A MIXED INTEGER NONLINEAR-PROGRAMMING MODEL FOR RETROFITTING HEAT-EXCHANGER NETWORKS

This paper addresses the problem of determining the optimal retrofit of an existing heat-exchanger network, considering the placement / reassignment of existing exchangers to different process stream matches, their need for additional area, the creation of new process stream matches, the cost of stream repiping, and other issues. A mixed integer nonlinear programming (MINLP) model is proposed that encorporates all possible process stream matches, network configurations, and existing exchanger reassignments in a single mathematical formulation. In addition, it is shown how heat-exchanger rating equations, repiping costs, pressure drop considerations, and varying heat-transfer coefficients can be included in this model. The proposed model simultaneously evaluates the required area of each existing and potential process stream match, the required additional area of each potential match-exchanger assignment, and the estimated repiping cost, while optimizing the selection of process stream matches, the network configuration, and the match-exchanger assignments. The resulting solution of this mathematical model gives the optimal selection of process stream matches, their heat loads, their total required area, the optimal network configuration, and the optimal assignment of existing heat exchangers to new or existing process stream matches, based upon the cost of moving an exchanger and the actual required additional area. It is shown that this mixed integer nonlinear programming formulation can be solved efficiently by applying the Generalized Benders Decomposition technique. The properties of the decomposition are discussed, and the method is illustrated with three example problems. © 1990, American Chemical Society. All rights reserved.