Introduction

Interest in anaerobic digestion of lignocelluloses for bioenergy production has surged in recent years among researchers worldwide. Corn yield throughout China exceeded 205.6 million tons in 2012 according to the China Statistical Yearbook 2013 [1]. The vast amount of corn stover generated from corn cultivation in China is either burned or abandoned in open fields, causing serious environmental pollution. To address this issue, anaerobic digestion of corn stover for bioenergy production has been proposed [2]. Despite obvious benefits of the procedure, some characteristics of corn stover, such as its lignocelluloses structure, result in poor methane production efficiency and low hydrolysis rate [2, 3].

Another potential resource that can be used for bioenergy production is vegetable waste. Annually, up to 1.3 × 107 ton of vegetable waste is generated during vegetable processing and transport in cities in China [4]. Vegetable wastes are characterized as easily biodegradable organic matter with high moisture contents and low C/N ratios [5, 6]. Anaerobic mono-digestion of vegetable waste is limited by rapid acidification [7], which causes failure of the anaerobic digestion process. As the substrate characteristics of corn straw and vegetable waste complement each other, these materials may be anaerobically co-digested to balance the C/N ratio, eliminate VFA accumulation, and increase the methane production efficiency of corn stover. Therefore, the performance of anaerobic digestion can potentially be boosted by anaerobic co-digestion of corn stover and vegetable waste.

Anaerobic co-digestion of lignocelluloses with other easily biodegradable waste has been considered to be a promising technology for improving the digestibility of lignocelluloses [8]. Zhong et al. [9] found that anaerobic co-digestion of blue algae and corn stover increased methane productivity by 46% and improved organic loading rates for the anaerobic process, because anaerobic co-digestion increases hydrolysis and methanogenic rates by eliminating volatile fatty acid (VFA) buildup and reducing free ammonia (FA) content. Similarly, efficiency and stability of the anaerobic process can be improved by co-digestion of corn stover and chicken manure [10]. Despite these promising results, the efficiency of anaerobic digestion process is largely dependent on the characteristics and type of co-digested biomasses. To the best of our knowledge, studies on anaerobic co-digestion using corn stover and vegetable waste have not been reported to date.

In general, semi-dry anaerobic digestion is defined based on total solid (TS) concentration of 10–15% [10, 11]. Compared with wet and dry anaerobic digestion, semi-dry anaerobic digestion not only has high mass transfer between organic matter and anaerobic microorganisms but also reduces water consumption and wastewater generation [12]. Therefore, semi-dry anaerobic co-digestion of corn stover and vegetable waste was carried out in the present study.

The main goal of this study was to investigate the feasibility of semi-dry anaerobic co-digestion of corn stover and vegetable waste. Six mixing ratios of corn stover and vegetable waste were anaerobically co-digested to achieve optimal methane production in batch reactors. The process stability of anaerobic digestion was assessed in two leaching bed reactors (LBRs). Comparisons of methane production and VS reduction, as well as digestate characteristics, were conducted. The modified Gompertz model was used to simulate methane production.

Materials and Methods

Corn Stover, Vegetable Waste, and Anaerobic Seed Cultures

Corn stover was collected from Nongcui Garden of Anhui Agricultural University, China, in September 2016 and chopped into 3–4 cm pieces with shears after air-drying. Pak choi (Brassica chinensis L. cv. “Shanghaiqing”) was purchased from a supermarket as vegetable waste and blended to homogeneity in a food blender before anaerobic digestion. The inoculums were well-digested rice straw obtained from our laboratory [13]. Anaerobic culture was concentrated by settling and used as inoculums. The chemical characteristics of the feedstock and anaerobic culture are shown in Table 1.

Table 1 Chemical characteristic of the feedstock and anaerobic culture
Full size table

Anaerobic Co-digestion Experiment

Anaerobic co-digestion was conducted in 12 batch reactors (1 L jars) and 2 identical LBRs. The inner diameter of the LBR was 14 cm, its height was 50 cm, and its working volume was 4.0 L Detailed descriptions of the batch reactor and LBR were provided elsewhere [12, 13].

A detailed experimental design of anaerobic digestion is given in Table 2. The initial TS concentrations of reactor for anaerobic co-digestion were approximately 10%, whereas those for mono-digestion of vegetable waste and corn stover were 3.6 and 12.1%, respectively. These may be due to the high water content of vegetable waste and high TS content of corn stover. The C/N ratio of digested substrate should be in the range of 20:1–30:1 for anaerobic digestion [3]. To keep the C/N ratio of 20–30 and reactor TS concentration of about 10%, six mixing ratios of corn stover and vegetable waste (i.e., 0:1 (R1), 1:0 (R2), 5.5:1 (R3), 8:1 (R4), 11:1 (R5), and 14:1 (R6) (TS, w/w)) were selected. The C/N ratio of feedstock was 27.4 (R2), 21.2 (R3), 22.7 (R4), 23.7 (R5), and 24.4 (R6), respectively, and mono-digestion of vegetable waste (R1) had a C/N value of 9.1.

