sign in sign up
<<
Home , Ore genesis

AGSO lournal of Australian & Geophysics, 17(4),207- 211 © Commonwealth of Australia 1998 BIF-hosted deposits-Hamersley style R.C. Morris CSIRO Exploration and Mining, Private Mail Bag, PO Wembley, 6014, Australia

EXPLORATION MODEL Examples Local geological criteria Metamorphosed : Mt Tom Price, Mt Whaleback, • Simple hydrodynamic (sub)-artesian systems with ElF as Paraburdoo, Channar, Giles (Western Australia); the aquifer; exposed to the atmosphere along strike (kms), Carajas (Brazil); Sishen (). with exits for dissolved gangue. Non-metamorphosed supergene: (Dominant Hamersley • Surface-expressed faults or fractures needed to initiate the resources.) Marandoo, BHP OB 29, Area C, Hope Downs, replacement process at depth ( grow upward). West Angela, Ophthalmias • Other pre-ore faulting and thrusting can add considerable Target structural complexity, but are not essential for enrichment. • New finds of outcropping ores are rare. Mineralisation features • Exploration now includes search for buried ores suitable for • Ores are almost entirely ferric iron. open pit mining. • Lamination and texture are preserved. • 5-300 Mt, depending on proximity to established • oxidises to martite (). infrastructure. • Pseudomorphs of goethite after gangue (includes P, • >62% Fe < 0.08% P. from BIF apatite) are characteristic of non-metamorphosed Mining and treatment ores. • BIF residues include chert, with kaolinite from silicate • Stratabound tabular bodies-mining is limited only by open• horizons. pit design needs. • Burial converts goethite to microplaty • Grade within narrow limits, by on-site and port blending of hematite, destroying internal texture, but beddingllamination lump (6- 30 mm >64% Fe); fines «6 mm >62% Fe). remains to high metamorphic grades. • Typical export specification of::; 0.075% P.

Section 6 Deposit Marra Mamba (,hr Water table ..,

o 500 m !

Figure I. Generalised section at MtTom Price, showing metamorphosed deposits (lIlp/ H) in the Brockman IF and non• metamorphosed deposits in Brockman and Marra Mamba I Fs.

Regional geological criteria Alteration • Synforms in BIF sequences of any age. • of ore produces 3 tiers: • Enrichment occurs at exposed to an oxic I. 1- 2 m of hematite-dominated outcrop, texture/bedding! atmosphere (post-2000 Ma). lamination preserved. • Two major periods of ore forming worldwide (Precambrian 2. Deep hydrated zone (hematite forms AI-goethite) with and Mesozoic) with possible Paleozoic (North America). destruction of texture and bedding, a type of lateritic • Burial metamorphism (goethite to hematite) is needed for ferricrete. premium ores. 3. Ore zone, with post-enrichment leaching of goethite increasing friability. 208 R.c. MORRIS

BANDED IRON FORMATION I I- HEMATITE + SILICATES leaching BLUE DUST + ~ ORES MAGNETITE SILICA, ~supergene enrichment - I oxidation I metasomatic replacement by goethite I , I burial I ~ HEMATITE-GOETHITE ORE .. regional metamorphism J

modi~by leaching re-exposure~r1 HEMATITE-RICH+ ORES l or ehydration I

contact+ [ metamorphism ] , , I VARIOUS ORE TYPES I -l MAGNETITE-RICH ORES I

[ surficial processes •- eluvial concentration J , Figure 2. Schematic of genetic • : .!.1:" I .. ;i'U :r'u::( .....{.UI . ' ...... modelling for the major BIF• hosted iron ores of the world.

A. System viewed in 3D B. Electrochemical Cathode cell 4e- + 02 +2H20 ..... 4(OHr

c. Transfer of Fe to anode D.Trans~rofFetoanode

H4SiO 4 Drainage E. Transfer of Fe to anode F. Final supergene ore body now subject to leaching by

Figure 3. Model of supergene iron ore growth from below, as removes the surface. BIF-HOSTED IRON ORE DEPOSITS - HAMERSLEY STYLE 209

