OPTIMAL DESIGN OF HEAT EXCHANGER

Abstract

/Deplishir4 energy resources and the resulting escalating cost has forced the scientific community all over the world to consider every possible way of conserving it. Chemical Industry being one of the most energy intensive industries, can contribute greatly towards this global aim by improving upon the efficiency of it's various operating units. Heat Exchanger, an essential unit in almost every production process, is the target of this thesis, which is aimed at minimizing the 'total annual cost' of the exchanger. Total annual cost is the sum of annual operating cost and a component of fixed or capital cost. The design variables of heat exchangers, in general, effect the two components a4c: of the cost in an opposite manner, increasing one and decreasing the other hence the variables are to be optimized in order to minimize the total cost. The problem associated with heat exchanger optimization is that if it is worked upon using exhaustive design correlations and detailed cost models, which predict the results (performance and capital cost estimate) with higher accuracy the problem becomes too complex for any further mathematical analysis. On the other hand, if relatively simple correlations and short-cut methods are adopted, the problem becomes quite simple but the accuracy of the predicted result becomes questionable. Hence, the decision on design correlations, cost model and the optimization technique is probably the most important step in the design optimization of the heat exchangers. In this analysis, an in-between approach has been adopted with slightly higher weightage to accuracy of the results. Three most commonly used exchanger have been covered in this work : Shell and Tube Heat Exchanger, Gasket Plate Heat Exchanger and Spiral Plate Exchanger. Shell and tube exchanger was designed and rated using the latest correlations on tube side and Bell's method for shell side. A detailed cost model was used for cost estimation. Optimization was done using Case study method guided by M. J. Box Complex method. In all, six variables were optimized : tube length and baffle space ratio as continuous variables and tube outside diameter, tube pitch, tube arrangement and tube side passes as discrete variables. In the first run of 11 trials the cost decreased from Rs. 478586 to Rs. 225,239 .