Practical Iterative Coupling of Geomechanics With Reservoir Simulation

Abstract

Abstract The use of reservoir simulation coupled with geomechanics to model physical phenomena such as compaction, subsidence, induced fracturing, enhancement of natural fractures and/or fault activation, SAGD recovery etc.., has been increasing. Different methods of coupling have been investigated by numerous researchers: fully implicit coupling, iterative coupling and one way coupling. So far, the iterative explicit method appears to be the prefered method for field-scale simulation. This method is a loose coupled approach between a reservoir simulator (finite volumes) and a geomechanical simulator (finite elements). At user-defined steps, the fluid pressures calculated by the reservoir simulator are transmitted to the geomechanical tool which computes the actual stresses and reports the modifications of the petrophysical properties (porosities and permeabilities) back to the reservoir simulator. In the classical iterative scheme, at each stress equilibration step, the reservoir simulation needs to be restarted from the previous converged step. This restart based scheme can be difficult to implement in practice within an industrial IT environment. This paper presents a new iterative scheme which allows: This is achieved by performing the pressure/stress iterations at the end of a complete reservoir simulation. Porosity and permeability modifications at various times are calculated, and the complete reservoir simulation is then repeated. Iterations are continued until convergence is achieved. The convergence of this new scheme is discussed and results are presented for two cases described below. Introduction The importance of geomechanics in problems such as wellbore stability, hydraulic fracturing and subsidence is well known. In recent years, there has been growing awareness of the importance of the link between fluid flow and geomechanics in the management of stress sensitive reservoirs1–9. New needs for coupled simulations appear such as assessing the integrity of the overburden for heavy oil recovery using thermal mechanism (SAGD technique, ..) or for acid gas injection. Standard reservoir simulation of compaction drive accounts for nonlinear porosity changes determined from uniaxial strain tests on cores. In many cases, laboratory-derived compressibility must be adjusted to match the contribution of compaction to total hydrocarbon recovery. Geomechanical effects such as stress arching and non-unique stress path are among the causes of discrepancy between laboratory-derived and field compressibility factors. If compressibility varies linearly with the mean reservoir pressure, then predictive reservoir modeling can be achieved without coupling between stress and flow. However, geomechanical effects are rarely linear for a number of reasons. These include load variations due to modification of pressure, temperature and saturation, change of the mechanism of production, progressive activation of faults and fractures that affect mechanisms such as stress arching and a non-linear stress path. Unlike standard compaction drive simulation, there is no simple linear method to account for the effects of stress on permeability especially for fractured systems, where the changes of permeability might be directional, localized and strongly non-linear. There are several ways to achieve the coupling between flow and stress 10–20. The most rigorous is done with fully coupled simulators, which not only solve the flow and the mechanical equations simultaneously, but also allow for anisotropy and non-linearity of the rock constitutive model. The feasibility and accuracy of such simulators, as far as complex and large scale reservoir systems are concerned, have yet to be proved. Partial coupling on the other hand consists of linking a flow simulator with a stress simulator allowing a good compromise between feasibility and accuracy. A one way link from flow to stress simulator is often used for subsidence forecasts. However, to solve the compaction drive problem, one-way coupling is not sufficient. To ensure the compatibility of pore volume calculations from the flow and the stress simulators, iterations must be performed within each stress analysis step before proceeding to the next stress step with or without permeability changes.