Download - Effects of lipid concentration on anaerobic co-digestion of municipal biomass wastes

Waste Management xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

Waste Management

journal homepage: www.elsevier .com/locate /wasman

Effects of lipid concentration on anaerobic co-digestion of municipalbiomass wastes

0956-053X/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.wasman.2013.07.018

⇑ Corresponding author.E-mail address: [email protected] (Y. Sun).

Please cite this article in press as: Sun, Y., et al. Effects of lipid concentration on anaerobic co-digestion of municipal biomass wastes. Waste Mana(2013), http://dx.doi.org/10.1016/j.wasman.2013.07.018

Yifei Sun a,⇑, Dian Wang a, Jiao Yan a, Wei Qiao b, Wei Wang c, Tianle Zhu a

a School of Chemistry and Environment, Beihang University, Beijing 100191, Chinab College of Chemical Science and Engineering, China University of Petroleum, Beijing 102249, Chinac School of Environment, Tsinghua University, Beijing 100084, China

a r t i c l e i n f o

Article history:Available online xxxx

Keywords:Anaerobic digestionLipidLCFAInhibitionMunicipal biomass waste

a b s t r a c t

The influence of the lipid concentration on the anaerobic co-digestion of municipal biomass waste andwaste-activated sludge was assessed by biochemical methane potential (BMP) tests and by bench-scaletests in a mesophilic semi-continuous stirred tank reactor. The effect of increasing the volatile solid (VS)concentration of lipid from 0% to 75% was investigated. BMP tests showed that lipids in municipal bio-mass waste could enhance the methane production. The results of bench-scale tests showed that a lipidsconcentration of 65% of total VS was the inhibition concentration. Methane yields increased with increas-ing lipid concentration when lipid concentrations were below 60%, but when lipid concentration was setas 65% or higher, methane yields decreased sharply. When lipid concentrations were below 60%, the pHvalues were in the optimum range for the growth of methanogenic bacteria and the ratios of volatile fattyacid (VFA)/alkalinity were in the range of 0.2–0.6. When lipid concentrations exceeded 65%, the pH valueswere below 5.2, the reactor was acidized and the values of VFA/alkalinity rose to 2.0. The amount of Brev-ibacterium decreased with increasing lipid content. Long chain fatty acids stacked on the methanogenicbacteria and blocked the mass transfer process, thereby inhibiting anaerobic digestion.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Over the past years, the output of municipal solid waste (MSW)in China has increased rapidly. The Yearbook of the National Bu-reau of Statistics shows that MSW exceeded 158 million tons in2010, almost 2.5 times the 1990 values (National Bureau of Statis-tics of China, 2011). As consumption habits and the fuel structurechange, the components and properties of MSW have also under-gone tremendous changes. About 50–60% of MSW is composed ofmunicipal biomass waste (MBW), such as food waste (FW) andfruit/vegetable residues (FVR) (Liu et al., 2012).

The main component of FW is animal fat, especially the case inChina. Chinese eating habits determine their way of cooking, lead-ing to a relatively high proportion of lipids in FW, including animalfats, vegetable lipids and other ingredients. According to theirform, food waste lipids can be classified into five categories asfloating, dispersed, emulsifying, soluble or oily solid substances(Ren et al., 2006). Among them, floating lipid droplets have a largediameter and can quickly float to form a continuous phase of oilfilm floating on the water. Dispersed lipids have bigger dropletswith diameter exceeding 1 lm, suspended inside the aqueousphase; the particle size of emulsified lipid is between 0.5 and

15 lm. Soluble lipid disperses its molecules in water to form ahomogeneous system which is difficult to separate. Oily solidwaste mostly combines solid lipid with solid garbage, and the lipidcan hardly be directly removed.

In recent years, high-solids anaerobic digestion (AD) technol-ogy, targeting biomass waste like FW, FVR, garden waste and oth-ers, has seen rapid development in Europe, the United States andother countries (De Baere and Mattheeuws, 2008). However, con-trary to common feedstock of traditional AD, FW has a high lipidcontent, which affects the AD. Lipids have a higher biochemicalmethane potential compared to other organic matter. Co-digestionof thickened waste activated sludge (TWAS) and fat, oil and grease(FOG) was studied previously. The results showed that specificmethane yields (SMY) of the TWAS and FOG co-digestion was137% higher than that of TWAS alone (Wan et al., 2011). Neverthe-less, the negative impact will becomes more obvious since the low-er degradation rate of lipid in comparison to e.g. sugar or protein,could easily lead to its accumulation. During the AD process, lipidwill accumulate and float as the lipid concentration increases(Cammarota et al., 2001; Zeng et al., 2007). Accordingly, at the mi-cro-level, the hydrophobic lipid, wraps or adsorbs onto the surfaceof microorganism and blocks the mass transfer process betweensoluble organic matter and microbial cells (Pereira et al., 2004).These soluble intermediates are not able to enter microbial cellsfor timely decomposition, which in turn exacerbates the

gement

http://dx.doi.org/10.1016/j.wasman.2013.07.018

mailto:[email protected]


http://www.sciencedirect.com/science/journal/0956053X

http://www.elsevier.com/locate/wasman


Table 1Characteristics of the raw materials and seed sludge.

