Anaerobic co-digestion of animal manures with corn stover or apple pulp for enhanced biogas production

Renewable Energy 118 (2018) 335e342

Contents lists available at ScienceDirect

Renewable Energy journal homepage: www.elsevier.com/locate/renene

Anaerobic co-digestion of animal manures with corn stover or apple pulp for enhanced biogas production Kun Li a, b, Ronghou Liu a, b, *, Shaofeng Cui a, b, Qiong Yu a, b, Ruijie Ma a, b a

Biomass Energy Engineering Research Centre, School of Agriculture and Biology, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, PR China b Key Laboratory of Urban Agriculture (South), Ministry of Agriculture, 800 Dongchuan Road, Shanghai 200240, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 December 2016 Received in revised form 27 July 2017 Accepted 11 November 2017 Available online 11 November 2017

Corn stover (CS) or apple pulp (AP) were used to improve the anaerobic digestion performance of chicken manure (CM) or pig manure (PM), with the aim of adjusting carbon/nitrogen ratio and increasing the system stability compared to animal manures alone. This study was conducted in batch and semicontinuously fed digester at laboratory scale. The results of batch tests showed that the optimal mixture ratios for CM/CS, CM/AP, PM/CS and PM/AP were 4:1, 2:1, 4:1, 4:1, respectively. In the semicontinuous mode, inhibition to methane generation occurred when organic loading rate (OLR) of manures mono-digestion was higher than 2.4 g VS L
1d
1. However, the co-digestion of chicken manure with apple pulp at ratio 2 allowed operation at OLR of 4.8 g VS L
1d
1 and obtained the highest specific methane production (0.34 L g
1 VSadded), due to its enhanced buffer capacity and nutrient balance. Fluorescence in situ hybridization revealed that the microbial community in digester fed with CM/AP mix was dominated by Methanosarcina and the remaining microorganisms mainly belonged to Methanobacteriales, both of which reflected the tolerance of inhibitors in this system. However, in digesters with other mixture (CM/CS, PM/CS, PM/AP), the negative impact of high levels ammonia and volatile fatty acids on sensitive Methanogenic Archaea resulted in serious decrease in the dominant species and finally caused the failure of anaerobic digestion. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Anaerobic co-digestion Stability Microbial community Fluorescent in situ hybridization (FISH)

1. Introduction In China, the biogas industry has grown quickly in recent years and has undoubtedly become one of the biggest biomass energy industry. Small rural biogas digesters were dominant in China especially from 2002 to the end of 2010 [1]. As large-scale livestock farms become more and more common, medium and large-scale biogas plants are needed to dispose large amounts of intensive animal manures. On the other hand, incentive policies were introduced by the Chinese government with the aim of supplying a sustainable alternative energy source, which significantly promoted the development of medium and large-scale biogas plants [2]. Scaled production of biogas requires a stable feedstock supply, and the shortage of single material therefore became constraints in

* Corresponding author. Biomass Energy Engineering Research Centre, School of Agriculture and Biology, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, PR China. E-mail address: [email protected] (R. Liu). https://doi.org/10.1016/j.renene.2017.11.023 0960-1481/© 2017 Elsevier Ltd. All rights reserved.

mono-digestion. In addition, manures can hardly be digested as mono-feedstock because the ammonia inhibition and even failure in some cases may happen [3]. Li et al. [4] reported operating limits of organic loading rate when livestock manures were digested as single wastes. Drawbacks of manure mono-digestion can be solved by co-digestion with low nitrogen biomass so as to increase the carbon/nitrogen ratio (C/N) as an approach of overcoming process limitations originated from ammonia [5]. But unwise selection of co-substrate would also cause problems in digestion process [6]. Thus, this study focuses on two types of co-substrate, corn stover (CS) or apple pulp (AP), for anaerobic co-digestion by using chicken manure (CM) or pig manure (PM) as based feedstock. China is rich in resource of corn stover which annually generated an amount of approximately 216 million metric tons [7]. Due to its abundance and high carbohydrate content, corn stover has a great potential for scaled biogas generation [8]. However, owing to the inherent limitations of lignocellulosic substrates (high C/N ratio), their digestion were greatly hindered. Studies showed that the problem associated with nitrogen deficiency in crop residues was liable to produce poor buffering capacity as well as unbuffered

336

K. Li et al. / Renewable Energy 118 (2018) 335e342

Nomenclature AD BMP COD C/N Cy3 DAPI FA FISH FW HRT

Anaerobic digestion Biochemical methane potential Chemical oxygen demand Carbon/nitrogen ratio Fluorochrome of Cy3 DAPI staining solution Free ammonia Fluorescent in situ hybridization Fresh weight Hydraulic retention time

volatile fatty acid (VFA) accumulation in the digestion process [9,10]. Then, activity of methanogens was inhibited, and the conversion of organic carbon was decreased [11]. An alternative co-substrate concerns rotten apple which was easily collected from market. Apple waste has been targeted for biogas production because of its high biodegradable organic matter content and moisture. However, this carbohydrate-rich waste is rapidly acidified to VFA when used as mono-feedstock [12]. Alkali addition has been suggested to correct the lower pH of apple waste. Without any measures, the pH quickly decreased and tended to inhibit methanogenesis [13]. The cause of the process disruption in mono-digestion of the above two feedstocks is considered to be excessive production of VFA and inadequate alkalinity levels [14]. Anaerobic co-digestion refers to the digestion of mixed substrates [15]. To prevent the drop in pH, these wastes could be buffered by co-digesting with a buffering substrate. Manures are high nitrogen substrates, and have wide range of nutrients and high microbial activity [16]. Codigestion of manure with corn stover or apple waste could not only achieve high buffer capacity and nutrient-balance, but balance the C/N ratio of each feedstock, thereby decreasing the risk of ammonia inhibition which likely occurred when manures are used as sole-substrate. This is also a cost-effective method for nutritional regulation compared with addition of nitrogen-rich substrates like urea or ammonium bicarbonate to corn stover. Although many co-digestion trials have been conducted by using livestock and poultry manure mixed with other crop-residues, studies on manures co-digested with corn stover which is a promising renewable feedstock for methane production are limited. And the optimum ratio for biogas production varied widely. For example, Li and Zhang et al. [17] found that chicken manure and corn stover only mixed at ratio 1:3 could acquire a methane yield of 298.2 mL g
1 VS which is higher than that of single chicken manure (291.1 mL g
1 VS), while in another batch tests, synergism could be seen in co-digestion of chicken manure and corn stover with a ratio reported as 1:1 [18]. On the other hand, rapid acidification had been found in mono-digestion of cellulosepoor wastes like apple waste, consequently inhibited the activity of methanogenic organisms. The feasibility of co-digestion of animal manures and apple waste still needs to be deeply explored. This paper presents the results of batch and semi-continuous fed digestion trials carried out with mixture of animal manure (chicken manure, pig manure) and corn stover or apple waste at ratios of 4:1, 2:1, 1:1, 1:2 and 1:4, with the purpose of optimizing feed ratio between CM/PM and CS/AW as well as investigating the co-fermentation performance of continuous digester with feedstock of CM/CS (4:1), CM/AP (2:1), PM/CS (4:1) and PM/AP (4:1) under increasing organic loading rate. This study also focused on microbial community analyses for the co-digestion process by

