Title of Invention: Statistical method for micro-scale rock damage

quantification and characterisation using x-ray micro-tomography.

Inventors: Penelope Clair Stewart

Affiliation: Petra Data Science Pty Ltd

Background

Micro-scale rock damage, is the creation, dilation and extension of microfractures within a rock matrix. The rock matrix is the intact rock between macrofractures (visible discontinuity planes) within the "rock mass". Rock mechanical behaviour and strength depends upon the properties of both the intact rock matrix, and discontinuity planes. Intact rock damage reduces the mechanical and dynamic strength properties of the rock. In minerals processing engineering, micro-scale rock damage is frequently referred to as "conditioning". Conditioning improves comminution efficiency, and improves heap leaching and in situ leaching metal recoveries. Geotechnical and rock mechanics engineers need to be able to estimate the effect of damage on intact rock strength. Estimates of micro-scale rock damage, or conditioning, are required for the following engineering design applications; geotechnical engineering, minerals processing comminution, blasting damage, rock cutting, rock conditioning, seismic propensity, numerical modelling and leaching of metals from rock.

Current Practice

Current methods for micro-scale rock damage assessment include;

• Geotechnical engineers assess macro-scale rock mass damage in the field for the purpose of estimating an empirical rock mass scale strength reduction factor. For example; these assessments of macro-scale rock mass damage are commonly incorporated into engineering design, through the empirically derived Disturbance factor, D (Hoek, 2012). However, Disturbance factor, D doesn't provide an estimate of micro-scale damage for the purpose of estimating the effect of micro-scale damage, on intact rock matrix strength properties.

• Rock dilation due to damage is a common proxy for rock mass damage, and can be

quantified using the following methods: inter-hole seismic tomography, photogrammetry, high precision survey, ground probing radar and visible changes in the rock structure from geotechnical mapping and core logging (e.g. fractures, joints, discontinuities). Whilst useful for assessing damage on a large scale, rock mass dilation, ground probing radar and visible changes in rock structure (obtained from geotechnical mapping and logging) relate to macro-scale rock mass properties. These methods provide no information as to whether the intact rock is damaged, or not.

• Borehole geophysics and imaging for rock masses (video, acoustic emissions, p-wave

velocity, porosity from density). Geophysical estimates of rock matrix density and fluid density have been used to develop porosity based rock strength correlations for

sedimentary rocks (Chang et al 2006). Borehole acoustic emissions and video are used to detect macro-scale rock structure. P-wave velocity correlations to micro-scale rock damage are associated with high levels of scatter due to other rock properties. Geophysical porosity estimates are derived using assumed rock matrix and fluid densities, and therefore, geophysics can't be used to develop pore size and volume distributions.

• Conventional laboratory vacuum porosity tests, only detect interconnected cracks and pores resulting in lower accuracy than is possible using x-ray tomography. Kilebrandt, Norrgard and Jern (2010) used laboratory vacuum porosities to detect order of magnitude change in porosity, despite high levels of scatter in the porosity data.

• Qualitative micro-scale damage descriptions of rock micro-scale damage are based on 2D petrography analysis and/or 3D CT tomography. Qualitative micro-scale rock damage has, until now, been limited to qualitative descriptions of the type of microdamage observed (Parra and Onederra, 2015), and sampling volumes were too small to be considered representative samples.

• Quantitative microfracture intensity/density has been used to quantify fault related micro- scale damage (Mitchell and Faulkner, 2012). Microfracture intensity/density using thin section petrography (e.g. cathodoluminescence), where the operator counts the number of microfractures to intersect a line. Microfractures counts, include those sealed by geological fluids. In many rocks, quartz sealed microfractures are insignificant structural features owing to quartz sealing, cement deposition, and temperature and pressure related pore closure (Laubach 2003, Laubach and Diaz-Tushman, 2009, Gale et al 2010). Inclusion of structurally insignificant sealed microfractures in microfracture intensity counting methods, reduces the sensitivity of these methods to intact rock damage detection. Because microfracture intensity is counted manually at the grain scale, in practice, it is difficult to justify the cost of obtaining sufficient representative grain-scale samples.

• Acoustic emissions for intact rock samples. Acoustic emissions are routinely used to quantify micro-scale intact rock damage for laboratory testing of core under load, but acoustic emissions instrumentation is not suitable for rocks under load in the field.

Description of I nvention

The Micro-scale Rock Damage Quantification and Characterisation method incorporates the following novel ideas to quantify rock damage:

• Quantifies and characterises damage through statistical analysis of structurally significant open microfractures and pores within rock samples. Specifically, the microfracture (cracks and pores) volume distributions are compared using statistical tests.

• X-ray microtomography scan data is processed using standard image segmentation

algorithms to produce microfracture statistics for rock samples e.g. microfracture size in volume.

• Microfracture statistics are automated using standard proprietary image segmentation algorithms with minimal sample preparation. Sample edge effects are removed digitally using standard proprietary software reducing operator sample preparation time.

• Damage is quantified in terms of the difference between the rock samples' microfracture volume distributions. These distributions may be either a single rock sample, or

microfracture volume statistics from multiple samples may be combined to generate a single distribution to represent the rock.

• Statistical tests are applied to test the significance of micro-scale damage:

o Statistical analysis of cumulative percent pore volume distributions is used to

distinguish damaged rock, from undamaged rock. For example: micro-scale damage is quantified in terms of the Kolmogorov-Smirnov test statistic, D - i.e. the maximum difference between two empirical distributions functions. o Microfracture volume is one of a number of microfracture statistics that could be used to quantify damage. Other microfracature statistics that can be used to quantify damage include pore size, length, and aperture.

o For example: The difference between the cumulative % volume distributions is used to quantify the degree of damage using the Kolmogorov- Smirnov test statistic, D. Empirical distribution functions estimate the true underlying cumulative distribution functions of the points in the sample. Note: The test checks whether the two data samples come from the same distribution. The p-value for this test is the probability that D is greater than the observed value under the null hypothesis of no difference between class levels or samples.

List of Drawings

Figure 1 - Workflow for new statistical method for micro-scale rock damage quantification and characterisation using x-ray micro-tomography.

Figure 2 - X-ray tomography section, showing x-ray micro-tomography 'Region of Interest' inside red outline. The denser the mineral grain, the darker the mineral. Standard x-ray micro-tomography image segmentation algorithms are applied to the Region of Interest for the purpose of obtaining 3D digital rock composition data, including the pore size and volume statistics used in the statistical analysis of the micro-scale rock damage.

Figure 3 - 2D digital rock image of Region of Interest rendered by density (Red open cracks/pores, purple high density veins, green rock matrix). The scale and location of the Region of Interest with respect to the rock sample is shown in Figure 2.

Figure 4 - 2D digital rock image of the Region of Interest showing microfractures/pores obtained from image segmentation. The scale and location of the Region of Interest with respect the rock sample is shown in Figure 2.

Figure 5 - Cumulative % Pore Volume for two rock samples. In this example; the Kolmogorov- Smirnov test statistic, D is the maximum vertical distance between the empirical distribution functions for each sample, and is measure of the amount of rock damage, or conditioning. The empirical distribution function estimates the true underlying cumulative distribution function of the points in each sample.

References

Batzle, M.L and Simmons, G., 1976. Microfractures in rocks from two geothermal areas. Earth Planet. Sci. Lett., Vol.30, p 71-93.

Chang, C, Zoback, M.D., Khaksar, A., 2006. Empirical relations between rock strength and physical properties in sedimentary rocks. Journal of Petroleum Science and Engineering, Vol. 51, p 223-237.

Cox, S.F. and Etheridge, M.A., 1983. Crack-seal fibre growth mechanisms. Tectonophysics, Vol. 92, p 147-170.

Gale, J.F.W., Lander, R.H., Reed, R.M., Laubach, S.E., 2010. Modeling fracture porosity evolution in dolostone, J. Struct. Geol. Vol. 32 (9), p 1201-1211 http://dx.doi.Org/10.1016/j.jsg.2009.04.018. Hoek, E. Martin, CD., 2014. Fracture initiation and propagation in intact rock - A review. Journal of Rock Engineering and Geotechnical Engineering. Vol. 6, p 287-300.

Kranz, 1983. Microcracks in rocks: a review. In: M. Friedman and M.N. Toksoz (Editors), Continental Tectonics: Structure, Kinematics and Dynamics. Tectonophysics, Vol. 100, p 449-480.

Laubach, S.E., 2003. Practical approaches to identifying sealed and open fractures. AAPG Bull. Vol. 87 (4), p 561-579.

Laubach, S.E., Diaz-Tushman, K., 2009. Laurentian paleostress trajectories and ephemeral fracture permeability, Cambrian Eriboll Formation sandstones west of the Moine thrust zone, northwest Scotland. J. Geol. Soc. Lond. Vol. 166 (2), p 340-362.

Parra, H. Onederra, I. Michaux, S. Kuhar, L. McFarlane, A. and Chapman, N., 2015. A study of the impact of blast induced conditioning on leaching performance. Mineral Engineering, Vol. 74. p 1-12.

Mitchell, T.M. and Faulkner, D.R, 2012. Towards the matrix permeability of fault damage zones in low porosity rocks. Earth and Planetary Science Letters, No. 339-340, p 24-31.

Padovani, E.R., Shirly, S.B. and Simmons, G., 1982. Characteristcs of microcracks in amphibolite and granulite facies grade rocks from southeastern Pennsylvania. J. Geophys. Res., Vol. 87, p 8605-8630.

Richter, D. and Simmons, G. 1977. Microcracks in crustal igneous rocks: microscopy. In: J.G. Heacock (Editors), The Earth's Crust. Am. Geophys. Union, Geophysics. Monograph 20. P 149-180.

Shirley, S.B., Batzle, M.L. and Simmons, G., 1978. Microfracture characteristics of geothermal systems. Geol. Soc. Am., Abstr. Programs, Vol 10, p 491-492.

Sprunt, E. and Nurm A., 1979. Microcracking and healing in granites: New evidence from

cathodoluminescence. Science, Vol 205, p 495-497.

Wang, H.F. and Simmons, G., 1978. Microcracks in crystalline rock from 5.3 km depth in the Michigan Basin. J. Geophys. Res., Vol. 83, p 5849-5856.