The ANZI strain cell and the overcoring method of stress relief have been used for in situ stress measurements in a wide range of rock types for more than 30 years. This instrument uses an inflatable packer system to bond strain gauges directly to the rock on the wall of the pilot hole and then test their integrity in situ prior to overcoring using a dilatometer style pressure test. Strains measured on six circumferential strain gauges located evenly around the circumference of the pilot hole have provided direct measurement of a phenomenon that causes the elastic modulus to vary with the first stress invariant; effectively the confinement provided to the rock fabric. Measurements indicate that the elastic modulus in coal measure strata (elastic modulus range 3 GPa to 40 GPa) typically doubles in magnitude from an unconfined state to a maximum at a first stress invariant broadly consistent with the long term state of stress in the rock. This paper presents measurements of this phenomenon from the in situ pressure tests, biaxial pressure tests, and laboratory core tests, and discusses the implications of modulus variation for interpretation of the in situ stresses from the strains measured during overcoring. The significance of this behaviour for modelling stress distributions about excavations in rock more generally is also discussed.
In vertical holes, and horizontal holes aligned with one of the horizontal principal stresses, the measured variation in elastic modulus around the borehole allows orientation of the other two principal stresses to be inferred from the in situ pressure test prior to overcoring. When the ratio of these other two stresses is known or can be estimated with reasonable confidence, the magnitude of the two stresses can also be inferred from the pressure test result. This approach provides confirmation of the stress field, particularly stress orientation, totally independent of the subsequent overcoring result. Agreement between the two builds confidence in the overall stress measurement. The approach is also useful in stress change monitoring applications where an overcore measurement is not conducted because it allows the initial stress state in the rock to be estimated.