IA OLR PA PVC SMP TA TAN TS VFA VMPR VS

Intermediate alkalinity Organic loading rate Partial alkalinity Polyvinyl chloride resin The specific methane productions Total alkalinity Total ammoniacal nitrogen Total solids Volatile fatty acid Volumetric methane production rate Volatile solid

fluorescent in situ hybridization. Insights gained from this work would enhance the comprehension of microorganisms involved in the anaerobic co-digestion of manures and other substrates as well as the factors that impact the performance and stability of anaerobic digester. 2. Methods and materials 2.1. Materials and digesters 2.1.1. Animal manures Chicken manure (CM) was obtained from a chicken farm in Heilongjiang province, China. Pig manure (PM) was collected from farm located in the Fengxian county of Shanghai city, China. Contaminants like stones and grass were picked away from manures and then they were homogenized, packed into 3-L plastic storage boxes, and frozen at
20  C. Manures were thawed before using, and stored at 4  C for no more than two weeks. Table 1 shows the properties and characteristics of inoculums, manures, corn stover and apple waste. 2.1.2. Corn stover and apple pulp The CS was obtained from experimental field of Agriculture and Biology School in Shanghai Jiao Tong University, Shanghai, China. The collected corn stover was ground into particles less than 1 mm. Rotten apples were collected from a market in Minhang District of Shanghai, China. Apples were firstly crushed with a blender, then homogenized and pass through 2.0 mm screen. Both CS and AP were stored in the same way as for manure. 2.1.3. Inoculum The inoculum for all digesters was obtained from a commercial biogas plant treating livestock manure under mesophilic conditions (continuous stirred tank reactor, operated by Sennong Company, Pudong District of Shanghai, China). Before use, the digestate was mixed thoroughly and sieved through 0.85 mm screen to remove large particles. 2.1.4. Digesters The batch digestion tests were carried out in 250 mL conical flasks each connected to a polyvinyl chloride resin (PVC) tube which was sealed one end using for biogas collection and volume determination. The temperature of flasks was kept at 37  C by a water-bath. The semi-continuous digester had a 2.5-L capacity with a 2-L working volume and was constructed of PVC pipe with airtight top and bottom. The top was sealed by water and fitted a feed port as well as a gas outlet. Drainage port was incorporated on the bottom. A stirrer was inserted through the top plate to allow mixing

K. Li et al. / Renewable Energy 118 (2018) 335e342

337

Table 1 Properties and characteristics of inoculums, manures, corn stover and apple waste. Characteristics

Sewage sludge

CM

PM

CS

AP

TS (% FW) VS (% FW) VS (% TS) pH TAN (mg/kg FW) Alkalinity (mg CaCO3/kg FW) VFAs (mg CH3COOH/kg FW) C (%TS) N (%TS) H (%TS) S (%TS) O (% TS) C/N

15.39 ± 0.05 10.38 ± 0.03 67.4 ± 0.17 7.44 ± 0.02 2698 ± 33.9 11501 ± 319 3356.4 ± 43.9 23.79 ± 0.16 2.99 ± 0.01 3.46 ± 0.18 3.39 ± 0.03 NA 7.9

25.12 ± 0.38 16.57 ± 0.03 65.99 ± 0.89 8.39 ± 0.04 7385 ± 43.8 84133 ± 589 19580.6 ± 81.4 30.29 ± 0.34 3.49 ± 0.08 5.31 ± 0.09 0.42 ± 0.02 32.03 ± 0.12 8.69

47.45 ± 0.63 36.49 ± 0.31 77.27 ± 0.38 7.29 ± 0.03 12510 ± 37.3 44398 ± 968 95502 ± 254 39.14 ± 0.05 3.92 ± 0.03 5.27 ± 0.27 0.79 ± 0.01 32.47 ± 0.39 10.00

24.1 ± 0.24 23.5 ± 0.18 97.6 ± 0.26 NA NA NA NA 30.32 ± 0.03 0.94 ± 0.01 6.16 ± 0.01 0.05 ± 0.00 NA 32.10

12.26 ± 1.33 10.91 ± 1.33 88.79 ± 1.55 4.33 ± 0.01 154 ± 11.3 3.89 ± 2.32 1128.1 ± 97.1 54.06 ± 0.05 0.66 ± 0.01 5.61 ± 0.03 0.02 ± 0.01 NA 82.35

Note: Values are presented as mean and data afterwards are standard deviations (n ¼ 3). FW: fresh weight; TS: total solids; VS: volatile solid; TAN: total ammoniacal nitrogen. NA: none analysis.

