A. Background

In the survey of calculation methods for shell-side flow in segmentally baffled exchangers given in Section 248. reference is made to the “flow stream analysis” method. In this method, the respective streams were the baffle/shell and tube/baffle leakage streams, the bypass stream and, of course, the cross-flow stream (see Section 248C). The flow split between the respective streams is determined on the basis equalisation of the pressure drop between the respective streams.

The concept of considering the various streams through the exchanger and estimating flows such that the stream pressure drops were equalised was originally proposed by Tinker (1951) with a more formal publication also by Tinker, appearing in Tinker (1958). However, the method did not become popular for several reasons including:

1. The flow resistances for the various streams were not well quantified and
   
   
2. The method was not well suited to hand calculation. To carry out hand calculations, it was necessary to introduce a range of simplifying assumptions that rendered the method inaccurate.

In the 1950’s and early 1960’s, an extensive series of experiments was carried out on shell-and-tube heat exchangers at the University of Delaware. These experiments confirmed the large effect of bypass and leakage on exchanger performance. The data were correlated (Bell, 1960 and 1963) by introducing correction charts for the effect of the leakage and bypass streams and this formed the basis for the widely-used Bell-Delaware method which is presented in detail in Section 249, Section 250, Section 251, Section 252, Section 253, Section 254, Section 255, Section 256 and Section 257. The method was suitable for hand calculation and is extensively described in many other handbooks and textbooks. Since the method involved the introduction of correction factors to a then-existing set of cross flow heat transfer and pressure drop correlations for ideal tube banks, these correlations are presented in Section 253 for completeness. The correlations differ from those given in Section 146 and Section 170 (though the differences are likely to be fairly small in the common ranges of application).