Recent studies at Osborne have highlighted a ... Ernest Henry, Eloise, Osborne, Mt Elliot and Selwyn mines, was selected to include occunences with either.
Eugen Stumpf! Memorial Symposium -
Session Two
New perspectives on IOCG deposits, Mt Isa Eastern Succession, northwest Queensland R. Mustard 1, T. Blenkinsopl, C. McKeagnei, C. Huddleston-Holmes l and G. Partington 2 'Predictive Mineral Discovery CRC and Economic Geology Research Unit. James Cook University, Townsville. QLD. 481 L Australia 2Kenex Knowledge Systems Ltd. PO Box 41136. Eastbourne. Wellington. New Zealand
A model for the formation of the iron-oxide-copper-gold (lOCG) deposits in the Mt Isa Eastern Succession involves fluids derived from late orogenic granites mixing with a fluid derived from an external source to fonn iron-rich alteration zones (commonly magnetite-rich) that contain vein stockwork, breccia, dissemination or replacement-style mineralisation. This process is inferred to be spatially and temporally associated with felsic pluton emplacement and cooling at around 1540 to 1500 Ma. This contrasts with an alternative model in which the fluids are entirely intra-basinal and amagmatic in origin. Recent studies at Osborne have highlighted a potential syn-peak metamorphic timing for mineralisation (based on 1595 Ma Re-Os dates on molybdenite and a 1595 ± 6 Ma U-Pb date on hydrothermal titanite), with no apparent proximal major intrusion. There is also a potential link between mineralisation and widespread mafic intrusive activity, which spans the entire range of known mineralisation ages. Methodology
In order to investigate the considerable range of potential geological controls on IOCG mineralisation, a prospectivity analysis was undertaken, aimed at evaluating the relative importance of a range of spatial variahles including: host rock type, proximity to granites or mafic intrusives, rockchip geochemistry (eu, Au, Co, Ni and As), structure and geophysics (including magnetics, gravity and wavelet-processed potentialfield data or 'worms'). A data-driven approach was taken in view of the considerable uncertainty in genetic models for IOCG deposits.
Cu-Au Occurrences Total contained Cu (Tonnes)
Study Area
N
Oto 10,000 (154) • 10,000 to 90,000 (10) (5) • 90,000 to 340,000 • 340,000 to 1,900,000 (3) .1,900,000 to 1,910,000 (1)
t
-50km
.Mt Oxide i
.. .i~ ,',"
•.
. . ... .-,.. :.,
• fit"
Australia
V
"
',.
", " .~Mt Ellio~
,
.... • Starra
•
•
Osborne
Figure 1,
Location of the Mt Isa study area and coppergold occurrences (n= 181 ).
281
SEG 2004 Predictive Mineral Discovery Under Cover Study area and datasets
The study area (- 82,750 km2) covers the Mt lsa Eastern Succession and part of the Western Succession. It extends from Mt Isa in the west, to Eloise copper-gold mine in the ea~t and Osborne copper-gold mine in the south (Fig. 1). Mineral deposit locations for hardrock copper-gold mineralisation were extracted from the MinOcc 2002 database (Queensland Department of Natural Resources and Mines 2(02). From the total of 567 copper-gold occunenees recorded in the database, a training data set (n= 181), including the Ernest Henry, Eloise, Osborne, Mt Elliot and Selwyn mines, was selected to include occunences with either recorded production or a quoted geological resource. Geological data, including the location of granites, mafic intrusives, Soldiers Cap Group, Corella Formation, and faults were derived from the Northwest Queensland Mineral Province Report (Queensland Department of Mines and Energy 2000). All faults within the study area were selected and plotted on a rose diagram. Based upon their distribution, they were subdivided into seven groups: N-S, NNE, ENE, E-W, ESE, SE and SSE (Table l). The regional magnetic digital dataset was obtained from Geoscience Australia. Surface rock chip geochemistry was obtained from the Geological Survey of Queensland geochemical database (Queensland Department of Natural Resources and Mines 2003). Weights of evidence
Spatial conelations were calculated using the Weights of Evidence technique (cf. Bonham-Carter 1994) using the Spatial Data Modeler extension developed for Mapinfo software. A unit area of 0.25 km 2 was used in these calculations, based on the assumption that thc known deposits have 3 0.25 km 2 area of influence. Table 1
Summary statistics for all faults (n=1854) from the Mt Isa project area. Sub-division based on fault populations plotted in rose diagrams. Constant buffer distance of 2 km used for all faults.
Orientation NS (350-15°) ENE (40-75°) SSE (15-40°) NNE (15-40°) 8N (75-100°) ESE (110-130°) SE (130-150°)
Contrast 1.45 1.53 0.95 0.91 0.56 0.53 0.39
Confidence 9.59 9.02 6.26 5.55 1.65 2.28 1.94
Fault Bends Orientation ENE SSE NS NNE SE ESE F3lV
Contrast 1.43 1.31 1.16 0.85 0.74 0.46 0.36
Confidence 6.29 6.62 6.10 3.06 2.55 1.18 3.15
Fault Intersections Orientation ENE & All SSE & All NNE & All N-S & All ESE & All E-W & All SE & All
Contrast 1.57 1.39 1.31 1.19 0.59 115 0.99
Confidence 7.29 7.63 6.54 6.67 1.71 2.98 4.17
Results
Weights-of-evidence analysis highlighted the strongest spatial association of copper-gold occurrences with N-S-oriented and ENE-orientalcd faults based 011 Contrast values of around I.S and Confidence values> 9 (Table 1). All significant copper-gold mincs are associated with at least one of thcse two orientations. The spatial association of copper-gnld occurrences with fault bends indicates a moderate to strong association with fault bends Oil ENE>trcnding faults (Ernest Henry), and SSE- and N-S-trending faults. The spatia! 282
Eugen Stumpf! Memorial Symposium -
Session Two
association of copper-gold occurrences with fault intersections indicates intersections with ENE faults are most important, followed hy SSE, l\NE and N-S faults (Table 1). Evaluation of the available surface rockchip geochemistry highlighted a strong spatial association of the copper-gold occurrences with Cu. Au. Co and Ni, whereas As had a poor correlation (Table 2). Table 2.
