(169c) Solving MINLP with Heat Exchangers: Special Structure Detection and Large-Scale Global Optimisation

Optimising heat exchangers networks (HEN) may increase efficiency in industrial plants; we develop deterministic global optimisation algorithms for a mixed-integer nonlinear optimisation (MINLP) model that simultaneously incorporates utility cost, equipment area, and hot / cold stream matches [1, 4]. In this work, we automatically recognise and exploit special mathematical structures common in HEN including log mean temperature difference and Chen approximation.

We computationally demonstrate the impact on the global optimisation solver ANTIGONE [3] and benchmark large-scale test cases against heuristic approaches. We also present, as an alternative, a Pyomo [2] model which solves the heat exchangers networks problems using iterative refinement and describe tradeoffs between the two techniques.

References

1. M. Escobar and I. E. Grossmann. Mixed-integer nonlinear programming models for optimal simulta- neous synthesis of heat exchangers network, 2010. Available from CyberInfrastructure for MINLP at: www.minlp.org/library/problem/index.php?i=93.

2. W. E. Hart, J.-P. Watson, and D. L. Woodruff. Pyomo: modeling and solving mathematical programs in python. Mathematical Programming Computation, 3(3):219–260, 2011.

3. R. Misener and C. A. Floudas. ANTIGONE: Algorithms for coNTinuous Integer Global Optimization of Nonlinear Equations. J. Glob. Optim., 59(2-3):503–526, 2014.

4. I. Quesada and I. E. Grossmann. Global optimization algorithm for heat exchanger networks. Ind. Eng. Chem. Res., 32(3):487–499, 1993.