The temperature of a zeotropic mixtures does not remain constant throughout a heat exchanger. Furthermore, zeotropes often exhibit a nonlinear temperature/enthalpy relationship. These factors contradict some of the assumptions that are made in deriving the Log Mean Temperature Difference (LMTD), a calculation that is used to compute the size of a heat exchanger (UA). In order to develop a more precise mean temperature difference, the derivation of the LMTD must be reëvaluated. Since the temperature is not an easily-determined function of enthalpy, the heat transfer process must be discretized and the properties determined at each point.
An ammonia-water mixture is examined first, since it is known to display a highly nonlinear temperature glide. Synthetic mixtures that are being studied by the refrigeration industry as replacements for HCFC-22 are also analyzed. In each case, the actual UA is compared to the UA found using the LMTD, and representative error scales are developed. It is found that these errors can cause a heat exchanger to be undersized by as much as a factor of fifty.
Finally, the advantages of zeotropes are also discussed, as are cycles that can utilize those advantages. Recommendations are made that the effect of the modified UA calculations on these cycles should be studied, and that the benefits of natural over synthetic refrigerants should also be investigated.