royalgold.compage ii june 2019 ni 43-101 technical report on resources and reserves, golden star...

June 2019

NI 43-101 Technical Report on Resources and

Reserves, Golden Star Resources, Wassa Gold

Mine, Ghana

Report Date: June 20, 2019

Effective Date: December 31, 2018

Golden Star Resources Ltd. 150 King Street West

Suite 1200

Toronto ON, M5H 1J9, Canada

Qualified Persons

Martin Raffield, P.Eng

Mitch Wasel, MAusIMM (CP)

Philipa Varris, MAusIMM (CP)

Page ii

June 2019

NI 43-101 Technical Report on Resources and Reserves, Golden Star Resources, Wassa Gold Mine, Ghana

Table of Contents

1 Executive Summary................................................................................................ 1

1.1 Introduction ..........................................................................................................................1

1.2 Property Description and Ownership ...................................................................................2

1.3 Geology and Mineralization ................................................................................................2

1.4 Exploration Status ................................................................................................................3

1.5 Mineral Resources ...............................................................................................................4

1.6 Mineral Reserves .................................................................................................................6

1.7 Mining … .............................................................................................................................7

1.8 Recovery Methods ...............................................................................................................8

1.9 Infrastructure ........................................................................................................................8

1.10 Market Studies and Contracts ..............................................................................................9

1.11 Social and Environmental Aspects ......................................................................................9

1.12 Capital and Operating Costs ..............................................................................................10

1.13 Economic Analysis ............................................................................................................10

2 Introduction .......................................................................................................... 12

2.1 Scope of Technical Report .................................................................................................12

2.2 Qualified Persons ...............................................................................................................13

2.3 Site Visits ...........................................................................................................................13

3 Reliance on Other Experts ................................................................................... 14

4 Property Description and Location .................................................................... 15

4.1 Location of Mineral Concessions ......................................................................................15

4.2 Mineral Titles and Agreements ..........................................................................................19

4.3 Surface Rights ....................................................................................................................19

4.4 Royalties and Encumbrances .............................................................................................19

4.5 Historic Environmental Liability and Indemnity ...............................................................20

4.6 Permits and Authorization .................................................................................................20

5 Accessibility, Climate, Local Resources, Infrastructure and Physiography .. 21

5.1 Accessibility .......................................................................................................................21

5.2 Physiography and Vegetation ............................................................................................21

5.3 Land Use and Proximity to Local Population Centres .......................................................21

5.4 Local Resources and Infrastructure ...................................................................................22

5.5 Climate and Length of Operating Season ..........................................................................23

5.6 Wassa… .............................................................................................................................23

5.7 Hwini-Butre/Benso/Chichiwelli ........................................................................................24

6 History ................................................................................................................... 25

6.1 Wassa … ............................................................................................................................25

6.2 Hwini-Butre, Benso and Chichiwelli .................................................................................26

6.3 Historic Mineral Resource and Reserve Estimates ............................................................27

6.4 Historic Mine Production ...................................................................................................28

7 Geological Setting and Mineralization ............................................................... 30

7.1 Regional Geology ..............................................................................................................30

Page iii

June 2019


7.2 Local Geology and Mineralization ....................................................................................33

8 Deposit Types ........................................................................................................ 49

8.1 Wassa… .............................................................................................................................49

8.2 Hwini-Butre .......................................................................................................................52

8.3 Benso…..............................................................................................................................53

8.4 Chichiwelli… .....................................................................................................................54

9 Exploration ............................................................................................................ 56

9.1 Introduction ........................................................................................................................56

9.2 Wassa… .............................................................................................................................56

9.3 Hwini-Butre .......................................................................................................................59

9.4 Benso and Chichiwelli .......................................................................................................60

10 Drilling ................................................................................................................... 62

10.1 Open Pit .............................................................................................................................62

10.2 Underground ......................................................................................................................65

11 Sample Preparation, Analyses and Security ...................................................... 66

11.1 Sample Preparation ............................................................................................................66

11.2 Sample Despatch and Security ...........................................................................................66

11.3 Laboratory Procedures .......................................................................................................66

11.4 Quality Control and Quality Assurance Procedures ..........................................................71

11.5 Specific Gravity Data .........................................................................................................71

12 Data Verification .................................................................................................. 74

12.1 Introduction ........................................................................................................................74

12.2 Data verification by GSR ...................................................................................................74

12.3 Analytical QA/QC .............................................................................................................74

13 Mineral Processing and Metallurgical Testing .................................................. 92

13.1 Historical Testing ...............................................................................................................92

13.2 Recent Metallurgical Testwork ..........................................................................................92

13.3 Testwork Findings .............................................................................................................95

14 Mineral Resources .............................................................................................. 107

14.1 Introduction ......................................................................................................................107

14.2 Resource Estimation Procedures ......................................................................................107

14.3 Resource Database ...........................................................................................................108

14.4 Grade Shell Modelling .....................................................................................................110

14.5 Statistical Analysis and Variography ...............................................................................118

14.6 Block Model and Grade Estimation .................................................................................130

14.7 Model Validation and Sensitivity ....................................................................................137

14.8 Mineral Resource Classification ......................................................................................140

14.9 Mineral Resource Estimate ..............................................................................................142

15 Mineral Reserves ................................................................................................ 144

15.1 Cut-off Grade Estimate ....................................................................................................144

15.2 Mineral Reserve Statement ..............................................................................................144

Page iv

June 2019


16 Mining Methods .................................................................................................. 147

16.1 Open Pit Mining ...............................................................................................................147

16.2 Underground Mining .......................................................................................................159

16.3 Combined Underground and Open Pit Mining Schedule ................................................177

17 Recovery Methods .............................................................................................. 179

17.1 Flow Sheet Description ....................................................................................................179

17.2 Historical Plant Production ..............................................................................................182

17.3 Future Plant Production ...................................................................................................184

18 Infrastructure ..................................................................................................... 185

18.1 Site Layout .......................................................................................................................185

18.2 Electrical Infrastructure ...................................................................................................185

18.3 Mine Services...................................................................................................................189

18.4 Dewatering .......................................................................................................................190

18.5 Workshops .......................................................................................................................192

18.6 Waste Disposal.................................................................................................................193

18.7 Tailings Storage Facilities................................................................................................196

18.8 Tailings Storage Facility 2 ...............................................................................................200

19 Market Studies and Contracts .......................................................................... 204

19.1 Market Studies .................................................................................................................204

19.2 Contracts ..........................................................................................................................204

20 Environmental Studies, Permitting and Social or Community Impact ........ 205

20.1 Relevant Legislation and Required Approvals ................................................................205

20.2 International Requirements ..............................................................................................211

20.3 Environmental and Social Setting ....................................................................................212

20.4 Environmental and Social Management ..........................................................................234

20.5 Environmental and Social Issues .....................................................................................237

20.6 Closure Planning and Cost Estimate ................................................................................239

21 Capital and Operating Costs ............................................................................. 240

21.1 Capital Costs ....................................................................................................................240

21.2 Operating Costs ................................................................................................................240

22 Economic Analysis .............................................................................................. 242

22.1 Inputs and Assumptions ...................................................................................................242

22.2 Taxes and Royalties .........................................................................................................242

22.3 Cash Flow Model and Project Economic Results ............................................................242

22.4 Sensitivity Analysis .........................................................................................................244

Page v

June 2019


23 Adjacent Properties ............................................................................................ 245

24 Other Relevant Data and Information ............................................................. 246

25 Conclusions and Recommendations ................................................................. 247

26 References ........................................................................................................... 251

27 Date and Signatures ........................................................................................... 257

List of Tables

Table 1-1 Mineral Resource estimate as of December 31, 2018 ..............................................5

Table 1-2 Mineral Reserve estimate as of December 31, 2018 ................................................6

Table 4-1 Mining leases, prospecting leases and mining permits ..........................................16

Table 5-1 Communities neighbouring Wassa Mine ...............................................................22

Table 6-1 Historic mineral reserve estimates 2012 to 2017 ...................................................27

Table 6-2 Satellite Gold Ltd. production history ....................................................................28

Table 10-1 Exploration data used for the Mineral Resource models .......................................63

Table 11-1 Specific gravity testing results ...............................................................................72

Table 11-2 Specific gravity from underground drill holes .......................................................72

Table 12-1 Summary of analytical quality control data from 2014 to early 2017 ...................78

Table 12-2 CRM for 2003 to 2007 (TWL) ...............................................................................85

Table 12-3 Geostats CRM for 2008 to 2012 (SGS) .................................................................85

Table 12-4 Gannet CRM for 2008 to 2012 (SGS) ....................................................................86

Table 12-5 Gannet CRM for 2013 (SGS) .................................................................................86

Table 12-6 Gannet CRM for 2014 to 2017 (SGS) ....................................................................87

Table 12-7 Gannet CRM for 2014 to 2017 (Wassa Site Lab) ..................................................87

Table 12-8 Gannet CRM for 2018 (Intertek) ............................................................................87

Table 12-9 Blank Sample Summary Statistics 2011 to Q1 2018 .............................................88

Table 12-10 Gannet CRM for Quarter Core Sample Analysis (Intertek) ...................................89

Table 12-11 Summary HARD Plot Results for Quarter Core Sample Analysis ........................89

Table 12-12 Summary HARD Plots 2013 Round Robin Results ...............................................90

Table 12-13 Round-robin Descriptive Statistics .........................................................................90

Table 12-14 Round-robin Descriptive Statistics 2017 ................................................................90

Table 12-15 Summary HARD Plots 2017 Round Robin Results SGS - TWL ..........................91

Table 13-1 Ore zones represented by the variability samples ..................................................93

Table 13-2 Summary and location of testwork samples ...........................................................94

Table 13-3 Screened head assay results ....................................................................................95

Table 13-4 Elemental and chemical analysis results ................................................................96

Table 13-5 Summary of diagnostic leach results ......................................................................97

Table 13-6 Results of Crushability Tests: UCS and CWi ........................................................99

Table 13-7 Results of 2015 BWi and Ai Tests .......................................................................101

Table 13-8 Gravity Gold Recovery Test Results ....................................................................103

Table 13-9 Whole Ore Leach and CIL test results .................................................................103

Table 13-10 Leach test results and reagent consumptions .......................................................104

Table 13-11 Overall gravity leach recoveries ...........................................................................105

Page vi

June 2019


Table 13-12 Reconciliation of assay and back-calculated head grades from testwork ............106

Table 13-13 Comparative settling test results ..........................................................................106

Table 14-1 Wassa drill hole database as of December 2018 ..................................................108

Table 14-2 Hwini-Butre drill hole database as of December 2018 ........................................109

Table 14-3 Benso drill hole database as of December 2012 ..................................................109

Table 14-4 Chichiwelli drill hole database as of 2012 ...........................................................109

Table 14-5 Modelling extents .................................................................................................112

Table 14-6 Modelling parameters ...........................................................................................113

Table 14-7 Summary Gold Statistics of Assays and Composites ...........................................118

Table 14-8 Comparison of Uncapped and Capped Gold Composite Grades .........................119

Table 14-9 Local variogram orientations and anchor point (AP) locations ...............................121

Table 14-10 Local variogram models by domain .....................................................................122

Table 14-11 Descriptive statistics for Hwini-Butre modelled domains (uncapped & capped) 122

Table 14-12 Variogram parameters for Hwini-Butre ................................................................127

Table 14-13 Descriptive statistics for Benso modelled domains (capped) ...............................127

Table 14-14 Descriptive statistics for simplified Hwini-Butre modelled domains (capped) ...128

Table 14-15 Variogram parameters for the Benso zones .........................................................128

Table 14-16 Descriptive statistics for Chichiwelli modelled domains (capped) ......................129

Table 14-17 Chichiwelli high grade capping ............................................................................129

Table 14-18 Variogram parameters for Chichiwelli zones .......................................................130

Table 14-19 Block model parameters .......................................................................................130

Table 14-20 Bulk density ..........................................................................................................131

Table 14-21 Block model definition using GEMS convention ................................................131

Table 14-22 Northern model estimation parameters ................................................................132

Table 14-23 Adoikrom block model parameters ......................................................................133

Table 14-24 Father Brown Zone block model parameters ........................................................133

Table 14-25 Hwini-Butre ellipsoid search neighbourhood parameters ....................................134

Table 14-26 Hwini-Butre rock density .....................................................................................134

Table 14-27 Benso block model parameters ............................................................................135

Table 14-28 Benso ellipsoid search neighbourhood parameters ..............................................135

Table 14-29 Benso rock density ...............................................................................................135

Table 14-30 Chichiwelli block model parameters ....................................................................136

Table 14-31 Chichiwelli ellipsoid search neighbourhood parameters .....................................136

Table 14-32 Chichiwelli rock density .......................................................................................136

Table 14-33 Mineral Resource estimate as of December 31, 2018 ..........................................143

Table 15-1 Cut-off Grade estimate .........................................................................................144

Table 15-2 Mineral reserve estimate as of December 31, 2018 .............................................145

Table 16-1 Wassa Pit Optimization Input Parameters ............................................................148

Table 16-2 Wassa Open Pit Design Geotechnical Parameters ...............................................150

Table 16-3 Total material movement by stage for open pit mining .......................................153

Table 16-4 Current equipment fleet ........................................................................................157

Table 16-5 Joint Sets used for Stope Design ..........................................................................159

Table 16-6 Q Index (Q’) estimate ...........................................................................................160

Table 16-7 Modified stability number (N’) for longitudinal stope .........................................161

Table 16-8 Hydraulic radius of stope geometry .....................................................................161

Table 16-9 Total Workforce by LoM Year ............................................................................174

Page vii

June 2019


Table 16-10 Equipment fleet ....................................................................................................175

Table 16-11 Estimated Ventilation Requirement .....................................................................177

Table 16-12 Open pit and underground production schedule ..................................................177

Table 17-1 Key Plant Design and Operating Parameters .......................................................180

Table 17-2 Overview of Historic Plant Performance .............................................................183

Table 18-1 Current and future underground loads .................................................................189

Table 18-2 Waste stockpile current capacities .......................................................................194

Table 18-3 TSF stage storage capacities ................................................................................202

Table 20-1 Primary Environmental Approvals Required for Mining Operations ..................206

Table 20-2 Environmental Approvals Obtained for the Wassa Mine ....................................209

Table 20-3 Baseline Study Identified Hydrogeological Units (MEL, 1996c) ........................214

Table 20-4 Interpreted Packer test results ..............................................................................215

Table 20-5 Overview of Local Communities .........................................................................232

Table 21-1 Capital cost schedule ............................................................................................240

Table 21-2 Operating costs .....................................................................................................241

Table 22-1 Economic Model ..................................................................................................243

Table 22-2 NPV5% Sensitivity ................................................................................................244

List of Figures

Figure 4-1 Location of Wassa Mine in Ghana, West Africa ...................................................15

Figure 4-2 Location of GSR operations and mining lease boundaries ....................................16

Figure 4-3 Location of operations and infrastructure in relation to concession boundaries ....18

Figure 6-1 Historic Wassa Mine gold production ....................................................................29

Figure 7-1 Location of the Wassa Mine on the Ashanti Belt ...................................................32

Figure 7-2 Total magnetic intensity reduced to pole of the Ashanti Belt ................................34

Figure 7-3 Compilation of geochronology dating from the Ashanti Belt ................................35

Figure 7-4 Deformational history of the Ashanti Belt .............................................................36

Figure 7-5 Mine geology ..........................................................................................................37

Figure 7-6 Eburnean folds and foliations from the Wassa Mine Starter Pit ............................39

Figure 7-7 Eburnean folds and foliations from the Wassa Mine B-Shoot Pit .........................40

Figure 7-8 Vertical section of the Wassa Main deposit (19975N) ..........................................41

Figure 7-9 Vertical section showing the tabular nature of the ore zones (20000N) ................43