Parameters Food waste Waste activated Sludge (WAS) Fruit/vegetable waste Mixture (2:1:1) Seed sludge

TS (g kg�1) 197.12 145.83 91.56 157.91 27.01VS (g kg�1) 170.43 106.31 77.28 131.11 10.45

Protein (g kg�1) 29.17 49.30 10.18 29.45 –Lipid (g kg�1) 40.72 14.73 2.39 24.64 –Carbohydrate (g kg�1) 100.54 42.28 64.71 77.02 –

VS/TS (%) 86.46 72.90 84.40 83.03 38.69SS (g kg�1) 79.82 136.03 55.31 87.75 24.97VSS (g kg�1) 75.64 105.03 53.46 80.29 9.71T-COD (g L�1) 106.01 124.72 114.64 115.00 33.29S-COD (g L�1) 97.50 2.75 62.81 56.47 1.82TN (g kg�1) 35.15 51.27 36.50 39.52 —TP (g kg�1) 3.94 17.23 4.97 7.53 —

Table 2Seed sludge, substrate added and methane prducitons in BMP tests.

Lipid concentration 0% 1% 2% 3% 4% 5% 10% 18%

Lipid added (g) 0.000 0.015 0.030 0.045 0.060 0.075 0.150 0.270Substrate after removing the lipid added (g) 13.62 13.49 13.35 13.22 13.08 12.94 12.26 11.17VS (g kg�1) 110.10 111.09 112.10 113.12 114.16 115.23 120.85 131.10Seed sludge added (g) 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0Methane productions (mL g�1 VS) 269.9 301.9 310.3 336.0 348.6 355.5 387.8 486.3

1-Peristaltic pump 2- Reactor 3-Magnetic stirrer

4-Biogas bag 5- Control valve 6- Discharge tank

7- Water circulating pump 8- Circulating water

Fig. 1. Schematic diagram of bench-scale reactor.

2 Y. Sun et al. / Waste Management xxx (2013) xxx–xxx

accumulation of intermediate products and therefore underminesthe continuity of the methane production process. A previousstudy of the mesophilic co-digestion of grease trap sludge (GTR)with sewage sludge, having a lipid volatile solid (VS) concentrationof 5–55%, was conducted by Luostarinen et al (2009). The resultsshowed that when the GTR concentration was below 46%, SMY in-creased with the increasing GTR concentration, but when GTR con-centration was raised to 55%, SMY began to decrease. Wang et al.(2013) studied the co-digestion of dairy manure, chicken manure,swine manure and rice straw, the results showed that methanepotentials of co-digestion were higher than that of single substrate.Zhang et al. (2013) proposed a dual solid–liquid (ADSL) system,and used this system anaerobic digested the food solid waste and

Please cite this article in press as: Sun, Y., et al. Effects of lipid concentration on(2013), http://dx.doi.org/10.1016/j.wasman.2013.07.018

food liquid waste: the results showed that the total methane pro-duction of ADSL system was 13.6% higher than that of raw foodwaste, resulting from the optimum C/N ratio in ADSL system.

Lipids co-digested with other organic matter could enhanceSMY owing to the higher biochemical methane potential. But whenthe lipid concentration was too high, the AD process would beinhibited. The present study tries to determine the lipid inhibitionconcentration for mesophilic AD and explores the inhibition mech-anism of lipids in AD. The co-digestion of FW, FVR, waste activatedsludge (WAS) and edible oil (EO) was conducted semi-continu-ously under mesophilic conditions. The main factors of inhibitionwere explained, including the variation of the amount and typeof bacteria.

anaerobic co-digestion of municipal biomass wastes. Waste Management


0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Met

hane

pro

duct

ion

ml/(

g V

S)

Digestion time (h)

LC 0% LC 1% LC 2%

LC 3% LC 4% LC 5%

LC 10% LC 18%

Fig. 2. Methane generation of municipal biomass wastes with different LC in biochemical methane potential tests.

