| A study was carried out to analyze the validity of a mathematical model previously developed to simulate the methanol steam reforming operation carried out in a membrane reactor, with a focus on the generation of ultra-pure hydrogen as a permeated product stream. A commercially available reactor featuring dense, unsupported Pd-23% Ag membranes was operated over a temperature range of 250 to 375°C, with pressures ranging from 50 to 300 psig. Elements of the model were addressed individually and a sequential validation process was adopted, progressing from analysis of membrane permeation, variation in reaction-side hydrogen partial pressure, effects of inert particle presence, effects of reformate components on permeation and, finally, results of catalyst conditioning and operation in the presence of a pure hydrogen stream. Data obtained from the reactor were compared with those predicted by the mathematical model and discrepancies were noted and analyzed. Elements of the model were corrected and/or improved based on the knowledge gained from these experiments. A complete model of the overall steam reforming reaction was not validated as several factors, including damage to the membrane tubes, prevented operation under realistic reforming conditions. Analysis of experimental data allowed for considerable progress towards a more accurate and practical model of steam reforming in a membrane reactor as well as the hydrogen permeation characteristics of the resident dense metal membranes. |