at speed of 20 rpm. Temperature of 36 ± 1  C in digesters was maintained by recirculating water from thermostatical water bath. The biogas daily collected by gas-impermeable bags was corrected to standard volume under temperature and pressure of 0  C,101.325 kPa. Biogas composition was determined once or twice per fortnight. 2.2. Batch screening tests The biochemical methane potential (BMP) assays were carried out in two series: (1) CM and PM were independently digested as mono-feedstock with increasing VS contents (8, 16, 32, 64 g VS L
1); (2) co-digestion of chicken manure with AP or CS independently at different ratios (4:1, 2:1, 1:1, 1:2, 1:4) on a VS basis at 32 g VS L
1 (this concentration was determined from results of batch tests 1); co-digestion of pig manure with AP or CS independently at the same ratios and loading. The working volume of conical flask was 200 mL, and 64 g inoculum (fresh weight) as well as corresponding substrate materials was added into each digester. Then digesters were flushed with 99.9% N2 for 3 min before sealing. Reactors only with inoculum and deionized water were also operated as blank to correct the biogas yield of substrates. Batch tests were carried out for more than 70 days at 37  C. All the batch tests were run in duplicate. The volume of biogas produced by each reactor was determined daily and the composition was measured every 2e3 days. 2.3. Semi-continuous digestion trials Four semi-continuous digesters were fed with mixture of animal manure and CS or AP at the optimal ratio acquired from batch tests for biogas production. In details, reactor 1 (denoted as R1) fed with CM and CS at ratio 4:1; reactor 2 (denoted as R2) fed with CM and AP at ratio 2:1; reactor 3 (denoted as R3) fed with PM and CS at ratio 4:1; reactor 4 (denoted as R4) fed with PM and AP at ratio 4:1. Reactor 5 and 6 were fed with solely CM and PM respectively as controls (denoted as R5 and R6). The two controls were operated to test independently the impact of increasing load on the biogas production from animal manures, irrespective of CS and AP addition. Once each digester was filled with 2 L inoculum (VS of 10.38%), the feeding strategy with different ratio of animal manures to CS or AP as well as sole manure was carried out at increasing organic loading rate (OLR) (2.4e9.6 g VS L
1d
1). Before use, the inoculum was mixed thoroughly and sieved through a 0.85 mm pore size screen to remove any large particles. Feedstocks were diluted to desired TS content and the substrate concentrations varied with

different mixture. The designed OLR of 2.4, 4.8, 7.2, and 9.6 g VS L
1d
1 at four phases was maintained by increasing the amount of feedstock and shortening hydraulic retention time (HRT). This paper mainly aims to obtain the most suitable materials to overcome the toxicity of high ammonia concentrations and process upset at high loading rate and short HRT. Table 2 shows operational regimes for semi-continuous co-digestion of animal manure with corn stover or apple waste. 2.4. Analytical methods TS, VS, pH, alkalinity and ammonium were determined using the Standard Methods [19]. pH was measured using a pH meter with sensitivity of 0.01 (ZDJ-4A, Shanghai INESA & Scientific Instrument Co., LTD., Shanghai, China). Alkalinity was determined by titration with 0.125 mol L
1 H2SO4 to endpoints of pH value 5.7 and 4.3 using an automatic digital titrator (ZDJ-4A, Shanghai INESA & Scientific Instrument Co., LTD., Shanghai, China) to allow calculation of total alkalinity (TA), partial alkalinity (PA) and intermediate alkalinity (IA) [20]. The minimum drip of titrator was set up to 0.02 mL. TAN was determined after the addition of NaOH using a KDN-103F ammonia analyzer (Shanghai QianJian instruments CO., China) and distillate is collected in boric acid indicator which then titrated by 0.125 mol L
1 H2SO4 [21]. Analyses of elements (C, N, H, O) were conducted using a Vario EL elemental analyzer (Model Vario EL III, Elementar, Germany). VFAs of C2-C5 were quantified in a gas chromatography (GC-7890A, Agilent Technology, USA) equipped with a flame ionization detector, and a 30 m  320 mm  0.25 mm capillary column. Nitrogen was the carrier gas. The temperatures of the initial oven, injection port, and detector were 60, 200, and 220  C, respectively. Biogas samples were analyzed using a GC7890A equipped with a thermal conductivity detector. A 2 m  3 mm column packed with TDX-01 (JK, China) was used to separate different gas composition in biogas. The carrier gas was helium. Fluorescent in situ hybridization: samples of six digesters were periodically collected for FISH analyses. The samples were

Table 2 Operational regimes for semi-continuous co-digestion trials. Time (days)

OLR (g VS/(L
1d
1))

HRT (d)

0e50 51e75 76e95 96e114

2.4 4.8 7.2 9.6

50 25 16.7 12.5

338

K. Li et al. / Renewable Energy 118 (2018) 335e342

Fig. 1. Cumulative methane production from anaerobic digestion of CM/CS (A), CM/AP (B), PM/CS (C) and PM/AP (D) with different mixing ratios.

pretreated according to the method described by Banks et al. [22], and then incubated overnight at 4  C with 4% paraformaldehyde fixative solution. Hybridization was carried out at 46  C for 2 h with hybridization solution containing 5 ng mL
1 of specific fluorescent probe. The probes for FISH were designed as previously described [22]. An Olympus IX71 epifluorescence microscope (IX71, Olympus, Tokyo, Japan) equipped with a 100 W mercury lamp and appropriate filter sets for DAPI and Cy3 was used to capture FISH images. 3. Results and discussion 3.1. Anaerobic digestion of batch tests CM and PM were used as mono-feedstock with increasing VS contents to investigate the performance of digesting manures alone. The cumulative methane yields of four substrate concentrations (8, 16, 32, 64 g VS L
1) were 126.6, 91.5, 76.0 and 66.2 mL g
1 VS for CM, and 124.4, 109.9, 88.9 and 79.6 mL g
1 VS for PM, indicating that inhibition from soluble compounds like TAN produced from degradation was developed at higher substrate concentration. Therefore, low nitrogen co-substrate was needed to improve C/N ratio, allowing enhanced process stability. Anaerobic co-digestion was carried out at relatively high organic load of 32 g VS L
1 determined from batch test of set-up one. Fig. 1 shows the cumulative methane production from anaerobic digestion (AD) of CM/CS, CM/AP, PM/CS and PM/AP with different mixing ratios. Table 3 shows the anaerobic co-digestion results of batch tests. It can be seen from Table 3 that both addition of CS and AP could improve the C/N ratio of animal manure, but methane production as well as VS removal rate, decreased as the ratio of cosubstrates in mixture improved. Similar results were observed in thermophilic anaerobic co-digestion of swine manure and maize stalk in which higher swine ratio showed higher total biogas yield [23]. One of the reason may be that energy density of organic matter decreases in the order of lipids > proteins > carbohydrates [24]. CS and AP are carbohydrate-rich substrates and have a lower gas yield than manures [25]. This phenomenon can also be accounted by the different chemical oxygen demand (COD) content in substrates. The COD of manure is generally in the range of 1.455e11.52 g g
1VS [26e28], and that is 1.03 g g
1VS [26] for wheat straw and 1.08e1.95 g g
1VS for fruit and vegetable waste [29e32]. Co-digestion of PM and AP at ratio 1:4 even stopped