Y. Sun et al. / Waste Management xxx (2013) xxx–xxx 3

2. Materials and methods

2.1. Material

The substrate used in this study was composed of FW, FVR,and WAS. The FW was first screened to remove bones, plastic,and metal and was then crushed. The FVR was directly crushedafter collection. The substrate was mixed with the crushed FW,the crushed FVR, and WAS in a 2:1:1 ratio. The FW was takenfrom a student canteen at Tsinghua University. FVR was collectedfrom a farmers market in the Changping district, Beijing. TheWAS was supplied by Beixiaohe municipal wastewater treatmentplant in the Changping district, Beijing, which treats 10 � 104 m3

municipal wastewater per day. The seed sludge was supplied bythe Xiaohongmen municipal wastewater treatment plant. Thecharacteristics of the substrate and seed sludge are shown inTable 1.

2.2. Methods

The biogas production was measured using a wet type gas me-ter (LMF-2, Changchun Automobile Filter, Changchun, Jilin, China).The gas composition was analyzed using a gas chromatography(GC)-thermal conductivity detector system (GC-17A, ShimadzuCorp., Japan). The system was equipped with a RT-Qplot(30 m � 0.53 mm, Restek, Bellefonte, PA, USA) column. Nitrogenwas used as the carrier gas. Methane production was obtained bymultiplying the biogas production by the corresponding methanecontent in the produced biogas. The values of methane productionwere converted to the standard temperature and pressure condi-tions (0 �C, 101.325 kPa). The volatile fatty acid (VFA) and longchain fatty acid (LCFA) were measured by GC (GC-17A, ShimadzuCorp., Japan) equipped with a flame ionization detector and aDB-FFAP column (30 m � 0.32 mm � 0.25 lm, Agilent Technolo-gies, California, USA). Total nitrogen (TN), total phosphorus (TP),pH, total solids (TS), suspended solids (SS), volatile solid (VS), vol-atile suspended solids (VSS) and alkalinity were measured accord-ing to US national standard (State EPA, 2005). Crude protein andcrude fat contents were determined according to the methods ofISO 1871:2009 and ISO 6492:1999, respectively. Organic com-pounds are composed of carbohydrate, crude protein, and crudefat, thus allowing the calculation of the carbohydrate content.The surface of the AD sludge was examined using scanning


electron microscopy (SEM) (Apollo 300, Camscan). For the denatur-ing gradient gel electrophoresis (DGGE) analysis, the DNA of theanaerobion was extracted using a Fast DNA SPIN Kit for Soil (MPBiomedicals, LLC, Illkirch, France) according to the protocoldescribed by the manufacturer. DGGE was carried out with theBio-Rad Dcode system (Bio-Rad Laboratories Ltd.). The thermocy-cling experiment conditions of thermocycling were as follows:94 �C for 4 min, then 20 cycles including 1 min at 94 �C, 1 min ofannealing, and 2 min at 72 �C. The annealing temperature wassubsequently decreased from 65 to 55 �C. When annealingtemperature was decreased to 55 �C additional 5 cycles wereperformed. Finally, a 10 min extension step at 72 �C wasconducted.

2.3. Biochemical methane potential tests

According to the Biochemical methane potential (BMP) test(Wang and Wang, 2005), 250 mL serum bottles with temperatureof 35 �C were used as anaerobic reactor, which contained the welladapted seed inoculums from the pilot-scale continuous stirredtank reactor. The biogas accumulated at the top of the water-filled500 ml graduated cylinder. The volume of gas produced was readfrom the graduated cylinder. All assays were shaken by hand twotimes per day and the volume of gas produced was recorded everyday. In order to compare the performance of all materials, the sam-ples added into the BMP reactors had the same amount of 1.5 g VSfor solid phase biomass waste. The quality of seed sludge, substrateand lipid added were shown in Table 2. To account for the accuracyof the experiments, triplicates were run for all samples, and aver-age values are reported.

2.4. Bench-scale tests

Semi-continuous AD was carried out in a 6 L reactor with aworking volume of 4 L. The reactor has a jacketed structure.The inner jacket of the reactor was used for the co-digestionof FW, FVR, WAS, and EO. The outer jacket was used as thermo-static system to maintain the reactor at 35 �C. The reactor wasconstantly mixed with an agitation time of 15 min per 2 hr.The bench-scale reactor set-up is shown in Fig. 1. The reactorwas filled with 4 L of seed sludge, and the headspace of the reac-tor was flushed with nitrogen for 5 min to create an anaerobicenvironment.



