International Journal of Hydrogen Energy, Vol.44, No.9, 4508-4523, 2019
Numerical modeling of an automotive derivative polymer electrolyte membrane fuel cell cogeneration system with selective membranes
Cogeneration power plants based on fuel cells are a promising technology to produce electric and thermal energy with reduced costs and environmental impact. The most mature fuel cell technology for this kind of applications are polymer electrolyte membrane fuel cells, which require high-purity hydrogen. The most common and least expensive way to produce hydrogen within today's energy infrastructure is steam reforming of natural gas. Such a process produces a syngas rich in hydrogen that has to be purified to be properly used in low temperature fuel cells. However, the hydrogen production and purification processes strongly affect the performance, the cost, and the complexity of the energy system. Purification is usually performed through pressure swing adsorption, which is a semi batch process that increases the plant complexity and incorporates a substantial efficiency penalty. A promising alternative option for hydrogen purification is the use of selective metal membranes that can be integrated in the reactors of the fuel processing plant. Such a membrane separation may improve the thereto-chemical performance of the energy system, while reducing the power plant complexity, and potentially its cost. Herein, we perform a technical analysis, through thermo-chemical models, to evaluate the integration of Pd-based H-2-selective membranes in different sections of the fuel processing plant: (i) steam reforming reactor, (ii) water gas shift reactor, (iii) at the outlet of the fuel processor as a separator device. The results show that a drastic fuel processing plant simplification is achievable by integrating the Pd-membranes in the water gas shift and reforming reactors. Moreover, the natural gas reforming membrane reactor yields significant efficiency improvements. (C) 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.