Introduction M artite-microplaty hematite Much of the following text is summarised from Morris (1980, The martite- mieroplaty hematite group includes the premium , 1985) and Harmsworth et al. (1990), which should be consulted low-P hematite-rich ores (often inaccurately called specularite for detailed citation. Though based mainly on research in the ores), the origin of which has attracted much debate and which Hamersley Province ('Hamersleys'), experience has shown that are here attributed to metamorphism of post-2000 Ma super• the basic concepts apply to BIF-hosted ores world-wide. gene ores. The group is characterised by the presence of well• preserved primary banding, but with loss of the internal texture Major ore types found in the non-metamorpho sed ores, as a result of the growth of abundant secondary microplaty hematite (mpl H) from the Although the dominant world resources of iron ore are associated goethite-pseudomorphed gangue. With increasing metamorphism with the major late Archean- BIF depositories, ores they range to granular or 'micaceous' hematite. The immense are known from the whole gamut these chemical . of size of some of these deposits (Mt Whaleback, Western Australia, for Because export specifications Australian bedded ores > 1500 Mt; N4E deposit, Carajas, Brazil, >3000 Mt) makes typically call for :0; 0.075% P, iron ores in Australia tend to be them the largest and most concentrated secondary accumula• considered in terms low phos or high ph as '. of' tions of any single metalliferous element in the 's , iron ores can be into BIF-hosted divided broadly three approached in scale only by some orebodies. groups: A subgroup comprises small deposits of magnetite-rich ore, I. i.e. residual concentrates Fe Leached BIF: of rarely exceeding a few million tonnes and generally associated lue dust Very low P. ('b ores'). with mpl H ores. These magnetite segregations can be attributed 2. Martite- goethite ores: moderate-high P, typically 0.07- to thermal metamorphism of normal ores by basic intrusives. 0.17% P. 3. Martite- microplaty hematite ores (± residual goethite): Low P, < 0.07%. Ore genesis Deposits of group 2 and 3 ores are shown in Figure I. There are five features of the BIF-hosted Fe-enrichment Weathering affects all three ore types in various ways, orebodies that merit special attention when considering their particularly in tropical areas, and early exploration was typically ong1l1. confined to the resulting complex profiles. This, compounded Scale, Individual deposits range from a few hundred to over by limited communication between rival groups, led to the con• 3 billion tonnes of >64°A, Fe. Thcy may extend along strike for flicting genctic modcls of the past. five or more kilometres and more than two kilometres down The mature three-ticred wcathering profile of ore in the structure. Typically, thcy grade abruptly from ore grade Hamcrslcys includcs: (>55'Yo Fc) to altered BIF «30% Fe) over distances that are ({II outcroppillg :;0111' of 12m of hematite-rich ore, the negligible compared with the size of the deposits- sometimes 'carapacc', with preserved BIF features resulting from less than a metrc. dchydration of the goethite-replaced matrix. underlain by: Purity of the ores, The orcs are dominated by ferric Fe oxides a highly modified Izl'lll'Uted :;Olle, up to 70 m, dominated by and only extremely rarely contain exotic components that cannot aluminous goethite, and marked by destruction of parental be attributed to residues from the parent rocks. features. Oxidatioll state o/'the ores. i\ fe\\ rclati\'ely small depo"its are mainly magnetite: others contain magnetite and hematite, These two tiers, comprising the 'hardcap', overlie but, quantitatively, it is the largely oxidised deposits, comprising the ore pmper, which retains the original BI F features, but hematite or hematite goethite ores, that dominate the world is typically affected by moderate scvcre post-enrichment scene. Despite their oxidised character, these e"tcnd to depths groundwater leaching of gocthitc and, Icss onen, ofhcmatitc. well below the normal influcnce of the atmosphere. but there arc no signi ficant mineralogical di fTerences bet\\ecn shallow Leached BIF (below hardcap) and deep ore. The leached BIF group is derived from BIF by direct dissolution Wall-rock alteration, It is generally agreed, even by advocates of the gangue minerals by groundwater to give residual concen• of hypogene models, that none of the major iron deposits con• trates of Fe oxides. The ore quality ranges from high-grade residues tains evidence for the upward passagc of fluids through the host ofmartite (oxidised magnctite = hematite) ± primary hematite rocks or associated faults. This is ofconsiderablc importance in (so-called 'blue dust orcs'), to highly siliceous but very friable view of