Effect Of Stress Anisotropy On Borehole Failure Initiation And Extent - A Laboratory Experimental Observation

Abstract

It is a common knowledge that failure of an openhole is affected by anisotropy of the two far field principal stresses acting perpendicular to the borehole, and to a less extent, by the principal stress acting parallel to the borehole. This knowledge is utilized in designing oil and gas well trajectory, wherever possible, to orient the borehole in the direction that minimizes the stress anisotropy. Linear elasticity, coupled with a strength criterion, such as Mohr – Coulomb strength criterion, is routinely applied to assess borehole stability and sand production. It is well known that such a simplistic model is almost always overly conservative and requires field (preferred) or laboratory calibration. The calibrated model is then applied to the boreholes oriented in the other directions. For example, for borehole stability analyses, the model would be firstly calibrated on a vertical exploration well, and the calibrated model is then applied to analyse borehole stability for deviated or horizontal development wells, taking into account other geomechanics factors likely to have an impact on the stability of the borehole, such as rock formation anisotropy. In general, formation rock mechanical behaviour is far more complex than that the idealistic linear elasticity can describe. Although model calibration will simplify most of this complexity, it remains questionable how reliable such a calibrated model is when applied to the boreholes with different orientations, hence, different stress anisotropies. Field evidence suggests that under a normal fault stress regime, the stability of horizontal well is not very sensitive to well orientation (Morita 2004).

To evaluate the effect of stress anisotropy on borehole stability, we performed a comprehensive literature review on available laboratory borehole stability experiments using true triaxial cells. In comparison with field observation, the advantage of laboratory experiments is obvious; the properties of the rock specimen can be well characterized, the boundary stresses are accurately controlled, borehole conditions are monitored, and failure initiation and extent can be accurately determined. The rock types ranged from weakly consolidated sandstones to competent sandstones and limestones. The specimen size was variable, most ranged from 80mm to 150mm cube with a central borehole. The specimen size to borehole diameter ratio was around 4 or greater. Failure initiation condition was determined either by visual observation or by borehole deformation measurements. Borehole failure extent was either evaluated by post-mortem CT scan or borehole camera. Each of these experimentally determined borehole failure initiation condition is compared with Kirsch solution coupled with the Mohr-Coulomb strength criterion.

The experimental results confirmed that the linear elastic model is overly conservative, in particular for the weakly consolidated rock materials. The trend between the borehole failure initiation condition and stress anisotropy demonstrated that the linear elastic model exaggerated the effect of stress anisotropy on borehole stability, which is dominated by the far field major principal stress perpendicular to the borehole. The weaker the rock specimen material, the less sensitive the borehole failure to stress anisotropy. Furthermore, borehole failure extent is more sensitive to the mean stress of the two far field principal stresses than to the stress anisotropy. These experimental observations shed some light on application of the linear elastic model to borehole stability analyses and sand production prediction.

AAPG Datapages/Search and Discovery Article #90324 © 2018 AAPG Asia Pacific Region GTW, Pore Pressure & Geomechanics: From Exploration to Abandonment, Perth, Australia, June 6-7, 2018