Investigating the impact of modeling assumptions and closure models on the steady-state prediction of solar-driven methane steam reforming in a porous reactor

Abstract

In this work, the effect of different modeling assumptions and closure models — frequently considered in the literature of solar thermochemical reactors and closely-related areas — is investigated. The application of different modeling assumptions and closure models may strongly affect the results accuracy and compromise a meaningful comparison across literature works. The following is herein investigated: (i) the role of upstream and downstream fluid regions to the two-phase reactor region; (ii) relevance of species and gas-phase thermal diffusion mechanisms; (iii) reaction heat accounted for in gas- or solid-phase energy balance equations; (iv) local thermal equilibrium (LTE) vs. local thermal non-equilibrium (LTNE) models; (v) model dimension (one-dimensional vs. two-dimensional axisymmetric models); (vi) local volumetric convection heat transfer correlations; and (vii) effective solid thermal conductivity correlations. The relevance of this work extends well beyond the current application (methane steam reforming in a volumetric solar reactor), since similar models and assumptions have also been widely applied for predicting the performance of volumetric solar absorbers and dry reforming solar reactors. The results show that an upstream fluid region should be considered while applying inlet first-type boundary conditions and diffusion transport in the corresponding governing equations to ensure full-conservation and simultaneously to account for developing profiles upstream the reactor inlet section. For the operating conditions considered, species diffusion and gas-phase heat conduction are particularly relevant at the reactor centerline but with a negligible integral effect. A significant difference in the reactor performance is observed while accounting for the reaction heat from surface reactions in the solid- or gas-phase energy balances — for the lowest inlet gas velocity herein considered, the thermochemical efficiency (methane conversion) is approximately equal to 70.8% (76.3%) and 77.3% (84.7%) assigning the reaction heat to the gas- and solid-phase energy balances, respectively. LTE model results are strikingly different from the results obtained with the LTNE model considering the reaction heat accounted for in the gas-phase energy balance but not significantly different from the results computed with the LTNE model with the reaction heat assigned to the solid-phase energy balance. One-dimensional reactor modeling provided with an average concentrated solar heat flux value results in a similar average performance as that given by a two-dimensional model but fails to predict the high temperatures observed at reactor centerline — for the lowest inlet gas velocity, the difference between the maximum solid temperatures predicted by one- and two-dimensional models is about 340K.