Solid state anaerobic digestion (SS-AD) is a promising technology to produce methane from various kinds of solid organic wastes. Compared with traditional liquid anaerobic digestion (L-AD), SS-AD has several advantages including higher organic loading capacity and lower operational energy consumption. In the integrated anaerobic digestion system (iADs, patent pending), effluent of L-AD is used as inoculum for SS-AD, thus both the effluent disposal problem of traditional L-AD and the inoculation problem of traditional SS-AD can be solved. In the iADs, the L-AD technology has been commercialized, but the SS-AD technology still needs further development. The instability at the initial stage and the low methane yields remain barriers for SS-AD. This research aimed to understand the effects of inoculum, feedstock, and total solids (TS) content on the performance of SS-AD, and to enhance methane production and SS-AD stability through experimental studies and mathematical modeling.
In the first study, effluent from three liquid anaerobic digesters, fed with sewage sludge, food waste, or dairy manure, respectively, were evaluated as inoculum and nitrogen source for SS-AD of corn stover in mesophilic batch reactors. The chemical composition and microbial communities of different L-AD effluents varied substantially, and the difference of L-AD effluent and the feedstock-to-effluent (F/E) ratio significantly affected methane yield from corn stover. At an F/E ratio of 2, the dairy manure effluent resulted in the highest methane yield of 238.5 L/kgVSfeed, which may be attributed to its larger population of hydrolytic bacteria. While the sewage sludge effluent, which had the lowest microbial populations and high ammonia concentration, resulted in the lowest methane yield of 118 L/kgVSfeed. At an F/E ratio of 4, the food waste effluent inoculated digester achieved the highest methane yield of 199.6 L/kgVSfeed, possibly due to its largest population of methanogens that helped to mitigate VFA accumulation from feedstock overloading.
The effects of physico-chemical properties (or non-microbial factors) of L-AD effluent on SS-AD performance were further studied. The microbial activity of the L-AD effluent was altered when a portion of the effluent was autoclaved. The unautoclaved portion of effluent was defined as active (“live”) effluent with viable microbes, and the mixture of active effluent with autoclaved effluent was defined as total effluent. The active effluent content represented the concentrations of anaerobic microbes, while the content of total effluent represented the non-microbial factors, such as macro- and micronutrient, pH and alkalinity. The feedstock-to-active effluent (F/Ea) ratio of 2.2, 4.4 and 6.6, and the feedstock-to-total (F/Et) effluent ratio of 2.2 and 4.4 were examined for their effect on SS-AD performance of corn stover. When F/Ea ratio was increased from 2.2 to 6.6, methane production was not significantly reduced; however reactors became acidified when the F/Et ratio was increased from 2.2 to 4.4. It was concluded that in the studied range, non-microbial factors, which was represented by F/Et, had a greater effect on methane yields than the microbial factors (represented by F/Ea).
To evaluate the effect of feedstock on SS-AD, easily digestible dog food was added as co-substrate for corn stover at F/E ratios of 2, 4, and 6. Co-digestion substantially improved methane yield of SS-AD, with the highest methane yield of 304.4 L/kgVSfeed obtained with substrate containing 50% corn stover and 50% dog food, which was 129% and 9% of that of digesting corn stover and dog food alone, respectively. However, high F/E ratios were less tolerable for easily digestible substrates, as at F/E ratio of 4, addition of 25% of dog food resulted in reactor failure. However, the high methane yield obtained from co-digestion was attributed to the degradation of dog food, and the cellulose and xylan degradation in corn stover decreased as dog food addition was increased.
Total solids (TS) content is an important operating parameter in SS-AD. The volumetric methane production rate in SS-AD was found to increase with the increase in TS until a threshold is reached, and then to decrease. This phenomenon cannot be explained by conventional understanding derived from L-AD. A new SS-AD model was developed to explain the TS effect based on mass transfer limitation and hydrolysis inhibition. This model was verified by experimental data from our own study as well as from the literature, and thus the explanation and the model developed might be applied broadly.
Based on the knowledge obtained from previous chapters, multiple linear regression (MLR) models and artificial neural network (ANN) models were developed to predict methane yield based on characteristics of lignocellulosic feedstock and process parameters of SS-AD. Out of the eleven factors analyzed in this study, lignin, cellulose and extractives contents in feedstock, as well as the inoculation size (F/E ratio) were found to be the most significant in determining the 30-day cumulative methane yield. Besides, the interactions between inoculation size and lignin content were also found to be significant. Both ANN and MLR were suitable methods for the prediction of methane yield, in terms of high R-sqared and low standard error of prediction. The verified models can provide guidance for future feedstock evaluation and process optimization in SS-AD.