Modelling of Multi-Tubular Fixed-Bed Reactor for Fischer-Tropsch Synthesis to Produce Synthetic Crude Using Syngas Obtained from the Work’s Arising Gases of an Integrated Steel Mill

Abstract

Steelmaking process is a highly carbon-intensive process. This is mainly due to the use of coke as a reducing agent in the blast furnaces to produce carbon-rich pig iron, which in turn, is used for the production of low-carbon steel in the basic oxygen furnaces. The exhaust gases from the blast and basic oxygen furnaces, which mainly contain CO and CO2, are utilised for electricity generation, and thus, these pollutants are released to the atmosphere. One of the possible ways to treat these work’s arising gases (WAGs) is to convert them into syngas, which can then be further converted into syncrude via Fischer-Tropsch synthesis (FTS). The FTS syncrude can then be further refined and processed to produce liquid fuels such as gasoline, kerosene, diesel, etc. These synthetic fuels are sulphur-lean and are essentially capable of replacing the existing fossil-derived liquid fuels, thus contributing to curbing the carbon emissions.

The main objective of this thesis was to develop a detailed model of a multi-tubular fixed-bed reactor (MTFBR) to produce synthetic crude from syngas via Fischer-Tropsch synthesis (FTS). The model was then used to simulate a reactor that utilises the syngas obtained from the processing of work’s arising gases of an integrated steel mill to produce synthetic crude. The FTS product distribution was modelled using the kinetic model based on CO-insertion mechanism, developed by Todic et al. A basic MTFBR model was initially developed using the equations and assumptions from the fixed-bed reactor model of Todic, and the basic MTFBR model was able to produce similar results as that of the Todic’s model, with slight deviations in the temperature and pressures profiles. The basic MTFBR model was then further improved to render it comparable with the commercial FT reactors. The main improvements in the model include the dynamic extraction of thermodynamic and transport properties of the system components using Aspen Properties; calculation of dynamic vapour-liquid equilibrium, liquid holdup and catalyst effectiveness factor; and the use of improved heat transfer and pressure drop equations.

A sensitivity analysis was performed to determine the effect of design and process parameters on the performance of the MTFBR model. The most crucial design parameter was observed to be the tube diameter as it had a considerable effect on the heat management and the pressure drop in the reactor bed. The most important process parameters for the reactor were observed to be the inlet temperature and the feed flow rate. A simplified FTS gas loop process was also modelled in Aspen Plus in order to introduce a recycle stream into the MTFBR. The effect of tail gas recycle for the recovery of unreacted H2 and CO was also studied, and it was observed that higher recycle ratios resulted in lower conversions per pass; however, overall CO conversions were observed to increase until a maximum, and then decrease thereafter. The optimum conditions for the simplified gas loop process were estimated to be with an inlet temperature of 484.5K and a total recycle of tail gas to the recovery section, for a inlet pressure of 30 bar. Optimised process conditions resulted in a CO conversion per pass of 46%, an overall CO conversion of 89%, a C5+ selectivity of 86.6%, a CH4 selectivity of 6.2%, and a C5+ productivity of 252,540 tonnes/y. The optimised model results, in terms of C5+ selectivity and overall CO conversion, were also pretty much inline with the available data from the Shell SMDS plant in Bintulu.