Principal component analysis and partial least square regression models to understand sorption-enhanced biomass gasification

Abstract Gasification represents a potential technology for the conversion of biomass into usable energy. The influence main gasification parameters, i.e. type used and its composition, as well composition outlet gas, were studied by multivariate statistical analysis based on principal component (PCA) partial least square (PLS) regression models in order to identify correlations between them contents methane, ethylene tar gas. In this work, experimental data input came from TRL-4 plant running under sorption enhanced conditions, using steam gasifying agent CaO bed material. feed played an important role quality gas composition. fact, biomasses with high ash sulphur (municipal solid waste) increased content, while those high-volatile matter content fixed C (wood pellets, straw pellets grape seeds) mainly CO 2 formation. By increasing temperature CaO/C ratio, it was possible reduce methane collected Other light hydrocarbons could also be reduced controlling T reactor FB. Methane, modelled, cross-validated tested new set samples PLS obtaining results average overall error 8 26%. statistically significant variables predict positively associated thermal negatively ratio. remarkable both variables, mentioned PCA analysis. As far which is undesirable all processes, decrease favoured temperature, low content. produce adequate (e.g. content), compromise should found balance sorbent-to-mass ultimate proximate analyses feed. Graphical abstract