Rock Physics

These are the properties that are routinely measured on the reservoir or crustal scale. Thus, we are able directly to link the changes in physical and transport properties to the state of stress and deformation in the rock under carefully controlled conditions. This cannot be done at the reservoir or crustal scale. Thus, we are able to provide insights into improved interpretation of reservoir and crustal scale geophysical data.

CREEP: Long-term time-dependent rock deformation in the laboratory and in a deep-sea observatory

Water-saturated rocks are ubiquitous in Earth’s upper crust and the chemistry of water-rock interactions leads to time-dependent deformation through the mechanism of stress corrosion that allows rocks to fail over extended periods of time at stresses far below their short-term strength. Recent theoretical models, based on mean-field damage mechanics, suggest that the growth and interaction of cracks leads to acceleration to failure once some critical damage threshold has been reached. Recent experimental observations in the Rock Physics laboratory at UCL (Baud & Meredith, 1997).support this approach.

We are therefore investigating this process by conducting constant stress, brittle creep experiments in our triaxial deformation apparatus. Because the process is highly non-linear, it is necessary to run experiments lasting from a few tens on minutes to a few tens of days in the laboratory.

In order to extend the range of observation even further, we would need to conduct experiments with durations of months or even years. Clearly, such experiments are not practicable in a normal laboratory. We have therefore constructed an apparatus to allow us to conduct ultra-long-term experiments at 2km depth in a deep-sea observatory in the Ionian Sea off Sicily. The main reason for considering the deep-sea environment is its stability; constant pressure and temperature throughout the year. The constant pressure at depth provides both the confining pressure and the constant creep stress (via a pressure intensifier).

To date, we have run experiments up to 6 months duration and achieved strain rates in the range 10-11 s-1; some two orders of magnitude lower than our slowest laboratory experiments. Importantly, these rates encompass the time-scales typical of the periods of precursory activity that precede major crustal failure events, and they bridge the gap between laboratory strain rates and those typical of the crust.

CREEP Data

Compaction bands are newly recognized but ubiquitous feature of the deformation of high porosity sedimentary rocks. They are generally manifested as densely packed, poorly sorted, low porosity bands in otherwise well sorted, high porosity rocks. They are important because they occur as low permeability bands in generally high permeability rocks, and thus act to compartmentalize reservoirs and aquifers.

Anisotropy of pore space is also a ubiquitous feature of sedimentary rocks, which leads to an anisotropy of the physical, transport and mechanical properties. We are therefore studying compaction band development in Diemelstadt sandstone; a bedded, anisotropic sandstone with an initial porosity of abut 23% and a mean grain size of about 0.3 mm .

We have quantified the pore-space anisotropy of Diemelstadt sandstone by measuring radial elastic S and P wave velocities as a function of azimuth around cores samples taken in three orthogonal directions. Fluid permeability has also been measured along these three principal directions and the technique of pAMS (Benson et al., 2003) has been used to quantify the anisotropic pore space geometry. We find a velocity anisotropy of around 10% for P-waves and 5% for S-waves, lower permeability normal to bedding, and a mean pore space geometry approximating to an oblate spheroid.

With the pore-space anisotropy established, we characterised its influence on mode of failure by performing triaxial deformation experiments on samples cored normal and parallel to bedding. Both orientations demonstrated a transition from brittle faulting at low effective pressure to the growth of discrete compaction bands at higher effective pressure. We found that the compactive yield envelope for the bedding-normal samples expanded more towards higher stress values than the envelope for bedding-parallel samples.

We find that compaction bands formed both normal and parallel to bedding act as barriers to flow and reduce bulk permeability, consistent with previous work (Vajdova et al., 2004). Microstructural comparisons of the discrete compaction bands formed parallel vs normal to bedding show that those formed normal are more tortuous and less extensive than those formed parallel.

A key consequence of the presence of microcracks within rock is their significant influence upon elastic anisotropy and transport properties. In this study, two microcracked rock types (a basalt and a granite) with known but contrasting microstructures have been investigated using an advanced experimental setup to measure sample porosity, P-wave velocity, S-wave velocity and permeability contemporaneously for effective pressures up to 100 MPa (Rock Fluids Laboratory).

Using the Kachanov (1994) non-interactive effective medium theory, the laboratory measured elastic wave velocities are inverted using a least square fit, allowing us to evaluate the evolution of crack density and crack aspect ratio with increasing isostatic pressure. Overall, the agreement between measured and predicted velocities is good, with average error less than 0.05 km/s. At larger scales above the percolation threshold, macroscopic fluid flow also depends on the crack density and aspect ratio. Using the permeability model of Guéguen & Dienes (1989) and the crack density and aspect ratio recovered from the elastic wave velocity inversion, we successfully predict the evolution of permeability with pressure for direct comparison to the laboratory permeability results. In addition, we calculate the influence of the crack porosity element, also based upon crack density and evolution, for direct comparison to the experimentally measured porosity change with isostatic pressure. These combined modelling/experimental results illustrate the importance of understanding the details of how the rock microstructure can change in response to an external stimulus in order to predict the common evolutions of rock physical properties.

Click Here For Experimental Data