solids (VS).b mg HAceq/mg CaCO3.

with the methane production yields at the same TS control levels(Fig. 4). Higher VS reduction values were obtained in the digestionsystems with higher methane yields. The highest methane yieldand VS reduction were observed at the TS level of 15%. By increas-ing the TS level from 5% to 15%, a 1.2-fold increase in VS reductionwas achieved. The continual increase of TS content from 15% to25% caused a 17.7% decrease in the VS reduction. Although themethane yield at 25% TS was 18.3% lower than that at 5% TS, theVS reduction of 39.6% at 25% TS was comparable to the VS reduc-tion of 40.4% at 5% TS. The lower methane yield but comparableVS reduction at 25% TS may be due to the conversion of VS intointermediate products such as high concentrations of VFAs (Brownand Li, 2013).

3.3.3. Digestion system characteristicsIt is in the interest of AD plant owners to run the plant at its

operational optimum to maximize methane production, whilemost biogas plants run at a less-than-optimum loading rate to pre-vent instability in the anaerobic digesters resulting from poormonitoring systems (Ward et al., 2008). Monitoring of the keychemical and physical parameters is necessary to achieve optimalcontrol of the AD process. The instability in a disturbed AD processis often caused by excess accumulation of VFAs, resulting in a dra-matic decrease in pH if the buffering capacity of the system is notsufficient (Brown and Li, 2013; Ward et al., 2008). The ratio of totalVFAs to the total inorganic carbonate (VFA/alkalinity) has been rec-ognized as a guide value for assessing process disturbances at anearly stage (Lossie and Pütz, 2010). The VFA/alkalinity ratio is amore reliable parameter for monitoring digester imbalance thansimple measurements of pH, since an accumulation of VFAs willlead to a significant decrease of the buffering capacity before thepH decreases. Therefore, VFA/alkalinity ratios and pH weremeasured in this study to monitor the AD process (Table 3).

The initial pH in the digestion systems at different TS levels ran-ged from 7.3 to 8.0, which were within or close to the optimum pHinterval of 7.0–8.0 recommended by Weiland (2010). After thereaction, a slight increase was observed for the final pH in all thedigestion systems with respect to the initial pH. The digestionsystems had final pH values ranging from 7.4 to 8.7.

As shown in Table 3, the initial and final VFA/alkalinity ratioswere below 0.3 for the digestion systems at TS levels in the rangeof 5–20%, except for the system at 25% TS, where the initial and fi-nal VFA/alkalinity ratios were 0.38 and 0.64, respectively. Accord-ing to the observations and recommendations provided by Lossieand Pütz (2010), VFA/alkalinity ratios between 0.3 and 0.4 aregenerally regarded as optimal for biogas production at a maximum,ratios below 0.3 are regarded as having deficient feedstock input,and ratios exceeding 0.6 are regarded as having excessive

aste and green waste.

VFA/alkalinity ratiob TAN concentration (mg/L)

Initial Final Initial Final

0.16 ± 0.03 0.35 ± 0.03 1865.6 ± 11.1 3456.5 ± 22.20.12 ± 0.09 0.22 ± 0.21 1865.3 ± 66.7 3228.1 ± 11.10.14 ± 0.00 0.17 ± 0.00 1963.5 ± 155.6 3117.4 ± 300.00.06 ± 0.00 0.22 ± 0.06 1946.6 ± 44.4 3423.6 ± 44.40.19 ± 0.06 0.13 ± 0.07 2078.3 ± 55.6 3504.8 ± 77.80.14 ± 0.12 0.14 ± 0.07 2360.2 ± 33.3 3254.9 ± 44.40.27 ± 0.02 0.25 ± 0.02 607.8 ± 32.3 769.2 ± 49.50.08 ± 0.00 0.12 ± 0.01 1488.1 ± 11.1 1853.6 ± 33.30.06 ± 0.00 0.22 ± 0.06 1946.6 ± 44.4 3423.6 ± 44.40.12 ± 0.01 0.15 ± 0.08 2214.7 ± 54.3 2764.9 ± 144.50.38 ± 0.03 0.64 ± 0.28 2954.6 ± 36.8 4243.4 ± 22.2

X. Chen et al. / Bioresource Technology 154 (2014) 215–221 221

feedstock input. As the feedstock/inoculum ratio in this study wasfixed at 1.0, more biogas production may be achieved by increasingthe feedstock/inoculum ratio for the digestion systems at TS levelsin the range of 5–20%, while for the digestion system at 25% TS,increasing the amount of inoculum may improve the gas produc-tion performance. For the HS-AD system at 25% TS, although thefinal pH was high (pH = 8.7), a high final VFA/alkalinity ratio of0.64 was found, which was also reflected by the low methane yield(Fig. 4). This was likely caused by the accumulation of VFAs due tooverfeeding.

