No abstract

No abstract

No abstract

Sequential Bayesian simulations are used to model the porosity distribution and assess the CO 2 storage potential in the Beauharnois Formation of the Saint-Flavien reservoir (Québec, Canada). The low porosity Beauharnois Formation is characterized by a complex geology, mostly composed of dolostones with a strong presence of limestone, sandstone, and shale. In such a complex geological environment, we transform the porosity distribution into a normal one to artifi cially stretch the range of porosity. This allows a clearer defi nition of the statistical relation between acoustic impedance (AI) and porosity, and a better identifi cation of petrophysical families in the reservoir unit. Guided by seismic derived AI cubes, 250 realizations of porosity are simulated by Bayesian sequential simulations (BSS), all respecting the initial porosity well logs, the a priori porosity distribution, and the statistical relation between AI and porosity. All realizations present different but realistic distributions of porosity. We estimate the connectivity between zones with porosity greater than 1.0%. The average porosity in the connected pockets is approximately 1.4% for all three selected realizations. We estimate 0.5Mt to 1.25Mt of CO 2 could be injected in the 3D model representation of the Beauharnois Formation in Saint-Flavien, with a CO 2 storage effi ciency factor of 27% to 36%.

No abstract