Table 3 Anaerobic co-digestion results of batch tests. Substrates C/N Methane yield mL/g VS

Methane content (%)

TAN mg L
1

Final pH

CM:CS 4:1 2:1 1:1 1:2 1:4 CM:AP 4:1 2:1 1:1 1:2 1:4 PM:CS 4:1 2:1 1:1 1:2 1:4 PM:AP 4:1 2:1 1:1 1:2 1:4

77.8 73.9 79.5 77.3 73.5 67.0 74.5 65.2 53.9 67.5 76.8 79.2 74.2 63.8 75.4 75.3 76.0 64.9 70.1 21.8

2832.6 ± 37.6 2480.7 ± 40.6 1732.7 ± 20.3 1571.7 ± 29.6 1626.1 ± 38.3 3155.2 ± 45.9 2828.4 ± 59.1 2205.3 ± 37.8 1631.2 ± 68.0 1486.7 ± 76.5 2014.1 ± 110.1 917.6 ± 17.5 1708.9 ± 26.8 909.5 ± 93.5 1079.5 ± 12.7 807.5 ± 42.5 1818.6 ± 35.5 1626.9 ± 19.1 1291.6 ± 15.2 1389.3 ± 20.7

8.02 8.04 7.91 7.92 7.55 8.01 8.55 7.94 7.84 7.97 7.80 7.63 7.96 7.20 7.47 7.80 7.85 7.82 7.80 6.64

25.8 35.2 47.0 58.8 68.2 15.8 18.5 21.9 25.3 28.0 24.5 34.1 46.2 58.2 67.9 14.4 17.4 21.0 24.7 27.7

79.3 71.5 67.7 67.8 56.5 80.9 82.7 67.7 70.8 56.6 83.2 81.8 65.4 60.1 66.5 102.1 86.8 79.0 54.6 2.27

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.04 0.02 0.09 0.01 0.01 0.02 0 0.03 0.01 0.03 0.02 0.03 0.05 0.01 0.03 0.02 0.02 0.02 0.01 0.02

Note: Values are presented as mean and data afterwards are standard deviations (n ¼ 3). CM is chicken manure, PM is pig manure, CS is corn stover, AP is apple waste.

producing methane at the beginning of the test. The maximum methane production obtained in CM/CS mix was 79.3 mL g
1 VS with a ratio of 4:1, and in CM/AP mix was 82.7 mL g
1 VS with a ratio of 2:1. The optimum ratio for PM/CS mix was 4:1, although the methane yield obtained was lower than mono-digestion of PM. The mixture of PM and AP with optimum ratio 4:1 produced a methane yield of 102.1 mL g
1 VS which is 14.8% greater than that produced by mono-digestion of PM. 3.2. Anaerobic digestion of semi-continuous trial 3.2.1. Methane yield From the results above, the optimal mixture ratios for CM/CS (4:1), CM/AP (2:1), PM/CS (4:1) and PM/AP (4:1) were determined. And the relative semi-continuous anaerobic co-digestion of these mixtures was carried out. Fig. 2 shows the specific methane production (A) and volumetric methane production rate (B) in control

K. Li et al. / Renewable Energy 118 (2018) 335e342

339

Fig. 2. Specific methane production (A) and volumetric methane production rate (B) in control and co-digestion reactors.

and co-digestion reactors. The specific methane productions (SMP) for control R5 (CM only) and R6 (PM only) at OLR of 2.4 g VS L
1d
1 were 0.267 L g
1 VSadded and 0.299 L g
1 VSadded, respectively. Values increased to 0.365 L g
1 VSadded (R2) and 0.322 L g
1 VSadded (R4) after the addition of AP. No obvious improvement in SMP was found when manures were co-digested with CS. The SMP of R1 (CM þ CS) and R3 (PM þ CS) were 0.251 L g
1 VSadded and 0.301 L g
1 VSadded, respectively. At day 51, OLR was increased to 4.8 g VS L
1d
1. Although this caused a decrease in all the SMP, this value in R2 was still 27% higher than that of control R5 in initial stage and 36% higher than R5 at present, which reflects the enhancement of methane production when chicken manure was co-digested with AP at an optimum ratio 2:1. From Fig. 2B, it can be seen that the volumetric methane production rates (VMPR) of control R5 and R6 increased by 56.9% and 47.4%, respectively, after the OLR rose to 4.8 g VS L
1d
1 to reach 1.139 L L
1d
1 and 1.211 L L
1d
1, which were respectively lower than that of CM/AP mix (1.578 L L
1d
1) and PM/ AP mix (1.287 L L
1d
1). Meanwhile, CM/CS mix (R1) and PM/CS mix (R3) got lower VMPR with values of 1.178 L L
1d
1 and 1.039 L L
1d
1, separately. The OLR was further increased to 7.2 g VS L
1d
1 on day 76, and digesters were overloaded at different level. Sharp drops in methane yield were found in controls (R5 and R6) and CS feeding digesters (R1 and R3) with a final SMP range of 0.05e0.11 L g
1 VSadded, while AP feeding digesters R2 (CM þ AP) and R4 (PM þ AP) obtained higher SMP of 0.21 L g
1 VSadded and 0.18 L g
1 VSadded, indicating the relative stability at high loading resulted from the addition of AP to manures. Rapid increase of TAN concentration can be observed in R1, R3, R5 and R6 in response of the improvement of OLR to 7.2 g VS L
1d
1 during day 76e95. The high TAN concentration caused an increase of pH as well as the inhibition of methane production by effecting the free ammonia concentration [33], and then pH decreased due to VFA accumulation accompanied with ammonia inhibition. In R2 and R4, the addition of AP caused

