Abstract

Unconventional gas and geothermal reservoirs are typically naturally fractured and therefore the challenging problem of hydraulic fracture (HF) growth through naturally fractured rocks is an important issue especially for the development of accurate predictive HF models. This paper presents results from laboratory experiments in which HFs have been created in both sandstone and crystalline (gabbronorite) rock specimens such that they impinge either orthogonally or at an oblique angle on an unbonded, machined frictional interface. Depending on the conditions of each experiment, the fractures either cross or are blunted at the interface. The growth behavior was ascertained for these experiments based on a synthesis of the injection pressure record, active ultrasound monitoring, and direct post-experiment inspection of fractured specimens. From these experiments we observe:

  1. A contrast between direct crossing and crossing with an offset for orthogonal and oblique intersection, respectively,

  2. Apparent independence of crossing/blunting behavior on the friction coefficient for orthogonal intersection,

  3. Dependence of the stresses required to promote crossing on the intersection angle between the HF and NF,

  4. Existence of cases in which the hydraulic fracture crosses but then essentially ceases propagation such that all fluid is diverted to the interface.

Introduction

Hydraulic fracturing is a vital technology used to increase the productivity of oil, gas, and geothermal wells. Predicting hydraulic fracture (HF) growth is, in turn, critical to design of hydraulic fractures that grow to the desired length, accept the desired amount of proppant, and remain contained in the reservoir without unwanted height growth into surrounding layers (e.g. Economides and Nolte, 2000). The importance of HF modeling has prompted 6 decades of research during which simulators have become more efficient, more accurate, and have addressed a growing range of complexities (see review of e.g. Adachi et al., 2007).

One of these complexities is the interaction between HFs and mechanical discontinuities such as natural fractures (NFs). Efforts in this area have been mainly driven by development of unconventional (i.e. low permeability) oil and gas resources because essentially all viable low permeability reservoirs, including shale gas/oil reservoirs, are naturally fractured to some degree (King, 2010). One of the most important questions for developing effective HF stimulation strategies in these reservoirs is whether the hydraulic fracture(s) is/are expected to be single-stranded or whether interaction with the NFs will lead to network-like growth. The basic building block for numerical simulators that can address this question (such as Kresse et al., 2013 and Weng et al., 2014) is whether an HF that intersects an NF will cross the NF or whether its growth will be captured by the NF so that additional fluid injection pressurizes the NF but does not generate ongoing growth of a new HF. These scenarios, including subcategories such as crossing with an offset, have been recognized for several decades (Daneshy, 1974; Warpinski and Teufel, 1987, and as recently reviewed and expanded by Chuprakov et al., 2013).

This content is only available via PDF.
You can access this article if you purchase or spend a download.