GEOCHEMISTRY AND GEOLOGY

I L!GEQ_lOGy C>EPT ~lU\Iers•ty f · - o Tasma . ma

MINERAqZED AND BARREN KOMATI ITES- WESTERN AUSTRALIA

BY

ROBERT D. McNEIL

A thesis submitted to the Department of Geology in fulfilment of the requirements for the degree of Master of Science.

University of Tasmania Hobart, Tasmania July, 1980

. I

This thesis contains no material which has been accepted for the award of any other degree or diploma at any university, and that to the best of my knowledge the thesis contains no copy or paraphrase of material previously published or written by another person except when due reference is made in the text of the thesis or as acknowledged in the Acknowledgements.

Robert D. McNeil

July 1, 1980

( i )

ABSTRACT 7

Western Australian Archaean komatiites which are associated with nickel

sulphide mineralization can be separated into two groups - Mineralized or

Barren, based on komatiite lithogeochemistry. Mineralized komatiites may

host nickel sulphide deposits whereas Barren komatiites do not. Chemical

relationships were determined from a data base of approximately 3300 samples

of fresh komatiite ult.r.amafic from four nickel pro¥inces and other greenstone

belts not known to contain nickel sulphides. Mean chemical values for each

group of komatiites were:

Category

Mineralized

(J) NiP

- (2) 1027

Ni CuP Cu Al Ca

2220 36 42 1.6 2.2

!i9_ Zn Cr Mn Fe CoP

19.2 69 1617 1057 6.1 49

Co

119

Barren 429 1530 29 39 2.2 2.8 16.4 76 2260 1128 7.0 32 119

Discriminant analysis, using the above thirteen chemical determinations

as variables, for each of 2775 samples from forty localities, indicated that

samples could be classified as either Mineralized or Barren with an expected

accuracy of greater than 80 percent. No single element or chemical determin-

ation is definitive, but collectively, Cr, Ni, Zn, Cu, Ni~, Mg, Fe and Co can

distinguish between the two groups of ultramafics. Critical elements are Cr,

Ni and Nif, assuming that values of Zn, Cu, Mg and Fe approximate the mean

value for all West Australian komatiites. The Ni to Cr ratio is always greater

than unity (1) in Mineralized komatiites and the Ni to NiP ratio is always

less than 3.5. Sulphur is not a diagnostic element.

NOTES:

1. f indicates a partial or sulphide analysis.

2. Al, Ca, Mg and Fe results are expressed in percentages; all others in

parts per million.

( i i)

Increasing Ni/Cr ratios and decreasing Ni/Ni~ ratios within a komatiite

can be regarded as indicative of increasing nickel sulphide potential.

Mineralized komatiites contain less Cr within the silicate lattice structure

and less chromite than Barren komatiites. However, the more important

relationship appears to be the lesser amount of Cr attached to the silicate

mineral lattice.

Correlation analysis showed that:

1. most correlations are much stronger in Barren than in Mineralized

ultramafics;

2. the chalcophile elements, Cu, Ni, Co and Fe (constituents of nickel

sulphide deposits), show moderate to strong correlations with the rock

forming elements, Mg, Mn, Ca, Al in Barren ultramafics, but only weak

or no correlation in the Mineralized ultramafics;

3. copper has moderate positive correlation with Fe, Mn, Ca, Al and

negative correlation with Mg in Barren ultramafics but shows no

correlation with these same elements in Mineralized ultramafics.

These correlation differences suggest that in Barren komatiites Ni, Cu,

Co and Fe are contained in the silicate mineral lattice whereas in Mineralized

komatiites they are presently partly as a separate sulphide fraction. In

addition they may also suggest that these sulphides were added or removed from

Mineralized komatiites after the formation of the komatiite magma, probably by

concentration and removal in arl immissible sulphide-oxide melt.

Komatiites can be divided into two separate suites called volcanic and

intrusive. Volcanic suites such as those at Kambalda and Windarra South may

contain many individual komatiite flows. The basal section of a komatiite

volcanic pile consists of a small number of thick units which may contain

sulphide mineralization whereas the central and upper parts of the pile consists

of multiple thin units. Both thick and thin units consist of an olivine

cumulate derived lower part overlain by a silicate liquid derived upper part.

In thick units the olivine cumulate section is dominant whereas in thin units the

(iii)

silicate liquid section is dominant. Spinifex texture is characteristic of

unmetamorphosed sequences. u,.., ~~IV'lJ ::>

In metamorphosed sequences such as Windarra South it is not possible to

identify individual komatiites using mineralogical or textural criteria but

it can be accomplished using chemical data.

