“…For methane, the extended hard-sphere model performs slightly less well than the original scheme at intermediate reduced densities, with significant negative deviations for reduced densities 1/V* between 0.3 and 0.5, but at higher densities, the model performs as well as the original (for comparison, the original scheme correlates the data at 1/V* ≥ 0.2, with ΔAAD = 2.6 % and ΔMAD = 8.7 %). It is notable that the data for van der Gulik [36] at T = 298.15 K, which extends up to p = 1000 MPa, are represented with relative deviations of between +3% and -6%, whereas van der Gulik's [37] data at T = 273 K show generally larger deviations of between +2% and -12%. The deviations for ethane, seen in Figure 2(b), also show some dependence on 1/V* but are generally satisfactory at all densities.…”