VFA accumulation and a drop in pH which provide a higher buffer capacity. Therefore, both pH and TAN were stable in the two digesters. However, the methane yield of AP feeding digesters was also reduced a lot when OLR increased to 9.6 g VS L
1d
1 on day 96, and the biogas production almost ceased in other digesters. This result indicated that in this semi-continuous trial the maximum OLR for CM/AP is 4.8 g VS L
1d
1. Compared with mono-digestion, the AD of CM/AP mix got higher methane yield especially at high loading. 3.2.2. Stability of semi-continuous digestion Fig. 3 shows the pH, alkalinity, ammonia and total VFA for all digesters. The pH value has a direct impact on acidogenic and methanogenic microorganisms and it is usually used as an indicator to monitor the stability of digestion process [34]. From Fig. 3A, it can be seen that the initial pH value was 7.44, and the value in six digesters increased to around 7.97 after 50 days digestion at the OLR of 2.4 g VS L
1d
1. When OLR improved to 4.8 g VS L
1d
1, the pH value in each digester maintained within range of 8.01e8.59, and then R1, R2, R4 and R6 stabilized around pH 8.37 with OLR of 7.2 g VS L
1d
1. But an obvious decline of pH value was observed in R3 and R5. During day 76e93, the pH in R3 and R5 respectively decreased from 8.54 to 7.77 and from 8.55 to 8.05. The drop of pH is mainly due to the VFA accumulation, as the TAN is increased to the inhibitory concentration and methanogen activity is inhibited. It can be seen that R3 and R6 had a great drop at loading 9.6 g VS L
1d
1 with a final pH value of 6.78 and 6.73, respectively. As a consequence of hydrolysis and acidification, the pH values of R1 (pH 7.49), R4 (pH 7.87) and R5 (pH 7.69) were also dropped, while pH of R2 was still maintained above 8 mainly due to its high buffer capacity and stability of anaerobic system. Alkalinity can be used to quantify the buffering capacity and evaluate system stability. Unlike pH which only reflects what has already happened, changes of alkalinity indicate what is happening and corrective measures can be applied to stabilize the process. The

340

K. Li et al. / Renewable Energy 118 (2018) 335e342

Fig. 3. pH (A), IA:PA (B), ammonia (C) and total VFAs (D) in control and co-digestion reactors.

main species consisting the proton accepting capacity in digestion system are the non-protonated forms of VFAs and the carbonate/ bicarbonate system. IA is attributed to the buffering capacity of VFA, and PA is a measurement of bicarbonate buffering. In robust anaerobic digesters, the ratio of IA/PA was typically below 0.3e0.5 [35,36]. In this study, the IA:PA ratios of all the digesters were below 0.5 under operation at 4.8 g VS L
1d
1 (Fig. 3B), whereas ratios of digesters except R2 and R4 were rapidly increased upon increasing the OLR to 7.2 g VS L
1d
1. At the highest loading of 9.6 g VS L
1d
1, only IA:PA ratio of R2 remained around 0.5, while ratios of other digesters have climbed up to range 1.34e2.5, indicating both the accumulation of VFAs and the instability of digestion system. Ammoniacal nitrogen is generated by the biological degradation of nitrogenous matter. Optimal ammonia concentration guarantees adequate buffer capacity and allows the system to recover after a VFA shock. But inhibition would occur if it is too high. Generally, the reported inhibitory concentration of TAN was 3e6 g L
1 [37,38]. The ammonia nitrogen concentration in each digester showed an increasing trend over the trial period (Fig. 3C). TAN concentration was around 1651 mg L
1 at loading 2.4 g VS L
1d
1, and higher TAN range of 1758 mg L
1 to 4475 mg L
1 was reached in response to the increase in OLR to 4.8 g VS L
1d
1. AP digesters R2 and R4 maintained lower TAN concentrations of 4602 mg L
1 and 4632 mg L
1 at OLR of 7.2 g VS L
1d
1, while seriously inhibition of methane production occurred in other digesters with TAN concentrations increased to the range of 5125e7774 mg L
1. From day 76e95, free ammonia (FA) concentration of R1, R5, R6 varied from 294.3 mg L
1 to 968.1 mg L
1 (not shown in figure), and lower FA was found in R3 as a result of the decreased pH [39]. When OLR further improved to 9.6 g VS L
1d
1, the ammonia nitrogen concentrations in R1, R3, R4, R5, R6 were beyond the higher limit of inhibitory threshold, meaning that serious inhibition occurred when manures digested alone or co-digested with CS. Whereas the TAN concentration of R2 (5811 mg L
1) maintained being the lowest. Previous studies showed that FA of 215 mg L
1 reduced the methane yield from manure fermentation by 50% [40]. However, a higher tolerance of 1000 mg L
1 has also been reported in the mesophilic methane fermentation of chicken manure [41]. In this study, the FA concentrations of primarily inhibition on methanogenesis for R1, R3, R5 and R6 were 394.5 mg L
1, 136.1 mg L
1, 441 mg L
1 and 490.4 mg L
1, respectively. Biogas production almost ceased when FA reached 500 mg L
1. Fig. 3D shows the VFA concentration profile during the semicontinuous process. The VFA accumulation in digesters started gradually from around day 83, and the VFA concentration was in