Figure 7-10 Vertical section showing the >1.5g/t Au shell .......................................................44

Figure 7-11 Vertical section showing the >0.4 g/t Au and >1.5g/t Au grade shells ..................45

Figure 7-12 Regional geology of the Hwini-Butre, Benso and Chichiwelli concessions ..........48

Figure 8-1 Geology of the Ashanti belt with location of major gold deposits .........................50

Figure 8-2 Syn-Eoeburnean veins from the B-Shoot, 242 and South-East zones ...................52

Figure 8-3 Different mineralization styles underlying the Hwini-Butre concession ...............53

Figure 8-4 Mineralized shear zones occurring on the Benso concession ................................54

Figure 8-5 Chichiwelli mineralization .....................................................................................55

Figure 9-1 Soil geochemistry anomalies ..................................................................................56

Figure 9-2 Wassa airborne magnetic interpretation .................................................................58

Figure 11-1 Transworld Laboratories sample processing flowsheet .........................................68

Figure 11-2 Intertek sample processing flowsheet ....................................................................70

Figure 12-1 HARD plot comparing fire assay and BLEG for field duplicates ..........................75

Page viii

June 2019


Figure 12-2 HARD plot of all coarse rejects (2011) from SGS .................................................76






Figure 12-8 HARD plot of all coarse rejects (2017) from SGS and Intertek ............................83

Figure 12-9 HARD plot of all coarse rejects (2018) from Intertek ...........................................84

Figure 13-1 View looking east-west of metallurgical sample locations ....................................93

Figure 13-2 Comparative indicated deportment of gold from diagnostic leach results .............98

Figure 13-3 Variation of UCS and CWi result with depth (relative level) ..............................100

Figure 13-4 2015 Ball Mill Bond Work Index against sample depth (relative level) .............101

Figure 13-5 2015 Abrasion Index against sample depth (relative level) .................................102

Figure 13-6 Leach recovery kinetic curves ..............................................................................105

Figure 14-1 Example of Structural ‘Form’ Surfaces ...............................................................111

Figure 14-2 North-facing cross sections showing structural form (18950N and 19170N) .....112

Figure 14-3 Oblique view of final volumes .............................................................................114

Figure 14-4 Mineral Resource wireframes and drill hole locations for the HwiniButre .........115

Figure 14-5 Mineral Resource wireframes and drill hole locations for the Benso deposits ....117

Figure 14-6 Mineral Resource wireframes and drillhole locations for Chichiwelli ................118

Figure 14-7 Probability Plot and Capping Sensitivity Plot ......................................................120

Figure 14-8 Adoikrom Log Probability .....................................................................................123

Figure 14-9 Adoikrom histogram ..............................................................................................124

Figure 14-10 Father Brown Log Probability .............................................................................125

Figure 14-11 Father Brown Histogram ......................................................................................126

Figure 14-12 South-North swath plot comparing various estimation sensitivities ....................137

Figure 14-13 Adoikrom swath plot ............................................................................................139

Figure 14-14 Father Brown Zone swath plot .............................................................................139

Figure 16-1 Wassa Pit Optimization Results ...........................................................................149

Figure 16-2 Wassa Main Topography as of December, 2017 .................................................149

Figure 16-3 Plan view of pits showing outlines of Cut 3 and 242 pushbacks .........................150

Figure 16-4 Pit design progression ..........................................................................................151

Figure 16-5 Cross section locations .........................................................................................152

Figure 16-6 Section A-A' (Fig. 16-5) showing block model ...................................................152

Figure 16-7 Section B-B' (Fig. 16-5) showing block model ....................................................153

Figure 16-8 Tonnes mined by month .......................................................................................155

Figure 16-9 Open pit schedule plan and isometric views ........................................................156

Figure 16-10 Wassa Main dump design in relation to final pit design ......................................158

Figure 16-11 Barton’s Q-Index Chart ........................................................................................160

Figure 16-12 Stope axes measurements .....................................................................................162

Figure 16-13 Stope stability graph .............................................................................................162

Figure 16-14 Phase 2 model: crown and sill pillar strength factor ............................................163

Figure 16-15 Long section looking east of open pit and underground as-built .........................164

Figure 16-16 Plan view of as-built development and stoping ...................................................165

Figure 16-17 Photograph of Starter Pit portal area ....................................................................166

Figure 16-18 Cross-section of main decline ..............................................................................167

Page ix

June 2019


Figure 16-19 As-built and planned ore reserve development (looking east) .............................168

Figure 16-20 Isometric view of as-built and planned development (looking NE) ....................168

Figure 16-21 As-built and planned development and stoping (looking east) ............................169

Figure 16-22 Isometric view of as-built and planned development and stoping (looking NE) .170

Figure 16-23 Process plant, tailings transfer and paste backfill plant locations ........................171

Figure 16-24: Paste Backfill Plant Process Flow Diagram ..........................................................173

Figure 16-25 Underground mining schedule .............................................................................174

Figure 16-26 Open pit and underground production schedule ..................................................178

Figure 17-1 Current Wassa plant flowsheet .............................................................................181

Figure 18-1 Wassa site layout ..................................................................................................186

Figure 18-2 UG Electrical substation, office and workshop area ............................................187

Figure 18-3 Current and future primary reticulation installations ..........................................188

Figure 18-4 Current dewatering long section (excluding F-Shoot for clarity) ........................190

Figure 18-5 Final dewatering long section ..............................................................................191

Figure 18-6 Pit catchments ......................................................................................................192

Figure 18-7 Current waste dump locations and volumes .........................................................194

Figure 18-8 Waste dump slope designs for operations and rehabilitation ...............................195

Figure 18-9 419 Waste dump location and elevation at LoM ..................................................196

Figure 18-10 GSWL TSF 1, TSF 1 extension and TSF 2 Cell 1 (March 2018) ........................197

Figure 18-11 TSF1 and TSF2 layout as per Knight Piesold report ...........................................199

Figure 20-1 Map of the Pra Basin showing the approximate location of the Project site .......213

Figure 20-2 Conceptual Groundwater Model ..........................................................................217

Figure 20-3 Groundwater flow direction .................................................................................218

Figure 20-4 Model boundaries and 3D model .........................................................................219

Figure 20-5 Modelled groundwater level recovery (Wassa Main) ..........................................221

Figure 20-6 Dewatering cone at life of mine ...........................................................................222

Figure 20-7 Paste pH vs NPR for pit, waste and underground samples ..................................225

Figure 20-8 NPR vs %S for pit, waste and underground samples ...........................................225

Figure 20-9 Conceptual Geo-environmental model (E-W cross section) ................................226

Figure 20-10 Conceptual Geo-environmental model (N-S layout cross section) ......................227

Figure 20-11 Ficklin diagram showing composition of underground mine leachate ................228

Figure 22-1 NPV5% Sensitivity ..............................................................................................244

Page x

June 2019


Abbreviations

AAS Atomic Absorption Spectroscopy

ALS ALS Minerals in Ghana-Kumasi

ANFO Ammonium Nitrate Fuel Oil

AP Acid Potential

ARD Acid Rock Drainage

Au Gold

BDG BD Goldfield Ltd

BLEG Bulk Leach Extractable Gold

BWi Bond Ball Mill Work Index

CIL Carbon-in-Leach

CIM Canadian Institute of Mining, Metallurgy and Petroleum

CMCC Community Mine Consultative Committees

COG Cut-off Grade

CRM Certified Reference Material

CYAP Community Youth Apprenticeship Program

CSA Canadian Securities Administrators

DD Diamond Drilling

EIA Environmental Impact Assessment

EIS Environmental Impact Statement

EMP Environmental Management Plan

EPA Environmental Protection Agency

FOS Factor of Safety

FS Feasibility Study

G&A General and Administrative

GEMS Gemcom Software

GSOPP Golden Star Oil Palm Plantation

GSSTEP Golden Star Skills Training and Employability Program

GSR Golden Star Resources Ltd.

GSWL Golden Star Wassa Ltd.

HARD Half Absolute Relative Difference

HDPE High-density polyethylene

HG High grade

HL Heap Leach

HBM Hwini-Butre Minerals Ltd

ICOLD International Commission on Large Dams

Page xi

June 2019


L Level

LG Low Grade

LoM Life of Mine

MDM MDM Engineering Group Limited

MSO Datamine Mining Shape Optimizer

NaCN Sodium Cyanide

NAG Non Acid Generating

NI 43-101 CSA’s National Instrument 43-101– ‘Standards of Disclosure for

Mineral Projects’

NPV Net Present Value

OK Ordinary Kriging

QA/QC Quality Assurance / Quality Control

QP Qualified Person pursuant to NI 43-101

RAB Rotary Air Blast

RAP Resettlement Action Plan

RC Reverse Circulation

SEDAR System for Electronic Document Analysis and Retrieval

SG Specific Gravity

SGS SGS Laboratories in Tarkwa/Lakefield

SGL Satellite Goldfields Ltd

SJR St Jude Resources (Ghana) Ltd

SRK or SRK

(Canada) SRK Consulting (Canada) Inc.

SRK (UK) SRK Consulting (UK) Limited

TSF Tailings Storage Facility

TSF1 Tailings Storage Facility No. 1

TSF2 Tailings Storage Facility No. 2

UCS Uniaxial compressive strength

UG Underground

U/Pb Uranium/lead

WGL Wexford Goldfields Limited

WRC Water Resources Commission

Page xii

June 2019


Units

deg. or degrees Degrees Celsius

g/t Grams per tonne

Ha Hectares

kg Kilogram

km Kilometre

kPa Kilopascal

kV Kilovolt

kW Kilowatt

kWh Kilowatt hour

kWh/t Kilowatt hour per tonne

l/s litres per second

m Metre

m/d Metres per day

m/s Metres per second

m3 Cubic metre

Ma Million years

ML Mining Lease in report mm millimeter

Mt Million tonnes

MPa Megapascal

mRL Meters Reduced Level

Mtpa Million tonnes per annum

MVA Mega-volt-ampere

oz Troy ounce

RL Reduced Level

t Metric tonne

t/d Tonnes per day

$ or US$ US Dollars

V Volt

June 2019


1 Executive Summary

1.1 Introduction

The Wassa Gold Mine is located near the village of Akyempim in the Wassa East District, in the

Western Region of Ghana. It is located 80 km north of Cape Coast and 150 km west of the capital

Accra. The property lies between latitudes 5°25’ and 5°30’ north and between longitudes 1°42’

and 1°46’ east. Golden Star Wassa Ltd. (“GSWL”) owns the rights to mine the Wassa, Benso and

Hwini-Butre concessions. Golden Star Resources Ltd. (“GSR”, “Golden Star” or the

“Company”), a Canadian federally incorporated gold mining and exploration company producing

gold in Ghana, owns a 90% interest in GSWL with the Government of Ghana owning the

remaining 10%.

This technical report summarizes the technical information that is relevant to support the disclosure

of a Mineral Reserve Statement for mine pursuant to Canadian Securities Administrators’ (“CSA”)

National Instrument 43-101 (“NI 43-101”). It presents the assumptions and designs at a level of

accuracy that is required to demonstrate the economic viability of the mineral resources.

Summary:

• Proven and Probable underground mineral reserves estimated at a $1,250/oz gold price,

as of December 31, 2018, are 7.5 Mt at an average grade of 3.95 g/t containing 949,000

ounces of gold.

• Proven and Probable open pit mineral reserves estimated at a $1,250/oz gold price, as of

December 31, 2018, are 9.9 Mt at an average grade of 1.57 g/t containing 500,000 ounces

of gold.

• Measured and Indicated mineral resources estimated at a $1,450/oz gold price, as of

December 31, 2018, are 43.8 Mt at an average grade of 2.40 g/t containing 3.4 million

ounces of gold. Measured and Indicated resources are inclusive of reserves.

• Inferred mineral resources estimated at a $1,450/oz gold price, as of December 31, 2018,

are 53.4 Mt at an average grade of 3.76 g/t containing 6.4 million ounces of gold.

• Stockpile processing of 1.2 Mt at an average grade of 0.63 g/t containing 24 thousand

ounces of gold.

• A 10 year production life.

• A metallurgical process recovery of 95%, yielding 1.4 million recovered ounces.

• Revenue calculations based on a gold price of US$1,300/ounce.

• Total underground development capital costs estimated at $50 million.

• Total open pit development capital costs estimated at $109 million

• Total underground sustaining capital costs estimated at $65 million.

• Total open pit sustaining capital costs estimated at $32 million.

• $218 million post-tax free cash flow.

June 2019


• $175 million post tax NPV at 5% discount rate.

• $671/oz life of mine (“LoM”) cash operating cost.

• $814/oz life of mine mine-site all in sustaining cost.

1.2 Property Description and Ownership

The Wassa area has witnessed several periods of local, small-scale and colonial mining activity

from the beginning of the 20th century. Mining of quartz veins and gold bearing structures are

evident from the numerous pits and shafts covering the Wassa lease area.

The Wassa Mine was originally developed as a 3 Mtpa open pit heap leach (“HL”) operation with

forecasted LoM gold production of approximately 100,000 ounces per annum. The first material

from the pit was mined in October 1998. After approximately one year of production, it became

evident that the predicted HL gold recovery of 85% could not be achieved, mainly due to the high

clay content of the resource and poor solution management.

In 2001, the project was put up for sale and GSR acquired the Wassa assets. Upon completion of

the acquisition of Wassa Mine by GSR, further exploration programs were undertaken. These

exploration programs formed part of a Feasibility Study (“FS”) that was completed in July 2003,

which demonstrated the economic viability of reopening and expanding the existing open pits and

processing the material through a conventional carbon-in-leach (“CIL”) circuit. The Wassa Mine

has been operating as a conventional CIL milling operation since April 2005.

1.3 Geology and Mineralization

The Wassa property lies within the southern portion of the Ashanti Greenstone Belt along the

eastern margin within a volcano-sedimentary assemblage located close to the Tarkwaian basin

contact. The eastern contact between the Tarkwaian basin and the volcano-sedimentary rocks of

the Sefwi group is faulted, but the fault is discrete as opposed to the western contact of the Ashanti

belt where the Ashanti fault zone can be several hundred meters wide.

The Wassa lithological sequence is characterized by lithologies belonging to the Sefwi Group and

consisting of intercalated meta-mafic volcanic and meta-diorite dykes with altered meta-mafic

volcanic and meta-sediments which are locally characterized as magnetite rich, banded iron

formation like horizons (Bourassa, 2003). The sequence is characterized by the presence of

multiple ankerite-quartz veins which are sub-parallel to the main penetrative foliation. The

lithological sequence is also characterized by Eoeburnean felsic porphyry intrusions on the south-

eastern flank of the Wassa mine fold.

The Wassa mineralization is subdivided into a number of domains, namely: F Shoot, B Shoot, 242,

South East, Starter, 419, Mid East, and Dead Man’s Hill (“DMH”). Each of these represents

discontinuous segments of the main mineralized system. The South- Akyempim (“SAK”) deposits

are located approximately 2 km to the southwest of the Wassa Main deposit on the northern end

of a well-defined mineralized trend parallel to the Wassa Main trend. The mineralization is hosted

in highly altered multi-phased greenstone-hosted quartz-carbonate veins interlaced with

sedimentary pelitic units. The SAK mineralization is subdivided into a number of domains as well,

SAK 1, 2 and 3.

June 2019


Gold mineralization within the Wassa Mine is structurally controlled and related to vein densities

and sulphide contents. The mineralization at Wassa is quite old and has been affected by several

phases of deformation since emplacement. Two major folding events effect the gold mineralization

which was likely emplaced early on in the deformational history of the deposit. Gold

mineralization was later remobilized into the hinges of a tight folding event and was later folded

around the deposit scale fold which influences the current day open pit configuration. The

remobilized gold mineralization in the hinges of the tight folding event are the primary

underground mining targets GSR is currently mining and are referred to as shoots, B and F being

the two main zones. These zones plunge to the south at approximately 20 degrees and it is this

controlling factor that has been and continues to be the main tool used to project the exploration

drilling targets to the south. GSR has now extended the known high grade (“HG”) mineralization

600 meters to the south of the current underground reserve and continues to drill the zone further

to the south where it remains open.

