Development of genetic programming (GP) models for gas condensate compressibility factor determination below dew point pressure

摘要

Gas compressibility factor plays a vital role in various engineering applications related to natural gas reservoir management, planning, transportation and processing. Compared to dry gases, gas condensates are thermodynamically complex and require thorough attention. The main challenge is phase segregation and compositional change during temperature variations or pressure depletion. Therefore, this study is focused on proposing novel compositional models based on a Genetic Programming (GP) framework for the accurate calculation of the gas condensate compressibility factor below dew point pressure. The new models are developed based on 1800 gas condensate datasets obtained from open literature. Both qualitative and statistical quantitative assessments were used to compare the precision and accuracy estimation of the new models to existing literature models. Moreover, the proficiency of the proposed models for compressibility factor calculations of gas condensate samples for sweet and sour gas samples was investigated. In addition, a sensitivity analysis based on Spearman and Pearson techniques was performed to carry out the degree of influence of each input parameter on the target value. It is expected that the developed models will pave the way for the accurate calculation of compressibility factors for gas condensates, which can be used by engineers for performance monitoring, optimization and production management in gas condensate systems.