Summary statistics for rockchip geochemistry. Element Cu Au Co Ni As
Samples (n) 30.161 20,431 17,511 7,717 14,038
Threshold >249 ppm >0.11 ppm >85 ppm >43 ppm >33 ppm
Contrast 2.50 2.38 1.85 1.65 0.26
Confidence 12.37 11.97 7.66 4.76 0.65
Five-layer model
A simple five-layer model was developed to allow a comparison of Contrast and Confidence for evidential layers considered to play an important role in the formation of IOCG deposits in the Mt Isa Inlier (Fig. 2. & Table 3). The results indicate a significant proportion of the mineral occurrences were located no more than 75(J m stratigraphically above the contact between the Corella Formation and Soldiers Cap Group. This contact is considered to represent a site of fluid mixing as well as a potential physical barrier to upwarddirected f1uid flow. Magnetic highs, which may reflect areas of magnetite ± pyrrhotite precipitation, also have a strong correlation with copper-gold occurrences. The ENE- and N-S-orientated fault sets also have a strong to moderate association with copper-gold occurrences. Mafic intrusives have a much closer spatial association with the occurrences than does the felsic intmsive of the Williams-Namku batholith. Table 3.
Summary statistics for the five-layer model.
Key Ingredients Corella FmnSoldiers Cap Group Magnetic Anomalies Faults (NS & ENE) Mafic Intrusives Felsic Intrusives
Buffer Distance (km) 0.75
Area (km
2
Contrast
Confidence
3,826
Minerai Occurrences 43
1.87
13.97
NA
1,123
43
1.82
14.36
0.60 0.75 1.50
9,423 4,300 754
64 29 5
1.45 1.24 0.73
17.20 7.47 2.38
)
Discussion
Two important outcomes for ore genetic models are the recognition that: The lithological contact between the Corella Formation and Soldiers Cap Group has a strong spatial association with IOCG deposits ENE- and N-S-orientated faults have a stronger spatial association with IOCG deposits than either mafic or felsic intrusives. These results imply that fluid pathways and sites of fluid mixing are much more important than fluid sources for controlling the distribution of IOCG deposits. This understanding can possibly explain some of the diversity in the range of IOCG deposit types and models. A common mineralising process could generate deposits in a variety of host rocks depending on the fluid pathways. The dominance of the fluid pathways means that fluid sources cannot be clearly recognised from the spatial associations of the deposits alone, and mineralising fluids may be complex and heterogeneous in view of their possible interactions with a variety of wallrocks. A detailed understanding of fluid pathways and structures at all scales is the most important direction for future research. Mechanicallllodelling directed at understanding fluid flow in the Mt [sa Eastern Succession based on thi.-, structural knowledge will also be an important tool Acknowledgements This work was funded by Ihe pmd*CRC 12/3: '[bta! systems analysis. Eastern Succession. Mt [sa Project. We w()uld like to thank Martin Higham (Avantra Geosystems) for continual upgrades to his MI-SDM software. 283
SEG 2004 Predictive Mineral Discovery Under Cover
5 Layer Model WofE_PosteriorProbability
II 0.108894 0.000621 365 0.000371385 0.000251279
-
5km
Figure 2.
A. five-layer model for the Mt Isa Inlier including significant copper-gold occurrences, incorporating structure. 1550-1500 Ma granites of the Williams-Naruku batholiths. mafic intrusives, magnetic highs (as a proxy for magnetite ± pyrrhotite) and Corelia Formation - Soldiers Cap Group contact. B. Ernest Henry copper-gold rnine and surrounding region. C. Selwyn region including the Starra and Mt Elliot copper-gold mines.
References Bonham-Carter G .. 1994. Geugraphic iniormation systems !'or geoscientisl',. lll(ldcling wilh (lIS. O\I("d. Eng!and. Pergamon. 1SI cd. Queensland Department or \lines and [))crgy. 20()O. Norlh\\TSI Qucenslaml "lillcr;!I I'n)\'ince Report. Queensland Department of Mines and Energy. L3risi1;1!]c. Queensland Department or '!atur;li Rcs()llrcc, and Mines. 20(J2. QJV11", ()uccn,I;1I1'; Mineral Resourcc Database MINOCC 2002 - Microsoft Access Dataha,c. Vcrsi'1ll1.IL December 200:'. DCjlan~wl1l "I' "'!;uuraJ Rl'S()UrCcs and 11'1 inc,. Queensland Depanrncill of ",,\lurai Resources ,md Mines, 200], Cc()scicncc dal