Fuzzy Logic-based Model for Prediction of CS2 and COS Concentrations in the Modified Claus Process

Chemical processes can be very complex and difficult to model with the traditional mathematical tools that are available. Fuzzy set theory offers a variety of alternative tools that can be used to approximately describe the behaviour of such complex systems with relatively little computational effort. The modified Claus process is one of the most used processes in the refining and gas processing industries for conversion of H2S to elemental sulphur and H2O. It is known that hydrocarbons, CO and CO2 are involved in the formation of COS and CS2. However, the reactions are not well known and the attempts to model their behaviour have resulted in very complex models that do not predict the concentrations accurately. A fuzzy logic-based (rule based) model to predict the concentration of COS and CS2 at the exit of the Waste Heat Boiler (WHB) in the modified Claus process was developed. The model considers the inlet composition of H2S, CO2 and CH4, and the residence time at a fixed temperature as the input variables that affect the process. The output variables of the model are therefore the concentration of COS and that of CS2. The model was validated with experimental data and it performs as well, if not better than the other more elaborate and more complex models available for this system.