Abstract

There is a tremendous increase in the demand for efficient and productive food supply chain due to rapid increase in the world population. At the same time, anaerobic digestion (AD) of discarded food waste residue (FWR) along with buffer substrate (BS) could be a sustainable way to transform waste to energy. In this study, substrate combinations (SCs) were prepared using pre (OFPre), and post (OFPost) processed FWR; anaerobic sludge (AS) as BS. The Biochemical methane potential test (BMPT) was performed for SCs prepared at varying proportion. The decline in substrate biodegradability ratio (SBR) and C/N ratio was also determined. The study was performed under three different temperature condition ranging from 35 °C to 73 °C over 90 days. It was found that SC of C5 and C6 had maximum biogas generation potential (BGP) under thermophilic (45 °C to 52 °C) temperature condition, and the cumulative CH4 production was 997 and 880 ml/g of VS. The artificial neural network (ANN) and random forest regression (RFR) were the two-machine learning (ML) tools used to predict and correlated the CH4 yield using input data series of SCs, SBR and C/N ratio. The ANN-CH4 prediction model was found to have best possible prediction efficiency.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call