the immense amount of solution necessary for enrich• BIF, readily upgraded by magnetic or methods. Ores of ment, whatcver mechanism is advocated. this derivation, often associated with enrichment ores, are found Stratigraphic situation, Most iron ores are strata-bound and in high-rainfall areas such as Brazil and India, but not in the arid there is a total absence of genetically related ' mineralisation' in Hamersleys. A small deposit of this type is known from the the associated rocks. wetter southern coastal area ofWestcrn Australia (South Downs deposit). SYI/gel/etic models These hypotheses have received minimal endorsement. Modelling Martite-goethite was based originally on a clastic origin for BIF, and on the The martite-goethite group includes a wide range of ores, in strata-bound character of the ore zoncs, 'Which werc thus deposits up to hundreds of millions of tonnes, which show considered to be a type of . With better access as residual BlF features, such as banding, as well as goethite• the result of mining, and with the acceptance of the concept of pseudomorphed microtextures. These goethite-rich ores range a chemical origin for B1F, thc models lost favour. Some unpub• from firmly indurated brown material to very friable yellow lished support for syngenetic models is still expressed in the ochre, all with martite and, often, primary hematite. Deleterious Hamersleys. phosphorus levels range from around a moderate 0.07% to a very high 0.17% P, depending on the content of the Supergene models parent BIF and post-enrichment leaching of goethite, the major Support for supergene models has becn strong in North America carrier of P. Because they lack evidcnce of metamorphism, are since the 1960s. However, Morey (1983) probably reflected the typically related to present erosion surfaces, and contain abundant North American attitude to the variety of local supergene and hydrous Fe oxides, these hematite-goethite ores are generally hydrothermal models when he stated that none had proven to accepted as resulting from supergene enrichment. be entirely satisfactory. Ore formation related to 'weathering 210 R.C. MORRIS crusts' of various ages and to metamorphism of early concentra• However, the earliest period of enrichment along exposed tions apparently had considerable support in the former USSR margins of the McGrath Trough probably resulted from the (Sokolov & Grigor' ev 1977). A general supergene and supergene• change to an oxic atmosphere at about 2000 Ma. Later, burial metamorphic model (Morris 1980, 1985, Harmsworth et al. to about 4 km resulted in low-grade 'metamorphism' (ca 100°C) 1990) is widely accepted in the Hamersleys, largely as a result of the deposits, converting goethite to form hematite-rich ores of detailed petrographic evidence. with some residual goethite. Subsequent erosion in the Mesozoic re-exposed the ancient ores to groundwater leaching, which Hypogene models removed the bulk ofthe residual goethite, leaving low-P, hematite There are three quite different ore types for which hydrothermal ore. These are the high-grade mp/ H ores of the Tom Price• origins are proposed. The first of these, the small magnetite Whaleback type. The Paraburdoo type is less leached, contain• deposits, such as those found with mp/ H ores in the Marquette ing remnant goethite and about 0.07% P. It is these mp/ H types, Range (USA), are possibly acceptable as hypogene, particu• with various metamorphic grades, that comprise the major high• larly with the evidence of massive cross-cutting magnetite veins grade ore resources outside Australia. and hydrothermal effects. More convincing are those, such as the specular, coarse, platy, micaceous or bladed hematite con• Six essential conditions for the genesis of supergene centrations with magnetite, in vertical pipe-like folds affecting iron ore (Figure 3). supergene ore at Koolyanobbing (Yilgarn Block, Western Aus• 1. An Fe-rich aquifer (i.e. BIF) both as a source of iron and tralia). These -like aggregates are associated with hydro• a target for enrichment. thermal and include hematite crystals up to 30 cm. These 2. Impervious layers, such as shales, above and below as 'hypogene' ore types are in reality supergene concentrations aquicludes. modified locally by hydrothermal solutions. 3. Suitable open structures, such as plunging synclines, Hypogene models have been used for the mp/ H ores, but together with deep-water access to initiate the system at are difficult to justify in view of, among other points, 1) the depth, e.g. -engendered fracture zones or fold-axis absence of conduits or wall-rock alteration below the orebodies; fractures. 2) the presence oflow-temperature phases, such as goethite, in 4. Suitable geochemical conditions, including exposure to deposits that have not been entirely leached or strongly meta• the atmosphere and exits for dissolved components. morphosed; and 3) the absence of hydrothermal minerals. 