the range from 379.4 mg L
1 (R2) to 1528.9 mg L
1 (R1). Thereafter, the total VFAs concentration increased sharply in R1, R3, R4, R5, R6, reaching 10338.1 mg L
1, 8556.6 mg L
1, 5145.3 mg L
1, 14836.4 mg L
1, 10891.3 mg L
1, respectively, on day 104. Due to VFA accumulation, decreasing pH value as well as methane production was observed. But the VFA concentration of R1 was higher than R3 from day 97 to day 114, while higher pH value in R1 (8.36e7.49) was measured. Similarly, R5 has higher VFA concentration and pH value (8.52e7.69) than R6 (8.40e6.73). It seems that the degradation of chicken manure released more ammonia than pig manure to enhance the buffer capacity and neutralize the VFAs. From another perspective, the pH of R1 and R5 was maintained in a suitable range for anaerobic digestion, while the biogas production was reduced a lot. This fact implies that methanogenesis step in R1 and R5 was inhibited by some factors other than pH. Ammonia inhibition was a key reason to explain this observation. This can be supported by the fact that adding AP (R2 and R4), readily fermented to organic acids, was beneficial to the AD performance because AP can reduce ammonia nitrogen concentration by neutralization and form buffer system in co-digestion. Compared with the above five digesters, the VFA concentration in R2 was consistently lowest, and the final value of VFA concentration on day 114 was 4694.7 mg L
1. The pH value of R2 was kept around 8.26 after increasing from initial 7.44 to 8.44 on day 83, much higher than others. This result shows that the addition of AP to chicken manure could stabilize anaerobic system. These results indicated that co-digestion of chicken manure and AP improved the buffer capacity of anaerobic system and process stabilities. The co-digestion of chicken manure and apple waste at ratio 2:1 was feasible with no occurrence of VFA accumulation and ammonia inhibition. The low C/N ratio of chicken manure (8.69) and pig manure (10.0) implied that they contained high quantity of nitrogen, mostly in organic forms of protein, nucleic acid and nitrogenous lipids etc. [39]. Ammonia inhibition was likely occurred when manures digested alone. AP is rich in carbohydrate which is converted into acetic acid after hydrolysis and acidogenesis [42,43]. The addition of AP could increase the organic acids in co-digestion system. High concentration of VFAs was neutralized by ammonia generated by protein breakdown, which forms an ammonium salt and enhances the buffer capacity of the digestion system [44]. Attributed to the balanced interaction among ammonia, VFAs and alkalinity, excessive production of dissolved NH3 which was toxic to methanogens was avoided in R2. R4 was in the same case of R2. The C/N ratios of mixture CM:CS (R1) and PM:CS (R3) were 25.8

K. Li et al. / Renewable Energy 118 (2018) 335e342

341

Table 4 The cellular concentration of methanogens in mesophilic anaerobic digestion reactors. Sample Eubacteria (Cell mL
1) (  109) R1 Day 0 Day 79 Day 94 Day 113 R2 Day 0 Day 79 Day 94 Day 113 R3 Day 0 Day 79 Day 94 Day 113 R4 Day 0 Day 79 Day 94 Day 113 R5 Day 0 Day 79 Day 94 Day 113 R6 Day 0 Day 79 Day 94 Day 113

Archaea (Cell mL
1) (  109)

Methanosaeta (Cell mL
1) (  109)

Methanosarcina (Cell mL
1) (  109)

Methanobacteriales (Cell mL
1) (  109)

0.05 0.15 0.02 0.06

± ± ± ±

0.01 0.06 0.05 0.10

0.39 0.32 0.04 0.03

± ± ± ±

0.03 0.02 0.01 0.01

0.09 ± 0.01 0.05 ± 0.02 0.00 ± 0.00 0

0.28 0.21 0.02 0.03

± ± ± ±

0.02 0.03 0.01 0.02

0.01 ± 0.01 0.01 ± 0.00 0.00 ± 0.00 0

0.05 0.15 0.03 0.03

± ± ± ±

0.01 0.06 0.05 0.10

0.39 1.07 0.48 0.87

± ± ± ±

0.03 0.04 0.01 0.10

0.09 0.05 0.00 0.01

± ± ± ±

0.01 0.01 0.00 0.00

0.28 0.63 0.41 0.67

± ± ± ±

0.02 0.03 0.02 0.05

0.01 0.09 0.04 0.17

± ± ± ±

0.01 0.02 0.02 0.03

0.05 0.11 0.07 0.08

± ± ± ±

0.01 0.06 0.05 0.10

0.39 0.45 0.05 0.02

± ± ± ±

0.03 0.05 0.01 0.01

0.09 ± 0.01 0.02 ± 0.01 0.00 ± 0.00 0

0.28 0.37 0.04 0.01

± ± ± ±

0.02 0.03 0.01 0.01

0.01 0.05 0.01 0.00

± ± ± ±

0.01 0.01 0.01 0.02

0.05 0.43 0.06 0.03

± ± ± ±

0.01 0.06 0.05 0.10

0.39 0.48 0.16 0.09

± ± ± ±

0.03 0.06 0.02 0.02

0.09 ± 0.01 0.01 ± 0.01 0.00 ± 0.00 0

0.28 0.42 0.09 0.08

± ± ± ±

0.02 0.06 0.02 0.03

0.01 0.02 0.04 0.01

± ± ± ±

0.01 0.01 0.01 0.00

0.05 0.21 0.07 0.02

± ± ± ±

0.01 0.06 0.05 0.10

0.39 0.18 0.03 0.01

± ± ± ±

0.03 0.02 0.01 0.00

0.09 ± 0.01 0.02 ± 0.01 0.00 ± 0.00 0

0.28 0.13 0.01 0.01

± ± ± ±

0.02 0.02 0.01 0.00

0.01 0.02 0.00 0.00

± ± ± ±

0.01 0.01 0.00 0.00

0.05 0.13 0.05 0.02

± ± ± ±

0.01 0.06 0.05 0.10

0.39 0.25 0.04 0.01

± ± ± ±

0.03 0.02 0.01 0.01

0.09 ± 0.01 0.01 ± 0.01 0.00 ± 0.00 0

0.28 0.21 0.00 0.00

± ± ± ±

0.02 0.01 0.00 0.00

0.01 0.01 0.03 0.00

± ± ± ±

0.01 0.00 0.01 0.00

and 24.5 respectively. As far as the C/N ratio was considered, both mixtures were suitable for biogas fermentation. However, compared with mono-digestion of chicken manure and pig manure, R1 and R3 did not show better performance like R2. The reason for this maybe that CS typically contains high content of lignocellulose which was difficult for microorganism to utilize. As a result, higher buffer capacity and system stability could hardly be obtained. The co-digestion of CS with animal manures was, therefore, easily failed because of the production of excessive ammonia in the fermentation process. For this reason, pretreatment of lignocellulosic materials before anaerobic digestion has been suggested.

