Successful prediction of the future performance of condensate reservoirs requires accurate values of dew point pressures of the interesting reservoirs. Although the dew point pressure can be determined experimentally from collected laboratory samples, these measurements are frequently not available. In these cases, fluid reservoir properties are determined with the use of empirical correlations or determined iteratively using an equation of state (EOS).
The objective of this paper is to present an application of genetic programming (GP)-Orthogonal Least Squares algorithm (OLS) to generate linear-in-parameters dew point pressure model represented by tree structures. The GP-OLS based gas condensate reservoir dew point pressure model was generated as a function of reservoir fluid composition (in terms of mol fractions of methane through heptanes-plus, nitrogen, carbon dioxide, and hydrogen sulphide, and the molecular weight of the heptanes-plus fraction), and reservoir temperature. The new model was developed using experimentally measurements of 245 gas condensate systems covering a wide range of gas properties and reservoir temperatures. One hundred-thirty five gas condensate systems that not introduced in building the new model were used to test and validate it against the other early published correlations. The validity test shows that the new model is more accurate than the other tested correlations, whereas the new model has the lowest average absolute relative error. Therefore, the new model can be considered an alternative method to estimate the dew point pressure when the experimentally measurement is not available.