(629c) Retrofit of Heat Exchanger Networks Using Superstructure Modeling and Global Optimization

Sung Young Kim, Debora Faria and *Miguel Bagajewicz

School of Chemical Engineering and Material Science, University of Oklahoma

100 East Boyd Street, T-335 – Norman, OK 73019-0628 USA

* Corresponding Author

Abstract 

In this paper, we present a robust global optimization method to design heat exchanger networks using superstructures and a recently developed method of domain and imgae partitioning of nonconvex functions. In previous work, we applied this method to the Synheat model. When using superstructures bilinear terms are also used, so we extend the bound contraction model to include these. The procedure we propose uses an MILP lower bound constructed using domain/image partitioning (Faria, Kim and Bagajewicz, 2012) and bilinear decomposition (Faria and Bagajewicz, 2011a,b). Then a newly developed bound contraction procedure is applied and compared to branch and bound as well as branch and bound with bound contraction at each node. We finally solve the retrofit problem, by proposing that certain heat exchangers in the superstructure exist, and new exchangers can be added.

1. D. Faria and M. Bagajewicz. A New Approach for Global Optimization of a Class of MINLP Problems with Applications to Water Management and Pooling Problems. AIChE Journal (2011a).
2. D. Faria and M. Bagajewicz. Global Optimization of Water Management Problems Using Linear Relaxations and Bound Contraction Methods. Industrial and Engineering Chemistry Research. 50, 7, 3738–3753 (2011b).
3. D. Faria, S.Y. Kim and M. Bagajewicz. Global Optimization of Nonconvex MINLP Problems by Domain and Image Partitioning : Applications to Heat Exchanger Network. Computers and Chemical Engineering, Submitted (2012)