1.4 Exploration Status

Exploration drilling commenced in February 1994 and, by March 1997, a total of 58,709 m of

reverse circulation (“RC”) and diamond drilling (“DD”) had been completed.

In March 2002, GSR started an exploration program as part of a due diligence exercise following

the ratification of a confidentiality agreement with the creditor of Satellite Goldfields Limited

(“SGL”), and the mining lease was purchased later in the year. The exploration program consisted

mainly of pit mapping and drilling below the pits to test the continuity of mineralization at depth.

Exploration drilling resumed in November 2002 under GSR with the aim to increase mineral

reserves and resources for the FS which was completed in 2003.

Simultaneous with the resource drilling program that targeted resource increases in the pit areas,

GSR also undertook grass roots exploration along two previously identified mineralized trends.

The 419 area was delineated south of the main pits and the SAK anomaly, a soil target that had

never been previously drilled, was discovered west of the main pits. Deep auger campaigns were

also undertaken in the Subri River Forest Reserve, located in the southern portion of the Wassa

Mining lease.

From 2002 to December 2018, exploration and grade control drilling completed approximately

28,000 holes, totaling just over 1 million meters of drilling.

In March and April 2004, a high resolution, aerial geophysical survey was carried out over the

Wassa Mining Lease and surrounding Prospecting and Reconnaissance Licenses. The surveys

consisted of 9,085 kilometres of flown lines covering a total area of 450 km2. Flight lines were

spaced at varying distances between 50 to 100 m depending on the survey type. The geophysical

surveys identified several anomalies with targets being prioritized on the basis of supporting

geochemical and geological evidences.

Drilling is carried out by a combination of DD, RC and RAB techniques. In general, the RAB

method, which has a depth capability of 30 m, is used at early stages for follow up to soil

geochemical sampling and during production for testing contacts and extensions of mineralization.

To further test the prospective structures and anomalies defined from soil geochemistry and RAB

drilling results, RC drilling is used. RC drilling is typically carried out along drill lines spaced

between 25 and 50 m apart with maximum drilled depths of 100 to 125 m, depending on the ground

water table. The DD method is used to provide more detailed geological data in those areas where

June 2019


more structural and geotechnical information is required. Generally, the deeper intersections are

also drilled using the DD method and, as a result, most section lines contain a combination of RC

and DD.

Sampling is typically carried out along the entire drilled length. For RC drilling, samples are

collected every meter and then combined into 3 m composites. Should any 3 m composite samples

return a significant gold grade assay, the individual 1 m samples are then sent separately along

with those from the immediately adjacent samples. DD samples are collected, logged and split

with a diamond rock saw in maximum 1 m lengths. The core is cut according to mineralization,

alteration or lithology and is split into two equal parts along a median to the foliation plane. The

sampling concept is to ensure a representative sample of the core is assayed. The remaining half

core is retained in the core tray, for reference and additional sampling if required.

Sample assays are then performed at either SGS Laboratories (“SGS”) in Tarkwa or Transworld

(now Intertek) Laboratories (“TWL”), also based in Tarkwa. Both laboratories are independent of

GSR. GSR has used both laboratories and regularly submits quality control samples to each for

testing purposes. Both laboratories are ISO certified (ISO/IEC 17025:2005) for testing and

analysis. SGS has been accredited since September 2015 and Intertek has been accredited since

December 2017. Specific gravity (“SG”) determinations were carried out by GSR at the core

facility using a water immersion method.

Quality control measures are typically set in place to ensure the reliability and trustworthiness of

exploration data, and to ensure that it is of sufficient quality for inclusion in the subsequent Mineral

Resource estimates. Quality control measures include written field procedures and independent

verifications of aspects such as drilling, surveying, sampling and assaying, data management and

database integrity.

The field procedures implemented by GSR are comprehensive and cover all aspects of the data

collection process. At Wassa, each task is conducted by appropriately qualified personnel under

the direct supervision of a qualified geologist. The measures implemented by GSR are considered

to be consistent with industry best practice.

1.5 Mineral Resources

The following section presents the combined open pit and underground Mineral Resource estimate

for the Wassa Main and satellite deposits. Mineral Resources are reported inclusive of the material

which makes up the Mineral Reserve. The Mineral Resource Statement is presented in accordance

with the guidelines of NI 43-101.

GSR commissioned SRK to construct a mineral resource models with estimated gold grades for

the Wassa Main and HBB deposits. The Wassa long- and short-range models were a joint effort

by GSR geologists and SRK’s offices in Canada, the UK and Moscow. The mineral resource

classification and statement was conducted by GSR under the supervision of S. Mitchel Wasel, a

Qualified Person (“QP”).

The “reasonable prospects for eventual economic extraction” requirement generally imply that the

quantity and grade estimates meet certain economic thresholds and that the Mineral Resources are

reported at an appropriate cut-off grade “COG”), taking into account extraction scenarios and

processing recoveries.

June 2019


In order to determine the quantities of material offering “reasonable prospects for economic

extraction”, GSR used a pit and underground optimizers and reasonable mining assumptions to

evaluate the proportions of the block model (Indicated and Inferred blocks) that could be

“reasonably expected” to be mined.

The optimization parameters are based on actual costs from the operations. The reader is cautioned

that the results from the pit optimization are used solely for the purpose of testing the “reasonable

prospects for economic extraction” and do not represent an attempt to estimate Mineral Reserves.

GSR considers that the blocks located within the conceptual pit envelopes show “reasonable

prospects for economic extraction” and can be reported as a Mineral Resource.

Table 1-1 shows the combined Mineral Resource statement for the Wassa Main and satellite

deposits.

Table 1-1 Mineral Resource estimate as of December 31, 2018

In declaring the Mineral Resources for the Wassa Main and HBB deposits, the following are noted:

• The identified Mineral Resources in the block model are classified according to the CIM

definitions for Measured, Indicated and Inferred categories and are constrained within a

Whittle pit shell using a gold price of US$1,450/oz and below December 2018

topographic surface. The Mineral Resources are reported in-situ without modifying

factors applied.

• The Wassa open pit Mineral Resource estimate is based on a COG of 0.4 g/t Au reported

within a conceptual Whittle shell. Pit optimization using industry standard software has

been undertaken on the Mineral Resource models using appropriate slope angles, process

recovery factors and costs.

• The Wassa underground portion of the Mineral Resource estimate is based on a COG of

2.1 g/t Au.

• The Father Brown Underground Mineral Resource has been estimated below the

US$1,450 per ounce of gold pit shell using a COG of 3.2 g/t Au.

• The Mineral Resource models have been depleted using appropriate topographic

surveys.

June 2019


• Block model tonnage and grade estimates were classified according to the CIM

Definition Standards for Mineral Resources and Mineral Reserves (December 2005).

The basis of the Mineral Resource classification included confidence in the geological

continuity of the mineralized structures, the quality and quantity of the exploration data

supporting the estimates, and the geostatistical confidence in the tonnage and grade

estimates.

• All figures are rounded to reflect the relative accuracy of the estimate.

• Mineral Resources are not Mineral Reserves and do not have demonstrated economic

viability.

1.6 Mineral Reserves

The Wassa Mineral Reserves were estimated based on the Mineral Resources that are classified as

Measured and Indicated. The Mineral Reserves are summarized in Table 1-2.

The Mineral Reserves have been prepared in accordance with CIM standard definitions for Proven

Mineral Reserves and Probable Mineral Reserves. The Indicated Mineral Resources reported

above include those Mineral Resources modified to estimate the Mineral Reserves.

The Mineral Reserves have been estimated using accepted industry practices for open pit and

underground mines, including the identification of the optimal final ore envelopes based on the

selected mining methods, appropriate modifying factors and COG estimates based on detailed cost

estimation. The identified ore bodies were subjected to detailed mine design, scheduling and the

development of a cash flow model incorporating the Company’s technical and economic

projections for the mine for the duration of the LoM plan. A gold price of US$1,250/oz was used

for the Reserve estimation.

Any mineralization which occurs below the COG or is classified as an Inferred Mineral Resource

is not considered as Mineral Reserves and is treated as mineralized waste for the purposes of the

LoM plan. The Wassa Mineral Reserve Statement is as of December 31, 2018.

Table 1-2 Mineral Reserve estimate as of December 31, 2018

Notes to Mineral Reserve estimate:

• Mineral Reserve estimates reflect the Company’s reasonable expectation that all

necessary permits and approvals will be obtained and maintained. Mining dilution and

mining recovery vary by deposit and have been applied in estimating the Mineral

Reserves.

June 2019


• Mineral Reserves are the economic portion of the Measured and Indicated Mineral

Resources. Mineral Reserve estimates include mining dilution at grades assumed to be

zero.

• The Mineral Reserve estimate was prepared under the supervision of Dr. Martin Raffield,

Chief Technical Officer for the Company and QP.

• The Mineral Reserves at December 31, 2018 were estimated using a gold price

assumption of $1,250 per ounce.

• The slope angles of all pit designs are based on geotechnical criteria as established by

external consultants. The size and shape of the pit designs are guided by consideration

of the results from a pit optimization program.

• COGs have been estimated based on operating cost projections, mining dilution and

recovery, government royalty payment requirements and applicable metallurgical

recovery. The marginal COG used for the open pit estimate is 0.7 g/t Au and the break-

even COG used for the underground estimate is 2.4 g/t Au.

• Numbers may not add due to rounding.

1.7 Mining …

The mine plan assumes underground mining will continue until the end of 2024 and that open pit

mining will start in 2023 and continue until 2028.

Total ore material amounts to 18.6 Mt at an average grade of 2.46 g/t Au estimated to recover 1.5

million ounces of gold. This consists of: 7.5 Mt at 3.95 g/t underground; 9.9 Mt at 1.57 g/t open

pit; and 1.2 Mt at 0.63 g/t from stockpiles..

The underground design utilizes a twin decline access from the base of the open pit which reduces

the access to the mineralization from the surface topography. The selected mining method is sub-

level open stoping utilizing unconsolidated waste backfill in the early stages and converting to

cemented paste backfill in 2020.

A staged approach to ventilation is taken, with raises to surface completed in line with the

development and production schedule.

Mining is undertaken using trackless, diesel powered equipment including twin boom jumbos for

development and long hole drills for production drilling. The approach to materials handling to

surface is through a combination of 17 t capacity loaders and 40 t capacity trucks.

The underground mine reached commercial production in January 2017 and, since that time, has

shown consistent improvement in ore tonnage generation capacity. By the end of 2018, it was

producing at a rate of 3,500 tpd, with plans to increase to close to 4,000 tpd in 2020.

Geotechnical conditions in the mine are very good with consistently low rates of unplanned

dilution being achieved in the stopes and development.

As with most underground operations, there is a risk of flooding of the underground operation and

significant effort has been expended to understand and mitigate this risk. Extensive

hydrogeological studies have been conducted to inform the mine hydrogeological and geo-

environmental models. The pit catchment areas have been reduced through earthworks and

June 2019


drainage diversions around the pit areas to ensure minimal water ingress into the pits from the

surrounding areas. The sump capacities in the pits below the holings into the underground have

been designed to contain 100-year storm events over 24 hour periods. The pumping systems from

the pits and underground are in the process of being upgraded to ensure enough capacity is

available to dewater the sumps and the underground mine during and following such rain events.

A conventional approach to open pit mining is envisaged, employing excavators and trucks which

are typical for this type and style of gold mineralization. Drilling and blasting of rock are conducted

over bench heights of 5 or 6 m and explosives are delivered to the hole by the manufacturer. Oxide

or weathered material is generally only required to be lightly blasted or, in some areas, can be

excavated without blasting. Hydraulic excavators are used in conjunction with conventional

blasting practice to mine a 2.5 or 3.0 m flitch height. Broken rock is loaded to 91 t capacity

off-highway haul trucks to a central stockpile or to the waste dump.

1.8 Recovery Methods

The Wassa Mine was originally developed as a 3 Mtpa (8,200 tpd) open pit HL operation which

operated from 1998 to 2001. In 2003, the plant was converted into a milling CIL circuit and

started operation in 2005, treating a mixture of oxidized, fresh ores and material reclaimed from

the HL pads.

From 2005 to the end of 2017, the plant processed at a rate 7,400 t/d primarily from oxide and

fresh open pit feed. The earlier part of this period was characterized by high availability of oxide

material and, from mid-2016, by the feeding of higher average grade underground material.

Currently, the treatment plant is processing ore from the underground operation in addition to low

grade (“LG”) pit stockpiles at a rate of about 4,000 t/d. The plant consists of a crushing, milling,

gravity and CIL gold recovery circuit.

Test work and recent experience indicates that the hardness and abrasiveness of the underground

feeds are no higher than the historic feed and, in fact, are indicated on the majority of the variability

samples to be slightly softer (lower Bond work index) and less abrasive, although the differences

are generally only minor. Variability and crushability test work shows that, although there are

slightly higher levels of sulphides present with depth, the processing characteristics of the ore from

the underground do not materially change.

1.9 Infrastructure

There are two tailings storage facilities (“TSF”) that will accommodate the anticipated tailings

production. Deposition now predominantly occurs in TSF 2, with TSF 1 largely complete.

The design of TSF 2 is cellular and a combination of compacted soil liner and high-density

polyethylene (“HDPE”) liner are incorporated in the design. Construction of TSF 2 Cell 1

commenced in July 2016, with the first deposition in May 2017. While TSF 2 [Cell 1] has received

both Environmental Protection Agency (“EPA”) and Minerals Commission permits, further

permitting is required for each new cell. A Supplementary Environmental Impact Assessment

(“EIA”) process is underway for Cell 2. The approved TSF 2 design provides a total storage

capacity of 41 Mt, although only 18.7 Mt is required for the current LoM, excluding the reduction

in TSF requirements as a result of paste backfill use. With 11 stages incorporated in the current

June 2019


engineering design, the capacity required for the current LoM will be accommodated by the first

six stages.

An assessment by SRK (2015) of the geotechnical investigation, embankment stability and

seepage analyses attests that these design elements are relevant to the size and scope of the project

and demonstrate satisfactory results for the storage facilities. The water management at the Wassa

site has been such that discharges to the receiving environment from the TSF have not been

required since 2010. The Wassa operation has an approved detoxification plant to treat elevated

cyanide concentrations that is available for the treatment of water should a discharge be required;

however, the water balance model for the current configuration of the site indicates that under

normal conditions the processing activities operate under net negative water balance.

Grid power from the national power supplier GridCo comes from a 161 kV line to local substation

where power is transformed down through a 33 MVA transformer to 34.5 kV. Two feeders

provide electricity to the GSR Wassa Mine substation at 34.5 kV. Backup generator power is

available for the plant and the underground mine.

1.10 Market Studies and Contracts

Gold is a freely traded commodity on the world market for which there is a steady demand from

numerous buyers. GSR and affiliates has a long-term sales contracts in place with RGLD Gold

AG, an affiliate of Royal Gold (“RGLD”) and a South Africa gold refinery. Under the purchase

and sale agreement with RGLD, GSWL is required to sell 10.5% of its gold production at 20% of

the London p.m. fix gold price to RGLD (“Stream”). When GSR meets the gold delivery threshold

GSWL the Stream will be reduced to 5.5% of gold production at 30% of the London p.m. fix gold

price. The remainder of the gold production is sold at prevailing market prices. The gold is shipped

in the form of doré bars. The sale price is generally based on the London p.m. fix on the day of the

shipment to the refinery.

GSR has a number of contracts in place with local, national, and international contractors for the

supply of materials and services.

