The advantages of well-planned stress measurement campaigns as a pre-excavation design tool are demonstrated through a case history involving the design of two large underground caverns for liquefied gas storage at relatively shallow depth under a foothill. By conducting measurements in two properly located test holes, and at elevations within but also above and below the planned cavern level, a variable stress regime was discovered that led to recommendations for drastic changes in the cavern design with respect to both depth of excavation and orientation of cavern axes.
The in situ stress regime is a critical rock parameter required for rationally designing underground openings in rock. It provides the initial conditions from which the actual stresses acting in the vicinity of the future excavation are derived. It plays a crucial role in each major design step such as selection of the most suitable location, cavern orientation and dimensions, cross sectional shape, excavation sequence, and estimation of the amount of support and reinforcement (Broch, 1981, Haimson, 1984). Before the advent of the hydraulic fracturing method, stress measurements were limited to overcoring and flat-jack techniques, which can typically be conducted only from the excavation itself because of their relatively short reach. Any unexpected stress condition revealed by such tests can be very costly if design changes have to be implemented after excavation is already underway. The hydraulic fracturing method of in situ stress determination has enabled measurements in deep boreholes, which could be drilled from the surface prior to excavation. Advance knowledge of the stress conditions brought about a revolutionary improvement in the planning of stable openings in that rational design could now be undertaken in the pre-excavation stage, avoiding major expenses resulting from potential remedial. I measures necessary when some or all of the underground works are already in existence (Haimson, 1977, 1978, 1992). Rock stress measurements are costly, and the extent of hydraulic fracturing tests in deep holes (reaching the depth of the future openings) is often closely scrutinized by management to minimize expenses. The result is that occasionally the number and depth of test holes and that of tests per hole are reduced resulting in insufficient information on the stress conditions throughout the future excavation. This may end up costing substantially more that the expense entailed it carrying out a well designed in situ stress characterization (Haimson, 1992). For Some years we have been recommending that pre-excavation in situ stress measurements in mountainous areas be properly designed so as to take in consideration possible changes in the stress regime within the project boundaries due to topographic or/and tectonic variations. For example, running all measurements in just one test hole may not be sufficient in some situations (Haimson, 1992). In this paper, we describe in some detail a case history in which well-designed stress measurements prior to excavation led to important recommendations regarding the proper depth and orientation of planned underground caverns.