3.2.3. Microbiological analysis for co-digestion of different feedstocks using FISH FISH was used to analyze the specific microbial composition in six reactors, and Table 4 shows the results of the cellular concentration of methanogens in mesophilic anaerobic digestion reactors. According to the results of FISH, the cellular concentration of methanogens in the reactors was negatively impacted by the increasing of the concentration of VFA and TAN. Methanosarcina was the predominant methanogen in inoculum by 72.9% of total archaea, followed by Methanosaeta (23.7%) and Methanobacteriales (1.7%). By day 79, the number of Methanosarcina and Methanobacteriales in R3 and R4, as well as Methanobacteriales of two controls (R5 and R6) increased. The predominant methanogens of R1-R6 were all Methanosarcina, with a proportion of 21%, 63%, 37%, 42%, 13% and 21%, respectively, of total archaea. But when the OLR further increased from 7.2 g VS L
1d
1 to 9.6 g VS L
1d
1, Methanosaeta, Methanosarcina and Methanobacteriales exhibited a decreasing trend in all reactors except R2. In R2, the continuous increase of Methanosarcina and Methanobacteriales throughout the experiment was detected. It seems that the co-

digestion of chicken manure and apple pulp maintained a suitable condition for the growth of Methanosarcina and Methanobacteriales. Considering the tolerance of the two methanogenic groups to inhibitor, R2 became more resistant towards process upset. In addition, the dominant species of R1-R5 over the trial period were consistently Methanosarcina. The reason for this may have been due in part to its tolerant of high inhibitors concentrations, and in part to originally enrichment of Methanosarcina in the inoculum which greatly facilitated its dominance [45,46]. In R6, the concentrations of VFA and TAN were significantly higher. As a result, Methanobacteriales dominated R6 instead of Methanosarcina by the end of the trial. 4. Conclusions The 4:1 ratio for CM/CS, PM/CS, PM/AP and 2:1 ratio for CM/AP showed better performance in batch test compared with other ratios. In the semi-continuous mode, mono-digestion of CM and PM at high OLR was unstable and inhibited by the accumulation of TAN and VFAs. Co-digestion of CM with AP at ratio 2 (R2) reduced TAN concentrations, and allowed improvement in OLR to 4.8 g VS L
1d
1 with an increased SMP and VMPR of 0.34 L g
1 VSadded and 1.631 L L
1d
1, respectively. Methanosarcina species dominated the archaeal populations of R2 with a gradually increasing of Methanobacteriales, which makes R2 more resistant towards process upset. The accumulated inhibitor in other digesters has a negative impact on the microbial community and the amount of methanogens declined rapidly. Acknowledgments Financial support from National Natural Science Foundation of

342

China through acknowledged.

K. Li et al. / Renewable Energy 118 (2018) 335e342

contract

(Grant

no.

51376121)

is

greatly

References [1] Z. Song, C. Zhang, G. Yang, Y. Feng, G. Ren, X. Han, Comparison of biogas development from households and medium and large-scale biogas plants in rural China, Renew. Sustain. Energy Rev. 33 (2) (2014) 204e213. [2] L. Chen, L. Zhao, C. Ren, F. Wang, The progress and prospects of rural biogas production in China, Energy Policy 51 (4) (2012) 58e63. [3] Q. Niu, Q. Wei, Q. Hong, Y.Y. Li, Microbial community shifts and biogas conversion computation during steady, inhibited and recovered stages of thermophilic methane fermentation on chicken manure with a wide variation of ammonia, Bioresour. Technol. 146C (10) (2013) 223e233. [4] K. Li, R.H. Liu, C. Sun, Comparison of anaerobic digestion characteristics and kinetics of four livestock manures with different substrate concentrations, Bioresour. Technol. 198 (2015) 133e140. [5] K. Li, R. Liu, C. Sun, A review of methane production from agricultural residues in China, Renew. Sustain. Energy Rev. 54 (2016) 857e865. [6] H.B. Nielsen, I. Angelidaki, Codigestion of manure and organic waste at centralized biogas plants: process imbalances and limitations, Bibliogr 107 (7) (2008) 1521e1528. [7] National Bureau of Statistics of China, China statistical Yearbook, China Statistics Press, Beijing. China, 2010. [8] Z. You, T. Wei, J.J. Cheng, Improving anaerobic codigestion of corn stover using sodium hydroxide pretreatment, Energy & Fuels 28 (1) (2014). [9] C.J. Banks, P.N. Humphreys, The anaerobic treatment of a ligno-cellulosic substrate offering little natural pH buffering capacity, Water Sci. Technol. 38 (4) (1998) 29e35. [10] X. Wang, G. Yang, Y. Feng, G. Ren, X. Han, Optimizing feeding composition and carbonenitrogen ratios for improved methane yield during anaerobic codigestion of dairy, chicken manure and wheat straw, Bioresour. Technol. 120 (2012) 78e83. [11] Z. Tong, C. Mao, N. Zhai, X. Wang, G. Yang, Influence of initial pH on thermophilic anaerobic co-digestion of swine manure and maize stalk, Waste Manag. 35 (7) (2015) 119e126. [12] H. Bouallagui, Y. Touhami, R.B. Cheikh, M. Hamdi, Bioreactor performance in anaerobic digestion of fruit and vegetable wastes, Cheminform 40 (3) (2005) 989e995. [13] P. Namsree, W. Suvajittanont, C. Puttanlek, D. Uttapap, V. Rungsardthong, Anaerobic digestion of pineapple pulp and peel in a plug-flow reactor, J. Environ. Manag. 110 (18) (2012) 40e47. [14] T. Thamsiriroj, A.S. Nizami, J.D. Murphy, Why does mono-digestion of grass silage fail in long term operation? Appl. Energy 95 (2012) 64e76. [15] Z. Song, Z. Chao, Anaerobic codigestion of pretreated wheat straw with cattle manure and analysis of the microbial community, Bioresour. Technol. 186 (2015) 128e135. [16] X. Wu, W. Yao, J. Zhu, C. Miller, Biogas and CH 4 productivity by co-digesting swine manure with three crop residues as an external carbon source, Bioresour. Technol. 101 (11) (2010) 4042e4047. [17] Y. Li, R. Zhang, C. Chang, G. Liu, Y. He, X. Liu, Biogas production from codigestion of corn stover and chicken manure under anaerobic wet, hemisolid, and solid state conditions, Bioresour. Technol. 149 (4) (2013) 406e412. [18] Y. Li, R. Zhang, X. Liu, C. Chang, X. Xiao, F. Lu, Y. He, G. Liu, Evaluating methane production from anaerobic mono- and Co-digestion of kitchen waste, corn stover, and chicken manure, Energy & Fuels 27 (4) (2013) 189e194. [19] APHA, Standard Methods for the Examination of Water and Wastewater, Public Health Association, Washington DC, 2005. [20] L.E. Ripley, J.C. Converse, Improved alkalimetric monitoring for anaerobic digestion of high-strength waste, J. - Water Pollut. Control Fed. 58 (5) (1986) 406e411. [21] A. Serna-Maza, S. Heaven, C.J. Banks, Ammonia removal in food waste anaerobic digestion using a side-stream stripping process, Bioresour. Technol. 152 (1) (2014) 307e315. [22] C.J. Banks, Z. Yue, J. Ying, S. Heaven, Trace element requirements for stable food waste digestion at elevated ammonia concentrations, Bioresour. Technol. 104 (1) (2011) 127e135.

