In the first part, it is explained how the notion of scale effect, the data acquisition and statistic analysis as well as the characterization of the morphology of a fracture under stress, the assessment of interactions between structure, stress and fluid flow and the knowledge of the joint filling can be employed to improve our understanding of rock discontinuities. At the beginning of the second part, the recent developments of three-dimensional modelling of a single and a fractured medium are presented. Then, it is underlined that the special distribution of discontinuities can be deduced from geophysical surveys.
Predicting the role played by rock discontinuities in fluid migration and stability of structures at great depth is an old challenge presented to scientists by many industrial organizations. These include firms involved in mining and civil engineering as well as those in the production of geothermal energy in nuclear waste storage and oil recovery which have sponsored the research presented at this conference.
In order to understand the behaviour of a discontinuous rock mass, Hoek and Brown (1980) introduced the notion of scale effects. So, a fractured rock with a spacing relatively large compared to the size of the project can be analyzed using a continuum approach, while the modelling of a continuous medium with a single discontinuity in the zone concerned requires the use of the displacement discontinuity method for example. A highly fractured rock can be treated as a continuum with equivalent properties, but in other cases the block interactions must be taken into account by using an approach based on the distinct element method or block theory. When a borehole collapses, thermal or poro-elastic effects induced by the mud circulation. Nevertheless, it is important to consider the scale effects presented above and to distinguish the major discontinuities from the minor ones in order to minimise the risks of borehole collapse. As Maury (1989) underlined in his case history, it is only the block structure of the rock mass that can explain the loss of wellbore Pont d'As 5 drilled at 4500 m depth below Pau. The re-pressurisation of a fault triggered small shear movements, thus weakening the casing so that it could not withstand a decrease of internal pressure. Joints or planes of weakness with about ten centimetres spacing can also lead to instabilities at a borehole wall. Goodman (1976) analyzed the effects of joints by inserting their stiffness and spacing into the elastic matrix of an anisotropic material; Amadei (1983) proposed a general analytical solution for the stress distribution induced by the drilling of a borehole in such a medium subjected to a homogeneous tectonic in-situ stress field and induced by the pressurisation of the borehole. By coupling Goodman's and Amadei 's solution to failure criteria for rocks, Aadnoy (1988) showed that borehole instability can result from certain choices of the borehole inclination and azimuth.
