Membrane reactor performances in H2 recovery under conditions relevant for the HCPB tritium extraction system

Abstract

The Tritium produced in the HCPB blanket is released in both elemental and oxidized form (the HT/HTO release rate is about 97/3). The reference process for removing the produced tritium relies on the use of Reactive Molecular Sieve Beds (RMSBs) and Cryogenic Molecular Sieve Beds (CMSBs). The open issues of such process are related to the regeneration procedure to be adopted in the RMSBs and on the suitable technique required for treating the gas stream leaving the RMSBs during the regeneration phase, which contains the released Q2O mixed with the He and H2 stream used for flushing the sieving beds. Membrane reactors represent a valid option due to their ability in performing Water Gas Shift (WGS) reaction and hydrogen separation directly in one device.In this study, the performances of a membrane reactor (a Pd/Ag membrane filled with 16 g of a commercial catalyst named ShiftMax® 120 HCF) in H2 recovery via Water Gas Shift reaction have been explored by performing both experimental tests and simulation analysis. Particularly two sets of experiments, differing solely in the H2O amount fed in the membrane reactor (5 and 10 g h-1), have been performed. In both, the H2O/CO feeding ratio was equal to 1 and four different conditions were investigated: i) no addition of H2 and He in the WGS feeding stream; ii) addition of H2 to the WGS feeding stream at several H2/H2O feed molar ratio; iii) addition of H2 and He to the WGS feeding stream at several H2/H2O and He/H2O feed molar ratio, iv) addition of He to the WGS feeding stream at several He/H2O feed molar ratio. The experimental results have been used to validate a novel simulation tool and in most of the cases a good agreement have been observed among the experimental and the simulation results.