Steam Surface Condenser Design based on Cost Optimization Using Genetic Algorithm

Abstract

Steam surface condenser is a shell and tube heat exchanger used in power plant to condense the steam. Cost minimization of this condenser design is a key objective for both designer and users. In the design of condenser, one need to check the thermal performance which satis es process condition, while rating depends on number of parameters like tube geometry, tube size, tube layout, and working uid condition. In the present work, rating of surface condenser has been carried out using appropriate condensation correlation for tube bundle and tubeside ow. The variation of thermo-physical properties of water and steam with temperature are also considered to evaluate the overall heat transfer coe cient for surface condenser. A Computer program has been prepared to solve the problem e ciently. The results of overall heat transfer coe cient calculated by the program are compared and found to be within 3.33% of deviation with the results obtained from HTRI software. The Results are also compared with the experimental results of HEI standard. These also shown good agreements with available experimental result for lower tubeside velocity (<2.3 m/s) and the deviation found is within 11.256%.The present study also explores the use of a non-traditional optimization technique; Genetic Algorithm (GA), for optimization of steam surface condenser. The code for GA is developed and successfully applied for the optimization of same by varying the design variables such as shell internal diameter, tube outer diameter, tube thickness and tube material. The two di erent ranges of design variables are used. One is Siemens range (dimension range of earlier parameters are used by Siemens Ltd., Baroda) and second is TEMA Range (dimension range of same parameters taken from TEMA standard). The objective function for capital cost and total cost is derived and same is used for optimization. The optimized results obtained from selecting these range are compared and it was found that by widening the design variables range using total cost as an objective function , the e ciency of GA for optimization improves.