The Provence Unit (part of Charbonnages de France group), located in the south-east of France, works a coal deposit at more than 1000 metres depth. On the 12/10/93, the RC2 pillar, located in the shaft protection area, was affected by a rockburst causing important damages. The mine management ordered several studies to research mechanisms behind this rockburst so as to identify the existant pillars which could be the object of the same phenomenon and then, to determine the geometric configurations to avoid. The calculations were realized with the SUIT3D code which uses the Boundary Element Method. These calculations had for objective the localization of zones with high stresses. Others calculations using the bidimensionnal FLAC code had for objective to explain the rupture mechanism of the RC2 pillar, then to determine the level of stresses acceptable by a pillar according to its dimensions. During 1994, a tridimensionnal modelisation with the help of the 3DEC code was realized. The object of this calculation was to understand the effects on the pillars of the winning of a nearby face and the influence of a fault crossing both the face panel and the shaft area. Besides, since the end of 1993, the shaft pillar area is monitored by means of deformation measurements, by pressure cells and by electro-magnetic measurements.
The Provence mine works a coal seam at great depth. The workings induce seismic activity in the colliery and tremors of various magnitude are recorded [1]). Substantial events (2.4) are often associated with rockburst and threaten the safety of the mine [2]. In 1993, an event of magnitude 2.67 caused substantial damage in a particularly sensitive zone. The pillar (RC2) located in the main shaft protection area was a violence rupture. The nearest face in activity was T07, at about 1000 metres from RC2 (figure 1). The mine management are concerned about the origins of such phenomena, their links with the mine workings and the conditions in which they might recur. The management undertook two types of action in order first to explain this type of phenomenon and, secondly, try to introduce predictive systems. To begin with, a number of experimental measurements were made: - determination of the most highly loaded pillars by means of test holes. - measurements of pressure and stress variations in the pillar by BOM cells; - measurements of deformation (expansion, convergence) in the roofs and walls of galleries; - stress measurements were used to determine which pillar must be treated (reinforcement by wood pillars,and destressing by sloting the floor) [3]. By analysing all the modelling processes it should be possible to identify the causes of the phenomenon which affected pillar RC2 and to understand the mechanisms leading to such a failure. Conversely, questions may be asked about the stability of other pillars in the shaft zone, with a view to taking preventive action. For the purposes of the numerical simulations, we considered the roof and wall as homogeneous, and isotropic.
