Proceedings: 9 - 11 September 2002 Perth, Australia
Proceedings: 9 - 11 September 2002 Perth, Australia
International
Conference on Iron
Ores and their host
rocks: genesis,
exploration, mining,
mineralogy, processing and environment in the
new millennium
IRON
ORE
2002
9 – 11 September 2002
Perth, Australia
PRINCIPAL SPONSORS
Iron Ore Genesis and Post-Ore Metasomatism at Mount Tom Price
R C Morris1
ABSTRACT
TABLE 1
About 90 per cent of the pre-mining BIF-hosted iron ore resource of the
Hamersley Province of Western Australia is of the Phanerozoic Basic classification of BIF-hosted iron ore genetic models,
supergene martite-goethite type. The remaining ten per cent formed as
simplified from Morris (1985).
supergene deposits at ~2000 ± 200 Ma, and were later modified by burial
SYNGENETIC
to form the Proterozoic martite-microplaty hematite ores. The
supergene-metamorphic segment of this general unified model has been Unmodified sedimentary/diagenetic concentrations
challenged by three recently published hypogene-based genetic models Modified by metamorphism or igneous activity
for the hematite ores, all requiring meteoric oxidation to produce the ore.
The most recent and detailed, by Taylor et al (2001), is based on data Modified by supergene processes
from the small Southern Batter and North deposits of the Mount Tom SUPERGENE (related to current or past erosion surfaces)
Price mine. The authors proposed that during Stage 1 ‘overpressured Residual concentration by oxidation of magnetite and leaching
bicarbonate-saturated basin brines from the dolomite aquifer of the
Wittenoom Formation’ at 150 - 250°C, dissolved free silica from BIF to Enrichment by replacement of gangue with oxidation
produce a stratigraphically thinned, magnetite-carbonate-silicate-apatite Enrichment with subsequent burial metamorphism
rock. In Stage 2, meteoric oxidation of siderite to secondary microplaty
hematite + ankerite, and magnetite to martite, at unspecified elevated HYPOGENE
temperatures and pressures, was followed by carbonate leaching (Stage Metamorphic
3). Finally, during Stage 4, weathering with removal of apatite, produced
Residual concentration by heated connate or recirculated meteoric
the typical low-phosphorus, porous martite-microplaty hematite ores of
fluids
the Tom Price-Whaleback type.
However, evidence of both microplaty hematite and martite reduced to Enrichment by replacement of gangue by hematite from connate or
magnetite, and of infill silica, apatite, and ferroan chlorite, with late-stage recirculated meteoric fluids
pyrite, in the pore space of martite-microplaty hematite ore, shows that Physico-chemical: differentiation or dynamothermal
the hydrothermal activity at Mount Tom Price is a post-ore event. Neither
Igneous
pyrite nor ferroan chlorite could resist the oxidation required to form
hematite ore. Recrystallisation of microplaty hematite in these infilled Residual concentration by heated juvenile or recirculated meteoric
zones to a typically coarser form than present in most of the normal ores, fluids
is a further supporting factor. Enrichment: replacement of gangue by iron in part derived from the
A more credible, post-ore alternative to the hypogene genetic model is intrusive
suggested here. Re-exposure to leaching by ground water of the
metamorphosed ore deposits in the Phanerozoic resulted in the partial
removal of remnant goethite. Localised modifications in the porous Most iron ore genetic models relate to the prominent
BIF/ore contact zones were driven by exothermic oxygen/pyrite/carbon hematite-rich type known as the ‘specularite ores’ in the
reactions in the footwall Mt McRae Shale. Cool descending oxygenated
Americas, and as the martite-microplaty hematite (M-mpl H)
meteoric water in the permeable ore horizons, acted as one limb of a
thermal convective cell. In the other limb, heated reducing solutions rose ores in Australia. Despite increasing use of mineragraphic data in
through the permeable contact zones, to produce the observed localised the last three decades, detailed mechanisms for the
modifications of BIF and ore. Expelled silica from the carbonatised BIF transformation of BIF to ore are still rare, particularly for the
invaded adjacent porous ore to form erratic silicified ore zones, and to origin of the secondary mpl H component.
produce the local quartz vein systems found in the associated overlying Syngenesis was the key starting point for the earliest genetic
BIF. models from at least 1883, largely on the assumption of a clastic
origin for BIF, but composite models of pre-metamorphic
INTRODUCTION syndiagenesis, and sedimentary–metamorphic were suggested in
Brazil up to at least 1982. Informal support for syngenesis in the
Genetic concepts for BIF-hosted iron ores, particularly since the Hamersleys in the past has not been backed by a published
1940s, have tended to recycle ideas from the past, but typically model.
with new twists. Many models were based on single deposits.
