Métodos geoestatísticos para definição do modelo geológico de um reservatório do pré-sal

Author(s)

Jean Carlos Rangel Gavidia

• Document type: Doctoral Thesis
• Press: Campinas, SP.
• Institution: Universidade Estadual de Campinas (UNICAMP). Instituto de Geociências
• Defense date: 2023-10-20

Examining board members

Alexandre Campane Vidal; Ulisses Miguel da Costa Correia; Bruno César Zanardo Honório; Michelle Chaves Kuroda; Manuel Gomes Correia

• Advisor: Alexandre Campane Vidal

Abstract

This thesis presents the results of the research project: Geostatistical Methods for the Definition of Geological Model of a Pre-salt Reservoir. The main objective of the project is to identify the properties of the reservoir and understand the factors that affect the high heterogeneity of the reservoir such as lithology, mineralogy, and reservoir properties, to know the parameters responsible for the abrupt increase of permeability at certain intervals, finally to represent these characteristics in a geological model, identifying the most appropriate algorithms for the representation of the facies and properties including the Representation of fractures and vugs. To achieve the main objective, the study was divided into three parts. The first section deals with understanding the relationships between mineralogical composition, facies, reservoir characteristics, and production. At this stage, a paper was published in the journal of Geoenergy Science and Engineering. "Utilizing Integrated Artificial Intelligence for Characterizing Mineralogy and Facies in a Pre-salt Carbonate Reservoir, Santos Basin, Brazil, using Cores, Wireline Logs, and Multimineral Petrophysical Evaluation". This work led us to the conclusion that the best zones within the reservoir are associated with shrubstone, and spherulitestone facies, a high volume of dolomites with abundant presence of secondary porosities (vugs). The second article refers to the identification of zones with extremely high permeability and production, which coincide with ranges identified in the first article (high amounts of vugs, high concentrations of dolomite and represented by facies such as shrubstones and spherulitestones) and can meet the requirements of super K zones. Traditional permeability determination approaches (core samples, NMR permeability, or regressions) cannot reproduce the high permeability values reported in Drill-stem test DST production tests, or the high production flows observed in PLT production profiles. This research provides a new methodology that includes static data and accounts for the porosity of the vugs and their impacts on increased permeability, to be calibrated with dynamic data such as DSTs, PLT, or Productivity Index (J) tests. as well as an innovative method to estimate the probability of finding super K zones. The last part corresponds to the construction of the geological model that can represent the heterogeneities of the reservoir. For facies modeling, we evaluated two of the traditional methods: Sequential Indicator Simulation SIS and Truncated Gaussian Simulation TGS, and we incorporated the use of Multi-Point Statistics MPS, to correct the deficiencies presented by methods based on two-point statistics (variograms) such as SIS and TGS that are incapable of reproducing the continuity of geological bodies, just as they are not capable of reproducing complex geometrical bodies. On the contrary, MPS can address the weaknesses of variogram-based methods, but it is supported by training images (TI) that are representations of conceptual geological models. The results of the facies modeling showed that the MPS method shows better results in the modeling of facies in complex reservoirs such as pre-salt carbonates. Still, the creation of the training image represents a real challenge (AU)

• FAPESP's process: 20/01306-0 - Geostatistical methods for definition of geological model of a pre-salt reservoir
• Grantee: Jean Carlos Rangel Gavidia
• Support Opportunities: Scholarships in Brazil - Doctorate