Start-up strategies of an experimental fuel processor

In this work, cold start-up of a methane fuel processor is explored. The experimental fuel processor is intended to provide hydrogen for a proton exchange membrane (PEM) fuel cell for the power generation (3 kWe). A dynamic model describing a series of reactors, the reformer, three water–gas shift reactors, and preferential reactor is constructed. Two important factors for rapid start-up are identified: speed of temperature front propagation and acceptable CO concentration. Steady-state analyses reveal that the fuel feed flow rate with fixed steam-to-carbon and air-to-carbon ratios is an ideal manipulated variable. Considering both large initial heat flux and gradual transition back to nominal operation, the shape of feed manipulation is determined. With the feed scenario available, the fuel processor start-up can be formulated as a constrained optimization problem and can be solved numerically. From optimization result, a heuristic is generated for rapid start-up of a fuel processor. This leads to a 25% improvement in the start-up time. Finally, issues of design modification are explored for further reduction in the start-up time.