Evidence from several ChevronTexaco deepwater wells suggests that reduction of fracture gradients due to cooling around the wellbore may be a major contributor to lost circulation events causing millions of dollars in additional well costs. A field test of the magnitude of the effect of temperature on the fracture gradient was performed in a shallow sand/shale sequence in a South Texas well. Pre -test modeling of thermal effects using a simple elastic model predicted that an increase in mud temperature of 33°C would result in an increase in the minimum stress of approximately 3.1 MPa at a depth of 915 m. A series of leak off tests was performed at different mud temperatures to test the predicted effect of temperature on the fracture gradient. The tests showed that an increase in the mud temperature of 33°C caused an increase in the fracture gradient of 1.0 MPa at the test depth. This result is much less than predicted by the elastic model and about half the effect predicted by a thermoporoelastic model. Differences could be due to experimental conditions, difficulty in obtaining accurate values for many model parameters, or inadequate model boundary conditions.
1. INTRODUCTION
An investigation of historical lost circulation events, particularly in deepwater environments, led to the consideration of thermal effects on formation stress as a possible cause. The evidence supported the possibility that high differential temperatures between the drilling mud and the formations might be causing high thermally- induced tensile stresses, reducing the formation breakdown pressure sufficiently to allow the formation of open fractures at the mud pressures that were used to drill the ells.
In order to test the theoretical predictions of temperature effects on near-wellbore formation breakdown pressures, a series of three leak off tests was performed at different well temperatures in an onshore well in south Texas. The goal of the tests was to determine if temperature had a significant effect on the breakdown pressure by directly measuring the magnitude of the change over a temperature range of 33°C (60°F). This paper presents the results of the experiment along with results of attempts to model the measurements using a linear elastic model with thermal stresses and a thermoporoelastic model.
2. EFFECT OF TEMPERATURE ON STRESS
2.1. Previous Published Work
The basic concepts relating to the effect of temperature on stresses around a cylindrical borehole have been known for many years. Timoshenko and Goodier [1] described the effect of thermal stresses around an infinitely long cylinder containing a circular hole in 1951. In it, they showed that heating the inner wall of the cylinder will result in an increase in the compressive forces around the hole. Perkins and Gonzales [2, 3] showed that injecting large volumes of liquid that is colder than the in-situ reservoir temperature can significantly reduce the fracturing pressures in the formation. Tang and Luo [4] presented model predictions of the effect on the near-wellbore stresses of differences in temperature between the mud in the wellbore and the formation.
More recent models which combine thermal effects with poroelasticity have been presented by several authors [5-8].