5. Suitable electrochemical conditions. To explain the absence of obvious conduits it is usually 6. Tectonic stability for long periods. suggested (e.g. Dorr 1965) that diffusing solutions selectively dissolved undetectable amounts of iron from the surrounding To form high-grade mp/ H ore of the Tom Price- Whaleback BIF, and transported this to the ore sites. Here the fluids changed type requires a further step: character to dissolve and remove all the non-Fe components, 7. Burial metamorphism of supergene ore with goethite precipitating hematite in their place, again without evidence of recrystallising to mp/ H, followed by later exhumation of their passage. Since at least greenschist conditions are stated or ore, and leaching of the remnant goethite with its included implied, this process must have occurred under as much as 10 km phosphorus. of cover. Ferric iron, in the absence of complexing agents, is The creation of ore in BIF requires the transfer of iron from virtually insoluble at pH >3, even at high temperatures. Data near surface, by biogenically aided dissolution, into the reacting from oceanic ridge systems and contemporary volcanic systems zone at depth. If, as in some models, the ore bodies grew from attest to the efficiency of iron transport in the ferrous form (~1% the surface downward, then the ore genesis must have been Fe2+) in supercritical aqueous fluid phases. But, equally, these much faster than surface erosion in order to generate the great fluids are not discriminating solvents. They invariably leave depths of ore typically observed. A more realistic appraisal evidence oftheir passage in the form of strongly leached residues, indicates growth upward as the surface BIF erodes (Fig. 3). a feature not associated with the typical BIF ores. Most deep-seated BIF-derived ferric ores have formed at Assuming that the absence of conduits can be explained, an depths beyond the likely reach of oxygenated water. Ferric iron even greater problem arises in the situation. Detailed is insoluble under most natural conditions. Thus, to have mechanisms for replacement processes are rare, because, as stated concentrated in its present position by pseudomorphous by Dorr (1965), a primary advocate of hypogene modelling, 'it is replacement of the BIF matrix, it must, presumably, have been difficult to imagine possible conditions under which large scale transported in the soluble ferrous form, followed by oxidation and continuing oxidation of ferrous iron at the site of deposition and precipitation as ferric iron at depth. Because magnetite is a might take place ... no large-scale and continuing source of oxygen good conductor of electrons, atmospheric effects in the sub• is easily found'. outcrop can drive massive electrochemical cells in BIF (Morris et al. 1980). Solution of iron near the surface leaves a friable residue of silica, colloquially known as 'denatured BIF'. This Supergene enrichment and metamorphic readily erodes, exposing further material for reaction. Iron is upgrading-modelling transferred in groundwater through pores and fractures to the Nearly two decades of CSIRO- AMlRA research into iron ore reacting zone at depth, replacing gangue minerals. Thus, the ore has shown that supergene enrichment ofBIF, followed in some body grows upward, as erosion removes the surface, which cases by burial metamorphism, is the only mechanism that, forms a topographic low. With the end of the enrichment phase, currently, can reasonably explain the known data from the iron weathering ofthe ore begins, while at depth, groundwater move• ore deposits, without the use of' geological theology rather than ment leaches goethite to produce porous ore. The typical ridges scientific knowledge', as Dorr (1967) put it so well. of the deposits today reflect the resistance to erosion of the Figure 2 is a general schematic diagram, modified from Morris hardcap. (1985), outlining the broad model. Essentially, oxidation of magnetite and metasomatic replacement of gangue minerals by goethite produces hematite-goethite ore. This ore may be modified Exploration by continued groundwater leaching or, by dehydration due to The known Hamersley resources of premium low-phosphorus surface exposure, to produce a range of porous to extremely mp/ H ores are likely to be exhausted by the middle of the next dense ore types. These non-metamorphosed ores of the Mesozoic century and, though the search continues for this ore type, in• comprise over 90% of the resources of bedded ore in the creasing attention is being focussed on the ancillary ores for Hamersleys, but are little known elsewhere. blending or development as discrete products. BIF-HOSTED IRON ORE DEPOSITS-HAMERSLEY STYLE 211

