Yale et. al.1 showed significant differences between "uncoupled" reservoir simulations of a highly compressible (but elastic) reservoir and the same model run with deformation and fluid flow "coupled" together. This paper utilizes a similar model but investigates the effect of plasticity of the reservoir rock and alteration of permeability with deformation on these coupled simulations.
The coupled simulations show that plasticity significantly increases the compressibility of the rocks and the compaction drive energy of the reservoir. The initial stress state of the reservoir is shown to have a large effect on the degree of plasticity and the degree of depletion for the same amount of fluid withdrawal and the same rock properties.
Modeling the change in permeability with deformation of plastic sands shows an extremely large effect on near wellbore pressure drawdown and deformation over normal reservoir simulations. The coupld geomechanis-fluid flow simulations show very strong interaction between pressure drawdown, plastic strain, and permeability.
Numerical simulation of flow in and from petroleum reservoirs is done to predict production so that optimal development plans for the reservoir can be developed. The more accurately the models represent the physics of the problem, the more useful the prediction made from the models can be. Most developments in numerical reservoir simulation have revolved around accurate modeling of fluid properties and interactions and accurate representation of the storage and transmissibility properties of the porous geologic material.
In many reservoirs, however, the storage and transmissibility properties of the reservoirs are strongly stress dependent and therefore there values are "coupled" with the pressure and flow of the fluids in the reservoirs. As such, numerical simulation of these types of stress-sensitive reservoirs requires the fluid flow equations from standard reservoir simulation to be solved simultaneously (i.e. coupled with) the equations for deformation of the porous media.
A number of different models have been developed to try and couple geomechancial models with fluid-flow simulators1–6. One difficulty in doing these types of simulations is that in addition to having to solve additional sets of non-linear equations (flow and deformation), a much larger volume of rock must be simulated (reservoir, overburden, underburden, sideburden) than in fluid flow only simulations. The increased computational effort (or decreased model resolution) requires a good understanding of the geologic conditions under which geomechanics-fluid flow coupling is required. It also raises questions as to the degree of coupling (partial coupling of separate models versus simultaneous solution of flow and deformation equations) and the degree on non-linearity (linear elastic versus plastic, constant permeability versus stress-dependent permeability, fracture flow, etc.) that needs to be included to realistically simulate a particular reservoir.
Yale et. al.1 and Gutierez and Hansteen4 have contrasted coupled and uncoupled simulations to understand when it is beneficial to apply coupled modeling. In both cases, linear elastic models of the reservoir were used to simulate the deformation. Partial coupling (running a geomechanics simulator every few timesteps of the fluid flow simulator to feed back porosity/permeability changes to the fluid flow simulator) has also been used to run large reservoir simulations with reasonable computational times.