Optimal Design of Sustainable Fischer-Tropsch Process Utilizing Micro-channel Technology

Abstract

Gas-to-liquid (GTL) has been considered a technology that converts natural gas into high value-added liquid fuel for decades. The produced fuel has less greenhouse gas emissions than the conventional gasoline or diesel after combustion, and the sulfur content is less than 0.5ppm. Therefore, it can be used as a clean fuel and environmentally friendly energy resource. GTL process can be divided into a reforming process for producing synthesis gas, which is a mixture of carbon monoxide and hydrogen, from natural gas, and a Fischer-Tropsch (FTS) process for synthesizing a hydrocarbon chain. The key process is the Fischer-Tropsch synthesis, which requires a structure to deliver an effective heat removal because of the highly exothermic reaction (ΔH = -165 kJ / mol). Generally, large scale commercial GTL processes are operated using a circulating fluidized bed reactor, a fluidized bed reactor, a multi-tubular fixed bed reactor, and a slurry column bed reactor. In recent years, research on the reactor development with microchannel technology has been highlighted because it has been pointed out that existing reactors are hard to be applied to marine conditions in the development of offshore plant and microchannel reactors take advantages of high economical efficiency for small and medium scale processes. Microchannel reactors can reduce the volume of existing reactors by a factor of 10 to 1000 times caused from reducing heat and mass transfer distances, increase the efficiency of the chemicals used in the process, make them environmentally friendly, easily control operation. By reducing the size of the reactor, space utilization can be widened through integration in the process, and productivity and process efficiency can be increased through modularization. In this thesis, optimal design of reactor with microchannel technology and economical structure of Fischer - Tropsch stand-alone process are obtained. The optimal design of the cooling layer distributor that is the key of the heat removal performance is derived. The optimal reactor design that can satisfy both the reaction safety and the miniaturization of the reactor core part is proposed. From the viewpoint of process, superstructure process modeling with microchannel reactor was carried out and optimization process was conducted to find the most economical process. The reactor model was validated using real reactor operation data. First, a simple trapezoidal-shaped guiding fin was optimized and applied to the manifold to ensure a uniform coolant flow to the cooling layer inside the stacked microchannel reactor. It was possible to achieve a stable distribution even over a large area of over 100 channels by using the principle of inducing mixing by appropriately transferring the refrigerant fluid introduced into the guiding fin to a space defined as a free mixing zone. Specifically, we performed optimization using the artificial neural network as a surrogate model for the structure of the guiding fin. Furthermore, we conducted a robustness test on the flow rate, fluid type, and operating temperature for the optimal structure. As a result, 500 ≤ Re GF ≤ 10800 Uniformity of the distribution could be maintained in a considerably large area. Next, the reactor core was modeled by introducing a cell-coupling method, and multi-objective optimization was performed on seven design variables with the maximum reaction temperature rise and reactor core volume as an objective function. The maximum reaction temperature rise is related to keeping isothermal condition (anti-hotspot) of the reactor and the reactor core volume is directly related to reactor compactness. As a result of the optimization, the Pareto optimal points can be obtained. As the maximum temperature rise increases, the reactor length becomes short and the width increases, and the height is generally constant. These optimization procedure can be used to determine two factors. First, the reasonable coolant flowrate can be obtained from the comparison among the optimal Pareto point set with the sensitivity analysis. Furthermore, it was possible to prioritize the design factors related with reactor durability. Finally, a superstructure FTS stand-alone process model with various processes such as single or multistage FTS, recycle and water gas shift reaction, which has been studied extensively in order to improve the process efficiency, is optimized to maximize the profitability. The optimal design was obtained by introducing Genetic Algorithm. As a result of further sensitivity analysis based on changes in raw material costs and product prices by deriving additional two representative systems (multi-stage and single-stage without recycle process), a single stage with recycle is absolutely superior in all cases. This study has a great contribution to the design and operation of an exothermic reactor using microchannel technology and a Fischer - Tropsch process. The both of optimal design procedures of the cooling layer with high robustness on uniform distribution and the reactor miniaturization with stable operation deliver great value in designing and operating the optimal microchannel reactor design. In addition, the proposed design procedure for economical Fischer Tropsch system can be used as a process design package with the reactor design methodology.