Abstract
Laboratory experiments have been conducted to determine the effect of in-situ stress variations on hydraulic fracture containment. Fractures were initiated in layered rock samples with prescribed stress variations, and fracture growth characteristics were determined as a function of stress levels. Stress contrasts of 300 to 400 psi (2 to 3 MPa) were found sufficient to restrict fracture growth in laboratory samples of Nevada tuff and Tennessee and Nugget sandstones. The required stress level was found not to depend on mechanical rock properties. However, permeability and the resultant pore pressure effects were important. Tests conducted at biomaterial interfaces between Nugget and Tennessee sandstones show that the resultant stresses set up near the interface because of the applied overburden stress affect the fracture behavior in the same way as the applied confining stresses. These results provide a guideline for determining the in-situ stress contrast necessary to contain a fracture in a field treatment.
Introduction
An under-standing of the factors that influence and control hydraulic fracture containment is important for the successful use of hydraulic fracturing technology in the enhanced production of natural gas from tight reservoirs. Optimally, this understanding would provide improved fracture design criteria to maximize fracture surface area in contact with the reservoir with respect to volume injected and other treatment parameters. In formations with a positive containment condition (i.e., where fracturing out of zone is not anticipated), long penetrating fractures could be used effectively to develop the resource. For the opposite case, the options would beto use a small treatment so that large volumes are not wasted in out-of-zone fracturing and to accept a lower productivity improvement, orto reject the zone as uneconomical.
These decisions cannot be made satisfactorily unless criteria for vertical fracture propagation are developed and techniques for readily measuring the important parameters are available.
Currently, both theoretical and experimental efforts are being pursued to determine the important parameters and their relative effects on fracture growth. Two modes of fracture containment are possible. One is the situation where fracture growth is terminated at a discrete interface. Examples of this include laboratory experiments showing fracture termination at weak or unbonded interfaces and theoretical models that predict that fracture growth will terminate at a material property interface.
The other mode may occur when the fracture propagates into the bounding layer, but extensive growth does not take place and the fracture thus is restricted. An example is the propagation of the fracture into a region with an adverse stress gradient so that continued propagation results in higher stresses on the fracture, which limits growth, as suggested by Simonson et al. and as seen in mineback experiments. Another example is the possible restriction caused by propagation into a higher modulus region where the decreased width results in increased pressure drop in the fracture, which might inhibit extensive growth into that region relative to the lower modulus region. Other parameters, such as natural fractures, treatment parameters, pore pressure, etc., may affect either of these modes.
Laboratory and mineback experiments have shown that weak interfaces and in-situ stress differences are the most likely factors to contain the fracture, and weak interfaces are probably effective only at shallow depths. Thus, our experiments are being performed to determine the effect of in-situ stresses on fracture containment, both in a uniform rock sample and at material properly interfaces.
SPEJ
P. 333^
Experimental Study of Stress-Seepage Coupling Properties of Sandstone under Different Loading Paths