Table 2 Experimental design on semi-dry anaerobic co-digestion of corn stover and vegetable wastes
Full size table

For 1 L batch experiments, appropriate amounts of corn stover and vegetable waste were mixed with 350 mL of anaerobic seed cultures in a 1 L jar, and distilled water was added to achieve the desired TS concentration. Then the feedstock was uniformly mixed. The reactor was finally sealed with butyl rubber stoppers after flushing with nitrogen gas for 2 min. Each bottle was manually mixed twice a day to avoid stratification during the 50 days anaerobic digestion period. Each experiment was conducted in duplicate. Methane yield under standard conditions is reported.

The result of 1 L batch experiment showed that the experimental group with a TS ratio of 14:1 had optimal performance. Therefore, in the LBR experiments, two TS ratios of corn stover and vegetable waste [1:0 (LBR1) and 14:1 (LBR2)] were selected to determine changes in process stability. For anaerobic digestion of LBR, corn stover and/or vegetable waste, and 1000 mL of anaerobic cultures were added to the bucket, and the appropriate volume of distilled water was added to achieve a TS concentration of 10%. Anaerobic digestion in the LBR was conducted for 45 days. During anaerobic digestion, 450 mL of the leachate was recycled to the top of the reactor by spraying the bed twice daily [13]. Biogas production was recorded daily by a wet tip gas meter and reported in units of liters per day under standard conditions. Biogas production and composition, as well as pH, COD, VFA, NH4+–N, and TA from the LBR leachate, were analyzed during anaerobic digestion.

Kinetic Assessment and Calculations

The modified Gompertz model has been widely used to describe hydrogen and methane production [14] and is expressed as indicated in Eq. (1) below.

$$M=P \times \exp \,\left( { - \exp \left[ {\frac{{{R_m} \times e}}{P}(\lambda - t)+1} \right]} \right)$$
(1)

where M is the cumulated methane yield at time t, P is the potential methane production, Rm is the maximum rate of methane production, e = 2.718281828, λ is the lag time, and t is the measured time. P, Rm, and λ were estimated by non-linear fitting using Origin 8.0.

The predicted methane yield was calculated for different mixing ratios of corn stover and vegetable waste using Eq. (2).

$$PMY=\frac{{{X_1}T{S_1}+{X_2}T{S_2}}}{{T{S_1}+T{S_2}}}$$
(2)

where PMY is the predicted methane yield of anaerobic co-digestion, X1 and X2 are respectively the experimental methane yields of corn stover and vegetable waste by anaerobic mono-digestion from 1 L batch experiment, and TS1 and TS2 are the TS weights of corn stover and vegetable waste added into the reactor, respectively.

Analytical Methods

Methane content was determined by gas chromatography (Ruipu SP-6890, China). Standard methods were used to measure the concentrations of TS, volatile solids (VS), COD, and NH4+–N [15]. Free ammonia (FA) concentration was calculated using a previously reported method [16]. VFA and TA were determined by previous methods [12]. Elemental C, H, N, and S contents were determined using an elemental analyzer (Vario EL Cube, Germany). Neutral detergent fiber, acid detergent fiber, and acid detergent lignin were analyzed in triplicate according to a previously reported method [17].

Statistical Analysis

One-way ANOVA with Fisher’s least significant difference (LSD) test was performed to evaluate significant differences at p < 0.05 using SPSS 14.0 for Windows.

Results and Discussion

Methane Production and Kinetic Assessment from Batch Experiment

The methane production efficiencies of semi-dry anaerobic digestion of corn stover and vegetable waste achieved in the 1 L batch reactors are summarized in Table 3. After 50 days of semi-dry anaerobic mono-digestion, methane yields reached 638.3 mL/g VS for vegetable waste and 289.5 mL/g VS for corn stover. These methane yields of vegetable waste were higher than those obtained from 23 samples of vegetable waste ranging from 190 to 400 mL/g VS [18], and the solid-state batch anaerobic digestion of corn stover [3]. These differences observed were attributed to the diversity of substrate characteristics [19] and other factors, such as inoculum source [20] and TS concentrations of reactor. The methane yields for anaerobic co-digestion of corn stover and vegetable waste ranged from 314.5 mL/g VS to 323.4 mL/g VS. Specific methane yield of anaerobic mono-digestion of corn stover (R2 experimental group) was 289.5 mL/g VS and significantly lower than those of anaerobic co-digestion.

