Modelling the energy harvesting from ceramic-based microbial fuel cells by using a fuzzy logic approach

Area of Science:

• Bioenergy and Wastewater Treatment
• Renewable Energy Technologies
• Electrochemical Systems

Background:

• Microbial fuel cells (MFCs) offer dual benefits of bioenergy generation and wastewater treatment.
• Optimizing power density is crucial for the large-scale application of MFCs.
• Ceramic-based MFCs utilizing human urine present a sustainable energy source.

Purpose of the Study:

• To simulate and maximize the absolute power output of ceramic-based MFCs fed with human urine.
• To evaluate the predictive accuracy of a fuzzy inference system (FIS) compared to nonlinear multivariable regression.
• To identify optimal operational parameters for enhanced energy harvesting.

Main Methods:

• A fuzzy inference system was developed to model MFC power output.
• Key parameters including membrane thickness, anode area, and external resistance were systematically varied (81 assays).
• Performance was assessed using R-squared and Variance Account For (VAF) to compare FIS and regression models.

Main Results:

• The FIS demonstrated superior prediction accuracy, achieving R-squared of 94.85% and VAF of 94.41%.
• Nonlinear multivariable regression yielded lower accuracy with R-squared of 79.72% and VAF of 65.19%.
• Maximum power output of 450 µW was predicted at specific anode area (160-200 cm²), external load (~900 Ω), and membrane thickness (1.6 mm).

Conclusions:

• Fuzzy inference systems provide a more reliable approach for predicting power output in ceramic-based MFCs.
• The study identified optimal conditions for maximizing energy harvesting from urine-fed MFCs.
• Membrane thickness showed minimal impact on power output within the tested range, simplifying optimization.