Spectral analysis of drill core from the Kanmantoo copper deposit
The Kanmantoo copper deposit describes a cluster of eight zones of Cu–Ag–Au mineralisation located ~2.5 km southwest of Kanmantoo township, 41 km southeast of Adelaide. Hydrothermal chalcopyrite–pyrite–pyrrhotite–magnetite mineralisation is concentrated in structurally controlled zones within biotite, quartz, andalusite, chlorite, garnet ± staurolite schist within the western limb of the Kanmantoo syncline. Copper production began in the late 1840s with underground mining of small high-grade lodes, followed by open pit mining in 1970–76 (Verwoerd and Cleghorn 1975) and most recently in 2011 to present, with expanded open pit operations by Hillgrove Resources. Past production and current resource estimates give a metal endowment at Kanmantoo of around 0.35 Mt of copper, 3 Moz of silver and 100 koz of gold (Rolley and Wright 2017). Estimated remaining total mineral resources, as at 31 December 2017, were 31.8 Mt at 0.6% Cu, 0.1 g/t Au, 1.3 g/t Ag, for cutoff grade 0.2% Cu (Hillgrove Resources 2018).
Continuous spectral analysis of selected drill core from various mineralised zones at Kanmantoo was completed recently by the Geological Survey of South Australia in collaboration with Hillgrove Resources to map mineralogy and mineral associations. The results have been used to assist with interpreting proximity to mineralisation. This article provides an overview of the project and findings. The approach may be useful in assessing the potential for further mineralisation within known ore systems at Kanmantoo and offers a means of acquiring high data density for evaluating patterns of hydrothermal activity identified in exploration drill samples from other copper targets in the district.
Geological setting and mineralisation
Copper mineralisation at Kanmantoo is hosted by the Tapanappa Formation within the Cambrian Kanmantoo Group, a 7–8 km thick package of dominantly marine turbidite sediments deposited in an extensional back-arc basin, the Kanmantoo Trough (Haines, Jago and Gum 2001). The sedimentary pile was accumulated over a comparatively short 8 ± 5 Ma period between 522 ± 2 to 514 ± 3 Ma (Foden et al. 2006). The Tapanappa Formation consists of a largely monotonous sequence of immature sandstone (greywacke) with muddy siltstone interbeds and minor pyritic mudstone (Toteff 1999). In addition to the Kanmantoo deposit, the sequence hosts several smaller deposits of Cu–Au mineralisation (e.g. Bremer, South Hill) and scattered occurrences and deposits of Pb–Zn–Ag, some of which have been mined (e.g. Angas, Wheal Ellen, Aclare, Strathalbyn, Scotts Creek) (Belperio et al. 1998; Both 1990; Gum 1998; Seccombe et al. 1985; Spry, Schiller and Ross 1988; Toteff 1999). Much of the Pb–Zn–Ag mineralisation is stratabound and coincident with sites of apparent alteration of the sediment, expressed as zones of quartz–biotite–garnet–andalusite ± staurolite rock that can be traced intermittently for ~30 km from north of Kanmantoo, south towards Strathalbyn (Fig. 1) (Seccombe et al. 1985; Toteff 1999; Pollock et al. 2018). The garnetiferous rocks are broadly anomalous in Pb, Zn, Cu and Mn (Pollock et al. 2018). The patterns of alteration and distribution of mineralisation, particularly Pb–Zn–Ag, in the Tapanappa Formation are interpreted as evidence of synsedimentary hydrothermal exhalative activity with metals deposited from heated basin fluids channelled along growth faults accompanying extension of the basin (Flöttmann et al. 1984; Seccombe et al. 1985; Toteff 1999).
