Multi-period design of heat exchanger networks

Heat exchanger networks are an integral part of chemical processes as they recover available heat and reduce utility consumption, thereby improving the overall economics of an industrial plant. This paper focuses on heat exchanger network design for multi-period operation wherein the operating conditions of a process may vary with time. A typical example is the hydrotreating process in petroleum refineries where the operators increase reactor temperature to compensate for catalyst deactivation. Superstructure based multi-period models for heat exchanger network design have been proposed previously employing deterministic optimisation algorithms, e.g. (Aaltola, 2002 and Verheyen and Zhang, 2006). Stochastic optimisation algorithms have also been applied for the design of flexible heat exchanger networks recently (Ma et al., 2007 and Ma et al., 2008). The present work develops an optimisation approach using simulated annealing for design of heat exchanger networks for multi-period operation. A comparison of the new optimisation approach with previous deterministic optimisation based design approaches is presented to illustrate the utilisation of simulated annealing in design of optimal heat exchanger network configurations for multi-period operation.

▸ A methodology is proposed for the design of heat exchanger networks that may remain optimally operable under multiple periods of operation. ▸ A multi-period simulation model for heat exchanger networks is developed. ▸ A simulated annealing algorithm is employed for optimization. ▸ Results of the proposed approach are compared with previous deterministic optimization based design approaches.