british geological survey - 1995 - j naden and p j henny - character is at ion of gold from fiji

8/3/2019 British Geological Survey - 1995 - J Naden and P J Henny - Character is at Ion of Gold From Fiji

1/58


2/58

British Geological Survey

TECHNICAL REPORT WC/95/41Overseas Geology Series

CHARACTERISATION OF GOLDFROM FIJI

J Naden and P J Henney

A Report prepa red for the Overseas Development Administration under theODABGS Technology Development and R esearch Programme, Project92/1

OD A ClussificutionSubse ctor: GeoscienceTheme: G I Prom ote environmentaly sensitive developmentof non-renewablenatural resourcesProject Title: Alluvial Gold Characterisation in Exploration PlanningReference number: R5549

Bibliographic reference:J. Naden and P J Henney, 1995

Characteristation of Gold from FijiBGS Technical Report WC/95/41

Keywordr:Gold , alluvial, soil, bedrock, mineralogy, electron microprobe, Fiji

Front cover illustration:Panning for gold - Tuvatu prospect, Viti Levu

ONERC 1995

Keyw orth, Nottingham, British Geological Survey,1995


3/58

Table of Contents

EXECUTIVE SUMMARY

1. INTRODUCTION

2. GOLD MINERALISATION IN FIJI

2.1. Epithermal gold mineralisa tion

2.1.1. High-sulphidation systems

2.1.2. Alkali systems

2.2. Porphyry-Cu-Au mineralisat ion

2.3. Alluvial gold

3. METHODOLOGY

3.1. Sampling

3.2. Laboratory studies and sample preparation

3.2.1 Bedrock gold samples

3.2.2. Placer gold and ore concentrate samples

3.2.3. Electron probe micro-analysis

4. MINERALISED AREAS

4.1. Mount Kas i , Vanua Levu

4.1.1. Bedrock gold

4.1.2. Alluvial gold

4.2. Emperor Gold Mine , Viti Levu

4.2.1 Bedrock gold

4.2.2 Alluvial gold

4.3. Tuvatu prospect, Viti Levu

1

2

2

2

2

3

3

3

3

3

3

4

4

5

5

6

7

7

7

9

9


4/58

4.3.1. Bedrock gold


4.4. Waimanu Alluvial gold, Viti Levu

5 . DISCUSSION

6 . CONCLUSIONS

BIBLIOGRAPHY

FIGURES

TABLES

APPENDIX 1 LIST O F SAMPLES

9

10

10

12

13

15


5/58


6/58

List of Tables.

Table 1.

Table 2.

Table 3 .

Table 4.

Table 5.

Table 6.

Table 7.

Table 8.

Table 9.

Analytical conditions and detection limits for electron micro-

probe analyses.

Electron micro-probe analyse s of en argite, goldfieldite, pyriteand calaverite from the Mount Kasi gold deposit.

Electron micro-probe analyses of bedrock gold for Mount Kasi.

Averaged electron micro-probe analyses of pyrite from Emperor

Gold Mine.

Electron micro-probe analyses of gold and other mineral

analyses of a table concentrate from Emperor Gold Mine.

Electron micro-probe analyses of alluvial gold and heavy

minerals collected in th e Emperor mine drain age catchment.

Electron micro-probe analyses of gold and associated mineralsfrom th e Tuva tu prospect.

Electron micro-probe analyses of alluvial gold from the

Waimanu river.

Electron micro-probe analyses of inclusions hosted in alluvial

gold from the Waimanu River.


7/58

EXECUTrVE SUMMARY

This is a study of the variation in chemistry and inclusion mineralogy ofbedrock an d placer gold from Fiji. It forms pa rt of a large project,undertaking gold characterisation from a wide range of geologicalenvironme nts in Ecuador, Zimbabwe, Malaysia and Fiji. The work was

carried out under the Overseas Development AdministratiodBritish

Geological Survey Technology Development and Research programme

(Project R5549) as pa rt of the British Governments provision of technicalassistance t o developing countries. For the Fijian component of theproject, samples were collected from river gravels, primary ore, and tableconcentrates.

In tota l thirty-five samples from five localities were examined. Da ta

collected from these samples are represented by over 100 point analyses ofgold, identification of associated minerals, and microgeochemical maps of

gold-mineral intergrowths.

A framework for identifying possible sources of alluvial gold is given. This

was achieved by characterising bedrock gold mineralisat ion from a varietyof epith erma l and porphyry environments. The environments studie d

included alkali (the Emperor Gold Mine), high-sulphidation (the MountKasi prospect) and telescoped (the Tuvatu prospect) systems.

A study of placer gold from the Waimanu alluvial deposit, th e only alluvialgold deposit in Fiji, showed two distinct sources of gold: one, a low-silvertype, associated with Cu-Fe sulphides, can be related t o the nearby

Namosi porphyry copper deposit. The other source, a high-silver type withabu nda nt tellurides, indicates an alka li epithermal association, suggesting

a source similar i n style t o the mineralisation observed at Emperor GoldMine.

The identification of two bedrock sources for the Waimanu alluvials

clearly shows that there must be a , yet unknown, alkali epithermal(Emperor) source within the W aimanu catchment. This demonstrates thepower of alluvial gold characterisation a nd it s role in gold exploration.

