Sensitivity and uncertainty analysis of heat-exchanger designs to physicalproperties estimation

We have used Monte Carlo methods to study the sensitivity and uncertainty o f heat exchanger designs to physical properties estimation, The determinati on of appropriate confidence intervals for the overall heat-transfer coeffi cient and total required heat-exchange area plays a very important role in heat-exchanger thermal designs. The physical property models used exhibit s ystematic and random errors, which can change depending on how the models a re used. Of particular interest in this work were the errors in properties estimation at high temperatures. Experimental physical properties of hot-ga s mixtures at high temperatures (i.e., combustion gases) are very limited, and heat-exchanger simulation and design relies heavily on predictive model s for physical properties. In this work, case studies of heat-exchanger per formance and design under the influence of random and systematic errors on property estimation are presented, The results show that the performance an d design can be very sensitive to errors in physical properties, The analys is method proposed can be used to identify when and which, properties play a significant role in the error propagation for this type of equipment, Fur ther, the methodology proposed can be used to study the uncertainty in heat -exchanger performance and cost introduced by these types of errors, (C) 20 01 Elsevier Science Ltd. All rights reserved.