How equation-of-state choice and property uncertainty influence numerical heat exchanger design: A case study of an n-butane/CO₂ mixture

Numerical design of heat transfer and pressure loss in heat exchangers depends strongly on the thermophysical property models used for the working fluid. This is particularly relevant for zeotropic mixtures, where saturation behavior, temperature glide, and transport properties determine local driving temperature differences during phase change. This work analyzes the interplay between heat exchanger modeling and thermophysical property models from user and developer perspectives, examining how model choice affects engineering predictions and analyzing how uncertainties in individual fluid properties propagate into application-level results. A one-dimensional spatially resolved double-pipe condenser model is applied to a zeotropic n -butane/CO₂ mixture with water as secondary fluid. The equations of state GERG-2008, Peng–Robinson, and a molecular-simulation-based Helmholtz model are combined with transport-property models based on the extended corresponding states principle and residual entropy scaling. Local heat-transfer coefficients and pressure gradients are evaluated using correlations suitable for zeotropic mixtures. Results show that fluid-model choice substantially affects temperature profiles and heat-exchanger sizing, mainly through differences in saturation states. Predicted heat exchanger lengths vary from −7.3% to + 11.8% relative to the mean; in a limiting case with small temperature difference full condensation is not always reached. Sensitivity analyses show that uncertainties in thermal conductivity, dynamic viscosity, and isobaric heat capacity dominate fluid-property uncertainty propagation into heat-transfer predictions, whereas pressure-drop-related properties have only minor influence on exchanger length under present conditions with relatively large tube diameters. Including empirical correlation uncertainties for heat-transfer coefficients and pressure loss makes them dominate the total output uncertainty; nevertheless, fluid-property uncertainties remain a relevant secondary contribution.