Assessment of uncertainty in 2D and 3D geostatistical models for use in steam assisted gravity drainage prediction

点击翻译按钮获取中文摘要

Steam assisted gravity drainage (SAGD) is used to produce hydrocarbons from the Athabasca Oil Sands. Geostatistical methods are used to model properties of the oil sands that are then used to forecast oil production, steam usage and energy consumption. Properties include porosity, bitumen content, permeability and rock type. Modeling these properties involves estimating or simulating unknown quantities at every unsampled location in the deposit on a grid and usually with uncertainty. These estimates are made using all available information such as geophysical surveys, well data, seismic, geochemical, geomechanical, etc. The overall goal of the proposed work is to accurately predict uncertainty in a SAGD environment by correctly accounting for uncertainty in all modeling aspects. When implementing current best practice uncertainty assessment methodologies, there is a discrepancy between uncertainty calculated using 2D or 3D models. Typically 2D models are used to calculate oil in place global resources/reserves while 3D models are used for oil production prediction with flow simulation. The goals of this project are to (1) reconcile differences between 2D and 3D uncertainty models (2) propose methodologies to effectively estimate global uncertainty consistently for 2D or 3D models including trend uncertainty and (3) investigate the impact of correctly modeling uncertainty on SAGD performance prediction, well placement and economic/environmental considerations for multiple SAGD pads.