An impor tant implication of these resu lts is that future exploration withinthe area should be focussed on locating this Emperor type source.


8/58

1. INTRODUCTION

This report documents the results of work undertaken on ODA/BGS TDRproject 92/1 R5549 entitled Alluvial Gold Characterisation i n Explorat ion

Planning. The project ha s three main aims. The first is characterisationof primary bedrock gold and associated placer gold from a wide range ofgeological and geographical settings. The second is t o establish whethe rvariations in the bedrock gold are inherited in placer gold. The third is

the investigation of how these variations can be used t o "fingerprint"

primary sources of gold from different geological environments.

The characterisation of bedrock and associated placer gold has two mainaspects: the use of scanning electron microscopy to study morphological

features (e.g., shape, size and growth patterns) and the application ofreflected light microscopy and electron micro-probe analysis t o identify ofheterogeneities, zoning and inclusions. The latte r includes detailed

microgeochemical mapping of gold grains on the sub-10pm scale.

This report documents and characterises some of the types of bedrock andalluvial gold minera lisation present in Fiji.

Fiji was chosen as an area suitable for study for the following reasons:

(i)

(ii)

(iii)

It has a well established operating gold mine (the Emperor Gold

Mine), the geology of which is well understood (Denholm, 1967;Forsythe, 1976; Ahmed, 1979; Ahmed e t al. , 1987; Anderson e t aZ.,1987; Anderson and Eato n, 1990; Kwak, 1990; Setterfield et al.,

1989; Setterfield, 1991).

Several distinct types of gold mineralisation with differentmineralogies are present on the islands (Colley, 1976; Colley and

Greenbaum, 1980). If the characterisation of alluvial gold allowsthe differentiation between these styles of mineralisation, then itcan be used as an exploration tool in delineating the type of

mineralisation responsible for placer gold in s tream sediments.

Possible source regions for alluvial gold and styles of mineralisation

for bedrock gold are limited in an island setting. Hence, th e

likelihood of mixed sources in alluvial samples are reduced

compared with very mixed populations in distal samples, tha t ar e

derived from large catchments with many sources (e.g., Zimbabwe).

1


9/58

2. GOLD MINERALISATION IN FIJI

The Fiji islands are located at a bend in the boundary between the Indo-

Australasian and Pacific plates, and occur midway between the west-

dipping Tonga-Kermadec subduction zone and the east-dipping Vanua tusubduction zone. The islands are separated from the subduction zones by

two discontinuous spre ading centres (t he Lau a nd North Fiji basins) .

Magmatism in Fiji was mainly related, t o a now inactive, southwest-dipping subduction zone, which was disrupted in the early Pliocene or late

Miocene. The oldest rocks on Fiji ar e island-arc volcanic and assoc iated

sedimentary rocks, which ar e late Eocene to early Oligocene in age. These

rocks were folded and then intruded by the 12-7.5 Ma Col0 plutonic suite.

Intru sive activity coincided with extensive sedimentation an d minor calc-

alkaline volcanism, which continued after intrusive activity ceased.

Subsequently, several shoshonitic volcanoes were active in northern Viti

Levu, the larger of th e tw o main islands. Gold mineralisation on Fiji iscommonly associated with this period of magmatism (Setterfield et al.,

1991).

Several distinct types of gold mineralisation a re present on Fiji. Thesecomprise tw o main styles of bedrock mineralisation, epithermal and

porphyry-copper-gold, with subsidiary concentrations of alluvial gold. A tleast tw o classes of epithermal mineralisation are present and includehigh-sulphidation and alka li systems.

2.1. Epithermal gold mineralisation

2.1.1. High-sulphidation systems

Epithermal gold mineralisation that is classified as being High-sulphidation in style is characterised by intense acid-sulphate alteration

and is commonly located close to the margin of volcanic craters.

2.1.2. Alkali systems

Alkali epithermal mineralisation has a close association with alkalinemagmatism (Bonham, 1988; Richards, 1993) and a distinctive mineralogy,

commonly dominated by a suite of telluride minerals. The Emperor GoldMine, on Fiji, is considered t o be a type example of this style ofmineralisation (Richards, 1993)

2


10/58

2.2. Porphyry-Cu-Au mineralisation

Porphyry gold mineralisation is not widespread, but a new mine exploiting

the Namosi porphyry copper deposit was recently opened and contains

minor amounts of gold (I lp pm ). This locality is of parti cular inte res t as itlies in t he drainage catchment of the Waimanu alluvial gold deposit.

2.3. Alluvial gold

Fiji is not noted for deposits of alluvial gold. The Waimanu gravels are

th e only known alluvial deposit. Reserves of gold ar e est ima ted at1.53 million m3 . The average grade is O.03ppm with a maximum value of

10.2ppm (Colley, 1976). Elsewhere, gold is generally very fine grained,

but locally ab unda nt (up to lOOppm in pan concentrates).

3. METHODOLOGY

3.1. Sampling

Four ar ea s of gold mineralisation were selected for sampling: Mount Kasi,

on Vanua Levu, plus Vatukoula (Emperor mine), Tuvatu-Mount Kingston,

and Waimanu on Viti Levu (Figure 1 ). In each are a a sui te of samples was

collected comprising bedrock ore an d it s weathered products (e.g., oxidised

ore an d residua l soil) an d alluvium. Where possible, th e unconsolidated

samples were sieved (c2mm) and panned on-site t o reduce th e amount ofmate rial. However, fur the r processing was left until the samples were

returned t o the laboratory. Thirty-five samples were collected for analysis(see Appendix 1 or details).

3.2. Laboratory studies and sample preparation

3.2.1 Bedrock gold samples

Obtaining samples with visible gold proved extremely difficult. At Mount

Kasi, samples were collected from localities at outcrop and drill-coreknown to have high gold values (>30ppm). At Vatukoula (Emperor mine),

a large (10kg) sample of crushed an d hand picked ore was collected and atTuvatu-Mount Kingston samples containing visible gold were provided by

Mr. B. Brookes of Geo Pacific Ltd. No bedrock gold was collected that wasdirectly associated with the Waimanu alluvial gold.

3


11/58

Polished thin sections of the high-grade samples from Mount Kasi

contained microscopic gold (e40pm) associated with en argite and ba rite.

Processing of the ore sample from Emperor mine (in the manne r describedbelow) revealed abund ant very fine gold (d00 pm ).

3.2.2. Placer gold and ore concentrate samples

Placer gold samples were wet sieved into +1000, 1000-500, 500-150, 150-

63 and -63 micron fractions. These fractions were then superpanned and

the concentrate from each fraction was the n magnetically sepa rat ed using

a Frantz magnetic separator into a further five fractions (magnet angle:

10"; magne t cur rent settings: 1.25A). Gold was concentrated in the non-magnetic fraction. Individual

gold grains were then hand picked and mounted in resin for subsequentpolishing and electron probe micro-analysis.

3.2.3. Electron probe micro-analysis

Polished thin sections of samples were examined using transmitted andreflected light microscopy t o identify ore and gangue minerals associated

with gold. Samples were then analysed using a Cameca SX50 electron

micro-probe. Four approaches were used t o provide data f o r

characte rising bedrock gold mineralisation:

quantitative analyses of gold; these were used t o assess thevaria tion and ran ge in gold composition.

Microgeochemical maps ; the se were used t o reveal complexintergrowths between gold and opaque minerals plus chemical

zoning in gold.

Associated minerals (e.g., tellurides) were identified using

quantitative analyses of inclusions in gold.

Quantitative analysis of the more common sulphides (e.g., pyriteand galena) was used to ascertain if they contained high a moun ts

of trace elements (e.g., Te, Sb, Ag, Au) that could prove diagnostic

for finger p rinting purposes.

Elements analysed included As, Se, S, Pd, Ag, Sb, Te, Fe, Cu, Au, Hg, Biand Pb. Although in some analyses values are quoted for a number of

trace elements these are very close t o the theoretical detection limits(Table 1) an d could be unreliable. Of the trace elements analysed, only

4


12/58

the concentrations for C u and Hg ar e thought t o be reliable as these arethe only elements tha t were consistently detected. However, even these

are generally below 1 weight percent.

Data are presented as series of tables of analyses for all mineral phases.For gold, 'SI-curves of silver content are used show the range anddistribution of data, plus a bivariate plot shows how silver concentration

varies with the trace element (Cu and Hg) content.

Microgeochemical maps were acquired by ma king point analyses on a 256

by 256 grid. The distance between points is dependent on the size of areabeing analysed but is generally of the order of 1-2pm. Mapping of goldgrains was underta ken using an acce lerating voltage of 20kV and a beam

current of 50nA. The dat a a re collected as uncorrected X-ray counts, butas the matrix is essentially the same in an individual map this causes

little error. Data ar e presented as a series of colour maps, each showing

the spatial distribution of a particular element, with red indicating thehighest concentration and mag enta the lowest.

4. MI N ER A LI SED AREAS

4.1. M o u n t K a s i, Vanua L e v u

The Mount Kasi deposit is located a t the western end of Savu Savu Bay on

the southern coast of Vanua Levu; it operated as a gold mine between1932 and 1946 producing a limited tonnage of ore (261,000t) at an averagegrade of 8g/t Au. Mount Kasi has been actively explored by a number ofmining companies since 1983, and is now about t o be exploited pending asuccessful application for a mining lease.

The epithermal mineral deposit at Mount Kasi is classified as a high-

sulph idati on gold-enargite system. It is located on the marg in of avolcanic caldera, and is Miocene to Pliocene in age. Mineralisation is

hosted in a strongly silicified structure t ha t strikes approximately north-

west and dips at 70 " t o the south; the country rocks comprise a suite of

calc-alkaline low-silica andesite flows, basalt dykes, and associated

volcaniclastic rocks (Natewa Volcanic Group). These are intruded by

hornblende-phyric andesite plugs (Nararo Volcanic Group) of Pliocene age(Turner, 1986) (Figure 1).

5


13/58


14/58


15/58

Emperor Gold Mine occurs on the western margin of a large caldera(Tavua Caldera). The structural regime of the area is dominated by a

complex pattern of N-S steep faults and intersect ing low-angle fracturescalled flatmak es. The location of mineralisation is controlled by these

structures and consists of thin, single o r multiphase, veins in st eep o r flatshears. The fracture infilling is typically multist age. Lode wid ths are

typically less than one metre, and throughout the life of the mine the

different types of mineralised structure have contributed equally t o goldproduction. Flatmakes are areally extensive and are the largest

mineralised structures in the mine, with strike lengths of up t o 2km anddip exten ts of lk m. A wide range of mineral s ar e recorded from the mine;

these comprise a suite of tellurides (sylvanite [AuAgTe,] , petzite

[Ag3AuTe,], krenner ite [(Au,Ag)Te,] h ess ite [Ag,Te], cal averi te [AuTe,]),auriferous pyrite and free gold; associated gangue minerals include

q u a r t z , a d u l a r i a , c a r b o n a t e , s e r i c i t e l i l l i t e a n d r o s c o e l i t e

[K(V,A1,Mg),(A1Si3)0,,(OH),]. oscoelite is commonly associated with

high-grade telluride mineralisation and is considered an important

indicator of high-g rade mineralisation. Accessory mineral s include

arsenopyri te , marcasi te , sphaleri te , nat ive tel lurium, tennanti te-

tetr ahe drit e, chalcopyrite and galena. Electron micro-probe analyses ofpyrite commonly showed it to be zoned and locally contain up to 8 wt%arsenic (Figure 4 and Table 4) . Point analyses across a zoned grainshowed that elevated levels of gold are associated with arsenic and

antimony-rich zones.

Within the mine, gold-silver ratios show a variation from c1 t o about 4with gold grades ranging between 25 and lOOOppm (Kwak, 1990;

Forsythe, 1967). In the polished thin sections examined, no visible or

microscopic gold was observed. However, in th e sample of ore concent rate ,

abu nda nt fine grained (c100pm) gold was observed. Here grains varied incomposition between 19.16 and 4.38 wt% Ag. The gold grains retri eved

from the table concentrate commonly comprised grains intergrown with a

variety of minerals such as pyrite, various gold-silver tellurides andcoloradoite [Hg,Te]. The vari atio ns in composition of the gold from thetable concen trate contrasts w ith th e average bullion fineness of between

700 and 800, which is equivalent t o approximately 20-30 wt% silver

(Setterfield e t al . , 1988). If analyses of gold (4 to 20 wt% Ag) from th etable concentrate are representative of the compositional variation of goldin th e mine, the n th e concentrated gold ore mus t eit her be contaminated

8


16/58

by silver-bearing minerals, or significant amounts of gold are present

combined with Au-Ag-tellurides. This shows th a t using production

records for estimating the composition of bedrock gold should be treated

with caution, as these may be a t variance with th e true compositional

ran ge of nat ive gold.

4.2.2 Alluvial gold

Alluvial samples were collected from four sites surrounding the mine.

Placer gold, suitable for analysis, was recovered from one sample (F J 18).The composition of alluvial gold from around the Emperor Gold Mine

varies between 15.38 and 6.15 wt% Ag, this is within the range and

matches the distribution found in bedrock gold (Figure 5). A number of

telluride minerals were identified in th e heavy mineral concentrates an dinclude Au-Ag tellurides, hessite [Ag,Te] , coloradoite and chalcopyrite(Table 1). All of these ar e recorded from th e mine.

The above data not only show that gold composition is strongly correlated

between bedrock a nd alluvial sources, but also the accessory mineralogy of

th e two environments is almost identical. Importantly, thi s dual

correlation shows th at , in Fiji, characterisation of alluvial gold accura tely

reflects compositional and mineralogical varia tion i n the primary source.

4.3. Tuvatu prospect, Viti Levu

4.3.1. Bedrock gold

The Tuvatu-Kingston ar ea is located on the western side of Viti Levu some

30km t o the north-west of Nadi. Mineralisation is hosted in the centreand on the margins of a caldera lying on the Emperor-Mount Kasi t ren d

(Fiji gold trend). The intra-ca ldera un its a re shoshonitic in composition

and comprise trachyandesitic volcanic rocks intruded by a monzoniticstock with associated late stage dyke swarms of basaltic and andesiticcomposition (Colley, 1976).

At Kingston, a telescoped multi-phase system of mineralisation occurs in

association with a volcanic caldera. It comprises a central Cu-Auporphyry hosted by monzonitic intrusive. High-level (adularia-sericite)

epithermal gold mineralisation is developed of the nor th-eastern margin of

th e caldera. The porphyry system ha s best grades of 12m at 1.29g/t Au.

The fairly extensive (1400 x 400 x 260m) epithermal vein system (Bana na

9


17/58

Creek) is zoned both vertically and horizontally. The lowest part of the

system is characterised by low-sulphide veins with a Au-Pb-Zn

association, while the higher levels are characterised by a low-

quartzhigh-sulphide system with a Au-As-Hg-Te association (Meares,

1990).

Samples were provided from the Tuvatu prospect by M r. Bill Brookes ofGeo Pacific Ltd., and consisted of thin microcrystalline quartz veins

cont ainin g sulph ides an d visible gold. Minerals identified included

arsenopyrite, chalcopyrite, pyrite, enargite [Cu,AsS,] , sphalerite, nativegold and r are altait e [PbTel (Table 7). Gold compositions are silver-poor

and contain between 0.15 and 6.84 wt % silver.


It was not possible t o recover placer gold suitable for analysis from thesamples collected within the Tuvatu prospect drainage catchment.

However, this does not preclude applying the information gained from the

bedrock gold t o othe r area s. In th e case of th e Tuva tu prospect, atelescoped porphyry Cu-Au system, associated alluvial gold would

comprise low-silver gold (1-7 wt% Ag) with an inclusion assemblage

dominated by sulphides rath er th an tellurides an d sulphosalts (e.g. pyrite,

chalcopyrite, spha lerite and arsenopyrite).

4.4. Waimanu Alluvial gold, Viti Levu

Alluvial gold was discovered in the Waimanu River about 20km west ofSuva in the early par t of this century. It occurs in river gravels unde r

river flats an d terr aces , overlying altered volcanic an d sed imen tary rocks

that have been intruded t o the south by a la rge gabbroic body. Gold

occurs in free flakes usually distributed in sinuous stringers in the

gravels. The exact bedrock source for the gold is not known, but

topographically above the gravels, approximately 15km to the north east,is th e recen tly discovered Namosi porphyry-Cu deposit, which contains a

smal l amoun t of gold (generally elppm) (Colley, 1976).

A sample of alluvial gold was provided by the Mineral ResourcesDepartment, Fiji.

The placer gold grains exhibit a variety of textures. Most grains exhibit

rims that have high gold contents compared to th e cores of grain s. These

10


18/58

vary from patchy and incomplete t o complete mantles. The thickness ofthe rims is about 5-20p.m; commonly the r ims ar e solid but in some cases

the y contain ab und ant pores. Also, a number of grains are internally

heterogeneous and contain patches and tracks of high-silver gold that is

locally replaced by low-silver gold (Figure 6). Within-grain variation insilver content (excluding silver-poor rims) is between 11.30 and 23.79 wt%

(Table 8). Overall, the composition of gold from the whole suite of veinsvaries from below detection t o 23.79 wt% silver. Modified S-curves

(Figure 7 ) show two distinc t populations: low-silver gold (0 to 15 wt %) an dhigh-silver gold (15 t o 25 wt%). However, detailed examination showsseveral sub-groups which each exhibit fine-scale variation in silver

conte nt. This fe ature of alluvial gold has been observed in the Raub

region of Malaysia (Henney et a l . 1995), where th e fine-scale variationobserved in t he alluvial gold could be directly linked t o variation within asingle bedrock source.

Othe r elements analysed were Se, Cu, Pd a nd Hg. Se a nd P d were below

detection limits , except for one analysis. Cu concentrations associated with

low-silver are up to 1.6 wt%. Hg concentrations ar e generally less th a n

0.12 wt%. A bivariate plot of Cu and Hg contents versus silver

concentration (Figure 8 ) shows tha t there is a weak correlation between

Cu and Ag concentrations; elevated Cu concentrations ar e associated withlow-Ag contents in gold.

Inclusions in the alluvial gold were observed in approximately 25% of

grains, and show a narrow range of minerals mainly comprising gold-

silver tellurides (Table 9). Hessite [Ag,Te] was the only mineral positively

identified, as analyses invariably contain a n unknown amount of gold and

silver from th e host gold gra in. The presence of gold from th e host grain

makes it difficult t o correctly assess t he stoichiometry of the gold-silvertelluride. Oth er miner als identified include pyrite and a silver-copper

sulphide, but these ar e rare . There is a strong relationship between th ecomposition of the alluvial gold and the type of inclusion. High-silver

values in the host mineral grain are associated with tellurium minerals

and silver sulphides, whilst pyrite is associated with low-silver gold

(Figure 7).

11


19/58


20/58

(iii) Telescoped porphyry system

Low silver content , simple mineralogy dominated by sulphides r athe rthan sulphosalts o r tellurides (Tuvatu).

(iv) Secondary goldVery low silver content, inclusions of supergene and hypogeneminerals (e.g. copper sulphides, enargite) (Mount Kasi).

The Waimanu alluvials, for which there is no known source, show two

distinct populations of gold: high-silver gold with common telluride

inclusions and rar er sulphosalt inclusions, and low-silver gold w ith rar e

pyrite inclusions. This suggests th a t the Waimanu alluvial gold ha s two

sources: derivation from an alkali epithermal system (medium to high-

silver gold); derivation from a porphyry system (low-silver gold).

6. CONCLUSIONS

1. A t Emperor mine, where data were obtained on paired bedrock andalluv ial gold samples, a n excellent correlation betw een gold

composition and accessory minerals was observed. These data, in

conjunction with studies by Styles (1995) and Henney (1994, 1995a

1995), show conclusively that many of the compositional and

mineralogical features observed in bedrock gold are preserved whengold is mobilised from primary ore t o alluvial deposit.

2. In Fiji, an important aspect in establishing the relationship between

alluvial and bedrock gold was the analysis of concentrated ore from

Emperor mine. This type of sample provides far more representative

compositional variation and mineralogical associations than single

point sampling (han d specimens and th in sections), o r bulk assay dat aobtained du ring mining operations. This observation reinforces those

made elsewhere during the TDR project (e.g., Styles, 1995; and

Henney, 1994; 1995a; 1995). I t shows tha t, wherever possible, analysis

of concentrated ore is the preferred method of charac terising bedrockgold, as single point sampling may miss key characteristics of the

bedrock gold.

3. In a regional study, it may not always be possible t o collect pairedalluvial an d bedrock gold samp les. However, knowing th a t key

identifiable features in bedrock gold are preserved during mobilisationfrom primary ore t o alluvial deposit allows unpaired bedrock and

13


21/58

alluvial samples to be used to characterise variation at a regional scale.