Sedimentation in the Kanmantoo Trough ceased with the onset of the Delamerian Orogeny during the mid-Cambrian (514 Ma). Compressive deformation was accompanied by emplacement of syntectonic I- and S-type granitoids along the eastern Kanmantoo Trough, with A-type granite and mafic and felsic dykes intruded during post-tectonic (490–480 Ma) relaxation and extension (Belperio et al. 1998; Foden et al. 2002). Three phases of deformation are recognised with D1 NW-directed thrusting imparting bedding parallel schistosity that was overprinted by axial plane crenulation cleavage developed during D2 upright open to tight folds with dominantly N–S axes and gentle southerly plunges (Offler and Fleming 1968). Kanmantoo Syncline is a D2 structure that includes an open synformal structure that plunges ~15°S. A parasitic syncline of the Kanmantoo Syncline occupies two-thirds of the main pit at the Kanmantoo mine (Rolley and Wright 2017; Schiller 2000). Later deformation, D3, developed open to tight folds about NW–SE axes, which are rarely evident at Kanmantoo where the D3 event is usually observed only as crenulations and kinks (Schiller 2000). Metamorphism was high temperature (550–600 °C), low pressure (3–5 kb) and reached amphibolite facies in the region of Kanmantoo with peak metamorphism possibly post D2, reflected in iron-rich garnet growth and recrystallisation of biotite in discordant zones that also contain chlorite and sulfides (Rolley and Wright 2017). This is in contrast with pre to early D2 peak metamorphism in the Karinya Syncline in Kanmantoo Group rocks some 80 km to the north of the mine area (Sandiford et al. 1995).
Mineralisation at Kanmantoo is discordant to bedding but is broadly aligned with D2 schistosity on the western synform limb (Schiller 2000). The discordance is explained in sedimentary exhalative models as a subsurface hydrothermal feeder zone with remobilisation of sulfides during metamorphism (Pollock et al. 2018; Spry, Schiller and Ross 1988; Toteff 1999). An alternative model favours emplacement of Cu–Ag–Au mineralisation post peak metamorphism with mineralising fluids introduced by igneous intrusion at depth and circulated along reactivated D2 and crosscutting NNE and NE structures; metal deposition resulted from interaction of fluids with reactive host rocks or a decrease in the thermal gradient (Arbon 2011; Lyons 2012; Oliver et al. 1998; Rolley and Wright 2017; Tedesco 2009).
Despite some lingering controversy with regard to the origin of the Cu–Au–Ag mineralisation at Kanmantoo, the data collected in previous and ongoing studies point to a role for spectral analysis and mineral mapping. In particular, attention was directed to:
- amphibolite-grade metamorphic minerals equated with possible zones of hydrothermally altered sediments (e.g. andalusite, garnet, staurolite, spinel)
- mineral alteration indicative of later hydrothermal fluid interactions (e.g. reduction in andalusite content, change in garnet or biotite composition, crystallisation and composition of chlorite, presence of carbonate or sulfate minerals)
- minerals most closely associated with copper mineralisation (chlorite–sulfides–magnetite).
The objective was to collect and analyse continuous spectral data of drill core to provide a mineralogical model that characterised alteration zoning around copper mineralisation at Kanmantoo, which could be used to assist in improving the effectiveness of subsequent brownfields drilling.
Spectral scanning and analysis
Continuous spectral data from drill core were collected using the HyLogger 3.3 instrument located at the South Australia Drill Core Reference Library at Tonsley. HyLogger 3.3 has a bank of four instruments – camera, visible-shortwave infrared (VSWIR; 400–2,500 nm) spectrometer, thermal infrared (TIR; 6,000–14,500 nm) spectrometer and a laser altimeter – under which core trays are moved on a robotic table. The collation of the data from these four instruments delivers an interactive digital file which enables the identification of a suite of minerals from a well characterised spectral library (Schodlock et al. 2016; Schodlock Green and Huntington 2016).
Fourteen legacy diamond drillholes were selected to encompass the breadth, lateral and vertical extent of the Kanmantoo copper mineralisation (Table 1; Fig. 2).
