Revisiting the Coefficient of Isothermal Oil Compressibility Below Bubble Point Pressure and Formulation of a New Model using Adaptive Neuro-Fuzzy Inference System Technique

Abstract

Isothermal oil compressibility coefficient is one of the physical properties that requires an exact description for applied and theoretical science applications, especially in the solution of petroleum reservoir engineering problems. Conventional empirical correlations are however inconsistent and yield high error due to high input parameters needed and regional crudes-based development. For a reservoir with pressure below bubble point, the effect of co to the fluid flow is insignificant as it is overshadowed by the presence of large gas compressibility (cg). This study aims to increase the range of applicability and accuracy of the formula used for estimating the co by eliminating the limitations that occur in existing correlations. A new formula for the estimation of the coefficient of isothermal oil compressibility below bubble point pressure is devised using Adaptive Neuro-Fuzzy Inference System (ANFIS). The approach is a combination of neural networks and fuzzy logic. This method targets to model imprecise mode of reasoning in order to make rational decisions in an environment of uncertainty and imprecision. A benchmark has been set based on the best model available in the literature using the current set of data. Trial-and-error approach was followed with the assist of the trend analysis to check a model that represents the true phenomenon. A total number of 369 data points were collected from worldwide fluid samples for the purpose of training and testing the model. Exhaustive trend analysis has been conducted to verify that the proposed ANFIS model honors the true physical behavior. The new proposed model found to follow the correct trend which implies its reliability. In addition, a comparative study was carried out using the best available correlations to confirm the significance of the results of the oil compressibility prediction using ANFIS. Different statistical analyses have been shown to verify the robustness of the newly developed model. The statistical analyses showed a positive outcome whereby the proposed model obtained the lowest average absolute percent relative error of 3.3976% and the highest correlation coefficient of 99.76%. The best model tested among the other models has five input parameters and average absolute percent relative error of 12.07% and a correlation coefficient of 98.27%. The new approach managed to produce the most accurate model to predict the coefficient of isothermal oil compressibility below the bubble point when compared to the best available models in the literature. The new proposed model overcome the limitations described by the locality of some correlations as they are depending on data from certain locations.