Integrated Reservoir Characterization and Modeling of the São Tomé Saline Aquifer (Brazil): A Case Study of a CCS Pilot Project
随着全球多个国家碳捕集与封存（CCS）技术的快速发展，巴西也在努力采纳这一关键解决方案以减少温室气体排放。巴西石油公司（Petrobras）作为全球最大的CCS参与者之一，已决定在巴西坎波斯盆地的圣汤姆水层盐下储层开展一个CCS试点项目。
该项目旨在测试此类项目的多个方面，并将该含水层本身作为更大规模CCS项目的候选地。预计在2027年开始注入，项目模拟已完成第一阶段审批。本文聚焦于针对这一CCS先导项目的地质建模、岩石物理特性、井测试、地质力学和流动模拟。
CMG软件应用情况
在本研究中，使用了CMG的GEM组分模拟器来建立模拟模型。模拟采用了角点网格，在注入井周围2.5公里半径范围内加密为50 x 50米，以提高模拟精度。模拟覆盖了3年的注入期，随后进行了175年的模拟以捕捉CO2与地层岩石相互作用的行为。模拟中使用了相对渗透率曲线来表示毛细管压力和滞后效应的影响，并使用了6种水相组分。
Abstract
As Carbon Capture and Storage (CCS) advances rapidly in several countries, Brazil is also working to adopt this crucial solution to reduce GHG emissions. Petrobras, one of the largest CCUS players in the world, has already decided to execute a CCS Pilot project in a saline reservoir in the Campos Basin – Brazil, the São Tomé reservoir. The idea is to test several aspects of this type of project and the aquifer itself as a candidate for larger CCS projects. The injection is scheduled to begin in 2027 and the project modeling has already completed its first stage for approval. This paper focuses on the geological modeling, petrophysics, well testing, geomechanics and flow simulation of this CCS Pilot Project. Geological modeling of saline aquifers presents several challenges due to limited availability of geological data. The São Tomé aquifer, with its large dimensions and proximity to the coast, lacks representative subsurface information for the entire depositional environment. To address this, conceptual geological models were employed to represent facies distribution and petrophysical properties more accurately. The reservoir rock, composed of siliciclastic sandstone, has porosity ranging from 22% to 27% and permeability ranging from 216 mD to 2080 mD. The facies modeling process was divided into three stages: subenvironments, architectural elements and electrofacies modeling. Porosity and permeability models used rock and well log data to define the parameters of normal and log-normal distributions, respectively. The spatial distribution trends considered the distance to the coastline for marine facies and the distance to channel axis for fluvial and deltaic facies. The well testing and reservoir evaluation aims to understand in detail the reservoir stratigraphy through open-hole logging, and the vertical fluid flow behavior in the porous medium through injection tests. This allows an assessment of the vertical migration process of the CO2 plume and helps to improve the accuracy of numerical models used for planning and monitoring the injection process. Logging-while-drilling (LWD) and wireline logging techniques are planned for the injection and monitoring wells. Core samples will be obtained from the injection zone and above, and fracture tests will be conducted to understand the geomechanical behavior of the rocks and ensure injection safety. Fluid samples from the aquifer will also be collected for chemical analysis to study the rock-fluid interaction. Water and CO2 injection tests will be performed and will help to understand the behavior of CO2 during injection and how it differs from single-phase injection cases. Pressure data, obtained from measurements in the injection zone and above, will provide insights into the vertical connectivity between the injection zone and upper zones.
Numerical simulation studies were used to establish the optimal depth for these sensors, which also guided the well completion design. The geomechanical analysis used data acquired from three wells in the nearby region and showed that the expected Bottom Hole Pressure (BHP) during injection is approximately 50 bars lower than the minimum horizontal stress. The temperature of the injected CO2 is 7°C higher than the reservoir temperature in the perforated interval, ruling out the possibility of thermally induced hydraulic fracturing. The simulation model used the Compositional GEM simulator from CMG. The injection rate is 200,000 m³/d (std) with the injection point located at a depth of -1350m. The simulation covers a period of three years of injection, followed by 175 years of simulation to capture the behavior of CO2 interaction with the formation brine. The good petrophysical characteristics of the reservoir rock, as well as the relative low salinity (35-65 g/L) of the aquifer, favors the CO2 injection into the São Tomé aquifer, where the main trapping mechanisms estimated by numerical simulations were Residual CO2 trapping and Solubility trapping. Despite the challenges of geological modeling of saline aquifers for CCS projects due to limited subsurface data in a vast region, utilizing conceptual models, appropriate data acquisition and monitoring plans, and advanced techniques such as geophysical measurements, injection tests, and numerical simulations, valuable insights can be obtained regarding the behavior of CO2 in the aquifer, ensuring the safety and success of future CCS projects.
Keywords: CCS, Numerical Modeling, Saline aquifer, Fluvial deltaic deposits
作者单位
- 巴西石油公司（Petrobras）
