Banca de DEFESA: IGOR VASCONCELOS DE LACERDA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : IGOR VASCONCELOS DE LACERDA
DATE: 30/01/2023
TIME: 09:30
LOCAL: Programa de Pós-Graduação em Engenharia Civil
TITLE:

Numerical Simulation of Compositional Flow in Petroleum Reservoirs Using a Fully Implicit Formulation

KEY WORDS:

Petroleum Reservoirs, Compositional Formulation, Fully Implicit Approach.

PAGES: 98
BIG AREA: Engenharias
AREA: Engenharia Civil
SUMMARY:

In the oil industry, the numerical simulation of reservoirs is a widely used tool to investigate the effect of different injection or production strategies, optimization of field production and evaluation of uncertainties, since the equations that govern the mathematical model of a reservoir cannot be solved by analytical methods for general problems. Of the most adopted fluid models, Black-Oil and compositional stand out. The Black-Oil model is widely used in the reservoir simulation due to its relative ease of implementation and relatively low computation-al cost. However, for the simulation of more complex problems, such as the application of Enhanced Oil Recovery techniques, volatile oil reservoirs (example of the Brazilian pre-salt) or condensate gas, it is necessary to apply the compositional model to achieve results more accurate and physically adequate. The solution of the compositional model involves a complex system of equations whose solution represents, until today, a great challenge from the computational point of view. Several solution algorithms arise from the combination of different primary variable selections and decoupling techniques. In this work, we present a Fully Implicit (FI) formulation using Cartesian meshes for the compositional simulation with the use of Equations of State (EOS) for modeling the phases behavior. We adopt the following assumptions: isothermal flow; there are up to three phases in the reservoir (water, oil and gas); there is no mass exchange between the water and hydrocarbon phases; capillary pressure and physical dispersion are neglected. For the diffusive terms of the equations describing the mathematical model, we discretize using the Finite Volume Method with the Two-Point Flux Approxima-tion (TPFA), while for the advective terms we apply the first-order upstream weighting. This simulator was created in an integrated environment together with other projects developed by members of the PADMEC group at UFPE. The Python language was used in the implementation, resorting to the vectorization of operations in the search for better computational performance. In this environment, two sequential formulations have already been elaborated (one IMPEC and one IMPSAT), and as a contribution to this work, a Totally Implicit formulation was added. Our formulation is evaluated by solving 1-D and 2-D problems found in literature and the results are promising, providing a basis for future implementation of more complex models, such as an expansion to non-isothermal formulation and suitability for implementation using unstructured meshes.

COMMITTEE MEMBERS:
Externo ao Programa - 2364065 - FERNANDO RAUL LICAPA CONTRERAS - nullExterno à Instituição - MÁRCIO RODRIGO DE ARAÚJO SOUZA - UFPB
Presidente - 2216046 - RAMIRO BRITO WILLMERSDORF