(596e) Dynamic Modeling and Model-Based Control Strategy for Dimethyl-Ether (DME) Production Reactor

Dimethyl ether (DME) is considered as a clean multi-source energy resource because it can be synthesized from low-grade coal, natural gas, or biomass without producing side products such as sulfur compounds and aromatic impurities. It has the potential to replace liquefied natural gas (LNG) or diesel fuel in the near future. DME production process consists of two major steps; the first step is methanol synthesis from syngas and the second is methanol dehydration to produce DME. Significant obstacles exist in commercialization of the direct DME synthesis since combining two highly exothermic equilibrium reactions makes separation and online measurements more complex. Therefore better understanding on the reactor is requisite for performance improvement by optimal operation and catalyst design.

 Due to the complexity in the DME synthesis models, most studies have been mainly focused on the yield estimation and optimization at steady state. In this work, we present a dynamic, distributed parameter model for synthesis DME reactor, a parameter estimation scheme, and a model-based optimal control strategy by adopting existing steady-state reaction rate models into the mass balance equation.

We combine the Vanden Bussche model for methanol formation reaction [1] and the Bercic model for methanol dehydration reaction [2]. The Bercic model is employed to include the effect of concentrations of non-methanol components in calculation of methanol dehydration rate. The model equations have highly complex polynomial structures because of the assumptions of Langmuir adsorption isotherm and existence of reversible reactions. To reduce the complexities and stiffness of the combined model, we simplify the modeling equations as an isothermal fixed-bed shell-and-tube reactor model. The resulting model equation is reduced to a first-order partial differential equation. The finite difference method is used to conduct sensitivity analysis between the manipulated inputs and the DME yield. The inputs are reactor temperature and feed flowrate. A model-based optimal control strategy will also be presented to propose an operation strategy to improve the DME yield while maintaining the process stability. This study is conducted with the support of Samsung Heavy Industries.

References:

[1]        K. Bussche and G. Froment, "A Steady-State Kinetic Model for Methanol Synthesis and the Water Gas Shift Reaction on a Commercial Cu/ZnO/Al2O3Catalyst," Journal of Catalysis, vol. 161, pp. 1-10, 1996.

[2]        G. Bercic and J. Levec, "Intrinsic and global reaction rate of methanol dehydration over. gamma.-alumina pellets," Industrial & engineering chemistry research, vol. 31, pp. 1035-1040, 1992.