Heat Exchanger Reliability Model And Bi-objective Optimization Of Heat Exchanger Network

As an indispensable system in industrial production,heat exchanger network is an important breakthrough to save energy.Through nearly 60 years’ research,the heat exchanger network has been optimized from the optimization of the heat exchanger numbers,heat transfer area and additional utilities,gradually to optimize more other indicators.Such as,system reliability,flexibility,operability and environmental impact.However,there are many constraints if optimizing three or more objectives simultaneously,the model is complex and difficult to calculate.The bi-objective optimization model is the most widely used one at present.The predecessors have done a lot of research work on optimizing the economics and flexibility,or economics and environmental impact of heat exchanger network,but there are few systematic studies on the economics and system reliability of heat exchanger network.Moreover,the predecessors in the study of system reliability generally regard the the heat exchanger reliability unit as a constant,which is not in accord with the actual industrial production.Therefore,this paper firstly proposes a heat exchanger reliability model considering fouling growth and then establishes a super-structure model of heat exchanger network for optimizing economics and system reliability simultaneously.which eliminate the unreasonable assumption heat exchanger reliability is constant throughout the operation time.The specific research content is divided into the following aspects.(1)A fault tree reliability analysis model of heat exchanger was established,taking into account various inherent factors of heat exchanger failure,including the heat exchanger materials,corrosion resistance,medium erosion,etc.In order to explore the influence of the physical properties of the streams on the heat exchanger reliability,the fouling growth as a basic event was added to the fault tree reliability analysis model.The fouling growth cycle is divided into several stages,then the heat exchanger reliability is solved in stages.According to the logical principle of FTA model,the minimal cut set of FTA model can be counted.The main reliability factors are determined based on the structural importance degrees of the each basic event obtained by quantity calculation and the order of the structural importance degrees.In this way,more targeted measures can be taken.The computation of heat exchanger reliability is the calculation basis for computing the reliability of heat exchanger network.(2)A superstructure model for the economics and system reliability of heat exchanger network is established,in which the unreasonable assumption that the heat exchanger reliability is constant has been amended.The economic objective function includes fixed investment costs and operating expenses.The system reliability is not only related to the heat exchanger reliability,but also to the attended mode of the heat exchanger in the heat exchanger network structure.In this paper,the largest uncorrelated subsystem in the heat exchanger network is searched by the structure matrix.According to the heat exchanger numbers and the heat exchanger reliability,the system reliability can be obtained.(3)The NSGA-Ⅱ algorithm with excellent algorithm is applied to optimize the superstructure model of heat exchanger network,which analyzes how to determine the initial feasible solution and how to fix the infeasible solution that occurs during the iterative operation.Using the fast non-dominated sorting,the comparison of congestion degree and the elite retention strategy of NSGA-Ⅱ algorithm,the bi-objective optimization program is developed.(4)Combining The heat exchanger fault tree reliability analysis model and the heat exchanger network bi-objective model to compute two cases.The variation trend of system reliability with time and the footing between economics and system reliability at each stage are obtained.Both cases show that the system reliability decreases with the operating time,and the rate of decrease in the later period is lower and lower,which is related to the slowing of the later fouling growth.Comparing with the literatures,the economics of case 1 decreased by 2.47% and the system reliability increased by 127%.For case 2,the chosen optimal solution enhances system reliability by 127% with TAC increased by 2.71%.In the situation of a small sacrifice of economics,the system reliability can be greatly improved,which can explain the effectiveness of the bi-objective optimization superstructure model of the heat exchanger network.