Control-oriented modeling of dynamic thermal behavior and two‒phase fluid flow in porous media for PEM fuel cells

The driving force behind research in alternative clean and renewable energy has been the desire to reduce emissions and dependence on fossil fuels. In the United States, ground vehicles account for 30% of total carbon emission, and significantly contribute to other harmful emissions. This issue causes environmental concerns and threat to human health. On the other hand, the demand on fossil fuel grows with the increasing energy consumption worldwide. Particularly in the United States of America, transportation absorbs 75% of this energy source. There is an urgent need to reduce the transportation dependence on fossil fuel for the purpose of national security. Polymer electrolyte membrane (PEM) fuel cells are strong potential candidates to replace the traditional combustion engines. Even though research effort has transferred the fuel cell technology into real‒world vehicle applications, there are still several challenges hindering the fuel cell technology commercialization, such as hydrogen supply infrastructure, cost of the fuel cell vehicles, on‒board hydrogen storage, public acceptance, and more importantly the performance, durability, and reliability of the PEM fuel cell vehicles themselves. One of the key factors that affect the fuel cell performance and life is the run‒time thermal and water management. The temperature directly affects the humidification of the fuel cell stack and plays a critical role in avoiding liquid water flooding as well as membrane dehydration which affect the performance and long term reliability. There are many models exists in the literature. However, there are still lacks of control‒oriented modeling techniques that describe the coupled heat and mass transfer dynamics, and experimental validation is rarely performed for these models. In order to establish an in‒depth understanding and enable control design to achieve optimal performance in real‒time, this research has explored modeling techniques to describe the coupled heat and mass transfer dynamics inside a PEM fuel cell. This dissertation is to report our findings on modeling the temperature dynamics of the gas and liquid flow in the porous media for the purpose of control development. The developed thermal model captures the temperature dynamics without using much computation power commonly found in CFD models. The model results agree very well with the experimental validation of a 1.5 kW fuel cell stack after calibrations. Relative gain array (RGA) was performed to investigate the coupling between inputs and outputs and to explore the possibility of using a single‒input single‒output (SISO) control scheme for this multi‒input multi‒output (MIMO) system. The RGA analyses showed that SISO control design would be effective for controlling the fuel cell stack alone. Adding auxiliary components to the fuel cell stack, such as compressor to supply the pressurized air, requires a MIMO control framework. The developed model of describing water transport in porous media improves the modeling accuracy by adding catalyst layers and utilizing an empirically derived capillary pressure model. Comparing with other control‒oriented models in the literature, the developed model improves accuracy and provides more insights of the liquid water transport during transient response.