Computational models capable of simulating non-ideal compressible flows are not yet sufficiently validated against measurement data. Furthermore, the wide variety of thermodynamic models available together with the large operational ranges make it difficult to correctly construct the simulation model with a sufficient balance between accuracy and computational speed. This research assessed the impact of the choice of various thermodynamic models on the accuracy of flow solvers, under a wide variety of operating conditions. This will allow for the generation of a thermodynamic map from which validation experiments can be designed. To that end, the results provided by polytropic (PPR) and non-polytropic (NPPR) Peng-Robinson EoS-based simulations in the SU2 flow solver were compared to those by a multi-parameter Helmholtz equation of state (HEOS), on a grid of 144 operating conditions, over the ORCHID nozzle domain. The pressure and Mach number responses were compared, as these can be measured experimentally, and previously obtained experimental uncertainties (assumed to be constant) were used as a distinguishing criterion. Mach number comparisons didn’t allow to distinguish between the models because of the high measurement uncertainty. In terms of pressure response, deviations were found to be negligible (lower than experimental uncertainties) below around \(\frac{p}{p_{\textrm{crit}}} = {0.31}\) . Further comparing across the whole grid, deviations increased to over 10 times experimental uncertainty values at supercritical conditions, increasing with decreasing compressibility factor. Additionally, Peng-Robinson EoS simulations were five times faster than HEOS, emphasising why one would choose them if the deviations are acceptably low. It is recommended to design validation experiments using the resulting map, and to further investigate the physical causes of these deviations.