S-SS-3-1 Stream Analysis Method for Prediction of Shellside Heat Transfer and Pressure drop in Segmentally-Baffled Exchangers

J. W. Palen, J. Taborek, and A. Yarden
1967; revised 1977; revised 2007; revised 2024

The major result of HTRI's four years' work on shellside flow is the development of a method representing a new plateau in accuracy and generality for prediction of no-phase-change heat transfer and pressure drop on the shell side of segmentally baffled shell-and-tube heat exchangers. This work was based on by far the largest, most varied, and most reliable shellside data set ever compiled, which was analyzed with unprecedented thoroughness through the aid of high-speed digital computers. The most notable features of the method are

  • The ability to calculate with reasonable accuracy the amount of flow leaking through baffle clearances and bypassing around the bundle, enabling realistic application of ideal tube bank correlations to the remaining crossflow stream.
  • The capability of accurately predicting the effect of exchanger geometry variations such as baffle cut, tube spacing, baffle-shell clearance, and number of sealing strips, on heat transfer, pressure drop, and overall unit efficiency.
  • The recognition and preliminary correlation of a temperature profile distortion effect, due to leakage streams undergoing different temperature changes than the crossflow stream. This effect, which reduces heat transfer efficiency as a function of the amount of leakage and the closeness of the temperature approach, may have been a major factor affecting may inoperable exchangers in the past, but heretofore had not been seriously considered.
  • The demonstrated ability to predict shellside heat transfer rates and pressure drops over a range of geometry variations far exceeding that of normal practice, and over a Re range of 2 to 100,000, within the tolerable accuracy limits of ± 20 to 30 percent, thus far surpassing the capabilities of previous methods.

Some additional work in this field, especially in the area of temperature profile distortion, appears justified and is suggested. Further major improvements, however, would not be expected without an exponentially larger additional effort initiated at the level of ideal tube banks.