In the case of the Fiji studies, data obtained from bedrock gold atMount Kasi and Tuvatu were used t o characterise the types of goldfrom high-sulphidation epithermal mineralisation and telescoped

porphyry system s. These da ta, coupled wit h information fromEmperor Gold Mine area, were integra ted with analyses of gold from

the Waimanu alluvial gold deposit and enabled identification of at least

two major sources for the Waimanu alluvials. One can be related t o

the nearby Namosi porphyry copper deposit, and the other has anumber of characteristics in common with alka li epithermal (Emperor)

style mineralisation; a source not yet discouered.

4. Th is study shows tha t the re is a probable alka li epithermal source of

placer gold in the Waimanu alluvial deposit. As Emperor Gold Mine(one of the type localities for alkali epithermal mineralisation) is aworld-class ore deposit, exploration efforts should be directed atlocating th is source. Exploration stra teg ies for locating it should befocussed on identifying key indicators in heavy mineral concentrates

collected from first and second order streams within the Waimanu

River catc hmen t. Also, as alluvial gold, in Fiji, is commonly fine

grained, it is impo rtant th at sufficient heavy mine ral conce ntrate for

ana lys is (5-10kg of wet concentrate ) is collected at each locality. Goodindicators for alkali epithermal mineralisation would be moderate- to

high-silver placer gold with an inclusion assemblage dominated by

tellur ides. Heavy mine ral concentrates would also be expected t ocontain significant amounts of tellurides. It is important to stress thatthese indicators cannot be identified by gold and base metal assaying of

pan concentrates alone, the individual grains must be examined

5. Fu tu re exploration for bedrock gold in Fiji, where minor alluv ial

occurrences have been identified, should include characterisation of

placer gold, as this will be able t o identify the type of goldmineralisation tha t is sourcing th e alluvial gold.

6. The dat a presented in this study are a n excellent demonstration of theapplication of alluvial gold chara cteri sation t o gold exploration. Theyalso provide very importan t background information for future stud ies

of alluvial and bedrock gold in Fiji and for studies of gold

mineralisation i n epithermal environments.

14


22/58

BIBLIOGRAPHY

Ahmad, M. (1979) Fluid-inclusion an d Geochemical Stu die s at the

Emperor Gold Mine, Fiji. PhD. Thesis University of Tasmania, 21Op.

Ahma d, M., Solomon, M. an d Walshe, J.L . (1987) Mineralogical an d

geochemical studies of the Emperor gold-telluride deposit, Fiji. Economic

Geology, 82, p.345-370.

Anderson, W.B. an d Eato n, P.C. (1990) Gold mine ralisa tion at theEmperor Mine, Vatakoula, Fiji. In: Hedenquist, N.C., White, N.C. and

Siddeley, G. (Editors). Epithermal gold mineralisation of the Circum-

Pacific; Geology, Geochemistry, Origin and Exploration, I I . Journal of

Geochemical Exploration, 36 , p.267-266.

Anderson, W.B., Antonio, M. Davis, B., Jones, F.P., Setterfield, T.N. andTua, P. (1987) The Emperor epithermal gold deposit, Vatukoula , Fiji. In:Proceedings Pacific Rim Congress 87. An International Congress on the

Geology, Structure, Mineralisation and Economics of the Pacific Rim. The

Australasian Institute of Mining and Metallurgy, Victoria Australia p.9-

13.

Colley, H.. an d Greenbaum , D. (1980) Mineral Deposits an dmetallogenesis of th e Fiji platform. Econ Geol. , 88, p.807-829.

Denholm, L.S. (1967) Lode structures and ore shoots at Vatukoula, Fiji.Proceedings of the Australa sian Inst itute of Mini ng and Metallu rgy, 222 ,p.73-83.

Forsythe, D.S. (1976) V ertical zoning of gold-silver tellurides in the

Emperor Gold Mine, Fiji. Proceedings of the Australasian Institute of

Mining and Metallurgy, 240, p.25-31.

Henney, P.J., Styles, M.T., Bland, D.J. and Wetton P.D. (1994)Characterisation of gold from the Lubuk Mandi area, Terengganu,Malaysia. Briti sh Geological Survey Technical Report WCl9412 1.

Henney, P.J., Styles, M.T., Bland, D.J. and Wetton P.D. (1995a)

Characterisation of gold from the Raub area, Pahang, Malaysia. British

Geological Surve y Technical Report WC/94/20.

15


23/58

Henney, P.J . , Styles, M.T., Bland, D.J. and Wetton P.D. (199513)Charact erisation of gold from th e Penjom are a, near Kuala Lipis, Pahan g,

Malaysia. Bri tis h Geological Survey Technical Report WCl95121.

Kwak, T.A.P. (1990) Geochemical an d tem pe rat ur e controls on oreminera l i sa t ion at th e Emper or Gold Mine, Vatuk oula, Fiji. In :

Hedenquist, N.C., White, N.C. an d Siddeley, G. (Editors). Epithermal gold

mineral isation of the Circum-Pacific; Geology, Geochemistry, Origin and

Exploration, I I . Journal of Geochemical Exploration, 36 p.267-338.