Table 1 Legacy drillholes selected for analysis from orebodies within the Kanmantoo copper deposit.
|Drillhole|| SA Geodata |
|Orebody or target|
|DDH KAN 2||206089||Distal North|
|DDH KAN 1||206088||Distal South|
* Number will be assigned in first quarter 2019.
Based on previous investigations identifying key mineral components of the deposit that showed spectral responses in the shortwave infrared (SWIR) and TIR the following minerals were selected for more detailed analysis: kaolin/jarosite, white mica, andalusite, garnet (almandine), biotite and chlorite.
Variation in andalusite content was associated with persistent, but minor amounts, of kaolin and jarosite, not previously reported but evident in SWIR data recorded by HyLogger and quantified using the Spectral Assistant (TSA) software in the Spectral Geologist (TSG) software (Schodlock et al. 2016). The significance of kaolin/jarosite was further investigated using scanning electron microscopy (SEM) on selected samples from KTDD149. Fragments of drill core were mounted on aluminium stubs using araldite and coated with carbon. These were examined at Adelaide Microscopy on a FE Quanta 600 SEM with energy dispersive X-ray analytical facility.
For white mica, the identification and relative abundance were calculated using TSA. Variation in the chemistry of the white mica was assessed based on the position of the wavelength minima for the 2,200 nm absorption feature.
Andalusite is a highly visible component in the core samples and forms white poikiloblastic porphyroblasts in biotite schist with varying degrees of coherence. The andalusite grains range from compact euhedral grains up to 2 cm in length overprinting the schistosity to somewhat diffuse masses, grading further to ghost outlines in the core.
Spectral absorption features for andalusite overlap with garnet. In order to map relative abundance of andalusite the TIR spectra were interpreted with the aid of polynomial fitting (PFIT). By focusing on the 10,365 nm absorption feature (Fig. 3) subtle contributions from andalusite were measured using the wavelength position. Andalusite abundance was calculated from the depth of the feature.
Garnet is a complex mineral with a wide range of chemical composition which may be present within a single grain, typically as concentric zones of varying composition. Major element compositions of garnet in the Kanmantoo deposit were measured by McPherson (2017) and results from the Kavanagh orebody are shown in Table 2. The analyses record garnet as dominantly almandine (Fe garnet), with lesser components of pyrope (Mg garnet) and spessartine (Mn garnet). The resolution of the HyLogger spectra (1 cm x 1 cm sample per spectra) means that the TIR spectra will be a composite of those individual species. This provides an opportunity to map geochemical gradients reflected in the change in overall garnet composition, determined by measuring wavelength shifts in the characteristic absorption features. To identify almandine, PFIT was used focusing on the wavelength of the 11,279 nm feature (Fig. 3) and the depth was used to estimate abundance. To measure changing chemistry, the wavelength of the 10,710 nm absorption feature was extracted. This was intended to map the proportion of Fe-garnet almandine (10,710 nm) relative to Mn-garnet spessartine (10,860 nm).
Table 2 Average major element compositions for garnet from Kavanagh orebody (from McPherson 2017).
|Number of atoms in formulae (oxygen basis 12)|
|Type of garnet|
In hot hydrothermal systems, trioctahedral micas may be recrystallised with modified chemistry due to interaction with the hydrothermal fluid. At the Yangyang iron oxide – apatite deposit, South Korea, Kim et al. (2018) report hydrothermally altered biotite with modified SWIR spectral response, where the wavelengths of key absorption features were used to map change in visible colour and Fe:Mg ratio. The abundance of biotite at Kanmantoo offered the opportunity to investigate similar correlations in relation to copper mineralisation. To calculate an appropriate measure of geochemical gradient the SWIR spectra were initially filtered using TSA to select only the dark mica mineral group. Two absorption features of biotite, 2,254 nm and 2,357 nm, appear to move in concert towards longer wavelengths with increasing iron content. The 2,254 nm Fe–OH absorption feature was most sensitive. Consequently, in addition to varying abundance of dark mica, a PFIT calculation on the 2,254 nm feature was used as a proxy to track variation in iron content.
