ex99-1.htm

ANNEXURE VII

COMPETENT PERSON'S REPORT ON THE CONTIGUOUS RIGHTS

Kalbasfontein and WA4
Cardoville and Wildebeestkuil Areas

Competent Person’s Report

Report Prepared for:

FREE STATE DEVELOPMENT AND INVESTMENT
CORPORATION LIMITED

Report No 381187

September 2007

Kalbasfontein and WA4
Cardoville and Wildebeestkuil Areas

Competent Person’s Report

FREE STATE DEVELOPMENT AND INVESTMENT
CORPORATION LIMITED

SRK Project Number 381187

265 Oxford Road

Illovo

2196

South Africa

P O Box 55291

Northlands
2116

South Africa

Tel: (011) 441-1111

Fax: (011) 880-8086

heyc@srk.co.za

September 2007

Compiled by

Collin Hey PrEng

MSc (Mining)

FSAIMM,AMMASA

_______________________

Project Consultants

Reviewed by:

Roger Dixon PrEng

BSc (Mining

FSAIMM,

_______________________

Partner

Contents

Page

Executive Summary		56

1 Introduction		57
1.1	Background	57
1.2	History	57
1.3	Competent Persons Report – Requirement, Compliance and Sources of Data	59
	1.3.1 Requirement	59
	1.3.2 Compliance	63
1.4	Preliminary Review – Sources of Data	63
	1.4.1 Sources of Data	63
	1.4.2 Effective Date, Valuation Date and Base Technical Information Date	63
	1.4.3 Verification, Validation and Reliance	63
	1.4.4 Technical Reliance	63
	1.4.5 Financial Reliance	63
	1.4.6 Legal Reliance	63
1.5	Basis of Valuation	64
1.6	Disclaimer, Limitations, Reliance on Information, Declarations, Consent and Copyright	64
	1.6.1 Disclaimer	64
	1.6.2 Limitations	64
	1.6.3 Reliance on Information	64
	1.6.4 Declarations	64
	1.6.5 Consent	65
2	Mineral Right Holdings	65
3	Geology	65
3.1	Ventersdorp Contract Reef (VCR)	65
3.2	Elsburg Sedimentary Units	65
4	Economic Geology of South Deep and Extensions	65
4.1	Upper Elsburgs	65
4.2	Facies Modelling	66
5	Sampling Method and Approach	73
5.1	Surface Drill Core Sampling Method	73
5.2	Underground Drill Core Sampling Method	73

5.3	Underground Chip Sampling Method	74
5.4	Sampling History	74
5.5	Other Experts	74
6	Sample Preparation, Analysis and Security	74
6.1	Results of South Deep QA/QC Program	76

Page

7	Data Verification	76
8	Mineral Resources and Reserves	77
8.1	Introduction	77
8.2	Datamine Modelling	77

8.3	Mineral Resource Estimate Results for the Upper Elsburgs	78
8.4	Methodology – Ventersdorp Contact Reef (“VCR”)	78
8.5	Mineral Resource Estimate - VCR	79
8.6	Cut-off Grades	81
8.7	Reserving Parameters	81

9	Mining	81
9.1	Mine Design	81
9.2	Conventional Stoping Design	82
	9.2.1 Destress Stope	82
	9.2.2 Skin Pillars	82
	9.2.3 Conventional Stoping	82

9.3

Trackless Stoping Design

	9.3.1 Drift and Fill with Benching	83
	9.3.2 Modified Drift and Fill with Benching	83
	9.3.3 Longhole Stoping Design	84
9.4	Project Areas	85
9.5	Bayesian Methodology	88
9.6	Application to Future Mining	89
9.7	Reserve Results - Upper Elsberg	90
10	Metallurgical	92
10.1	South Deep Processing Plant Flow Sheet	93
11	Tailings Disposal	94

12	Infrastructure	94
12.1	Project Infrastructure	94
13	Environmental Management	95
14	Technical Economic Parameters	96

14.1	Operating Cost	96
14.2	Capital and Sustaining Cost	96
14.3	Production Profiles	96
14.4	Recovery Profiles	96
14.5	Unit Revenue Profiles	96
14.6	Exchange Rate Profiles	96

Page

15	Material Assets Valuation	98
15.1	Valuation Methodology	98
15.2	Royalties	98

15.3

Taxation

15.4	DCF Valuation Results	98
	15.4.1 SRK Comments on Scenario Results	99
	15.4.2 Interpretation of DCF Results	99
	15.4.3 Comparable Transaction Valuation Results	100
	15.4.4 Interpretation of Comparable Results	100
15.5	Valuation Summary	102

15.6	Sensitivity	104
16	Conclusions	104
17	References	104

List of Tables

Table 1.1	History of Drilling	57
Table 1.2	History of Major Transactions at South Deep	58
Table 1.3	Kalbasfontein Mineral Rights	59
Table 1.4	WA4 Mineral Rights	60
Table 1.5	Cardoville Mineral Rights	61
Table 1.6	Wildebeestkuil Mineral Rights	63
Table 8.1	Tonnage/Grade Distribution (High, Moderate and Low Geozones)	78
Table 8.2	Tonnage/Grade Distribution at Various Grade Cut offs - VCR	80
Table 8.3	Reserving Parameters: Dilution & Losses	81
Table 9.1	Mining Method per Corridor	81
Table 9.2	Conventional Design Standards	83
Table 9.3	Modified Drift and Fill with Benching Design Standards	84
Table 9.4	Longhole Stoping Excavation Dimensions	85
Table 9.5	Mining Methods per Concatenation	89
Table 9.6	Probable Mineral Reserves in the Pillars	90
Table 9.7	Kalbasfontein Concatenations – South Deep Current Mine Equivalents	90
Table 9.8	Probable Mineral Reserves per Mining Method	90
Table 9.9	Probable Mineral Reserves by mining method and project area	91
Table 9.10	Total Probable Mineral Reserves by Mining Project Area	92
Table 14.1	Estimated Operating Costs	96
Table 14.2	Base Kalbasfontein Production Profile, Yield and Cost/milled Tonne	97
Table 15.1	Summary NPV Scenario Values	99
Table 15.2	South Deep Look Through Transaction	100
Table 15.3	Comparable Transaction Summary	100
Table 15.4	Comparable Transactions	101
Table 15.5	Valuation Summary of FSD’s 44% Share	102
Table 15.6	Gold Price Sensitivity on FSD’s 44% Share of the Kalbasfontein & WA4 Project	104

Page

List of Figures

Figure 4.1	The Six Reef Subdivisions Identified within the Upper Elsburg Reef Package	67
Figure 4.2	Schematic Section Showing Primary Reefs	67
Figure 4.3	Plan View of MBT Geozones	68
Figure 4.4	Plan View of MBB Geozones	69
Figure 4.5	Plan View of MIB Geozones	70
Figure 4.6	Plan View of MA Geozones	71
Figure 4.7	Plan View of ECT Geozone	72
Figure 4.8	Plan View of ECB Geozone	73
Figure 6.1	Laboratory checks	75
Figure 6.2	Results of South Deep QA/QC Program	75
Figure 8.1	West- East Section of the Upper Elsburgs	77
Figure 8.2	VCR Polygons within Western Areas WA4 Project Area (Not to scale)	79
Figure 9.1	Kalbasfontein Project Area, Mining Projects and Corridors	82
Figure 9.2	Modified Drift and Fill with Benching	84
Figure 9.3	Trackless Longhole Stoping	85
Figure 9.4	Project Areas in Kalbasfontein Area	86
Figure 9.5	Pillar Layout in Kalbasfontein Project Area	86
Figure 9.6	Development Accessing the Kalbasfontein Project Area	87
Figure 9.7	Three Dimensional View of the Wrench Fault in the Kalbasfontein Project Area	87
Figure 9.8	Application of Mining Designs to the Kalbasfontein Project Area	88
Figure 9.9	West – East Section through Future Mine Showing Regional Pillars	89
Figure 9.10	Oblique View of the Kalbasfontein Project Area and South Deep Pillars	89
Figure 10.1	Process Plant Flowsheet	93
Figure 15.1	Valuation Summary of FSD’s 44% Share	103

Cross reference of Section 12 Listing requirements

Report location	Section 12 Number
Front Cover	12.9(a) & (b)
Sec 1.1	12.3(e)&12.9(f)
Sec 1.2	12.10(b)(vi)
Sec 3	12.10(b)(i)
Sec 4.1	12.10(a)(xi)
Sec 5.1	12.10(a)(i&v))
Sec 5.2	12.10(a)(i)
Sec 5.3	12.10(a)(iii,iv&ix)
Sec 5.5	12.10(b)(vi)(9)
Sec 6	12.10(a)(vi) & (f)(i&ii)
Sec 7	12.10(a)(vii)
Sec 8.2	12.10(a)
Sec 8.3	12.10(b)(ii). 12.10(b)(vi)(4)
Sec 8.4	12.10(b)(ii,iii&iv) 12.10(b)(iv)(4)
Sec 8.5	12.10(b)(ii)
Sec 8.6	12.10(b)(vi)(4)
Sec 8.7	12.10(b)(vi)(5)
Sec 9.1	12.10(a)(xvi)
Sec 13	12.10(c)

Glossary

Geozone

Area with specific gold distribution pattern and economic potential

Mining Corridor

Subdivision of the South Deep Mining Authorisation area with specific sedimentological and grade characteristics and suitable for a particular mining method.

Proximal facies

The positional environment relatively close to the source of the sediments (used in association with distal facies)

Distal facies

The positional environment relatively far from the source of the sediments (used in association with proximal facies)

Abbreviations

IRP	Independent Review Panel.

Executive Summary

The Mineral Resources and Reserves for the Kalbasfontein, Cardoville, Wildebeestkuil and the WA4 Project areas have been estimated using a Bayesian methodology. Estimations procedures are similar to those methods used at South Deep Mine.

The estimate yields an Indicated Mineral Resource of 85.3Mt at an insitu grade of 7.36g/t gold containing 20.2Mozs within the Kalbasfontein Project Area, and an Inferred Mineral Resource of 10.3Mt at an insitu grade of 14.27g/t gold yielding 4.7Mozs gold within the WA4 Project Area.

Probable Mineral Reserves in the Kalbasfontein Project Area are estimated to be 58Mt with a head grade of 4.95g/t gold containing 9.2Mozs gold.

The probable reserves are defensible as they follow the guidelines and suggestions by the panel of experts comprising the Independent Review Panel.

SRK has valued the Kalbasfontein property using DCF techniques and comparable transactions and the WA4 area using comparable transactions.

SRK project valuation indicates that the Kalbasfontein property would not be economically viable as a stand alone project. In SRK’s opinion the best use of the Kalbasfontein property is by access through South Deep. When comparing the reserves to the current South Deep reserves, mining of the Kalbasfontein reserves in place of the current South Deep reserves reduces the South Deep NPV. The Kalbasfontein reserves maximum value is realized when these reserves are mined at the end of the South Deep mine life. It is SRK opinion that South Deep overall production risk profile would be reduced with the acquisition of the Kalbasfontein project, as this would allow more operational flexibility.

In SRK’s opinion the DCF value for 44.4% of the Kalbasfontein property including the WA4 lies between ZAR278m and ZAR327m.

In SRK’s opinion the Comparable value for 44.4% of the Kalbasfontein property including the WA4 lies between ZAR300m and ZAR394m.

Using a 33:66 weighting SRK values the 44.4% of the Kalbasfontein and WA4 property between ZAR293m and ZAR372m.

1 Introduction

1.1 Background

Goldridge Gold Mining Company (Proprietary) Limited, a subsidiary of Free State Development and investment Corporation (“FSD”) which in turn is jointly owned by Randgold And Exploration Company Limited (55.1%) and JCI (44.9%) wishes to dispose of its 44.4% share of ownership of the Kalbasfontein mineral rights contiguous to the South Deep Mine. In order to meet Listing requirements a Competent Person’s Report (CPR) and a valuation report need to be submitted to the JSE Limited. SRK have prepared this CPR from previous JCI reports based on a recent valuation completed by Mr. Dexter Ferreira, the competent person signatory, who has agreed that his work and findings be included. SRK remain responsible for the correctness of this CPR in its entirety. The Competent Person responsible for the declared Mineral Resources and Reserves is Mr. Dexter Ferreira. The Competent Person with overall responsibility for the report is Mr. Roger Dixon, who is a Partner and Corporate Consultant with SRK.

· The scope of this CPR includes:

· An assessment of Ferreira’s earlier work and its suitability for this project;

· A valuation of the contiguous rights in accordance with international best practice.

All Section 12 listing requirements have been complied with except those highlighted in thecross-reference. This report shows that the Kalbasfontein and WA4 mineral rights areas although containing some 18.7 million ounces of gold resources cannot justify a stand alone mine due mainly to the capital cost of a shaft system. The tenements do however add value to the contiguous and currently operating South Deep Mine. SRK estimates that the value of FSD 44.4% share in the Kalbasfontein and WA4 property lies between ZAR293m and ZAR372m.

1.2 History

12.10(b)(vi)(4)

South Deep started production in 1961 as the Western Areas Mine after a period of exploration activity led by the Johannesburg Investment Company Limited (“JCI”). Expansion into the southern down dip of the ore body was identified into a feasibility study completed in 1986. The South Deep Exploration Company Limited was formed and later merged with WAGM in 1995 when development of the South Deep Main shaft started.

Restructuring of the company after 1995 culminated in the formation of the Placer Dome Western Areas Joint Venture in 1999. The South Deep Shaft Complex was commissioned in 2005, signalling the start of the ramp-up to full production.

In March 2006, Barrick announced a take-over of Placer Dome Inc’s share of ownership in the South Deep Mine, resulting in the formation of the Barrick Gold Western Areas Joint Venture.

In February 2007, Gold Fields Limited bought out both South Deep JV partners.

Table 1.1: History of Drilling

Borehole No.	No. of intersections drilled	Drilled by	Date Completed	Comments
MD16	5	Jhb Consolidated Investments	1962	Within Current Mine
DP2	8	Jhb Consolidated Investments	1963	Inside lease Area Phase 1
MD24	9	Jhb Consolidated Investments	1965	Within Current Mine
MD42	4	Jhb Consolidated Investments	1965	Inside Lease Area
DP8	3	Jhb Consolidated Investments	1982	Inside lease Area Phase 2
KDP1	20	Anglo American Corporation	1982	On Southern Extremity of Lease Area
KDP3	11	Anglo American Corporation	1988	1 400m South of Lease Area

Table 1.2: History of Major Transactions at South Deep

Year	Event
1958	Johannesburg Consolidated Investment Company Limited (“JCI”) formed a prospecting consortium that acquired the prospecting rights over the farms Waterpan, Modderfontein and Jachtfontein in the Westonaria district.
1959	A formal application for mining lease was made, which resulted in the Western Areas Gold Mining Company Limited (“Western Areas”) being incorporated as a public company on September 8, 1959.
1961	Commercial production of the Western Areas Mine commenced in September with the ore treated in the nearby process plant at Harmony - Randfontein. Ore treated in a new 2959 t/d process plant at Western Areas.
1974	The Elsburg Gold Mining Company merged with Western Areas on July 1, 1974.
1982	Initial trials of trackless mining commenced.
1986	Initial Feasibility for the South Deep Mine completed.
1987	Trackless mining accounted for 72% of production at Western Areas.
1989	Revised Feasibility Study issued. Economics were improved after the boundary with WAGM was moved further to the north.
1990	Western Areas shareholders approved the transfer, cession and assignment of certain land and mineral rights to South Deep Exploration Company Limited in exchange for its shares.
1992	Reduction in the amount of trackless mining started due to high dilution and operating costs.
1995	Western Areas and South Deep Exploration Company merged on January 1, 1995. Pre-sink started for the South Deep Main shaft.
1997	In January 1997, the mine was split up into the Western Areas North Division and the Western Areas South Division. In July 1997 the North Division including the SV₄ sub-vertical shaft and the North Division gold plant were sold to Harmony - Randfontein Gold Mining Company Limited.
1998	WAGM changed its name to Western Areas Limited (WAL)
1999	On April 1^st, Placer Dome and WAL formed a 50:50 joint venture, Placer Dome Western Areas Joint Venture, to develop and operate the South Deep Mine in February 2000.
2001	South Deep Ventilation shaft sunk to a depth of 2759 m below collar. New mill with a capacity of 7233 t/d commissioned to replace the old plant.
2002	South Deep Main shaft commissioned for hoisting from mid shaft.
2003	South Deep Main hoisting shaft sunk to 2991 m below collar.
2004	A rock fall damages SV1 sub vertical shaft between 60 and 63 Levels on 2 January.
2005	Official opening of the South Deep shaft complex on 7 February.
2006	Barrick take-over of Placer Dome’s interest in the South Deep Joint Venture,
2007	Gold Fields Limited acquire a controlling share of South Deep.

1.3

Competent Persons Report – Requirement, Compliance and Sources of Data

1.3.1 Requirement

SRK was commissioned by Free State Development and investment Corporation (“FSD”) to prepare a CPR In order to meet Listing requirements a Competent Person’s Report (CPR) and a valuation report need to be submitted to the JSE Limited. SRK has been informed that a copy of this CPR will be filed with the Circular to JCI Shareholders and included in the prospectus lodged by JCI Limited in support of its proposed transaction.

