Geomechanical Modeling of Complex Reservoirs: Recent Lessons and Current State of the Art
J. T. Fredrich
Sandia National Laboratories, Albuquerque, NM 87185-0750
Cost-effective improvements in the technology needed to develop and manage reservoirs in challenging environments require an increase in our understanding of geomechanical behavior. The local stresses relevant to reservoir
-scale processes are strongly affected by stratigraphic and structural features at that scale, and the relationship between hydrocarbon production and the mechanical behavior of the
reservoir
and overburden can be complicated and difficult to discern from field data directly. Numerical simulation provides a means to achieve critical insight into the behavior of complex reservoirs and to advance understanding of both the subsurface environment before drilling as well as the relationship between fluid flow and geomechanical behavior during production. This paper discusses recent work aimed at utilizing large-scale geomechanical modeling as a field development and
reservoir
management tool. We describe constitutive models developed for important classes of geomaterials and describe field cases where detailed understanding of the rheological behavior of the
reservoir
and adjacent formations is critical for successful field development. The central issue in the first case relates to unique geomechanical effects associated with basic rheological contrasts between
reservoir
sands and adjacent or overlying formations. The specific setting is the deepwater Gulf of Mexico where the cost of drilling a single well is tens of millions of dollars, so that the economics
drive
developments with a minimum number of wells with a service lifetime of 20-30 years. Our work focuses on understanding the large-scale geomechanical loading conditions that exist in and around the massive salt bodies that pervade this region. Non-linear finite element modeling is applied to enable better planning of well paths by providing estimates of vertical and horizontal stresses within, below, and adjacent to massive salt bodies for wellbore stability analyses so as to avoid areas of potential geomechanical instability, identify potentially unstable well trajectories, and enable accurate fracture gradient prediction while entering, drilling through, and exiting salt bodies. The second field case relates to geomechanical effects induced by near-wellbore pore pressure drawdown in low-permeability, compactable reservoirs. Non-linear finite element modeling is applied to investigate the causative agents of casing damage, and to identify operating policies and infill drilling strategies to mitigate against future casing damage. Historical (~20 years of production) simulations reveal the important role of material contrasts in influencing the geomechanical behavior during production.
The case studies illustrate the difficult issues encountered in practical application of large-scale three-dimensional non-linear finite element geomechanical modeling and suggest areas where improvements and/or research advances could be beneficial. These areas include: development of faithful (and tractable) solids models and finite element meshes from disparate geologic and/or numerical data; development of realistic constitutive models and robust and efficient implementation of these material models in finite element codes to achieve reasonable solution times; determination of material models from sparse log and/or and rock mechanics and the natural heterogeneity of geosystems; inclusion of coupled deformation-fluid flow processes; implementation of far-field (tectonic) stresses and stress initialization; and integration of geomechanical modeling results with other analysis tools.
Copyright © 2004. The American Association of Petroleum Geologists. All Rights Reserved.