Supergene models in which the ores are believed to be
Table 1, shows eleven basic models for iron ore genesis,
residual concentrates from ground water leaching of BIF are
though there are many variants. The list illustrates the quandary
common. Excellent evidence for this mechanism exists in high
that we face in genesis paradigms, a problem put well by Richard
Kuhn, an Austrian chemist who was forced to decline a Nobel rainfall areas such as Brazil and India, where these ‘blue dust’
prize by the Nazi government. ‘In general, scientists debating the ores are often associated and confused with very leached M-mpl
relative merits of competing paradigms simply talk right through H types. But most supergene models for the mpl H ores invoke
each other, using the same words to discuss different realities’. enrichment by connate or recirculated meteoric waters. In 1886
Dorr (1967) put it more bluntly: many assumptions used to for example, a combination of in situ oxidation and precipitation
support the genetic models have been based largely on of iron oxides from solution was suggested. In 1897, in what was
‘geological theology rather than on scientific knowledge’. Sadly, possibly the earliest detailed mechanism, oxygen-bearing water
some authors misrepresent previous concepts and arguments to and a separate flow of ferrous iron-bearing water, converged to
improve the credibility of their own ideas, reminiscent more of precipitate iron oxides at depth. More complex mechanisms have
political lobbying than of science. been proposed, requiring pyramiding of unusual geologic
conditions. However, most writers have tended to ignore the
difficulties of the redox conditions needed for synchronous
1. Honorary Research Fellow, Division of Exploration and Mining,
Commonwealth Scientific and Industrial Research Organisation.
mobilisation and precipitation of iron. They have instead used the
(CSIRO). obvious signs of weathering, such as the oxidized mineral
assemblage, the association of ores with present or past erosion preserved. The original BIF microscopic textures are superbly
surfaces, and the juxtaposition of the ores with the most logical represented by martite (± primary hematite) and phosphorus-
source of oxidation potential, the atmosphere, to support a bearing goethite pseudomorphs of chert, carbonates, and
supergene origin. But there are other technical problems with the silicates, but no secondary mpl H. Hundreds of deposits exist,
supergene formation of the mpl H ores, in part overcome by and these range from small pockets of enrichment to over 1.5
adding the effects of burial and metamorphism to a primary billion tonnes. They are typically associated with the current
supergene product. The earliest of these compound models may erosion surface, particularly those in the Marra Mamba IF, but
be that of Van Hise and Leith (1911), and the latest, Morris some extend to >250 m in depth. The phosphorus content, held
(1980, 1985, 1998). The dominant ores of the Hamersleys, the mainly in the goethite, ranges from ~0.07 per cent to over 0.17
M-G type, without secondary mpl H, are generally accepted as per cent P.
supergene enrichment ores without superimposed metamorphism. Martite-microplaty hematite (M-mpl H) ores comprise the
Hypogene modelling is typically the most complex, related premium, low-phosphorus (<0.06 per cent P) hematite-rich types.
initially to the effects of metamorphic or igneous hydrothermal These are dominated by hematite, largely as martite, but with
solutions on BIF. Models range from direct leaching of gangue to significant other secondary hematite, much of it in the mpl H
produce residual ores, to a variety of enrichment processes, with form, occupying former gangue mesobands. This is the only real
introduced iron either directly replacing BIF gangue with textural difference between the two primary ore types, since all
hematite or, indirectly by some more intricate process. levels of the banding from macrobanding down to the finest
Eventually, mixing of hypogene and meteoric water is required to laminae are typically well preserved, as are the hematite-
produce the required oxidisation to hematite. J W Gruner is replaced, ex-cherty BIF-varves. Further stratigraphic thinning
probably the best known of the early hypogene advocates. His during Proterozoic and Phanerozoic re-exposures, resulted from
basic model involved oxidation and concentration of iron leaching of remnant non-hematite components, particularly
minerals by leaching of silica from BIF in hot magmatic water. goethite, with a final stratigraphic thickness of around 50 per
Later, using Yellowstone National Park as an example, he cent. As a result, the mpl H ores have a higher relative level of
suggested leaching and oxidation by the reaction of rising martite than their stratigraphically equivalent unmetamorphosed
vapours with descending meteoric waters (Gruner, 1937). M-G ores. Only about 20 of these deposits have been found in
Replacement of BIF matrix by hematite at considerable depths the Hamersleys, ranging from a few hundred thousand tonnes, to
was integral to many of the later metamorphic hypogene models giants such as Mount Tom Price (900 million tonnes) and Mount
such as those of Guild (1953), and Dorr (1962) for Brazilian Whaleback (1.7 billion tonnes). Most are in the Brockman IF,
ores. Dorr (1965) suggested that the lack of apparent changes to and rarely in the Marra Mamba IF. The mpl H deposits are
the country rocks was the result of diffusion of supercritical typically deeper than their equivalent sized Phanerozoic M-G
fluids that selectively dissolved only minute amounts of iron deposits. A notable exception is the main Mount Tom Price
from the surrounding BIF, and precipitated this iron at the ore deposit that, like the M-G ores, has been largely exposed at the
sites. But he offered no explanation of how the vast amount of surface by deep erosion.
introduced ferrous iron converted to the oxidised state as mpl H.
