3.3 Process performance under stable and disturbed states targeting HRT
During the HRT shortening stage (Fig. 3), the volumetric biogas production registered was stable at around 4.4 L L-1 d-1 under 5 d HRT (Fig. 3a). The specific methane yield was 354±27 mL CH4 g-1 VS-1, and it represented the 65% v/v of the biogas (Table 2). This result fell within the range reported, i.e. 182-368 mL CH4 g-1 VS-1, in previous studies treating straw and food waste in batch reactors (Liu et al., 2018); however, it was higher than that reported for the mono digestion of food waste at longer HRT under thermophilic conditions (Kim et al., 2006), implying a synergistic effect of the co-digestion strategy. The methane yields declined by 29% (251±63 mL CH4 g VSin-1) alongside with the decrease of the HRT to 3 and by 78% (78±12 mL CH4 g VSin-1) with HRT shortened to 1.5 d, respectively.
With the decrease in the HRT, the propionate concentration only slightly changed while the acetate concentration accumulated from 333±197 mg L-1 to 951±577 and 2533±496 mg L-1 at HRT of 5, 3 and 1.5 d, respectively (Fig. 3c and Table 1). The concentration of propionate at HRT of 1.5d was stable at about 224±75mg L-1 and so it was 2-fold lower than that reported in HRT of 5d. It was therefore obvious that the consumption rate of acetate was the primary rate-limiting factor at HRTs of 3 and 1.5 d. Although the concentration of VFAs sharply increased during the shortest HRTs, it was much lower than VFAs concentration at OLR of 17.1 g-1 VS L-1 d-1 in the first stage of current study. Therefore, the reason behind the reduction in gas production rate at this stage was most probably due to the excessive microbial washout caused by the shortest HRTs, and partially because of the impact of VFA accumulation on methanogens activity. The phenomena of the washout effects and the inhibition of VFA have been previously reported under HRT shortened to 5d (Algapani et al., 2016).
The VS removal efficiency with the shortening of the HRT was consistent with the methane production results and this further verified the VFA accumulation. Alongside with the decrease of the HRT, the removal efficiency of VS decreased down to 5.2% at the end of experiment, which was 11-fold less than the removal efficiency attained at an 8d HRT. The variation of pH between different HRTs was negligible and varied between 7.2 and 7.3. To conclude, the low methane production during the shortest HRT can be explained by i) the finite digesting time; ii) the microbiome washout effect; and iii) methanogenesis inhibition by VFA accumulation.