An improved procedure is presented for analyzing heat exchanger data where a fixed thermal resistance (e.g. fouling) is to be determined from two separate sets of data. The data are usually collected on two heat exchangers one of which is characterized by an additional thermal resistance. The heat exchangers are identical in all other aspects.
The procedure consists of using the so-called covariance regression model and is shown to provide more accurate results than the usual method of constructing two Wilson plots.