The Hamersley hypogene-supergene model of Barley et al Weathering
(1999) and Taylor et al (2001), is the latest of the composite
type. Surficial alteration of all ore deposits in the Hamersleys,
generally accepted as totally unrelated to ore genesis (Morris,
1985), has resulted in a 1 - 2 m, hematite-dominated carapace,
Genetic modelling in the Hamersleys preserving the original features of the ore, and thus the BIF
Intensive exploration for iron ore in the Pilbara began in the banding. This is underlain by a hydrated weathering zone
early-1960s, with early ideas based mainly on macroscopic generally 15 - 30 m but up to 70 m thick, typically dominated by
examination of the weathered zones, followed by more detailed vitreous aluminous goethite, with destruction of most of the
concepts as drilling increased. In the progress of published ideas parent ore textures and banding. In the ore below this hardcap,
documented by Morris (1986), the initial ideas of simple leaching of the supergene goethite by groundwater has resulted
supergene leaching of BIF, gave way to variants involving in increasingly friable ore that today still retains its BIF banding
leaching with void filling by goethite. Then came the suggestion at all scales and, in the case of the unmetamorphosed ore type, its
that only high temperature solutions were capable of oxidising microscopic pseudomorph textures, though in a more skeletal
primary siderite to magnetite and then to martite, and that this mode compared with remnant specimens of unleached ore.
occurred in the late Precambrian. In the northern Pilbara, a
supergene–burial metamorphism model was discarded in favour CURRENT GENETIC MODELLING IN THE
of hydrothermal-metamorphic re-concentration of iron during or HAMERSLEYS
after Precambrian folding. By 1975, supergene modelling was
dominant in the Hamersleys, but there was also minor support for
hypogene processes, and informal support for syngenetic origins. Commonwealth Scientific and Industrial Research
An industry backed CSIRO supergene and supergene- Organisation – Australian Mineral Industries
metamorphic concept was widely accepted in the 1980s and Research Association (CSIRO-AMIRA) — A unified
1990s, and this model is currently under challenge by a new supergene, and supergene-metamorphic model
hypogene-supergene paradigm.
This unified ore genesis model was introduced in Morris (1980),
and refined in Morris (1984, 1985, 1987, 1998), Morris et al
BIF-HOSTED IRON ORES IN THE HAMERSLEY (1980) and Harmsworth et al (1990). These papers resulted from
PROVINCE nearly two decades of co-operative research between CSIRO and
the Hamersley iron ore industry through AMIRA.
There are essentially two basic BIF-hosted ore types in the
Hamersleys.
Supergene modelling
Phanerozoic, non-metamorphosed supergene martite-goethite
(M-G) ores totalled about 90 per cent of the pre-mining In the supergene stage, BIF gangue was largely pseudomorphed
Hamersley BIF-hosted resource (Harmsworth et al, 1990). by goethite, preserving the original microscopic and macroscopic
Despite stratigraphic thinning of up to ~35 per cent, and severe detail of the parent BIF. Magnetite was eventually oxidised to
post-ore leaching, all levels of BIF banding are consistently kenomagnetite (an informal term for the range of metastable
FIG 1 - Model of supergene BIF enrichment with iron ore growth from below, as erosion removes the surface (A-G), followed by burial
metamorphism to form martite-microplaty hematite-goethite ore (H). Later erosion with leaching of goethite, results in martite-microplaty
hematite ore (L). Note that E, from the martite-goethite Section 6 mine at Tom Price, shows friable silica residues resulting from BIF
leaching, as an example of the leached ‘loose’ silica zone shown in C, D, and F.
