The influence of fracture density and burial depth on the static and dynamic elastic properties of crystalline rocks

Fracture in rock is a major factor that affects the rock's elastic properties. Elastic properties can be measured statically where stress and strain data are recorded during slow loading of a specimen, or dynamically, where the elasticity can be calculated from P- and S- wave velocity. During crustal deformation, rocks deform nearly statically, hence the relationship between the static and dynamic elastic properties must be known so that the dynamic elastic properties can be converted to static elastic properties to allow geomechanical and geodynamic modelling. In this study, the dynamic and static elastic properties were measured for dry crystalline rocks (Westerly granite) that were thermally treated to 250, 450, 650 and 850°C. Increasing the temperature produces an increased fracture density that is isotropically distributed. Experiments were carried out under confining pressure up to crack-closure pressure, 130MPa (~8km burial depth under hydrostatic pore pressure conditions). Increased fracture density within the rock results in a reduction in Young modulus and an increase in the Poisson's ratio, in both the static and dynamic case which is very significant above 573°C. The reduction and increase are retarded with increasing confining pressure. At crack-closure pressure the fracture density, in terms of effective medium models, is zero even though the rock still contains cracks. The crack-closure pressure is independent of fracture damage incurred in the rock. We compared the static and dynamic measurements and found a linear relationship between the static and dynamic Young's modulus with very high correlation and a gradient of one which is independent of confining pressure and the amount of fracturing incurred in the samples from thermal treatment. We also found that the static and dynamic Poisson's ratio are in agreement for values less than 0.34. Above this value, the static Poisson's ratio is much higher than the dynamic Poisson's ratio. Voigt-Reuss-Hill averaging technique, which inherently assumes zero crack density, was used to calculate the elastic properties so that a comparison could be made with what we assumed are zero crack density at 130MPa. The results from the Voigt-Reuss-Hill averaging technique are similar to the dynamic elastic properties at crack-closure pressure. Hence the dynamic elastic properties at this pressure could be incorporated into an effective medium model to predict the crack densities at pressure below the crack closure pressure. Changes in crack density are interpreted as creation or opening of cracks. We found that the crack density increased from ~0.2 to 1.8 at 10MPa confining pressure. The increase in the crack density is also retarded with increasing confining pressure.