Novos modelos computacionais hierárquicos para reservatórios fraturados carstificados

• Ano de defesa: 2018
• Autor(a) principal: Lopes, Tuane Vanessa
• Orientador(a): Murad, Marcio Arab
• Banca de defesa: Murad, Marcio Arab, Loula, Abimael Fernando Dourado, Pires, Adolfo Puime, Falcão, Flávia de O. Lima, Bezerra, Francisco Hilario Rego
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Laboratório Nacional de Computação Científica
• Programa de Pós-Graduação: Programa de Pós-Graduação em Modelagem Computacional
• Departamento: Coordenação de Pós-Graduação e Aperfeiçoamento (COPGA)
• País: Brasil
• Palavras-chave em Português: Reservatórios; Rochas sedimentares; Método dos elementos finitos; Homogeneização autocoerente; Modelagem hierárquica
• Área do conhecimento CNPq: CNPQ::ENGENHARIAS::ENGENHARIA CIVIL::ENGENHARIA HIDRAULICA::HIDROLOGIA
• Link de acesso: https://tede.lncc.br/handle/tede/286

In this work, we construct a new computational model to describe flow in karstified and fractured carbonates with particular emphasis on the pre-salt reservoirs. We consider the reservoir composed of intact rock and geological features that frequently occur in carbonates, such as the objects generated by the karstic process (caves, sinkholes and vugs) and a network of fractures (joints and veins). In this scenario we develop a Hierarchical Karst- Fracture Modeling (HKFM), where larger geological objects that directly influence flow and geomechanics are explicitly incorporated within the model whereas smaller structures which occur at higher density, are homogenized and replaced by equivalents continua with hydromechanical properties computed through an analytical upscaling procedure. The homogenization to the mesoscale is performed using two different procedures. The first consists in computing, through an analytical upscaling technique, equivalent properties of the porous matrix containing a network of micro-fractures. In the second, we discuss in detail the caves filled by collapsed breccias, which arise with the collapse of the rock after chemical dissolution processes, giving rise to several sub-regions. Among them we highlight (i) chaotic microstructures composed of rock fragments (clasts) and sediments (sand, clay) characterizing the chaotic breccias, (ii) the adjacent region with mosaic breccias, and (iii) the rock disturbed by the collapse and dissolution, called “damage zone” or crackle breccias. In this scenario, we explore techniques based on self-consistent theory, to estimate equivalent properties, such as permeability and elastic modulus, associated with inclusions in the form of ellipses with high aspect ratio. In the hydrodynamical formulation fractures of larger dimension with high connectivity are treated as objects of reduced dimension and incorporated in the hydrodynamics within the framework of the Discrete Fracture Modeling (DFM). Our approach incorporates the resistivity of fractures, allowing for the description of a barrier mechanism that occurs in the presence of filling material inside the fractures which gives rise to jumps in the pressure field. The discretization of the hydrodynamical model at the meso scale by the finite element method requires the triplication of the degrees of freedom in the pressure variable within the subdomains occupied by the fractures, in order to capture the jumps in the pressure field. Computational simulations are performed at different scenarios, ruled by different properties of the geological objects, giving rise to equivalent permeabilities at the macro scale by flow based upscaling techniques. Such equivalent properties can be incorporated into the pre-salt carbonate reservoir flow models, bringing important information about the finer scales consequently establishing a precise bridge between the macroscopic models and the microstructure. The reiterate homogenization (micro-meso and meso-macro) proposed in this work coupled with the hierarchical modeling of the geological complexities brings a new avenue for the exploration of the sensitivity and the constitutive dependence of equivalent properties in different scenarios which are subsequently used as input parameters in commercial simulators adopting coarse computational meshes.