Final Report of FORGE WP4.1.2: verification of critical stress theory applied to repository concepts

This report describes an experimental study of 48 separate experiments examining the validity of critical stress theory. Two main types of experiment were conducted: 1). Loading-unloading tests, where fracture flow was monitored at constant injection pressure as normal load was increased in steps to a given level and then reduced back to the starting stress state; 2). Gas breakthrough experiments, where gas injection pressure was increased in a pressure ramp at constant vertical load. These were conducted with and without active shear. It was found that critical stress theory is valid in predicting the preferential flow of gas in relation to the orientation of the fracture plane with respect to the maximum horizontal stress direction. However, loading unloading experiments showed that understanding the stress history of the rocks is of paramount importance and a mere knowledge of the current stress state is insufficient in accurately predicting the nature of fluid flow. A total of 17 loading-unloading experiments were conducted, all on a 30° slip-plane. The main conclusions of this part of the study were; a). During a loading (vertical stress) and unloading cycle considerable hysteresis in flow was observed signifying the importance of stress history on fracture flow; b). For the case of gas injection the change in flow is chaotic at low normal loads, whereas for water injection the flow reduces smoothly with increased normal load; c). Hysteresis in horizontal stress observed during unloading demonstrates the importance of the ratio between horizontal stress and vertical stress and its control on flow; d). Differences have been observed between injection fluids (water and helium), especially the hysteresis observed in flow. For water injection flow is only partially recovered during unloading, whereas for gas enhanced flow is seen at low normal loads. A total of 26 gas breakthrough experiments were conducted on 0°, 15°, 30°, and 45° discontinuities; both with and without active shear. All tests were conducted in an identical manner. The main conclusions of this part of the study were; a). During gas breakthrough experiments episodic flow/fault valve behaviour was seen with a decrease in subsequent peak pressures and the form of the pressure response was different during subsequent breakthrough events; b). Repeat gas injection testing had shown a consistent gas entry pressure but considerably different, non-repeatable, gas peak pressures; c). Differences in gas entry pressure were seen dependent on the orientation of the fracture; d). Shear can be seen to reduce the gas entry pressure, suggesting that shearing in kaolinite has the opposite effect of self-sealing to gas. Other general observations of gas flow along fractures included; a). The flow of fluids through clay filled fractures is non-uniform and occurs via localised preferential pathways; b). The pressure recorded within the slip-plane showed a negligible fracture pressure and did not vary much in all tests