0

200

400

600

800

1,000

1,200

0 5 10 15 20 25 30

Bio

gas

(L)

Digestion time (d)

(a)LC 60%

LC 55%

LC 40%LC 25%LC 18%

LC 65%LC 70%LC 75%

0

100

200

300

400

500

600

700

800

0 5 10 15 20 25 30

Met

hane

pro

duct

ion

(ml /

g V

S)

Digestion time (d)

(b)LC 60%

LC 55%

LC 40%LC 25%LC 18%

LC 65%

LC 70%

LC 75%

0

100

200

300

400

500

600

700

800

900

18% 25% 40% 55% 60% 65% 70% 75%

Met

hane

yei

lds

Met

hane

pro

duct

ion

(ml/g

VS)

Lipid concentration (%VS)

(c)Actual Methane yields

Fig. 3. Biogas and methane production in the anaerobic co-digestion (a) cumulative biogas production (b) daily methane production (c) actual and methane yields.

Table 3Gas productions of different lipid concentration substrate.

Lipid concentration 18% 25% 40% 55% 60% 65% 70% 75%

Biogas cumulative production (L) 617.2 712.9 801.3 978.1 1031.8 271.6 154.3 85.8Enhanced compared to the raw MBW (%) 100 116 130 158 167 44 25 14Daily methane productions mL�(g VS d)�1 420.3 446.3 515.4 624.5 706.1 35.1 29.5 30.6Theoretical daily methane productions mL�(g VS d)�1 497.4 581.5 668.0 754.5 783.3 812.1 841.0 869.8Methane contents (%) 64 64 64 64 64 56 55 54


3. Results and discussion

3.1. BMP tests of municipal biomass wastes with different lipidconcentration

The BMP test was carried out to evaluate the methane productionpotential of raw biomass waste before and after lipid elimination.


Lipid extracted by the hydrothermal process diluted and put intomixture biomass waste after lipid elimination. Each vial was added100 mL seed sludge, the quality of the mixed MBW and lipid addedwas shown in Table 2. The lipid concentration (of total VS) of sam-ples was set as 0%, 1%, 2%, 3%, 4%, 5%, 10%, and 18%. Fig. 2 showsthe methane generation amount of BMP for mixed biomass wasteadded lipid with different concentration. During the first 24 hr, the



53

56

59

62

65

68

CH

4(%

)

Digestion time (d)

LC 18%

LC 25%LC 40%

LC 65%

LC 75%

LC 70%

LC 55%

Fig. 4. Variation of methane content of municipal biomass wastes with different LC in bench-scale tests.


methane productions of all lipid concentration increased rapidlyand the amounts of methane production were nearly the same.But methane production differences appeared as the reaction pro-gressed. From 25 hr on, the methane production of all lipid concen-tration increased gradually and tended to stabilize. When thedigestion reacted 143 hr, there was no biogas produced in all reac-tion vials. The CH4 production of different lipid concentrations wereshown in Table 2, the corresponding average deviations were nothigher than 7.5, 12.5, 5.0, 12.5, 15.0, 10.0, 17.5, and 17.5. It indicatedthat with the increment of lipid concentration, methane generationamounts increased. This was because content of carbohydrate, pro-tein, and lipid were different. Theoretical CH4 productions of carbo-hydrate, protein, and lipid could be calculated by Eq. (1) (Buswelland Neave, 1930).

CnHaOb þ n� a4� b

2

� �H2O! n

2þ a

8� b

4

� �CH4

þ n2� a

8þ b

4

� �CO2 ð1Þ

Theoretical CH4 productions of carbohydrate, protein, and lipidwere 415, 496, and 1014 mL g VS�(m3 d)�1. (Angelidaki and Sand-ers, 2004). It indicated that the BMP of lipid was higher than thatof carbohydrate and protein. The BMP test results showed thatmethane production increased with the increasing lipid concentra-tion, which was consistent with the above conclusion.

BMP tests showed that when lipid concentration did not exceed18%, AD process proceeded well and methane production increasedwith the increment of lipid concentration. The lipid concentrationof raw mixture biomass waste was 18%. So we can conclude thatMBW with lipid concentration below 18% could be directly di-gested anaerobically and there was no inhibition phenomenon.Theoretically, the more lipids were added, the more methanewould be produced. To assess whether the lipid concentrationcould be increased limitlessly or not, additional bench-scale testswere conducted with adding oil to the mixed biomass wastes.

3.2. Effects of lipid concentration on biogas and methane production

The reactor was started with a relatively low organic loadingrate (OLR), and then with increasing OLR of 1.5, 2.5, 4.0, 6.0, 8.0,and 10.0 kg VS�(m3 d)�1. The operation period was three timesthe corresponding hydraulic retention time (HRT) for each OLR.


The reactor was started up with MBW of FW, FVR and WAS. Whenthe reactor stabilized at 10.0 kg VS�(m3 d)�1, the seed sludge waskept at 35 �C in an anaerobic environment for 2–4 days beforethe tests started until there was no significant biogas produced.Then, feeding of different lipid concentration was conducted onceper day. At this time, MBW with different lipid concentrationswas used as substrate.

