3.1 Daily biogas production and average methane content
The daily biogas production (DBP) of MAD and TAD digesters of 40°C and 60°C pretreatment condition are shown in Figure 1(a) and (b). The results revealed that in MAD system (Figure 1(a)), the DBP of CS dropped slowly from 6.25 L·d-1 to 5.62 L·d-1 as the OLR increased from R1 to R3 of two treatment conditions while untreated one was lower than that and showed no obvious increment. During MAD processes, the average methane content showed respective increment from 52.49% to 56.90% and from 53.05% to 57.52% of 40°C and 60°C pretreatment condition as shown in Table 2. However, in TAD system (Figure 1(b)), the DBP of CS gradually improved from 6.12 L·d-1 to 7.63 L·d-1 as the OLR increased from R1 to R3 of two treatment conditions while untreated one showed a little improvement. During this TAD processes, the average methane content respectively kept stable at 52.48% and 52.09% of 40°C and 60°C pretreatment condition for all OLRs as shown in Table 2.
The relatively fast DBP rates of CS in TAD systems was due to the thermophilic hydrolytic bacterial population which would accelerate the hydrolysis process, mainly at a temperature of around 55°C [20]. In addition, high rate of hydrolysis increases the production of volatile fatty acids (VFAs) from CS is also more efficient at TAD reactors than at mesophilic digestion [21]. Therefore, the daily biogas productions at TAD conditions were higher than MAD conditions.
3.2 Influence of OLR on reactor performance
A change from low OLR to high OLR may lead to system instability, such as accumulation of VFAs, a lower pH, and a mismatch between the growth rates of VFA-producing and consuming microbes and thus resulted in low biogas production [22, 23]. This was consistent with the results of Figure 1(a), which demonstrated the production rates of biogas decreased in MAD systems as OLR increased. Nevertheless, the production rates of biogas improved in TAD systems as OLR increased. At the same time, in TAD system, the feeding rate of 2.0 g·L-1·d-1 was 1.25 times of 1.6 g·L-1·d-1, which produced enhanced biogas production rate by 24.59% and 38.18% than that of MAD and untreated conditions, respectively. This was because the consuming speeding of soluble substance in TAD system higher than in MAD systems. As a result, the TAD systems achieved higher biogas production rate relative than MAD and untreated systems as OLR increased. Besides, the OLR (1.6 g·L-1·d-1) of feeding in this study were higher than that of another study of Tian (1.44 g·L-1·d-1) [24], and the DBP in this study were also higher than that study, which only attained 4.83-5.30 L-1·d-1. This illustrated that the OLRs included in this study were appropriate for AD digesters.
3.3 Comparison of different pretreatment and AD temperatures
3.3.1 Comparison of different pretreatment conditions
Pretreatment temperatures of biomass materials usually play an important role in batch experiment but may not intrigue decisive effects of CSTR study [25]. In this work, at 40°C pretreatment condition, the rates of biogas production under MAD were 6.25 L·d-1, 5.54 L·d-1 and 5.62 L·d-1 for three respective OLRs. At 60°C pretreatment condition, the rates of biogas production under MAD were 6.24 L·d-1, 5.57 L·d-1 and 5.60 L·d-1 for three respective OLRs. The same trends also presented in TAD system, and likewise, the methane production of different pretreatment conditions indicated similar results as well. As shown in Table 2, the gaps of methane yield between 40°C and 60°C pretreatment condition were small under MAD conditions which were 5, 9 and 3 mL·gvs-1 for three OLRs, respectively. These also showed the same trends of the gaps under TAD which were 12, 8 and 0 mL·gvs-1 for three OLRs, respectively. This could be directly related to the gap of production of TVFAs hydrolyzed from CS between 40°C and 60°C. As shown in Figure 2 and Table 3, the TVFAs and ethanol produced under 40°C and 60°C pretreatment showed no obviously difference, even the production of acetic acid under 40°C was a little bit more than that under 60°C, which not harshly inhibit the production of biogas.
