Carbon to nitrogen ratio and organic loading rate optimization of sewage sludge and rice straw: Economic analysis and anaerobic digestion process understandings through machine learning

Biowaste is considered an abundant renewable energy resource which can ensure the energy security of the rural community. In the current study, the anaerobic digestion of sewage sludge and rice straw was carried out in batch and semi-continuous mode while carbon to nitrogen (C/N) ratio and organic loading rate (OLR) polynomial regression models were developed for commercial biogas plants. During the batch fermentation, C/N of 30 was optimized with 26 % methane enhancement, while during semi-continuous mode, an OLR of 3.5 g.VS/L.day was concluded as optimum with 282.36 mL/g.VS.d methane generation capability. The statistical co-relation analysis was carried out according to the Pearson co-relation matrix to understand the relationship of total volatile fatty acids, chemical oxygen demand, total ammonia and free ammonia towards methane generation. Machine learning techniques like feature engineering and machine learning models were applied to understand parametric influences in methane generation. The random forest model has provided a higher goodness of fit value (0.99). A detailed economic feasibility analysis of the biogas plant dealing biowaste of 10 tons/day was also executed, with an energy generation potential of 112 GJ/day and 6.93 years was deduced as the breakeven point for the commercial scale adaptation.