Intrusive suite komatiite sequences such as Forrestania or Perseverance

usually consist of a small number of high Mg, homogeneous peridotites and/or

dunites. Equigranular, equant olivine textures are characteristic. These

komatiites are often continuous over strike lengths of the order of tens of

kilometers and contain relatively little internal chemical variation.

Volcanic komatiites such as those at Windarra South and Kambalda are

considered to be ultramafic lavas. Chemical differences between volcanic

and intrusive sequences have been defined. Typical chemical values for the

cumulate section of a volcanic komatiite and for intrusive komatiites, both

with moderate to high mineralization coefficients are:

Classification NiP Ni CuP & Cu Al !:19. Zn Cr Mn Fe CoP Co - -Volcanic Komatiites 1000 2100 30 - 90 1-2 17-24 60 1300 1000 5.5 55 120

Intrusive Komatiites 1200 2500 5 - 60 0.5 20-26 60 1000 900 6 60 125

In general, if Nif or Ni are less than 500 and 1800 ppm respectively, or Cr

greater than 2100 ppm, a komati ite can be regarded as Barren.

It has been possible to define sections of greenstone belts as prospective

for nickel sulfides and other parts as unprospective. For example, the

Forrestania section of the Forrestania-Southern Cross greenstone belt has a

different chemical signature to the Southern Cross section. The latter section

is unlikely to contain economic nickel sulphide accumulations.

Ci v)

ACKNOWLEDGEt~ENTS

The writer acknowledges the support of the Tenneco Australia Inc. -

Minops Pty. Ltd. Joint Venture and Union Oil Development Corporation.

Numerous companies and individuals allowed the writer to collect samples,

most notably the Tenneco-Minops Jofnt Venture, Poseiden Ltd. and Amax

Exploration (Australia) Pty. Ltd.

Most of the geologic maps used for location purposes or for geological

descriptions of specific areas are based on field mapping by other geologists.

The names of many of these geologists are unknown but M. Lennox, J. Noakes

and M. Woodhouse, all formerly of Tenneco Australia Inc. deserve specific

acknowledgement. However, all solid geology interpretive maps and sections

presented here are the responsibility of and have been compiled by the

writer.

The writer especially wishes to thank the above three geologists for

their assistance and for many stimulating discussions. Thanks are also due

to E.A. Rugg formerly of Tenneco Australia Inc. and E.H. Lindsey of Union

Oil Co. of California for their support throughout much of the study.

W.R. Guthrie, also formerly of Tenneco Australia Inc., was responsible

for most of the analytical procedures and results. The description of

analytical procedures are based on personal communications and notes from

Guthrie.

P. Walker and N. Campbell deserve thanks for explaining the statistical

procedures to the writer. The description of Mulvar in Appendix E is based

on unpublished notes by J. Keays.

Finally I express my sincere thanks to Pamela Strauss for the final

typing of this manuscript.

(v)

CONTENTS

ABSTRACT

ACKNOWLEDGEMENTS

TABLE OF CONTENTS

LIST OF TABLES

LIST OF FIGURES

CHAPTER I - INTRODUCTION

1.1 Format of Thesis 1. 2 Data Base 1.3 Analytical Procedures

CHAPTER 2 - GEOLOGICAL ENVIRONMENT

2.1 Topography and Weathering 2.2 Greenstone Belts 2.3 Komatiites 2.4 Metamorphism and Alteration 2.5 Geological Subdivision of the Yilgarn Block 2.6 Nickel Provinces

CHAPTER 3 - MINERALIZED AND BARREN KOMATIITES

3.1 Mean Geochemical Results for each Locality 3.2 Discriminant Analysis

3.2. 1 First and second stage analyses 3.2.2 Third stage analyses

3.3 Principal Component Analysis 3.4 Correlation Analysis 3.5 Relative Importance of Each Variable

CHAPTER 4 - CHEMICAL CHARACTERISTICS OF VOLCANIC AND INTRUSIVE KOMATIITES

4.1 Volcanic Komatiite Suite 4 . 1. 1 K amb a 1 d a 4. 1.2 Windarra South 4. 1.3 Trough Wells 4. 1.4 Eureka Greenstone Belt 4. 1. 5 Red We 11

- 4. 1.6 Airport - Yilmia 4.2 Intrusive Komatiite Suite

4.2. 1 Queen Victoria Rocks 4.2.2 Forrestania 4.2 .3 Bullfinch 4.2.4 Mistake Creek

PAGE

(i)

(iv)

( v)

( i X)

(xi i i )

3 5 9

11

12 14 16 20 23 28

30

31 33 34 36 41 42 47

55

56 61 73 92 96

104 108 114 115 121 131 132

(vi )

