D. Rhett

Phillips Petroleum Company, Bartlesville, Oklahoma

ABSTRACT: Pore Pressure Controls on the Origin of Regional Fractures: Experimental Verification of a Model

Regional fractures have intrigued geoscientists for many decades. These fractures commonly occur as bedding normal joints and/or (conjugate) shear fractures in essentially undeformed, flatlying, sedimentary strata, and they may occur over large regions with remarkable persistence and little variation in strike. Regional fractures frequently show little or no shear displacement and often display surface textural features that suggest they are dominantly opening mode fractures.

The mechanism of formation of regional fractures has been a topic of considerable interest and much debate for many years. Two dominant theories addressing mechanisms of formation of regional fractures appear in recent geotechnical literature. The "hydraulic fracturing" model (Pollard and Aydin, 1988) contends that regional joints initiate when tensile stresses develop around local flaws in rock subjected to farfield compressive stresses, and they propagate as hydraulic fractures when the pressure of pore fluids within the fracture exceeds the minimum horizontal stress. Conceptual problems arise, however, when one must provide for the numerous local sources of high pressure pore fluids to drive each fracture. One must also reconcile the propagation mechanism with the fact that pore fluids leak off through the fracture surfaces; fluid loss due to leakoff through the two fracture faces increases with fracture length.

A second model of joint development, the "extension fracture" mechanism (Lorenz, Teufel, and Warpinski, 1991), also relies on high pore pressures for fracture formation. According to this model, increasing pore fluid pressure throughout the rock mass results in reduction of all three effective principal stresses according to the Terzaghi Effective Stress Law. As pore pressure increases, the minimum horizontal effective stress, which is the minimum effective confining pressure on the rock, approaches very low values before the other two. At very low effective confinement, the rock mass is essentially unconfined and the effective overburden stress may be sufficiently great to induce brittle failure of the rock. The resulting fractures are "axial splitting" fractures (Gramberg, 1965) and are dominantly opening mode fractures that form perpendicular to the minimum horizontal stress direction. Results from many uniaxial strain/repressuring tests of several different lithologies, conducted under simulated in-situ conditions of temperature, fluid saturations, pore pressures, and loads, support the "extension fracture" mechanism of regional fracture and joint development.

AAPG Search and Discovery Article #90906©2001 AAPG Annual Convention, Denver, Colorado