Voltage Balancing Model for Series Stacked Microbial Fuel Cell

Abstract

Microbial Fuel Cell (MFC) is an emerging technology that can support future global energy demand if it is ideally harvested. Even though many studies, suggestions, structures, and designs have been proposed, the commercialization of MFC remains unachieved due to persistent challenges such as low power output and high operational costs. Connecting two or more cells in series can multiply the total output power of the MFC. However, voltage reversal often occurs in weaker cells when the voltage drops below 0 V, leading to power loss and potential damage to the entire stack, especially under fuel starvation or imbalance in cell performance. It has been suggested that a continuous supply of anodic substrate can prevent fuel starvation and improper conditions. However, the method to find the weak cells and the amount of anodic substrate to add has not yet been discussed. In this study, for the first time, a Matrix Laboratory (MATLAB) model is proposed to identify the weak cell in a series-stacked MFC and determine the appropriate amount of substrate to add to the anodic chamber of that cell using an Artificial Neural Network (ANN). The model finds the weak cells and then modifies their Open Circuit Voltage (OCV) by adding or reducing the Total Dissolved Solids (TDS) of the Anodic chamber until all the cells reach their optimal condition. This approach helps balance the voltages between cells and prevents voltage reversal. The model works fine with OCV ranging from -1.1 V to 1.1 V, TDS 1193 ppm to 3370 ppm, Temperature 296 K to 302 K, and pH 4.66 to 8.23. The model was tested and validated using 144 sets of actual data, yielding a Root Mean Square Error (RMSE) of 8.05229 and a coefficient of determination (R²) of 0.98257.