Mapping and optimization of an industrial steam methane reformer by the design of experiments (DOE)

Natural gas is an important precursor to produce synthesis gas through the steam reforming process. This work aims to map the optimal operating conditions of a steam methane reformer based on the design of experiments (DOE) approach and response surface methodology (RSM) relying on a phenomenological model validated with industrial data. A screening factorial design 2IV8−4is initially conducted to identify the most significant process variables, followed by a 25 full factorial design for a more accurate analysis of the main and interaction effects of the input variables on the methane conversion and syngas quality. The most significant variables are the steam to carbon ratio, the flow rate of the stream fed to the reforming tubes, and feed gas and combustion air temperatures. A Box-Behnken design (BBD) with 25 experiments is carried out and polynomial models are proposed and validated through the analysis of variance with great confidence. An optimization problem is formulated to maximize the methane conversion, ensure the quality, and tube wall temperature constraints. This problem is solved in Excel, using the generalized reduced gradients method and the calculated optimal point is analyzed through contour plots.