Image-based Permeability Assessment for Oil Sands and Application to Reservoir Forecast

点击翻译按钮获取中文摘要

Optimizing reservoir development and resource management requires an accurate forecast of recovery performance. Petrophysical properties are essential to the evaluation of the reservoir formation. Porosity and permeability are the most important parameters since they control resource estimation and flow rate, respectively. Over the life of the reservoir, many crucial decisions depend on the accuracy of these predictions. However, the estimation of such properties, especially permeability, is challenging as it is difficult and expensive to acquire representative rock samples for permeability testing in oil sand deposits. This research proposes a methodology to estimate accurate high-resolution permeability (and porosity) for improving reservoir forecasts.The aim of this research proposal is to predict permeability and porosity by machine learning techniques. CT-Scans and image logs will be combined with the conventional well logs to build a 3D indicator model and representative porosity-permeability relationships. Next, a high-resolution porosity-permeability relationship will be built. Finally, permeability (horizontal and vertical), facies, porosity and water saturation are modeled for the entire reservoir domain. The calibrated porosity-permeability relationships are applied to construct properties at individual grid block. This research will be primarily beneficial to the Canadian company PetroChina Canada Ltd. (formerly known as Brion Energy Corporation). The results of this project could lead to reduced steam consumption, optimized production design, minimized environmental footprint and improved reservoir forecast by improving numerical reservoir models. This will lead to substantial environmental and economic advantages to Canada by mitigating water resource usage, and natural resource could be extracted with more attention to environment.