Ammonia accumulation is potentially encountered during AD ofN-rich feedstock due to the degradation of proteinaceous materials(Sung and Liu, 2003). An optimal ammonia concentration ensuressufficient buffering capacity in the AD system thus increasing thestability of the AD process, while high ammonia is reported as astrong inhibitor of biogas production (Chen et al., 2008; Rajagopalet al., 2013). Chen et al. (2008) reported that the inhibiting totalammonia nitrogen (TAN) concentration that caused a 50% decreasein the methane yield varied much from 1.7 to 14 g/L. Most of thestudies on ammonia inhibition have been focused on traditionalL-AD over the past few decades. As seen in Table 3, the initialand final TAN concentrations in the digestion systems generally in-creased with the increase of TS levels from 5% to 25%. The digestionsystem at 25% TS had a 5.5-fold higher final TAN concentrationthan that of the system at 5% TS. It seems that ammonia inhibitionis more likely to be encountered in HS-AD, since lower watercontent affects dilution (Wang et al., 2013). Considering that themethane yields in HS-AD at 15–20% TS were higher or comparableto the yields in L-AD, and that the methane yield started todecrease with further increasing TS level to 25% (Fig. 4), it can beconcluded that a higher TAN concentration (4.2 g/L) in the diges-tion system at 25% TS may initiate inhibition of methanogenesis,leading to lower methane yields. Consequently, careful consider-ation should be taken to avoid ammonia inhibition of the HS-ADprocess.

4. Conclusion

The optimal performance for co-digestion of food waste andgreen waste was achieved at their mixing ratio of 40:60. Under thispreferred ratio, the effect of TS content (5–25%) on anaerobic co-digestion was investigated in batch systems. The results indicatethat methane yields from HS-AD (15% and 20% TS) were higheror comparable to the output of L-AD (5% and 10% TS), while 25%TS content corresponded to a threshold at which methane produc-tion was inhibited. Considering the volumetric productivity, HS-ADsystems (15–25% TS) showed increases of 278–357% compared tothat of L-AD at 5% TS.

Acknowledgements

This work was financially supported by National Science &Technology Pillar Program of China (No. 2012BAC17B02). Theauthors also would like to thank Mrs. Eileen Deaner (Departmentof Design Sciences, Lund University) for proofreading and languagecorrection.

References

Alvarez, R., Liden, G., 2008. Semi-continuous co-digestion of solid slaughterhousewaste, manure, and fruit and vegetable waste. Renew. Energy 33, 726–734.

APHA, 1998. Standard Methods for the Examination of Water and Wastewater,20th ed. American Public Health Association, Washington DC, USA.

Brown, D., Shi, J., Li, Y., 2012. Comparison of solid-state to liquid anaerobic digestionof lignocellulosic feedstocks for biogas production. Bioresour. Technol. 124,379–386.

Brown, D., Li, Y., 2013. Solid state anaerobic co-digestion of yard waste and foodwaste for biogas production. Bioresour. Technol. 127, 275–280.

Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: areview. Bioresour. Technol. 99, 4044–4064.

Fernández, J., Pérez, M., Romero, L.I., 2010. Kinetics of mesophilic anaerobicdigestion of the organic fraction of municipal solid waste: influence of initialtotal solid concentration. Bioresour. Technol. 101, 6322–6328.

Fdez-Güelfo, L.A., Álvarez-Gallego, C., Sales, D., Romero García, L.I., 2012. Dry-thermophilic anaerobic digestion of organic fraction of municipal solid waste:methane production modeling. Waste Manage. 32, 382–388.

Forster-Carneiro, T., Pérez, M., Romero, L.I., 2008. Influence of total solid andinoculums contents on performance of anaerobic reactors treating food waste.Bioresour. Technol. 99, 6974–7002.

Guendouz, J., Buffiere, P., Cacho, J., Carrere, M., Delgenes, J.P., 2008. High-solidsanaerobic digestion: comparison of three pilot scales. Water Sci. Technol. 58,1757–1763.

