From property uncertainties to process simulation uncertainties – Monte Carlo methods in SimSci PRO/II process simulator

Abstract

Abstract This study presents a methodology to apply Monte Carlo methods for property uncertainty propagation in the process simulation software SimSci PRO/II. The aim of this work is to integrate advanced uncertainty and sensitivity analysis tools into commercial process simulators. The property uncertainty and sensitivity analysis tools were applied to a heat pump system with cyclopentane as a working fluid. Monte Carlo sampling technique was used to generate property samples of the SRK equation of state parameters critical temperature, critical pressure and acentric factor. The samples were subsequently evaluated in the heat pump flowsheet built in SimSci PRO/II. This allowed describing the process model output uncertainty in a distribution and with the 95% confidence interval. Furthermore, Monte Carlo based standard regression could be used to analyse the sensitivity of the respective fluid properties. The results showed that property uncertainty propagation strongly depends on the correlation between the property parameters. The sensitivity analysis showed that the acentric factor is the most sensitive SRK parameter. This works demonstrates that Monte Carlo methods are a simple and useful tool, which can be used in commercial process simulators by industrial users.