278113 Nonlinear Control of a Water Gas Shift Membrane Reactor for Hydrogen Production
Hydrogen is a promising solution for carbon free energy production in fuel cells and gas turbines. The majority of the worldwide hydrogen production is based on fossil fuels (e.g., coal, natural gas and oil) while, a small but increasing percentage of hydrogen production is based on renewable energy resources (e.g., biomass). Light and heavy hydrocarbons generate hydrogen through steam reforming, while coal- or biomass-derived syngas (a mixture of mainly carbon monoxide and hydrogen) produces hydrogen through the water gas shift (WGS) reaction.
The need for high-purity hydrogen production via WGS typically requires a packed-bed reactor operating at high conversions followed by a separation unit for hydrogen purification. However, the merging of the reaction and separation processes into one unit, also known as process intensification, increases process efficiency and reduces costs, despite the higher process complexity. This can be achieved by using membrane reactors to replace both unit operations.
In addition to the advantage of process intensification, membrane reactors enable the continuous removal of hydrogen from the reaction zone through the membrane, leading to higher conversions when compared to conventional packed-bed reactors under the same operating conditions. However, higher conversions (which correspond to more released heat for exothermic reactions) may cause steeper temperature gradients (hot spots) along the membrane reactors than in conventional reactors. Therefore, a proper control strategy should be developed to ensure a safe operation at a predefined temperature range and preclude catalyst deactivation during the transition between different operating points and under the effect of unexpected disturbances.
In this study, a WGS membrane reactor is considered consisting of three different compartments: the reaction zone (tube), where the moderately exothermic WGS reaction takes place, the permeation zone (shell), where the produced hydrogen is collected and the cooling zone (inner tube). A distributed model for the WGS membrane reactor is developed using a hydrogen-selective (zeolite-based) membrane and a high-temperature catalyst (Fe/Cr – based). This membrane reactor achieved higher carbon monoxide conversions than the ones achieved by conventional packed-bed reactors. A steady-state analysis under different flow configurations revealed the presence of hot spots along the reactor. Open-loop simulations demonstrated the necessity for a control strategy not only to suppress the hot spot but also to maintain the operation of the reactor under a specified temperature. A nonlinear controller is derived on the basis of the PDE model of the process which is a distributed analogue of nonlinear inversion-based geometric control methods for nonlinear ODEs. To assess the performance of the nonlinear controller, a case study is performed where a large step in the tube inlet volumetric flow rate (representing an increase in H2 production) is imposed. The results revealed a superior performance of the nonlinear controller when compared to a linear PI controller in suppressing the hot spots and attaining an operating temperature under the maximum accepted temperature constraint.