A MODIFIED FRACTURE GRADIENT RELATION AND ITS APPLICATION TO EAST TEXAS AND THE TIMOR SEA

Abstract

Theoretical fracture gradient relations are generally based on the assumption that the sedimentary sequence behaves elastically under conditions of lateral constraint. Hence the minimum horizontal stress (σ_hmin) is given by:

where V is Poisson's ratio, σ_v is overburden stress, p_p is pore pressure, and at is far -field tectonic stress. In driling practice, fracture initiation, or leak -off pressures, which are related to σ_hmin are most commonly predicted by the application of empirical stress /depth relations such as that proposed for offshore Western Australia by Vuckovic (1989):

Leak -off pressure (psi) = 0.197D¹¹⁴⁵, where D is depth in feet.

A modified form of the uniaxial elastic relation for the prediction of σ_hmin is proposed, such that:

where the constants c and d are straight line regression constants derived from cross -plotting effective minimum horizontal stress and effective vertical stress. This relation, as opposed to previous empirical approaches to fracture gradient /σ_hmin determination, yields regression coefficients of physical significance: c represents the average Poisson's ratio term, v /(1 -v), and d represents an estimate of the tectonic (and inelastic) component of the minimum horizontal stress. This application of the modified fracture gradient relation, termed the effective stress cross -plot method, is tested successfully against published data from experimental wells in the East Texas Basin where independent estimates of Poisson's ratio are available.

Leak -off pressures have been compiled from 61 wells in the Timor Sea. Leak -off pressures in the Timor Sea are somewhat lower than predicted by Vuckovic's (1989) stress /depth relation for offshore Western Australia, and a new, empirical stress /depth relation, which better fits the Timor Sea data is proposed:

The effective stress cross -plot method is also applied to the Timor Sea data, yielding:

Detailed pore pressure data were not available for the Timor Sea data -set and the effective stress cross -plot method does not fit the observed data any better than the new empirical stress /depth relation. However, the regression constants suggest an average Poisson's ratio of 0.26 and a relatively insignificant tectonic stress of 1 MPa for the Timor Sea.