CONTENTS PAGE

CHAPTER 5 - EVALUATION OF WONGANOO-BANDJAWARN GREENSTONE BELT 136

5.1 Geological Setting 138 5.2 Evaluation of individual areas within the greenstone belt 142

5.2. 1 Dingo Range West 142 5.2. 1.1 Geology 142 5.2. 1.2 Komatiite Sequence 143 5.2. 1.3 Geochemistry 149

5.2.2 Devines 151 5.2.3 Dingo Range East and Lalor North 151 5.2.4 Mt. Step and Collin Well 153

5.3 Discussion 159

CHAPTER 6 - COMPARISON OF FORRESTANIA NICKEL PROVINCE WITH THE CONTIGUOUS SOUTHERN CROSS GREENSTONE BELT

6.1 Forrestania Nickel Province 6.2 Southern Cross Greenstone Belt

6.2. 1 Marvel Lock A 6.2.2 Trough Wells 6.2.3 Southern Cross Drill Holes 6.2.4 Marvel Lock B 6.2.5 Marvel Lock C 6.2.6 Ennuin 6.2.7 Bullfinch

6.3 Discussion

161

164 166 169 171 171 174 176 178 180 181

CHAPTER 7 - APPLICATION OF GEOCHEMICAL CRITERIA TO NEW AREAS AND 183 SAMPLE GROUPS WITH HIGH DEGREE OF MISCLASSIFICATION

7.1 Application of geochemical criteria to new areas 184 7. 1.1 Area B 184 7.1.2 Area C 191

7.2 Barren groups with high percentage of samples misclassified 193 7.2. 1 Area A 193 7.2. 1 Yerilla 195 7.2.3 Heather Hill 197

CHAPTER 8 - DISCUSSION AND CONCLUSIONS

8.1 Mean Geochemical Values 8.2 Mineralized and Barren Komatiite Characteristics

8.2. 1 Volcanic Suite 8.2.2 Intrusive Suite 8.2.3 Application of Mineralized- Barren Criteria

8.3 Chemical Gradients 8.4 Regional and Stratigraphic Chemical Differences 8.5 Comments on Individual Areas 8.6 Significance of Sulphur 8.7 Genetic Aspects

202

203 204 208 211 212 214 215 216 219 222

(vii )

CONTENTS

REFERENCES

APPENDIX A - GEOLOGY AND GEOCHEMISTRY OF KALGOORLIE-NORSEMAN NICKEL PROVINCE

A. 1 Geological Setting A. 1.1 Stratigraphy A.l.2 Structure A. 1.3 Intrusive Rocks

A.2 Widgiemooltha Dome- Spargoville Nickel Occurrences A.2. 1 Spargoville A.2.2 Mt. Edwards A.2.3 Widgiemooltha A.2.4 Wannaway A.2.5 Redress

A.3 Geology and geochemistry of areas sampled A.3. 1 Bouchers A.3.2 Mt. Jewel, Mt. Jewel North and Red Dam A.3.3 Scotia A.3.4 Nepean A.3.5 Jubilee A.3.6 Creede A.3.7 South Bulong/North Bulong A.3.8 Wongi South

APPENDIX B - GEOLOGY AND GEOCHEMISTRY OF LEONORA-WILUNA NICKEL PROVINCE

B. 1 Descriptions of main nickel sulphide areas B.l.l Mt. Keith B.l.2 Six Mile Deposit B. 1.3 Perseverance B.l.4 Mt. Clifford

B.2 Geology and Geochemistry of areas sampled B.2. 1 Weebo Bore B.2.2 Sir Samuel B.2.3 Mestondab

APPENDIX C - GEOLOGY AND GEOCHEMISTRY OF THE WINDARRA NICKEL PROVINCE

C. 1 Geology and Geochemistry of areas sampled C. l. 1 Windarra

PAGE

232

234 234 240 240

242 242 246 246 249 250

251 251 266 273 277 283 285 287 289

291

296 296 299 299 303

305 305 308 308

312

314 314

(viii)

CONTENTS PAGE

APPENDIX D -ULTRAMAFIC ROCK ANALYSIS BY ATOMIC ABSORPTION SPECTRO- 321

D. 1 D.2 D.3 D.4 D.5 D.6 D.7 D.8 D.9 D. 10 D. 11 D. 12

PHOTOMETRY USING A HYDROFLUORIC/PERCHLORIC ACID DECOMPOSITION TECHNIQUE

Introduction Sampling and Preparation Instruments and Apparatus · Method of Decomposition Standard Solutions Reference Samples Cu, Ni, Zn, Co, Cr and Mn - Standard Conditions Ca - Standard Conditions Al - Standard Conditions Fe - Standard Conditions Mg - Standard Conditions Ti02 - Standard Conditions

APPENDIX E - MULVAR - DISCRIMINANT ANALYSIS

VITA

322 324 325 326 328 329 330 331 333 335 337 339

340