lacunar spinel phases between magnetite and maghemite), and Burial metamorphism of supergene M-G ore to form
martite (a textural-mineralogical term for hematite after M-mpl H ore
magnetite). Partial loss of gangue to ground water during the
process, resulted in stratigraphic thinning of some of these In this section of the model (Figure 1H and I), the M-mpl H
martite-goethite orebodies of around 32 - 36 per cent (Baldwin, deposits represent metamorphosed supergene ores of the M-G
1976). type. The earliest of these M-G ores formed along the folded and
The iron metasomatism occurred in massive, gravity- exposed margins of the McGrath Trough at about 2.0 (±0.2) Ga,
controlled hydrodynamic electrochemical cells (Morris et al, possibly toward the end of the Beasley River Quartzite. The
1980), driven by cathodic reactions in the upper BIF zone during metasomatic goethite was converted in part to mpl H during
wet seasons (4e- + O2 + 2H2O → 4OH-) (Figure 1). Ferrous iron burial to ‘diagenesis’ levels (ca 100°C), before re-emergence and
from biogenic reactions in the vadose zone was carried to depths erosion of the deposits in places to form ore conglomerates in the
well beyond the normal reach of the atmosphere. Here Mt McGrath Formation of the upper Wyloo Group. Exposure of
pseudomorphous replacement of chert, silicates, and carbonates the mpl H orebodies since at least the Mesozoic has resulted in
by ‘goethite’ occurred by anodic oxidation at depth (Fe2+ → Fe3+ the dissolution by groundwater of most of the unconverted
+ e-) followed by ferrolysis (Fe3+ + 3H2O → Fe (OH)3 + 3H+). goethite in the deposits, leaving porous hematite ore, in places
Magnetite/kenomagnetite layers in the BIF acted as electron extremely friable due to partial solution of the hematite itself, but
conductors, with groundwater as the ionic transfer agent. Silica some deposits still retain large areas of residual recrystallised
at depth was released to drainage at accelerated rates as a result goethite in the mpl H ore.
of cycling of seasonally driven iron redox reactions with quartz
(Morris and Fletcher, 1987). The martite-goethite zones grew Evidence for the metamorphic model
upward along conductive strata from the deep reacting zones, as
erosion removed the friable silica outcrop (Figure 1E), General
colloquially known as ‘denatured BIF’, which resulted from the
leaching of iron. Fluid transport down surface-expressed faults, Many similarities between the apparently disparate M-G and
thrusts or fracture zones, or through cross folds, initiated the M-mpl H deposits relate to their common heritage of banding
deep cells; once started, the ore forming process generated its and chemistry. But there are other factors compatible with a
own fluid pathways and thus these fracture zones became related genetic mechanism. The overall evidence of the
redundant. relationship of ore to plunging structures from the deposits of
In this model there is no requirement for oxygenated ground both types supports the concept that they all represent the eroded
water at depth, and there is no relationship to lateritisation remnants of confined gravity controlled hydrodynamic systems,
processes. ie flowing aquifers.
Unlike many of the M-G deposits, the mpl H deposits seldom Springs Basalt feeders, puts their claim for a pre-Wyloo origin
conform to the recent weathering surface (Tom Price is a notable for mpl H ore in doubt. Barley et al (1999) also dismiss the
exception). However even apparent ‘blind’ deposits, such as the association of hematite mineralisation with the Ophthalmian
Southern Batter deposit, or the No 2 lens at Paraburdoo, always orogeny, preferring a later date, circa 2.0 Ga, associated with
have connecting channels to the surface, consistent with original Wyloo rifting, consistent with the 2.0 ± 0.02 Ga suggested by
supergene access. Morris (1985).
The mpl H deposits are stratigraphically thinner than the M-G Support for orogenic meteoric water is cited by Powell et al
type. This is also consistent with the modelling – initial thinning (1999) from modelling of uplift at ~5 Ma of the Southern Alps of
during supergene enrichment, followed by volume loss from the New Zealand. Here, evidence for oxidation is based on
supergene goethite to hematite partial transformation (~27 per interpretation of subtle variations in oxygen isotope ratios in
cent), due to burial, with further compaction following leaching quartz veins. Proterozoic oxidation in the BIF ore zones is
of the remnant goethite after re-exposure. anything but subtle, and yet probably occurred at notably lower
The major structural difference between the two ore types is atmospheric oxygen levels (Morris, 1985) than in the 5 Ma New
that the mpl H deposits tend to be significantly deeper than Zealand environment. Evidence of hydrothermal minerals,
comparably sized M-G deposits (the main Tom Price deposit is conduits, or veining actually associated with the majority of the
again an exception), and associated with major faulting. This ores, or of connective aquifers for suitable long distance
may reflect the Proterozoic structural setting when the bulk of transport of hot (>250°C), leaching fluids is not discussed.
the Brockman IF would have been buried beneath about 5 km of Powell et al (1999) further speculated that as the solutions
strata. Thus, exposure to surface-driven processes at that time cooled away from the orogenic front, they produced some M-G
could have resulted only from major structural discontinuities, deposits (unspecified), though all the mpl H deposits are
consistent with the evidence in the current M-mpl H deposits. geographically closely associated with M-G ores. The latter are
particularly common in the Marra Mamba IF, a stratigraphically
Residual goethite in M-mpl H ore deeper unit than the Brockman IF that hosts the bulk of the mpl
H ores. Though most researchers agree that genesis of the mpl H
Significant relic goethite from the Proterozoic supergene ores took place between F2 and F3 folding (≡ F3 and F4 of
enrichment, contributes to the mid phosphorus (≥0.075 per cent Taylor et al, 2001), all current field structural evidence indicates
P) levels of M-mpl H + g ores along the Paraburdoo Range. This
the M-G ores are post F3 (Kneeshaw et al, 2002).
limits the burial temperature to about 140°C, probably below
100°C (Morris, 1980, 1985, 2002). Kaufman et al (1990), from The minimal data on ore genesis mechanisms by Powell et al
carbonate isotope data, later supported burial temperatures ‘near’ (1999) make it hard to discuss this paper in any depth.
