System characterisation of a 1 MW advanced dual fluidised bed steam gasification demonstration plant by a design of experiments approach

• The study shows the first application of DoE methodology in demonstration scale advanced dual fluidised bed gasification. • ·The concrete influence of four main operating parameters on the process is quantified and interactions are discussed. • The resulting system maps describe the relevant output parameters for downstream Fischer-Tropsch synthesis. • From the analysis, two operating points for optimised Fischer-Tropsch synthesis-suited product gas generation are deduced. The advanced dual fluidised bed steam gasification process can provide a versatile product gas from biomass residues and waste materials, applicable for different downstream syntheses processes. In Austria, current research utilises a demonstration scale plant to investigate fuel flexibility and optimise key performance parameters in an industrial environment. However, the dual reactor principle is a system with many potential interactions between its operating parameters which are difficult to distinctly evaluate. The design of experiments method as a statistical approach to experimental investigation can help to identify and quantify these interactions, as shown in multiple fields and industries. With the use of this method, a two-week experimental campaign (111 h) with the demonstration plant was planned and executed. In total, 19 individual operating points were investigated, and the resulting data was analysed to evaluate suitable settings for the generation of a product gas suited for downstream Fischer-Tropsch synthesis, with the aim of increasing conversion efficiency and product gas quality. Furthermore, results should provide plant operates with easy-to-understand instructions for plant operation with respect to specific in- and outputs. The data evaluation led to a first rudimentary system mapping of the reactor system, over a wide range of four main operating parameters. This first ever application of statistical methods in this process’ scale shows that the approach is suited for this technology. The presented results provide a base for further system analysis and help future experimental campaigns and the ongoing technology optimisation process.