Stress and Fracture Prediction with Reservoir-Scale Geomechanical Models

Andreas Henk
Geologisches Institut, Universitaet Freiburg, Albertstr. 23 b, D-79104 Freiburg, Germany
[email protected]

Knowledge of the tectonic stress field and fracture pattern in a reservoir is essential for optimal well location and the planning of secondary and tertiary recovery, e.g. hydraulic fracture treatments. Particularly the design of horizontal wells with respect to borehole stability and multiple fracs relies on a robust pre-drilling prediction of the recent subsurface stresses. However, the orientation and magnitude of the stress field in sedimentary basins and, hence, the geometric and hydraulic properties of the natural and induced fracture network can be highly variable. Particularly near faults and due to lithological changes (e.g., salt, shale) stresses and fracture orientations can differ significantly from the regional trend. The study explores the potential of geomechanical models as a tool to predict tectonic stresses and fractures in hydrocarbon and geothermal reservoirs on the basis of mainly seismic and only sparse well data.

Geomechanical modeling utilizes three-dimensional finite element techniques to calculate stresses and strains in inhomogeneous structures with complex geometries and non-linear material properties. Special focus is on local stress perturbations in fault-controlled reservoirs. The numerical simulations describe brittle rock deformation by the Mohr-Coulomb law using specific values for cohesion, angle of internal friction and dilatancy angle for each lithology. If ductile rheologies, e.g. salt, are involved, their mechanical behavior can be described by temperature- and / or strain rate-dependent creep laws. Existing faults are incorporated into the model using so-called contact elements which are capable to transmit normal and shear stresses and allow for independent meshing of the individual fault blocks. Friction coefficients can be assigned to these elements representing the fault surfaces. The geomechanical models are reservoir- (kilometer-) scale and can describe either the present subsurface geometry or past evolution stages. In the first case, fault geometries and lithological boundaries are imported from seismic interpretations and the boundary conditions applied to the finite element model represent the present-day regional stress field. If the reservoir geometry has changed with time, like in fold-and-thrust belts, the numerical simulations have to be based on geometrically restored subsurface models representing individual stages in the tectonic history of the reservoir. Similarly, boundary conditions can be varied to study variable regional stress field orientations. The numerical simulation provides the complete stress tensor, so that, for example, the magnitude and orientation of the principal stresses can be derived for each part of the subsurface model. Calculated stresses and strains can be combined to predict fracture type and orientation as well as relative fracture density.

In order to assess the predictive potential of the modeling approach it is applied to the real world and model predictions are compared to observed stress and fracture data from a gas reservoir and various surface outcrops, respectively.

 

AAPG Search and Discover Article #90066©2007 AAPG Hedberg Conference, The Hague, The Netherlands