100°C for BIF below the ore at Paraburdoo, in contrast to ‘near’
300°C in BIF at Wittenoom. Taylor et al (2001) claim that this University of WA (2) and Taylor et al (2001) –
intersticial goethite is redeposited superficial material, though Hypogene-supergene modelling
this weathering process apparently has not added comparable
goethite to the ‘clean’ hematite ores of Mt Tom Price or Barley et al (1999) and Taylor et al (2001) have presented similar
Mt Whaleback. The residual Proterozoic phosphorus-bearing hypogene-supergene genesis models for M-mpl H ore derived
goethite is texturally very different from redeposited goethite. from banded iron-formation (BIF). Their models are based on
The latter rarely contains detectable phosphorus, except where it evidence of hydrothermal features in the two smallest of ten Tom
is found as clearly recognisable late-stage colloform material Price hematite deposits, originally recorded as erratic carbonate
(Dukino et al, 2001). The Paraburdoo Range contains the only metasomatism in BIF core from Tom Price (Ewers and Morris,
BIF-hosted ores still associated with remnants of Proterozoic 1981), and as localised post-ore metasomatism by Harmsworth et
cover, implying relatively recent re-exposure, and hence some al (1990) and Morris (1998).
explanation for the common preservation of the Proterozoic
goethite in these deposits. Barley et al (1999)
Further supporting data for the goethite to hematite transition The brief paper by Barley et al (1999) gives little detail on the
is found, for example, in a zone of alternating ore and proposed mechanism of the transformation of BIF to ore, mostly
unenriched, oxidised BIF within the Whaleback M-mpl H describing the petrographic features of the various phases of
deposit. In the oxidised BIF, goethite in veins and after alteration in the North deposit at Mount Tom Price (Figure 2).
carbonate, contains mpl H, which in the case of the vein material Ore formation ‘involved initial high temperature crystallization
at least, must be of secondary origin, and probably coeval with of magnetite-siderite-iron silicate assemblages. This was
the mpl H of the adjacent ore strata. Mpl H of the type found in followed by development of hematite and ferroan dolomite
the mpl H ores, is not found in goethite in the normal supergene assemblages with subsequent oxidation of magnetite, leaching of
environment.
carbonate and silicates and crystallization of further hematite.’ A
‘range of fluid types most likely reflecting interaction of
University of Western Australia (1) – hydrothermal brines with descending meteoric fluids’ produced
Orogenic-meteoric modelling initial hematite crystallization at about 250°C, and the ore
Powell et al (1999) suggested deep penetration and heating of formed ‘as a response to heat and fluid flux during Wyloo rifting
oxygenated meteoric water during early Paleoproterozoic uplift at about 2.0 Ga’. There are minor differences with the Taylor et
or during orogenic collapse of the Ophthalmian (F2) fold and al (2001) model in that only partial loss of silica from BIF is
thrust belt of the southern Hamersley Province. The fluids were involved, with recrystallisation of remnant chert, and that only
channelled to a few of the available structural sites, producing magnesium, calcium and carbonate were added. The transition
massive localised hydrothermal dissolution of silica and from (1) magnetite-chert BIF (25 - 30 per cent Fe) to (2)
oxidation of the BIF host to generate mpl H ore. The mechanism enriched magnetite (± hematite)-siderite-(± iron silicate) BIF (50
for generation of mpl H is not discussed. They suggested that per cent Fe) to (3) magnetite-hematite mineralisation (~60 per
northwest trending dolerite dykes that are associated with some cent Fe) to (4) martite-hematite (± goethite) ore (65 per cent Fe)
ore deposits, potentially supplied additional heating. Since the occurred over a few metres with an abrupt transition from phase
northwest dolerites of the orebodies are clearly pre-ore (Morris, 2 to phase 3. The formation of hematite following the primary
1985), their interpretation that these are Wyloo Group, Cheela stage (2), is described as: ‘Increasing oxidation (also at elevated
FIG 2 - Cross-section through the North deposit, Mount Tom Price mine, from Taylor et al (2001).
