| Summary: | 碩士 === 國立成功大學 === 化學工程學系 === 103 === A well-designed heat-exchanger network (HEN) can often be adopted for maximum heat recovery in any modern chemical process. However, as time goes on under the normal operating conditions, fouling on the heat-exchange surface is unavoidable. If the heat-transfer units in a HEN are not cleaned regularly, the originally envisaged thermal efficiency can only be maintained for a short period of time.
Based on the published fouling models, Lavaja and Bagjewicz (2004) constructed a mixed-integer nonlinear program (MINLP) for synthesizing the optimal cleaning schedule of any given heat-exchanger network. Although this method could be used to produce schedules that effectively reduce the additional utility costs caused by fouling, the required cleaning operations still result in unnecessary utility consumption since the corresponding exchangers must be removed from duties.
The objective of this study is therefore to modify the aforementioned existing MINLP model so as to optimally assign spares to replace the units that are taken out of service for cleaning. Specifically, a set of binary variables are introduced and each is used to reflect whether a removed exchanger should be substituted with a spare. The corresponding optimal solution includes not only the cleaning schedule but also the total number of spares, their heat transfer areas and the substitution schedule. Finally, this thesis also provides the optimization results of a series of case studies to verify the feasibility and effectiveness of the proposed approach.
|
|---|
