Nonsharp separation sequences are designed using an algorithmic approach. The design process is to separate a multicomponent feed stream into several multicomponent product sets. The proposed superstructure contains sharp, semisharp, and nonsharp separation options for all columns. Also embedded are all possible sequences and stream splitting, mixing, and bypassing alternatives. All the components may distribute between the overhead and bottom products. The component recoveries are modeled by using the Fenske equation and are explicitly treated as optimization variables. The search space of the algorithmic separation sequence is limited by the recovery lower bounds. These lower bounds also determine the minimum separation required. The superstructure is formulated as a mixed-integer nonlinear programming problem (MINLP). The objective is to minimize the total annual cost. Shortcut simulations and regression analysis are used to develop the cost function for the formulation. The solution of the MINLP formulation results in an optimal separation sequence.This algorithmic approach has been applied to five design problems. The results are compared to design via algorithmic solution using sharp separation, algorithmic solution using nonsharp separation without nonkey component distribution, and knowledge-based solution with nonkey component distribution. The solutions proposed by this study show saving improvements of 3-58.8%.
展开▼
