Detection of fracture dilatancy on the cliff top using the azimuthal apparent resistivity technique

Hard rock cliffs erode through an initial catastrophic collapse along pre-existing discontinuities in the rock mass. These may be ancient faults or fractures, orientated at a variety of angles to the cliff face, or relatively new tension fractures formed during cycles of cliff recession, sub-parallel to the cliff face. It is likely that an approaching cliff fall will be associated with increasing fracture dilatancy within the fracture network. Hence if the change of dilatancy can be measured then it may be possible to generate alerts of impending cliff collapse. Since fractures often occur in sets with a preferred orientation they impose anisotropic physical properties on the rock mass. Hence, the apparent resistivity of the rock will vary with azimuth reflecting the dominant fracture orientation. Measures of anisotropy can be calculated from the measurements and would be expected to vary with time if the fractures are dilating. Work package one of the 5th Framework co-funded project ‘PROTECT’ (PRediction Of The Erosion of Cliffed Terrains) was to detect fracture dilatancy. Azimuthal apparent resistivity data were collected at five research sites in the UK, France and Denmark, all situated on outcropping chalk. At each research site, data were collected with the Square array at three locations near the cliff edge and at a Control site set back from the cliff edge by about 50 m. Data were collected approximately every two months for two years to create a temporal data set. After processing the data to remove the effect of the infinite resistance afforded by the cliff face, the data were fitted to an ellipse in order to test for anisotropy. Measures of anisotropy were then calculated from these data. The anisotropy has been interpreted as fracturing and indicates a number of tectonic fracture orientations that agree with geological mapping. At several of the research sites a cliff-parallel fracture set was identified in a zone 10 to 20 m wide adjacent to the cliff edge. It is assumed that this fracture set develops in response to the stress relief at the cliff face. At the Birling Gap research site a cliff collapse within the zone of resistivity measurements produced a dramatic drop in the magnitude of the post-collapse calculated measures of anisotropy. However, other cliff falls that occurred outside of the immediate zone of resistivity measurements did not generate appreciable changes in the calculated measures of anisotropy. It appears that the tectonic fractures that limit the lateral extent of the cliff fall may also limit the fracture dilatancy within the cliff parallel fracture set. At some sites there was a seasonal variation in the measures of anisotropy with peaks in the summer and troughs in the winter. It appears that the most likely driver for these variations is rock temperature that is itself controlled by the external air temperature. Overall, the research has been successful in establishing that there are measurable changes in the rock mass prior to a collapse. However, the methodology is not yet advanced enough to be able to develop technology for the reliable early warning of a cliff fall. The next stage of any research would be to install a system for continuous monitoring in order to establish the magnitude of the changes in the measures of anisotropy immediately prior to a cliff collapse.