Process intensification in hydrogen production via membrane-based reactive separations

Our results show that the empirical model and the so-called direct oxidation microkinetic model display the best fits to our experimental data. ❧ We chose the CMS membrane for our MR process due to its relatively low fabrication cost and its robustness under our reaction conditions. In our research we have investigated the pure and binary/multi-mixed gas permeabilities of the CMS membrane for the common gases present in coal-derived and biomass-derived syngas.

The results show that the CMS membranes show significant selectivity between H₂ and other slower gases, thus making the CMS membrane an excellent candidate for hydrogen separation in the WGS-MR process The transport of fast gases like H₂ and He obeys an Arrhenius-type equation, and we have calculated the activation energies and pre-exponential factors by fitting the experimental data. For slow gases we have shown that they transport via Knudsen diffusion, with their permeances showing a 1/T0.5 dependence.

During the transport of mixtures, the permeances of the fast gases initially drop in the presence of slower gases in the mixture, whereas the permeances of the slow gases do not decrease notably in the presence of faster gases. ❧ We have used the sour-shift catalyst and CMS membrane in a WGS-MR system and have shown that the MR displays higher conversion than a conventional PBR setup for temperatures up to 300° and pressures up to 200 psig.

In some of the MR experiments, steam has been used as a sweep stream in the permeate side in order to increase the conversion and to improve the hydrogen purity and hydrogen recovery. We have also developed an isothermal 1-D model and have used the kinetic model generated (see Chapter 2) to describe the behavior of the MR with different sweep ratios and temperatures up to 300 ℃ and pressures up to 200 psig. ❧ In Chapter 5, we developed an adiabatic 1-D model for both the PBR and MR, and we showed how the temperature rise along the length of the PBR and MR can reduce the conversions obtained from these reactors.

Additionally, our adiabatic MR models have demonstrated how the use of cool steam as a sweep can help keep the membrane temperature within the safe operational temperature range. ❧ We have also developed an isothermal 2-D model for the MR in order to investigate the radial concentration profiles and possible mass transfer limitations of gases (especially hydrogen) to the surface of membrane. Our 2-D model shows that the radial concentration profiles are negligible for the lab-scale reactor.

We also have shown that the final conversion of the MR obtained from the 1-D model is virtually the same as that from the 2-D model. ❧ Additionally, we have introduced the multistage reactor as an alternative option to the MR and have explained some of the advantages of the multistage reactor over the MR. We have shown that at certain flow rates (with the same initial operating conditions, the same weight of catalyst, and the same membrane surface area), the multistage reactor can give higher conversions than the MR.