temperatures) then resulted in crystallization of prismatic Stage 1 (primary stage): virtually all free silica was removed
hematite and ferroan dolomite’, with later ‘oxidation of at temperatures of 150 - 250°C by hydrothermal solutions, with
magnetite to martite and further crystallization of hematite as partial carbonate replacement. Retention of original magnetite,
prismatic and microplaty forms’. carbonate, and silicate, without oxidation, resulted in a typical
final assemblage of magnetite-siderite-stilpnomelane-pyrite-
Barley et al (1999) repeatedly misrepresented the Morris
apatite. ‘At the end of this stage the iron formation had lost
supergene modelling as being related to lateritic processes. For
considerable volume and was thinned’ to form a general, and
example in their final statement, they say that the concept that the
potentially world-wide hematite ore precursor.
hematite ores ‘are simply the result of ancient deep lateritic-style
weathering (± metamorphism) should be re-evaluated’. Stage 2 (deep meteoric circulation and oxidation at
‘elevated’ temperatures or in Tertiary weathering profiles):
Indeed, if anyone has actually claimed that particular genesis magnetite was oxidised to martite, and siderite converted to mpl
for the M-mpl H ores, it most certainly should be re-evaluated! H + ankerite, with essentially no further loss of volume at this
Lateritic weathering is essentially a superficial concentration stage.
mechanism that leads to almost total destruction of the primary
Stage 3 (leaching stage): the remaining carbonate was almost
textures and banding in the enriched zone, completely different
completely leached, leaving a porous magnetite-apatite or
from the well documented superb preservation of BIF texture hematite-apatite assemblage. Further compaction resulted,
deep below the zone of weathering in the M-G ores. As stated in increasing the residual enrichment of Fe, but apparently no
Morris (1985), ‘… supergene metasomatic replacement contrasts further oxidation, enabling preservation of magnetite.
strongly with a texture-destroying supergene process that occurs
Stage 4 (weathering – shallow meteoric waters): removal of
in the vadose zone. An extreme example of the latter is what is
all components other than martite and secondary hematite, to
loosely called lateritisation…’.
produce the high grade low phosphorus M-mpl H ore.
Taylor et al (2001) Modelling inconsistencies
These authors were more specific about mechanisms in their
There is one major advantage in the current conflict between the
model, using data from the North (Figure 2) and the Southern hypogene-supergene and the supergene-metamorphic modelling
Batter deposits (Figure 3) at Mount Tom Price. They also for mpl H ore — they differ so materially in their primary
suggested four stages in the alteration of BIF to produce M-mpl mechanisms that objective examination of the data should enable
H ore. This was initiated by leakage of ‘overpressured a clear judgment on their relative merits. This, of course, does
bicarbonate-saturated basin brines from the dolomite aquifer of not necessarily mean that either is right, but the resulting
the Wittenoom Formation’ through faults into the lower part of discussion from this contentious issue can only be good for iron
the Brockman Iron Formation. ore geology.
FIG 3 - Cross-section through the Southern Batter deposit (grey) and the Centre pit (black), Mt Tom Price mine from Taylor et al (2001).
Morris (2002), in criticising the modelling of Taylor et al POST-ORE ALTERATION OF BIF AND ORE
(2001), touched briefly on problems with deep meteoric
oxidation, iron transport and enrichment, contradictory thickness Hypogene model for Mount Tom Price
data on BIF to ore relationships, contradictory chemical and
textural considerations, inconclusive data on dolerite alteration, The hypogene modelling for Mount Tom Price iron ore stands or
the nature of ‘jasperoidal BIF’, evidence of post-ore falls essentially on just one assertion:
metasomatism, source of the hydrothermal fluids, and other • that the hydrothermal alteration of BIF is pre-ore.
points. The reply by Taylor, Dalstra, and Harding (2002)
conceded one important point, ignored some, and failed to Mineragraphic sample data
satisfactorily explain the others.
Further data on some of these discussion points will be given The following discussion is based on the writer’s examination of
here in support of the alternative view that the hydrothermal some 250 - 300 polished thin sections and blocks mostly from
alteration of BIF and ore at Tom Price is a post-ore feature, and Tom Price core samples supplied during the CSIRO-AMIRA
co-operative program on iron ore. About 200 of these are from
thus did not contribute to the genesis of the mpl H ores of the
the critical zones in the hydrothermally affected deposits, and
Hamersleys.
one of the cores, ND 93/80, was described in some detail in by
Some of this mineragraphic data was also included by Barley Taylor et al (2001).
