Experimental design and statistical evaluation of a full-scale gasification project

Abstract Sasol in South Africa gasifies approximately 30 million tons of bituminous coal per year to synthesis gas, which is converted to fuels and chemicals via the Fischer–Tropsch process. Three years ago, Sasol embarked upon a unique project to optimise the gas production of the Sasol–Lurgi fixed bed dry bottom coal gasification process. Optimisation of the gasification process was carried out by utilising the method of factorial experimental design on the process variables of interest obtained from a specifically equipped full-scale test gasifier. A response surface model in the process variables was fitted for each of the performance variables. The models developed were validated and proved to be valid when considering different data sets. This paper discusses the application of factorial experimental design and the combination of process variables and mixture components in optimising a production process.