1.11 Social and Environmental Aspects

The GSWL operational area is within the moist tropical rainforest area of the Western Region of

Ghana. The mean annual rainfall is in the order of 1,750 mm. The natural vegetation at the mine

site had been disturbed prior to mining by logging and farming activities. The total area of the

Wassa Mining Lease is 5,289 Ha, with approximately 595 Ha of disturbance having occurred as a

result of GSWL’s activities.

In 2015, GSWL commenced an EIA for an expansion to the operations to accommodate the Wassa

underground, Pit 3 expansion and associated expansion in waste rock storage. As the infrastructure

and operations permitted under this expansion occur virtually entirely within the footprint of the

existing Wassa operations and compensated land buffers, the impact assessment found it unlikely

that any further impact to flora and fauna would result from the expansion. Following submission

of the Environmental Impact Statement (“EIS”) to the EPA in September 2016, the Environmental

Permit was invoiced in January 2017 and issued in October 2017.

The mine is in a rural area and there are no major urban settlements within 50 km by road. The

villages of Akyempim, Akyempim New Site (formerly Akosombo, which was resettled by the

June 2019


Company) and Kubekro are the closest to the Wassa operational area. The total population of these

communities is about 3,000. GSWL has undertaken the necessary compensation and resettlement

activities required for access to 1,293 Ha of land in the lease area, including the current operational

area, the site of TSF 2 and buffer zones.

As required by the Environmental Assessment Regulations, GSWL has submitted the required

three-year Environmental Management Plan (“EMP”) for its operations. The most recent EMP,

for the period 2018-2020, was submitted to the EPA in December 2017 and the Environmental

Certificate was invoiced in June 2018. The EMP covers the Wassa, Hwini-Butre, and Benso Mines

and all associated infrastructure, including the Hwini-Butre Benso Access Road. The previous

approved EMP continues to apply until such time as the EPA issues the new permit.

As required by its permitting conditions, a Reclamation Security Agreement was established

between the Company and the EPA in 2004. To make way for the TSF 2 construction, resettlement

of Togbekrom village and surrounding hamlets was undertaken in accordance with an approved

Resettlement Action Plan (“RAP”).

1.12 Capital and Operating Costs

Total capital of $255 million is comprised of:

• Underground development capital costs estimated at $50 million.

• Open pit development capital costs estimated at $109 million

• Underground sustaining capital costs estimated at $65 million.

• Open pit sustaining capital costs estimated at $32 million.

Mine operating costs include:

• $37/t-ore underground stoping cost;

• $6.00/t-ore paste backfill cost;

• $3,400/m underground development cost;

• $3.35/t open pit mining costs;

• $15/t to $25/t processing costs depending on throughput; and

• $5/t to $9/t G&A costs depending on throughput.

1.13 Economic Analysis

The mine has been evaluated on a discounted cash flow basis. The cash flow analysis was prepared

on a constant 2019 US dollar basis. No inflation or escalation of revenue or costs has been

incorporated.

Using a long-term gold price forecast of $1,300/oz, the post-tax free cash flow is $218 million

and the post-tax NPV5% is $175 million.

Life of mine cash operating cost is estimated at $671 per ounce and mine-site all-in sustaining

cost at $814.

June 2019


The NPV5% is most sensitive to changes in the gold price, plant head grade and operating costs

and least sensitive to capital costs.

June 2019


2 Introduction Golden Star is a Canadian federally-incorporated international gold mining and exploration

company, producing gold in Ghana, West Africa. The Wassa Gold Mine is located near the village

of Akyempim in the Wassa East District, in the Western Region of Ghana. It is located 80 km

north of Cape Coast and 150 km west of the capital Accra. The property lies between latitudes

5°25’ and 5°30’ north and between longitudes 1°42’ and 1°46’ east. GSWL owns the rights to

mine the Wassa, Benso and Hwini-Butre concessions. GSR owns a 90% interest in GSWL with

the Government of Ghana owning the remaining 10%.

This technical report summarizes the technical information that is relevant to support the disclosure

of a Mineral Reserve Statement for mine pursuant to NI 43-101. It presents the assumptions and

designs at a level of accuracy that is required to demonstrate the economic viability of the mineral

resources.

The Wassa underground mine achieved commercial production in January 2017 and open pit

mining was halted temporarily in January 2018.

The LoM plan reflects underground mining continuing until the end of 2024, and open pit mining

starting in 2023 and continuing until 2028.

Total ore material mined amounts to 18.6 Mt at an average grade of 2.46 g/t Au estimated to

recover some 1.4 million ounces of gold. It should be noted that the underground grade is

significantly higher than the open pit grade.

The underground design utilizes a twin decline access from the base of the open pit which reduces

the access to the mineralization from the surface topography. The selected mining method is sub-

level open stoping, utilizing unconsolidated waste backfill in the early stages and converting to

cemented paste backfill in 2020.

Mining is undertaken using trackless, diesel powered equipment including twin boom jumbos for

development and long hole drills for production drilling. The approach to materials handling to

surface is through a combination of 17 t capacity loaders and 40 t capacity trucks.

A conventional approach to open pit mining is envisaged, employing excavators and trucks which

are typical for this type and style of gold mineralization. Drilling and blasting of rock is conducted

over bench heights of 5 or 6 m and explosives are delivered to the hole by the manufacturer. Oxide

or weathered material is generally only required to be lightly blasted or, in some areas, can be

excavated without blasting. Hydraulic excavators are used in conjunction with conventional

blasting practice, to mine a 2.5 or 3.0 m flitch height. Broken rock is loaded to 91 t capacity

off-highway haul trucks to a central stockpile or to the waste dump.

2.1 Scope of Technical Report

This technical report is intended to support the December 2018 Mineral Resource and Reserve

estimate for GSWL. This Technical Report has been prepared in accordance with the requirements

of NI 43-101 – ‘Standards of Disclosure for Mineral Projects’, of the CSA and for filing on CSA’s

System for Electronic Document Analysis and Retrieval (“SEDAR”).

June 2019


2.2 Qualified Persons

Dr. Martin Raffield is the QP responsible for Sections 1 to 3, 6, 13, 15 to 19, and 21 to 27 of this

report. He is based in Toronto, Canada and employed by GSR as Chief Technical Officer. Dr.

Raffield has overall responsibility for the report.

S. Mitchel Wasel is the QP responsible for Sections 7 to 12 and Section 14 of this report. Mr.

Wasel is based in Takoradi, Ghana and is employed by GSR as Vice President of Exploration.

Philipa Varris is the QP responsible for Sections 4, 5 and 20 of this report. Ms. Varris is based in

Bogoso, Ghana and is employed by GSR as Vice President of Corporate Responsibility.

2.3 Site Visits

Martin Raffield visited site in April 2019.

Mitchel Wasel visited site in June 2019.

Philipa Varris visited site in June 2019.

June 2019


3 Reliance on Other Experts The preparation of this technical report has been undertaken by GSR staff; however, there are

disciplines where GSR was not the sole author or relied on specialists in a particular field. In these

cases, GSR’s QPs have reviewed and approved the work of other experts as follows:

• Environmental impact assessment studies undertaken by Golder Associates (Ghana and

South Africa);

• Geological long-range model wireframe prepared by SRK (UK);

• Geological short-range model wireframe and resource estimation prepared by SRK

(Moscow);

• Resource estimation and block model prepared by SRK (Canada);

• Tailings storage facility design and geotechnical assessments by Knight Piésold Ghana

(the engineer of record); and,

• Paste backfill studies carried out by Outotec (Canada) Ltd.

June 2019


4 Property Description and Location

4.1 Location of Mineral Concessions

The Wassa Mine is located near the village of Akyempim in the Wassa East District, in the Western

Region of Ghana. It is located 80 km north of Cape Coast and 150 km west of the capital Accra.

The property lies between latitudes 5°25’ and 5°30’ north and between longitudes 1°42’ and 1°46’

east. The location of the Wassa Mine is shown on Figure 4-1 and Figure 4-2.

The total area of the Wassa Mining Lease is 5,289 Ha, with approximately 595 Ha of disturbance

having occurred as a result of GSWL’s activities.

Figure 4-1 Location of Wassa Mine in Ghana, West Africa

(Source: United Nations 2008)

June 2019


GSWL currently holds three mining leases, namely Wassa, Hwini-Butre, and Benso. In addition,

GSWL holds several prospecting leases in the region. The mining and prospecting lease details

are summarized in Table 4-1.

Figure 4-2 Location of GSR operations and mining lease boundaries

(Source: GSR, 2018)

Table 4-1 Mining leases, prospecting leases and mining permits

Permit / Lease Permit No. Agency Date of Issue Expiry Date Comments

Wassa Mining Lease LVB 87618/94 Minerals

Commission 17/9/1992 16/9/2022

GSWL Mining

Operating Permit for

LVB 87618/94

00002281/19 Inspectorate

Division 2018-11-01 31/12/2019 Updated annually

Benso Mining Lease LVB 2681/07 Minerals

Commission 31/12/2012 30/12/2019

Renewal process scheduled to

start 3 months before expiry.

GSWL Mining


LVB 2681/07



Hwini-Butre Mining

Lease LVB1714/08

Minerals

Commission 31/12/2012 30/12/2018

Renewal process concluded.

Feasibility report and site plans

submitted to Mincom in Sep-

18.

GSWL Mining


LVB 1714/08



Dwaben (Safric)

Reconnaissance LVB1624/06

Minerals

Commission 02/02/2006 N/A

Renewal under application.

2019 Prospecting permit fees

paid Feb-19.

June 2019


Wassa Prospecting

/Exploration permit in

the Subri River Forest

Reserve

Part of the Wassa

ML

Forestry

Commission 14/12/2006 30/06/2007

Applied for Renewal – only

required when work is to be

done in forest.

Benso (Chichiwilli-

Amantin) Prospecting PL.2/1550

Minerals


Recommendation for renewal

forwarded to the sector

Minister from Mincom May-19

Ateiku - Twifo RL 3/31 Minerals



paid Feb-19.

Abura RL 2/135 Minerals

Commission 04/02/2010 12/12/2021

Esuaso (Kobra) PL 2/379 Minerals


Extension application

submitted Apr-15. 2019

Prospecting permit fees paid

Feb-19.

Manso PL 2/378 Minerals



submitted Sep-11. Processing

fees paid Dec-11. 2019


Feb-19.

Manso (Pacific) PL 2/337 Minerals



submitted-Mar-12. 2019


Feb-19.

Oseneso 1 & 2

Prospecting LVB 13975/06

Minerals

Commission 08/09/2006 01/03/2011

Extension under application.


paid Feb-19.

The map in Figure 4-3 shows the location of the various GSWL exploration properties and mining

leases, which incorporate the deposits as follows:

• Wassa mining lease: Wassa Main is an operating open pit gold mine comprising the

following mineralization domains: F Shoot, 419, B Shoot, 242, Starter, South-East, Mid-

East and Dead Man’s Hill. SAK comprises a number of deposits to the West of Wassa

Main, which are no longer mined.

• Benso mining lease: comprising the Subriso East (“SE”), Subriso West (“SW”), G-Zone,

C-Zone and I-Zone deposits.

• Hwini-Butre mining lease: comprising the Father Brown, Adoikrom and Dabokrom

deposits with only Father Brown having remaining open pit reserves.

• Chichiwelli exploration property: comprising two mineralized zones, Chichiwelli West

(“West Domain”) and Chichiwelli East (“East Domain”).

• Manso exploration property: located adjacent to Benso, to the east.

The properties and leases are spread out along a line trending approximately 80 km southwest of

the Wassa mine complex. There are sufficient access and surface rights for the operation with no

risks to site access or to the title of the leases.

June 2019


Figure 4-3 Location of operations and infrastructure in relation to concession boundaries

(GS Exploration, 2012)

June 2019


4.2 Mineral Titles and Agreements

A corporate body duly registered in Ghana can apply to the Minerals Commission for a renewable

exploration licence granting exclusive rights to explore for a particular mineral in a selected area

for an initial period not exceeding three years. When exploration has successfully delineated a

mineral reserve, an application may be made to the Minerals Commission for conversion to a

mining lease, granting a company the right to produce a specific product from the concession area.

Mineral rights in the Wassa concession have been granted to GSWL under the Minerals and

Mining Act, 2006 (“Act 703”) which is the governing legislation for Ghana’s minerals and mining

sector. As defined by Act 703, every mineral in its natural state in, under or upon land in Ghana,

rivers, streams, water-courses throughout the country, the exclusive economic zone, and an area

covered by the territorial sea or continental shelf, is the property of the Republic of Ghana and is

vested in the president in trust for the people of Ghana. By means of Act 703, land in the country

may be made the subject of an application for a mineral right in respect of a mineral specified in

the application.

The Wassa concession is a mining lease that was transferred to GSWL on September 17, 1992 by

the Government of Ghana with land registry number LVB 7618/94.

In order to confirm the Company’s title in its material mineral properties, the Company will from

time to time obtain legal opinions from its local Ghanaian counsel regarding such title. On

February 7, 2017, the Company received title opinions from Ghanaian counsel with respect to the

Wassa properties which confirmed that GSWL (as applicable) is the holder of the applicable

mineral rights in each property and that such mineral rights are in good standing and are subject

only to those statutory rights and options conferred on the Government by Act 703. In order to

render such opinions, Ghanaian counsel reviewed, among other things, the mining leases relating

to the material resource properties, and conducted official title searches at appropriate

governmental registries. In addition, the Company relies on its in-house tenement officers and the

services of local experts, including local external legal counsel, to ensure that its operating

subsidiaries in Ghana comply with applicable legal and regulatory requirements relating to the

ownership and operation of its material mineral properties and assets in Ghana.

4.3 Surface Rights

The Subri-Akyempim exploration concession was granted to Wassa Mineral Resources Ltd.

“WMRL”) in 1988. The current mining lease was assigned to SGL under Land Registry number

2033/1994 and transferred to Wexford Goldfields Limited “WGL”) in October 2002. The present

day Wassa mining lease is valid until September 16, 2022. The mining lease comprises an area of

52.89 km2 lying to the North and South of Latitudes 525’ and 530’ respectively and bounded to

the East and West by Longitudes 142’ and 146’ respectively. The lease is within the Wassa East

District of the Western Region of the Republic of Ghana.

4.4 Royalties and Encumbrances

The Wassa Mining Lease stipulates that a 5% royalty on gross revenue be paid on a quarterly basis

to the government as prescribed by the legislation. Royalties are based on production and are to be

paid through the Commissioner of Internal Revenue within thirty days from the end of each

quarter.

June 2019


In the press release dated May 7, 2015, GSR announced the securing of a $150 million financing

with Royal Gold, Inc. (“RGI”) and its wholly-owned subsidiary RGLD Gold AG (“RGLD”). The

$150 million financing consists of a $130 million stream transaction with RGLD and a further $20

million term loan from RGI.

The stream transaction requires Golden Star (a) to deliver 8.5% of all production to RGLD at a

cash purchase price of 20% of spot gold until 185,000 ounces have been delivered; (b) thereafter,

to deliver 5% of all production to RGLD at a cash purchase price of 20% of spot gold until an

additional 22,500 ounces have been delivered; and (c) thereafter, to deliver 3% of all production

to RGLD at a cash purchase price of 30% of spot gold.

4.5 Historic Environmental Liability and Indemnity

The Wassa operations were permitted under an EIA developed for SGL in 1998. The operation

was a HL operation fed by the Main pits complex comprising the interconnected South-East, 242,

F and B Shoots, South and Main South, and 419 pits.

In 2002, Golden Star purchased the fixed assets of the project and liabilities for the operations

transferred. In late 2005, Golden Star acquired St. Jude Resources (Ghana) Limited (“SJR”) and,

with it, the Hwini-Butre and Benso properties and their associated liabilities.