et al (1999) and by Taylor et al (2001), but they did not discuss Samples are from 15 different diamond core holes with depths
the implications of the information they presented. ranging from 97 to 249 m in the North deposit, and 130 to 366 m
Unfortunately, permission to examine the Barley and Taylor in the Southern Batter deposit, together with a grab sample from
material for this paper, and to visit the Mount Tom Price mine, a blast at the North West deposit. A nine metre length of core
was withheld. starting in BIF macroband 5 of the Dales Gorge Member from
DDH 1970/10, about 1 m below the ore zone, was supplied as (2002) that it is found only as a selvedge of a few centimetres to
the then best available ‘fresh BIF’ from Tom Price for the ‘jasperoidal BIF’. The material also occurs in the North West
comparison with BIF from Paraburdoo and Wittenoom (Ewers deposit. The association of silicified ore and the presence of
and Morris, 1981). This core is possibly from the North deposit. altered BIF with 150 - 350°C quartz veins of Taylor et al (2001),
suggests the silica infill came from the adjacent metasomatised
Mount Tom Price Centre Pit and Synclines BIF.
deposits — ‘normal’ mpl H ores
These samples show the normal features of the M-mpl H ores
from around the Province, for the most part consisting of fine to
medium mpl H typically in ex-gangue mesobands between
martite horizons, generally porous, but occasionally associated silica
with goethite. However, samples from the footwall zone of
enriched BIF of the Mt McRae Shale, show apatite and pyrite in
M-mpl H ore. This zone is reported as consistently rich in
phosphorus.
IRON TRANSPORT AND ENRICHMENT The fresh Dales Gorge Member BIF core from Paraburdoo
described by Ewers and Morris (1981) comes from directly
From their data and computer calculations, Taylor et al (2001) below the ore sequence in the Paraburdoo 4E mine. It contains no
claimed that ‘volume loss and not a net addition of iron is the key obvious evidence of metasomatism.
process involved in the progressive build-up of iron to ore grade’
In Figure 3, ore in the main Tom Price deposit is shown
at Tom Price. Following criticism of this modelling by Morris
(2002), Taylor, Dalstra, and Harding (2002) conceded the need terminating against a dolerite intrusive, whereas unenriched BIF
for substantial additional iron, but made no attempt to explain a (except for a small basal enrichment) still exists along the
hypogene source for this iron. This is an unsolved problem that Southern batter fault, the suggested main conduit for the primary
has been the basis of much speculation since the earliest BIF modification fluids. Assuming this cross-section is
hypogene models were suggested. Much of the discussion of this representative of the main deposit, why was the ore formed on
problem has come from North American-Canadian deposits the far side of the dolerite ‘batter’ away from the fault?
where the destruction of potential clues imposed by deep What is the main fluid conduit for the North deposit BIF
weathering, is minimal. alteration (Figure 2)?
A final point here is what happens to the vast amount of silica
A SPECULATIVE SOURCE OF HEATED FLUID – expelled from the BIF following the metasomatic changes,
TAYLOR et al (2001) assuming that all the 900 Mt of ore of Mount Tom Price is
formed by this hypogene mechanism? Massive silica vein
The source of heat for the carbonate metasomatism of the BIF is systems should pervade the surrounding rocks, and the upper
not entirely clear from the Taylor et al (2001) paper, which sections of the faults should be extensively silicified. Much of the
suggested the responsible solutions were related to ore is exposed at the surface and erosion could have removed the
‘overpressured bicarbonate basin brines from the dolomite evidence in the overlying rocks, but there are still abundant areas
aquifer of the Wittenoom Formation’, shown in their Figure 23 as of potential silica metasomatism. Despite extensive mining and
the ‘sealed Paraburdoo Member’. This presumably happened at exploration this expected feature has not been reported.
the ambient temperatures of the strata, since no major thermal
anomaly is represented in the ore (p 859), though ‘ … the
Southern Batter fault may have been associated with a local AN ALTERNATIVE SPECULATIVE SOURCE OF
thermal pulse’. Depth of burial, at temperatures of 150 - 350°C, LOCALISED HEATED FLUID – POST-ORE
based on fluid inclusions in quartz veins outside the ore, would SUPERGENE OXIDATION OF CARBON AND
have presumably been about 5 - 10 km. PYRITE IN THE FOOTWALL MT McRAE SHALE
Problems with the hypogene concept In this alternative model, the localised post-ore metasomatism of
BIF was driven by exothermic reaction of oxygenated meteoric
Whether a ‘sealed’ dolomite-shale sequence, some 600 million water with the rich fine grained pyrite and reactive carbon which
years after it was first deposited and deeply buried, still occurs extensively in the Mt McRae Shale that forms the
contained the vast amounts of basinal solution needed to produce immediate footwall to the Brockman IF. The pyrite in this unit in
the 2000 Ma, 900 million tonnes of ore at Tom Price or 1700 Mt fresh mine exposures tends to be of the ‘poppy seed’ type, but in
at Whaleback, is at best, speculative. An informal estimate of many areas of the mines rich segregations of coarsely
9000 cu km for Tom Price was suggested to the writer by one of recrystallised and marble-sized globular pyrite are present.