The assets included in this transaction are presented in Tables 1.3 to 1.6

Table 1.3: Kalbasfontein Mineral Rights

Kalbasfontein Area	Farm Name	Farm No.	Portion	Area(Ha)
	Driehoek	343 IQ	Ptn 1	35.6966
	Driehoek	343 IQ	Ptn 2	35.6914
	Driehoek	343 IQ	RE of Farm	35.6966
	Modderfontein	345 IQ	Ptn /4	17.1306
	Modderfontein	345 IQ	Ptn 11/2	171.9859
	Modderfontein	345 IQ	Ptn 12/2	171.9859
	Modderfontein	345 IQ	Ptn 34/15	48.8687
	Modderfontein	345 IQ	Ptn 52	92.5052
	Doornpoort	347 IQ	Ptn 20/1	153.058
	Doornpoort	347 IQ	Ptn 19/4	19.3368
	Doornpoort	347 IQ	Ptn 22/1	3.383
	Doornpoort	347 IQ	Ptn 50/21	30.6116
	Doornpoort	347 IQ	Ptn 51/21	30.6116
	Doornpoort	347 IQ	Ptn 52/21	30.6116
	Doornpoort	347 IQ	Ptn 53/21	30.6116
	Doornpoort	347 IQ	RE of Ptn 1	139.5922
	Doornpoort	347 IQ	RE of Ptn 21/1	30.6116
	Kalbasfontein	365 IQ	Ptn 26/1	171.3064
	Kalbasfontein	365 IQ	Ptn 70 and 78	90.7924

Table 1.4: WA4 Mineral Rights

Western Areas 4 Area	Farm Name	Farm No.	Portion	Area(Ha)
	Davonia	363 IQ	Ptn of RE (REMA 4/3)	3.3127
	Rietfontein	349 IQ	Ptn 26/10 (RE/MA 29)	3.6655
	Rietfontein	349 IQ	Ptn 27/4 (RE/MA 30)	2.2786
	Doornpoort	347 IQ	RE Ptn 18/4 (RE/MA 18)	15.0805
	Doornpoort	347 IQ	RE Ptn 4/2 (RE/MA 14)	4.1024
	Doornpoort	347 IQ	Ptn 11/5 (RE/MA 11)	22.1419
	Doornpoort	347 IQ	Ptn 5/2 (RE/MA 15)	24.1581
	Doornpoort	347 IQ	Ptn 37	29.6989
	Doornpoort	347 IQ	Ptn 14/2	10.3549
	Doornpoort	347 IQ	Ptn 10/5 (RE/MA 7)	87.1785
	Doornpoort	347 IQ	RE/Ptn 7/2 (RE/MA 5)	144.5882
	Doornpoort	347 IQ	Ptn 16/2 (RE/MA 24)	91.9041
	Doornpoort	347 IQ	RE/Ptn 2 (RE/MA 22)	98.8154
	Doornpoort	347 IQ	RE/Ptn 15/2 (RE/MA 23)	135.2447
	Elandsfontein	346 IQ	Ptn 25 (RE/MA 10)	13.3481
	Elandsfontein	346 IQ	Ptn 2	4.0765
	Elandsfontein	346 IQ	Ptn 8/1	128.0144
	Elandsfontein	346 IQ	RE/Ptn 4/1 9 (RE/MA 9)	247.6808
	Elandsfontein	346 IQ	RE/Ptn 27/5	6.0994
	Elandsfontein	346 IQ	RE 1 (RE/MA 4)	264.9097
	Elandsfontein	346 IQ	RE/Ptn 3/1 (RE/MA 8)	110.2797
	Elandsfontein	346 IQ	Ptn 28/27	194.0303
	Elandsfontein	346 IQ	Ptn 18/3 (Part of Ptn 34)	77.4962
	Elandsfontein	346 IQ	Ptn 19/3 (Part of Ptn 34)	77.4962
	Elandsfontein	346 IQ	Ptn 29/27	194.0303
	Elandsfontein	346 IQ	Ptn 17/3	77.4962
	Waterpan	292 IQ	Ptn 19/ 7	64.0605
	Modderfontein	345 IQ	RE/Ptn 53/Ptn1	85.6727
	Modderfontein	345 Iq	Ptn 67/5 (RE/MA 5)	79.2796

Table 1.5: Cardoville Mineral Rights

Cardoville Area	Farm Name	Farm No.	Portion	Area(Ha)
	Cardoville	358 IQ	Ptn of RE (REMA 4/3)	3.3127
	Cardoville	358 IQ	Ptn 26/10 (RE/MA 29)	3.6655
	Cardoville	358 IQ	Mineral Area 3 on Portion 3	76.1623
	Cardoville	358 IQ	Mineral Area 1 on Portion 1	43.7201
	Cardoville	358 IQ	Portion 2	52.8842
	Cardoville	358 IQ	Portion 4	75.5954
	Cardoville	358 IQ	Remaining Extent	127.3952
	Cardoville	364 IQ	Remaining Extent of Portion 1	321.2110
	Cardoville	364 IQ	Portion 9 of 3 (Now portion of 11)	103.9578
	Cardoville	364 IQ	Portion 10 of 5 (Now portion of 11)	69.1629
	Cardoville	364 IQ	Remaining Extent of Portion 3	358.1697
	Cardoville	364 IQ	Remaining Extent of Portion 5 of 3	139.4112
	Cardoville	364 IQ	Portion 4 of 3	2.5696
	Doornpoort	347 IQ	Portion 51 of 21	30.6116
	Droogheuvel	521 IQ	Portion 10 of 5	40.6924
	Droogheuvel	521 IQ	Remaining Extent	117.1551
	Droogheuvel	521 IQ	Portion 2	362.3287
	Elandsfontein	346 IQ	Portion 12 of 6	176.6954
	Elandsfontein	346 IQ	Portion 10 of 5	176.6926
	Foch	150 IQ	Portion 3	66.4098
	Foch	150 IQ	Portion 6 of 4	29.9786
	Foch	150 IQ	Remaining Extent of Portion 4	55.6746
	Foch	150 IQ	Portion 13 of 5	13.9612
	Kraalkop	147 IQ	Remaining Extent of Portion 11 of 2	205.8168
	Kraalkop	147 IQ	Portion 36 of 11	21.9708
	Kraalkop	147 IQ	Portion 16 of 11	89.5076
	Kraalkop	147 IQ	Remaining Extent of Portion 11 of 2	330.5128
	Leeuwspruit	148 IQ	Portion 5 of 1	332.2351
	Rietfontein	519 IQ	Portion 17 of 4	108.6559
	Rietfontein	519 IQ	Remaining Extent of Portion 4	108.9537
	Rietfontein	519 IQ	Remaining Extent of Portion 8 and Portion 21 of 8 (Now portion 8 of 1)	245.4964
	Rietfontein	519 IQ	Portion 18 of 4	107.9963
	Rietfontein	519 IQ	Remaining Extent of Portion 1	245.1738
	Rietfontein	519 IQ	Portion 16 of 4	171.3064
	Springbokkraal	359 IQ	Portion 1	133.6190
	Wildebeeskuil	360 IQ	Portion 1	209.5834
	Wildebeeskuil	360 IQ	Remaining Extent	12.3369
	Wildebeeskuil	360 IQ	Mineral Area 2 on 18	211.2965
	Wildebeeskuil	360 IQ	Mineral Area 23on 18 (17 of 5)	14.2298

Table 1.5: Cardoville Mineral Rights

Cardoville Area	Farm Name	Farm No.	Portion	Area(Ha)
	Wildebeeskuil	360 IQ	Portion 7	229.5977
	Wildebeeskuil	360 IQ	Portion 2	209.5834
	Kaalplaats	394 IQ	Portion 37	262.5102
	Kaalplaats	394 IQ	Remaining Extent of Portion 8 (Portion of Portion 7)	252.7097
	Kaalplaats	394 IQ	Remaining Extent of Portion 9 (Portion of Portion 7)	501.0681
	Kaalplaats	394 IQ	Mineral Area 2/Mineral Area 3 (On Portion 7)	228.0873
	Rietfontein	399 IQ	Mineral Area 1/Mineral Area 2 (On Portion 1)	41.2823
	Taaiboschspruit	400 IQ	Portion 1	36.1456
	Taaiboschspruit	400 IQ	Portion 6 of 3	193.7718
	Taaiboschspruit	400 IQ	Portion 4	85.6076

Table 1.6: Wildebeestkuil Mineral Rights

Wildebeestkuil Area	Farm Name	Farm No.	Portion	Area(Ha)
	Wildebeestkuil	360 IQ	Portion 6	3.3127
	Wildebeestkuil	360 IQ	A Portion of Portion 6	3.6655
	Wildebeestkuil	360 IQ	Mineral Area 4 on Portion 4	2.2786
	Cardoville	358 IQ	Carol 362 IQ	15.0805
	Cardoville	364 IQ	Portion 7	4.1024

1.3.2 Compliance

This report has been prepared under the direction of the competent person (the “CP”) who assumes overall professional responsibility for the document. The report however is published by SRK, the commissioned entity, and accordingly SRK assumes responsibility for the views expressed herein. Consequently with respect to all references to CP and SRK: 'all references to SRK mean the CP and vice-versa'.

1.4 Preliminary Review – Sources of Data

1.4.1 Sources of Data

Details of data/information used to prepare this report were provided by FSD.

1.4.2 Effective Date, Valuation Date and Base Technical Information Date

The effective date (the “Effective Date”) of this report is deemed to be 1 August 2007.

1.4.3 Verification, Validation and Reliance

This report is dependent upon technical input. The technical information as provided to and taken in good faith by SRK has not been independently verified by means of re-calculation. SRK has however; conducted a preliminary review and assessment of some of the material technical issues likely to influence the future performance of the assets.

·	Satisfied itself that such limited information provided by FSD is appropriate and valid as reported herein. SRK cannot consider that with respect to all material technical-economic matters it has undertaken all necessary investigations to be SAMREC compliant.

Where fundamental base data has been provided (LoM plans, capital expenditures, operating budgets etc) for the purposes of this preliminary review, SRK state that it has not performed all necessary validation and verification procedures deemed appropriate in order to place an appropriate level of reliance on such information.

1.4.4 Technical Reliance

SRK places reliance on the personnel at FSD that all technical information provided to SRK as at 1 August 2007, is both valid and accurate for the purpose of compiling this report. The information with respect to Mineral Resources and Mineral Reserves as stated by FSD. The Competent Person responsible for the declared Mineral Resources and Reserves is Mr. Dexter Ferreira.

1.4.5 Financial Reliance

In consideration of all financial aspects relating to the assets and the valuation thereof, SRK has placed reliance on the Financial Officers of the Company that the following information is appropriate as at 1 August 2007:

1.4.6 Legal Reliance

In consideration of all legal aspects relating to the valuation of the Assets, SRK has placed reliance on the legal representatives JCI Limited that the following are correct as at 1 August 2007:

·	That the legal ownership and of all mineral and surface rights has been verified;

·	That no significant legal issue exists which would affect the likely viability of the company and/or on the estimation and classification of the Mineral Reserves and Mineral Resources as reported herein; and

·	That no significant legal issue exists from existing or future environmental compliance which would affect the likely viability of the company.

1.5 Basis of Valuation

The valuation of the assets comprises the following:

·	The valuation date is 1 August 2007.

The following documents as provided by JCI and Gold Fields:

·	Western Areas – Kalbasfontein and WA4 Project Area, Dexter S. Ferreira, 2/27/2007;

·	Total CMSTMod (Dexter Ferreira on Uncle Harry's) (Excel Document); and

·	Uncle Harry’ s Ground(SRK) (Excel Document).

Although SRK has not validated the underlying data contained therein, SRK has carried out a high level audit and are satisfied that the information used will not lead to a material difference to the conclusion presented.

1.6 Disclaimer, Limitations, Reliance on Information, Declarations, Consent and Copyright

1.6.1 Disclaimer

The observations and comments presented in the report represent SRK’s opinion as of 1 August 2007.

The standard practices adopted by JCI and FSD were developed over many years and compared favourably with international best practice. SRK has therefore no reason to doubt the underlying data used to derive the Mineral Resource and Reserve estimates.

The underlying data for the model generated in this report was accepted in good faith and SRK do not accept any liability either direct or consequential for the veracity of this information.

1.6.2

Limitations

The Company has agreed that, to the extent permitted by law, it will indemnify SRK and its employees and officers in respect of any liability suffered or incurred as a result of or in connection with the preparation of this report. This indemnity will not apply in respect of any negligence, wilful misconduct or breach of law. The Company has also agreed to indemnify SRK and its employees and officers for time incurred and any costs in relation to any inquiry or proceeding initiated by any person except where SRK or its employees and officers are found liable for, or guilty of, gross negligence, wilful misconduct in which case SRK shall bear such costs.

The Company has confirmed in writing to SRK that to its knowledge the information provided by it was complete and not incorrect, misleading or irrelevant in any material aspect. SRK has no reason to believe that any material facts have been withheld.

The achievability of LoM Plans, budgets and forecasts are neither warranted nor guaranteed by SRK. The forecasts of the assets as presented and discussed herein have been proposed by management and adjusted where appropriate by SRK, and cannot be assured; they are necessarily based on economic assumptions, many of which are beyond the control of FSD. Future cashflows and profits derived from such forecasts are inherently uncertain and actual results may be significantly more or less favourable.

1.6.3 Reliance on Information

SRK’s value for the Assets is effective as at 1 August 2007 and is based on information provided by JCI Limited throughout the course of SRK’s investigations, which in turn reflect various technical-economic conditions prevailing at the date of this report. In particular, the value is based on expectations regarding the commodity prices prevailing at the date of this report. These and the underlying Technical Economic Parameters (TEPs) can change significantly over relatively short periods of time. The value could be materially different in these changed circumstances. Further, SRK has no obligation or undertaking to advise any person of any change in circumstances which comes to its attention after the date of this technical report or to review, revise or update the technical report or opinion.

1.6.4 Declarations

SRK will receive a fee for the preparation of this report in accordance with normal professional consulting practice. SRK will receive no other benefit for the preparation of this report. SRK does not have any pecuniary or other interests that could reasonably be regarded as capable of affecting its ability to provide an unbiased opinion in relation to the Mineral Resources, the Mineral Reserves and the value of the assets.

SRK does not have at the date of this report, and has not had within the previous two years, any shareholding in or other relationship with FSD. SRK considers itself to be independent.

In this report, SRK provides assurances to the Directors of the Company that the TEPs, including production profiles, operating expenditures and capital expenditures, of the assets as provided to SRK by FSD.

This report includes technical information, which requires subsequent calculations to derive subtotals, totals and weighted averages. Such calculations may involve a degree of rounding and consequently introduce an error. Where such errors occur, SRK does not consider them not to be material.

1.6.5 Consent

SRK consents to the issuing of this report in the form and content in which it is to be included in documentation distributed to the directors and shareholders of EH. Neither the whole nor any part of this report nor any reference thereto may be included in any other document without the prior written consent of SRK as to the form and context in which it appears.

Copyright of all text and other matter in this document, including the manner of presentation, is the exclusive property SRK. It is a criminal offence to publish this document or any part of the document under a different cover, or to reproduce and/or use, without written consent, any technical procedure and/or technique contained in this document. The intellectual property reflected in the contents resides with SRK and shall not be used for any activity that does not involve SRK, without the written consent of SRK.

2 Mineral Right Holdings

An evaluation of the area was conducted in December 1997 by JCI for the exchange of mineral rights with AAC. This involved farms in the Fochville, Kalbasfontein, Cardoville, Doornkop, Edom and Witrand areas. The outcome was that the Kalbasfontein rights, lying east of South Deep, was acquired by Kalbasfontein Mining Company (Kalbasfontein), jointly owned by JCI and WAGM.

In November 2004, the Kalbasfontein rights as well as the WA4 rights on the west and contiguous to South Deep Gold Mine were reviewed. As a result JCI now owns 44,4% of these mineral rights with WAL holding the balance. In 2007, Gold Fields Limited acquired the WAL holding.

The Mineral rights are shown in Table 1.3 to 1.6 and were provided by Venmyn Rand.

3 Geology

The area is underlain by strata of the Waterpan and Modderfontein Members of the Mondeor Formation which occurs towards the top of the Central Rand Group and this is unconformably overlain by the Ventersdorp Formation. Gold concentrations occurring in the area are associated with the Ventersdorp Contact Reef (VCR) and the Elsburg reefs.

3.1 Ventersdorp Contract Reef (VCR)

12.10(b)(i)

The VCR represents a major unconformity developed at the base of the Ventersdorp Supergroup, which in the report area, becomes less pronounced to the east of the Upper Elsburg subcrop position. The VCR is complex in its origin, and morphology, being characterised by well developed terraces on the Kloof Gold Mine to the west of South Deep. Apart from the structural discontinuities that exist between Kloof and South Deep Mine (Witpoortjie Horst), the VCR can effectively be regarded as being continuous by virtue of the fact that South Deep is “down-palaeoslope” from Kloof.

3.2 Elsburg Sedimentary Units

12.10(b)(i)

The larger upward-coarsening Elsburg sedimentary Units (EA to MB) represent prograding braid plain deposits, with smaller, upward-fining cycles representing fluvial channels migrating on these larger prograding braid plains. The palaeocurrent direction, as determined on the mine, is from the west-northwest to east-southeast. The more proximal deposits are hence situated to the west (immediately east of the Upper Elsburg subcrop position), with more distal deposits occurring to the east. In a distal environment, the conglomerate units tend to be more discrete, and have a higher percentage quartzite development per unit thickness.

4 Economic Geology of South Deep and Extensions

4.1 Upper Elsburgs

12.10(a)(xi)

Initially, the South Deep Mining Authorisation was subdivided into mining corridors by the Strategic Planning department at South Deep, in order to identify zones suitable for a particular mining method. The corridor boundaries parallel the Upper Elsburgs subcrop position (where Upper Elsburg reefs converge and die out towards the west against the VCR) and correspond to total Upper Elsburg reef package thickness, thereby assisting with the selection of an appropriate mining method. The positions of the corridor boundaries were ratified by both the South Deep Geological Department and the Independent Review Panel (IRP), sponsored by South Deep.

With a 3D block model having been completed for South Deep, it was given that volumes for each of the reef units could be determined for the Kalbasfontein Project Area (Kalbasfontein). A Bayesian method would then be used to assign an estimated gold distribution curve to the volumes (area underneath the curve) modeled in the Kalbasfontein Project Area on a per reef basis.

Mr. D Muntingh of the IRP proposed that reef units, as identified in South Deep Current Mine, could not simply be utilized as a general footprint for assigning grade values to the same reefs in areas to be mined in the future. His studies revealed the sedimentological characteristics of the reefs changed from north to south, and thus care should be taken when making grade assignments. He, along with the South Deep Geology Department, analysed the logs and drill cores of some 250 underground drillholes as well as a number of surface drillholes. He proposed a number of contour lines denoting the potential of certain reef units based on the nature and behaviour of the conglomerates. In striving to incorporate a high level of geological detail into the current Mineral Resource and Reserve Estimate at South Deep, it was decided that a detailed sedimentological model for the Upper Elsburg orebody be incorporated in the modelling process, with specific reference to Phase 1 and 2 of the South Deep Mine.

This objective was twofold: to distinguish the main stratigraphic units to a resolution where the primary unconformities and main quartzite and conglomerate (reef) lithological units are honoured to model the lateral geological facies changes within the reef zones, and to relate to gold distribution, in order that areas (geozones) of equal economic potential could be delineated.

