A hybrid three-dimensional agent-based computational framework for simulating multiscale microbial fuel cell dynamics

Microbial fuel cells (MFCs) convert the chemical energy stored in organic substrates into electrical energy through the metabolic activity of electroactive microorganisms. Their performance is governed by coupled biological, electrochemical, and mass-transport processes operating across multiple spatial scales, making mechanistic prediction challenging. This study presents a hybrid three-dimensional agent-based computational framework that integrates individual microbial behaviour with spatial substrate transport, local pH variation, biofilm development, and reactor-scale electrochemical response. Individual microbial agents respond to their local microenvironment through growth, substrate consumption, attachment, and extracellular electron transfer represented using effective model relationships. The framework was implemented using GPU-based parallel computing to enable large-scale simulations while preserving individual-cell heterogeneity and spatially resolved environmental interactions. Simulations qualitatively reproduced the characteristic temporal behaviour of a published acetate-fed microbial fuel cell, including the initial lag phase, voltage increase, stabilization, substrate depletion, and response to substrate replenishment. It produced a peak voltage of 0.708 V, compared to 0.693 V experimentally (2.2% error), while five independent stochastic realizations showed low variability in the predicted voltage response. Numerical verification through spatial convergence, stochastic realizations, polarization analysis, inoculum-loading studies, and local sensitivity analysis demonstrated stable model behaviour and identified the principal factors influencing voltage generation and coulombic efficiency. The proposed framework provides an extensible computational platform for investigating multiscale microbial fuel-cell processes and may support future model-based analysis and design of bioelectrochemical systems.