A detailed model deduction and parameter values for the Tennessee-Eastman benchmark

Abstract This paper presents a detailed deduction of the model introduced in the recognized Tennessee-Eastman (TE) benchmark process. The purpose is to show the majority of process operation considerations and the not so obvious assumptions applied by Downs and Vogel during the model statement in their seminal paper. This objective is achieved by providing the details of the deduction of a phenomenological-based semi-physical model (PBSM) for the TE plant. This implemented modelling approach is completely described whereas the equations are derived individually for each relevant piece of equipment (reactor, condenser-flash separator, stripping tower, and mixing point). Then, such equations are integrated to construct the final model of the plant. It was feasible to simulate the four fundamental modes of plant operation, achieving alignment with the steady-state values documented in prior research. Furthermore, the values of additional parameters that were not previously available in the literature were determined, providing other researchers with the opportunity to readily simulate those operation modes of the plant. Finally, some relevant simulation results are presented, including the system responses to disturbances, along with a discussion about determining the process initial conditions.