Microcracks in the Earth's crust

The presence or absence of cracks within in situ crustal rocks is open to wide misunderstandings because of the inaccessibility of the material and the difficulty of reproducing in situ conditions in the laboratory. There is now evidence from a wide range of results that most crustal rocks are pervaded by aligned liquid-filled microcracks, where the liquid is usually water but may be oil in hydrocarbon reservoirs. The recognition that there are aligned liquid-filled microcracks deep within the crust ties together a number of previously contradictory phenomena. Liquid-filled cracks in the Earth's crust, although generally recognised as being present (Brace 1972, 1980), are little understood and the implications of their behaviour have not been explored. This is partly because it is impossible to reproduce in laboratory conditions more than a few of the large range of independent phenomena controlling the existence and behaviour of cracks in rock in situ. The liquid in these cracks is usually a water solution but may be oil in hydrocarbon reservoirs. The other major source of misunderstandings about cracks in the crust is that the principal technique for examining the properties of rocks at depth in the crust has been the analysis of travel times of seismic bodywaves, and reflection and refraction experiments by the exploration industry yield consistent structural interpretations without any need to assume the existence of widespread cracking. We now recognise that this is because almost all seismic experiments in the past have used P-waves, and P-wave travel-times are very little affected by liquid-filled cracks with low aspect-ratios. In contrast, shear-wave splitting (shear-wave birefringence) is very sensitive to distributions of aligned cracks (Crampin 1978). Recent observations of shear-waves (Crampin et al. 1980; Crampin 1985a) suggest that parallel, vertical, water-filled micro cracks pervade much of the brittle, upper 10-20 km of the crust. We call such distributions of aligned fluid-filled cracks extensive dilatancy anisotropy (EDA) (Crampin, Evans and Atkinson 1984). EDA is a unifying concept that allows a variety of phenomena from geology, geophysics and the rock-mechanics laboratory to be correlated for the first time. The existence of such cracks has wide implications for all deformatory processes in the crust, and the ability to monitor crack geometry by shear-wave propagation has applications to many currently important activities ranging from determining preferred directions of flow in hydrocarbon reservoirs to earthquake prediction.