The biogas cumulative production (BCP) was used to measurethe amount of biogas produced since the digestion started. Asshown in Fig. 3(a), BCPs of different lipid concentration increasedgradually. When the lipid concentration was below 60%, no differ-ences of BCP with various lipid concentration were found in thefirst 5 days. From the sixth day, BCP increased with increasing lipidconcentration. When the lipid concentration exceeded 65%, BCPdecreased with increasing lipid concentration. Steady states of allreactors were achieved after approximately two HRT from initialfeeding. After 25 days, BCPs of different lipid concentration sub-strate was shown in Table 3. The results showed that when the li-pid concentration was below 60%, AD reactors ran stably, but whenthe lipid concentration exceeded 65%, the AD process was inhib-ited. Previous studies (Wan et al., 2011), showed that the digesterfailed at a lipid VS concentration of 74%, which almost coincidedwith the present study Daily methane production, expressed inmL CH4�(m3 d)�1, was used to show how effectively the MBWwas converted to CH4. Daily methane production of varying lipidconcentration was indicated in Fig. 2(b). When the lipid concentra-tion was below 60%, the daily methane production increased rap-idly in the first 5 days and gradually became stable. During thefirst 10 days, the daily methane production was lower than thatin the following days. When the lipid concentration exceeded65%, the daily methane production increased during the first10 day and then decreased rapidly. The cycle of lipid degradationwas long and the degradation process had a given lag. The degra-dation rate of lipid was lower than that of protein and carbohy-drate. When the lipid concentration was raised to 65%, the dailymethane production was lower than that of substrates with low li-pid concentration. Daily methane productions of substrates withhigher lipid concentration decreased rapidly between day 10 and20 owing to high levels of VFA and LCFA resulting from hydrolysisof lipids, as further detailed in Section 3.4. All reactors reached astable state between day 21 and 25. During this period, daily meth-ane productions stabilized at different values and that of substratewith 60% lipid concentration was the highest.



pH

Digestion time (d)

(a)

LC 18%

LC 25%

LC 40%

LC 55%

LC 60%

LC 65%

LC 70%

LC 75%

LC 65%LC 75%

LC 70%

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0 2 4 6 8 10 12 14 16 18 20 22 24 26

VF

A (

mg/

l)

Digestion time (d)

(b)LC 18% LC 25%LC 40% LC 55%LC 60% LC 65%LC 70% LC 75%

LC 65%

LC 70%

LC 75%

5000

6000

7000

8000

9000

10000

11000

12000

0 5 10 15 20 25 30

Alk

alin

ity

(mg/

L a

s C

aCO

3)

Digestion time (d)

(c)LC 18%LC 25%LC 40%LC 55%LC 60%

LC 70%

LC 75%

LC 65%

pHDigestion time (d)

0.0

0.4

0.8

1.2

1.6

2.0

VF

A/A

lkal

init

y

Digestion time (d)

(d)LC 18% LC 25%LC 40% LC 55%LC 60% LC 65%LC 70% LC 75%

LC 65%LC 75%

LC 70%

Fig. 5. Variation of pH (a), volatile fatty acids (VFA) (b), alkalinity (c), and VFA/alkalinity (d) of municipal biomass wastes with different LC in bench-scale tests.


Please cite this article in press as: Sun, Y., et al. Effects of lipid concentration on anaerobic co-digestion of municipal biomass wastes. Waste Management(2013), http://dx.doi.org/10.1016/j.wasman.2013.07.018


Myr

isti

c ac

id (

mg/

L)

Digestion time (d)

(a)LC 18%

LC 60%

Ole

ic a

did

(mg/

L)

Digestion time (h)

(b)

LC 18%

LC 60%

Fig. 6. Variation of myristic acid (a) and oleic acid (b) of municipal biomass wastes with 18% and 60% LC in bench-scale tests.


Fig. 3(c) shows the actual daily methane production, and varia-tion of methane yields, with increasing lipid concentration. Dailymethane productions at different lipid concentration of substratewere shown in Table 3. Based on Table 1 and equation (2), the the-oretical daily methane production of different lipid concentrationswas calculated. Theoretical daily methane productions of substratewere also shown in Table 3. The organic material was usually de-graded into its most oxidized form CO2 and its most reduced formCH4. The oxygen-to-carbon ratio of carbohydrate, protein, and lipidwere 0.83, 0.40, and 0.11. If this ratio was smaller, the degradationproducts would be more degraded into its most reduced form CH4.Thus, Theoretical daily methane productions increased withincreasing lipid concentration. In fact, when lipid concentrationwas raised to 65%, daily methane production decreased rapidlyand showed the opposite trend compared to theoretical dailymethane productions. Methane yields of substrates with 65%,70%, and 75% lipid concentration were only 4.3%, 3.5%, and 3.5%.The results showed that when lipid concentration was below60%, the AD reactor ran well and methane yields were over 75%,but when lipid concentration exceeded 65%, the AD process wasinhibited severely.

