Long Lake Oil Sand SAGD History Matching Research

Abstract

AbstractLong Lake oil sand has more than 10 years of production history. As the production is decreasing, the residual oil distribution and production forecast is needed to support sustaining the production. History matching were considered in the oil sand development residual oil distribution analysis as an effective method in conventional oil field. Because of the long simulation time of field model, synthetic CMG model were set up first based on field parameters. Extensive sensitivity analyses were conducted within proper parameter ranges for two groups of parameters. One with parameters that are almost fixed before history matching, mainly include geological parameters. The other with parameters that can be changed during the matching, mainly include fluid and rock properties. Field model history matching was done by tuning local static parameters, operational parameters, and global rock and fluid physical properties. Take cumulative oil as the objective function, the result shows that in geological parameters initial oil saturation and formation thickness have the largest effect. Porosity and permeability follows. In rock and fluid properties, relative permeability curvature and oil viscosity have the largest effect. Reservoir rock thermal diffusivity, compressibility, heat capacity, thermal conductivity, and overburden heat capacity, thermal conductivity have less effect but also significant. Geological parameters have strong impact on the steam chamber (SC) shape, at curtain circumstance the SC may deviate from standard Butler SC shape. The production profile changes accordingly. Wells with similar behaviors are found. Deterministic geomodeling method is thought to be insufficient when dealing with multi-wellpair SAGD history matching. Statistical geomodeling combined with automatic history matching will be a trend in the future. This article does sensitivity analysis on geological and reservoir engineering parameters to the SAGD SC development and production profile, points out the problems during history matching and provides a preliminary trial.KeywordsLong LakeOil sandSAGDNumerical simulationSensitivity analysisHistory matching