Robust intelligent tool for estimating dew point pressure in retrograded condensate gas reservoirs: Application of particle swarm optimization

• A sophisticated approach to monitor dew point pressure in retrograde gas condensate reservoirs was developed.
• The robustness of fuzzy approaches was compared with the PSO-ANN approach.
• Extensive dew point pressure data in gas condensate reservoirs was managed using the PSO-ANN model.

Liquid production from gas condensate reservoirs, which is an important economic and technical issue, depends on the thermodynamic conditions underlying the porous media. Accurately estimating the relevant parameters is an incentive for researchers to develop and propose a diversity of correlations; however, certain correlations are not sufficiently precise compared with correlations that are routinely applied to determine the dew point pressure (Pd). Due to numerous misunderstandings in Pd estimations, which are typically observed in upstream industries, great effort was expended herein to produce a high-performance method to monitor the Pd. The solution was produced by creating a hybrid of two effective and robust methods, the swarm intelligence and artificial neural network (ANN) models. The proposed model was extended using precise dew point pressure data reported in previous studies; moreover, based on these data, the evolved intelligent approach and conventional schemes were compared. The statistical results show a notable performance by the smart model in determining the dew point pressure of condensate gas reservoirs. Based on the reliable results, which are highly accurate and effective, it can logically be inferred that implementing the proposed approach, PSO-ANN, can aid in better understanding reservoir fluid behavior through reservoir simulation scenarios.