Genetic algorithms and neural networks in optimization of sorbent enhanced H2 production in FB and CFB gasifiers

The paper introduces artificial intelligence (AI) approach for the optimization study of a hydrogen concentrations in syngas via CaO Sorption. The performed model allows estimating the hydrogen content in the syngas produced from biomass in different types of facilities. Bubbling fluidized bed (FB) and circulating fluidized bed (CFB) are taken into account in the study. Comparing with the previous FB gasification results, CFB gasification was capable of producing syngas with higher hydrogen concentration. The model considers and covers a broad range of conditions, influencing the hydrogen rich-gas production. The presented non-iterative approach gives quick and accurate results as an answer to the input data sets. The H-2 concentration in the gas, estimated using the developed model, is in good agreement with the experimental data. Maximum relative error between measured and calculated data is lower than +/- 8%. The model allows also studying the influence of operating parameters on the hydrogen concentration in the gas. The method constitutes an easy to employ and useful complementary technique in relation to the other ways of data handling, including experimental procedures. The model can be used by scientists and engineers for optimizations purposes and can be applied as a submodel or a separate module in engineering calculations, capable to predict the H-2 concentration in the syngas from biomass via the CaO sorption both in FB and CFB gasifiers.