Global sensitivity analysis to identify influential model input on thermal risk parameters: To cottonseed oil epoxidation

Abstract

Thermal runaway is still an important cause of accident in chemical industry. To evaluate the risk of such events, thermal risk assessment, which is a part of process safety, must be done. This assessment determines the safe operating conditions of a process by evaluating the thermal risk of an exothermic system. Nevertheless, based on thermal risk assessment, it is not possible to know which model inputs have the most influence on the thermal risk parameters. The knowledge of the most influential model inputs on thermal risk parameters is important to establish adequate safety barriers. Global sensitivity analysis was used to evaluate the influence of model inputs and their interaction on thermal risk parameters. It was performed on the exothermic system: epoxidation of cottonseed oil by performic acid in semibatch mode under isoperibolic conditions. The maximum reaction temperature and the time to reach this maximum reaction temperature were chosen as thermal risk parameters. We have also studied the influence of model inputs on the temperature rise. In the operating conditions of this study, it appeared that two parameters have the most influence on maximum reaction temperature and the temperature rise: the initial concentration of hydrogen peroxide and the jacket temperature, and one parameter for the time to reach this temperature: the jacket temperature.