فهرست نشریات علمی دسترسی آزاد ایران

Precision in computational analyses and simulations is paramount in the oil and gas sector. A critical aspect of these calculations involves ascertaining the molar density of gases and liquids under varying pressure and temperature conditions. For this purpose, the Benedict-Webb-Rubin (BWR) equation of state (EOS) stands out as a robust instrument for approximating the fluid's pressure-volume-temperature (PVT) properties. This article aims to introduce an effective technique for refining the BWR EOS by employing Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO) algorithms. The Super-Halley method was used to determine the molar volume or molar density. This article details how, using a dataset of 360 experimental observations, the most effective constants (comprising eight parameters) were ascertained utilizing the PSO and ACO algorithms. Furthermore, an additional set of 50 experimental data points was analyzed to assess the accuracy of the BWR EOS. The results indicated an error margin of 2.164% for PSO and 2.768% for ACO, respectively. Subsequently, the predictive error associated with the 11-parameter Benedict-Webb-Rubin-Starling (BWRS) equation was found to be 3.75%. In contrast, the simpler, three-parameter Soave-Redlich-Kwong (SRK) equation exhibited a notably higher error rate of 9.866%. These findings underscore a general trend: as the number of parameters in an EOS increase, the prediction error tends to decrease. Interestingly, its error rate is approximately double even though the BWRS equation features three additional parameters compared to the empirical BWR equation. This observation suggests that experimental approaches significantly enhance predictive accuracy despite the increased computational demands in determining optimal constants.