PROPAGATION OF FLUID-DRIVEN FRACTURES IN JOINTED ROCK .2. PHYSICAL TESTS ON BLOCKS WITH AN INTERFACE OR LENS

As introduced in Part 1 (Int. J. Rock Mech. Min. Sci. & Geomech. Abstr. 27, 243-254, 1990) of this paper, we have pursued a dual-numerical and experimental-modelling program to gain insight in the interaction of induced, fluid-driven fractures with natural discontinuities in rock masses. This is a topic of interest in the simulation of oil and gas reservoirs, as well as geothermal fields. It is also relevant to the disposal of liquid wastes underground and the estimation of in situ stresses, both by hydrofracturing. The experimental program was composed of two series of tests on blocks of rock simulants. The first one is denoted "interface tests". It was performed at the time of the development of the steady-state, coupled fracture and flow numerical model, for the purpose of validation. The blocks were hydrofractured while loaded biaxially. These tests provided information only as to whether cracks had crossed the interface between the two parts of the blocks, or not. No time-dependent data or fracture path data were obtained during the tests. The fracture trajectories were determined from post-test dissections of the blocks. The second, more sophisticated, test series will be denoted "lens tests". Sandstone lenses were embedded in hydrostone blocks which were fractured with single-wing fluid-driven cracks, while under a triaxial external loading. Fluid pressure and crack front were tracked as a function of time. This series had a two-fold purpose: to provide a validation basis for the new, time-dependent, numerical developments, and to provide forward diagnostics, on the pressure-time records, of the interaction of hydrofractures with embedded lenses. The physical testing was successful on both counts. In particular, the pressure-time records showed clear pressure discontinuities associated with the passage of the hydrofracture into the lens. © 1990.