The chlorite investigation used tools provided by TSA in TSG to identify variations in the abundance of Fe-, Fe–Mg- and Mg- chlorites. In addition, the wavelength of the 2,252 nm absorption feature was calculated as a geochemical gradient indicator with an increasing iron content shifting the feature to longer wavelengths (Pontual, Merry and Gamson 1997).
In the Kanmantoo data many of the calculated gradient factors showed a high degree of variance. To clarify the overall trends, a smoothing filter of a moving average over a 5 m interval was applied.
The trends identified from the six mineral species showed systematic changes moving from distal to proximal locations in relation to copper mineralisation. The results of two holes (Fig. 4) were chosen to best illustrate the patterns of mineralogical changes within 300 m of mineralisation.
White mica (not shown) tends to be far-distal to mineralisation – up to 300 m away from copper – and absent closer. Where the chemistry of the white mica tends towards phengitic composition there is some association with gold mineralisation in late stage structures.
Andalusite is present mostly inboard of white mica but remains near-distal to the copper mineralisation, i.e. it is not usually found in close proximity to copper. The distribution of kaolinite appears to be antithetic to andalusite and is not related directly to copper mineralisation.
Electron microscopy of samples from KTDD149 confirmed dissolution and kaolinisation of andalusite was more intensive below the zone of copper mineralisation at ~280–340 m. Small clusters of poorly define jarosite crystals were associated with thin kaolinite coatings on mineral grains throughout the drillhole.
Almandine has greater abundance proximal to mineralisation and the wavelength of the 10,710 nm absorption feature trends to shorter wavelengths closer to copper.
For biotite, both the 2,254 nm (short) and the 2,356 nm (long) absorption features move in concert and copper is associated with longer wavelengths, and with an overall decrease in biotite abundance (Fig. 5).
For chlorite the chemical gradient is towards more iron-rich species being associated with copper mineralisation.
The combination of varying mineral abundance and change in mineral chemistry for key minerals identified from spectral data provides an indication of geochemical gradients that show a correspondence with proximity to copper mineralisation for the Kanmantoo copper orebodies. The absolute values vary from one drillhole to the next but evidence of a geochemical gradient is consistently observed. In hydrothermal systems, chemical change signalled by the presence of a geochemical gradient may be significant in ore forming processes, irrespective of the magnitude of the gradient (Keith Scott, CSIRO, pers. comm., 1997).
In the case of garnet, the shift in wavelength of the 10,710 nm absorption feature suggests a proximal almandine and more distal spessartine composition, although an increase in andalusite content and overlap of spectral features may be a factor in the apparent longer wavelengths of absorption for more distal garnet. Irrespective of the explanation, the gradient remains consistent with longer wavelength features for garnet being distal and shorter proximal. The increase in garnet content proximal to mineralisation appears to be at the expense of andalusite and biotite.
Andalusite dissolution with associated kaolinite precipitation (Fig. 6) is interpreted as the result of a relatively low-temperature acidic fluid, post copper mineralisation. Topotactic crystallisation of kaolinite on biotite, aligned along biotite cleavage (Fig. 7) is consistent with alteration by a late hydrothermal fluid. The relatively minor kaolinite/jarosite alteration is not considered to be part of the copper mineralising system, but may be indirectly related in that the same fluid pathways were accessed and the acidity due to partial oxidation/dissolution of sulfides in this part of the system.
An outcome of the observations described above is a decision tree that incorporates spectral analysis to inform a future drill program (Fig. 8). The two gradients being measured are: (i) relative mineral abundance; and (ii) relative change in mineral chemistry. Moving through the diagram from left to right and top to bottom the presence of white mica places the sample distal to copper mineralisation by the order of 200 m. If the white mica shows a shift in the wavelength of the 2,200 nm feature (2200W) towards phengitic composition, gold may be associated. If white mica is absent but andalusite is present this places the sample in-board of white mica but still distal to mineralisation. In-board of andalusite, almandine garnet, Fe rich and Mn poor, forms proximal to mineralisation. Examining the more pervasive biotite and chlorite species, the shift to longer wavelengths of key absorption features is indicative of iron enrichment. In addition to the change in chemistry, biotite also is usually less abundant proximal to mineralisation.