the authors. Assuming a generous porosity of ten per cent this Spontaneous combustion and emission of acrid SO2 fumes from
requires an aquifer capacity of ~90 000 cu km. Given a dolomite fractured material exposed during mining, is an indication of the
thickness of around 100 m this represents about 1 000 000 sq km, potential for less dramatic heating in a semi-closed environment.
about ten times the total outcrop area of the Hamersley Province, The pyrite-rich material along the Southern Batter fault (Taylor
without even considering the fluid needed for the ores of Mount et al, 2001, p 849) suggests that considerable sulphur mobility
Whaleback, Paraburdoo, Giles, Channar, etc. Morris (2002) in was achieved.
discussing this fluid source used a very conservative 1000 cu km
based on silica solubility of 0.1 per cent at 350°C needed to In the North deposit (Figure 3) fluid access is not clear, but
remove about one billion tonnes of chert from Tom Price. In may be related to the dolerite shown in the section. It is
reply, Taylor, Dalstra and Harding (2002), cited the Tertiary and important to note that the BIF alteration does not appear in the
present day aquifer characteristics of the Wittenoom Dolomite, to basal zone of the Dales Gorge member here, as would be
suggest recharge by meteoric fluid. Would recharge be feasible in expected under the Taylor et al model, but, consistent with the
the Proterozoic, in a ‘sealed overpressured’ carbonate aquifer, at post-ore alteration concept, it occurs along the permeable
greenschist burial levels, where carbonates tend to ‘heal’ rather pre-existing ore-BIF contact zone.
than maintain porosity? A possible mechanism for the localised flow of heated fluids is
The suggested dolomite aquifer of the Paraburdoo Member described. Following exposure of the buried ore and leaching of
directly overlies BIF of the Marra Mamba IF. In an much of the intersticial goethite under reducing conditions, cool
‘overpressured’ system it would be expected that hydrothermal oxygenated water descended through the permeable ore
fluids would have directly affected the upper levels of this BIF sequences, and reacted with pyrite and carbon in fractured areas
throughout the province, and following erosion and exposure to of the footwall Mt McRae Shale. This is modelled as the cold
deep seated meteoric solutions this unit too should show signs of
down-flowing limb of a convective cell. Heated solutions from
mpl H enrichment. The only reported mpl H ore in the Marra
Mamba IF is in an overturned sequence in the BHPB OB 43 in black shale reactions, passed into the permeable partly oxidised
the Newman area. BIF/ore contact zone of the Southern Batter Deposit, to produce
the metasomatism of the BIF. Thus, this zone acted as the
Under the hypogene model, the BIF (Bruno’s Band) of the Mt
upward-flowing limb of the convective cell. Expelled silica from
Sylvia Formation that overlies the Wittenoom Formation should
also be expected to always show M-mpl H mineralisation, or at the carbonatised BIF was transported through the reported high
least carbonate metasomatism of the BIF, where comparable temperature vein system (150 - 250°C) of Barley, Taylor et alia,
deposits occur in the overlying Brockman IF. This does not invading adjacent porous ore, depositing silica as it cooled, into
appear the case, since at Mount Whaleback and Mount Tom the pore space in the overlying M-mpl H ore. Other components
Price at least, the enrichment of this BIF to both M-G and mpl H such as apatite, chlorite, and pyrite, were also deposited into the
ore is sporadic. ore pore space.
Morris, R C and Fletcher, A B, 1987. Increased solubility of quartz Taylor, D, Dalstra, H J and Harding, A E, 2002. Genesis of high-grade
following ferrous-ferric reactions, Nature, 330:558-561. orebodies of the Hamersley Province Western Australia—a reply,
Morris, R C, Thornber, M R and Ewers, W E, 1980. Deep-seated iron Economic Geology, 97:179-181.
ores from banded iron-formation, Nature, 288:250-252. Trendall, A F and Blockley, J G, 1970. The iron formations of the
Powell, C M, Oliver, N H S, Li, Z X, Martin, D M B and Ronaszecki, J, Precambrian Hamersley Group, Western Australia; with special
1999. Synorogenic hydrothermal origin for giant Hamersley iron ore reference to the associated crocidolite, West Aust Geol Surv Bull,
bodies, Geology, 27:175-178. 119, 336 pp.
Taylor, D, Dalstra, H J, Harding, A E, Broadbent, G C and Barley, M E, Van Hise, C R and Leith, C K, 1911. The Geology of the Lake Superior
2001. Genesis of high-grade hematite orebodies of the Hamersley Region, US Geological Survey Monograph, 52: 641 pp.
Province, Western Australia, Economic Geology, 96:837-873.