The predominant liabilities of the original SGL operations, including HL area and waste dumps,

have now been fully encompassed by the GSWL operations. Likewise, the development of the

HBB operations by GSWL saw the establishment of infrastructure that fully encompassed the

previous areas of disturbance of SJR. The establishment of the Reclamation Security Agreement

with the EPA in 2005 and the associated bond with the EPA addresses security for reclamation

and closure. There are no other legacy issues associated with the GSWL site.

4.6 Permits and Authorization

In addition to those specified on Table 4-1, GSWL currently holds the following major approvals

related to the Wassa, Benso and Hwini-Butre operations:

• Wassa operations (EPA/EIA/112) and expansions (EPA/EIA/322) including South

Akyempim pits (EPA/EIA/190).

• Hwini-Butre and Benso operations (EPA/EIA/175) and expansion (EPA/EIA/247).

• TSF 2 (EPA/EIA/383) and renewal (EPA/EIA/442).

• Wassa Expansion project including Wassa underground, and Main pits and waste dump

expansion (EPA/EIA/508).

• (EPA/EIA/508).

GSWL currently also holds the following permits; subject to renewal:

• Explosive Purchase for Mining Operations.

GSWL has undertaken a series of EIA studies on its concessions to support the permitting of its

various mining projects and therefore has considerable background data relating to the mining

areas to support required environmental permitting processes.

June 2019


5 Accessibility, Climate, Local Resources, Infrastructure and Physiography

5.1 Accessibility

The Wassa Mine is located near the village of Akyempim in the Wassa East District in the Western

Region of Ghana. It is 62 km north of the district capital, Daboase, and 40 km east of Bogoso. It

is located 80 km north of Cape Coast and 150 km west of the capital Accra. The main access to

the site is from the east, via the Cape Coast to Twifo-Praso road, then over the combined road-rail

bridge on the Pra River. There is also an access road from Takoradi in the south via Mpohor.

5.2 Physiography and Vegetation

The project area is characterized by gently rolling hills with elevations up to 1000 and 1100 m RL,

incised by an extensive drainage network. The natural vegetation is an ecotone of the moist, semi-

deciduous forest and wet rainforest zones. It has been degraded due to anthropogenic activities,

giving way to broken forest, thickets of secondary forest, forb re-growth, swamps in the bottom of

valleys, and cleared areas. Extensive subsistence farming occurs throughout the area, with

plantain, cassava, pineapple, maize, and cocoyam being the principal crops. Some small-scale

cultivation of commercial crops is also carried out, with cocoa, teak, coconuts and oil palm being

the most common. Forests patches are present on the steep slopes and in areas unsuitable for

agriculture.

Environmental assessments carried out in the project area over the last two decades (SGS 1996

and 1998, WGL 2004, GSR 2015, Geosystems 2013, and Golder 2016) indicate that the

biodiversity of the Wassa operational area is of low ecological significance and conservation

status.

5.3 Land Use and Proximity to Local Population Centres

The Wassa Mine, and its associated processing plant and TSF, is in a rural area and there are no

major urban settlements within 50 km by road. The villages of Akyempim, Akyempim New Site

(formerly Akosombo, which was resettled by the Company), Kubekro and Nsadweso are the

closest to the mine. The total population of these communities is about 3,000. The community of

Togbekrom has been resettled as part of the development, construction and operation of TSF 2.

The Benso and Hwini-Butre Mines are about 65 km and 35 km, respectively, north-north-west of

the Port of Takoradi and south-east of Tarkwa. The key communities within and outside the

concession are Subriso, Odumase, Ningo, Akyaakrom, Mpohor, Benso, and Anlokrom. The total

population of these communities is about 10,000. The Benso Township is approximately 5 km

from the Benso mine site to the south and the Mpohor Township is approximately 2 km west of

the Hwini-Butre mining site.

The population data/estimates for the larger communities located within the Wassa concession

boundaries are given in Table 5-1.

June 2019


Table 5-1 Communities neighbouring Wassa Mine

Community Divisional Area /

Paramountcy

Estimated Population

(SGS 1996)

Population

(GSS 2012)

Population

(WEDA 2013)

Akyempim

Mamponso

2,500 3,303 2,533

Akosombo N/A 166

Kubekro

Anyinabrem

300 772 335

Nsadweso 2,400 1,872 1,541

Togbekrom NM 674

NM= not measured in survey

Land uses in the vicinity of GSWL operations are predominantly rural with agricultural, forestry,

agroforestry (palm oil and rubber plantations), and unauthorized small-scale mining operations.

5.4 Local Resources and Infrastructure

There are four other mines in the vicinity of Tarkwa; namely, Ghana Manganese Company – Nsuta

andAnglo Gold Ashanti Iduapriem mines and Goldfields Ghana Limited - Damang and Tarkwa

mines.

The Wassa Mine itself is located in the Wassa Mining Lease, which covers an area of 52.89 km2.

Wassa Mine is an operating underground mine, with the required services, infrastructure, and

community support already in place. The following are relevant to the assessment of resources and

infrastructure:

• access to the project is via the public road network that extends on to the site;

• electricity and water are available;

• surface infrastructure in the area consists of a variety of government, municipal, and

other roads with good overall access;

• processing is carried out at the existing GSWL processing plant;

• tailings are stored in the existing GSWL TSF with deposition now primarily to TSF 2

with deposition to TSF 1 expected to be completed in 2019;

• waste rock generated at the site is placed in existing waste dumps adjacent to the Wassa

open pit with additional waste dump footprint expansion permitted; and

• the extensive history of mining in Ghana provides opportunities to obtain skilled

underground mine workers.

June 2019


5.5 Climate and Length of Operating Season

The climate in the project area is classified as wet semi-equatorial. The Inter Tropical Convergence

Zone (“ITCZ”) crosses the area twice a year, resulting in a bi-modal rainfall pattern, with peaks

in March to July and September to October. During the dry season months of November to

February, the climate is heavily influenced by the dry, dust-laden, northwest trade wind, known

locally as the Harmattan, which blows from the Sahara desert.

Analysis of available rainfall data, obtained from the Ateiku Meteorological survey (1944 to 2009)

indicates that the average annual rainfall is 1,996 ± 293 mm. The wettest month of the year is

generally June, with an average rainfall of about 241 ± 85 mm, whilst January is the driest month

of the year with an average rainfall of about 31 ± 35 mm. The wettest month on record is June

2009, when 475 mm of precipitation was recorded. Rainfall is mainly influenced by south-west

monsoon winds, which blow from the south-western part of the country towards the north-east.

Long term data from Ateiku compare well with the Wassa site (2003-2013) and the adjacent TSF 1

rainfall data (2007-2014), which also demonstrate the bi-modal rainfall pattern. Using data from

the GSWL weather station, the average annual rainfall has been estimated at about 1,750 mm and

the wettest month on record was June 2014, when 511.8 mm of precipitation was recorded at

TSF 1. A drier period, which is influenced predominantly by a sweep of the north-east traded

winds, is experienced between the month of November and February.

Annual potential evapotranspiration is estimated to be approximately 1,337 mm/year, indicating a

minimum precipitation excess of 288 mm/year. Rainfall exceeds potential evapotranspiration from

March to July and September to October, and groundwater recharge is most likely to be prevalent

during these periods. Relative humidity is fairly constant throughout the year, ranging from 88%

to 90%.

Under such climatic conditions, surface mining operations can continue year-round with short

breaks during storms, most of which are short-lived and may be experienced throughout most of

the year. Underground mining operations will not be directly affected by storms as long as effective

storm water management infrastructure is in place at surface to divert runoff from mine accesses.

5.6 Wassa…

The area is characterized by gently rolling hills with elevations up to 1000 and 1100 m RL; incised

by an extensive drainage network. The area comprises tropical rainforest and is relatively wet, with

many low-lying swampy areas. Extensive subsistence farming occurs throughout the area, with

plantain, cassava, pineapple, maize, and cocoyam being the principal crops. Some small-scale

cultivation of commercial crops is also carried out, with cocoa, teak, coconuts and oil palm being

the most common.

The mine is strategically located 38 km due east of the Bogoso and Prestea mines, which are also

owned by Golden Star. Paved roads are complete from the Coast to Twifu-Praso, some 28 km

from the project site. The Takoradi/Kumasi railway line passes through the village of Ateiku,

10 km north east of Wassa, but it has not been operational for several years.

June 2019


5.7 Hwini-Butre/Benso/Chichiwelli

These concessions can be accessed by tarred highway from Accra to Takoradi (approximately 4

hours) and from Takoradi to the southern boundary of the concessions by tarred highway (Takoradi

to Tarkwa road) and finally by dirt road. The southern portion of the Hwini-Butre concession is

covered extensively by large scale commercial palm oil plantations and is, therefore, crossed by

many roads and tracks which provide access to all areas.

The Chichiwelli vein deposits are located on the east side of the Bonsa river, just inside the Subri

River Forest Reserve, and about 7.5 km SE of Wassa Nkran (or about 25 km due east of Tarkwa)

on the Bonsa River. Access is via the dirt road branching to the SE from Abosso.

Road access to within 12 km of Hwini-Butre is good from the main Takoradi-Tarkwa highway

and then on unpaved but serviceable roads. From Hwini-Butre, the GSWL haul road provides good

access northwards through Benso, Chichiwelli, and on to Wassa.

The area is well populated, with the large villages of Mpohor and Edum Banso having populations

of 10,000 and 5,000, respectively. Outside of these, the area is extensively farmed by small scale

or family enterprises.

June 2019


6 History

6.1 Wassa …

The Wassa area has witnessed several periods of local small-scale and colonial mining activity

from the beginning of the 20th century with the mining of quartz veins and gold bearing structures

being evident from the numerous pits and adits covering the Wassa lease area.

From 1988, the property was operated as a small-scale mining operation with a gravity gold

recovery circuit by WMRL, a Ghanaian company. In 1993, WMRL was looking for a capital

partner to further develop the mining lease and invited the Irish companies Glencar Exploration

Limited (“Glencar”) and Moydow Ltd. to visit the concession. Following this visit, SGL was

formed between WMRL, Glencar and Moydow Ltd. The mining lease, which is valid for a 30-

year period expiring in 2022, was assigned by WMRL to SGL.

Extensive satellite imagery and geophysical interpretations were carried out and identified a strong

gold target (>1 g/t Au). Exploration drilling commenced in February 1994, and, by March 1997, a

total of 58,709 m of RC and DD had been completed. Construction of the Wassa Mine was initiated

in September 1998, after Glencar secured a US$42.5 million debt-financing package from a

consortium of banks and institutions.

The Wassa mine was originally developed as a 3 Mtpa open pit HL operation with forecasted LoM

gold production of approximately 100,000 ounces per annum. The first ore from the pit was mined

in October 1998.

After approximately one year of production, it became evident that the predicted HL gold recovery

of 85% in the oxide ore could not be achieved, mainly due to the high clay content of the ore and

poor solution management. After a number of attempts to improve the recovery, including

doubling the leach solution application rate, it was concluded that the achievable gold recovery on

oxide ore by HL was between 55 and 60%.

The combined effect of the lower than planned gold recovery and lull in the gold price at the time

resulted in the company not being able to service its debt to the banks. In early 2001, the banks,

together with Glencar, decided to sell the project to recover some of the accumulated debt.

GSR started negotiations to purchase the fixed assets of the project in mid-2000 and, by December

2000, the negotiations were at an advanced stage. In March 2001, a drilling program was initiated

as part of GSR’s final due diligence. The program was designed to test the GSR geological model

developed during its due diligence and to test the extensions to some of the HG ore bodies.

In April 2002, Golden Star concluded that the mineable reserve at Wassa was 30% lower than the

648,000 oz stated by SGL. This resulted in the renegotiation of the conditions of purchase of the

property. Agreement was finally achieved in early September 2002. GSR immediately commenced

an intensive exploration program with the aim to increase the in pit reserve to above 650,000 oz

and gain further confidence in the current reserves.

GSR has been actively mining and exploring the Wassa area since it took control of the property

in 2002 and has drilled over 28,000 RC, surface and underground DD holes totaling over 1 million

meters. As of the end of 2018, GSR has produced 1.91 million ounces of gold from this operation

and currently has a mine life of approximately 5 years.

June 2019


6.2 Hwini-Butre, Benso and Chichiwelli

The alluvial gold deposits in the immediate vicinity of Mpohor (Hwini-Butre) were important

historically and some early European reports indicate that the Dabokrom area may have been a

major source for gold sold at Elmina to the Portuguese explorers who first came to the region in

the late 1400s.

Direct European interest in the area probably dates to the late 1800s because this was a known

source of gold and it was close to Sekondi-Takoradi, which was to become a major port and

railhead city to service the inland gold operations at Tarkwa, Prestea and Obuasi. The area was

covered by exploration licences in the gold boom of 1898-1902 and the 1930s saw much more

sustained interest when virtually the whole area was under license; in many cases, to local

Ghanaian businessmen and entrepreneurs.

At Dabokrom, a vertical and inclined shaft was sunk by Oceania Consolidated in the mid 1930s to

intersect and follow the shallow dipping quartz veins. They continued to work on the property for

several years but stopped at the beginning of WW2 in late 1939. Earlier, a shaft was sunk just after

WW1 (1918) on a quartz vein at the Chichiwelli prospect at the very north end of the Benso

concession, just along the boundary of the Subri River Forest Reserve. Many collapsed adits and

shallow shafts are scattered over several parts of the concessions and they attest to European

activities, dating mainly to the 1930s.

It was not until the late 1980s that exploration attention was again directed to this area. The

Dabokrom concession was acquired BD Goldfields Limited (“BDG”). This group invited Danish

Company (Lutz Resources Limited) to work on the property. They carried out preliminary

exploration work in the early 1990s and then had the property transferred to Hwini-Butre Minerals

Limited (“HBM”), also controlled by Scandinavian investors. Shortly thereafter, HBM entered a

joint venture with Placer-Outokumpu who drilled several vertical holes in 1993 around the

Dabokrom area with a view to assessing the large-scale potential of the vein systems. They

concluded that the veins were too widely spaced and the intervening diorite host rock contained

little gold, so that large scale potential seemed limited.

SJR acquired Dabokrom in late 1994 and explored the area and managed the project up until early

2006; however, there was about a three-year hiatus on the work as the result of a legal dispute

between BDG, the Government of Ghana, and HBM. The dispute was finally resolved in

December 2005, prior to GSR’s acquisition of SJR. In March 2006, the concession was transferred

to First Canadian Goldfields Limited, a subsidiary of SJR, which in turn was a subsidiary of GSR.

To the north, extensive reconnaissance work (1989-92) by BHP Billiton (“BHP”) identified

significant soil geochemical anomalies at Chichiwelli, Subriso, Denerawah and Amantin. Some

follow-up work was carried out, especially at Chichiwelli where twelve drill holes were completed.

None of the targets were deemed large enough to meet BHP’s size threshold and they relinquished

all of their interests. Shortly thereafter, a local company, Architect Co-Partners, acquired a 150

km2 prospecting concession covering Amantin, Subriso and Chichiwelli. This also included a large

part of the Subriso River Forest Reserve, which was closed to exploration after 1996.

Fairstar Exploration Limited of Canada (“Fairstar”) took over the Benso concession in 1995 and

carried out extensive work, especially at Subriso and Amantin, where considerable drilling was

carried out under the management of the consulting company, CME (Ghana) Ltd. of Accra and

Vancouver, Canada. By the end of the decade, work on the concession had largely ceased because

June 2019


of a lack of funds. By mid-2001, SJR completed an agreement with Fairstar and took over the

exploration work.

From early 2002 to about mid-2004, SJR’s focus was in the Subriso area where substantial mineral

resources were outlined at two important prospects, Subriso East and West. Numerous other

prospects were located nearby, which were drill tested, as was the Amantin area, which had also

been drilled to a considerable extent by Fairstar. By early 2004, SJR was able to recommence work

on the Hwini-Butre concession. Work priorities included further evaluating existing targets and

identifying new prospects in the vicinity of Abada and Guadium at the north end of the Hwini-

Butre concession. For much of 2005, drilling was focused at the southern end of the concession.

