Geostatistical modeling accounting for variation of reservoir and geomechanical properties in Montney formation, Alberta

摘要

Spatial distributions of reservoir and geomechanical properties such as porosity, slowness, and rock moduli are key components of static modelling and production optimization and depend on the shape of variogram models. Incorrect determination of search neighbourhood and anisotropy in variogram modelling result in non-optimal fracture design, inaccurate assessment of hydrocarbon in place, uncertain oil recovery, and poor net present value estimation in unconventional reservoirs. Here, we investigate the selection of variogram models with respect to anisotropy and data trend. A three-dimensional (3D) earth model was built by integrating both petrophysical and geological log data as well as dynamic elastic properties of the Montney Formation in Alberta, Canada. Sequential Simulation (SS) is used to populate reservoir and geomechanical properties honoring hard data and the histogram and spatial continuity of the data. The geomodelling results demonstrate the heterogeneity of geomechanical properties. In the method developed here, bulk density and slowness were used to simulate petrophysical and geomechanical attributes and variogram anisotropy was captured in both vertical and horizontal directions by using a nested structure variogram model. On the other hand, hole effect and decreasing trend of the variogram plots suggest repeating stratigraphic units in the vertical direction. Consequently, the appropriate search neighborhood was selected based on the trend obtained from the variogram plots. Although, density/porosity values had almost normal distribution, the spatial continuity of the data differs from well to well. On the other hand, the results showed higher vertical continuity for slowness distribution. Sensitivity analysis depicted the important effect of variogram determination on static model and development plan in the study formation. In addition to above, mechanical properties such as Young’s modulus and Poisson’s ratio were estimated and used to populate the geomodel. We defined fracability based on Young’s modulus, Poisson’s ratio, tensile and shear stress level, and fracture toughness. The suggested workflow in this paper can be used to optimize well placement and number of fracture stages and locations based on fracability.