Re-design of the Tennessee Eastman Challenge Process: an Eigenvalue Optimization Approach

The Tennessee Eastman Challenge Process (TECP) represents an interesting case study within PSE for process control and process optimization purposes. It has been widely addressed by the chemical engineering research community since its publication. The TECP is an open-loop unstable recycle reactor. The eigenvalues of the Jacobean matrix of the system at the base-case steady-state conditions, range from –1968 to +3.07. In absence of feed-back, small perturbations cause large transients reaching shut-down limits within an hour. Operation at open-loop unstable steady-states, even if feasible with feed-back control, is undesirable and should be avoided by proper design. In this contribution, the redesign problem of the TECP is addressed in order to generate an open-loop stable operating condition, by making use of eigenvalue optimization techniques. The objective is to find a new steady state point, such that the whole spectrum of the Jacobean matrix of the system lies in the left half of the complex space. Such a problem corresponds to the design-for-operability discipline, of outstanding importance within the PSE research.