This work included upgrading and expanding resources at Adoikrom and Father Brown and testing

other prospect areas such as Semkrom and Adoikrom North. In addition, in late 2004 and for much

of 2005 and early 2006, efforts were directed towards carrying out engineering, metallurgical and

environmental studies needed in an application for a mining lease to cover the main Benso and

Hwini-Butre prospects.

In 2005, considerable attention was also directed towards clearing all legal and title issues that

held up progress on the project. These efforts were finally successful in late 2005

contemporaneously with the acquisition of SJR by GSR. Since then, GSR has carried out more

detailed drilling in the areas of the main known occurrences.

The two Chichiwelli prospects are approximately 2 km apart and although the Bremang occurrence

appears to be a LG quartz vein, which has received little attention in the past, the Chichiwelli vein

deposit saw considerable work starting in the very early 1900s. In the early 1920s, fairly extensive

underground workings were established, including a decline to an inclined depth of about 260ft

(79m), and several crosscuts. Work was abandoned in 1924 after the mine was flooded.

6.3 Historic Mineral Resource and Reserve Estimates

In September 1997, consulting engineers Pincock, Allen and Holt completed a FS, which

determined a proven and probable mineable reserve of 17.6 Mt at 1.7 g/t Au, for a total of 932,000

contained ounces of gold.

Golden Star’s initial Resource statement for Wassa was released on December 31, 2002. The

Indicated Mineral Resource was 17.8 Mt at 1.3 g/t Au, for a total of 737,000 contained ounces of

gold and the Inferred Mineral Resource was 28.8 Mt at 1.2 g/t Au.

The initial Proven and Probable Reserve statement for Wassa was not released until the following

year and, as of December 31, 2003, totaled 16.1 Mt at 1.3 g/t Au, for a total of 656,000 contained

ounces of gold.

Table 6-1 shows Wassa Mines’ published mineral reserve estimates between 2012 and 2017.

Table 6-1 Historic mineral reserve estimates 2012 to 2017

Year End Tonnes [millions] Grade [g/t] Ounces Contained

[thousand]

Gold price

assumption [$/oz]

2012 32 1.44 1.47 1,450

2013 35 1.75 1.97 1,300

2014 24 2.04 1.58 1,200

2015 20 2.27 1.49 1,100

June 2019


2016 17 2.37 1.33 1,100

2017 19 1.98 1.20 1,250

6.4 Historic Mine Production

The Wassa Mine was originally developed as a 3 Mtpa, open pit, HL operation with forecasted

LoM gold production of approximately 100,000 ounces per annum. The first material from the pit

was mined in October 1998. After approximately one year of production, it became evident that

the predicted HL gold recovery of 85% could not be achieved, mainly due to the high clay content

of the resource and poor solution management. After a number of attempts to improve the

recovery, including increased agglomeration and doubling the leach solution application rate, it

was concluded that the achievable gold recovery by HL was between 55 and 60%. The combined

effect of the lower than planned gold recovery and lull in the gold price at the time resulted in

Glencar not being able to service its debt to the creditors. In early 2001, the creditors, together

with Glencar, decided to sell the project to recover some of the accumulated debt. Mining was

stopped at the end of October 2001. Irrigation of the HL with cyanide solution continued until

March 2002, after which rinsing of the heaps with barren solution continued until August 2002.

SGL’s mining operations at Wassa commenced in October 1998. Annual historic production from

October 1998 to October 2001 is as follows:

Table 6-2 Satellite Gold Ltd. production history

Year

Oct'98-Dec'99

Jan'00-Dec'00

Jan01-Oct'01

Total

Waste

Mined

(BCM)

2,956,000

3,038,000

1,379,000

7,372,000

Ore Mined

(BCM)

1,422,000

1,316,000

979,000

3,717,000

Ore Stacked

(t)

3,010,000

3,051,000

2,052,000

8,113,000

Stacked

Grade

(g/t)

1.75

1.63

1.57

1.66

Gold

Poured

(ozs)

87,034

98,010

64,019

249,063

The cash operating costs during the three years of mining operations were US$184 per ounce for

1999, US$223 per ounce for 2000 and US$252 per ounce for Q1 of 2001.

The high clay content of the weathered ore resulted in lower than the anticipated 85% gold

recoveries. As at October 2002, the overall gold recovery for the project was 57.5%. Substantial

changes to the HL operations were made during the three years of operation in an attempt to

improve on recoveries, which included lowering of stacking height and doubling of solution

application rates without much improvement in gold recoveries.

The combination of the low gold recoveries and slump in the gold price in the late nineties resulted

in inadequate cash flow from the operation to service debt and to pay the mining contractor on site.

The primary lenders liquidated the company and mining operations were halted in October 2001.

GSR started gold production from the Wassa – Hwini-Butre Benso (“Wassa-HBB”) operations in

2005 and has produced a total of 1.91 million ounces as of the end of 2018.

June 2019


Figure 6-1 Historic Wassa Mine gold production

(Source: GSR)

0

50,000

100,000

150,000

200,000

250,000

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

20

17

20

18

Ou

nce

s p

ou

red

June 2019


7 Geological Setting and Mineralization

7.1 Regional Geology

The regional geological setting of the Ashanti belt has been described by several authors

previously. The most recent publication describing the geological setting of the sub-region was

from Perrouty et al., in Precambrian Research in 2012.

The Ashanti greenstone belt in the Western Region of Ghana is composed primarily of

paleoproterozoic metavolcanic and metasedimentary rocks that are divided into the Birimian

Supergroup (Sefwi and Kumasi Groups) and the Tarkwa Group. Both units are intruded by

abundant granitoids (Figure 6-1) and host numerous hydrothermal gold deposits such as the Wassa,

Obuasi, Bogoso and Prestea mines and paleoplacer deposits such as the Tarkwa and Teberebie

Mines.

Allibone et al. (2002) separated the Paleoproterozoic Eburnean orogeny into two distinct phases

known as Eburnean I and II. This classification was revised by Perrouty et al. in 2012 who

proposed two distinct orogenic events, the Eoeburnean orogeny and the Eburnean orogeny. The

Eoeburnean orogeny predates the deposition of Tarkwaian sediments and is associated with a

major period of magmatism and metamorphism in the Sefwi Group basement. The Eburnean event

is associated with significant post-Tarkwaian deformation that affected both the Birimian

Supergroup and overlying Tarkwaian sediments. The Eburnean orogeny is associated with major

north-west to south-east shortening that developed major thrust faults, including the Ashanti Fault

along with isoclinal folds in Birimian metasediments and regional scale open folds in the

Tarkwaian sediments. These features are overprinted by phases of sinistral and dextral

deformational events that reactivated the existing thrust faults and resulted in shear zones with

strong shear fabrics.

The Birimian series was first described by Kitson (1918) based on outcrops located in the Birim

River (around 80 km east of the Ashanti Belt). Since this early interpretation, the Birimian

stratigraphic column has been revised significantly. Before the application of geochronology, the

Birimian super group was divided in an Upper Birimian group composed mainly of metavolcanics

and a Lower Birimian group corresponding to metasedimentary basins. Subsequent authors have

proposed synchronous deposition of Birimian metavolcanics. Most recently,

Samarium/Neodymium and U/Pb analyses have reversed the earlier stratigraphic interpretation

with the younger metasediments overlying the older metavolcanics. Proposed ages for the

metavolcanics vary between 2,162 ± 6 Ma and 2,266 ± 2 Ma. Detrital zircons in the metasediments

indicate the initiation of their deposition between 2,142 ± 24 Ma 2,154 ± 2 Ma. The Kumasi Group

was intruded by the late sedimentary Suhuma granodiorite at 2,136 ± 19 Ma (U/Pb on zircon,

Adadey et al., 2009).

The Tarkwa super group was first recognized by Kitson (1928) and consists of a succession of

clastic sedimentary units, which have been divided in four groups by Whitelaw (1929) and Junner

(1940). The Kawere Group located at the base of the Tarkwaian super group is composed of

conglomerates and sandstones with a thickness varying between 250 m and 700 m. The unit is

stratigraphically overlain by the Banket Formation, which is characterized by sequences of

conglomerates interbedded with cross-bedded sandstone layers, the maximum thickness of this

group being 400 m. The conglomerates are principally composed of Birimian quartz pebbles

June 2019


(>90%) and volcanic clasts (Hirdes and Nunoo, 1994) that host the Tarkwa Placer deposits. The

Banket formation is overlain by approximately 400 m of Tarkwa Phyllites. The uppermost unit of

the Tarkwa super group is the Huni Sandstone, comprised of alternating beds of quartzite and

phyllite intruded by minor dolerite sills that form a package up to 1,300 m thick (Pigois et al.,

2003). U/Pb and Pb/Pb geochronology dating of detrital zircons provide a maximum depositional

age of 2,132 ± 2.8 Ma for the Kawere formation and 2,133 ± 3.4 Ma for the Banket formation

(Davis et al., 1994; Hirdes and Nunoo, 1994). These ages agree with the study by Pigois et al.

(2003) that yielded maximum depositional age of 2,133 ± 4 Ma from 71 concordant zircons of the

Banket formation. According to all concordant zircon histograms (161 grains) and their

uncertainties, a reasonable estimation for the start of the Tarkwaian sedimentation could be as

young as 2,107 Ma.

Abundant granites and granitoids intruded the Birimian and Tarkwaian units during the

Paleoproterozoic. Eburnean plutonism in south-west Ghana can be divided into two phases

between 2,180 to 2,150 Ma (Eoeburnean) and 2,130 to 2,070 Ma (Eburnean) that is supported by

the current database of U/Pb and Pb/Pb zircon ages. Most of the granitoids intruded during both

phases correspond to typical Tonalite–Trondhjemite–Granodiorite suites. However, in the

southern part of the Ashanti Belt, intrusions within the Mpohor complex have granodioritic,

dioritic and gabbroic compositions.

Dolerite dykes oriented north-south and East northeast to West south-west that are generally less

than 100 m in thickness are abundant across the West African craton where they cross-cut Archean

and Paleoproterozoic basement. In south-western Ghana these dykes are well defined in magnetic

data where they are characterized by strong magnetic susceptibility. Dolerite dykes are observed

to cross-cut undeformed K-feldspar rich granites that formed during the late Eburnean, and are

overlain by Volta basin sediments with a maximum depositional age of 950 Ma (Kalsbeek et al.,

2008). These relationships constrain dyke emplacement to between 2,000 Ma and 950 Ma. In

contrast some older dolerite/gabbro dykes and sills were deformed during the Eburnean orogeny

and are dated at 2,102 ± 13 Ma (U/Pb on zircon, Adadey et al., 2009).

With the exception of some late Eburnean granitoids, dolerite dykes and Phanerozoic sediments,

all other lithologies have undergone metamorphism that generally does not exceed upper

greenschist facies. Studies on amphibole/plagioclase assemblages suggest the peak temperature

and pressure was 500 to 650C and 5 to 6 kbar (John et al., 1999), dated at 2092 ± 3 Ma (Oberthür

et al., 1998).

June 2019


Figure 7-1 Location of the Wassa Mine on the Ashanti Belt

(Perrouty et al., 2012)

June 2019


7.2 Local Geology and Mineralization

7.2.1 Introduction

The Wassa property lies within the southern portion of the Ashanti Greenstone Belt along the

eastern margin of the belt within a volcano-sedimentary assemblage located at proximity to the

Tarkwaian basin contact. The eastern contact between the Tarkwaian basin and the volcano-

sedimentary rocks of the Sefwi group is faulted, but the fault is discrete as opposed to the western

contact of the Ashanti belt where the Ashanti fault zone can be several hundred meters wide.

Deposition of the Tarkwaian sediments was followed by a period of dilation and the intrusion of

late mafic dykes and sills.

The lithologies of the Wassa assemblage are predominantly comprised of mafic to intermediate

volcanic flows which are interbedded with minor horizons of volcaniclastics, clastic sediments

such as wackes and magnetite rich sedimentary layers, most likely banded iron formations. The

volcano-sedimentary sequence is intruded by syn-volcanic mafic intrusives and felsic porphyries.

The magnetic signature of the Ashanti belt is relatively high in comparison to the surrounding

Birimian sedimentary basins such as the Kumasi basin to the west of the Ashanti belt and the

Akyem Basin to the East as illustrated in Figure 7-2.

Rock assemblages from the southern area of the Ashanti belt were formed between a period

spanning from 2,080 to 2,240 Ma as illustrated in Figure 7-4, with the Sefwi Group being the

oldest rock package and the Tarkwa sediments being the youngest. The Ashanti belt is host to

numerous gold occurrences, which are believed to be related to various stages of the Eoeburnean

and Eburnean deformational event. Structural evidences and relationships observed in drill core

and pits at Wassa would suggest the mineralization to be of Eoeburnean timing while other known

deposits in the southern portion of the Ashanti belt such as Chichiwelli, Benso and Hwini-Butre

are considered to be of Eburnean age.

The Eoeburnean deformation is best observed at Wassa where the deformational event has

produced a penetrative foliation with an associated lineation which is defined by mineral

alignments. A period of extension occurred between the Eoeburnean and Eburnean deformational

events which resulted in the formation of the Akyem Basin (Kumasi Group) to the northeast of the

Wassa Mine and the Tarkwa group to the west of the Wassa concession. Both metasedimentary

sequences of the Tarkwa and Kumasi group have not been affected by the penetrative foliation

observed at Wassa.

The Eburnean deformation is divided in multiple events which vary in number depending on the

authors as summarized in Figure 7-4. All deposits underlying the Wassa concession have been

affected by the Eburnean deformational events, the main penetrative foliation has been affected by

at least three Eburnean folding events which have resulted in a large scale refolded synform. The

main foliation is sub-vertical and oriented northeast to south-west on the south-eastern flank of the

Wassa mine fold whereas it is dipping at around 45° to the south-southeast on the north-west flank

of the Wassa mine fold.

June 2019


Figure 7-2 Total magnetic intensity reduced to pole of the Ashanti Belt

(Modified from Perrouty et al., 2012)

June 2019


Figure 7-3 Compilation of geochronology dating from the Ashanti Belt


June 2019


Figure 7-4 Deformational history of the Ashanti Belt


June 2019


7.2.2 Wassa

The Wassa lithological sequence is characterized by lithologies belonging to the Sefwi Group and

consisting of intercalated meta-mafic volcanic and meta-diorite dykes with altered meta-mafic

volcanic and meta-sediments which are locally characterized as magnetite rich, banded iron

formation like horizons (Bourassa, 2003), as illustrated in Figure 7-5. The sequence is

characterized by the presence of multiple ankerite-quartz veins which are sub-parallel to the main

penetrative foliation. The lithological sequence is also characterized by Eoeburnean felsic

porphyry intrusions on the south-eastern flank of the Wassa mine fold.

Figure 7-5 Mine geology

(Modified from Bourassa, 2003 and Perrouty et al., 2013)

June 2019


The first deformational event (D1) at Wassa is of Eoeburnean timing and consists of North-South

Shortening. This pre-Tarkwaian event resulted in a penetrative foliation which transposed

lithological contacts along this main foliation. Early, gold bearing, syn-D1 quartz-ankerite veins

were also formed during the Eoeburnean event.

The second event of deformation (D2) is an extension period with no local deformation at the mine

scale at Wassa. Regionally, this event separates the Eoeburnean and Eburnean orogeny by an

extension period of approximately 40 Ma which resulted in the sedimentation of the Birimian and

Tarkwaian basins.