4.2 Facies Modelling

Based on information from 250 underground and 20 surface boreholes, this investigation conclusively correlated 6 unconformity-bounded units across the South Deep Mining Authorisation and the contiguous Kalbasfontein Lithologically these unconformity-bounded units were further split into 6 conglomeratic reef and 5 sandy units, totalling 11 stratigraphic units. Refer Figure 4.1.

Typical of the Witwatersrand, the 6 auriferous Upper Elsburg conglomeratic reef units at South Deep also occur directly above the 6 main unconformities. Gold-poor sandy units gradationally overlie five of these reef units.

Within the sub-region, all 6 auriferous conglomerate reef units display significant lateral variation with respect to the amount (cumulative thickness and percentage) of conglomerate development, as well as associated clast size. These lateral variations range between two end member facies, namely:

·	A coarse conglomeratic facies comprising a high proportion (>70%) and/or thickness of conglomerate with large to cobble sized clasts. These facies normally occur as lobes that parallel the Upper Elsburg Subcrop;

·	A sandy conglomerate facies generally comprising a low proportion of conglomerate (<30%) and small to medium-sized clasts. This facies gradationally bounds the conglomeratic lobes to the north, east and south,

Within all the reefs, economic gold mineralisation shows a good positive spatial correlation with the coarse conglomeratic facies, while the sandy conglomerate facies generally host uneconomic gold concentrations. Based on these associations, the high potential geozones closely follow the lobes of coarse conglomerates, while areas of poor and fine conglomerate development have been delineated as low potential geozones. The moderate potential geozones represent the transition between these high and low potential geozones. This work has indicated the vertical coincidence of the economically important high potential geozones of several reefs to form an area of prolific economic bulk tonnage potential, stretching from the south of South Deep Current Mine through to South Deep Phase 1 to the northern 20% of South Deep Phase 2, (future mining areas to the south) and into the Kalbasfontein Project Area.

The southern 80% of South Deep Phase 2 is dominated by facies of moderate economic potential. D.L. Arthur (1990), based on her investigation of surface boreholes in the region, first defined the stratigraphy of the Upper Elsburgs at South Deep. She identified 5 main stratigraphic units, namely the EA, EB, EC-ED, MA and MB units. She further split each of the EB, EC – ED and MA units into a lower conglomeratic sub-unit and an overlying sandy sub-unit.

Arthur’s stratigraphic subdivision was subsequently refined in agreement with WAGM original definitions. Focusing mainly on the economically important EC and M packages, these adaptations include the addition of the EC9 unit at the base of the EC’s, the identification of the MIB conglomerate between the MA and MB units and the recognition of an MB Top sandy unit above the MB conglomerate unit. Based on this work an eleven-fold stratigraphic subdivision was adopted. Refer to Figure 4.1.

Six primary unconformity surfaces within the Upper Elsburgs, each progressively eroding deeper into the underlying units towards the west, have been identified. The westerly convergence of unconformities against the VCR resulted in a westerly tapering, wedge-shaped, preservation of the Upper Elsburg ore body, which is the primary control on ore volumes. These 6 unconformity surfaces can be correlated throughout the South Deep Mining Authorisation and the Kalbasfontein Project Area . Refer to Figure 4.2.

Figure 4.1: The Six Reef Subdivisions Identified within the Upper Elsburg Reef Package

Figure 4.2: Schematic Section Showing Primary Reefs

The lithological sequences between the unconformities reveal a cyclical occurrence of conglomerate and quartzite typical of the Central Rand Group of the Witwatersrand Supergroup. The main conglomeratic units characteristically, directly overlie the unconformities, which grade upwards into a more quartzitic zone with subordinate finer conglomerates. Such upward-fining unconformity-bound sequences are considered to each be genetically related to a major tectonic event. In this context, the lower coarse conglomeratic portions represent high-energy deposition during peak tectonism, while the relatively finer sandy overlying zones represent the waning of the same tectonically controlled depositional event. These gradationally upward-fining waning flow cycles are typical of fluvial depositional environments.

Core from a representative intersection of each of 20 surface boreholes covering the Kalbasfontein Project Area, the South Deep Mining Authorisation, and south of the mine’s lease boundary were inspected. With the exception of boreholes DP12 and DP13, which occur within the composite reef zone near the subcrop, the 11-unit stratigraphic subdivision could be confidently applied to all the remaining surface boreholes. The conglomeratic units were assessed and modelled in terms of average clast size, percentage conglomerate and maximum clast size. These parameters were correlated to gold grade, and on which basis three zone types were indicated, namely: High, Moderate and Low Economic Potential. The MBT unit has an area of low economic potential over the Kalbasfontein Project Area, with a small area of moderate economic potential. Refer to Figure 4.3.

Figure 4.3: Plan View of MBT Geozones

The MBB Unit is, on average, the most conglomeratic unit within the Upper Elsburg package, with maximum conglomeratic development in the central sector of South Deep. Due to the lobate geometry of this unit, both the associated high and moderate potential zones extend well into the Kalbasfontein Project Area. Considering the areal extent and great widths (up to 21m thickness in the distal areas) of mineralisation within this package, it is evident that the MBB Reef unit has exceptional resource potential, potentially hosting the bulk of the total gold resource within the sub-region. Although lobe geometry is representative of the geozones, in reality the smooth curves shown would be demarked by a series of interfingering conglomeratic sequences. Some conglomerate waning quicker away from the subcrop whilst others would continue further. Given the magnitude of the scale of the deposit, this simplification of the lobate geometry is reasonable.

Figure 4.4: Plan View of MBB Geozones

The MIB conglomerate is generally poorly developed in the northern sector of South Deep but becomes of greater economic significance in the southern sector of South Deep (Phase 2) due to the more prolific development of the conglomerate. As such, the southern sector of South Deep is dominated by the moderate potential zone. The moderate potential zone extends into the Kalbasfontein Project Area slightly, with the bulk of the Project Area being dominated by the low economic potential zone. Refer to Figure 4.5.

Figure 4.5: Plan View of MIB Geozones

The MA unit is also lobate in nature, with its associated high and moderate potential zones partly entering the Kalbasfontein Project Area. However, the majority of the Project Area is dominated by the low economic potential zone (Figure 4.6).

Figure 4.6: Plan View of MA Geozones

The ECT unit is more linear in its distribution by way of its associated high, moderate and low potential zones and hence conglomeratic development.

As regards to the Kalbasfontein Project, the area is dominated by the low potential zone of the ECT unit (Figure 4.7).

Figure 4.7: Plan View of ECT Geozone

The ECB unit is most conglomeratic primarily in the Upper Elsburg subcrop, for which a narrow (±300m) low potential zone has been delineated. Accordingly, the majority of South Deep is characterised by the low potential zone, as is the case for the Kalbasfontein Project Area (Figure 4.8).

Figure 4.8: Plan View of ECB Geozone

5 Sampling Method and Approach

5.1 Surface Drill Core Sampling Method

12.10(a)(i&v)

Drill core is split using either a mechanical core splitter or by using a diamond saw. Prior to splitting the core in half, the geologist logging the core marks the position along which the core will be halved, so as ensure that sedimentological features that potentially control gold mineralization are sampled proportionately. Sample widths for the reef intersections vary from 15cm to 30cm and for the intervening barren quartzite from 50 cm to 100cm. Each sample is collected and placed in a plastic bag with a corresponding sample number and then submitted for analysis. Once the samples are taken the remaining core is transferred to a core library for storage and future reference.

Core recovery is generally very high, typically greater than 95%.

5.2 Underground Drill Core Sampling Method

12.10(a)(i)

Reef units and structural contacts are marked on the core and sampling is done based on the geology contacts such that the sample intervals do not cross stratigraphy or structural features. Whole core is taken, with sample lengths ranging from 10cm to 100cm centimetres, with the average sample length being about 25cm. Two to three samples are routinely taken from the hanging wall lava and the footwall quartzite. Non-sampled core from the footwall quartzite units or the lava units is discarded. The core is broken up using a hammer and placed into individual clear plastic bags along with the appropriate bar coded sample ticket. The individual samples are grouped into their relative batches and placed into larger thick plastic bags with a batch summary docket for submission to the assay laboratory.

Since 2003, blank and standard samples have been routinely submitted for assay with each batch dispatched such that for each 22 or less samples in a batch, one blank and one standard are included.

Core recovery is generally very high, typically greater than 90%.

5.3 Underground Chip Sampling Method

12.10(a)(iii&iv)

Underground chip sampling is carried out on all stope faces and on any on-reef development drives, generally raises and winzes. Sample section intervals in the stopes of the Upper Elsburg reefs are spaced every 6m on dip and every 6m along strike. In the VCR stopes, sample sections are spaced every 6m on dip and every 6 m on strike. In development ends, chip sampling is done on dip at 5 m intervals.

The area to be sampled is washed prior to sampling, so that any possible contamination is removed. Sample sections and individual sample widths are then marked off in chalk with the sample widths being determined geologically. The average sample length is typically 20cm. Sample sections are orientated perpendicular to the dip of the reef to achieve a true intersection thickness and are taken from the hanging wall to the footwall, such that the whole exposure length is sampled.

Samples are collected by being chipped from the face using a hammer and chisel, with the rock fragments being collected directly into a metal tray held below the chipped section. Oversize fragments are broken by hand using a thick cloth or rubber pad and a second hammer. The rock fragments are then bagged individually in plastic bags with the appropriate sample ticket and sample weights are generally around 500g.

12.10(a)(ix)

Chip sample data is used for the month-end evaluation of the broken ore and the grade of material sent to the mill. The gold predicted by the mine or “called for” compares favourably to the monthly gold accounted for by the mill. However, a 10% to 30% bias (M. Harley, SRK) has been identified in the chip samples when compared to the diamond drillholes and as such the chip samples are not used for resource estimation. Due to this known bias and the desire for quick laboratory turn around for production support decisions, no blank or standard samples are submitted with chip sample batches.

5.4 Sampling History

In 2002, Dr. M. Harley of SRK reviewed the quality of the chip sample data collected from underground workings and statistical comparison completed by Mr. D. Ferreira (January 2002). As a result, the chip samples were excluded from the data used for resource estimation.

5.5 Other Experts

12.10(b)(vi)(9)

R. Lewis of Placer Dome Asia Pacific validated the SV1 drillhole database used for resource estimation using DatamineÒ macros. SRK’s review has relied upon this validation and SRK has not completed an independent validation of this data.

Henry Lombard an independent consultant has reviewed and validated the surface drillhole data in March 2000. SRK’s review has relied upon this validation and SRK has not completed an independent validation of this data.

Andre Deiss (Pri. Sci. Nat.) of Geologix Mineral Resource Consultants (Pty) Limited validated the Current Mine drillhole database used for resource estimation using manual techniques as well as DatamineÒ macros. SRK’s review has relied upon this validation and SRK has not completed an independent validation of this data which was signed off by the former acting South Deep Manager of Geology, Mr. J. Kleynhans.

6 Sample Preparation, Analysis and Security

12.10(f)(i)

South Deep Mine uses the Harmony - Randfontein; Cooke Laboratory located approximately 25km north of the property for all sample analysis. Until November 2000 the laboratory was owned by Harmony - Randfontein Limited part of the Harmony Gold Mining Group. At the time of the sampling and assay analysis undertaken by Mr. D. Ferreira, Harmony - Randfontein; Cooke Laboratory was not an accredited SANAS laboratory. Currently the laboratory is operated and managed by Performance Laboratories, part of the Time Mining Group. The Cooke Laboratory is accredited for gold fire assay in terms of ISO 17025.

South Deep Mine uses the SGS Lakefield Laboratory in Johannesburg as an umpire laboratory for quality assurance and quality control purposes. The SGS Lakefield Laboratory is accredited for gold fire assay in terms of ISO 17025.

12.10(a)(vi)

Samples are transported to the laboratory by mine staff with no further handling on the mine site. Upon arrival at the laboratory the samples follow the process as outlined below. A separate, dedicated stream is used for all South Deep Mine samples.

The current sample preparation protocol is as follows:

·	Samples are sorted into respective batches and placed into numerically ordered metal trays;

·	Samples are dried under overhead electric air dryers;

12.10(f)(ii)

·	Entire samples are jaw crushed to a nominal 10mm;

·	Crushed samples are pulverised in a ring and puck pulveriser;

·	Pulps are homogenized by manual shaking in cocktail shakers;

·	500g homogenized pulps are placed in metal containers with appropriate tags.

All samples are fire assayed and assay results contained in electronic files from the laboratory information management system are imported directly into the mine sample management database. The assay results are electronically and manually validated upon import into the database.

Historically, samples were crushed by making use of a disc pulveriser, while Aztec was used for assaying instead of fire assay. The following quality control procedures are in place:

·	Five pulp samples are randomly selected each day and a screen analysis carried out to determine the portion passing 75 microns;

·	Two samples are randomly selected from each tray of twenty samples and re-assayed;

·	The results of the internal laboratory check samples and size analysis are reported with the results of each batch submitted. A monthly laboratory quality control report is also submitted to the mine;

·	Standards are routinely assayed;

·	To check for any contamination, Geologists routinely send blanks for assaying;

·	To check for precision, Geologists routinely submit blind standards for assaying;

·	Pulp rejects are routinely re-submitted for re-assay to measure repeatability since 2003.

Data was extracted from DiGS Database and Excel spreadsheets updated internally by the Geology Department for the period under review. Since 2003 the laboratory has produced a grind average of 78.56 micron.

The South Deep QA/QC Program started in earnest in June 2003 after the Bloom Audit. To date 2,407 internal laboratory check samples have been submitted and show an excellent correlation (Refer to Figure 6.1), while 512 blind precision repeat check samples have been sent to the laboratories, of which 300 were submitted during 2005. The reduced correlation is evident when compared to the internal laboratory check samples correlation coefficient.

The graphs in Figure 6.2 depict the repeatability of the assays.

Figure 6.1: Laboratory checks

Figure 6.2: Results of South Deep QA/QC Program

6.1 Results of South Deep QA/QC Program

The mine has documented and operates a quality assurance program to monitor, identify and rectify quality issues at the main assay laboratory. The program includes site inspections and monitoring, communication and reporting activities, and audit and review of operations.

Any batch of samples whose control samples do not meet the required quality standards are not used for resource estimation purposes. During 2005, only 32 boreholes were excluded from the resource estimation database for this reason.

This was based on a failure of some 306 samples out 11,160 samples, equivalent to 2.74% of samples.

Monitoring of the performance of the assay laboratory using the Quality Control and Quality Assurance program is done daily and reported to mine management on a quarterly basis. Results show that assays are reliable and variances are within acceptable norms.

Performance Laboratories was thoroughly audited by the Perth based Resource Services Group in 2001, and again by Lynda Bloom of Analytical Solutions ltd. in 2003. Both Performance Laboratories and SGS Lakefield participate in the Geostats. Pty. round robin program (an international quality assurance program).

Enhancements were to be made to the sampling, assaying , reporting and security as part of a Quality Assurance and Quality Control program, (refer to pages 27 and 28 of December 31, 2001 Draft South Deep OSC Report which lists the changes and improvements to be undertaken). These were as follows:

·	Dust contamination to be controlled by installation of extractor fan in the crushing and pulverising sections;

·	Quartz silica flushing for pulverising equipment plus sample tray cleaning to be standard;

·	Pulps to be stored in sealed bags with appropriate tags;

·	Open access to all internal check systems and round robin results;

·	Samples to be prepared using ring and puck pulverising equipment;

·	All samples to be analysed by fire assay with parting;

·	Detection limits for the fire assays with parting to be 0.04g/t Au.

A large number of specific gravity (“SG”) determinations have been carried out on drill core from South Deep over the life of the mine. The latest campaign using a mass air/mass water method was carried out in October 2004. The average value obtained of 2.71g/cm3 compares well with other campaigns.

7 Data Verification

Mr. Henry Lombard, an external geological consultant with expert knowledge of the South Deep area, was responsible during March 2000 for completing the data verification of the twenty one surface drillholes.

South Deep Mine personnel verified the underground drillhole data for inclusion in the resource estimate during January 2005.

The following fields in the drillhole database were verified against the original log sheets, original assay sheets (available for drillholes drilled in the last 1.5 years), and geological plans and sections:

·	Drillhole number (BHID);

·	Drillhole collar information (X, Y, Z co-ordinates);

·	Down-hole survey information (core length from the collar, bearing, dip);

·	Geological information (reef type, rock type, structure type);

·	Assay information (Au values, U₃O₈ values).

This methodology was used for all data pre-dating the installation of the electronic sample database management system.

Since January 2002 South Deep has used a commercial sample RDBMS to collect, validate, analyse and store drillhole logging and sample data. The DiGS software is Microsoft SQL Server based and allows controlled access to data and storage, maintenance and tracking of data. Data is validated on entry into the software using a combination of drop down options lists, rules and filters. Once entered into the system, checked and approved, data is digitally “locked” so that that it may not be altered without approved access. “Locked” data is user, date and time stamped so changes can be tracked. A hard copy log of each drillhole in the database is also created and filed together with a hardcopy of the assay results file and manifests and the down hole survey file. Each drillhole log also contains a control sheet checklist detailing the data entry, checking, validation and approval carried out.

Approved data for resource estimation purposes is queried from the software system in a format identical to that of the older data. The two sets of data are then merged and both subjected to the verification procedures described above.

All new drill holes are entered into the DiGS software and the validation process is now the standard practice for all new drillholes.

Chip sample data was not verified, as it does not appear to be representative of the reef material based on work done by Dr. M. Harley of SRK. It appears that there is a 30% relative bias with respect to the drillhole assays. This bias is thought to be due to mechanical nature of the sampling process using a hammer and chisel. The softer sulphide material is taken preferentially over the harder quartz pebble material. Staff at the South Deep Mine consider that chip samples provide a general snapshot view of the mined grade at any specific time, however, they are not suitable for resource estimation.

The drillhole database also contains assays that were done using an Aztec Analyzer XRF machine. It appears there is a 10% negative bias between the fire assays and Aztec analysis based on limited test work. That is the Aztec assays are lower than the fire assays by 10%. Since 2002 both drillhole and chip samples have been analyzed using fire assay only, and only those older drillhole and chip sample data may have been affected by this underestimation of grade.

8 Mineral Resources and Reserves

8.1 Introduction

With the completion of South Deep Mineral Resource and Reserve evaluation in January 2006, under the guidance of the Independent Review Panel (IRP) comprising experts from various fields, Western Areas decided to evaluate their Mineral Resource and Reserve holdings using a similar methodology. This report outlines the methodology utilized in estimating the Mineral Resources and Reserves within the Kalbasfontein Project Area to the east and south of South Deep Gold Mining Authorisation, and the WA4 Project Area, to the west. The methodology and logic applied is identical to that applied to achieve South Deep 2005-year end Mineral Resource and Reserve estimates for statutory reporting.

