Varicol process optimization strategy based on the solution of a MINLP

Varicol is a continuous chromatographic separation process which is based on the simulated moving bed principle with an asynchronous commutation of the inlet and outlet streams. The objective of this work is to develop a systematic procedure to optimize the operating conditions of the Varicol process, including the commutation sequence and the internal flow rates. To this end, the continuous dynamics are represented using a conventional mass-balance partial differential equation model, whereas the discrete dynamics are described by a timed transition Petri net. A Mixed-Integer Non-Linear Problem (MINLP) is formulated to maximize the process productivity under constraints on the desired purities of both the extract and the raffinate components. The MINLP is solved by constructing a set of candidate integer parameters and solving the resulting nonlinear problem for a subset of candidates selected on the basis of an approximate evaluation using the (true moving bed) triangle theory. Numerical results show the feasibility of this approach.