The spectral mineralogy study of the near-mine environment at the Kanmantoo copper deposit identified geochemical/mineral gradients defined by mineral associations showing consistent patterns proximal to copper mineralisation. Useful minerals for spectral analysis included kaolin/jarosite, white mica, biotite, andalusite, almandine and chlorite.
A geochemical gradient of increasing iron content in biotite, garnet and chlorite most closely correlates with zones of copper mineralisation. Mineralogical changes in proximity to sulfide mineralisation include increased almandine abundance at the expense of andalusite and biotite.
The persistent presence of kaolinite/jarosite alteration is interpreted to result from circulation of a moderately low temperature hydrothermal fluid at comparatively shallow crustal level. This may be associated with the mineralisation, or alternatively is a younger fluid event, but either are consistent with late stage hydrothermal activity of a mineral system that developed post-metamorphism. Spectral results, however, provide few additional insights on the origin of the copper mineralisation.
Continuous spectral data of drill samples using the HyLogging system was shown to be effective in identifying patterns in the mineralogy that have consistent spatial relationship with known copper mineralisation. Consequently, this offers a useful tool that could be applied systematically to the interpretation of future drill sampling in the vicinity of the Kanmantoo Mine.
The HyLogging system refers to a group of instruments developed by CSIRO for hyperspectral measurement and imaging of mineral specimens. The trademark is now owned by Corescan Pty Ltd. TSG and TSA remain the trademarks of CSIRO. National Collaborative Research Infrastructure Strategy (NCRIS) funding, managed by AuScope, enabled the acquisition of the HyLogger 3-3 owned and operated by the Geological Survey of South Australia.
Sam Williams and Georgina Gordon (Geological Survey of South Australia) are gratefully acknowledged for contributing to the acquisition and preliminary processing of the Kanmantoo HyLogger spectral data. Stuart McClure assisted with SEM analyses at Adelaide Microscopy.
Arbon H 2011. Bismuth distribution in the Cu-Au mineralisation of the Kanmantoo deposit, South Australia. Hons thesis, University of Adelaide.
Belperio AP, Preiss WV, Fairclough MC, Gatehouse CG, Gum J, Hough J and Burtt A 1998. Tectonic and metallogenic framework of the Cambrian Stansbury Basin – Kanmantoo Trough, South Australia. AGSO Journal of Australian Geology and Geophysics 17:183–200.
Both RA 1990. Kanmantoo Trough — geology and mineral deposits. In FE Hughes ed., Geology of the mineral deposits of Australia and Papua New Guinea. Australasian Institute of Mining and Metallurgy, Melbourne, pp.1195–1203.
Flöttmann T, James PR, Rogers J and Johnson T 1994. Early Palaeozoic foreland thrusting and basin reactivation at the Palaeo-Pacific margin of southeastern Australia Precambrian craton: a reappraisal of the structural evolution of the southern Adelaide fold-thrust belt. Tectonophysics 234:95–116.
Foden JD, Elburg MA, Dougherty-Page J and Burtt A 2006. The timing and duration of the Delamerian Orogeny: correlation with the Ross Orogen and implications for Gondwana assembly. The Journal of Geology 114:189–210.
Foden JD, Elburg MA, Turner SP, Sandiford M, O’Callaghan J and Mitchell S 2002. Granite production in the Delamerian Orogen, South Australia. Journal of the Geological Society 159:557–575.
Gum J 1998. The sedimentology, sequence stratigraphy and mineralisation of the Silverton Subgroup, South Australia. PhD thesis, University of South Australia, Adelaide.
