Performance correlations for biomass gasifiers using semi-equilibrium non-stoichiometric thermodynamic models

Biomass gasification has been a viable alternative for decentralized electricity generation in developing countries. The efficiency of the biomass gasification process for operation of the engine-generator set is mapped in terms of quality and quantity of the producer gas. In this study, we have attempted to devise generalized correlations for four principal parameters that form the benchmark for the performance of the gasifier. These parameters are lower heating value and net yield (per unit biomass) of producer gas, and volume fractions of CO and CO2 in the gas resulting from biomass gasification process. The correlations have been constituted using simulations of gasification of three common biomass feedstocks (viz. rice husk, saw dust and corn cobs) using semi-equilibrium non-stoichiometric thermodynamic model. The independent variables used in the simulations are air ratio, carbon conversion, gasification temperature and three elemental ratios in the gasification mixture, viz. H/C, O/H and O/C. As many as eight expressions of linear and non-linear type have been evaluated to best fit the simulations data for each performance parameter. On the basis of statistical indicators, the compatibility of the correlations for best fit of the data has been assessed. Finally, the predictions of the correlation have been tested against experimental data on gasification of different biomass. The best correlation for each performance parameter was chosen on the basis of least average absolute error and highest (absolute) regression coefficient. It was found that the set of best correlations could predict the values of performance parameters within engineering accuracy of +/- 1020%. The correlations proposed in this work are independent of the type of biomass gasifier. These correlations can form a useful tool for design and optimization of fixed or fluidized bed gasifier for any biomass feedstock. Copyright (c) 2011 John Wiley & Sons, Ltd.