For the last decade, the integrity of caprock was the dominant topic in SAGD aspects, especially after Josyln creek steam release accident (ERCB, 2010). Lots of works had been done to identify the failure mechanisms (Carlson, 2012) and massive progress had been made(Collins, Walters et al., 2011; Yuan, Xu et al., 2011; Walters, Wang et al., 2012). One of these efforts was to determine the maximum operating pressure from geomechanical study (Walters, Wang et al., 2012). Another issue in SAGD operation falled into geomechanical category as well was inter-bedded shale (IBS), which largely impeded the vertical growth of SAGD chamber (ITO, T. HIRATA et al., 2004). IBS generally yielded low operation performance in heterogeneous reservoir and gradually draw the attention of the industry and academia. Yang (Yang and Butler, 1992) was the first to conduct physical simulation to study the effect of relative position and size of IBS on SAGD performance. He concluded IBS above the injector would force steam chamber to detour which largely delayed the rising process of chamber. So breaking IBS to facilitate fluid flow has been an objective pursuit by the industry.

Actually the efforts to prevent caprock from failure was essentially the reverse purpose of breaking IBS. The conditions leading to caprock integrity issue were exactly those to promote potential failure of IBS. Conventional ideas like drilling slim hole, vertical well, tunnel or fishbone wells (Chang, Ivory et al., 2012; JOHN L. STALDER, 2012; PTAC, 2012; TUNNEY CATHAL, 2012) might sound great but failed to meet the challenge of identifying location of IBS precisely. Alternately in situ method was more feasible and most works reported in public domain fell into this category. As is well known, rock failure happened normally in the forms of shear failure and tensile failure (Carlson, 2003; Collins, 2005; Dusseault and Collins, 2008). Shear failure was especially common in anisotropic stress condition. Accompanying with this concept, uniaxial boundary condition was widely integrated (Walters, Wang et al., 2012). It was rather simple and ideal model. In addition, failure criteria was another important ingredient for shear failure study. Models like Mohr-Coloumb (Carlson, 2003; Collins, 2005; Dusseault and Collins, 2008) and Drucker-Prager (Yuan, Xu et al., 2011; Yang, Xu et al., 2013) were generally used in academic research. The former one was widely used in engineering activities while the latter one was commonly adopted for numerical simulation (IPEK, FRAUENFELD et al., 2008; Du and Wong, 2010). Tensile failure was also possible in certain unloading condition (XU, YUAN et al., 2011) and should be concerned properly. It's necessary to mention Li's work (Li and CHALATURNYK, 2006) in which a relation between permeability and volumetric strain was derived. This essentially bridged the gap between reservoir engineering and geomechanical research and helped understand permeability enhancement after failure.

Apart from numerical simulation, physical experiments also provided valuable thoughts to address IBS issue. Collins & Carlson (Carlson, 2003; Collins, Walters et al., 2011) taught proper methods to core, store, and prepare core plugs for experiments. Collins (Collins, 2005) reviewed the work at UTF project and concluded that stress front propagated far ahead of temperature and pressure front. This finding indicated IBS would be broken by means other than fluid injection. One of the most feasible approach was thermal stimulation which was already there in SAGD operation. Heat was one of the most important component in thermal stimulation process and heat transfer always exist in IBS no matter steam or hot water entered. In contrast, fluid wouldn't penetrate IBS with extremely low porosity and permeability and hence alter the stress condition inefficiently. Dr yuan (XU, YUAN et al., 2011) conducted lab tests and showed both shear failure and tensile failure of IBS during thermal loading. He argued that the stress path and loading history was important and might lead to thermal contraction or thermal expansion. Thermal expansion was critical in thermal stimulation process(AGAR, MORGENSTERN et al., 1983; BUTLER, 1986) which might result in pore pressure increase and anisotropic stress status. To which extent this factor might affect the strength of IBS depended on the competition between thermal expansion and pressure dissipation. On the other hand, thermal contraction of shale was also extendedly studied (Bois, CurisTec et al., 2011; Li and Wong, 2013) and fractures would developed in certain condition. Hueckel (Hueckel and Baldi, 1990; Hueckel and Borsetto, 1990) studied saturated clay heating under undrained condition. He presented yield surface change in p'- q - T plot and found overconsolidation ratio played an important role in the failure behavior. Wong (Wong, 1998) proved that swelling caused by water imbibition was an alternative to weaken shale and improve effective permeability.

You can access this article if you purchase or spend a download.