Haines PW, Jago JB and Gum J 2001. Turbidite deposition in the Cambrian Kanmantoo Group, South Australia. Australian Journal of Earth Sciences 48:465–478.
Hillgrove Resources Limited 2018. Hillgrove Resources annual report for the year ended 31 December 2017. Hillgrove Resources Limited, Adelaide.
Kim YH, Choi S-G, Seo J, Ko K-B, Lee YJ 2018. Application of SWIR spectrometry to the determination of biotite compositions in hydrothermally altered units of the Yangyang iron-oxide-apatite (IOA) deposit, South Korea. Ore Geology Reviews 99:303–313.
Lyons N 2012. Evidence for magmatic hydrothermal mineralisation at Kanmantoo copper deposit, South Australia. Hons thesis, University of Adelaide.
McPherson MV 2017. The origin of the sediment-hosted Kanmantoo Cu-Au deposit, South Australia: mineralogical considerations. MSc Thesis, Iowa State University, Ames, Iowa.
Offler R and Fleming PD 1968. A synthesis of folding and metamorphism in the Mt Lofty Ranges, South Australia. Journal of the Geological Society of Australia 15:245–266.
Oliver NHS, Dipple GM, Cartwright I and Schiller J 1998. Fluid flow and metasomatism in the genesis of the amphibolite-facies, pelite-hosted Kanmantoo copper deposit, South Australia. American Journal of Science 298:181–218.
Pollock MV, Spry PG, Tott KA, Koenig A, Both R and Ogierman J 2018. The origin of the sediment-hosted Kanmantoo Cu-Au deposit, South Australia: mineralogical considerations. Ore Geology Reviews 95:94–117.
Pontual S, Merry N and Gamson P 1997. Spectral interpretation field manual, Vol 1, P1-84. G-Mex Version 1.0. AusSpec International Pty Ltd, Melbourne.
Rolley P and Wright M 2017. Kanmantoo copper deposits. In GN Phillips ed., Australian ore deposits. The Australian Institute of Mining and Metallurgy, Melbourne, pp. 667–670.
Sandiford M, Fraser G, Arnold J, Foden J and Farrow T 1995. Some causes and consequences of high-temperature, low-pressure metamorphism in the eastern Mt Lofty Ranges, South Australia. Australian Journal of Earth Sciences 42:233–240.
Schiller JC 2000. Structural geology, metamorphism and origin of the Kanmantoo copper deposit, South Australia. PhD thesis, University of Adelaide.
Schodloc MC, Green A and Huntington J 2016. A reference library of thermal infrared mineral reflectance spectra for the HyLogger-3 drill core logging system. Australian Journal of Earth Sciences 63:941–949.
Schodloc MC, Whitbourn L, Huntington J, Mason P, Green A, Berman M, Coward D, Connor P, Wright W, Jollivet M and Martinez R 2016. HyLogger-3, a visible to shortwave and thermal infrared reflectance spectrometer system for drill core logging: functional description. Australian Journal of Earth Sciences 63:929–940.
Seccombe PK, Spry PG, Both RA, Jones MT and Schiller JC 1985. Base metal mineralization in the Kanmantoo Group, South Australia: a regional sulfur isotope study. Economic Geology 80:1824–1841.
Spry PG, Schiller JC and Ross RA 1988. Structure and metamorphic setting of base metal mineralisation in the Kanmantoo Group, South Australia. The AusIMM Bulletin and Proceedings 293:57–65.
Tedesco A 2009. Late-stage orogenic model for Cu-Au mineralisation at Kanmantoo mine: new insights from titanium in quartz geothermometry, fluid inclusions and geochemical modelling. Journal of Geochemical Exploration 101:103.
Verwoerd PJ and Cleghorn JH 1975. Kanmantoo copper orebody. In CL Knight ed., Economic Geology of Australia and Papua New Guinea. Australasian Institute of Mining and Metallurgy, Melbourne, pp. 560–565.
For more information, contact: