J UL ULY 2 0 0 9 • N o 7 8
S EG E G N E W S L E TT TT E R
13
Discovery and Geology of the Platinum Group Element Depos Depo sits of the Bus ushve hveld ld Com plex , South Afr Afric ica a ELL L R O G E R N . S C O O N (SEG 1985 F),† Postne t Suite Suite 291, Pr Priva iva te Ba g X3 X31, 1, Knysna Knysna 6570, Sou th A frica , an d A NDREW A . M ITCH EL (SE (S EG 200 2001 1 F), F), Dep a rtme nt o f G eo log y, Univers University ity of Kwa Zulu-Nata l, Pri Priva va te Ba g X540 54001, 01, Durb Durb a n 4000, 4000, Sou Sou th A fri fric c a
INTR INT RO DUC TIO N
(Hall, 1932). Here, the Tran sv svaa aa l Supergroup Supergroup The Bush veld Complex, wh ic ich h ha s been been tilted tilted and includes the largest known up li lifted fted in response to mafic intrusion in the world, vertical tectonics reveals som som e specta specta cular geolinduced by intrusion of ogy, includin includin g th e classic classic layerthe m a fi ficc rocks rocks (Da (Da ly, ing of dark ultram afic aficss and 1926) to form form a 2,000light-colore lig ht-colored d a north osi ositic tic rocks m-high range, whereas (Fig. (F ig. 1). Th Th ree distinct gro up s of the Rustenbu rg La La yered rocks roc ks ma ke up the comp lex Suite crops out in wide (Taa nka rd et al., 1982), (T 1982), the oldvalleys associa associa ted with est of which is the Rooiberg the Olifants River sysGroup Gro up (domina ted by rh yol yolite ite)) tem, a nd a ls lso o forms forms a and the youngest, the Lebowa rugged, 250-km-long Gran ite Suite Suite (includin (includin g a su ite escarp esc arp m ent. This area of granophyres). The majority influenced infl uenced deve development lopment of the ore deposits are, howof the Great Escarpment, ever, restric restricted ted to th e interven FIG URE 1. Classic layering at the contact between the Lower Critical which whic h in tu rn formed by ing group of ultram ultram a fic ic-ma -ma fi ficc and Upper Critical zones in the Olifants River section, Eastern limb. upliftt during the uplif rocks, or Rustenburg Layered Dark-colored Dark -colored laye rs of feldspathic orthopyroxenite and chrom iti itite te a re Cenozoic era (Du Toit, Suite. The Rustenburg Layered interca inter ca lated with light-c light-c olored laye rs of a northos northosite. ite. 1933) and ha s resulted resulted Suite is intrusive into the sediin a disc discrete rete geograph ic ic-m enta ry roc rocks ks an d subordina subordina te with wi th a nu m ber of sm sm aller satelli satellite te bodbotan ic sy syste stem m known a s the volcaa nic rocks of the Tra volc Tra nsva a l ies.. The ies The ea stern limb is particularly well Middleveld (located Supergroup (ca. 2.5–2.1 2.5–2.1 Ga) an d h a s to page 14 known owing to the excellent exposures between betwe en th e centra centra l been da ted at 2.055 Ga (Scoates (Scoates and Friedma n, 2008). The The Bushveld conta conta ins the greatest concentration concentration of min eral wealth on th e plan et an d includes, includes, in in addition to the platinum group elemen elemen t (PGE (P GE)) ores, base m etals (e.g., (e.g., chrom ium , iron, ir on, tin, tin, titanium titanium , and van adium ) and industrial minerals (e.g., andalusite, dimension stone, stone, an d m agn es esite ite), ), as described desc ribed by Willem Willem se (1969) an d Wilson and Anhaeusser (1988). Mining operations ha ve spa spa wned extensive extensive towns an d a ss ssoci ociated ated in dustri dustrial al com plexes, including down stream process process-ing plan ts with with smelte smelters rs an d precious precious metal refineries, many of which use technology speci specifi fically cally designed for the Bushveld ores.
. . .
TH E RUS RU STEN BURG LAYERED SUITE The Rustenb urg La yered Suite Suite form s three p rincipal lim bs (Fig. (Fig. 2) together Corres Cor responding ponding au thor: e-ma il, rnscoon @iafrica.com †
FIG URE 2.
Sche ma tic ma p o f the Bus Bushveld hveld Co mp lex de pic ting the three ma in li limb mb s of the intrusion with mines and other localities referred to in the text: 1 = Winnaarshoek, 2 = Maandagshoek, Mooihoek and Driekop, 3 = Onverwacht, 4 = Brits, 5 = Kroondal, 6 = Union, 7 = Ama nd elbu lt, 8 = San dsl dsloo oo t. Out liliers ers an d sate llllites ites not show n.
14
. ..
S EG N E W S L E TTE R
from
13
N o 7 8 • J ULY 2 0 0 9
Disco very and Geo logy o f the PGE Depo sits of t he Bushveld Com plex, South Africa (Continued )
plateau a nd coastal lowlan ds). In contrast, the western an d northern lim bs a re covered by th ick soils on th e Highveld. Knowledge of th ese area s is largely based on m ining a ctivity an d extensive drilling programs, most nota bly in the western limb, where m ining h a s revea led details of som e of the econom ic reefs over m a ny ten s of kilometers. Despite excellent overviews of the Rustenbu rg La yered Suite (Wager a nd Brown, 1968; Von Gruen ewaldt et al., 1985) our u nderstanding is inh ibited by subjective interpretion of field relationships between widely separated cham bers which a re un likely to be interconnected. The ea stern a nd western limbs are subdivided into sectors on th e ba sis of regional differences in the stratigraphy, althou gh this has never been form alized. Sectors are sepa ra ted by structurally complex bounda ry areas. Regular cha nges in the stratigraph y of the south ern sector of the western limb led Ea les et al. (1988) to re cognize different facies (broadly referred to a s proxima l an d distal) on the ba sis of (theoretical) feeder localities. This interpretation is probably widely applicable an d can a ssist with our un derstan ding of similarities between, for exam ple, the distal components of both the western a nd ea stern limbs. Sectors a re furth er subdivided into subchambers, typically from deta iled knowledge of chrom itite layers an d PGE reefs. In the n orthern sector of the western lim b, for exa m ple, the Union (Viljoen et a l., 1986a ) an d Ama nd elbult (Viljoen et a l., 1986b) mines comprise discrete subchambers. Facies chan ges ma y in p art b e a scribed to synBush veld tectonism (Scoon a nd Teigler, 1994), despite a conflict with a widely held belief tha t layered intru sions occur in stab le contin enta l settings (e.g., Irvine, 1982). Structural disrup tions in the Bushveld include diap irism of the floor rocks (Uken a nd Watkeys, 1997), and development of the Rustenbu rg La yered Suite on th e fla nks of domes wa s influen ced by associated episodic uplift (Scoon, 2002). This is a scribed to m ultiple pha ses of ma gma replenishm ent—an importan t tenet of our un derstan ding of the Rustenbu rg Layered Suite, a s proposed a nd discussed by Ea les et al. (1988). The Rustenbu rg La yered Suite a tta ins a m ax imum thickness of some 9 to 12 km within th e western a nd ea stern
lim bs, whereas th e sequence in the on th e som ewhat h igher grade layers in northern limb is significantly truncated. the Lower Critica l zon e an d lower parts On a sma ller scale, each sector (and of the Upper Critica l zone (Schurm a nn each subcham ber) reveals considerable et al., 1998). The Ma in zone th a t overvaria bility (South Africa n Com m ittee lies the Critical zone is cha ra cterized by for Stra tigraph y, 1980). The u ncon relatively monotonous sequences of formable relationship with the floor norite an d ga bbronorite, a lbeit with rocks, which is most appa rent in th e prom inent la yers of an orthosite. The Upper zone includes num erous Ti ma geastern limb (Sha rpe, 1981), results in netite la yers intercalated with m a gnetite the lowerm ost parts of the Suite sequence bein g spa tially restricted in gabbro, anorthosite, and ferrodiorite. their development, with on ly the upp erm ost la yers being la terally extensive. Various parental m agm a types have PLATIN UM DI SC O V ERI ES been iden tified to explain th e developDetails of the original discoveries of ment of the Suite (e.g., Eales, 2002), platinum are summ arized here from over an d a bove the effects of upwa rd the review by Scoon a nd Mitchell fractional crystallization (Wager and (2004a), which in turn was ba sed on Brown, 1968), althou gh th e m echa a rticles an d editorial com m ents in the nisms of m agm a intrusion are poorly popular mining press of the time (e.g., constrain ed. Scoon a nd Teigler (1994) Merensky, 1925), as well as on a biog an d Uken a nd Watkeys (1997) sugra ph y of Han s Merensky (Fig. 4) by gested tha t the therma l and loading Lehmann (1955). Exploration started effects of new influx es are im porta nt in enhancing floor rock irregularities, which must be accounted for in the interpretation of lateral va riation s between sectors an d subcha mbers. The Rustenb urg La yered Suite is subdivided into zones on th e ba sis of a repetitive cyclicity an d la terally exten sive m a rker layers (Fig. 3). Th e Mar gina l zone consists of relatively fine grained norite and feldspa thic pyroxen ite with little discernible layering. The Lower zone is a lmost entirely ultram afic an d is domin ated by layers of dun ite, ha rzburgite, an d orthop yroxenite. It is the Critical zon e (Hall, 1932), h owever, tha t reveals the m ost spectacular layering and contains the PGE deposits an d chrom itite layers. The m odern p ra ctice is to recogn ize a Lower Critica l zone, domina ted by feldspathic orthopyroxenite, an d an Upper Critical zone, with m ore complex la yering of lithological un its, typically including feldspa thic orthopyroxenite an d norite-anorthosite. Chromitite layers, which are a defining cha ra cteristic of both the Upper FIG URE 3. Generalized vertical section of the laya nd Lower Critica l zones, are ered sequence from the eastern limb of the subdivided into Lower, Middle, Bushveld Co mp lex, dep icting zona l subd ivisions an d Upper groups. Chrom ite and impo rta nt ma rker layers. m ining is m ostly concentrated
J ULY 2 0 0 9 • N o 7 8
FIGURE 4.
Photograph of Dr Hans Merensk y “ p o i nt i ng o u t a d i a m o n d i n m a t r ix from Alexan de r Bay (from Lehm ann , 1955).
when a sample pann ed by Andries Lomba ard from an ephemeral stream on h is farm Ma an dagshoek, located some 40 km from Lydenbu rg, was dispa tched in June 1924 to the Joha nn es burg office of Merensky. The assay reported n ative Pt an d Au, together with iron oxide a nd traces of Rh an d Ir. Merensky imm edia tely un dertook a field visit, during which h e requested Lomba ard, a n experienced gold prospector, together with his cousins Schalk a nd Willem Schoem an , to continue the search. The “Lydenburg Platinu m Syndicate” wa s formed by Merensky, privately funded by close friends, with the objective of locating alluvial an d h a rd-rock PGE ores. After acquiring some mineral rights titles, Merensky returned to the field to find that th e Schoem an brothers had pa nn ed Pt in soils to the ea st of the stream. Three days later, on August 15, 1924, the Syndicate located Pt in ou tcrops of “dark, lustrous crystalline pyroxenites and ultrabasic rocks” on Mooihoek to the ea st of Maa nda gshoek. The mineralization occurred in a discorda nt bod y (pipe) over which th e synd icate h a d to a pply for claim s. Discovery of the Driekop pipe is credited to Willem Schoem a n, wh o recollected seeing similar rocks on a sma ll hill to the north of Maa nda gshoek. The Syndicate also loca ted the low-gra de Twyfelaa r pipe, but the Onverwacht pipe was discovered by Ra nd Mines (in October 1924), with geologist F.W. Blain e un derta king the field program . No additional platiniferous pipes h a ve been discovered in the Bushveld, despite the subsequ ent recognition of ma ny th ousan ds of discordant bodies! In Septem ber 1924, the Synd ica te m ade the “far m ore importa nt finding” of a layered reef at Maa nda gshoek. This
S EG N E W S L E TTE R
was initially credited to Lom ba a rd, as he u ndertook the rock chip sam pling, but the prospecting team insisted it be na m ed the “Merensky reef.” Merensky was a ware tha t the pipes were of limited size an d was convinced tha t layered rocks, specifica lly ultra m a fics with seconda ry Cu, were a far m ore im portan t ta rget. Wagn er (1929) referred to the Merensky reef a s the “Moth er Lode” despite the pipes being successfully m ined, an d despite problems with exploitation of the Merensky reef as described below. The Syndicate delineated the reef over much of the eastern lim b, an d several m onth s later found the reef in the western limb ; Merensky also assisted with discovery of the Platreef in the northern limb. Addi tiona l fun ding was acquired an d the company was renamed “Lydenburg Platinum Ltd.” Evaluation of the three pipes, the Merensky reef, an d some alluvial concentrations was un dertaken, an d ma ny of the trenches an d underground workings from this period can still be exam ined, including several reef declines an d drives at th e Winna arshoek locality (Mitchell and Scoon, 2007). In 1925, the compa ny was purcha sed by the Gold Fields group a nd floa ted on both the JSE a nd LSE durin g a shortlived boom when the Pt price was five times tha t of gold, driven by shorta ges and stockpiling. It is interesting th a t Merensky’s pa rents, who were German missionaries, pa ssed through the discovery a rea while fleeing a n u prising in Sekhukhu nelan d prior to settling a t Botsha belo, not far from the ea stern limb of the intrusion, where Hans was born a nd lived a s a young child. Merensky played a pivotal role in a nu m ber of additiona l discoveries, includin g the west coa st dia m ond fields, the chrom ite deposits a t Jagdlust, the apatite orebody at Phalaborwa, and the southern extension of the Wit watersra nd gold fields, an d severa l of the Lombaard and Schoeman families beca m e successful geologists.
EXPLO RATIO N M ETHO DO LO G Y Our research ha s led us to conclude th at South Africa n geologists, includin g Merensky, ha d speculated for ma ny years a bout the p ossibility of find ing Pt in u ltram a fic rocks of the Bushveld Complex. Merensky first sampled and a ssa yed rocks from the ea stern limb in 1904, including chrom itite layers. The association of PGE with the Bushveld
15
chromitites was discovered by Bettel (1925) in 1906 an d it should be noted tha t chrom ite was mined from th e Bushveld long before the platinum discoveries, includin g a t th e Winterveld m ine, where the Onverwacht pipe is loca ted. Merensky also provided sa m ples for Hall and Humphrey (1908), who reported tha t one chrom itite layer yielded a gra de >6 g/t (th e UG2?). This did not con stitute a “discovery,” as the finen ess of the PGM mea n t extraction problems were not resolved for many years (Verm a a k, 1985). Discovery of the nickel-rich Vlakfontein pipes in 1923 was also im portan t, an d the a ssociation of Pt a nd chrom itite as well as sim ila rities between the p ipes an d the deposits of the Russian Urals were widely discussed. During an nou ncemen t of the discoveries, Merensky was su pported by both A. L. Hall and P. A. Wagner, an importa nt point as finan cial sca nda ls resulting from earlier “discoveries,” together with p roblems with th e Waterberg vein deposits, ha d ha m pered previous investmen t. In summ ary, we recognize three im portan t com ponen ts to the exploration m ethodology: knowledge of field relation ships, an exploration m odel focusing on ultram afic rocks, an d use of stream sediment sam pling despite th e polygenetic source (Oberthu r et al., 2004).
SO M E EA R LY M I N I N G H I STO RY Prior to min ing of the Bushveld ores, the m ain supply of Pt was from a lluvial deposits, m ostly located in Russia. The Bush veld pipes a re the oldest un derground Pt m ines (Onverwacht was opened in 1925), with the exception of sm a ll-scale workin gs at Solovyov Hill in the Urals. The p ipes were evalua ted by core drilling and trial mining. They yielded spectacular grades, notably at On verwacht, where resources were initially ca lcula ted as 55,000 t at 16 g/t PGE (>90% Pt) to a d epth of 76 m (min ing eventu ally attained a depth of 320 m ). They p roved rela tively easy to m ine a nd process, as ore m inerals, dom ina ted by sperrylite a nd Pt-Fe a lloy, were coarse-grained a nd a m ena ble to gravity concentra tion (Wagner, 1929). Production costs am ounted to h alf the Pt price. In comparison, development of the Merensky reef proved fa r m ore difficult, owing to h igher m ining costs, m etallurgical p roblems cau sed by oxidation of near-surface ore, and the presence of to page 16
. . .
16
. ..
N o 7 8 • J ULY 2 0 0 9
S EG N E W S L E TTE R
from
15
Disco very and Geo logy o f the PGE Depo sits of t he Bushveld Com plex, South Africa (Continued )
ba se-m etal sulfides in deeper ore (Wagn er, 1929). Som e area s were trial mined, but low metal prices in 1929 resulted in closure of m ost operation s in the Bushveld, including those on the pipes. The on e exception wa s Kroonda l, where minin g of the thin yet extraordina rily high gra de Merensky reef (ore was ha nd sorted prior to milling) proved to be viable (Wagn er, 1929).
THE PG E D EPO SITS
The UG 2 re e f The UG2 reef is a textbook ex a m ple of a stratiform deposit: mineralization is constrained to a layer of chromitite, a lbeit loca lly with one or m ore barren partings of feldspathic orthopyroxenite, which m ay va ry from a few centimeters to severa l m eters th ick (Fig. 5A). The UG2 is persisten t in som e sectors of the eastern and western limbs for tens of kilometers along strike and to depths of a t least several kilometers. The footwa ll an d h an ging-wall rock un its reveal considerable lateral variation, but, typically, the ch rom itite occurs in a ssocia tion with feldspa thic orthopyroxenite an d noriteanorthosite. Mining of a narrow reef, pa rticularly in h igh tonna ge operations,
FIGURE 5.
is a specialized business an d in th e case of the UG2 is exa cerbated by prob lems caused by potholes (Lom berg et al., 1999), area s from a few m eters to several hu ndreds of m eters in dia m eter where pa rt of the footwa ll sequence is absent. Dilution resulting from h an ging-wall collapse due to p lan es of weakness associated with thin chromitite stringers (“lea der seam s”) is a n a dditiona l problem. Ore reserve calculation s, excluding losses due to poth oles a nd d ilution, are, h owever, relatively sim plistic, despite evidence of a pron oun ced nu gget effect. The double peak of minera lization (bottom a nd either middle or top loaded, depen ding on layer th ickness) is persistent throughou t the intrusion, and is defin ed by total PGE a nd in terelem ent ratios (Hiemstra, 1986). The antipathetic relationship between grade an d thickness has been widely remarked upon (e.g., Viljoen a nd Schurm a nn , 1998). The footwall lithology of the UG2 varies from a north osite, leuconorite, or orthopyroxenite to pegmatoidal feldspa thic pyroxenite. Minor a m ounts of PGE are located in the footwall, specifically if it is pegm a toidal, but a re typically directly associated with dissem ina tions or stringers of chrom itite. The irregular ba sal contact of the chromitite
Vertical profiles of some typical reefs: (A) UG2 reef from the Marula mine, Winnaa rshoe k; (B) M erensky “ thin-ree f” from the Brakspruit sec tion, Rusten bu rg Pla tinum mine (a fter Viljoe n and Hieb er, 1986); (C) Me rensky “w ide-reef” from the Ma rula m ine, Winnaarshoek.
when the footwall is pegm atoidal contrasts sharply with the plan ar u pper contact. The chrom itite layer ma y be thinn er in a reas where the pegma toid contains a bunda nt chromitite, an observation the a uthors have m ade a t the Crocodile River min e, where la rge a reas of the UG2 are un derlain by leuconorite. The grad e of th e UG2 is typically 4 to 8 g/t PGE+Au. The content of base-metal sulfides is typically very low (avg <200 ppm Ni an d Cu). The PGE occur a s discrete, very fine PGM (Kinloch, 1982); th e ba se-m etal sulfides are domina ted by pentlandite and cha lcopyrite (McLaren and De Villiers, 1982). Fortuitously, the higher va lue PGE (Pt an d Rh) together with Pd occur in PGM located in interstitial sites, ena bling recovery by fine grinding an d sophisticated flotation techniques, as opposed to the bu lk of the Ru an d Ir, which occurs in lau rite tha t rema ins locked within the gra ins of chrom ite.
Th e M e r e n sk y r e e f The Merensky reef is also latera lly contiguous through out large area s of the intru sion, but wh a t is less well known is tha t in deta il it is extraordina rily complex, particularly in compa rison to th e UG2, an d over la rge area s of the ea stern a nd western limbs th e grade is too low to justify m ining. Resource estim a tion m odels are frau ght with un certainty. This is over an d a bove problem s ca used by estima ting loss of resources due to poth oles (Viring a n d Cowell, 1999), some of which form catastrophic or regional structures, and discordant bodies of iron-rich ultramafic pegmatite (Viljoen and Scoon, 1985). Cryptic layering of PGE in th e reef is a n en igma tic feature a nd presents difficulties with optim izing m ining cuts. PGE typically occur with in a comp osite series of la yers severa l m eters wide, including two chrom itite stringers (typically <10 m m wide), with th e h ighest gra des restricted to a na rrow part, or pa rts, of this reef zone, as typified at th e Ama ndelbult m ine (Viljoen et a l., 1986b). Where a so-called thin-reef facies is developed (Fig. 5B), as described at the Brakspruit section of th e Rustenburg m ine (Viljoen a nd Hieber, 1986), or where the en tire reef is condensed into a single, relatively th ick, chrom itite stringer, a s in pa rts of Imp a la m ine (Leeb-Du Toit, 1986), cryptic la yering is difficult to identify and some spectacular grades
J ULY 2 0 0 9 • N o 7 8
m a y be preserved (>50 g/t). In som e a reas, PGE a re a ssocia ted with in terstitial base-metal sulfides that extend into the footwall (Cawth orn, 1999). In som e pa rts of the ea stern limb, cryptic layering occurs on a sca le of centim eters in a wide-reef facies (Fig. 5C), a s described in the Winnaarshoek locality by Mitchell a nd Scoon (2007). In the wide-reef facies, two ch rom itite stringers constrain the width of the mineralized reef zone, as is typically the case, but th ey are sepa rated by a layer of feldspath ic orthop yroxenite tha t is relatively thick (1.8 m on a verage) an d includes a ba rren m iddling. Moreover, the lithology between th e chrom ite stringers throughout m ost of the western limb is a pegma toidal orthopyroxenite, an unu sual assemblage the origin of which ha s been m uch debated (e.g., Cawthorn a nd Boerst, 2006), wherea s in th e ea stern lim b th e principa l layer of pegm atoid occurs below the m inera lized zone. The average grade of the Merensky reef is sim ilar to th e UG2, a lthough it is far m ore variable. Typically, a m uch larger n um ber of PGM species occur (Kinloch a nd Peyerl, 1990) an d latera l varia tion of the PGM contra sts with the regula rity of the prima ry la yering. PGM are spatially associated with base-m etal sulfides (1–2 %), constituting a pprox im ately equa l abun da nces of pyrrhotite, pentlan dite, an d cha lcopyrite; the relative paucity of pyrrhotite is unusual (Liebenberg, 1970). Base-metal sulfides are a n im portan t by-product from the Merensky a nd a ssist with down stream smelting operations, whereas smelting of sulfide-poor UG2 concentrates is problematic.
The Pla tre e f The Platreef is restricted to a rela tively sma ll area of the northern limb, an d is typically located a few tens of m eters above th e ba se of the intrusion, where the floor ran ges from Archean gran ite to sedimentary rocks of the Transvaal Sup ergroup. The lowermost zones of the Rustenbu rg La yered Suite a re n ot developed in th is a rea. The Platreef is fa r m ore irregular th an the UG2 an d Merensky, is typica lly mu ch th icker (up to 100 m or m ore), an d contains PGE tha t a re invariably a ssociated with ba se-m etal sulfides throu ghou t (Viljoen an d Schurm an n, 1998). The Platreef is also chara cterized by a m uch lower Pt/Pd ra tio than the Merensky an d UG2 reefs, a lthough the lower average grad es a re offset by the width of the reef. The
17
S EG N E W S L E TTE R
Platreef m a y constitute a loca lized equivalent to the Merensky reef, an d a t the Tweefontein loca lity, a few kilometers to the south of the San dsloot m ine, the m inera lized sequence is rela tively thin a nd in cludes both chromitite stringers an d a p egma toida l pyroxenite (Viljoen a nd Schurm an n, 1998). We find th e importa nce of footwall contam ination to th e Platreef ha s been oversta ted, in pa rt as ea rlier studies were restricted to an area where the floor rocks are dolomite (Wh ite, 1994). A new termin ology ma y assist with explaining this: the sequence below the (barren) Main zone shou ld be categorized as a Platreef un it rath er than the “Platreef.” This consists of gabbronorite, feldspa thic websterite, feldspa thic harzburgite, and reconstituted feldspathic pyroxenite, the latter revealing a pegmatoidal or glassy texture. These rock units are in discorda nt in trusive rela tionship with on e an other an d with the overlying Main zone. Thus the sequential stratigraphy (A, B, and C reefs of ea rlier workers) is ina pprop riate, yet despite this, a Main mineralized layer (an a logous to the “B” reef?) can generally be identified. At the Akanani loca lity, where deep drilling h a s revealed a relatively thick Pla treef un it downdip from the open pit San dsloot deposit, we found PGE to be m ost abu ndan t in sinuous layers of ha rzburgite a nd , to a lesser extent, reconstituted pyroxenite; the gabbronorite and feldspa thic websterite are rela tively weakly mineralized.
The pipe de po sits The p ipe deposits a re no lon ger of com m ercial interest a nd, p a radoxically, ha ve a n egative affect on reef-type m ines, as they disrupt th e layered wall rocks. PGE were concentra ted in sma ll core zones (ma x diam 24 m ) of coarsegrained, iron-rich dun ite an d wehrlite (Wagner, 1929), now m ostly mined ou t. The bu lk of each p ipe (diam >300 m ), however, is domina ted by barren m agnesian dun ite. Barren outer envelopes also occur, an d our unpu blished ma pping a t Mooihoek dem onstrates this compon ent is even m ore extensive tha n the m agn esian dun ite. Bushveld pipes tha t do not reveal this zona tion are typica lly barren (Viljoen a nd Scoon, 1985).
LO W- G RA DE PG E DEPO SITS An un derstan ding of the econom ic deposits is incom plete withou t m ention
of low-gra de ores in the Bush veld. These include all the chrom itite layers located below the UG2 reef, a s well as th e Pseudoreefs tha t are situa ted between the UG2 an d Merensky in th e northwestern Bushveld. The Bushveld chromitites reveal regular upwa rd trends in both their comp osition (e.g., decrease in the Cr/Fe ratio) an d the PGE grade and tenor (Scoon and Teigler, 1994). These chrom itite layers, h owever, report very low con tent s of sulfide. These data are important a s they demonstrate tha t the UG2 an d Merensky reefs do not occur in isolation bu t represent the culm ination of a general upward increase in th e PGE content of m inera lized layers within th e Rustenb urg Layered Suite. This trend is disrupted by th e low-grad e, ha rzburgitic Pseudoreefs. The bifurcation a nd subsequent elimina tion of discrete layers of ha rzburgite (boun ded top a nd b ottom b y stringers of chrom itite) is indicative of rejuvenation of more primitive ma gma s that were intruded latera lly an d which m ay be interpreted as precursors to the Merensky reef (Scoon a n d De Klerk, 1987).
G EN ESIS A N D C O N C LU DI N G REM A RK S The h olistic approa ch a dvocated by Wagn er (1929) is im portan t to our un derstan ding of th e PGE deposits in the Bushveld Com plex. There is a m a rked spatia l associa tion of PGE with chromitite a nd other ultram a fic rocks that ha ve sha rply defined and demonstrably discordan t ba sal contacts, a nd not uncommonly sharp upper contacts. Norite-anorthosite is almost invariably ba rren, except in isolated ca ses. These relationsh ips, in a ddition to the restriction of econom ic ores to the Upper Critical zone, a re un likely to be coincidenta l. Rath er than the m odel of Cam p bell et al. (1983), in wh ich the resident m agm a colum n provides the PGE, we believe field relationships and mass bala nce considerations a re consistent with the “lateral m ixing hypothesis” of Scoon a nd Eales (1989) an d Scoon a nd Teigler (1994). Thin, hot layers of fresh, ultram a fic m agm a, enriched in PGE (an d chrom ite: Ea les, 2000) stream ed laterally into the different chambers of the intrusion. In the case of the Merensky reef, as well as some oth er un its, the new, U-type ma gma was intruded into an earlier-formed crystalline substrate of to page 18 norite-anorthosite
. . .
18
. ..
N o 7 8 • J ULY 2 0 0 9
S EG N E W S L E TTE R
from
17
Disco very and Geo logy o f the PGE Depo sits of t he Bushveld Com plex, South Africa (Continued )
(Mitchell a nd Scoon, 2007), an imp orta nt consideration if, as suggested by Ea les et a l. (1986), the lowermost pa rt of the Main zone was emplaced after the reef. A sim ila r h ypoth esis is app lica ble to the Platreef: U-type m agm a intruded earlier-formed gabbron orite a nd websterite, forming mineralized harzburgite an d pegma toid, which in turn wa s emp laced prior to the Main zone. Chromitite layers formed by mixing of U-type an d m ore evolved m agm a (Atype), a hypothesis presented by Sharpe a nd Irvine (1983), a lbeit we ha ve suggested the la tter, at least in the Upper Critica l zone wa s derived by pa rtial m elting of plagioclase cumu lates under the in fluen ce of the n ew influx es of Utype ma gma . The spectacular grade of some ch rom itite stringers, a s well as PGE-bearing chromitites with low sulfide contents, suggests two processes— nu cleation of PGM triggered by crystallization of copious amounts of chromite a nd S-satu ra tion—were necessa ry to genera te econom ic deposits. U-type m agm as also intruded the complex in vertical cond uits (Scoon a nd Mitchell, 2004b). Pa rtial m elting of ea rlierformed reefs, triggered by h eat a ssociated with th e ma gnesian dun ite, genera ted a PGE- an d iron-rich m elt which drained downward into the cores of specific pipes, a hyp othesis sup ported by sim ilarities with so-called repla ced Merensky reef (reef invaded by discordan t iron-rich ultram afic pegma tite: Scoon a nd Mitchell, 2004a ). In sum m ary, we find tha t hydrotherma l fluids are of little im portan ce in forma tion of PGE deposits in the Bushveld, as discussed by Barnes and Campbell (1983), an d recognition of orthom a gma tic processes is fundamental. The uniquely layered Upper Critica l zone, conta ining th e world’s prem ier PGE ores, resulted from episodic replenishm ent b y U-type m agm as tha t persisted long a fter the onset of the crysta llization of noriteanorthosite.
A C KN O W LED G M EN TS We a cknowledge reviews by Wolfga ng Maier an d an a non ymou s reviewer, an d a re pa rticularly grateful to the form er for his constructive com m ents. We pa y tribute to Hugh Ea les for his m entorship ea rly on in our careers, an d for his ongoing interest in our research.
REFEREN C ES Ba rnes, Stephen J., an d Cam pbell, I.H., 1983, Role of la te m a gm atic fluids in Merenskytype pla tinum deposits: Geology, v. 16, p. 488–491. Bettel, W., 1925, Occurren ces of pla tinu m in the Tran svaa l: South African Mining a nd Engineering Journa l, no. 36 (March 14), p. 35 . Cam pbell, I.H., Naldrett, A.J., a nd Barn es, S.J., 1983, A mo del for the origin of the platin um -rich sulfide horizons in th e Bushveld a nd Stillwater Com plexes: Journal of Petrology, v. 24, p. 133–165. Cawthorn , R.G., 1999, Perm eab ility of the footwall cum ulates to th e Merensky reef, Bushveld Com plex: South Africa n Journ a l of Geology, v. 102, p. 293–302. Cawth orn , R.G., an d Boerst, K., 2006, Origin of the pegma titic pyroxenite in th e Meren sky Unit, Bushveld Complex, South Africa: Journal of Petrology, v. 47, p. 1509–1530. Daly, R.A., 1926, Our mobile Earth: New York, Scribners, 324 p. Du Toit, A.L., 1933, Crustal m ovemen ts as a factor in th e geograph ica l evolution of South Africa: South Africa n Geograph ic Journal, v. 16, p. 3–20. Ea les, H.V., 2000, Imp lication s of the chromium budget of the Western limb of the Bushveld Com plex: South Africa n Journal of Geology, v. 103, p. 141–150. ——2002, Caveats in defining the magmas pa renta l to the m a fic rocks of the Bushveld Complex, an d the ma nn er of their empla cement: review and commen tary: Mineralogical Maga zine, v. 66, p. 815–832. Ea les, H.V., Ma rsh , J.S., Mitch ell, A.A., de Klerk, W.J., Kru ger, F.J., a n d Field, M., 1986, Some geochemical constraints upon models for the crysta lliza tion of the upp er Critical zone-Main zone interva l, northwestern Bushveld Com plex: Mineralogica l Maga zine, v. 50, p. 567–582. Ea les, H.V., Field, M., De Klerk, W.J., a n d Scoon, R.N., 1988, Regional trends of chemical variation a nd th erma l erosion in th e Upper Critical Zone zone, western Bushveld Comp lex: Mineralogical Maga zine, v. 52, p. 63–79. Ha ll, A.L., 1932, The Bushveld Ign eous Com plex of the centra l Tran svaa l: Geological Survey of South Africa Memoir 28, 560 p. Hall, A.L. a nd Hum ph rey, A.W., 1908, On the occurrence of chrom ite deposits along the southern an d eastern m argins of the Bushveld Plutonic Comp lex: Tran sactions of the Geological Society of South Africa, v. 11, p. 75–76. Hiemstra, S.A., 1986, The distribution of chalcophile an d platinum -group elem ents in th e UG-2 chrom itite layer of the Bushveld Complex: Economic Geology, v. 81, p. 108–1086. Irvine, T.N., 1982, Term inology for layered intrusions: Journal of Petrology, v. 23, p. 127–162. Kinloch, E., 1982, Regional trends in the platinum -group m inera logy of the critical zone of the Bushveld Com plex: Econom ic Geology, v. 77, p. 1328–1347.