8.2 Datamine Modelling

12.10(a)

The 6 main reef units (VCR, MB, MA, ED, EC, EC9) were further discretized into 11 units (greater detail) and contours denoting high potential, moderate potential and low potential were constructed for South Deep extending from the Current Mine in the north to the southern boundary of the Mining Authorisation. Muntingh (IRP) proposed that these geozones, along with the corridors (as used in the determination of mining reserves) should be used to determine the appropriate footprint within South Deep Current Mine in order to assign values to South Deep Future Mine, which was similarly, adopted the Kalbasfontein Project Area. depicts a west-east section through South Deep Current Mine model coloured per reef type.

The geozone strings generated by the IRP, in conjunction with the South Deep Geological Department were imported into Datamine/Guide™ and added to the mining corridors and finally to the project areas as supplied by Western Areas.

Figure 8.1: West- East Section of the Upper Elsburgs

Mr Dexter Ferreira, supplied models for both South Deep Current Mine and Future Mine containing Corridor, Reef and Geozone fields. It is beyond the scope of this document to discuss the derivation of these models and the reader is referred to the appropriate Resource Estimation Document (D.S. Ferreira, Placer Dome - Western Areas Joint Venture Resource Model Sept 2006 titled “A report on the methodology employed in the estimation of the mineral resource model within the mine lease area, January 2006”). After studying the model and discussing the model with Mr. Ferreira, SRK is convinced of the soundness of such model and agrees with its application to the areas under consideration.

Once completed, the entire geology model was verified in Datamine/Guide™ against the appropriate tagged drillholes. The model was also delivered to the South Deep Geology Department in order to allow a final check by site geologists. After verification, the geozones, as drafted by Muntingh (IRP), were exported from Surfer™ and imported into Datamine/Guide™, and wireframes were built. These wireframes were built large enough to encompass the entire South Deep Current Mine blockmodel, as well as the Kalbasfontein Project Area. The corridors were then added as wireframes and the entire estimated blockmodel for South Deep Current Mine was parsed into individual groups, evaluated within Datamine/Guide™ using MODRES, and exported into an Excel spreadsheet for further analysis.

Tonnage grade distributions were then derived for each modelled geozone, for each mining corridor for the following reef sub-units: MBT, MBB, MIB, MA, ECT and ECB. Reefs that had no geozone were used such that the entire reef unit within South Deep Current Mine was used as a footprint for Future Mine. An identical approach was utilized in the determination of the mineral resources at Kalbasfontein Project Area. The tonnage-grade distribution for these blocks was used to estimate the volume contained within corridor #1, MBB, and low potential. This was carried out for every defined geozone.

The blocks and their values were exported into an Excel spreadsheet and sorted on estimated gold grade. A grade-tonnage curve was then generated. Volumes, which had previously been determined on a per corridor, per reef and per geozone potential zone basis, were then assigned the appropriate grade distribution. The resultant Bayesian estimate therefore utilizes actual gold grade estimates, as derived from surface and underground drilling information, to populate known volumes. Once all the footprints were collated and assigned, there were 5 separate cases where there were no equivalent footprints within Current Mine that could be assigned to South Deep Future Mine. These were given to Muntingh who in cooperation with Paul Obermeyer (South Deep geologist), recommended alternative footprints. These are shown below:

·	Corridor 1 MBT Moderate should be comparable with Corridor 1 MBB Moderate;

·	Corridor 2 MBT High should be comparable with Corridor 2 MBB High;

·	Corridor 2 MBT Moderate should be comparable with Corridor 2 MBB Moderate;

·	Corridor 1 MIB Moderate should be comparable with Corridor 2 MA Moderate;

·	Corridor MIB Moderate should be comparable with. Corridor 3 MA Moderate.

With the Bayesian methodology, the surface drillhole data that exists both within the South Deep Mining Authorisation and inside the Kalbasfontein Project Area are only used to model the individual reef units (given their pierce points) and not for grade estimation.

8.3 Mineral Resource Estimate Results for the Upper Elsburgs

Once the alternative footprints were assigned, the Bayesian estimate was finalized. Ultimately, 4 geozone types were recognized: high potential, moderate potential, low potential and undefined geozones. This latter category refers to reef units where Muntingh did not define a geozone (i.e. VCR, ED, EC9 and MAD). Volumes were determined using Datamine/Guide™, and the appropriate representative footprint used to assign a grade-tonnage distribution. The final mineral resource estimates for the Upper Elsburgs in the Kalbasfontein Project Area is shown in Table 8.1.

12.10(b)(ii)

Table 8.1: Tonnage/Grade Distribution (High, Moderate and Low Geozones)

Cutoff (Au g/t)	Tonnes (Mt)	Au (g/t)	MOz
0.00	436.4	2.60	36.5
1.00	284.7	3.69	33.8
2.00	181.3	4.96	28.9
3.00	121.4	6.21	24.2
4.00*	85.3	7.36	20.2
5.00	59.8	8.59	16.5
6.00	43.1	9.81	13.6
7.00	32.3	10.92	11.4
8.00	24.9	11.94	9.6
9.00	16.9	13.57	7.4
10.00	12.8	14.88	6.1
*Indicated mineral resources

12.10(b)(vi)(4)

SRK recommends that a cut-off grade of 4.0g/t be taken to determine the Indicated Mineral Resources on the basis that this represents the lowest mine design cut-off (longhole stoping) (see section Figure 9.3).

8.4 Methodology – Ventersdorp Contact Reef (“VCR”)

12.10(b)(iii&iv)

The area held by Western Areas, west of the South Deep Mining Authorisation, is known as the WA4 Project Area, and is targeting the VCR. The structure of the VCR has been delineated by the recent 3D seismic survey in the area. The refined VCR surface was translated into 3D strings and individual polygons were constructed, which were trimmed to accommodate the South Deep Mining Authorisation boundary (Figure 8.2).

Figure 8.2: VCR Polygons within Western Areas WA4 Project Area (Not to scale)

The estimation of the VCR, west of the subcrop, was completed using a Bayesian approach as well. The geological polygons were defined, as above, and verified by the site geologists at South Deep. The area constrained by these polygons (excluding fault losses) were tallied using Datamine/Guide™ and multiplied by the average channel width (CW) observed for the VCR within South Deep Current Mine (west of subcrop) of 1.46m. Thereafter, the estimated grades for the unvoided (no mining activity) blockmodel for VCR in South Deep Current Mine were assigned proportionately to this tonnage value. Once complete, the entire model was discounted in order to account for possible overbank material within the area (24.1% in terms of tonnage at 0.0g/t gold cut-off grade).

The figure shows a horizontal line depicting the approximate location of 110 level; the bottom-most level accessible by the Twin Shaft system. The individual VCR reef polygons created were limited by the recent 3D seismic surface along the western and southern flanks.

8.5 Mineral Resource Estimate - VCR

12.10(b)(ii)

The VCR Mineral Resource, estimate at various cut-off grades, is shown in Table 8.2 below. The results are split showing Mineral Resources above and below infrastructure.

Table 8.2: Tonnage/Grade Distribution at Various Grade Cut offs - VCR

Cutoff Au g/t	Above Infrastructure			Below Infrastructure			Total
Cutoff Au g/t	Tonnes (Mt)	Au g/t	MOz	Tonnes (Mt)	Au g/t	MOz	Tonnes (Mt)	Au g/t	MOz
0.00	7.88	12.58	3.1875	4.89	12.58	1.9790	12.77	12.58	5.1665
1.00	7.88	12.58	3.1875	4.89	12.58	1.9790	12.77	12.58	5.1665
2.00	7.87	12.59	3.1870	4.89	12.59	1.9788	12.76	12.59	5.1658
3.00	7.86	12.60	3.1863	4.88	12.60	1.9783	12.75	12.60	5.1646
4.00	7.75	12.73	3.1727	4.81	12.73	1.9699	12.56	12.73	5.1425
5.00	7.35	13.17	3.1145	4.57	13.17	1.9337	11.92	13.17	5.0482
6.00	6.89	13.69	3.0323	4.28	13.69	1.8827	11.16	13.69	4.9151
7.00*	6.38	14.27	2.9260	3.96	14.27	1.8167	10.34	14.27	4.7426
8.00	5.96	14.74	2.8255	3.70	14.74	1.7543	9.67	14.74	4.5798
9.00	5.44	15.33	2.6834	3.38	15.33	1.6661	8.82	15.33	4.3494
10.00	4.96	15.90	2.5357	3.08	15.90	1.5744	8.04	15.90	4.1101
* Inferred Mineral Resources

Given the lack of underground and surface drillhole data, the resources for the VCR within the WA4 Project Area have been classified as an Inferred Mineral Resource. SRK recommends that the Inferred Mineral Resource be taken at a cut-off of 7g/t on the basis that this represents the mine design cut-off for conventional mining. (see Section 8.6).

8.6 Cut-off Grades

The cut-off grades used for determining Mineral Reserves were estimated by South Deep mining engineers using historical costs, direct and indirect, sustainable capital costs and modifying factors.

12.10(b)(vi)(4)

The average annual direct costs, as determined in the South Deep Strategic Business Plan (SBP) were used. More specifically, these were the projected direct costs for the period of 2006 to 2029 for the planned production profile of 330ktpm. Initially, the costs were subdivided in accordance with the various mining methods: trackless (drift and fill, benching, and longhole) and conventional. The costs were then averaged over both a short term, 5-year period (2006 to 2010), and a long term, 19-year period (2011 to 2029).

Indirect costs of R205.50/t were then added to these direct costs. These costs represent production and mine overheads.

The only capital included in the derivation of the cut-off grades was sustainable capital at 5% of the total working cost, inclusive of indirect and overhead costs.

The gold price used is USD550/oz and the Rand-Dollar exchange was R7.18 to the US Dollar.

Two final parameters were applied in the determination of cut-off grades, the mine call factor (MCF) and the plant recovery factor (PRF). The MCF applied was 90% for the VCR and 95% for trackless mining. The MCF values are based on the historical 60-month progressive average values at South Deep. The PRF value of 97.2% applied was based on the 21-month progressive average values at South Deep.

Based on these costs and parameters, the following cut-off grades were determined and subsequently used for the reserving process:

·	Conventional Mining: 7.4g/t

·	Trackless Mining: 4.9g/t

·	Long hole Stoping: 3.9g/t

8.7 Reserving Parameters

12.10(b)(vi)(5)

Other parameters used in the mine design process, such as dilution and losses per specific mine design, are listed in Table 8.3.

Table 8.3: Reserving Parameters: Dilution & Losses

Mining Method	Volume	Grade	Material Loss	TYPE
Ramping	5.2%	0.0	1.8%	TRACKLESS
Drifting	8.3%	0.0	1.6%
Longhole	3.8%	0.0	1.5%
Conventional	7.9%	2.0	0.0%	CONVENTIONAL
DD	7.9%	2.0	0.0%
Destress	7.9%	2.0	0.0%
Skin Pillars	7.9%	2.0	50.0%

9 Mining

9.1 Mine Design

The area covered by the South Deep Current Mine model was redesigned according to South Deep Future Mine parameters, identical to those used for the Kalbasfontein Project Area, since the existing Current Mine was not suitable as it is dependent on existing infrastructure and pillar placements. Table 9.1 describes the mining method applied to each corridor. The mining corridors and the relevant Kalbasfontein mining projects are shown in Figure 9.1, along with the geozones defined for the MBB reef unit.

The design philosophy for each mining method will be elaborated on further in this document.

Table 9.1: Mining Method per Corridor

12.10(a)(xvi)

Corridor	Mining Method
1	Longhole stoping, Drifts & Drives
2	Longhole stoping, Drifts & Drives
3	Drifting & Benching

The Kalbasfontein Project is situated east of the current South Deep Mining Authorisation and is accessible from the South Deep 95 to 110 level (Phase I) infrastructure. Accordingly, the reserving process followed was similar to that adopted by South Deep, for its 2005 reserve estimate.

Figure 9.1: Kalbasfontein Project Area, Mining Projects and Corridors

Figure 9.1 shows the high potential, moderate potential and low potential geozones for the MBB reef package. This is the main target area for trackless mining as defined in the reserving process for the Kalbasfontein Project Area.

The following sections will detail the design criteria that were utilized for the determination of reserves within the Kalbasfontein Project Area.

9.2 Conventional Stoping Design

9.2.1 Destress Stope

To enable trackless mining in an area, the area must be destressed first. The destress stope is mined conventionally. The maximum efficiency envelope of a destress horizon is in the range of 50m to 60m above and below the destress cut. Therefore once the orebody reaches and exceeds this limit, the destress horizon position is adjusted to optimise and maximize the amount of destressed resources available. The design of the destress horizon was done by viewing the rotated block model in a YZ direction (NNE-SSW), in conjunction with the pillar triangulations, void triangulations and the geological structure. Lines were digitized in Vulcan every 20m for each of the mining projects applying the maximum destress efficiency considerations as discussed previously. This 20m is defined by the block model cell size, which is 20m in the X direction, 20m in the Y direction and 1m in the Z direction. Each of the projects’ polylines was projected 7.5m forward and 7.5m backward and converted into a triangulated surface. The projection of these lines produced a better-fit triangulated surface, minimizing the edge effect of the block model, therefore preventing under-evaluation. Each of the triangulated surfaces were projected 1.7m downwards, perpendicular to the dip of the destress horizon (bearing of 30°), to account for the dip and strike of the orebody. Closing the surfaces along their edges produced a solid.

9.2.2 Skin Pillars

The skin pillar lies directly below the destress stope and is designed at a 4m vertical thickness. Once mining in the area is complete the bottom 2m can be mined at a 50% extraction rate of the material above the conventional cut-off grade. The destress stope footwall was projected vertically downwards by 2.0m and 4.0m to produce the hangingwall and footwall surfaces of the top 2m pillar, of the 4m skin pillar zone. Their open edges were closed to produce two individual solids representing the two components of the skin pillar. Where the destress stopes hole into one another the skin pillars were cut against the voids to prevent volume duplication.

9.2.3 Conventional Stoping

This design category includes all Upper Elsburg blocks above the conventional cut-off grade (7.4 g/t) that are not incorporated in any other designs.

Digitising polygons in a YZ orientation from the block model outlined these areas. A minimum stoping width of 1.5m was assumed, therefore if only one cell existed vertically then a cell could only be selected if the grade of the stope envelope equalled or exceeded 7.4 g/t. At minimum selection of 2 cells in a NNE – SSW orientation formed part of the envelope definition. These polygons were projected 7.5m forward and 7.5m backwards and converted into triangulated solids as per the destress stope. The projection of these polygons produced a better-fit triangulated solid, minimising the edge effect of the block model, therefore preventing under evaluation.

A minimum pillar of 4m is required between existing voids or planning and the hangingwall of planned conventional stopes. Table 9.2 below shows the design standards applied to conventional mine design.

Table 9.2: Conventional Design Standards

Destress
Description	Measure
Stoping width	1.7m
Face length	30m
Gully & Raise dimensions (portion in stope footwall)	1m(deep) x 1.6m (w)
Conventional Drifts & Drives
Description	Measure
Drift height	1.5m
Drift width	1.8m
Drift length	30m
Drive height	Total drift height + 1m
Drive width	1.8m
Drive length	Determined by project size

9.3 Trackless Stoping Design

Three types of trackless mine design were employed in the planning of South Deep Current Mine, which had been applied to the Kalbasfontein Project Area, namely: drift and fill with benching, modified drift and fill with benching and long hole stoping.

9.3.1 Drift and Fill with Benching

Common to the drift and fill and the modified drift and fill is the selection process. In both cases, stepping through the block model in an YZ orientation at 20m intervals and outlining stacks of cells that exceeded a thickness of 5m with an overall grade of 4.9g/t identified target areas within each project. These target polygons were translated to the edges of the cell (±10m) and triangulated. The resultant triangulations are independent of existent voids, pillars and destress stopes and represent the maximum design potential of the model. The designer was then able to use the triangulations rather than a view of the block model in the design process, a more efficient design technique.

Access ramps were designed from existing infrastructure, wherever possible, to provide an attacking point to the ore pocket. The drifts and benches (if any) were evaluated along with the access ramps to ensure that the planned area remained payable.

9.3.2 Modified Drift and Fill with Benching

This method utilizes a layered approach with the design of drifts and benches. It is seen as a more optimized mining method that avoids ramping inside the high-grade target areas, therefore minimizing secondary ramp pillars (reserves that can only be mined at low efficiencies). The designing of the drifts and/or benches was done in a 2-step process. Firstly, the payable selections were made as per the drift and fill method described above. The second stage was to step through the newly created triangulations in 5.5m plan view (XY) sections. New polygons were designed to ensure no overlap with designed pillars, destress stopes, and destress stope skin pillars.

The resulting polygons were used to create triangulated solids by projecting them 5.5m downwards from the top of the polygon. The design of sub-benches is basically footwall lifting in areas below drifts where selections could be made without overlapping any other designs at a vertical thickness of 2.75m. This process lessens the potential of underestimating by minimising the edge effect that the 5,5m steps can create. The selection of sub-benches occurred after the main drifts and benches were designed.

The ramps and access ramps for the modified drift and fill with benches were designed according to existing mine design parameters, following on from existing development. Table 9.3 shows the design standards applied to this trackless mine design as shown in Figure 9.2.

Table 9.3: Modified Drift and Fill with Benching Design Standards

Modified drift & fill with benching
Description	Measure
Drift height	5.5m
Drift width	6.0m
Drift length	10-80m
Ramp height	5.5m
Ramp width	5.5m
Bench height	5.5m
Bench width	6.0m
Bench length	<60m (remote loading)

Figure 9.2: Modified Drift and Fill with Benching

9.3.3 Longhole Stoping Design

South Deep Current Mine area was divided into 4 corridors (running parallel to the Upper Elsburgs). These corridors were defined irrespective of existing infrastructure and mining; Figure 9.8 reveals that corridors 1, 2 and 3 intersect the Kalbasfontein Project Area. The design of the longhole stoping was completed in each of corridors 1 and 2 within the Kalbasfontein Project Area. All designs were undertaken in the South Deep Current Mine blockmodel, and utilized a cut off grade of 3.9g/t for longhole stoping.

