Optimal Operation of Heat Exchanger Networks

Abstract

Optimal operation of heat exchanger networks is much less studied than optimal design. The control objective of a heat exchanger network is to control the temperature out of the heat exchanger network within a certain range. Often the control objective is to maximize the end temperature, and this is the focus of this study. A typical example is feed preheating in a crude oil fractionator. There seem to be no simple systematic ways to maximize the end temperature and the practical solutions are often suboptimal. Alternatively, RTO is used which is both challenging and expensive. This study investigates the performance of a self-optimizing control strategy proposed by PhD candidate Johannes Jäschke. The method maxmimizes the end temperature and relies only on cheap temperature measurements, i.e. no flow measurements or technical data (heat exchanger area, heat transfer coefficients, heat capacities etc.) are necessary. The method has been demonstrated in four different cases; two theoretical cases and two real cases. The real cases are from Perstorp in Perstorp, Sweden and Statoil Mongstad outside of Bergen, Norway. A dynamic model has been made for the Perstorp case. Also, it has been looked into if the same self-optimizing control strategy can be used for an LNG process. The self-optimizing control strategy perform well in all the cases investigated. At Perstorp it improves the performance, while at Statoil Mongstad the performance is just as good as the existing RTO. The results are presented for Perstorp and Statoil Mongstad which both are optimistic about the method.