Table 3 Comparison in the methane production efficiency of anaerobic digestion of corn stover and/or vegetables wastes
Full size table

The PMYs for anaerobic co-digestion were calculated using Eq. (2) and ranged from 275.2 to 283.9 mL/g TS (Table 3). As shown in Table 3, the ratios of experimental methane yield (EMY) and PMY were in the range of 0.99–1.04, and increasing methane yield was not observed in the anaerobic co-digestion of corn stover and vegetable. However, an increase in mean methane production rate was found in anaerobic co-digestion compared to anaerobic mono-digestion, particularly compared to R1, suggesting synergistic effects during anaerobic co-digestion [10]. These effects may be attributed to balance of nutrients and alleviation of VFA accumulation in the reactors [8]. Moreover, the mean methane production rate increased with decreasing amount of vegetable waste. This trend may be explained by the easily degradable characteristics of vegetable waste, which produces accumulation of VFA during anaerobic digestion, leading to instability of the anaerobic digestion process [6]. The mean methane production rate of anaerobic mono-digestion of corn stover (R2) was higher than those of R3 and R4, but lower than those of R5 and R6.

The mean methane contents observed for semi-dry anaerobic co-digestion ranged from 62.4 to 65.8% (Table 3). The technical digestion time (T80) refers to the amount of time during which cumulative biogas production reaches 80% of the total biogas yield [21]. The value of T80 in the present work ranged from 19.9 to 29.5 days. VS removal efficiencies for corn stover ranged from 86 to 88.7% after 50 days of digestion. The TS and VS removal efficiencies of R2 group were close to those of R3, R4 and R5 groups. These values were similar to those obtained for anaerobic co-digestion of livestock and vegetable processing waste (86% VS reduction) [22] but slightly higher than that for anaerobic co-digestion of fruit/vegetable waste with other organic waste (73.1–85.4% VS reduction) [5]. R6 exhibited the highest methane production efficiency and rate, as well as VS removal efficiency among the reaction systems observed; this system also yielded the lowest T80 among the four mixing ratios of anaerobic co-digestion studied. Thus, the optimal mixing TS ratio of corn stover and vegetable waste was 14:1.

The values of the parameters calculated by the modified Gompertz model are presented in Table 4. The coefficient of determination (R2) was higher than 0.98 and the error rates were less than 8.4%, which indicated that the modified Gompertz model is suitable for simulating the experimental data. The maximum methane production rate (Rm) increased with decreasing amount of vegetable waste for anaerobic co-digestion, and R6 had the highest Rm compared with those of mono-digestion of corn stover and anaerobic co-digestion. The lowest lag time (λ) occurred in R6. Thus, results of kinetic assessment also indicated that the optimal mixing TS ratio of corn stover and vegetable waste was 14:1.

Table 4 Values of the parameters calculated by the modified Gompertz equation
Full size table

Methane Production from LBR Experiment

The profiles of specific biogas production rate in LBRs during semi-dry anaerobic digestion are presented in Fig. 1. Biogas production rates rapidly decreased in the first 5 days of digestion in both LBR1 and LBR2. After continuous alkaline (lime) supplementation for pH adjustment to 6.8 for the LBR leachate, biogas production rates gradually recovered after the 12th day, and peak values occurred on the 26th day (at 18.5 mL/(d. g TS)) for LBR1 and 30th day (at 16.7 mL/(d. g TS)) for LBR2. These results reveal severe inhibition at the initial stage of mono-digestion and co-digestion.

Fig. 1
figure 1

Biogas production rate of semi-dry anaerobic co-digestion of corn stover and vegetable wastes by LBR

Full size image

The specific methane yields were 185.4 mL/g VS, respectively, for LBR1 and 182.4 mL/g VS, respectively, for LBR2. The VS removal efficiency of corn stover reached 83.5% for LBR1 and 84.3% for LBR2. These values are slightly lower than those obtained from the batch reactor, suggesting that the scaling-up process from batch reactor to LBR decreased the efficiency of anaerobic digestion. The T80 (30.5 days) for LBR1 was lower than that for LBR2 (33.3 days). This finding suggests that the biogas production rate of anaerobic co-digestion of corn stover is lower than that of mono-digestion of the same substrate.