Once the corridor outline limits had been set, the destress design was completed so as to destress the entire mining area of the corridor. Destress stoping was designed at 1.7m stoping width, positioned above the ED reef package. The designed area requiring destressing is then cut from the surface triangulation so that a surface area is determined for the ground to be destressed. This surface triangulation is then translated 1.7metres so that a solid can be created of the void that will be mined. This triangulation is then evaluated in the block model.

The main access ramps to the project areas are then designed in the corridor. These are the down dip access ramps from the level above to open up the ground earlier for mining purposes (i.e. access to ground that is available due to the lag time in capital development for the level). The ramps have the dimensions of 5.5 x 5.5m and 4.5 x 4.5m. Twin ramps are designed in the EC package (below the destress horizon). These ramps are designed at a maximum gradient of 8 degrees to comply with equipment parameters and optimal productivities of the equipment that will operation in the ramps. Cubbies have been designed in each of the ramp systems to ensure drawpoints and travel way / storage areas are available. These cubbies are placed at 25 metre intervals (and have the same height as the ramp from which they are developed but are developed at a width of 5.5m). The twin ramps sit below the destress stoping.

The design process focuses on one priority, the delineation of longhole stoping reserves. The blockmodel (utilizing a cut off grade of 3.9g/t), and destress stoping is loaded on the screen and changed into a XY view (section perpendicular to dip of orebody). In the section view, all blocks from the blockmodel are outlined on 20-metre intervals. A minimum vertical height for this selection / outlining of longhole areas is 15 metres. The trackless outline selections

are then triangulated to create a solid triangulation. This represents the initial selection of the trackless areas. Once completed, detailed stoping is designed (based on the initial selections). The selection triangulation is sectioned on 15 metre intervals to represent the centerlines of each individual stope. Each stope is then individually assessed for practical mining capability and the achievement of cut-off parameters. To ensure access to each of the longhole stopes, the “in stope” hangingwall and footwall ramps are designed. These ramps sit within the designed stoping area. Two hangingwall and footwall ramps have been designed to enable adequate material and equipment movement and to satisfy all ventilation requirements.

To satisfy all geotechnical requirements, a 15-metre pillar is maintained around these instope ramps. These will be removed at the end of the project life (on a retreat basis). The accesses to each of the individual longhole stopes are then designed. Once all longhole mining has been designed, secondary trackless mining is then identified and outlined, producing drifting areas reporting to the mineral reserves estimate. Secondary conventional mining is now addressed. All triangulations of designed trackless and destress stoping are loaded on the screen and using the conventional mining cut-off, areas are identified for conventional mining.

The final stage of the reserving process is the determination of the reserves that can be extracted from the skin pillars located below the 1.7m destress stope. Two skin pillars are designed directly below the destress stope for geotechnical purposes. Skin 1, sits directly below the destress stope. Skin 1 is designed at 2.0m thicknesses and is not converted to reserves, as it will not be mined at any stage. Skin 2 is designed directly below Skin 1 (2m below bottom of destress stope). This stope is evaluated at a cut-off of 7.4g/t and only 50% of this evaluated material is converted to reserves. Table 9.4 shows the design standards applied to this trackless mine design and is presented in Figure 9.3.

Table 9.4: Longhole Stoping Excavation Dimensions

Longhole stoping
Description	Measure
Stope access drift height	4.5m
Stope access drift width	4.5m
Stope access drift length	78m
Ramp height	4.5m
Ramp width	4.5m
Stope height	17m – 35m
Stope width	15m
Stope length	60m

Figure 9.3: Trackless Longhole Stoping

9.4 Project Areas

The moderate potential MBB geozone was used as the driver in defining projects in the Kalbasfontein Project Area as shown in Figure 9.4. The projects in red depict South Deep Phase1, green is South Deep Phase 2 and magenta for projects in the Kalbasfontein Project Area.

Figure 9.4: Project Areas in Kalbasfontein Area

The regional pillar design criteria applied in Kalbasfontein Project Area was the same as was applied in South Deep Phase I. This includes placing a continuous pillar along the large wrench fault that is the main structural entity in the area. This is shown in Figure 9.5 along with a three dimensional view as an inset.

Figure 9.5: Pillar Layout in Kalbasfontein Project Area

The pillar orientation changed in the Kalbasfontein Project Area due to the change in dip within that area; from east to slightly west. The regional pillars were therefore oriented to follow the structure. The infrastructure design applied in Kalbasfontein Project Area is similar to the design philosophy adopted for the Massives for South Deep Phase I. The area is accessed by extending the haulages designed in Phase I (100-110 Levels) with a dual return airway (RAW) leading back to 95 Level. The proposed development for accessing the Kalbasfontein Project Area is shown in Figure 9.6 whilst in Figure 9.7 shows the wrench fault along with its measured throws in three dimensions.

Figure 9.6: Development Accessing the Kalbasfontein Project Area

Figure 9.7: Three Dimensional View of the Wrench Fault in the Kalbasfontein Project Area

The mine design for the Kalbasfontein Project Area is based on a modular approach with extensions on South Deep current haulage designs on 100, 105 and 110 levels. A dual haulage system on 95 level will allow access to destressing, and once mechanized mining is progress on the lower levels, this access will act a return airway to be linked to 94 level.

The schedule has been arranged so that a throughput of 330,000tpm can be achieved by 2011, which is the scheduled production start date of 95 level ore from the Kalbasfontein Project Area. Most of this early ore is to come from destressing on 95 and ramp accesses (for the longhole stopes) within ore. Longhole ore is scheduled to begin in 2017 and to peak to over 2.2mtpa but decline until depleted in 2024. The waning of ore availability from longhole stoping, beginning in 2020, is to be supplemented by ore from drifts and drives.

Destressing on 100 level is to begin 2016 with mine production from longhole stoping beginning in 2021. Ore from drifts and drives will supplement this material until 2034. Ore from 105 level will begin in 2021 with destressing starting two years earlier. The longhole stopes will be exhausted by 2030 leaving an additional year of ore being mined from drifts and drives.

In-ore development on 110 level will begin 2024 with longhole stopes beginning in 2031. Production from these stopes is scheduled to peak at 1.9mt/yr and running until its final production year of 2041. Ore will be continuously mined from drifts and drives until 2045 with a final production year of 2046 from material mined conventionally.

The idea behind the Bayesian approach is that of a priori; what is known about something can be applied to what is assumed to be similar. Therefore, the estimated gold distribution in South Deep Current Mine can be applied to South Deep Future Mine, and the Kalbasfontein Project Area immediately adjacent to South Deep, providing a consistent geological framework between the both regions. In this regard, the corridor, geozone and reef fields common to both regions could be used to allocate grade into the estimated areas in Kalbasfontein. In order to facilitate this transference, a new field was created from a concatenation of the corridor, geozone and reef fields. For example, a VCR model cell occurring in the moderate geozone of corridor 1 would be coded 1ModerateVCR.

9.5 Bayesian Methodology

The mine designs for the Kalbasfontein Project Area were evaluated against the combined geology / resource model. Figure 9.8 shows the Bayesian assignment of mine designs (drifting and longhole) to the Kalbasfontein Project Area within the MBB geozones within corridors 1,2 and 3. The designs were classified according to a general mining method and a specific mining method. Three general mining methods were defined: trackless, conventional and destress.

Figure 9.8: Application of Mining Designs to the Kalbasfontein Project Area

The destress cut was separated from conventional mining due to grade considerations. Specific mining methods describe each aspect of the general mining method. Thus, the trackless general method, for example, comprises of ramping, drifting, benching, sub-benching, longhole stoping and longhole access.

The colour-coding shown in Figure 9.8 reveals that reserving figures that resulted in corridor 1, from longhole stoping (and all associated mining activity, i.e. ramping, raising etc.), within the moderate potential geozone of the MBB, were assigned to the volumes within the Kalbasfontein Project Area within the moderate potential geozone of the MBB (magenta colour) of corridor 1.

Longhole stoping reserved within the moderate potential geozone of the MBB (red colour) was assigned to the volume obtained of the same geozone within the Kalbasfontein Project Area within corridor 2.

Reserves from drifts and drives mining activity, designed within corridors 2, were assigned to the volume obtained within corridors 2 within the Kalbasfontein Project Area (green colour).

Table 9.5 shows the various mining methods planned for the MBB reef of the high potential geozone in Corridor 1. The evaluation of the mine design triangulations was tabulated to report volume and grade per concatenation and mining method. A resource to insitu reserve conversion ratio was calculated for each category. There, a reserve extracted from resources, in terms of percentage converted, on a per mining method, per mining activity, per corridor, per geozone and per reef unit was obtained. This percentage was then assigned to a similarly tagged resource within the area to be evaluated. This Bayesian methodology was applied to South Deep Phase 2 and identically to the Kalbasfontein Project Area.

Table 9.5: Mining Methods per Concatenation

Concatenation	General Method	Specific Method	Corridor	Geozone	Reef
1HighMBB	Conventional	Conventional	Corridor 1	High	MBB
1HighMBB	Conventional	Destress	Corridor 1	High	MBB
1HighMBB	Trackless	Drifting	Corridor 1	High	MBB
1HighMBB	Trackless	LH Access	Corridor 1	High	MBB
1HighMBB	Trackless	LH Stope	Corridor 1	High	MBB
1HighMBB	Trackless	Ramping	Corridor 1	High	MBB

9.6 Application to Future Mining

Before the conversion ratios could be applied to the Kalbasfontein Project Area, the updated pillar design had to be taken into account so that only mineable ground would be converted to reserves. Cells in the Kalbasfontein model were flagged as either part of a pillar or not. Figure 9.9 shows a west-east section through the Future Mine model with the pillar outline in black and the pillar cells in green. The mineable ground is shown in blue.

Figure 9.9: West – East Section through Future Mine Showing Regional Pillars

A three-dimensional view of the pillar layout in the Kalbasfontein Project Area is shown in Figure 9.10.

Figure 9.10: Oblique View of the Kalbasfontein Project Area and South Deep Pillars

A summary of the reserves locked in pillars within the Kalbasfontein Project Area is shown in Table 9.6, on a per mining method basis; these ”locked” reserves are not included in the final reserve tabulation for the Kalbasfontein Project Area.

Table 9.6: Probable Mineral Reserves in the Pillars

Mining Method	Tonnes (Mt)	Grade (g/t)	MOz
Conventional	0.40	10.55	0.14
Destress	1.18	2.18	0.08
Trackless	23.19	4.97	3.71
TOTAL	24.77	4.93	3.92

In addition to the pillars, the reserving model was also labelled with a project identifier to allow for detailed reporting. Due to the nature of the ore body, and the geometry of some geozones, some the Kalbasfontein Project Area and South Deep Phase 2 concatenations had no South Deep Current Mine equivalents. If these were not accounted for, reserves for the Kalbasfontein Project Area and South Deep Phase 2 would be lost. P. Obermeyer (South Deep Geology) and D. Muntingh (IRP) hence proposed the equivalent South Deep Current Mine concatenations listed in Table 9.6 that could be used as a footprint to define concatenations within the Kalbasfontein Project Area and South Deep Phase 2.

Table 9.7: Kalbasfontein Concatenations – South Deep Current Mine Equivalents

Kalbasfontein Project Area	South Deep Current Mine Equivalent
Corridor 1 Moderate MBT (1ModerateMBT)	Corridor 1 Moderate MBB (1ModerateMBB)
Corridor 2 High MBT (2HighMBT)	Corridor 2 High MBB (2HighMBB)
Corridor 2 Moderate MBT (2ModerateMBT)	Corridor 2 Moderate MBB (2ModerateMBB)
Corridor 1 Moderate MIB (1ModerateMIB)	Corridor 2 Moderate MA (2ModerateMA)
Corridor 2 Moderate MIB (2ModerateMIB)	Corridor 3 Moderate MA (3ModerateMA)

With these concatenations in place, the Bayesian exercise was completed for all mining project areas within the Kalbasfontein Project Area.

9.7 Reserve Results - Upper Elsberg

The application of a Bayesian approach, assigning a grade range to the appropriate volume modelled for the Kalbasfontein Project Area and assigning reserving parameters, and assuming that the Kalbasfontein Project Area will be included in South Deep Mining Authorisation, has resulted in the mineral reserve estimate shown in Table 9.8.

Table 9.8: Probable Mineral Reserves per Mining Method

Mining Method	Tonnes (Mt)	Grade (g/t)	MOz
Conventional	1.11	10.26	0.37
Destress	1.70	2.28	0.12
Trackless	55.12	4.93	8.73
TOTAL	57.93	4.95	9.22

Given that the estimated Mineral Resources have been classified as being indicated, these mineral reserves have been classified as Probable Mineral Reserves. The Reserves can be further broken down onto mining activity types for both the high and moderate geozone areas as shown in Table 9.9 and per mining project area in Table 9.10.

Table 9.9: Probable Mineral Reserves by mining method and project area

Mining Project	Conventional Mining			Destress Mining			Trackless Mining
Mining Project	Tonnes (kt)	Au g/t	kOz	Tonnes (kt)	Au g/t	kOz	Tonnes (kt)	Au g/t	kOz
95 P1	184.62	10.39	61.70	190.51	2.02	12.375	7 765.17	4.75	1 186.25
95 P2	151.01	10.53	51.11	116.71	2.24	8.422	4 651.22	4.37	652.74
95 P3	98.25	10.56	33.34	69.97	2.36	5.313	2 995.37	4.38	421.75
95 P4	33.39	10.57	11.35	25.41	2.46	2.006	1 011.98	4.39	142.76
100 P1	29.81	9.41	9.02	130.40	2.36	9.891	4 306.65	5.62	777.81
100 P2	68.02	10.65	23.28	66.01	2.24	4.754	2 061.18	4.31	285.84
100 P3	87.07	10.61	29.69	99.58	2.28	7.288	2 656.77	4.32	369.21
100 P4	77.38	10.52	26.18	72.93	2.38	5.591	2 375.27	4.38	334.58
105 P1	14.35	9.36	4.32	62.58	2.96	5.949	1 642.47	5.37	283.41
105 P2	65.31	10.42	21.89	133.69	2.66	11.417	2 163.97	4.19	291.20
105 P3	52.90	10.52	17.89	81.12	2.26	5.901	1 652.11	4.32	229.45
110 P1	113.68	10.06	36.78	167.93	2.00	10.778	7 320.01	5.11	1 203.60
110 P2	91.74	9.23	27.23	167.27	1.77	9.499	8 816.74	5.40	1 531.01
110 P3	26.54	8.90	7.60	180.62	2.51	14.591	3 439.88	5.56	614.96
110 P4	15.05	9.22	4.46	135.86	2.54	11.094	2 265.07	5.58	406.66
Total	1 109.11	10.26	365.84	1 700.59	2.28	124.869	55 123.86	4.93	8 731.22
Note: A MCF of 95% has been applied therefore the quoted grade is a head grade.

Table 9.10: Total Probable Mineral Reserves by Mining Project Area

Mining Project	Total Mining
Mining Project	Tonnes (Mt)	Au g/t	MOz
95 P1	8.14	4.82	1.260
95 P2	4.92	4.50	0.712
95 P3	3.16	4.53	0.460
95 P4	1.07	4.53	0.156
100 P1	4.47	5.55	0.797
100 P2	2.20	4.45	0.314
100 P3	2.84	4.44	0.406
100 P4	2.53	4.51	0.366
105 P1	1.72	5.31	0.294
105 P2	2.36	4.27	0.324
105 P3	1.79	4.41	0.253
110 P1	7.60	5.12	1.251
110 P2	9.08	5.37	1.568
110 P3	3.65	5.43	0.637
110 P4	2.42	5.44	0.422
Total	57.93	4.95	9.222

A summary of the reserves within the Kalbasfontein Project Area, per mining method is shown in Table 9.8.

At a cut-off of 4.0g/t, the conversion of Mineral Resources to Mineral Reserves is in the order of 45%. This conversion percentage is similar to that for South Deep of 44%.

10 Metallurgical

The metallurgical process has been well established and developed over an operating period in excess of thirty years. There were two processing facilities at South Deep Mine, one near to the South Main shaft (“South Plant”) and the other close to the South Deep shaft (“South Deep Plant”). The older South Plant utilized a Merrill-Crowe circuit; while the newer South Deep Plant utilizes a carbon-in-pulp (“CIP”) circuit.

The South Deep Plant was commissioned in June 2002 and will handle the production for the current life of mine schedule. Plant capacity is 7,233 tonnes per day (“tpd”). The metallurgical recovery has consistently been between 96.5% and 97.5% since 2002. Future recoveries are forecast to be in the same range with a figure of 97.2% used in the LOM plans.

The new South Deep Plant utilizes standard gold processing procedures. It utilizes a crushing plant and a semi-autogenous mill for size reduction and incorporates a gravity circuit. The slurry is thickened ahead of cyanidation and carbon-in-pulp processing. Gold from the carbon handling circuit is sludged in five electro-winning (EW) cells. Separate doré bars are produced from the EW sludge and from the gravity-recovered gold. The leached pulp from the CIP circuit is sent to the backfill plant where a coarse fraction is cycloned off and sent underground. The remaining fine fraction is pumped to the tailing ponds.

Minor modifications are being made to fine-tune the South Deep Plant. The South Deep Plant includes additional space to allow for expanding the required sections of the plant to ultimately treat 10,520 tpd. This will include, but not be limited to:

·	Recycle crusher and return conveyor belt;

·	Additional ball mill;

·	Primary gravity unit;

·	Three additional leach tanks.

A new backfill plant will eventually replace the current plant which is located at the South Shaft complex.

Water from the metallurgical process runs through a series of cascade dams for cyanide destruction. All decant solution is returned to the plant. Ferrous-sulphate is added to backfill to complex cyanide.

There are currently two slimes dams in operation. The “old” slimes dam has been in use since 1968. The “new” dam has been operational since 1982. The dams are nearing the end of their lives. The mine is in the process of securing land for constructing a new slimes dam. The Environmental Impact Assessment (“EIA”) for a new tailings dam is nearing completion and it is expected to be submitted shortly.

For budgeting purposes and for long range forecasting in the Life of Mine Plan, gold recovery has been fixed at 97.2%, which compares to an actual recovery of 97.3% in 2005.

Metallurgical recoveries were estimated based on historic and forecasted values.

Metallurgical processes have been well established and developed over an operating period in excess of thirty years. The South Deep Plant utilizes a carbon-in-pulp (“CIP”) circuit. The South Deep Plant was commissioned in June 2002 and will handle the production for the current life of mine schedule. Figure 10.1 shows a simplified flow sheet for the new South Deep Plant.

10.1 South Deep Processing Plant Flow Sheet

The ore processing facilities at the new South Deep Plant include crushing, milling, gravity concentration, thickening, CIP leaching and tailing disposal. A coarse tailing product is recovered for use as underground backfill and the remaining tailing is sent to the tailing ponds. The plant was commissioned in the second quarter of 2002 and has a treatment capacity of 7,233 tpd or 2,640,000 t/a. With the installation of a recycle crusher, second ball mill and extra leach tanks, the plant will be capable of treating 10,879 tpd or 3,960,000 t/a. (Refer to figure 10.1).

Ore from the new South Deep Main shaft is crushed underground to <300 mm, hoisted to surface and conveyed from a bin in the head frame to a 22,000 t open stockpile. Any ore hoisted up the old South Main shaft is trucked to the open stockpile. Ore from the stockpile discharge chute is conveyed to the mill as feed to the SAB (SAG/Ball) milling circuit. The SAG mill circuit consists of a variable speed 5,000 kW SAG Mill (7.9 m diameter by 4.5 m long), discharge and vibrating screens and provision to install a recycle crusher. Mill rejects are fed back onto the SAG mill feed conveyor.

Figure 10.1: Process Plant Flowsheet

Secondary milling is conducted in a fixed-speed closed-circuit trunnion 5,000 kW Overflow Ball Mill (5.5 m diameter by 8.8 m long). Classification of the ball mill discharge is done using a cluster of 12 mounted 420 mm cyclones fed from a dual sump with dedicated variable speed pumping units. The final milled product is 75% passing 74 microns.

The underflow and overflow from the cyclones are handled separately. A portion (20 to 25%) of the cyclone underflow is screened on a mounted vibrating screen with the +2 mm oversize being recycled to the feed of the secondary mill. The remainder of the cyclone underflow portion (80 to 75%) is fed into two XD48 Knelson Concentrators for the recovery of the free gold. The concentrate from the Knelson Concentrators is fed to storage in the smelter house, and then fed to an ILR2000 high intensity cyanidation inline leach reactor. The solid residue from the leach reactor is screened and washed free of residual cyanide before being returned to the ball milling circuit. The leach liquor is passed through a dedicated electro-winning reactor and smelted to produce gold bullion.

The cyclone overflow gravitates to two of three 20 m2 linear screens, with one as a standby unit, for wood chip/trash removal. The wood chips are de-watered on one common vibrating screen and collected in a telecon hopper. The final pulp gravitates into a 30 m diameter high-rate thickener (“HRT”). Flocculent is added to the HRT feed from a flocculent make up system. The overflow water reports as process water for re-use in the circuit. The thickener underflow, at a mean density of 1.56 t/m3 (1.48 to 1.64 t/m3), is pumped through a cross-stream primary sampler and a secondary Vezin sampler to the 3,000 m3, twin mechanically agitated 1:1 aspect, leach tanks. Currently, there are six tanks operating with enough space for four more. Cyanide and lime are added to the first leach tank and the system has facilities to add more cyanide and lime further down the train, if needed. The slurry is leached for a nominal period of thirty-six hours to ensure the dissolution of the gold.

The leach pulp gravitates through a feed box, to the first 200 m3 pump cell of an eight pump cell carousel circuit arrangement. The circuit allows for a single cell to be off line to process the contained pregnant carbon while the others cells remain on line. The contents from the off line cell are pumped over a vibrating screen to separate the loaded carbon from the pulp. The pulp then gravitates back to the CIP feed launder, while the loaded carbon is screened, acid treated and eluted in 16 m3 Hastelloy elution columns to produce a pregnant gold solution and eluted carbon products. The pregnant gold solution is passed through a 24 m3/h electro-winning circuit, with the product back-washed, dried and then smelted in an induction furnace to produce gold bullion assaying 80 to 85% gold. The bullion bars are then transported to the Rand Refinery for final treatment.

The eluted carbon is fed into an electrically fired rotary kiln for regeneration. The regenerated carbon is fed over a vibrating screen to separate the fines from the oversize. The empty, offline absorption cell is re-filled with the oversize and recycled pulp prior to being brought back on line. The newly filled cell becomes the last adsorption stage in the next cycle. Undersized carbon is fed to a fine carbon recovery plant that consists of a conical tank operating as a settler. Settled carbon is discharged into bulk bags for disposal.

The leached pulp is discharged, after passing through the eight pump cells, onto a vibrating screen to remove carbon from the final residue stream. The carbon reports to a carbon recovery tank from where it will be pumped to a carbon settling facility. The final residue is fed to the Backfill Plant by the South Main shaft where Coarse Classified Tailing (“CCT”) backfill is prepared. The Backfill Plant utilizes a cluster of cyclones to deslime the tails producing a coarser underflow product that is pumped to the shaft area for further preparation prior to being sent underground. The cyclone overflow product and any unused tailing is discharged as the final waste product in to the tailing dams.

Make up water for processing comes from the return water from the tailing ponds and from storage tanks fed from underground water discharge lines, as well as surface water drillholes drilled to the aquifer. The gravity circuit is fed from a dedicated water supply tank with facilities to add water from the Rand Water Board (“RWB”) and from underground. Potable water is provided from the South Deep Shaft infrastructure.

The air requirements are supplied from a dedicated 1,000 m3 compressor station. All instrument air is cleaned and dried to the required quality.

11 Tailings Disposal

A coarse tailing product is recovered for use as underground backfill and the remaining tailing is sent to the tailing ponds. The plant was commissioned in the second quarter of 2002 and has a treatment capacity of 7,233 tpd or 2,640,000 t/a. With the installation of a recycle crusher, second ball mill and extra leach tanks, the plant will be capable of treating 10,879 tpd or 3,960,000 t/a.

12 Infrastructure

No change is envisaged to the infrastructure and services of the mine as ore production remains constant hoisted through existing shafts.

The joint venture partners own equal proportions of the farms Modderfontein 345 IQ and Doornpoort 347 IQ totalling 2469 hectares. The mine infrastructure is situated on parts of these farms. As of February 20, 2006 the mine also owns 437 houses, including 84 flats within the residential areas of Westonaria and Fochville. In addition, the mine has a hostel, which has 388 rooms, and 115 guesthouses located at the mine site.

12.1 Project Infrastructure

Power is supplied to the mine through the ESCOM national power grid by a ring main feed from two independent high-tension power supply systems. Potable water is supplied to the mine by Rand Water (a semi-parastatal water supply company) and has been a very reliable source during the operating life of the mine.

Shaft structures consists of the old South Shaft servicing SV1, 2 and 3 sub-vertical shafts, Twin Shafts (Main and Ventilation) accessing the Current Mine (90 to 95 level), and Phase 1 (100 to 110a level).

Compressed air is supplied from surface by means of three 30,000 cfm sulzer machines situated at the South Shaft with the future plan of two additional 30,000 cfm machines situated at Twin Shafts Complex.

Refrigerated water is supplied by means of five refrigeration plants situated on surface. The water used by these installations is from underground pumped water and Rand Water Board make up water.

The newly commissioned process plant is currently equipped for 220,000 tonnes per month (“t/m”), and will be upgraded to a capacity of 330,000 t/m, as specified in the mine plan.

The lease area as described in Section 3 is sufficient for the planned mining operations and provides ample space for existing and potential future planned waste rock and tailings dump sites which are capable of providing waste storage for the planned duration of operations. The South Deep Plant is located adjacent to the Main Shaft complex and includes surface area to allow for expansion of the plant capacity.

South Shaft infrastructure: consists of workshops, offices, sewage plant, security and ancillaries, hostels, married quarters, shops and general area.

13 Environmental Management

The South Deep Mine operates under a sustainability policy, which commits the operation to a high standard of environmental stewardship. This involves protecting human health, reducing the impact of mining on the ecosystem, and returning the site to a state compatible with a healthy environment. A series of management systems have been implemented for maintenance, environmental health and safety.

South Deep Mine does not comply with regards to the National Water Act. Water is currently being discharged into public streams without a Water Use Licence. An application has been submitted to Dept. Water Affairs & Forestry and is currently being reviewed

South Deep has been recognised as a world leader by the World Bank for its treatment program for employees living with HIV/Aids, and providing marketable skills for former employees and their families affected by economic downscaling.

South Deep operates a human resources development program to provide Adult Basic Education to the illiterate and semi-literate portions of the workforce.

The mine is actively involved in providing suitable housing for employees and their families in Westonaria and its surrounds, and is also committed to a number of community outreach programs such as the Ipeleng Primary and Thusa Setshaba Senior Secondary School feeding projects.

12.10(c)

South Deep is currently contributing to an environmental rehabilitation trust fund to be used for environmental closure obligations at the end of the mine life including costs for decommissioning and ongoing maintenance. The total closure costs are estimated to be R56 million. The current provision in the trust is R35 million and the required contribution for the next 10 years is R2.1 million/a followed by a contribution R0.4 million per annum for the rest of the mine life.

The following programme is considered the most important from the rehabilitation viewpoint

Tailings Dams Rehabilitation (Existing 2 dams): The following rehabilitation actions were taken into account:

·	Seal Penstocks;

·	Rock Cladding of sides;

·	Rehabilitation and cladding of paddocks;

·	Re-vegetation;

·	Stability monitoring;

·	Removal of pipeline;

·	Removal of pump houses;

·	Rock cladding already done.

South Shaft Waste Rock Dump: The waste rock dump at South Shaft can be rehabilitated in various ways, and current figures reflect a worst case scenario, with drainage channels and in situ rock encapsulation as options;

South Shaft - Infrastructure: Include all workshops, offices, sewage plant, security and ancillaries, hostels, married quarters, shops and general area;

Twin Shaft Complex -Shafts: Includes all structures associated with the shafts;

Twin Shaft Complex – Metallurgical Plant: Includes demolition and removal of all structures, transformers, workshops, ponds, pump houses, sub station, smelter house, concrete paving, tar roads, security fencing and radiological hazard clean-up;

After Care Monitoring – After care and maintenance of the closure plan objectives once implemented, is required to prove to the Government that these controls are sustainable and include:

·	Ground water rehabilitation cost;

·	After Care monitoring for a period of 5 years.

14 Technical Economic Parameters

14.1 Operating Cost

The operating costs used are based on those achieved at South Deep and are shown in Table 14.1.

Table 14.1: Estimated Operating Costs

	Cost/ Milled Tonne (ZAR/t)
Manpower Cost	151.9
Consumables	121.3
Contractors	15.1
Utilities (Power and Water)	50.2
Business Support	7.9
Overheads	8.6
Total	355.2

14.2 Capital and Sustaining Cost

The capital cost is dependant on the valuation scenario; please refer to tables for details. Sustaining capital is based on the requirements at South Deep which is 5% of operating cost.

14.3 Production Profiles

The production profiles (tonnes, metals, and grade) are dependant on the valuation scenario; please refer to tables for details. The base production profile, yield and cost/milled tonne for the Kalbasfontein property is shown in Table 14.2.

14.4 Recovery Profiles

Recoveries are based on those achieved currently at South Deep.

14.5 Unit Revenue Profiles

A constant real value of USD550.00/oz of Au has been used for valuation purposes.

14.6 Exchange Rate Profiles

A constant real exchange rate of ZAR7.18:1USD has been used for valuation purposes.

Table 14.2: Base Kalbasfontein Production Profile, Yield and Cost/milled Tonne

	Units	Y 1	Y 2	Y 3	Y 4	Y 5	Y 6	Y 7	Y 8	Y 9
Tonnes Milled	kt	3	185	220	456	348	290	883	2 106	2 372
Yield	g/t recovered	2.0	1.5	1.6	2.4	2.8	3.5	4.0	4.3	4.5
Operating Cost	ZAR/t	540	388	429	432	444	417	348	331	311.

		Y 10	Y 11	Y 12	Y 13	Y 14	Y 15	Y 16	Y 17	Y 18
Tonnes Milled	kt	2 676	2 603	2 584	2 389	2 529	2 354	2 545	2 664	2 350
Yield	g/t recovered	4.3	4.8	5.0	5.3	5.1	4.9	4.4	4.6	4.8
Operating Cost	ZAR/t	322	333	340	328	342	345	353	363	382

		Y19	Y 20	Y 21	Y 22	Y 23	Y 24	Y 25	Y 26	Y 27
Tonnes Milled	kt	2 253	2 192	2 370	2 588	2 560	1 777	1 901	1 728	1 718
Yield	g/t recovered	4.6	5.0	4.8	5.2	5.3	5.3	5.7	5.6	5.6
Operating Cost	ZAR/t	367	409	372	358	367	370	359	349	341

		Y 28	Y 29	Y 30	Y 31	Y 32	Y 33	Y 34	Y 35	Y 36
Tonnes Milled	kt	1 875	1 954	1 789	1 236	751	751	576	312	44
Yield	g/t recovered	5.6	5.7	5.8	5.4	5.2	5.2	5.2	5.9	9.2
Operating Cost	ZAR/t	341	346	349	370	381	381	381	405	540

15 Material Assets Valuation

15.1 Valuation Methodology

Please note that unless stated otherwise, the values, for the purposes of understanding should be interpreted as FSD holding 100% of the 4 properties.

In SRK’s opinion Kalbasfontein and WA4, although not development properties, are too far down the project stage timeline to warrant the use of cost approaches to determine fair values. The Resources on the WA4 property are classified as Inferred Resources thus SRK, in accordance with the SAMREC Code, has valued the WA4 area using comparable transactions only. SRK has used Discounted Cash Flow techniques (“DCF”) and comparable transactions to value the Kalbasfontein property. Cardoville and Wildebeeskuil Projects which lie to the south are outside the scope of this report. They will have economic potential in the future as they are part of the same trend found in Kalbasfontein but due to a paucity of data at present they have not been assessed. They do form part of the transaction however.

The Net Present Value (“NPV”) for the Kalbasfontein property is determined in real terms. The NPV is derived from post tax, pre-debt real cashflows. The real cashflows are based on Technical Economic Parameters (“TEP’s”), commodity price and macro-economic projections. To accommodate taxation a nominal financial model (“FM”) has been populated and the final free cash flow has been de-escalated to real terms prior to discounting. A discount factor of 8% is applied to the real cashflows to determine the NPV. The discount factor is not based on FSD’s weighted cost of capital, but in SRK’s experience is a reasonable discount factor for this type of project.

To populate the nominal FM the commodity prices have been escalated over the LoM by US inflation which is estimated to be 2% per annum. All capital and operating costs have been escalated from Year 1 to the end of the LoM by RSA inflation which is estimated to be 4.9% per annum. The predicted exchange rate is determined from the differential between the US and RSA inflation which in this case is 2.9% per annum. The final cashflow is converted to real terms by de-escalating the cashflow by RSA inflation prior to discounting. The FM is based on calendar years running from June to July. The Kalbasfontein valuation is based on a zero starting balance with respect to unredeemed capital expenditure and assessed losses.

In addition SRK has chosen to value the Kalbasfontein and the WA4 property using comparable market transactions. Applicable transactions have been used to calibrate a yard stick (value per Au oz in the ground) which has been applied to the property Resources to determine property values. All transactions have been adjusted to take into account the time value of money.

The valuation is based on information provided by JCI and Gold Fields, in particular SRK has relied on information in the following documents:

·	Western Areas – Kalbasfontein and WA4 Project Area, Dexter S. Ferreira, 2/27/2007;

·	Total CMSTMod (Dexter Ferreira on Uncle Harry's) (Excel Document); and

·	Uncle Harry’ s Ground(SRK) (Excel Document).

The valuation date can be considered as 1 June 2007.

The valuation is based on the farms as shown in Table 2.1 and Table 2.2., and the Resources therein as depicted in Section 8 and 9. The valuation is based on both properties and cannot be split, prorated or inferred.

15.2 Royalties

Royalties have been deducted from metal content as per SRK’s understanding of the South African Royalties Bill (Draft). A royalty of 1.5% has been added to the operating costs.

15.3 Taxation

Company tax has been deducted from profits in line with the South African legislation. The FMl takes into consideration a tax shield consisting of unredeemed capital and assessed losses. Taxable income is determined from operating profit less unredeemed capital expenditure in the year the capital expenditure occurs. In cases of an operating loss the capital expenditure is carried forward and added to unredeemed capital expenditure. The effective tax rate is determined from the South African gold taxation formula as follows:

Effective Tax Rate = 45% – 225/Profitability Ratio (“P/R”), where the P/R = Profit / Revenue

15.4 DCF Valuation Results

SRK has used 3 DCF scenarios to determine project values. The scenario’s are as follows:

Stand alone: all necessary infrastructures are required to be funded by the project to realize the production profile. The infrastructure and capital cost is based on similar infrastructure to that of South Deep.

Maximize the South Deep Net Present value: the project is commenced at a time which would maximize South Deep NPV.

No Capacity Constraint: In this scenario the hoisting and processing constraints existing at South Deep is ignored and the Kalbasfontein area is accessed as soon as possible in conjunction with the current South Deep LoM.

The values from these scenarios are interpreted to determine appropriate NPV values for the project. The summary values for the scenarios are shown in Table 15.1.

Table 15.1: Summary NPV Scenario Values

	Units	Value Added	Stand alone	No Capacity Constraint
NPV@8%	ZARm	93	(1,554)	1 070
NPV@10%	ZARm	41	(1,662)	606
NPV@12%	ZARm	22	(1,664)	309

Tonnes Milled	T	57,933.6	57,933.6	57 933.6
Yield	g/t Recovered	-0.7	5.0	5.0
Gold Produced	Kg	286,833.6	286,833.6	286 833.6

Cash Costs	ZAR (1000)	21 121 723	20 575 636	20 575 636
Infrastructure Capital	ZAR (1000)	Included in Sustaining	5 600 000	Included in Sustaining
Sustaining Capital	ZAR (1000)	1 904 560	1 091 470	1 904 560
Total Costs	ZAR (1000)	23 026 283	27 267106	22 480 196

15.4.1 SRK Comments on Scenario Results

Stand alone: The results indicate that the Kalbasfontein property would not be economically viable as a stand alone project given the current economic assumptions.

To be economically viable access capital would have to reduce from ZAR5.6bn to ZAR2.4bn, which in SRK’s opinion is unlikely or the real long term gold price would have to increase from USD550.00/oz to USD793.00/oz which in SRK’s opinion can be considered optimistic.

In this scenario first production would commence in 2013.

Maximize the South Deep Net Present value: Due to the lower overall grade of the Kalbasfontein Reserves when compared to that of Phase 1 and Phase 2 of the South Deep project, and Hoisting constraints at South Deep, South Deep NPV is reduced if Kalbasfontein Reserves are mined in preference to that of Phase 1 or Phase 2.

