Measurement-related uncertainties in overcoring data at the Äspö HRL, Sweden. Part 2: Biaxial tests of CSIRO HI overcore samples

This paper is the second dealing with measurement-related uncertainties of overcoring data undertaken at the Äspö Hard Rock Laboratory and focuses on the biaxial test data from CSIRO HI overcore samples. The first paper dealt with measurement-related uncertainties in connection with the overcoring phase [1].The uncertainties identified in connection to biaxial tests of CSIRO HI overcore samples include too large applied pressures and poor sampling frequency. At the Äspö HRL, the results yield that most overcore samples fractured during biaxial testing, meaning that a significant part, 56%, of available strain gauge combinations were removed from calculations of the elastic parameters. Remaining strain gauge combinations indicate average values of 62±5 GPa for Young's modulus and 0.25±0.01 for Poisson's ratio, which are considerably lower than previously published values [2–5], and are in good agreement with results from biaxial tests on Borre Probe overcore samples [6–10]. The stress calculations were obtained from re-analyzed elastic parameters and strains, and show primarily a reduction in stress magnitudes. Overall, the stress field obtained with different stress measurement methods and its variation with depth is now quite well resolved. The overcoring data suggest that the principal stresses are inclined with a vertical component dipping about 65° from the horizontal over the investigated rock volume (140–420 m depth). This is interpreted as a result of influence from the sub-vertical NE-2 Fracture Zone that divides the stress data into two stress domains [11], although it may also be an artefact because the σ2σ2- and σ3σ3-magnitudes are of the same order of magnitude.