Integrated Asset Modeling for Deep Gas Reservoirs. Case Study in UAE

The primary objective of reservoir management is the maximization of plateau duration and reserves recovery, while minimizing the project risks and maintaining the highest environmental and safety standards. It is critical to identify upfront all the risks and bottlenecks that may occur through the entire life of the assets, to prevent them from happening or minimize the reaction time in case of occurrence. Integration of all asset elements (subsurface dynamic models and a surface facility network) provides a fit-for-purpose tool to optimize assets development. This article describes fully coupled surface and subsurface modeling done for 6 reservoir formations (3 fields) sharing the surface facility and including 6 gas reservoirs with different PVT properties. A fully integrated model describes all technical processes occurring in the subsurface and surface systems and has to be further calibrated using measurements done as per the reservoir monitoring plan.

Computer Application to Estimate PVT Conditions in Oil Wells in the Ecuadorian Amazon

A reservoir behaviour's characterisation is determined by analysing the fluids' physical properties, reported in Pressure, Volume and Temperature (PVT) tests. These tests are performed in the laboratory or are estimated by mathematical correlations with the well's basic properties. The eastern basin of Ecuador is considered a hydrocarbon zone, and the analysis of the physical properties of the fluid from oil wells is essential. The aim is to develop the PVTTESTSYSTEM software to estimate PVT conditions when there are no laboratory tests. The study methodology is based on (i) Compilation of 10 PVT laboratory tests of oil wells in the eastern basin of Ecuador; (ii) Analysis of mathematical correlations; (iii) Development of the PVTTESTSYSTEM software, with the wells' initial conditions' input, selecting the mathematical correlation and estimation of results, based on the relationship of the properties of oil and gas; iv) Comparison of data obtained by laboratory tests and PVTTESTSYSTEM software reports. The software used with a graphical interface presents a registration and login platform and five modules that allow: inserting company and field data, initial oil well data, selecting correlations, calculating PVT properties and generating a graphic report. The results show that the mathematical correlations that estimate PVT properties were systematised, which approximate the laboratory tests' real results. The approximation of the calculated results with the actual results establishes a high confidence level for the PVTTESTSYSTEM software.

Machine Learning-Based Improved Pressure–Volume–Temperature Correlations for Black Oil Reservoirs

Pressure–volume–temperature (PVT) properties of crude oil are considered the most important properties in petroleum engineering applications as they are virtually used in every reservoir and production engineering calculation. Determination of these properties in the laboratory is the most accurate way to obtain a representative value, at the same time, it is very expensive. However, in the absence of such facilities, other approaches such as analytical solutions and empirical correlations are used to estimate the PVT properties. This study demonstrates the combined use of two machine learning (ML) technique, viz., functional network (FN) coupled with particle swarm optimization (PSO) in predicting the black oil PVT properties such as bubble point pressure (Pb), oil formation volume factor at Pb, and oil viscosity at Pb. This study also proposes new mathematical models derived from the coupled FN-PSO model to estimate these properties. The use of proposed mathematical models does not need any ML engine for the execution. A total of 760 data points collected from the different sources were preprocessed and utilized to build and train the machine learning models. The data utilized covered a wide range of values that are quite reasonable in petroleum engineering applications. The performances of the developed models were tested against the most used empirical correlations. The results showed that the proposed PVT models outperformed previous models by demonstrating an error of up to 2%. The proposed FN-PSO models were also compared with other ML techniques such as an artificial neural network, support vector regression, and adaptive neuro-fuzzy inference system, and the results showed that proposed FN-PSO models outperformed other ML techniques.