Extensive experimental data from drained triaxial tests on oilsands at different confining pressures and ambient temperatures have led to the development of a non-associated generalized plasticity model in 3-D stress conditions. Many of the identifiable behavioural aspects such as hardening, softening, plastic volumetric expansion and contraction have been incorporated in the formulation. When implemented into the general purpose Finite Element program ABAQUS, numerical results compare reasonably well with measured response. In trying to weigh merits of the model, it is interesting to note that practical applications such as the computation of well casing/oil sand formation interaction during insitu thermal recovery of bitumen become amenable to analysis.
The use of insitu thermal methods such as steam injection through vertical well bores to deep seated oil sand reservoirs is common and relatively effective for the recovery of heavy oil and bitumen. There has been, however, numerous well casing failures reported during field injection trials. Due to the presence of two types of materials, namely shale and oil sand in the geological strata, strong pore fluid and stress gradients are created around the cavity. In particular, high differential shear stresses developed at the sand-shale interface during steam injection may cause instability and collapse of the well casing which compromise the good progress of oil production. The obvious solution to the casing deformation has been to stiffen the well casing installation, but in order to arrive at a rational and economical design, it is supreme to accurately evaluate the force distribution around the structure and deformation behaviour of oil sand.
Numerical schemes such as the Finite Element method are readily available in several program packages, one of them being ABAQUS. Although, they offer a wide range of applications which cover most common structural and geotechnical problems, it seems, however, that there is no suitable geological material model which can specifically capture most of the constitutive behavioural aspects of oil sand. On the other hand, the need for a comprehensive characterization of the Stress-strain behaviour of oil sand is fully acknowledged for a correct finite element analysis.
The deformation behaviour of oil sand is governed by many factors. However, it can be conveniently considered as comprised of basically two main constituents: the pore fluids (water, bitumen, and gas) and the sand skeleton. The theoretical relationships governing the behaviour of the pore fluid, as to describe gas exsolution and other related aspects, have been first introduced by Harris and Sobkowicz (1977) and subsequently extended by Byrne and Grigg (1980). The present paper does not treat the pore fluid response but rather focusses on the mathematical modelling of the sand skeleton as an elastic-plastic continuum.
The capability to realistically analyze oil sand behaviour is a challenging task mainly due to its complexity and variability in response. As such, the inherent non-linear and history dependent nature of the problem makes it necessary to resort to elaborate theories such as plasticity instead of simple linear elasticity for its modelling.