Moreover, the productions of methane in this research were higher by 32.7% than previous study which gained 211 mL gvs-1·VS at MAD using CS pretreated in 20°C [26]. The results were also higher by 15.7% than the study of Liu et al., which gained 242 mL·gvs-1 under TAD with 20°C treated CS [7]. These results revealed that under MAD or TAD condition, enhancing the pretreatment temperature from 40°C to 60°C was not highly conducive to improve the biogas and methane yield. Based on cost benefit and effectiveness, 40°C pretreatment temperature could be suggested as the optimum pretreatment temperature for MAD and TAD processes.
3.3.2 Comparison of different AD conditions
The process performance of anaerobic reactors often depend on temperature, which plays its part in making favorable conditions for degradable substance in reactors especially in long-term operating condition [27]. Therefore, the Figure 3 and Table 2 showed significant daily methane production (DMP) and methane yields difference between MAD and TAD systems. With the OLR increased, the gap in methane yield between the two AD temperatures was gradually becoming greater. This was because excess VFAs accumulated in reactors inhibited MAD processes but stimulated TAD processes to result different methane yields. In R3 phase (highest OLR phase), the methane yield of TAD were enhanced by 30.38%, 28.26% and 52.54% than that of MAD with 40°C, 60°C treatment and untreated one. The methane yield of TAD under 40°C and 60°C pretreatment conditions were higher than that by 31.11% compared with untreated CS.
The higher methane yield in TAD system was attributed to the high rate of methanogenesis under thermophilic condition, so that VFA could be easily converted to methane after hydrolysis and acidogenesis process in the thermophilic reactors [28]. Jiménez et al. also reported that specific methane activity (SMA) of thermophilic methanogens was higher than mesophilic methanogens [29]. In a study using wheat straw as raw material, methane yield under thermophilic conditions was higher than that under mesophilic conditions, where the results are in line with the present work [30]. Consequently, the results revealed that thermophilic system can tolerate high OLR feeding and could be an important insight for the future biogas industry using CS as feedstock.
3.4 Changes of main compositions
Content of LCH after pretreatment was shown in Table 4. Prior to treatment the main components of CS, lignin, cellulose and hemicellulose, accounted for 40.36%, 20.07% and 11.41% respectively giving 71.84% in total. Following 40°C and 60°C treatments, the proportion of hemicellulose significantly decreased to 13.15%, which was 52.62% lower than that of untreated control. The content of lignin also decreased by 5.6%-7.4%, which suggested partial lignin removed via NaOH pretreatment. Lignin degradation could release more cellulose and hemicellulose, thus increasing the biodegradability of the CS.
LCH were the main carbon sources for anaerobic microorganisms during AD of CS. The biogas and methane production were attributed to biological degradation of cellulose and hemicellulose. The conversion rates of cellulose, hemicellulose, TS and VS during the process of AD were used to evaluate the biodegradability improvement and digestion performance of the system [31]. Conversions rate of cellulose, hemicellulose, TS and VS after AD were analyzed from the effluent discharged after steady-state. As shown in Table 5, compared to the untreated, the reductions of cellulose, hemicellulose and lignin after AD with 40°C and 60°C pretreatment improved by 12.63%-35.37%, 8.13%-32.59% and 26.92%-67.95%, respectively. In addition, the TS and VS removal rate of 40°C and 60°C pretreatment amounted to 46.3%-50.6 and 53.55%-61.9% while the untreated one kept 43.3%-47.5% and 53.5%-58.7%. At the same time, the reductions of all components in TAD systems showed relative higher than in MAD systems. These results were also in agreement with the fact that under thermophilic digestion the rate of organic matter removal is more efficient [32]. The main reason for the improvement of conversion rate at TAD might be the biodegradation of more feedstock which accounts for more methane produced from excess VFAs that released directly from the thermophilic reactor. The high conversion of substrates was also corresponded to the biogas and methane production which presented a better conversion rate of CS under TAD condition from R1 to R3.