Kinloch, E., and Peyerl, W., 1990, Platinum group m inerals in va rious rock types of the Merensky Reef: Genetic interpretations: Econom ic Geology, v. 85, p. 537–555. Lehm a nn , O., 1955, Look beyond the wind: the life of Dr Hans Meren sky: Cape Town, Howard Tim m ins, 183 p. Liebenberg, L., 1970, The sulphides in the layered sequence of the Bushveld Igneous Complex: in Symp osium on th e Bushveld Igneous Complex a nd other layered intrusions, in Visser, D.J.L. a nd Von Gruen ewa ldt, G., eds., Geological Society of South Africa Special Publication 1, p. 108–207. Leeb-Du Toit, A., 1986, The Impa la p latinu m mines, in Anh aeu sser, C.R., an d Ma ske, S., eds., Mineral deposits of southern Africa: Johannesburg, Geological Society of South Africa, v. 2, p . 1091–1106. Lom berg, K.G., Ma rtin, E.S., Patt erson, M.A., an d Venter, J.E., 1999, The m orphology of potholes in the UG2 chrom itite layer an d Merensky reef (pothole facies) at Union Section, Rustenburg p latinu m m ines: South African Journal of Geology, v. 102, p. 209–220. McLa ren, C.H., an d De Villiers, J.P.R., 1982, The platinum -group chemistry an d m ineralogy of the UG2 chromitite layer of the Bush veld Comp lex: Econom ic Geology, v. 77, p. 1348–1366. Merensky, H., 1925, How we discovered platinum : Mining a nd Industrial Maga zine of South ern Africa, v. 1 (Dec 2), p. 265–266. Mitchell, A.A., an d Scoon, R.N., 2007, The Merensky reef at Winn aa rshoek, Eastern Bushveld Complex: A prima ry ma gma tic hypoth esis ba sed on a wide reef facies: Econom ic Geology, v. 102, p. 971–1009. Oberth ur, T, Melcher, F., Gast, L., Wohrl, C., an d Lodziak, J., 2004, Detrital pla tinum group m inerals in rivers dra ining the ea stern Bushveld com plex, South Africa : Can a dian Mineralogist, v. 42, p. 563–582. Schurm an n, L.W., Grabe, P-J., an d Steenka m p, C.J., 1998, Chromiu m , in Wilson, M.G.C., an d Anha eusser, C.R., eds., The m ineral resources of South Africa: Handbook, Coun cil for Geoscience, v. 16, p. 90 –105. Scoa tes, J.S., a nd Friedm a n, R.M., 2008, Precise age of the platiniferous Merensky reef, Bushveld Complex, South Africa, by the U-Pb Zircon chemical abrasion ID-TIMS technique: Econom ic Geology, v. 103, p. 465–471. Scoon, R.N., 2002, A new occurren ce of Merensky reef on th e fla nks of the Zaa ikloof dome, northea stern Bushveld Complex: Rela tionship between diapirism an d ma gma replenishm ent: Economic Geology, v. 97, p. 1037–1049. Scoon, R.N., a nd De Klerk, W.J., 1987, Th e re lationship of olivine cumulates and mineralization to cyclic un its in pa rt of the Upper Critical zone of the western Bushveld Com plex: Ca na dian Mineralogist, v. 25, p. 51–77. Scoon, R.N., and Eales, H.V., 1989, Platinumgroup element mineralization in Critical zone of Bushveld Complex, South Africa: Distribution an d m ineralization types [ext. a bs.]: 28 th Interna tiona l Geologica l Con gress, Wash ington, Extended Abstracts, v. 3, p. 61–62.