The Eburnean orogeny is divided in three distinct deformational events, D3 is a Northwest-

Southeast shortening event which resulted in the inversion of regional detachment faults into thrust

faults. At the mine scale, this event generated a second penetrative foliation at Wassa and a first

phase of Eburnean folding. The D4 deformational event, a North Northwest-South Southeast

shortening event resulted in the sinistral reactivation of earlier faults at the regional scale and

severely buckled the Wassa stratigraphic sequence into moderately steeply dipping, tight fold

patterns (F4 Fold) and a third penetrative foliation (S4).The last deformational event, D5, is the

result of sub-vertical compression which resulted in open recumbent folds at Wassa and a fourth

foliation located in the axial plane of the F5 folds and is generally sub-horizontal, shallowly

plunging to the South. The various phases of Eburnean deformations and their effect on the host

rocks are illustrated in Figure 7-6 and Figure 7-7. Also, large scale F3 and F4 folds can be observed

on vertical sections in Figure 7-8.

June 2019


Figure 7-6 Eburnean folds and foliations from the Wassa Mine Starter Pit

Top picture syn-D1 veins and S1 foliation folded by an F3 fold, bottom picture, syn-D1 veins and

S1 and S3 foliations affected by a mesoscopic F4 fold.

June 2019


Figure 7-7 Eburnean folds and foliations from the Wassa Mine B-Shoot Pit

Top picture syn-D1 veins folded and buckled by the S5 foliation, bottom picture, syn-D1 veins

affected by both the S4 and S5 foliations.

June 2019


Figure 7-8 Vertical section of the Wassa Main deposit (19975N)

The Wassa mineralization is subdivided into a number of domains, namely; F Shoot, B Shoot, 242,

South East, Starter, 419, Mid East and Dead Man’s Hill. Each of these represents discontinuous

segments of the main mineralized system which extends for approximately 3.5 km along strike

from surface and is still open at depth. The SAK deposits are located approximately 2 km to the

southwest of the Wassa Main deposit on the northern end of a well-defined mineralized trend

parallel to the Wassa Main trend. The mineralization is hosted in highly altered multi-phased

greenstone-hosted quartz-carbonate veins interlaced with sedimentary pelitic units. The SAK

mineralization is subdivided into a number of domains as well, SAK 1, 2 and 3.

June 2019


Mineralization within the Wassa Mine is structurally controlled and related to vein densities and

sulphide contents. In detail, the mineralization generally consists of broadly tabular zones

containing dismembered and folded ribbon-like bodies of narrow quartz vein material, zones are

typically 10 m to 50 m wide within a 900 m mineralized corridor as illustrated in Figure 7-9. Three

vein generations have been distinguished on the basis of structural evidence, vein mineralogy,

textures and associated gold grades. Evidence further relates the majority of gold mineralization

to the earliest recognized vein generation which is believed to be syn-Eoeburnean. Gold grades

broadly correlate with the presence of quartz-dolomite/ankerite-tourmaline bearing quartz veins

and the presence of sulphide minerals (predominantly pyrite) within and around the quartz veins.

Gold grades appear to be spatially restricted to the quartz veins, vein selvages and the immediate

wall rocks. The alteration haloes developed around the veins and pervasively developed within the

core of the Wassa Fold contain lower grade mineralization. The combined and overprinted

Eburnean deformational events (D3 to D5) render precise prediction of the vein geometries and

localities difficult in areas with wider spaced or little drillhole data. However where drilling density

is tighter (12.5 m x 10 m), as with in the immediate underground mining areas it is possible to

construct both hanging and footwall contacts of the economic gold mineralization, Figure 7-10.

The higher grade zones of gold mineralization are constrained with in broader lower grade

mineralized zones that can be defined reasonably well with the wider spaced surface drillhole data,

but to delineate the geometry of the higher grade zones tighter underground grade control drilling

is required, Figure 7-11.

June 2019


Figure 7-9 Vertical section showing the tabular nature of the ore zones (20000N)

June 2019


Figure 7-10 Vertical section showing the >1.5g/t Au shell

Underground

Development

Underground

Drilling

Short Range model

HG >1.5 g/t Au

Surface

Drilling

Pit

Design

$1450

Pit Shell

WEST EAST

Dec 17

Year

end pit

Mined

stopes

June 2019


Figure 7-11 Vertical section showing the >0.4 g/t Au and >1.5g/t Au grade shells

Underground

Development

Underground

Drilling

Long Range model

LG >0.4 g/t Au

HG >1.5 g/t Au

Surface

Drilling

Pit

Design

$1450

Pit Shell

WEST EAST

Dec 17

Year

end pit

June 2019


7.2.3 Hwini-Butre

The Hwini-Butre concession is underlain by three deposits: Adoikrom, Dabokrom and Father

Brown, which are all characterized by different styles of mineralization. The Hwini-Butre deposits

are hosted within the Mpohor mafic complex, which consists mainly of gabbroic and gabbro-

dioritic intrusive horizons as illustrated in Figure 7-12.

The timing of the mineralization at Hwini-Butre is considered to be of late to post Eburnean age

with the period of hydrothermal activity likely to have spanned a considerable length of time. At

Father Brown and Dabokrom, mineralization is associated with quartz vein systems which are

locally surrounded by extensive, lower grade, disseminated quartz stockwork bodies, especially at

Dabokrom. The Father Brown deposit is characterized by well-developed fault-filled quartz veins

which are, as is the case for Dabokrom, light grey with carbonate and mica accessory minerals and

minor tourmaline and feldspar. Wallrock alteration is commonly associated with elevated gold

grades and consists of silicification with carbonates, muscovite and sericite. Secondary strain

fabrics are also present, with mylonitic and cataclastic fabrics common in the heavily altered zones.

Visible gold occurs as disseminations in discrete quartz veins and within zones of silicification

associated with pyrite. Gold is medium to coarse grained and generally occurs with pyrite and

appears to be free milling. As at Benso, arsenopyrite is largely absent from the Hwini-Butre

deposits.

At Adoikrom, the mineralization is shear hosted and characterized by the absence of quartz veins;

gold is associated with fine grained pyrite and intense potassic alteration.

7.2.4 Benso

The Benso concession is underlain by four main deposits: Subriso East, Subriso West, G Zone and

I Zone. All the deposits are characterized by similar style of mineralization. As with Hwini-Butre,

the Benso deposits are hosted within mafic intrusive rocks of gabbroic to dioritic composition,

which intrude a thick volcano-sedimentary sequence mainly composed of mafic volcanic flows.

Mineralization at Benso is associated with late deformational stages of the Eburnean orogeny and

deposits are shear hosted along subsidiary structures.

Mineralogy is relatively simple with fine grained but visible gold disseminated in the shear fabric

and associated with pyrite which can be locally abundant. Zones of intense alteration with chlorite,

carbonates and epidote are common. Arsenopyrite is absent from the deposits and in microscopic

section the gold would appear to be free milling.

7.2.5 Chichiwelli

The Chichiwelli deposit consists of two sub-parallel mineralized trends which hosts two distinct

types of mineralization. The Chichiwelli West trend is a shear zone hosted deposit with a quartz,

carbonate, sericite and potassic alteration assemblage, the mineralization is associated with pyrite.

The Chichiwelli East trend is a quartz vein associated deposit with an ankerite and sericite

alteration assemblage. Mineralization is also associated with pyrite along vein selvages and in the

wall rocks.

The lithological assemblage at Chichiwelli West consists of mainly fine to medium grained dioritic

intrusives with local intercalation of basalt and feldspar porphyritic intrusives. Lithologies are

moderately to strongly foliated adjacent to the shear zone, the mineralization is bounded to the

June 2019


shear zone and associated with a strong shear fabric. The shear zone mineralization is characterized

locally by boudinage quartz and calcite stringers with fine disseminated sulphides, mainly pyrite,

and associated with a sericite and potassium alteration assemblage with minor silicification. The

Chichiwelli East lithological sequence is comprised mainly of deformed diorite with local strain

zones. The mineralization is characterized by milky white quartz veins associated with potassium

alteration and euhedral coarse grained pyrite.

June 2019


Figure 7-12 Regional geology of the Hwini-Butre, Benso and Chichiwelli concessions

(PL = Prospecting License ML = Mining Lease)

Chichiwelli PL

Haul Road to

Wassa Plant

Benso Pits

Hwini Butre ML

Amantin PL

Manso 1 PL

Manso 2 PL

Adoikrom/Father Brown Pits

June 2019


8 Deposit Types

8.1 Wassa…

The Wassa deposit is located on the eastern flank of the northeast trending Ashanti Belt, a

Paleoproterozoic greenstone belt which was formed and deformed, along with the dividing

Birimian and Tarkwaian sedimentary basins during the Eoeburnean and Eburnean orogeny. Most

deposits found within the Ashanti belt can be classified as lode gold deposits or orogenic

mesothermal gold deposits, with the exception of the Tarkwaian paleoplacer deposits which have

a sedimentary origin. Orogenic gold deposits are the most common gold systems found within

Archean and Paleoproterozoic terrains, in the West African shield, these deposits are typically

underlain by geology considered to be of Eburnean age and are generally hosted by volcano-

sedimentary sequences.

B. Dubé and P. Gosselin of the Geological Survey of Canada described these deposits as

greenstone-hosted quartz-carbonate vein deposits in the 2007 special publication No. 5 entitled

Mineral Deposits of Canada. The authors described these deposits as typically occurring in

deformed greenstone belts and distributed along major compressional crustal scale fault zones

commonly marking the convergent margins between major lithological boundaries. The

greenstone-hosted quartz-carbonate vein deposits correspond to structurally controlled complex

deposits characterized by networks of gold-bearing, laminated quartz-carbonate fault-fill veins.

These veins are hosted by moderately to steeply dipping, compressional brittle-ductile shear zones

and faults with locally associated shallow-dipping extensional veins and hydrothermal breccias. In

these deposits, gold is mainly confined to the quartz-carbonate veins but can also occur within

iron-rich sulphidised wall rocks or within silicified and sulphide-rich replacement zones.

The Ashanti belt is considered prospective for orogenic mesothermal gold deposits and hosts

numerous lode gold deposits and paleoplacer deposits. As illustrated by Figure 8-1, several major

gold deposits are found within the Ashanti belt which can be classified into six different deposit

types:

• sedimentary hosted shear zones;

• fault fill quartz veins;

• paleoplacer;

• intrusive hosted;

• late thrust fault quartz veins; and

• folded veins system.

The sedimentary hosted shear zone deposits are localised principally along a steep to sub-vertical

major crustal structures located along the western margin of the Ashanti belt referred to as the

Ashanti trend. The Ashanti trend shows a range of mineralization styles associated with graphitic

shear zones, which represents the principal displacement zone of a regional-scale shear zone that

defines the mineral belt. These styles include highly deformed graphitic shear zones containing

disseminations of arsenopyrite as the principal gold bearing phase and disseminations of sulphides

in mafic volcanic rocks generally found in the footwall of the main shear zones. The sedimentary

June 2019


hosted shear zone deposits which occur along the Ashanti trend include Bogoso, Obuasi, Prestea

and Nzema.

Figure 8-1 Geology of the Ashanti belt with location of major gold deposits


June 2019


The second type of deposit found within the Ashanti belt are laminated quartz vein deposits

containing free gold. Fault filled quartz vein deposits also occur along the Ashanti trend but are

only present at Obuasi and Prestea. The third type of deposit are paleo-placer deposits within the

Tarkwaian sedimentary basin which are hosted within narrow conglomerate horizons intercalated

with sandstone units characterized by iron oxides cross beddings. Paleoplacer deposits occur in

the southern portion of the Tarkwa basin and examples include Tarkwa, Teberebie and Iduaprim.

The fourth type of deposit found within the Ashanti belt are intrusive hosted deposits which occur

along second order structures such as the Akropong trend in the Kumasi basin and the Manso trend

in the Southern portion of the Ashanti belt. These deposits can be hosted both within felsic and

mafic intrusives and are characterized by a penetrative fabric where gold is associated with pyrite

and arsenopyrite. Examples of such deposits include Edikan and Pampe along the Akropong trend

and Benso and Hwini-Butre along the Manso trend. The fifth type of deposit found within the

Ashanti belt is late thrust fault associated quartz vein deposits. The Damang mine which is located

just west of Wassa is the only known thrust fault related deposit in the Ashanti belt. The deposit is

characterized by low angle; undeformed extensional and tensional veins associated with low angle

thrust faults. This type of deposit contrasts with the last type of deposit found with the belt, the

multi-phase folded Wassa vein deposit. The Wassa mineralization consists of greenstone-hosted,

low sulphide hydrothermal deposits where gold mineralization occurs within folded quartz-

carbonate veins, as illustrated in Figure 8-2. The Wassa deposit can therefore be classified as an

Eoeburnean folded vein system and is the only such deposit recognized to date within the Ashanti

belt.

Host rocks in the Wassa mine area have been affected by at least four phases of ductile

deformation, producing a polyphase fold pattern at the mine scale. Discrete high-strain zones

locally dissect this fold system. The structural history of the Wassa area is important in that the

various deformational events have been responsible for the emplacement of the gold

mineralization as well as the geometry of the zones themselves. Mineralized zones at the Wassa

Mine are related to vein swarms and associated sulphides that formed during the Eoeburnean

deformational event. All rock types underlying the Wassa Mine appear to be altered to variable

degrees with the most common alteration consisting of a carbonate-silica-sulphide assemblage.

June 2019


Figure 8-2 Syn-Eoeburnean veins from the B-Shoot, 242 and South-East zones


8.2 Hwini-Butre

The Hwini-Butre deposits can be characterized as mafic intrusive hosted, orogenic shear zones.

The deposits are hosted within diorite and granodiorite intrusive rocks of the Mpohor complex.

The Father Brown deposit is characterized by well-developed fault-filled quartz veins as illustrated

in Figure 8-3, whereas the Adoikrom deposit is a shear zone hosted deposit characterized by

intense potassium and silica alteration assemblage.

Analysis of geophysical surveys and topographical features have identified several north to north-

northeast trending regional features running through the area which are tentatively interpreted as

boundary faults along the margins of the Ashanti Belt. The Mpohor complex exhibits the

underlying north-south trends but also has extensive cross cutting features present particularly in

the north-west orientation. These structural features are second order or subsidiary structures

splaying from primary structures.

The Adoikrom, Father Brown and Dabokrom deposits occur in the south portion of the Mpohor

complex and appear to be controlled by a series of shallow to moderately dipping faults and shear

structures with dips varying from 20° to the south at Dabokrom and steepening to 65° to the

northwest at Adoikrom.

June 2019


Figure 8-3 Different mineralization styles underlying the Hwini-Butre concession

(Top is a fault fill smoky quartz vein characterizing the Father Brown deposit, while the bottom

picture represents the potassic alteration at Adoikrom shear zone).

8.3 Benso…

The Benso deposits can also be characterized as mafic intrusive hosted, orogenic shear zones

deposits, which are hosted by Birimian metavolcanics into which coarse plagioclase porphyry units

have intruded and are generally conformable with the volcaniclastic units.

At Subriso East, the metavolcanics host complex quartz vein systems associated with intense

shearing and abundant sulphide mineralization, as shown in Figure 8-4. At Subriso West, the

presence of intermediate porphyry intrusive appears to play a more significant role (Figure 8-5)

and quartz veining is less extensive and broad scale silicification is more common. The contacts

between metavolcanics and porphyry have been identified as potential targets for higher grade

gold mineralization.

The mineralization hosting structures generally dip steeply towards the west with foliation

generally parallel to the bedding. The aeromagnetic interpretation reveals a north to north-

northeast striking fault system along the course of the Ben River with several other fracture

systems also evident with strikes varying between the northwest and northeast. The Subriso East

deposit is interpreted to dip less steeply to the west at approximately 50°.

