A novel approach for infeasible path optimization of distillation-based flowsheets

In this work, a new approach for simultaneous steady state simulation and optimization of distillation-based flowsheets is presented. The process simulation is embedded in an optimization problem with a small number of optimization variables and constraints. Hence, no specialized solvers are needed. Moreover, an arbitrary number of specifications in the form of equalities or inequalities can be set in the problem formulation. This matches the process engineering demands usually much better than specifying a given quantity of fixed numbers. The remaining process variables for distillation columns are calculated from stage-to-stage using fixed-point iterations as developed and analyzed in previous work of the authors (Hoffmann et al., 2017). To illustrate the new approach, typical examples for process simulation and optimization of distillation-based flowsheets including several columns and also recycle streams are presented. For such distillation-based flowsheets an algorithm is presented that determines a possible calculation sequence of units within one iteration of the optimization algorithm.