Modeling, estimation, and control of proton exchange membrane-based electrochemical systems

To reduce emissions and meet the rapidly growing global energy demand, affordable and efficient methods of electrical energy storage and generation are needed to exploit renewable energy sources more effectively. Proton exchange membrane (PEM) based electrochemical systems, such as vanadium redox flow batteries (VRFB) and PEM fuel cells, are playing an increasingly important role because they have a fast response rate, high efficiency, and small environmental impact. However, widespread commercial viability of these technologies in the future heavily depends on further improvements in their performance and reliability. Accordingly, this dissertation focuses on developing new methodologies to predict and control the behavior of these PEM-based electrochemical systems. In the first part of this work, a control-oriented physics-based model of a VRFB system is developed. This model can predict the transient response of the cell voltage under different operating conditions and inputs such as current, flow rate, and temperature. The significance of this study is having the ability to predict the short and long term effects of membrane crossover on the system performance. One major challenge of operating VRFB systems is that monitoring the state-of-charge (SOC) in real-time using traditional measurement techniques is expensive and impractical. To address this problem, an SOC estimator is developed based on a constrained extended Kalman filter that can be used for real-time optimization and control because it requires only simple voltage measurements and a low-order model. Simulation results demonstrate the ability to predict the vanadium concentrations of a VRFB system without knowledge of the crossover dynamics. A major obstacle preventing the widespread commercialization of VRFBs is excessive capital costs. This issue is addressed by developing a methodology to optimally size a VRFB system using the minimum amount of materials required for the intended power range. For PEM fuel cells, proper water and thermal management is critical to optimizing performance and longevity. However, this can be a challenging task due to strong system interactions between multiple input and output variables. In the final part of this work, these system interactions are studied in detail and a suitable controller is designed to regulate the stack voltage, stack temperature, and relative humidity during load transients.