The acquisition of the Kalbasfontein property only improves the South Deep NPV as a LoM extension commencing after the completion of Phase 1 and Phase 2, i.e. capital expenditure to access the Kalbasfontein area would start in 2040 and first production would commence in 2043.

It is SRK’s opinion this indicates the best property use and thus the scenario can be considered valid. But the value of ZAR93m is not an appropriate sales value for the Kalbasfontein property as South Deep does not pay for the opportunity to access the area at an earlier stage if required.

No Capacity Constraint: In this scenario capital expenditure would commence in 2008 and first production would commence in 2011.

It is SRK’s opinion that ZAR1,070m is not an appropriate sales value for the Kalbasfontein area because given the assumptions it is impossible for South Deep to realize this value if they acquired the Kalbasfontein property.

SRK is of the opinion that it is not appropriate to determine a value for the South Deep infrastructure with respect to the Kalbasfontein property and offset this value against the project value.

It is SRK’s opinion that the infrastructure would be considered in a joint venture deal, i.e. South Deep infrastructure would be considered in a funding scenario, not to determine the value of the property.

15.4.2 Interpretation of DCF Results

Due to South Deep hoisting constraints, South Deep to maximise its NPV, would only access the Kalbasfontein area in 2040.

If it is assumed that Gold Fields would pay ZAR1,070m (2007 money terms) in 2035 for the project. (5 years prior to capital infrastructure development due to security of tenure), and that the Kalbasfontein owners require that the sale happens in 2007, the appropriate value to be paid now is that value, which if invested at a Risk free rate will realize ZAR1,070m in 2035.

If it is assumed that a Risk free investment yields a real rate of return of 2% a value of ZAR627m invested for 28 years will realize ZAR1,070m in 2035.

SRK is of the opinion that this is an appropriate sale value for the Kalbasfontein area as the Kalbasfontein owners would receive ZAR1,070m (2007 money terms) in 2035 without having to engage any additional risk and the cost of the opportunity for Gold Fields to access the area is ZAR627m – ZAR93m (additional NPV from South Deep Maximum NPV scenario) = ZAR534m. The Kalbasfontein value of ZAR627m can be considered as the Risk free value.

It is SRK’s opinion that the overall risk profile associated with South Deep production build-up (Phase 2) would reduce if South Deep accessed the Kalbasfontein area as soon as possible. To simulate this scenario SRK has reduced the grade from Phase 2 to the average weighted grade of Phase 2 and Kalbasfontein (4% grade reduction).

In addition SRK has included 40% of the Kalbasfontein capital expenditure commencing in 2008. Lowering the required return of 8% to 7% (reflecting the lower risk profile), results in an additional NPV of R738m when compared to Phase 1 and Phase 2 only. The Kalbasfontein value of ZAR738m can be considered as the Risk Adjusted value.

15.4.3 Comparable Transaction Valuation Results

The transactions listed in Table 15.4 were considered in the calibration of a yard stick (value per Au oz in the ground) which has been applied to property resources to determine property values. These transactions have been extracted from the Metal Economics Group database and are deemed comparable as they are underground Witwatersrand gold property transactions and at the time of the transaction the projects where at a similar development stage.

SRK has not considered 3 of the transactions due reasons given in the “Applicable/ Transactions” column. SRK considers the Target North transaction as the most similar to that of the Kalbasfontein property.

In addition to the database transactions SRK has considered a “Look Through” transaction where according to Gold Fields part of the total transaction for South Deep was an additional payment of R1bn to Western Areas which was made up as follows

Table 15.2: South Deep Look Through Transaction

	ZAR
Cash Portion	320 000 000
Mineral Rights (29.6%)¹	200 000 000
Royalty	480 000 000
Note: ¹The mineral rights included Kalbasfontein, WA4, Cardoville and Wildebeestkuil. SRK has accepted the validity of the transaction from Gold Fields in good faith

15.4.4 Interpretation of Comparable Results

The database transactions and the Look Trough transactions are summarized in Table 15.3.

Table 15.3: Comparable Transaction Summary

	Min	Mean	Max
Look Through¹		676 ²
WA4		55 ³
Kalbasfontein	889 ⁴		966 ⁵
Note: ¹The mineral rights included Kalbasfontein, WA4, Cardoville and Wildebeestkuil.. SRK has accepted the validity of the transaction from Gold Fields in good faith ² R300m x 100/ 29.6 = R676m ³ USD7.7m x 7.18 = R55m (WA4 Applicable Transactions) ⁴ USD123.8m x 7.18 = R889m (Kalbasfontein Most Applicable Transaction) ⁵ USD134.6m x 7.18 = R966m( Kalbasfontein Applicable Transactions)

100

Table 15.4: Comparable Transactions

				Grade	Content	Transactions			All Transactions	Applicable /Transactions	Most Applicable Transaction
		Date	Mt	Au (g/t)	Au (koz)	Payment (USD) (2007)	Share (%)	Total Value (USDm)	Cost/Au ounce (USD/Au oz)	Cost/Au ounce (USD/Au oz)	Cost/Au ounce (USD/Au oz)
WA4	Total Reserves and Resources		6	7.78	1 426
All Transaction Value								43.5	30.5
Applicable Transactions value								7.7		5.4
Kalbasfontein	Total Reserves and Resources		96	8.09	24 893
All Transaction Value								760.2	30.5
Applicable Transactions value								134.6		5.4
Most Applicable								123.8			5.0
Burnstone	Total Reserves and Resources	2002	34.7	15.53	17 326	3.52	80.0	4.4	0.3	0.3
Burnstone	Total Reserves and Resources	2007	46	7.08	10 471	36.8	26.0	141.5	13.5	13.5
Doornkop	Total Reserves	2003	20	10.17	6 539	29.5	26.0	113.5	17.4	Reserves only
Malmani	Total Reserves and Resources	1994	1.6	6.86	353	1.7	100.0	1.7	4.9	4.9
Modder East	Total Reserves and Resources	2002	6	0.35	68	9.6	92.1	10.5	154.9	Outlier
Target North	Total Reserves and Resources	1996	309	5.86	58 217	57.9	20.0	289.5	5.0	5.0	5.0
Weltevreden	Total Reserves and Resources	2005	21.42	4.75	3 271	11.0	100.0	11.0	3.4	3.4
South Deep	Total Reserves and Resources	2007	288	7.20	66 668	1 500.0	50.0	3 000.0	45.0	Operating Mine

Average						21.4		81.7	30.5	5.4	5.0

101

15.5 Valuation Summary

The valuation summary of FSD’s 44% share in the properties are shown in Table 15.5 and Figure 15.1.

SRK has chosen the following values as preferred values

·	The Risk Free value.

·	The Risk adjusted value.

·	The Target North Transaction.

·	The Look through Transaction.

·	SRK has chosen not to include the comparable value associated with WA4 due to the resolution of the results, and consider the value of WA4 to offset some of the risk associated with the Kalbasfontein area.

It is SRK’s opinion that in this case the confidence associated with the DCF values and those of the comparable values are similar and thus should have similar influence on the final value. The range associated with the comparable value is twice that of the DCF values, thus the weighting is 66% comparable and 33% DCF.

Table 15.5: Valuation Summary of FSD’s 44% Share

Summary		FSD	Peotona	WAL	Total
Summary		44.4%	26.0%	29.6%	100.0%
Discounted Cash Flow
Stand alone		(690)	(404)	(460)	1,554)
Maximise South Deep Business Plan		41	24	27	93
No Capacity Constraint		475	278	317	1,070
Risk Free		278.4	163.0	185.6	627.0
Risk Adjusted		327.5	191.8	218.3	737.6
Comparable
Look Through	Mean	300.0	175.7	200.0	675.7
WA4 (No Outliers)	Mean	24.6	14.4	16.4	55.3
Kalbasfontein (No Outliers)	Max	429.0	251.2	286.0	966.3
Kalbasfontein (Most Applicable)	Min	394.5	231.0	263.0	888.6
SRK Preferred Values
Discounted Cash Flow	Max	327.5	191.8	218.3	737.6
Discounted Cash Flow	Min	278.4	163.0	185.6	627.0

Comparable	Max	394.5	231.0	263.0	888.6
Comparable	Min	300.0	175.7	200.0	675.7

102

Figure 15.1: Valuation Summary of FSD’s 44% Share

103

15.6 Sensitivity

The sensitivity of the gold price to the value of FSD’s 44% share in the properties are shown in Table 15.6.

Table 15.6: Gold Price Sensitivity on FSD’s 44% Share of the Kalbasfontein & WA4 Project

Gold Price (USD/oz)	Min (ZARm)	Max (ZARm)
525	277.0	358.1
550	292.9	372.4
575	308.8	388.2
600	324.7	402.5
625	340.6	418.4
650	357.4	432.8
675	373.2	447.2

16 Conclusions

SRK has valued the Kalbasfontein property using DCF techniques and comparable transactions and the WA4 area using comparable transactions.

SRK financial evaluation indicated that the Kalbasfontein property would not be economically viable to be operated as a stand alone project. In SRK’s opinion the Kalbasfontein property maximum financial value can be obtained as an extension of South Deep.

Comparing the reserves to the current South Deep reserves, mining of the Kalbasfontein reserves are of lower value to the current South Deep reserves and in SRK opinion cannot be mined in place of the current South Deep reserve production schedule without reducing the South Deep NPV. The Kalbasfontein reserves maximum value is realized when these reserves are mined at the end of the South Deep mine life.

It is SRK opinion that South Deep overall production risk profile would be reduced with the acquisition of the Kalbasfontein project, as this would allow more operational flexibility.

·	In SRK’s opinion the DCF value for 44.4% of the Kalbasfontein property including the WA4 lies between R278m and R327m.

·	In SRK’s opinion the Comparable value for 44.4% of the Kalbasfontein property including the WA4 lies between R300m and R394m.

·	Using a 33:66 weighting SRK values the 44.4% of the Kalbasfontein and WA4 property between R293m and R372m.

17 References

The Evaluation of Western Areas’ Kalbasfontein (Goldridge) & WA4 Projects Using the Bayesian Approach - Author: Dexter S. Ferreira, May 2006

104

Competent person	Sponsor and corporate adviser


Independent reporting accountants and auditors	Corporate law adviser

Company Secretary and registered office Mr R P Pearcey, FCIS, FCIMA 13th Floor 28 Harrison Street Johannesburg, 2001 (PO Box 11165, Johannesburg, 2000) Telephone: +27 11 688 5100 Facsimile: +27 11 492 1070 Website: www.randgold.co.za	South African transfer secretaries Computershare Investor Services 2004 (Proprietary) Limited (Registration number 2004/003647/07) Ground Floor 70 Marshall Street Johannesburg, 2001 (PO Box 61051, Marshalltown 2107) Telephone: +27 861 100 634 or +27 11 370 5000
Corporate law adviser Taback and Associates (Proprietary) Limited 13 Eton Road Parktown Johannesburg, 2193 (PO Box 3334, Houghton, 2041) Telephone: +27 11 358 7700 Facsimile: +27 11 358 7800	United Kingdom secretaries St James’s Corporate Services Limited 6 St James’s Place London SW1A 1NP United Kingdom Telephone: +44 (20) 7499 3916 Facsimile: +44 (20) 7491 1989
Sponsor and corporate adviser Sasfin Capital a division of Sasfin Bank Limited (Registration number 1951/002280/06) Sasfin Place North Block 13 – 15 Scott Street Waverley, 2090 (PO Box 95104, Grant Park 2051) Telephone: +27 11 809 7500 Facsimile: +27 11 809 7726	United Kingdom registrars Capita Registrars The Registry 34 Beckenham Road Beckenham Kent BR3 4TU United Kingdom Telephone: 0870 162 3100 (from inside the United Kingdom) Telephone: +44 (20) 8639 3399 (from outside the United Kingdom) Facsimile: +44 (20) 8639 2342
Reporting accountants and auditors KPMG Inc. (Registration number 1999/021543/21) KPMG Crescent 85 Empire Road Parktown, 2193 (Private Bag 9, Parkview, 2122) Telephone: +27 11 647 7111 Facsimile: +27 11 647 8000	Communications Brian Gibson Issue Management Brian Gibson 23 Sutherland Avenue Craighall Park, 2196 (PO Box 406, Parklands, 2121) Telephone: +27 11 880 1510 Facsimile: +27 11 880 1392
Competent person SRK Consulting Engineers and Scientists (Registration number 1995/012890/07) 265 Oxford Road Illovo, 2192 (PO Box 55291, Northlands, 2116) Tel: +27 11 441 1111 Fax: +27 11 880 8086

Corporate information	Inside front cover
Forward-looking statement disclaimer for R&E		1
Definitions and interpretations		3
Important dates and times		6
Circular to R&E shareholders
1. Introduction		7
2. The proposed transaction		7
3. Opinion of the R&E board		10
4. Financial information		10
5. Major shareholders		12
6. Material changes		12
7. Directors		12
8. R&E and its directors’ interests and dealings		14
9. Future strategy of R&E		14
10. History of changes		15
11. Working capital statement		15
12. Litigation statement		15
13. Directors’ responsibility statement		15
14. Significant contracts		15
15. Borrowings		15
16. Expenses		15
17. Consents		16
18. Notice of R&E general meeting		16
19. Documents available for inspection		16
Annexure I Provisional unaudited and unreviewed results for the two years ended 31 December 2005
and restated provisional results for the year ended 31 December 2003 for R&E		17
Annexure II Shareholder update on the settlement and/or merger negotiations between JCI and R&E
and renewal of cautionary		34
Annexure III Pro forma income statement and balance sheet of R&E		38
Annexure IV Report of the independent reporting accountants’ on the pro forma financial effects of the
proposed transaction and the pro forma income statement and balance sheet of R&E		40
Annexure V(a) Cardoville Prospecting Rights		42
Annexure V(b) Kalbasfontein Prospecting Rights		44
Annexure V(c) WA4 Prospecting Rights		45
Annexure V(d) Wildebeestkuil Prospecting Rights		46
Annexure VI Litigation statement		47
Annexure VII Competent Person’s Report on the contiguous rights		49
Notice of R&E general meeting		105
Form of proxy for use by certificated R&E shareholders and “own name” dematerialised R&E
shareholders only	Attached

	Date
Circular and notice of the R&E general meeting posted to R&E shareholders on	Monday, 15 October 2007
Last day for lodging forms of proxy with the transfer secretaries for the R&E general meeting by no later than 10:00 on	Monday, 29 October 2007
R&E general meeting to be held at 10:00 on	Wednesday, 31 October 2007
Results of the R&E general meeting released on SENS on	Wednesday, 31 October 2007
Results of the R&E general meeting published in the press on	Thursday, 1 November 2007

D M Nurek	(non-executive Chairman)
P H Gray	(CEO)
M Steyn	(Financial)
D I de Bruin	(Independent non-executive)
M B Madumise	(Independent non-executive)

Percentage holding in WAP	Percentage
R&E	24.46
JCI	19.94
FSD effective interest	44.40
GFO	29.60
Sub-total	74.00
BEE participant	26.00
	100.00

R&E shareholders	Number of R&E shares held	Percentage holding of R&E shares entitled to vote
Trinity Asset Management (Pty) Limited	10 500 000	14.03
Allan Gray Limited¹	9 054 786	12.10
Nedbank Securities	8 673 000	11.59
Clear Horizon Capital	3 804 572	5.09
JCIIF	3 250 000	4.34
JCI	1 696 868	2.27
CMMS	1 250 000	1.67
Oryx Investment Management	1 049 856	1.40
	39 279 082	52.5

	Before (cents)¹	After the proposed transaction (cents)²	Change (%)
EPS per R&E share – (cents)⁴	193	456	136.21
HEPS per R&E share – (cents)⁵	10	10	–
NAV per R&E share – (cents)⁶	1 969	2 217	12.62
TNAV per R&E share – (cents)⁶	979	1 228	25.37
Number of R&E shares in issue⁷	74 813 128	74 813 128	–
WANOS in issue⁸	74 813 128	74 813 128	–

	Before¹ R’000	Value after the proposed transaction² R’000	Difference %
Prospecting rights	250 000	217 684³	(12.93)
Taxation	–	(30 750)⁴	–
Net number of shares in issue	71 813 128	71 813 128⁵	–
NAV	827 000	763 934^6.1	(7.63)
NAV per R&E share (cents)	11.52	10.64^6.1	(7.63)
NAV – assuming a settlement of R1.2 billion	2 027 000	1 963 934^6.2	(3.11)
NAV per R&E share – assuming a settlement of R1.2 billion (cents)	28.23	27.35^6.2	(3.11)
NAV – assuming a settlement of R1.5 billion)	2 327 000	2 263 934^6.3	(2.71)
NAV per R&E share – assuming a settlement of R1.2 billion(cents)	32.40	31.53^6.3	(2.71)

R&E shareholders	Number of R&E shares	Percentage holding of R&E shares
Allan Gray Limited	9 274 817	12.40
Nedbank Group	9 224 327	12.33
Bank of New York (ADR’s)	8 664 040	11.58
Clear Horizon Capital	5 058 185	6.76

Name	Designation	Appointment date	Resignation date
R B Kebble *	CEO	24.07.2003	24.08.2005
R A R Kebble	Non-executive chairman	05.03.1998	24.08.2005
H C Buitendag	Financial	01.03.2000	24.08.2005
M B Madumise	Independent non-executive	24.07.2003
L R Ncwana	Non-executive	24.07.2003	24.08.2005
A C Nissen	Non-executive	24.07.2003	01.04.2007
J C Lamprecht	Financial	24.08.2005	16.05.2006
P H Gray **	CEO	24.08.2005
D M Nurek***	Non-executive chairman	07.10.2005
J Blersch	Independent non-executive	14.08.2006	09.03.2007
T G Dale	Independent non-executive	14.08.2006	09.03.2007
M Steyn	Financial	13.12.2006
D I de Bruin	Independent non-executive	01.04.2007

Director	Salary R	Bonus R	Fees R	Sign-on incentive R	TOTAL R
D M Nurek	–	–	62 500	–	62 500
P H Gray *	700 000	100 000	–	2 600 000	3 400 000
J C Lamprecht *	700 000	100 000	–	6 956 250	7 756 250
M B Madumise	–	–	78 750	–	78 750
A C Nissen	–	–	22 500	–	22 500