As shown in Fig. 4, methane contents decreased in the first3 days, and then increased gradually to a normal level when the li-pid concentration was below 60%. During the reactor set-up stage,significant amounts of H2, CO2, and acetic acid were produced dueto hydrolysis and acidogenesis of the substrate. The increasing ace-tate promoted the acetoclastic methanogenesis process. Theseprocesses led to an increase in the CO2 concentration and a


decrease in the CH4 concentration. But when the lipid concentra-tion exceeded 65%, methane contents decreased continuously dur-ing the whole AD process. From day 21 to 25, methane contentsstabilized at a given concentration. During this period, methanecontents were shown in Table 3. The relative standard deviationswere ±0.6%, 0.4%, 0.4%, 0.3%, 0.9%, 0.3%, 0.4%, and 0.5%. It was dem-onstrated that lipid concentration below 60% has no effects on bio-gas content. But when lipid concentration exceeded 65%, methanecontents were lower than normal level. Substrates with high lipidconcentration produced plenty of CO2 as a result of hydrolysiswhich was not utilized by methanogens because the methanogen-esis was inhibited owing to the excess LCFA.

3.3. Effects of lipid concentration on the characteristics of AD fluid

As shown in Fig. 5(a), when the lipid concentration was below60%, the pH remained in the optimum range (6.9–7.8) during thewhole AD process. The pH values in the reactors with 65%, 70%,and 75% lipid concentration began to be lower than the optimumrange in 11d, 7d, and 4d, respectively. The pH value continues todrop during the future days. When the lipid concentration was be-low 60%, pH values decreased in the first 7 days and then revertedto the normal levels. With lipid concentrations increasing from 18%to 60%, pH values decreased in turn. During the first 7 days, sub-strates hydrolyzed and produced abundant VFA, resulting in thereducing pH values. Along with utilization of VFA by methanogens,pH values recovered to normal levels. LCFAs were significantly pro-duced owing to hydrolysis of lipid. LCFAs converted to VFAs which



Fig. 7. Scanning electron microphotographs of anaerobic sludge after digested 25day in bench-scale tests: (a) lipid concentration of 18%, (b) lipid concentration of60%, (c) lipid concentration of 75%.


mainly contributed to the pH values of the system. Thus, pH valuesdecreased with increasing lipid concentration.

The VFA concentrations of all reactors are shown in Fig. 5(b).When the lipid concentration was lower than 60%, the productionrate of acids was faster than its consumption rate during the first7 days owing to the hydrolysis of the readily degradable organicmaterials. Thus, VFA concentrations increased rapidly. Between


day 7 and 25, along with the utilizing of VFA by methanogens,the VFA concentration decreased and stabilized gradually. Whenthe lipid concentration exceeded 65%, the levels of VFA increasedcontinuously. The production rate of acids was faster than the con-sumption rate during the first 20 days. Combined with Fig. 3(b), weknew that daily methane production of substrates with 65%, 70%,and 75% lipid concentration began to decreased from day 10. Itcould be concluded that methanogenesis was inhibited betweenday 10 and 20. From Fig. 5(b), when the lipid concentration ex-ceeded 65%, VFA concentrations remained in relatively high levelsand changed little during the following days. It could be obtainedthat acidogenesis was also inhibited between day 21 and 25.

As shown in Fig. 5(c), alkalinity decreased firstly and then in-creased gradually and stabilized in the end when lipid concentra-tions were below 60%. But when the lipid concentrationexceeded 65%, the alkalinity decreased continuously. The ratio ofVFA/alkalinity was typically used as a measurement to evaluateanaerobic system stability. When the ratio is between 0.2 and0.6, the process is stable and without the risk of acidification (Los-sie and Pütz, 2008). As shown in Fig. 5(d), when the lipid concen-tration was lower than 60%, VFA/alkalinity was less than 0.6, andstable operation was obtained. The system would endure higherlevels of VFAs if the concentration of alkalinity was high. The sys-tem has better buffering capacity. In this research, the alkalinity atlipid concentrations of 70% and 75% were 7801.1 and8225.6 mg L�1 in 11d. Especially the maximum concentration ofVFA, which the system could endure were 4680.6 and4935.3 mg L�1. However, the actual concentrations of VFA were6004.8 and 6712.7 mg L�1. The balance of the system was de-stroyed and daily methane production began to decrease.