Meares, R.M.D. (1990) SP L 1218 - Kingston (FGi). Annual Report for the

Period to 30 June 1990. Open File Report 1218-9. Minera l Resources

Department, Fiji.

Richards, J.P. and Ledlie, I. (1993) Alkalic intrusiv es associated with th e

mou nt Kare deposit, Pa pu a New Guinea: Comparison wit h th e Porgera

intrusive complex. Econ. Geol. , 88, p.755-781.

Richards, J. P. (1992) Magmatic-epithermal trans itions i n alkalic systems;

Porgera gold, deposit, Pap ua New Guinea. Geology, 20, p.547-550.

Richards, J.P., McCulloch, M.T., Chappell, B.W. and Kerrich, R. (1991)

Sources of metals in the Porgera gold deposit, Papua New Guinea:Evidence from alte ration, isotope, and noble meta l chemistry. Geochemica

et Cosmochimica Acta , 55, p.565-580.

Rodda, P. an d Kroenke, L.W. (1984) Fiji: a fragmented arc. I n : Kroenke,L.W. (Editor) Cenozoic Development of t h e Southwes t Pac if ic .

CCOPISOPAC Technical Bulletin No. 6, p.87-109.

Setterfield , T.N. (1991) Evolution of the Tavua Caldera and Associated

Hydrothermal Systems , FGi. PhD. Thesis, University of Cambridge, 23 l p .

Setterfield , T.N., Anderson, W.B., Jon es, G.F.P., Antonio, M and Tua, P.(1989) Geology of th e Emperor MinelTavua Caldera Area. I n : Jones,

D.G. (Editor) South Pacific Gold Deposits. Bicentennial Gold '88

Excursion Guide No. 10, p.7-13.

Setterfield, T.N., Eaton, P.C., Rose, W.J. and Spar ks , R.S.J. (1991) The

Tavua Caldera, Fiji: a complex shoshonitic caldera formed by concu rrent

faulting and downsagging. Journal Geological Society, London , 148 ,p. 115-127.

16


24/58


25/58

Figure la. Geological map of Fiji (after Colley and Greenbaum, 1980).

179.E 100. lis. WSUCCESSION FOR VA N U A L E V U e TAVEUNI

D

9

117' f I7 WL

-.0 7 .

.- #0

N O RT Y E R N L A U G RW P '

W I T- O R O G C N K ROCKS1

0

100. 179.1I

SUCCESSON FOR VlTl LEV U , YASAWAS, LOMAIVITI8 XANOMJ

P O S T - O W N I C S UO SU ON IT IC 19 0 TnOLCllTlC RUCKS

I O4 - KOROIMAVUA S UI T E I

4)sl-oRo6cNK CALCALKALINL ROCKS I N A U O Y A ~ S I T I ,VAT14 . N O C I l C . I ( A N 0 AV U I

P OS 1- 010GCN l C S COI ME NT S I NA NOI . NAVOS A.V C I ATA . R A .T Y U V U 5 C D M E N l A R 7GROUPS. d M C N 0 I ) A U W T W W 0 ) C l


26/58

1 ,

v)aa

L"-I

a

sz 0 B

0

w

t!\

0

' D

Q

h0


27/58


28/58


29/58


30/58


31/58


32/58


33/58


34/58


35/58

Table 1 . Analytical conditions and detection limits for electron micro-

probe analyses. Operating condition of the Cameca SX50

were 30kV high voltage and a beam cu rren t of 20nA.


36/58

c- I L L

wI I I I I I


37/58


38/58

Table 4.

NicuZ nHg

Averaged electron micro-probe analyses of pyrite from

Emperor Gold Mine. Elevated gold levels measured on the

gold M, confirm other workers observations (e.g., Ahmad etal . , 1987) that significant sub-microscopic gold occurs inpyrite.

0.06 0.00 0.19 0.05 0.030.11 I 0.04 0.21 0.08 0.030.05 0.04 0.06 0.01 0.03

Bi ~ 0.17 0.16

I33 0.35-- 0.33 -Total 99.20

0.18 0.01 0.14

0.36 0.02 - 0.281.22


39/58


40/58

O W b dhllw101 c

I ,

I9 : I

3 h"'II I

C

b V+:!I

' I '

I -%i :, ' I 'I

gI , I

II~

~i I

i 1 I

I; i1


41/58

I ! '

I

t


42/58


43/58


44/58

0 : I

~

q 'Tii

I . - t -

I ,

I-- tr

- 1 - t iI1 I

O r

m v :~:I I - A -

C

[T i - i-I ,+ - - 4 - tt

o c

$0 0- t - - i

I /

1


45/58

I ,

I ,

I I

2C

I

I ,

I I I I

I ,1p0 I1 I I 1 I

20bn

U3


46/58

I

I , I

i~ '0

I

l i

-+

+++

II / I

5 C

I

I II1 I I1 I I

@a

,5

I


47/58

mm4 05 0 0 4 0

l l l l 1 l I I 1

,' i

-t- 1 -t

--tt ! --t


48/58


49/58


50/58

Table 8 . Continued.