3.5 Evaluation of system stability
During the AD processes, the activities of microbial communities can be deeply influenced by pH, ammonia nitrogen (AN) and total volatile fatty acids (TVFAs). More specifically, the loss of methanogens, high accumulation of TVFAs and low total alkalinity concentration (TAC) can inhibit the process of AD. Additionally, the ratio of TVFAs to TAC (TVFAs/TAC) can be used as significant indicators of the stability of AD [33].
3.5.1 pH
The value of pH is an important parameter in determining the stability of the AD system. In this study, all the pH values during entire AD period were in the range of 6.8-7.7 (Figure 4(a)). This shows that the pH values were at optimum level, which was believed to be suitable for the growth of methanogens [9]. From two AD conditions applied in this test the TAD systems had a relatively higher pH than the MAD systems during all the OLRs. This was mainly due to high consumption rate of VFAs produced at optimum pH under thermophilic condition.
3.5.2 Ammonia nitrogen and TVFAs/TAC
Optimal ammonia concentration ensures sufficient buffer capacity of methanogenic medium in AD thus increasing the stability of the digestion process. However, high ammonia is regularly reported as the primary cause of digester failure because of its direct inhibition of microbial activity [34]. As shown in Figure 4 (b), The AN of the process effluent of six CSTRs was high after starting feeding but gradually reduced to 200 mg·L-1 at the end of the first OLR (R1) and less than 200 mg·L-1 in the second to third OLRs (R2 and R3), which didn’t inhibit the process [35]. At the same time, there were two small AN peak values by the end of later two OLRs for more substrates were degraded by microorganism.
In this study, TVFAs from the effluent of six CSTRs at the OLRs of 1.6, 1.8 and 2.0 g·L-1·d-1 were determined. As shown in Figure 4 (C), the time of TVFAs accumulation of 40°C pretreated samples were 10 days faster and had more TVFAs than the samples pretreated at 60°C temperature. In addition, in R3 phase, the TVFAs reached its peak level for reactors with the samples at 40°C and 60°C pretreatment temperature both at the 200th day, this was because the gap of amount of TVFAs was small in two treatment conditions. Moreover, the TVFAs produced in TAD were relatively less than that of MAD during R1 to R3 processes while the pH value in TAD were relatively higher than that of MAD, which showed the consistency of TVFAs and pH value.
The ratio of TVFAs/TAC can be applied to evaluate the buffer ability of anaerobic digestion. Generally, the digestion process could keep stable when the ratio of TVFAs/TAC is below 0.4, and however, the system shows instability when the ratio exceeds 0.8 [27]. TVFAs and TVFAs/TAC of this work were shown in Figure4 (C). The results showed that the ratios of VFA to alkalinity ration of the effluents from six reactors were from 0.01 to 0.2. This showed that the operating condition in each CSTR was at optimum level. Besides, the TVFAs/TAC of TAD was lower than MAD during the last two OLRs, which indicated that TAD showed better buffer ability at high OLR.
3.6 Energy balance
As AD is an energy production system, it is important to evaluate the energy balances of the system for the characteristics of the six systems digestates were different (Figure 4). In order to evaluate comprehensively about six AD reactors, the energy input of the systems should also be considered. In this paper, three processes (pretreatment, pumping and mixing) were considered (Eq. (1)) and the results are summarized in Table 6.
The input energy includes the electricity and heat input [36]. The input electricity of the TAD systems was slightly higher than those of the MAD systems because the power and heating of TAD reactors have more demands. The input heat of the 40°C-treatment was slightly lower than that of 60°C-treatment and Ei,h of the two temperatures were the same. The energy balance analysis of AD alone (Table 6) showed the ΔE of the three MAD systems were higher than that of TAD systems, which was consistent with the value of Ro/i of six different reactors. The results suggested the MAD of CS in this study recovered more energy from the degraded organic matter of CS than the thermophilic digestion although it produced less energy from methane gas. In addition, the ΔE of 40°C -treatment CS were relatively higher than that of 60°C-treatment while Ro/i also showed high, this demonstrated the AD of 40°C-treated CS recovered more energy. As a result, 40°C was recommended as a preferable pretreatment temperature.