Non-intrusive Polynomial Chaos Expansion Based Uncertainty Analysis of Bioethanol Production Process

Ethanol production has been a topic of great interest since the world found the significance of renewable biofuels. Process efficiency and sustainability are main points of focus for ethanol production. Uncertainty in input parameters and its effect on the process outcome, i.e., ethanol production, has been a challenge in realizing efficient operations for the process. This study aims quantification of the effect of uncertainty in process inputs on the production of ethanol. The study is based on an Aspen PLUS (R) lowsheet of ethanol production from corn stover. MATLAB (R)-Excel (R)-Aspen (R) interfacing is used to estimate ethanol production for different values of input variables. The data generated through the interfacing was used to develop a data-driven model. The data-driven model based on the idea of ensemble learning was used within a Polynomial Chaos Expansion to quantify the accumulative effect of uncertainty in process input variable on process output, i.e., ethanol production.