3.4. Effect of the lipid concentration on the Inhibition mechanism

LCFA is the hydrolysis product of lipid. As shown in Fig. 6, con-centrations of myristic acid increased from 351.6 to 482.6 mg L�1

during the first 8 days and then decreased to a normal level of304.8 mg L�1. When the lipid concentration was increased to 60%,concentrations of myristic acid merely changed. It could be con-cluded that myristic acid was not the main LCFA which inhibitedthe AD process. Oleic acid increased sharply when the lipid concen-tration was increased from 8% to 60%. At day 11, oleic acid concen-trations were 1403 and 3207 mg L�1 for lipid concentration with18% and 60%. Previous studies showed that a reduction of 50% ofmethanogens activities was found when the oleate concentrationwas raised from 50 to 200 mg L�1 (Alves et al., 2001). The resultsshowed that oleic acid was the most toxic LCFA. Previous researchpresented similar conclusions (Angelidaki and Ahring, 1992).

During the AD process, microorganism played an importantrole. Scanning electron microphotographs (SEM) of AD fluid ofday 25 with 18%, 60%, and 75% lipid concentration were shownin Fig. 7. Brevibacterium was determined by SEM. As shown inFig. 7, abundant, few, and no Brevibacterium could be visualizedwith lipid concentration of 18%, 60%, and 75%. This was becausewhen lipid concentration was too high, lipid hydrolyzed and plentyof LCFA was produced. The microorganisms were ‘‘encapsulated’’by the accumulated LCFA, thereby being blocked towards masstransfer.

DGGE profiles on eubacterial 16S rDNA from samples of lipidconcentration of 18% and 60% were shown in Fig. 8. The degrada-tion process of different organic materials required different kindsof microorganisms. The concentration of lipid, proteins, and carbo-hydrates were 18%, 22%, 60% and 60%, 11%, 29% for Fig. 8(a) and (b),respectively. The dominant bacterial community was different ow-ing to the differences of the concentration of organic material forlipid concentration with 18% and 60%. The species of microorgan-ism decreased a little with increasing lipid concentration. This



e

a

b

c

f

g h

m n

z

b

h

(a) (b)

d

Fig. 8. DGGE profiles on eubacterial 16S rDNA of anaerobic sludge with lipid concentration of 18% (a) and 60% (b) after digested 25 day in bench-scale tests.


was because when lipid concentration was increased to 60%, the li-pid content was the dominant material and lipid degradation bac-teria was the main microorganisms in the AD reactors. Stripe a, c, f,g nearly disappeared when lipid concentration was raised to 60%. Itshowed that these four types of bacteria could degradate proteinsand carbohydrates better. Stripe b, d, e, m, n, z appeared when lipidconcentration was increased to 60%, which indicated that these sixtypes of bacteria could degradate lipid better. When lipid concen-tration was enhanced to 60%, Stripe h barely changed. It revealedthat this type of bacteria could degradate the three kinds of organicmatter better. A bacteria did not die while changed to another bac-teria when lipid concentration was increased.

In conclusion, it was LCFA that inhibited the AD process and ledto AD reactor failure when the lipid concentration was increased to65%. LCFA trapped the microorganisms and hindered the transferprocess of acetic acid, hydrogen and methanogens. This inhibitionprocess was reversible.

4. Conclusions

BMP tests showed that AD process proceeded well and methaneproduction increased with the increment of lipid concentrationwhen lipid concentration did not exceed 18%. MBW with lipid con-centration not higher than 18% could be AD directly. Bench-scaletests showed that when lipid concentration was below 60%, dailymethane production was enhanced with increasing lipid concen-tration. Lipid concentration has no influence on biogas contentand the concentration of methane was 64%. The pH decreasedand VFA concentration increased during the first 7 days. Fromthe 8 day, pH and VFA began to gradually revert to the normal le-vel. The ratio of VFA/Alkalinity was between 0.2 and 0.6. The sys-tem was operated stably. When lipid concentration exceeded65%, daily methane production decreased with increasing lipidconcentration. Daily methane production of lipid concentrationwith 65%, 70%, and 75% was decreased to 8%, 7%, and 7% compared


to the raw materials. Methane content decreased along withincreasing lipid concentration. The methane contents of lipid con-centration with 65%, 70%, and 75% were 55.3%, 55.0%, and 54.9%.When lipid concentration exceeded 65%, the pH dropped to 5.18in day 25. Lipid concentration of 60% of total VS was the highest li-pid concentration which would not inhibit the AD process. LCFAaccumulated on the surface of the metabolites and inhibited theAD process. The amount of lipid degradation bacteria increasedwith increasing lipid concentration.