With possible rare exceptions, all significant enrichment References deposits have been found by their generally extensive outcrop, Dorr, J. van N., 1965. Nature and origin of high-grade hematite which largely explains why there have been no important premium ores of Minas Gerais, Brazil. , 60,1-46. ore discoveries in the past two decades of intensive exploration Dorr, 1. van N., 1967. The Fort Gouraud, Mauritania, iron ore in the known BIF fields. Nevertheless, there is still potential for deposits-discussion. Economic Geology, 62, 567- 572. economically viable buried bodies, and the search continues. Gair, IE., 1975. Bedrock geology and ore deposits of the Palmer The obvious targets, using supergene-metamorphic modelling, Quadrangle, Marquette County, Michigan. U.S. Geological are Precambrian unconfonnities, but if the ores have not been Survey, Professional Paper 769, 159 pp. sufficiently metamorphosed or have not undergone a leaching Gross, G.A., 1965. Geology of iron deposits in Canada: Volume stage, they may retain a significant goethite(+P) matrix. Though I, General geology and evaluation of iron deposits. it could be expected that some phosphorus might have been Geological Survey of Canada, Economic Geology Report exhaled with water from goethite during metamorphism, the ore 22. may not reach the hoped for quality of the best exposed ores. Harmsworth, R.A., Kneeshaw, M., Morris, R.C., Robinson, C.J. Systematic regional mapping is important for this explor• & Shrivastava, P.K., 1990. BIF-derived iron ores of the ation, and recognition offeatur es equivalent to the Precambrian Hamersley Province. In: FE. Hughes (editor), Geology of McGrath Trough and their relationship to post-2000 Ma expo• the deposits of Australia and Papua New Guinea. sure and later metamorphism will help to focus the search areas Australasian Institute of Mining and Metallurgy , for the buried mp/ H ores. Following this, the choice for more Monograph 14,617 - 642. detailed exploration should be centred on structural situations Kneeshaw, M., 1984. iron ore classification-a proposal that can be related to suitable pre-ore sub-artesian systems. for a common classification for BIF-derived supergene ores. Geophysical techniques have been used more as an adjunct to Australasian Institute of Mining and Metallurgy, Proceed• evaluation than for exploration in the past, but the development ings, 289, 157-162 . of airborne gravity methods could add a significant tool to the MacLeod , W.N. , 1966. The geology and iron deposits of the search for buried deposits. Hamer sley Range area, Western Australia . Geological The original cover over mp / H bodies might perhaps be Survey of Western Australia, Bulletin 117, 170 pp. expected to show some evidence of an aureole of hydration, Morey, G. B. , 1983. Animikie Basin, Lake Superior Region, possibly with some increase in trace phosphorus , from the U.S.A. In: Trendall, A.F & Morris, R.C. (editors), Banded exhalation of water from goethite. However, since the cover iron-formation: fact s and probl ems. 13-6 8. El sevier, sedim ents themsel ves were originall y saturated with water Amsterdam. before this low level of metamorphi sm (ca. 100DC), thi s is Morri s, R.C., 1980. A textural and mineralogical study of the arguable. Some increase in quartzihematite veining mi ght be a relati onship of iron ore to banded iron-formation in the more realistic expectation. Hamersley Iron Provincc of Western Australia. Economic The best hope for viable buried mp / H deposits in the Geology, 75, 184- 209. Hamersleys is those eroded to base level and concealed beneath Morris, R.C. , 1985. Genesi s of iron ore in banded iron-formation all uvium in the intermontane areas. C lues to these lie in the by supergene and supcrgcne - metamorphic proccss a identification of IIlp/ H texture in all uvials around the margins conceptual model. In: Wolf, K.H. (cditor), Handbook of of the McGrath Trough(s). strata-bound and stratiform ore deposits. 13, 73 235. An alternative exploration model could be based on hypo• Elsevier, Amsterdam. gene modelling. As discussed already, there are major difficul• Morris , R.C. , Thornber, M.R. & Ewers, W.E. , 1980. Deep• ti es in app lyi ng this conce pt to any known ores in the seated iron ores from banded iron-formation. Nature, 288, Hamersleys, but that does not invalidate the concept for a new 250-252. ore type so far (1997) not seen. However, care should be taken Sokolov, G.A. & Grigor'ev, V.M., 1977. Deposits of iron. In: to avoid mistaking post-ore metasomatic events in current ore v.I. Smirnov (editor), Ore deposits of the USSR. 1, 7- 113. types, such as the secondary magnetite, apatite, and of Pitman, London , the Southern Ridge deposit at Mt Tom Price, for hydrothermal Trendall, A.F. , 1983. The Hamersley Basin. In: Trendall, A.F. & processes genetically related to iron ore fonnation. Morris, R.C. (editors), Iron-fo rmation: facts and probl ems. 69- 129. Elsevier, Amsterdam. Acknowledgements I would like to thank the referees, R.A. Harmsworth and Received 30 October J 996; accepted /2 May J 997 M. Kneeshaw both for improvements to the ms, and for their active support and input during the CS IRO- AMIRA research programs.