[23] Z. Tong, C. Mao, N. Zhai, X. Wang, G. Yang, Influence of initial pH on thermophilic anaerobic co-digestion of swine manure and maize stalk, Waste Manag. 35 (7) (2014) 119e126. [24] K.H. Hansen, B.K. Ahring, L. Raskin, Quantification of syntrophic Fatty Acid-bOxidizing bacteria in a mesophilic biogas reactor by oligonucleotide probe hybridization, Appl. Environ. Microbiol. 65 (11) (1999) 4767e4774. [25] V.N. Gunaseelan, Anaerobic digestion of biomass for methane production: a review, Biomass & Bioenergy 13 (1) (1997) 83e114. [26] G. Wang, H.N. Gavala, I.V. Skiadas, B.K. Ahring, Wet explosion of wheat straw and codigestion with swine manure: effect on the methane productivity, Waste Manag. 29 (11) (2009) 2830e2835. [27] R. Li, S. Chen, X. Li, Anaerobic Co-digestion of kitchen waste and cattle manure for methane production, Energy Sour. Part A Recovery Util. Environ. Eff. 31 (20) (2009) 1848e1856. [28] Q. Wang, Y. Yang, D. Li, C. Feng, Z. Zhang, Treatment of ammonium-rich swine waste in modified porphyritic andesite fixed-bed anaerobic bioreactor, Bioresour. Technol. 111 (1) (2012) 70e75. [29] H. Bouallagui, Y. Touhami, R.B. Cheikh, M. Hamdi, Bioreactor performance in anaerobic digestion of fruit and vegetable wastes, ChemInform 36 (19) (2005) 989e995.  mez, A. Mora n, Vegetable processing wastes [30] B. Molinuevo-Salces, X. Go addition to improve swine manure anaerobic digestion: evaluation in terms of methane yield and SEM characterization, Appl. Energy 91 (1) (2012) 36e42. [31] H. Bouallagui, M. Torrijos, J.J. Godon, R. Moletta, R.B. Cheikh, Y. Touhami, J.P. Delgenes, M. Hamdi, Two-phases anaerobic digestion of fruit and vegetable wastes: bioreactors performance, Biochem. Eng. J. 21 (2) (2004) 193e197. mez, A. Mora n, M.C. Garcíagonza lez, Anaerobic co[32] B. Molinuevosalces, X. Go digestion of livestock and vegetable processing wastes: fibre degradation and digestate stability, Waste Manag. 33 (6) (2013) 1332e1338. [33] H. Tian, N. Duan, C. Lin, X. Li, M. Zhong, Anaerobic co-digestion of kitchen waste and pig manure with different mixing ratios, J. Biosci. Bioeng. 120 (1) (2015) 1e20. [34] H. Bouallagui, H. Lahdheb, E. Ben Romdan, B. Rachdi, M. Hamdi, Improvement of fruit and vegetable waste anaerobic digestion performance and stability with co-substrates addition, J. Environ. Manag. 90 (5) (2009) 1844e1849. lez, X. Font, T. Vicent, Alkalinity ratios to identify process [35] L. Martín-Gonza imbalances in anaerobic digesters treating source-sorted organic fraction of municipal wastes, Biochem. Eng. J. 76 (28) (2013) 1e5. zquez, X. Font, Long term operation of a thermophilic anaerobic [36] I. Ferrer, F. Va reactor: process stability and efficiency at decreasing sludge retention time, Bioresour. Technol. 101 (9) (2010) 2972e2980. [37] K.H. Hansen, I. Angelidaki, B.K. Ahring, Anaerobic digestion of swine manure: inhibition by ammonia, Water Res. 32 (1) (1998) 5e12. [38] P.N. Hobson, B.G. Shaw, Inhibition of methane production by Methanobacterium formicicum, Water Res. 10 (10) (1976) 849e852. [39] M. Kayhanian, Ammonia inhibition in high-solids biogasification: an overview and practical solutions, Environ. Technol. 20 (4) (1999) 355e365. [40] E.H.T. Benabdallah, S. Astals, A. Galí, S. Mace, J. Mata-Alvarez, Ammonia influence in anaerobic digestion of OFMSW, Water Sci. Technol. A J. Int. Assoc. Water Pollut. Res. 59 (6) (2009) 1153e1158. [41] Q. Niu, Q. Wei, Q. Hong, T. Hojo, Y.Y. Li, Mesophilic methane fermentation of chicken manure at a wide range of ammonia concentration: stability, inhibition and recovery, Bioresour. Technol. 137 (11) (2013) 358e367. [42] I.W. Koster, Liquefaction and acidogenesis of tomatoes in an anaerobic twophase solid-waste treatment system, Agric. Wastes 11 (4) (1984) 241e252. [43] A. Veeken, S. Kalyuzhnyi, H. Scharff, B. Hamelers, Effect of pH and VFA on hydrolysis of organic solid waste, J. Environ. Eng. 126 (12) (2000) 1076e1081. [44] C. Zhang, X. Gang, L. Peng, H. Su, T. Tan, The anaerobic co-digestion of food waste and cattle manure, Bioresour. Technol. 129 (2) (2013) 170e176. [45] X. Guo, W. Cheng, F. Sun, W. Zhu, W. Wu, A comparison of microbial characteristics between the thermophilic and mesophilic anaerobic digesters exposed to elevated food waste loadings, Bioresour. Technol. 152 (152C) (2014) 420e428. [46] V.P. Tale, J.S. Maki, C.A. Struble, D.H. Zitomer, Methanogen community structure-activity relationship and bioaugmentation of overloaded anaerobic digesters, Water Res. 45 (16) (2011) 5249e5256.