Fracture Modelling as Part of Integrated Reservoir Characterization

Abstract

Abstract Natural fractures play an obvious role in many reservoir performances; in others their influence is more subtle, but yet still significant, especially in secondary or enhanced recovery schemes. It is more often than not the case that incorporation of fractures and small scale faults in a reservoir heterogeneity model will improve the validity and value of that model, and enable reservoir flow simulation and prediction to be on a surer basis. Numerical geomechanical models provide a means of predicting realistic strain fields associated with fracturing, and can incorporate general heterogeneities in mechanical properties, but have degrees of freedom in lateral boundary conditions that cannot be resolved by matching surface contours alone. They are also cumbersome to condition to data available from well. This paper introduces a geostatistical methodology for modelling strains that makes use of spatial covariances in the components of deformation: it provides a simpler, more flexible tool for conditioning strain to well data, in addition to surface data. It can be looked upon as an improved type of curvature analysis, but additionally capturing the 2 shear modes of strain that are neglected by that technique. This paper describes a set of stages for modelling strain that makes use of all the available structural information and taking the results forward to incorporate permeability due to fractures into the overall integrated reservoir characterisation and dynamic fluid flow prediction.