Geostatistical Evaluation of Rock-Quality Designation and its link with Linear Fracture Frequency

Serge Antoine Séguret (1) , Cristian Guajardo Moreno (2)

1 GEOSCIENCES - Centre de Géosciences
2 Codelco

Rock Quality Designation (RQD) is an important attribute used in geotechnics for quantifying the rock quality. It measures the borehole core recovery percentage incorporating only pieces of solid core that are longer than 100 mm measured along the centerline of the core. The presentation examines the behavior of this attribute in a Chilean porphyry copper deposit by analyzing more than 60,000 1.5-meter long samples. The drill holes have different directions and the nature of RQD requires accounting for the sample direction if the fracture network is anisotropic, a concept different from the geostatistical anisotropy which measures the variability along a direction set by two samples. A directional analysis is conducted and shows different variograms associated with different sample direction classes. This leads to different maps, calling into question the usual practices which do not account for the sample direction. The second part of the presentation concerns the link with the linear Fracture Frequency (FF), another important attribute which measures the number of discontinuities per meter. Under the assumption that the discontinuities along a line follow a Poisson process, Priest &Hudson established in 1976 a formula which expresses RQD as a function of FF. This formula is compared to E[RQF|FF], the mathematical expectation of RQD given FF, deduced from the data. The result of the comparison depends on whether FF is or is not corrected by the sinus of the angle between the sample direction and the fracture, as recommended by Terzaghi in 1965. When applied to FF, this correction breaks a natural correlation with RQD which appears when no correlation is applied. In the latter case, the Priest & Hudson formula is acceptable, in the first case it is not. So there is a dilemma: on the one hand, Terzaghi looks necessary to correctly calculate FF, on the other, the correction systematically increases FF and reduces the relative influence of RQD when both attributes are incorporated into an overall rating like the Rock Mass Rating (RMR), for example. A discussion follows, pointing out that RQD is subject to the same directional bias as FF and should be corrected in the same way by a sinus of the angle between the sample and the fracture, but such a correction is difficult, if not impossible; tests are presented. Finally, a correction of RQD is proposed, based on the Priest & Hudson formula.

Mots clés: Mining; geostatistics; geotechnic; RQD; FF; RMR

Domaines: Sciences de la Terre; Mathématiques appliquées

Serge Antoine Séguret, Cristian Guajardo Moreno. Geostatistical Evaluation of Rock-Quality Designation and its link with Linear Fracture Frequency. IAMG 2015, International Association of Mathematical Geosciences, Sep 2015, Freiberg, Germany. ⟨hal-01187731⟩