Preliminary Guidelines For Construction Of Large, Deep Underground Caverns

Introduction Caverns of 50 m in linear dimension at depths below 2300 m with a life of more than 25 years are proposed for the next generation of water-filled neutrino detectors. Little precedent exists for excavations in such an extreme environment, although some guidance is available from mining experience at great depths, for example, in South Africa, in Canada and at the former Homestake Mine in the U.S.A. Unlike mining for ore, there are choices available in the construction of such caverns that may mean the difference between success and failure. These choices should be informed by rock engineering principles. Recently proposed guidelines for orientation of brick-shaped caverns where the long dimension is a small multiple of the short dimension indicate that the most favorable orientation is with the long and short dimensions parallel to major and minor preexcavation principal stresses (Pariseau, 2005). Numerical analyses at reduced strength indicate some yielding at cavern walls should be expected but that an equilibrium state is quite possible and is aided by arching the back into a spherical dome when the cavern is an empty cylinder (Tesarik and others, 2002). The same indications of stability occur with brick-shaped caverns. However, the most dangerous situation occurs during the excavation sequence to final cavern dimensions. There are any number of excavation sequences that lead to the same final cavern shape, so there is an opportunity to partially optimize the sequence with construction guidelines that minimize peak stress concentrations during construction. This contribution discusses three practical excavation sequences that lead to the same brick-shaped cavern in the strong, massive Yates formation at the Homestake Mine where considerable rock mechanics data has accumulated through the years(e.g., Pariseau, 1985, Pariseau and others, 1995) and where actual mining on the 8000 Level indicates a strong possibility of successful investment in the next generation of large, deep neutrino detectors. The main results are: (1) a favorable brick-shape is one with dimensions in the same proportion as the preexcavation principal stresses (compressionpositive), and (2) favorable orientation of the most extreme but temporary shape of excavation is also best for the construction sequence. Analysis results from a mix of Homestake Mine data and possible excavation geometry illustrate the advantages of these guidelines.