A 2-D model of the mechanism of borehole instability due to crack propagation parallel to the free surface is considered. Pre-existing cracks start to grow under the action of the compressive stress concentrations at the borehole boundary. The interaction with the boundary eventually makes the crack propagate unstably and separate a thin rock layer which subsequently buckles. This initial step of failure slightly moves the borehole boundary, enabling the next crack to propagate. Such a process repeats itself until the change in the borehole shape is sufficient to suppress the unstable phase of crack growth. This is regarded as the ultimate extent of the failure. The smaller the sizes of the pre-existing cracks, the larger the extent. This implies that in the cases when the considered failure mechanism prevails, larger failures can be expected for less microcracked rock masses.
Borehole breakouts have been used for many years as an indicator of the in-situ stress field (e.g., Bell and Gough, 1983; Shamir and Zoback, 1992). Although the orientation of such breakouts can be used to determine the in-situ stress tensor orientation in the case of a hole drilled parallel to one of the principal stress components (e.g., Mastin, 1988), any attempt to correlate the area or volume of broken rock to the stress magnitude will require a full understanding of the failure mechanism(s) responsible for the final borehole geometry (Detournay and Roegiers, 1986).