Karthikeyan, O.P., Visvanathan, C., 2013. Bio-energy recovery from high-solidorganic substrates by dry anaerobic bio-conversion processes: a review. Rev.Environ. Sci. Biotechnol. 12, 257–284.

Kim, D.H., Oh, S.E., 2011. Continuous high-solids anaerobic co-digestion of organicsolid wastes under mesophilic conditions. Waste Manage. 31, 1943–1948.

Labatut, R.A., Angenent, L.T., Scott, N.R., 2011. Biochemical methane potential andbiodegradability of complex organic substrates. Bioresour. Technol. 102, 2255–2264.

Li, Y., Park, S.Y., Zhu, J., 2011. Solid-state anaerobic digestion for methaneproduction from organic waste. Renew. Sust. Energ. Rev. 15, 821–826.

Liu, G., Zhang, R., El-Mashad, H.M., Dong, R., 2009. Effect of feed to inoculum ratioson biogas yields of food and green wastes. Bioresour. Technol. 100, 5103–5108.

Lossie, U., Pütz, P., 2010. Targeted control of biogas plants with the help of FOS/TAC.Practice Report Hach-Lange.

Luning, L., van Zundert, E.H.M., Brinkmann, A.J.F., 2003. Comparison of dry and wetdigestion for solid waste. Water Sci. Technol. 48, 15–20.

Martin, D.J., Potts, L.G.A., Heslop, V.A., 2003. Reaction mechanisms in solid-stateanaerobic digestion: II. The significance of seeding. Process Saf. Environ. 81,180–188.

Pantaleo, A., Gennaro, B.D., Shah, N., 2013. Assessment of optimal size of anaerobicco-digestion plants: an application to cattle farms in the province of Bari (Italy).Renew. Sust. Energ. Rev. 20, 57–70.

Pohl, M., Heeg, K., Mumme, J., 2013. Anaerobic digestion of wheat straw –performance of continuous solid-state digestion. Bioresour. Technol. 146, 408–415.

Rajagopal, R., Massé, D.I., Singh, G., 2013. A critical review on inhibition of anaerobicdigestion process by excess ammonia. Bioresour. Technol. 143, 632–641.

Schievano, A., D’Imporzano, G., Malagutti, L., Fragail, E., Ruboni, G., Adani, F., 2010.Evaluating inhibition conditions in high-solids anaerobic digestion of organicfraction of municipal solid waste. Bioresour. Technol. 101, 5728–5732.

Shi, J., Wang, Z., Stiverson, J.A., Yu, Z., Li, Y., 2013. Reactor performance andmicrobial community dynamics during solid-state anaerobic digestion of cornstover at mesophilic and thermophilic conditions. Bioresour. Technol. 136, 574–581.

Sung, S., Liu, T., 2003. Ammonia inhibition on thermophilic anaerobic digestion.Chemosphere 53, 43–52.

Van Soest, P.J., Robertson, J.B., Lewis, B.A., 1991. Methods for dietary fiber, neutraldetergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J.Dairy Sci. 74, 3583–3597.

Voß, E., Weichgrebe, D., Rosenwinkel, K.H., 2009. FOS/TAC–Deduction, Methods,Application and Significance. Available from: <http://www.ve-gmbh.de>.

Wan, C., Zhou, Q., Fu, G., Li, Y., 2011. Semi-continuous anaerobic co-digestion ofthickened waste activated sludge and fat, oil and grease. Waste Manage. 31,1752–1758.

Wang, Z., Xu, F., Li, Y., 2013. Effects of total ammonia nitrogen concentration onsolid-state anaerobic digestion of corn stover. Bioresour. Technol. 144, 281–287.

Ward, A.J., Hobbs, P.J., Holliman, P.J., Jones, D.L., 2008. Optimisation of the anaerobicdigestion of agricultural resources. Bioresour. Technol. 99, 7928–7940.

Weiland, P., 2010. Biogas production: current state and perspectives. Appl.Microbiol. Biotechnol. 85, 849–860.

Wu, G., Healy, M.G., Zhan, X., 2009. Effect of the solid content on anaerobic digestionof meat and bone meal. Bioresour. Technol. 100, 4326–4331.

Xu, L., Li, Y., 2012. Solid-state co-digestion of expired dog food and corn stover formethane production. Bioresour. Technol. 118, 219–226.