Process Stability of LBR Experiment

The pH profile (Fig. 2) observed during semi-dry anaerobic co-digestion reflects the relative stability of the anaerobic process. The pH profiles decreased sharply in the first 3 days of digestion from 6.72 (1st day) to 5.87 (3rd day) in LBR1 and from 6.48 (1st day) to 5.83 (3rd day) in LBR2. These profiles then gradually increased after alkaline supplementation and remained over 7.5 after the 29th day of digestion. The optimum pH for methanogens archaea ranges from 6.5 to 7.2 [23], which suggests that acidification at the initial stage of digestion inhibited methane production during the anaerobic process. No significant difference in pH was observed between the two LBRs.

Fig. 2
figure 2

Change in pH of LBR leachate from semi-dry anaerobic co-digestion of corn stover and vegetable wastes

Full size image

VFAs are important intermediates in the anaerobic degradation of organic wastes. The COD and VFA profiles obtained during semi-dry anaerobic digestion are shown in Fig. 3. COD and VFA accumulations were observed at the initial stage of anaerobic digestion, which was the cause of rapid decrease in pH. The maximum COD values of 34.4 g/L (7th day) for LBR1 and 36.9 g/L (7th day) for LBR2 were obtained. VFAs ranged from 710 to 3960 mg/L for LBR1 and from 760 mg/L to 5910 mg/L for LBR2. There was no significant difference in VFA and COD values between the two LBRs.

Fig. 3
figure 3

Change in COD (a) and VFA (b) of LBR leachate from semi-dry anaerobic co-digestion of corn stover and vegetable wastes

Full size image

The VFA and alkalinity ratio may be used to evaluate the stability of anaerobic processes [5, 24]. The TA and VFA/TA ratio profiles observed during semi-dry anaerobic co-digestion are presented in Fig. 4. Increase in TA was observed in the first 7 days of digestion, likely due to NH4+–N generation and alkali supplementation, and TA values ranged from 7560 to 9410 mg CaCO3/L for the two LBR experiments. VFA/TA ratios remained between 0.4 and 0.75 during the first 10 days of digestion in the two LBRs, which indicates some instability upon initial digestion [24]. No significant difference in TA contents and VFA/TA ratio was found between the two LBRs.

Fig. 4
figure 4

Change in TA and VFA/TA ratio of LBR leachate from semi-dry anaerobic co-digestion of corn stover and vegetable wastes

Full size image

High NH4+–N concentration can inhibit microbial activity during anaerobic digestion of vegetable waste [5]. The NH4+–N contents ranged from 516 to 789 mg/L for LBR1 and from 582 to 826 mg/L for LBR2 (Fig. 5). Levels of NH4+–N ranging from 200 to 1000 mg/L showed no antagonistic effect on the anaerobic digestion process [25]. FA is more toxic than NH4+–N because FA can penetrate through the cell membrane and accumulate in the cell, thereby inhibiting enzymes or causing proton imbalance [16, 25]. FA was correlated with the pH, temperature, and NH4+–N concentration [16], and pH values varied between 7.5 and 7.7 after 38 days of digestion. FA contents reached peak values on the 38th day, yielding 27.1 mg/L for LBR1 and 31.3 mg/L for LBR2. Calli et al. [26] reported that FA levels above 200 mg/L inhibit propionate-degrading acetogenic bacteria at pH 7.7 and 35 °C. Statistical analysis revealed no significant difference in NH4+–N and FA contents between the two LBRs.

Fig. 5
figure 5

Change in NH4+–N and FA contents of LBR leachate from semi-dry anaerobic co-digestion of corn stover and vegetable wastes

Full size image

Based on the above analysis, inhibition occurs at the initial phase of semi-dry anaerobic co-digestion at the LBR. This was due to acidification during the initial stage of digestion, as reflected by pH less than 6.0, and accumulation of COD in the two LBRs. However, no significant differences in pH, COD, VFA, NH4+–N, and TA levels were observed between the two LBRs. This suggested that the difference in two anaerobic processes, mono-digestion and co-digestion, was not the main causes of acidification. Therefore, the main causes of acidification were deduced from high TS concentration of semi-dry anaerobic digestion and easy biodegradation characteristics of the digested substrate.

Conclusions

After 50 days of semi-dry anaerobic co-digestion in batch reactors, methane yields reached 314.5–323.4 mL/g VS, and synergistic effects in production rate were observed during anaerobic co-digestion. The optimal TS ratio of corn stover to vegetable waste was 14:1; at this ratio, the highest methane production efficiency and production rate, and VS removal efficiency were obtained. Methane yields in LBRs were lower than those in batch reactors because of strong acidification during the initial stage of digestion, which was attributed to the high TS concentration of semi-dry digestion and easy biodegradation characteristics of digested substrates.