HENEA-3: A SECOND-LAW BASED EXPERT ASSISTANT FOR THE SYNTHESIS OF HEAT EXCHANGER NETWORKS

摘要

The present paper describes the latest development of an Expert System for the design and optimization of Heat Exchange Networks (HEN). The code described herein, HENEA (acronym for Heat Exchanger Network Expert Assistant) is a Knowledge Based System capable of interacting with a Process Engineer to generate an optimal configuration for a HEN under several different objective functions. Contrary to most of the currently available methods, the code is able to produce such solutions by making explicit use of a set of Second Law Analysis criteria and of some of the basic principles of the "Pinch Method". With respect to the previous version, HENEA now includes an additional application, "HEN-SPLIT" that accounts for the possibility of splitting one or more of the participating streams according with a general "minimal entropy generation" criterion. In its present version, HENEA can be used both for process design (new HEN) and for process structural optimization (retrofit problems). The optimization procedure is based on Sama's "Second Law Insight Method" and is directly and explicitly finalized to the reduction of the global entropy generation. The code is also endowed with a cost calculations utility, and can be thus employed to search for the optimum using two alternative optimization criteria: it generates in fact both a Maximum Energy Recovery (MER) and Minimum Exchange Area configuration (MCC, Minimum Capital Cost). Since the code is though intended for practical industrial use, and it is well known that a thermodynamically optimal configuration is not always the "most convenient" one from an engineering point of view, HENEA presents to the user also a certain number of additional sub-optimal configurations. An interactive graphic user interface makes HENEA very user-friendly: furthermore, the user can (indirectly) interrogate HENEA about its choices and understand the underlying physic in an unambiguous and immediate way. Several tests conducted on some of the available benchmarks have proven that the solutions generated by HENEA are always comparable to those produced by other optimization criteria (Pinch Method or Second Law method): in fact, for large problems (many participating streams, several splittings) they are often better than those obtained with its competitors.