Acknowledgments

The authors are grateful to the Ministry of Science and Technol-ogy of China (No. 2010DFA22770), the National Science and Tech-nology Support Program of China (No. 2010BAC66B04), and theKey Laboratory for Advanced Technology in Environmental Protec-tion of Jiangsu Province (No. AE201003) for providing financialassistances.

References

Alves, M. M.; Mota Vieira, J. A.; Álvares Pereira, R. M.; Pereira, M. A.; Mota, M., 2001.Effects of lipids and oleic acid on biomass development in anaerobic fixed-bedreactors. Part II: Oleic acid toxicity and biodegradability. Water Research. 35(1),264–270.

Angelidaki, I., Ahring, B.K., 1992. Effects of free long-chain fatty acids onthermophilic anaerobic digestion. Appl. Microbiol. Biotechnol. 37 (6), 808–812.

Angelidaki, I., Sanders, W., 2004. Assessment of the anaerobic biodegradability ofmacropollutants. Rev. Environ. Sci. Biotechnol. 3 (2), 117–129.

Buswell, A.M., Neave, S.L., 1930. Laboratory Studies of Sludge Digestion. JeffersonspriNting & Stationery Company.

Cammarota, M.C., Teixeira, G.A., Freire, D.M.G., 2001. Enzymatic pre-hydrolysis andanaerobic degradation of wastewaters with high fat contents. Biotechnol. Lett.23 (19), 1591–1595.

De Baere, L., Mattheeuws, B., 2008. State-of-the-art 2008—anaerobic digestion ofsolid waste. Waste Manage. World 9.

Liu, X., Gao, X.B., Wang, W., Zheng, L., Zhou, Y.J., Sun, Y.F., 2012. Pilot-scale anaerobicco-digestion of municipal biomass waste: focusing on biogas production andGHG reduction. Renewable Energy 44, 463–468.


http://refhub.elsevier.com/S0956-053X(13)00346-2/h0010
















Lossie, U., Pütz, P., 2008. Targeted Control of Biogas Plants with the Help of FOS/TAC.Practice Report Hach-Lange.

Luostarinen, S., Luste, S., Sillanpää, M., 2009. Increased biogas production atwastewater treatment plants through co-digestion of sewage sludge withgrease trap sludge from a meat processing plant. Bioresour. Technol. 100 (1),79–85.

National Bureau of Statistics of China, 2011. China Statistical Yearbook. <http://www.stats.gov.cn/tjsj/ndsj/2011/indexch.htm>.

Pereira, M.A., Sousa, D.Z., Mota, M., Alves, M.M., 2004. Mineralization of LCFAassociated with anaerobic sludge: kinetics, enhancement of methanogenicactivity, and effect of VFA. Biotechnol. Bioeng. 88 (4), 502–511.

Ren, L.H., Nie, Y.F., Liu, J.G., Jin, Y.Y., 2006. Impact of hydrothermal process on thedewaterability and degrease performance of restaurant garbage. Environ. Sci.Technol. 27 (9), 1906–1911.


Wan, C.X., Zhou, Q.C., Fu, G.M., Li, Y.B., 2011. Semi-continuous anaerobic co-digestion of thickened waste activated sludge and fat, oil and grease. WasteManage. (Oxford) 31, 1752–1758.

Wang, Z.J., Wang, W., 2005. Enhancement of sewage sludge anaerobic digestibilityby thermal hydroly-sis pretreatment. Environ. Sci. 26 (1), 68–71.

Wang, X.J., Yang, G.H., Li, F., Feng, Y.Z., Ren, G.X., Han, X.H., 2013. Evaluation of twostatistical methods for optimizing the feeding composition in anaerobic co-digestion:mixture design and central composite design. Bioresour. Technol. 131, 172–178.

Zeng, K., Wan, H.Y., Meng, F.C., Liu, S.J., 2007. Discussion on oil and grease removalin slaughterhouse and soybean processing wastewater treatment. Environ. Sci.Technol. 30 (4), 91–92 (in Chinese).

Zhang, C.S., Su, H.J., Tan T.W., 2013. Batch and semi-continuous anaerobic digestionof food waste in a dual solid–liquid system. Bioresour. Technol. <http://dx.doi.org/10.1016/j.biortech.2013.03.030>.






http://www.stats.gov.cn/tjsj/ndsj/2011/indexch.htm

http://www.stats.gov.cn/tjsj/ndsj/2011/indexch.htm


















http://dx.doi.org/10.1016/j.biortech.2013.03.030

http://dx.doi.org/10.1016/j.biortech.2013.03.030


Download - Effects of lipid concentration on anaerobic co-digestion of municipal biomass wastes

Top Related