grain No.39404142

~~~43

444546

~- _ _

inc lus ion s Ag-rich tracks Seabsent absentabsent - presentabsent absentabsent absentabsent absent -absent __ absentabsentAg, Au tellu rides present

absent -:d

0.28 103.24 103.529.650.67

I0.11 1102.98l 0.09

I03.85

~~ 9.84 0.03 94.02 0.09- 103.984.15 1.63 98.01 012 103.91

85.69 104.02


51/58


52/58

t

cro

- r -n 1 0

709

10N

0 0

O - :

I3

c

o r -

3 c? ?

o v09c- c c

t

c

7-z4m v ,4 4

I


53/58


54/58


55/58

APPENDIX 1

LIST O F SAMPLES


56/58

Sample

number

Locality Sample type Sample descr iption

F J 01 Mount

Kasi

Pan

concentrate

Alluviudc olluv ium (assa ys of colluvium shol

up to 30gh gold).

Alluviudc olluv ium (assa ys of colluvium shol

up t o 30gh gold).

High-grade barite in core of breccia collectec

from zone at south east ern e nd of Mount Kas

oDen cut.

FJ 02 MountKasi

Mount

Kasi

Pan

concentrate

RockJ 03

FJ 04 Mount

Kasi

Drill core DC4 box 25, high-grade enargite ore fron

Done Creek prospect, adjacent to Mount Kas

Mine.

F J 05 Mount

Kasi

Drill core R5 47m. High-grade breccia ore.

F J 06 Mount

Kasi

Rock High-grade breccia material collected fron

zone at sou th eastern end of Mount Kasi ope1cut.

High-grade breccia material collected fron

zone at sou th east er n end of Mount Kasi ope1

cut.

FJ 06 Mount

Kasi

Rock

FJ 08 MountKasi

Panconcentrate

Heavily weathered breccia from the Nandun,Creek prospect. Gold content not known.

Done Creek immediately below M t Kasi mineRare, small (


57/58

Samplenumber

FJ 13

Locality Sample type Sample description

Vatukoula Rock Samples from the 1800' t o 1950' levels of thcMatanag ata workings of Emperor Gold Mine.

Samples from the 1800' to 1950' levels of theMatanag ata workings of EmDeror Gold Mine.

RockJ 14 Vatukoula

FJ 16 Vatukoula Rock Samples from the 1800' to 1950' levels of the

Matan agata workings of Emperor Gold Mine.

1 km below mine collected from above

confluence where a small creek which drain

tailings pond mee ts the Nasivi River. Rare

FJ 18 Vatukoula Pan

concentrate

flakes of visible gold, plenty of magnetite.

lkm below mine collected from old alluviaJ 19 Vatukoula Pan

concentrate ter rac e. No obvious visible gold, plen ty o

magnetite.

1 km below mine collected from belov

confluence where a small creek which drain

the tailings pond meets the Nasivi River

Rare flakes of visible gold, plenty o

FJ 20 Vatukoula Pan

concentrate

magnetite.

FJ 21 Vatukoula Panconcentrate

Millconcentrate

Pan

concentrate

From creek which drains tailings pond lknbelow main mine

Mill concentrate from Emperor Gold Mine.J 22 Vatukoula

F J 23 Vuda Collected from unconsolidated river gravels i1the Vunda River lkm below main Vund:

prospect. 2-4 colours per pan , wi th finc

grained flour gold.

F J 24 Tuvatu Drill core DDH 8 62.7m. Thin veins ( l cm ) ochalcedonic qu ar tz in altered monzonite. Gel

Pacific's Tuvatu prospect

DDH 8 62.4m. Thin veins ( l c r n ) o

chalcedonic quartz in altered monzonite. Gel

Pacific's Tuvatu DrosDect

F J 25 Tuvatu Drill core


58/58

- ~

Sample Locality Sample type Sample descrip tionnumberP J 26 Tuvat u Drill core DDH 8 67.2m. Thin ve ins ( lc rn) o

chalcedonic quar tz i n alte red monzonite. Gec

Pacifics Tuvatu prospect

DDH 8 64.5m. Thin ve ins ( lc rn) o

chalcedonic qu ar tz in al te red monzoni te. Gec

Pacifics Tuvatu prospect

DDH 8 60.2m. Thin ve ins ( lc rn) 0

chalcedonic qu ar tz in alt ered monzonite. Gec

Pacifics Tuva tu prospect

DDH 9 66.0m. Th in veins (l crn) of chalcedonic

qu art z in alte red monzonite. Geo Pacific:Tuvatu DrosDect

J 27 Tuvatu Drill core

PJ 29 Tuvatu Drill core

P J 30 Tuv atu Drill core

?J 31 Tuvatu Drill core DDH9 68.7m. Th in veins (l crn ) of chalcedonii

quartz in al te re d monzonite. Geo Pacific;

Tuvatu prospect

Pan concentrate from western creek; mucl

magnetite with rar e grains of gold

Pan concentrate from Tuvatu creek; mucl

magnetite and rar e grains of gold.Collected upst rea m from the confluence of i

smal l c reek wi t h Vunda River. Thc

confluence is lOOm downstream from localit:

FJ23.

P J 32 Tuvatu Pan

concentrate

?J 33 Tuvatu Pan

concentrate?J 34 Vuda Pan

concentrate

~

?J 35 Waimanu Alluvial gold Sa mp le of al lu vi al gold (-30-40 grains

d o n a t e d by t h e M i n e r a l R e s o u r c e ,

Department, Fiji.

british geological survey - 1995 - j naden and p j henny - character is at ion of gold from fiji

Documents