A two-layer capacitor-based power management system for a stack of microbial fuel cells

Abstract

While the penetration rate of renewable and sustainable energies is growing daily, the demand for newer sources and more efficient energy harvesting has been growing. Microbial fuel cells (MFC), as a sustainable energy source, have gained a lot of attention recently from scientists in different fields. This alternative source of energy can use any organic matter including urban or industrial wastewater as fuel to produce energy and does not emit any greenhouse gas as a byproduct. Given the low power density of MFCs, it is very important to design efficient and adaptive power management systems (PMS). These systems will produce usable power and voltage for load while managing high fluctuations of the MFCs. This study aims to design a new PMS that will increase the mean output power of the stack of MFCs by detecting faulty and lowperforming MFC(s) and disconnecting them from the stack. In this approach mean power output of various prospective actions is estimated and the action that results in higher output power is selected by the PMS. Although removing MFC units from the stack causes small potential losses over a brief period after disconnection, it is important to evaluate the PMS in long-term operation. To control the frequency, accuracy, and robustness of the PMS and avoid potential losses, a hysteresis control method, a simple moving average approach and a cooldown timer have been implemented. These additional control methods help the PMS to adjust its tolerance to low-performing MFCs, base its decision on multiple cycles to eliminate the effect of noise in measurement and avoid repetitive connections and disconnections. The harvested energy is then sent to a boost converter to deliver a usable voltage to the load. This boost converter will help the PMS to increase the magnitude of the output voltage and regulate the voltage using a proportional integral (PI) controller. To evaluate the performance of this PMS, more than 100 measurements were taken from four MFCs in the lab and two test cases were constructed to replicate their behavior by a simple electrical equivalent model based on the measurements. These measurements were taken in different ambient and operational conditions, so the test cases can represent a wide variety of MFC behavior. This testing framework was implemented in MATLAB Simulink® and the proposed PMS was tested under this testing framework. Although the proposed PMS can theoretically manage an infinite number of MFCs in a stack, for practical reasons a stack of four MFCs was chosen as an example. The results show that the proposed PMS is able to successfully detect and disconnect faulty and low-performing MFC from the stack, is resistant to electrical noise and disturbances and manages to increase the mean output power compared to a similar PMS proposed by a former study (Nguyen et al., 2019). The proposed PMS presents an increase in mean output power of up to 67% without voltage regulation and an increase of up to 23% with voltage regulation compared to the former study (Nguyen et al., 2019).

Date	18 Nov 2022
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Lyne Woodward (Supervisor)