Modeling suction of unsaturated granular soil treated with biochar in plant microbial fuel cell bioelectricity system.

There is an initiative driven by the carbon-neutrality nature of biochar in recent times, where various countries across Europe and North America have introduced perks to encourage the production of biochar for construction purposes. This objective aligns with the zero greenhouse emission targets set by COP27 for 2050. This research work seeks to assess the effectiveness of biochar in soils with varying grain size distributions in enhancing the soil-water characteristic curve (SWCC). This work further explores the effect of different combinations of biochar content (0 to 15 mass %) on the bioelectricity generation from biochar-improved plant microbial fuel cells (BPMFC). Additionally, different machine learning models such as the "Gradient Boosting (GB)", "CN2 Rule Induction (CN2)", "Naive Bayes (NB)", "Support vector machine (SVM), "Stochastic Gradient Descent (SGD)", "K-Nearest Neighbors (KNN)", "Tree Decision (Tree)", "Random Forest (RF)", and "Response Surface Methodology" (RSM), have been developed to predict SWCC based on soil suction, electric current, electrical potential, volumetric water content, temperature, and bulk density. The newly established model demonstrates a reasonable ability to predict SWCC and a cheaper technology in predicting the suction of unsaturated soils in relation to the studied bioelectric factors of the BPMFC. Overall, in this research paper, the GB, SVM and CN2 outclassed the other regression techniques in this order thereby proposing the cheapest technology with the highest performance index to predict the SWCC behavior of unsaturated soils in a BPMFC system.