This study is concerned with the propagation of a hydraulic fracture propagating near the free surface of a rock mass. This problem has a number of industrial and scientific applications: the industrial applications include the recent utilization of hydraulic fracturing to induce caving near mine openings and to improve the performance of remediation work at shallow environmental waste or spill sites, while the scientific applications deal with a number of near-surface geological processes involving hydraulic fracturing, such as the saucer-shaped magmatic sill intrusions. Fundamental questions that arise in connection to this problem are related to the influence of the in-situ stresses and gravity on the shape of the fracture (including the breakthrough radius, the distance from the fluid source at which the fracture daylights), the evolution of the fracture shape and size, the variation of internal fluid pressure with time and the dependence of the process on the rock and fluid properties. This research is a joint theoretical and experimental investigation involving an international cooperation between the University of Minnesota (UMN) and CSIRO Petroleum in Australia. The modeling effort, the focus of the activity at the UMN, deals with the development of computational models capable of simulating the growth of a bowl-shaped fracture, and with the tabulation of solutions in a parametric space to ensure most generality of the obtained numerical solutions. The laboratory work, which is carried out at CSIRO Petroleum, involve near-surface hydraulic fracturing experiments in Polymethylmethacrylate (PMMA) and in silica glass, under confined and unconfined conditions. The work combining development of rigorous solutions and experimental validation of these solutions is expected to yield quantitative means to analyze and predict the propagation of a hydraulic fracture near a free-surface, under a variety of conditions. The introduction of dimensionless control parameters allows collapsing countless specific cases into a single point of the parametric space, and opens the possibility of an economical tabulation of the computed solutions. The experimental research will also yield valuable data for benchmarking theoretical solutions of deep hydraulic fractures, which are virtually nonexistent despite the substantial economic value of hydraulic fracturing stimulations of oil and gas reservoirs. Finally, the experimental setup provides an unusual opportunity to observe and quantify the curved trajectory of a tensile (mode I) fracture.