Director	Salary	Bonus	Retrenchment	Directors’ fees	TOTAL
	R	R	R	R	R
D M Nurek	–	–	–	250 000	250 000
P H Gray *	1 254 000	1 156 000	–	–	2 410 000
J C Lamprecht *	695 442	300 000	600 000	–	1 595 442
M Steyn *	400 000	–	–	–	400 000
J Blersch	–	–	–	75 000	75 000
T G Dale	–	–	–	75 000	75 000
M B Madumise	–	–	–	84 375	84 375
A C Nissen	–	–	–	103 125	103 125

Expenses	R’000
Printing and distribution – Ince (Pty) Ltd	420
Documentation and inspection fee – JSE	25
Reporting accountants report – KPMG	25
Legal expenses – Taback and Associates	500
Sponsor and corporate adviser – Sasfin Capital	300
Competent Person’s Report – SRK Consulting	240
Estimated total	1 510

R000	Provisional Year ended 31 December 2005	Provisional Year ended 31 December 2004	Restated Year ended 31 December 2003	Previously Audited Year ended 31 December 2003
Income
Dividends and interest received	12 049	28 962	8 248	9 357
Profit on sale of portion of investments in associate	-	138 327	142 340	241 429
(Loss)/Profit on sale of investments	404	23 112	5 398	(34 151)
Profit on sale of property, plant and equipment	-	178	772	-
Fair value adjustment of investment property	-	1 550	-	-
Claims	322 624	331 174	117 791	-
Fair value adjustment of investments	(4 016)	7 577	(16 417)	-
Other income	2	7 869	14 215	14 987
Profit on sale of subsidiary	1 421	-	-	-
Expenses
Interest expense	(18)	(7 867)	(6 545)	(6 447)
Depreciation and amortisation	(156)	(203)	(179)	(994)
Impairments	(48 725)	(179 413)	-	(24 398)
Write-off of listed investments and other long-term assets	(100 552)	(235 506)	(268 276)	-
Derecognition of investments in Kabusha/Viking Pony	-	-	(132 983)	-
Foreign exchange loss	(945)	(139)	(132 067)	(132 067)
Commission on sale of property	-	(500)	-	-
Provision for bad debts	-	(30 963)	-	-
Exploration and corporate expenditure	(27 811)	(11 473)	(27 924)	(27 972)

R000	Notes	Provisional Year ended 31 December 2005	Provisional Year ended 31 December 2004	Restated Year ended 31 December 2003	Previously Audited Year ended 31 December 2003
Profit/(Loss) before taxes, equity income and minority interest	1	154 757	72 685	(295 627)	39 744
Taxation	2	(38 628)	(47 310)	(29 571)	(29 593)
Profit/(Loss) before equity income and minority interest		116 129	25 375	(325 198)	10 151
Equity (loss)/income from associate	4	-	(11 778)	166 351	166 351
Minority interest		28 268	(6 364)	(261)	702
Profit/(Loss) for the year		144 397	7 233	(159 108)	177 204
(Earnings)/Loss per share (cents)		193	12	(330)	367
Headline earnings/(loss) per share (cents)		10	(40)	(50)	(12)
Reconciliation between earnings/(loss) and headline earnings/(loss)
Earnings/(Loss) for the year		144 397	7 233	(159 108)	177 204
Add back:
Profit on sale of portion of investment in associate		-	(138 327)	(142 340)	(241 429)
Impairment of goodwill		-	-	-	24 398
Impairment of fixed assets		48 275	179 413	-	-
Claims		(322 624)	(331 174)	(117 791)	-
Write-off investments		100 522	235 506	268 276	-
Write-off of investment in Kabusha		-	-	132 983	-
Fair value adjustment of investment property		-	(1 550)	-	-
Profit on sale of property, plant and equipment		-	(178)	(772)	-
Profit on sale of investments		(404)	(23 112)	(5 398)	-
Profit on sale of subsidiary		(1 421)	-	-	-
Tax effect of adjustments		36 628	47 310	-	33 810
Headline earnings/(loss) for the year		7 373	(24 879)	(24 150)	(6 017)

R000	Provisional Year ended 31 December 2005	Provisional Year Ended 31 December 2004	Restated Year ended 31 December 2003 ⁽¹⁾	Previously Audited Year ended 31 December 2003
Balance at beginning of the year	1 311 442	707 196	1 021 669	547 629
Other	5 016	(4 324)	29 320	-
Shares issued during the year	-	314 081	-	304 588
Deferred tax	(36)	(52 480)	-	-
Net movement in reserves – fair value adjustments	12 250	339 736	(7 481)	(7 752)
Profit for the year	144 397	7 233	(336 312)	177 204
Balance at end of the year	1 473 069	1 311 442	707 196	1 021 669

2. TAXATION The income tax expense recognised in the income statement reflects a best estimate for Capital Gains Tax ("CGT") at 14,5% in 2005 (2004: 15%) on the deemed profit in terms of the disposals of investments for which a recoverable claim receivable has been recognised and on profits on sale of investments. A deferred tax liability was also raised on the mark-to-market adjustments relating to investments recorded in equity at CGT rates at 31 December 2004 and 2005.
R000	Provisional Year ended 31 December 2005	Provisional Year ended 31 December 2004	Restated Year ended 31 December 2003	Previously Audited Year ended 31 December 2003
3. PROPERTY, PLANT AND EQUIPMENT Opening balance Additions Disposals Depreciation Impairments	53 863 - (4 999) (156) (48 275)	39 225 194 254 - (203) (179 413)	14 713 24 754 (63) (179) -	14 713 65 477 (63) (994) -
Balance at year end	433	53 863	39 225	79 133
As previously reported Fictitious mineral rights reversed – Phikoloso transaction			79 133 (39 908)
At 31 December 2004 the total for property, plant and equipment included R5 million relating to investment property, carried at fair value. Such property was disposed of during 2005.
4. INVESTMENT IN ASSOCIATE RRL	-	-	419 311	432 339
As previously reported Adjust holding from 36.86% to 35.75%			432 339 (13 028)
Equity (loss)/income from associate		(11 778)	166 351	166 351
Randgold's shareholding in RRL decreased from 35,75% (per restated provisional results) at 31 December 2003 to below 20% during 2004. As a result, Randgold ceased to equity account for its investment in RRL. At 31 December 2004 and 2005 the investment in RRL is mark-to-market and reflected under investments.

5. INVESTMENTS	Number of shares	Provisional Year ended 31 December 2005 R000	Provisional Year ended 31 December 2004 R000	Restated Provisional Year ended 31 December 2003	Previously Audited Year ended 31 December 2003 R000
Listed– at market value
Kelgran Pan Palladium WAL JCI Ltd ordinary shares JCI Ltd debentures RRL	2 324 830 13 100 000 5 361 613 37 560 613 50 651 4 000 000 ⁽¹⁾	488 28 166 227 064 6 010 63 412 500	- - - - - -	- - - - - -	- - - - - -
		674 291	-	-	-
⁽¹⁾ Pledged as security. Refer note 11.
Aflease Kelgran Pan Palladium WAL JCI Ltd ordinary shares JCI Ltd debentures RRL	53 049 442 2 324 830 13 100 000 5 361 613 37 560 613 50 561 8 650 000	- - - - - -	94 959 372 28 165 134 040 10 141 42 582 578	- - - - - - -	- - - - - - -
		-	850 297
Aflease Kelgran Simmer and Jack WAL JCI Ltd ordinary shares	8 100 000 2 324 830 40 000 000 82 500 55 867 389	- - - - -	- - - - -	26 325 255 8600 3 424 43 577	- - - - -
		-	-	83 181	-
Aflease Kelgran Simmer and Jack WAL JCI Ltd ordinary shares Amplats DRD Harmony	7 300 000 2 321 115 40 000 000 4 042 500 55 867 389 235 000 500 000 315 000	- - - - - - - -	- - - - - - - -	- - - - - - - -	23 724 255 9 600 167 765 43 577 68 503 10 300 34 177
Aflease shares held in Kabusha	23 000 000	- -	- -	- -	357 901 73 308
		-	-	-	431 209
Restated provisional				83 181
As previously reported Adjusted for: - Aflease - DRD - Amplats - Harmony - WAL - Aflease - Aflease	7 300 000 500 000 235 000 315 000 3 960 000 23 000 000 8 100 000			431 209 (23 724) (10 300) (68 503) (34 177) (164 340) (73 309) 26 325

Aflease	The Afrikander Lease Limited
Amplats	Anglo American Platinum Corporation Limited
DRD	DRDGold Limited (formerly Durban Roodepoort Deep, Limited)
Harmony	Harmony Gold Mining Company Limited
Inkwenkwezi	Inkwenkwezi Gold (Pty) Limited
JCI Gold	JCI Gold Limited (a wholly-owned subsidiary of JCI Limited)
JCI Group	JCI Limited (The JCI Group and its subsidiaries)
Kelgran	Kelgran Limited
Phikoloso	Phikoloso Mining (Pty) Limited
RRL	Randgold Resources Limited
Simmer and Jack	Simmer and Jack Mines, Limited
Viking Pony/Kabusha	Viking Pony Properties 359 (Pty) Limited/Kabusha Mining and Finance (Pty) Limited
WAL	Western Areas Limited

Investment	Market value R000
13 312 480 shares in RRL (48.2%)⁽¹⁾	1 671 166
3 521 513 shares in DRD Gold Limited ("DRD") (Formerly Durban Roodepoort Deep, Limited)	115 566
202 500 shares in Western Areas Limited ("WAL")	8 894
2 321 115 shares in Kelgran Limited ("Kelgran")	1 741
37 139 298 shares in JCI Limited ("JCI")	21 169
2 613 615 JCI Limited 50% Unsecured Convertible Redeemable Debentures ("JCI Debentures")	2 718
Total	1 821 254

Investment	Market value R000
12 360 000 shares in RRL (44.7%)⁽¹⁾	1 551 598
21 513 shares in DRD	706
202 500 shares in WAL	8 894
2 321 115 shares in Kelgran	1 741
37 139 298 shares in JCI	21 169
2 613 615 JCI Debentures	2 718
Total	1 586 826

Investment	Market value R000
4 000 000 shares in RRL^{(1) (2)}	412 500
5 361 613 shares in WAL	227 064
2 321 115 shares in Kelgran	488
37 560 613 shares in JCI	6 010
50 651 JCI Debentures	63
13 100 000 shares in Pan Palladium	28 166
Total	674 291

Recipients	Number of shares (Pre-split)	Number of shares (Post-split)	Proceeds R’000
Disposals - 2002 ⁽¹⁾ - Other ⁽²⁾	952 481	1 904 962	64 326
Balance at 1 January 2003	12 360 000	24 720 000
Disposals - 2003 - Randgold - Other ⁽³⁾	1 575 000 325 000	3 150 000 650 000	218 298 55 060
Disposals - 2004 - Randgold - The JCI Group ^{(2) (3)} - Other ⁽³⁾	1 606 000 3 405 000 1 124 000	3 212 000 6 810 000 2 248 000	200 205 294 912 154 693
Disposals - 2005 - The JCI Group ^{(2) (3)} - Other ⁽³⁾	1 875 000 450 000	3 750 000 900 000	311 726 92 813
Balance at 31 December 2005	2 000 000	4 000 000 ⁽⁴⁾	1 392 033
Disposal post year-end ⁽⁴⁾	2 000 000	4 000 000	412 500
	-	-	1 804 533

Net asset value at 31 December 2005	(R000)
4 000 000 shares in RRL ⁽¹⁾	412 500
5 361 613 shares in WAL	227 064
2 321 115 shares in Kelgran	488
37 560 613 shares in JCI	6 010
50 651 JCI Debentures	63
13 100 000 shares in Pan Palladium	28 166
Investments	674 291
Other long-term assets	202 158
Loans: - JCI Group	122 780
Claims:
- JCI Group	720 006
- Other	20 441
Property, plant and equipment	433
Total current assets	9 218
Total assets	1 749 327
Total long-term liabilities	96 529
Total current liabilities (including taxation)	133 420
Total liabilities	229 949
Outside shareholders’ interests	46 309
Net assets	1 473 069
Shares in issue	74 813 128
Net asset value per share (cents)	1 969

	R&E	JCI
	(R million)	(R million)
Goldfields Limited (“Goldfields”)	230	1,644
R&E	-	186
Matodzi Resources Limited	-	51
JCI	78	-
Other listed investments	12	29
Boschendal wine estate	-	140
Other long-term assets	154	133
Prospecting Rights	400	350
Jaganda	-	284
Investec Loan Agreement Profit Share	-	(388)
Taxation	(16)	(220)
Net current assets	4	(163)
Post Retirement provision	(34)	-
Sub - Total	827	2,046

	R&E	JCI
Total (carried from above)	827	2,046
Settlement	1,200	(1,200)
Total after settlement	2,027	846

	R&E	JCI
Ordinary shares in issue	74,8	2,224,8
Estimated number of shares to be cancelled	(3,0)	(317,0)
Net shares in issue	71,8	1,907,8

	R&E	JCI
NAV per share at settlement of R1.2 billion	28.22	0.44
NAV per share at settlement of R1.5 billion	32.40	0.29

	R&E	JCI
- Assuming a R1.2 billion settlement	71%	29%
- Assuming a R1.5 billion settlement	81%	19%

	JCI’s share-holding in R&E	R&E’s share-holding in JCI
number of shares (millions)	6.2	223.4
Share value in Rands	30.00	0.35
Valuation (R’million)	186	78

	Before ¹ Published	Adjustments ³	After the proposed transaction ²
	R’000	R’000	R’000
Income
Dividends and interest received	12 049	27 650 ⁴	39 699
(Loss)/Profit on sale of investments	404	-	404
Claims	322 624	-	322 624
Fair value adjustment of investments	(4 016)	-	(4 016)
Other income	2	-	2
Profit on relinquishment of contiguous rights	-	395 000 ⁵	395 000
Profit on sale of subsidiary	1 421	-	1 421
Expenses
Interest expense	(18)	-	(18)
Depreciation and amortisation	(156)	-	(156)
Impairments	(48 275)	-	(48 275)
Write-off of listed investments and other long-term assets	(100 522)	-	(100 522)
Foreign exchange loss	(945)	-	(945)
Exploration and corporate expenditure	(27 811)	(1 510) ⁶	(29 321)
Net profit before taxation	154 757	421 140	575 897
Taxation	(38 628)	(63 378) ⁷	(102 006)
Net profit after taxation	116 129	357 762	473 891
Minority share	28 268	(161 080) ⁸	(132 812)
Profit attributable to ordinary shareholders	144 397	196 682	341 079
Profit per share (cents)	193	263	456
Headline earnings per R&E share (cents)	10	-	10
Weighted average number of shares	74 813 128	74 813 128	74 813 128

	Before ¹ Published	Adjustment ³	After the proposed transaction ²
	R’000	R’000	R’000
ASSETS
*Non-current assets*	1 740 109	-	1 740 109
Property, plant and equipment	433	-	433
Investment in associate	674 291	-	674 291
Investments	720 006	-	720 006
Claims – JCI Group	20 441	-	20 441
Claims other	122 780		122 780
Other long-term assets	202 158	-	202 158
*Current assets*	9 218	395 000	404 218
Trade and other receivables	1 697	-	1 697
Cash and cash equivalents	7 521	395 000 ⁴	402 521
Total assets	1 749 327	395 000	2 144 327
EQUITY AND LIABILITIES
*Capital and reserves*	1 519 378	338 130	1 857 508
Shareholders’ equity	1 473 069	185 862 ⁵	1 658 931
Minorities	46 309	152 268 ⁶	198 577
*Non-current liabilities*	96 529	-	96 529
Provision for post-retirement benefits	43 969	-	43 969
Deferred taxation	52 560	-	52 560
*Current liabilities*	133 420	56 870	190 290
Trade and other payables	8 993	1 510 ⁷	10 503
Taxation	124 427	55 360 ⁸	179 787
Total equity and liabilities	1 749 327	395 000	2 144 327

Earnings per R&E share (cents)	193	263	456
Headline earnings per R&E share (cents)	10	-	10
Net asset value per R&E share (cents)	1 969	248	2 218
Net tangible asset value per R&E share (cents)	979	248	1 228
Number of R&E shares in issue	74 813 128	74 813 128	74 813 128

FOCHVILLE WEST & EAST PROJECT
PROSPECTING RIGHT
WITH DEPARTMENT OF MINERALS AND ENERGY, GAUTENG REGION
REFERENCE NO GP 30/5/1/1/2 (56) PR
EXECUTED ON 04.11.05 AND VALID FOR 5 YEARS
		REGISTRATION		AREA
REF	FARM	DIVISION	PORTION	(HECTARES)
1	Cardoville 358	IQ	Mineral Area 3 on Portion 3	76.1623

2	Cardoville 358	IQ	Mineral Area 1 on Portion 1	43.7201

3	Cardoville 358	IQ	Portion 2	52.8842

4	Cardoville 358	IQ	Portion 4	75.5954

5	Cardoville 358	IQ	Remaining Extent	127.3952
	Area for CARDOVILLE 358 IQ in Ha			375.7571
6	Cardoville 364	IQ	Remaining Extent of Portion 1	321.211

7	Cardoville 364	IQ	Portion 9 of 3 (Now portion of 11)	103.9578

7	Cardoville 364	IQ	Portion 10 of 5 (Now portion of 11)	69.1629

8	Cardoville 364	IQ	Remaining Extent of Portion 3	358.1697

9	Cardoville 364	IQ	Remaining Extent of Portion 5 of 3	139.4112

10	Cardoville 364	IQ	Portion 4 of 3	2.5696
	Area for CARDOVILLE 364 IQ in Ha			994.4822
11	Doornpoort 347	IQ	Portion 51 of 21	30.6116
	Area for DOORNPOORT 347 IQ in Ha			30.6116
12	Droogeheuvel 521	IQ	Portion 10 of 5	40.6924
13	Droogeheuvel 521	IQ	Remaining Extent	117.1551
14	Droogeheuvel 521	IQ	Portion 2	362.3287
	Area for DROOGEHEUVEL 521 IQ in Ha			520.1762
15	Elandsfontein 144	IQ	Portion 12 of 6	176.6954
16	Elandsfontein 144	IQ	Portion 10 of 5	176.6926
	Area for ELANDSFONTEIN 144 IQ in Ha			353.388
17	Foch 150	IQ	Portion 3	66.4098
18	Foch 150	IQ	Portion 6 of 4	29.9786
19	Foch 150	IQ	Remaining Extent of Portion 4	55.6746
20	Foch 150	IQ	Portion 13 of 5	13.9612