Global sensitivity and uncertainty analysis of a Fischer-Tropsch based Power-to-Liquid process

Abstract

A workflow to conduct a combined uncertainty and global sensitivity analysis supplementing techno-economic process analysis is depicted and applied in this work. Furthermore, different estimators for conducting the variance-based global sensitivity analysis are compared with respect to their computational effort. The model under investigation is a Fischer-Tropsch based Power-to-Liquid process that allows to produce syncrude from electricity, water and carbon dioxide. One of the major contributors to the uncertainty of the net production cost is found to be the full load hours of the electrolyzer. Reducing the full load hours of the electrolyzer reduces the cost for electricity, as the average wholesale electricity price decreases assuming that the electrolyzer is operated at the cheaper hours within the year. However, it also results in increased storage cost and cost for oversizing of the electrolyzer. The results indicate, that the reduced costs for electricity do not outweigh the additional costs for storage and electrolyzer oversizing. Thus, the lowest net production cost is observed at maximum full load hours of the electrolyzer. The base case syncrude net production cost for the plant situated in Germany is 3.89 €2019 /kgC5+ (5.47 €2019/kgH2). The resulting 95 % confidence interval, after fixing the full load hours of the electrolyzer to 8 000 h/a, is between 2.79 and 5.29 €2019/kgC5+.