R.D. Barree, Barree & Associates; M.W. Conway, Stim-Lab Division of Core Laboratories;
J.V. Gilbert, Barree & Associates; and R.A. Woodroof, ProTechnics Division of
For years, uncontrolled fracture heights have been modeled for layered reservoirs
using Mode-I tensile failure driven threedimensional (3D) and pseudo-three-dimensional
(P3D) fracture geometry models. In these models stress contrast between layers has
been assumed to provide the dominant height containment mechanism. Meanwhile, many
of the completion diagnostic tools were revealing more contained fracture heights
and were thus being discounted. As more diagnostic tools have been employed, it
has become apparent that the frac models may have been incorrectly over-predicting
fracture heights in many cases. Mine-backs and post-frac coring revealed more contained
fracture heights, as did adioactive tracer logs, treating pressure profiles, fall-off
data, production modeling, and the absence of perforated interval communication.
This paper will present some of the diagnostic data that support more contained
fracture height development. These will include case histories from tight gas sands
and shales in which restricted fracture height observations are supported by diagnostic
observations. It will also go one-step further, by proposing mechanisms which contribute
to fracture height containment. These include poroelasticity, abnormally high "trapped"
pore pressure, and complex shear in layered or anisotropic media. It will present
some modeling methodologies that place more emphasis on poroelasticity and Mode-II
shear failure than current conventional fracture models. The need for investigation,
and adoption, of these methodologies will be corroborated by diagnostic data.