"Brittleness Index" has become a popular parameter in unconventional plays, used in an attempt to describe the behavior of rocks under in-situ hydraulic stimulation. However, with a multitude of different definitions and more critical assessments of the brittleness index (BI) emerging in recent years, the shortcomings in the interpretation of the brittleness index as a fracability parameter are increasingly evident. Calculating brittleness from both rock elastic properties and mineralogy has become commonplace due to the almost ubiquitous acquisition of those data in "shale" plays. The temptation is then to directly apply them in reservoir stimulation decisions. However, taking into account many shortcomings in the computation and physical interpretation of the brittleness index, questions arise about the meaning of rock brittleness when defined as a combination of different rock properties. In this paper we look into three different brittleness index definitions based on dynamic elastic moduli and mineral composition, and critically appraise their similarities, and potential for application in hydraulic fracture programs.
Effective exploitation of tight oil and gas shale resources is possible due to a combination of horizontal drilling, and hydraulic fracturing with slick water. To enhance flow from these extremely low permeability reservoirs, pressurized water is pumped into a borehole initiating fractures that propagate away from the borehole into the reservoir. The right combination of treatment conditions, rock properties, and stress regime allows not only the propagation of new complex fractures that enhance permeability, but also results in cross-cutting and activation of pre-existing fractures. Furthermore increased pore pressure in the reservoir induces slippage on pre-existing faults. As rock properties are an important part of this mix, a great amount of research has been focused on searching for the association between effective hydraulic fracturing and brittleness of rocks.
The concept of brittleness evolved as an idea of a rock's ability to fail in a brittle manner, i.e. the point at which elastic strain that dominates a rocks deformation up to the point its strength is exceeded and the rock fractures (Altindag, 2003). Brittle failure was therefore opposed to ductile failure, when inelastic strain prevails over elastic strain, causing rocks to deform irreversibly and ultimately fail. In an attempt to classify rocks based on their ability to undergo brittle failure, or conversely, showing varied degree of ductility, a brittleness index (BI) can be defined. Today, three main groups of brittleness definitions exist, referring to different concepts of rock properties or behavior. The first group focuses on the static mechanical properties of rocks, by comparing different rock strength (compressive, tensile, residual, fracture toughness) and strain relations. In this paper we will not discuss these definitions, but they are reviewed elsewhere (Hu et al., 2015a, Yang et al., 2013, Zhang et al., 2016)). The second group focuses on different aspects of rock composition, comparing weight or volume fraction of minerals that are supposed to favor brittle failure, to those that are supposed to be less favorable to brittle failure. The main concept behind mineralogy based brittleness is to highlight the right combination of mineralogy that would represent zones that require relatively low pumping pressures in order to overcome a fracture gradient and ensure that the propagated fractures remain open. Brittleness indices based on mineralogy and proposed by different authors are not consistent, attributing varied importance to different rock components. For instant Jarvie et al. (2007) introduced an index that would facilitate to identify the most quartzose and best producing intervals in the Barnett Shale, and therefore quartz is their most desired component as opposed to calcite and clays. Wang and Gale (2009) introduced a second desired parameter, dolomite, as another brittle component. The mineralogy concept of classifying rocks is very convenient to apply, as it only requires knowledge of rock's composition, accessible from logs or cuttings. Even more convenient in producing continuous brittleness information is an idea of using elastic parameters calculated from Pand S-wave sonic logs or seismic survey, forming a basis of the third group of brittleness definitions. Rocks undergo elastic deformation when subject to sonic waves, but despite that, little information about inelastic deformations is generated in this manner. However with the right calibrations to static moduli, one can successfully calculate such parameters as Young's modulus or Poisson's ratio at the depth of interest. The application of elastic moduli in describing rock's brittleness is still regarded as lacking a physical meaning because they describe different rock characteristics, i.e. deformation vs failure (see Bai, 2016 and references therein). Moreover, there is an objection to treating brittleness as a rock property, but rather a rock's behavior, dependent on the stress regime in the reservoir (Hu et al., 2015a), variability in the stress gradient, and an extent of time-dependent rock deformation and stress relaxation (Sone and Zoback, 2014). The purpose of this paper is therefore to compare popular brittleness index definitions and reconcile any discrepancies between different approaches for the estimation of rocks brittleness.