The use of hydraulic fracturing for distressing rock formations is not a new idea. Two decades ago, the U.S. Geological Survey seriously considered massive injections to lubricate the San Andreas Fault in an attempt to avoid locking and associated energy buildup. More and more, hydraulic fracturing treatments are performed worldwide in gassy coalfields for degasification purposes. Those stimulations have recently reached sizes that make methane production also commercially viable. Burst-prone hard rock mines are also contemplating hydraulic fracturing ahead of the face in an attempt to locally relieve the high stress concentrations.
Based on a newly-developed three-dimensional numerical simulator, this paper analyzes the Potential energy release and its extent for various stimulation schemes. It addresses the influence of the following parameters: job size, pumping rate, fluid viscosity, and proppant density. In addition, effects of frictional interfaces and material heterogeneities are considered. Based on these parametric studies, the practical aspect of stress relief is discussed in detail and field recommendations are given.
Hydraulic fracturing has been widely used in the hydrocarbon industry since the 40's as a way of increasing the productivity of a well. Related techniques and models have been discussed by Howard and Fast (1970) and Economides and Nolte (1987). In most sandstone reservoirs, and under the assumptions that there are no local variations in the orientation of principal stresses, no large scale natural discontinuities, and that the hydraulic fracture is initiated from a borehole drilled parallel to one of the principal stress directions, the propagation path will be planar (horizontal, if the minimum stress σ3 is vertical; or vertical if σ3 is horizontal). The physical processes involved in the propagation of large scale fractures have been discussed by Cleary (1980). During this last decade, with the advent of in- creased computational power, the understanding of hydraulic fracturing processes has considerably improved and models have evolved from simple one-dimensional equations to complex three dimensional ones. Such models and their domain of application are discussed by Mendelsohn (1984) and BenNaceur and Roegiers (1990).
The experience gained in the oil and gas industry has transferred to other domains of energy extraction. The geothermal industry has attempted to connect wells by way of hydraulic fracturing (Armstead, 1979; Batcheloret al., 1980). Although attempts have been made to monitor the propagation of these man-induced fractures and eventually control their path (Batchelor et al., 1983), the inherent complexity of these crystalline reservoirs resulted in compounded fracture patterns, and the resulting inter-well communication are far from being routinely predictable.
Over the 80's, coalbed methane stimulation has become an economical process, and it has attracted an increased interest from operators and researchers as witnessed by the contributions to two symposia on coal bed methane extraction (Proceedings of the Coal bed Methane Symposium, 1987 and 1989). The mechanisms of recovery of gas from coalbed methane has been discussed in detail by King and Ertekin (1989). One important difference between stimulating coal bed formations and sandstone reservoirs lies again in the complexity of the hydraulically-induced
