Tese e Dissertação

Tese: Development of a compositional reservoir simulator based on a plugin architecture

This work describes the development of a compositional simulator of oil reservoirs. It consists of a computational tool for numerical simulation of the dynamic behavior of reservoirs, aiming at increasing the efficiency of oil and gas extraction activities. The most well-known simulation model is the black-oil, in which the migration of hydrocarbon components between different phases (e.g., liquid or gas) is not considered. However, there are cases where such modeling simplification does not allow for obtaining realistic results, especially in the presence of volatile oils that can easily present phase migrations or also when CO2 is injected into the reservoir. On the other hand, to solve this problem, several compositional simulation models were developed, where such migrations of hydrocarbon components were considered. In this context, the main objective of this work is to develop a computational system for simulating compositional oil reservoirs based on the Coats model. The implementation will consist of an alternative to the black-oil model, already existing in the GSIM framework for reservoir simulation, developed as a partnership between Petrobras and PUC-Rio. The proposed compositional modeling uses a discretization of the domain through orthogonal Cartesian meshes and the finite difference method to approximate the differential equations that govern the problem. Furthermore, aiming to increase the efficiency of simulations, GSIM was developed in C++ language using a plugin architecture, which is optimized to allow faster and more efficient treatment of the data structure associated with reservoir problems. Using the plugin system allows greater versatility in developing new functionalities. It also enables using the proposed simulator for commercial and research purposes, representing an advantage over available commercial applications. The proposed computational system will be tested in three different reservoir simulations, and the results will be compared with those obtained by a commercial simulator called CMG-GEM.

Ver todos