June 2019


Oxidation associated with weathering is variable but generally limited. The weathering forms a

layer of lateritic clay rich material grading into a soft saprolite. The vertical depth is generally 10

m or less but can reach depths of 30 m in places. There is a sharp boundary between oxide and

fresh material with a narrow and poorly developed transition zone.

Figure 8-4 Mineralized shear zones occurring on the Benso concession

(left, sheared siltstones with fine grained pyrite found at Subriso East; on the right, sheared

volcanic flows hosting the Subriso West mineralization)

8.4 Chichiwelli…

The Chichiwelli deposits can also be characterized as mafic intrusive hosted, orogenic shear zones,

the deposits are hosted within diorite and granodiorite intrusive rocks. The mineralization zones at

Chichiwelli are similar to those observed at Benso, with the mineralized hosting structures

generally dipping to the east.

The Chichiwelli deposit consists of two sub-parallel mineralized trends which hosts two distinct

types of mineralization, as shown in Figure 8-5. Mineralization at the Chichiwelli West zone is

shear zone hosted with a carbonate, sericite and potassic alteration assemblage, while

mineralization along the Chichiwelli East trend is quartz vein associated with an ankerite and

sericite alteration assemblage. Mineralization is spatially associated with pyrite at both deposits.

June 2019


Figure 8-5 Chichiwelli mineralization

(On the left, drill samples from the Chichiwelli West trend where mineralization is shear hosted,

on the right, drill samples from the Chichiwelli East trend where mineralization is hosted within

hydrothermal veins).

June 2019


9 Exploration

9.1 Introduction

In addition to the drilling described in Section 9, extensive exploration work has been conducted

on and around the Wassa concession. Previously, several airborne and ground geophysical surveys

consisting of aero-magnetics, radiometrics and Induced Polarization (“IP”) were conducted on the

properties. The geophysical surveys targeted geochemical anomalies, which had previously been

identified following multiple stream and soil geochemical sampling programs.

9.2 Wassa…

Modern exploration programs on the Wassa concession began in the early 1990s with satellite

imagery and geophysical surveys which identified geophysical lineaments and anomalies over

small scale and colonial mining areas. Stream and soil geochemistry sampling programs were

conducted over the geophysical anomalies and identified two linear gold in-soil anomalies as

illustrated in Figure 9-1.

Figure 9-1 Soil geochemistry anomalies

June 2019


Exploration drilling commenced in February 1994 and, by March 1997, a total of 58,709 m of RC

and DD had been completed. In September 1997, consulting engineers Pincock, Allen and Holt

completed a FS. Only minimal exploration work was conducted by SGL between the completion

of the FS in 1997 and the 2001 bankruptcy.

In March 2002, GSR started an exploration program as part of a due diligence exercise following

the ratification of a confidentiality agreement with the creditor of SGL. The exploration program

consisted mainly of pit mapping and drilling below the pits to test the continuity of mineralization

at depth. The concession was acquired later that year by GSR following the completion of the due

diligence exercise. Exploration drilling resumed in November 2002 under GSR with the aim to

increase the quoted reserves and resources for the FS, which was completed in 2003.

Simultaneously to the resource drilling program that targeted resource increases in the pit areas,

GSR also undertook grass roots exploration along two previously identified mineralized trends.

The 419 area was located south of the main pits and the SAK anomaly was a soil target that had

never been previously drilled and was located west of the main pits. Deep auger campaigns were

also undertaken in the Subri forest reserve, which is located in the southern portion of the Wassa

Mining lease.

In March and April 2004, a high resolution, helicopter geophysical survey was carried out over

the Wassa Mining Lease and surrounding Prospecting and Reconnaissance Licenses (Figure 9-2).

Five different survey types were conducted, namely: Electromagnetic, Resistivity, Magnetic,

Radiometric and Magnetic Horizontal Gradient. The surveys consisted of 9,085 km of flown lines

covering a total areal of 450 km2. Flight lines were flown at various line spacing varying between

50 to 100 m depending on the survey type. The geophysical surveys identified several anomalies

with targets being prioritized on the basis of supporting geochemical and geological evidences.

The exploration program in 2005 continued to focus on drill testing anomalies identified by the

airborne geophysical survey as well as infill drilling within the pit area to expand the reserve and

resource base. The resource definition drilling program focused mainly on SAK, South-East and

the 419 area. The following years were subject to more infill and resource definition drilling in the

pit areas at Wassa. In 2011, exploration drilling programs shifted towards drilling deep HG targets

below the pits; this drilling continued until 2015. Drilling was limited in 2016 with rigs in filling

the first planned stoping areas to increase confidence in the resource prior to underground mining.

The 2017 drilling programs were two-fold, infilling gaps in the previous drilling with in the

proposed expanded open pit as well as testing the B shoot underground mineralization both north

and south, up and down plunge respectively. The southern extension drilling initiated in 2017

continued into 2018 and utilized larger drill rigs to conduct directional wedging and downhole

motor work to delineate the deeper southern extensions of B and F shoot HG mineralization.

June 2019


Figure 9-2 Wassa airborne magnetic interpretation


June 2019


9.3 Hwini-Butre

The Hwini-Butre concession began to be subject to modern exploration programs in the early

1980s, the Dabokrom concession was acquired by BD Goldfields Ltd. which entered into an

agreement with Danish Company Lutz Resources Limited. Preliminary exploration work was

conducted in the early 1990s and the property was transferred to HBM, which was also controlled

by Scandinavian investors. In 1993, HBM entered into a joint venture with Placer-Outokumpu

who drilled several vertical holes around the Dabokrom area to assess the large-scale potential of

the vein systems. The drilling program totalled approximately 300 m in 3 diamond drill holes and

610 m over 13 RC holes.

SJR acquired the Hwini-Butre concession in the mid 1990s and began exploring the concession in

February 1995. Exploration programs undertaken by the SJR represented the first sustained

exploration program on the concession. SJR undertook ground geophysical surveys which

included magnetic, radiometric and induced polarization surveys; soil geochemical surveys were

also completed on the concession area, resulting in the identification of numerous targets.

Trenching and pitting were conducted in areas of geophysical and geochemical anomalies and over

historical prospects or old workings in an attempt to outline near surface mineralization.

Subsequent drilling of the surface targets resulted in the delineation of the Adoikrom, Father

Brown and Dabokrom prospects along a combined strike length of 900 m. Further exploration

conducted in 2005 identified the Adoikrom North prospect. A total of some 22,100 m over 267

drill holes were completed on the main mineralized zones and the exploration targets.

GSR acquired the Hwini-Butre concession in late 2005 and commenced exploration work in early

2006. GSR exploration activities concentrated on the previously defined mineralization at

Adoikrom North, Adoikrom, Dabokrom and Father Brown. The drilling program focused mainly

on infill drilling and extending the continuity of the deposits at depth. The previous drilling by SJR

reached a maximum vertical depth of approximately 130 m, whereas GSR extended the modelled

mineralization at vertical depths of over 250 m.

GSR also undertook regional exploration programs over the concession by targeting a number of

geochemical and geophysical anomalies previously identified by SJR, these anomalies were

mainly tested by use of rotary air blast drilling. A combination of 4 m deep auger and shallow

auger at a grid spacing of 400 m by 50 m was also carried out to further test the existing gold in

soil anomalies and gaps in the geochemistry sampling over the Hwini-Butre concessions.

In 2007 and 2008, GSR focused its Hwini-Butre exploration activities on the northern portion of

the concession where several colonial gold occurrences such as Breminsu, Apotunso, Abada,

Whinnie and Guadium are located. Previous soil sampling in these areas identified several

anomalies and the follow up programs included deep auger and rotary air blast drilling. A total of

1,384 auger holes and 41 RAB holes totalling 725 m were completed.

In 2009, 5,992 m RC (83 holes) and 2,100 m DD (21 holes) were completed on the Hwini-Butre

property (Father Brown, Adoikrom and Dabokrom) to test the strike extensions of the zones and

also upgrade the existing quoted resources. The drilling program also identified potential

underground target beneath the Subriso West pit. Also, 86 RAB holes, totalling 2,195 m were

drilled at Abada to test coincidental gold in soil and geophysical anomalies. On the Benso

concession, resource delineation and definition drilling was undertaken on the Subriso East,

Subriso West and G Zone deposits. A total of 3,159 m RC (35 holes) and 2,538.4 m DD were

June 2019


completed. Induced Polarization geophysical surveys were conducted over the Hwini-Butre and

Benso concessions in 2009. The program generated targets that were coincidental with lithological

trends and gold in soil anomalies.

The resource definition drilling program continued in 2010 at Father Brown, Adoikrom and

Dabokrom where 5,075 m of RC drilling (72 holes) and 5,207.3 m of DD drilling (24 holes) were

completed. The drilling program also tested the underground potential of the deposits with

significant success. A deep auger program totalling 746 m over 205 holes to test IP geophysical

anomalies at Essaman was also completed.

Exploration activities conducted at Hwini-Butre in 2011 included the testing of deeper targets at

Father Brown and Adoikrom to evaluate the underground potential of the deposits. In all, 13 DD

holes totalling 3,689.6 m were drilled at Father Brown and Adoikrom. RAB drilling, totalling

2,941 m (174 holes) were undertaken at Semkrom on the Hwini-Butre property to test IP and

aeromagnetic/radiometric anomalies. In 2012, exploration at Hwini-Butre concentrated on Father

Brown and Adoikrom infill and step out underground drilling program, with 33 DD holes totalling

10,094 m being completed. In 2018, exploration drilling resumed at Father Brown and Adoikrom

to continue evaluating the underground potential. The program combined RC and DD holes

totalling 8,236.2 m.

9.4 Benso and Chichiwelli

The first exploration program at Benso and Chichiwelli was conducted by BHP between 1989 and

1992. The work consisted of regional soil sampling, with a total of 5,400 samples collected and

several significant soil geochemical anomalies identified at Chichiwelli, Subriso, Denerawah and

Amantin. BHP also undertook some advanced exploration work, especially at Chichiwelli where

twelve drill holes were completed, but one of the targets were deemed large enough to meet BHP’s

size threshold and they relinquished all of their interests in the concessions. Shortly thereafter, a

local Ghanaian Company called Architect Co-Partners acquired a 150 km2 prospecting concession

covering the Amantin, Subriso and Chichiwelli prospects. This also included a large part of the

Subriso River Forest Reserve, which was closed to exploration after 1996.

In 1995, Fairstar took over the Benso concession and carried out extensive work, especially at

Subriso and Amantin, under the management of the consulting company, CME (Ghana) Ltd. of

Accra and Vancouver, Canada. The work program between 1995 and 1997 consisted of 800

prospecting pits averaging 4.5 m depth and 100 trenches totalling 4,245 m, plus 1,400 m of old

trenches were re-opened and mapped. Also, approximately 8,000 m of diamond drilling was

carried out and almost 10,000 drill samples were logged and assayed. By the end of the decade,

work on the concession had largely ceased because of a lack of funds.

By mid-2001, SJR completed an agreement with Fairstar and took over the exploration work. From

early 2002 to about mid-2004, SJR focused mainly on the Subriso area where substantial mineral

resources were outlined at two prospects, Subriso East and West. Numerous other prospects,

namely Subriso Central, I Zone and G Zone were identified and drill tested, as was the Amantin

area, which had also been drilled to a considerable extent by Fairstar.

GSR acquired the Benso and Chichiwelli concessions in late 2005 and commenced exploration

work in early 2006, with exploration activities focusing on the previously defined mineralization

at Subriso East, Subriso West, I Zone and G Zone. The drilling program focused mainly on infill

drilling and extending the continuity of the deposits at depth. The 2006 exploration program was

June 2019


also the focus of regional exploration programs over the concession by targeting a number of

geochemical and geophysical anomalies previously identified by SJR, these anomalies were

mainly tested by use of rotary air blast drilling. A combination of 4 m deep auger and shallow

auger at a grid spacing of 400 m by 50 m was also carried out to further test the existing gold in

soil anomalies and gaps in the geochemistry sampling over the Hwini-Butre concessions.

Exploration on the Benso property in 2007 and 2008 concentrated on drill testing new zones of

mineralization delineated by the RAB drilling in 2006. A total of 81 holes and 10,232.3 m of RC

and DD drilling was completed at Subriso East, Subriso West, G Zone and I Zone. At Amantin,

follow-up programs included deep auger sampling on a 200 by 50 m grid and RAB drilling was

undertaken to test the previously defined soil anomalies. A total of 3,717 m of RAB drilling from

178 holes and 1,683.9 m of deep auger drilling over 487 holes were completed at Amantin.

The 2009 exploration program at the Benso concession focused on resource delineation and

definition drilling at the Subriso East, Subriso West and G Zone deposits. A total of 3,159 m RC

(35 holes) and 2,538.4 m DD were completed. Induced Polarization geophysical surveys were

conducted over the Benso concessions in 2009 and the program generated targets that were

coincidental with lithological trends and gold in soil anomalies.

The 2010 exploration activities at Benso included the continuation of the resource delineation and

definition drilling in and around the pits and also drilling off the potential underground target at

Subriso West. A total of 8,815 m RC (112 holes) and 8286.2 m DD (18 holes) were completed. A

deep auger program totalling 1,114 m over 319 holes was undertaken to test IP targets at Subriso

West.

In 2011, 12 DD holes, totalling 4,557 m, were drilled on the Benso property at Subriso West to

close up the spacing along strike and down dip of the HG zone of mineralization intersected

beneath the pit. At Amantin, a shallow RC program. totalling 1,177 m (22 holes). was completed

to follow up on widely spaced RAB and RC intersections from earlier drilling programs. A deep

auger (6 m) program totalling 907.5 m from 174 holes were completed at K Zone and I Zone to

test additional targets generated by IP survey program.

Exploration activity at Benso in 2012 was limited to structural interpretation of the controls on

mineralization to determine the underground potential at Subriso West.

June 2019


10 Drilling

10.1 Open Pit

10.1.1 Drilling

Drilling is carried out by a combination of DD, RC and RAB techniques. In general, the RAB

method is used at early stages for follow up to soil geochemical sampling and, during production,

for testing contacts and mineralization extensions around the production areas. RAB has a

maximum drilling depth of around 30 m. The RC pre-collar diamond core tail drilling is used as

the main method for obtaining suitable samples for Mineral Resource estimation and is carried out

along drill lines spaced between 25 and 50 m apart along prospective structures and anomalies

defined from soil geochemistry and RAB drilling results. RC drilling is typically extended to

depths of in the order of 100 to 125 m. The DD method is used to provide more detailed geological

data and in areas where more structural and geotechnical information is required. Generally, the

deeper intersections are also drilled using DD and, as a result, most section lines contain a

combination of RC and DD drilling.

RC and DD drilling were carried out in double shifts and during every shift a GSR geologist was

on site to align the drill rig and check the drill head dip and azimuth. Downhole surveying was

conducted using a single shot camera, for RC and DD holes at the bottom of holes exceeding 30

m depths and then taken progressively every 30 m up hole. The single shot camera recorded the

dip and azimuth for each of the surveys on a film image which was validated and recorded by the

GSR geologists or was recorded by a Reflex survey instrument and captured in the database as

well as being filed in the respective drillhole file folders on site.

Drilling depths at Wassa Main have generally been less than 250 m but with the discovery of

higher grades below the Wassa Main pit in late 2011, hole depths have increased. In the 1st half of

2014, two gyro survey instruments were utilized to resurvey several of the deeper holes. In total,

153 holes, drilled during 2012 to 2014, were resurveyed. The gyro survey readings were conducted

every 10 m both in and out of the hole and the values were then averaged. The 153 gyro surveyed

holes were updated in the database and subsequently used for the resource estimates. The gyro

surveys showed that there was some deviation in the holes below 250 m drilled depth. Deviations

varied from location to location depending on drill orientation with a general tendency for the hole

to steepen and swing to the north.

Drilling of the deeper targets at Wassa has required the use of directional drilling methods. The

deeper holes, often exceeding 1000 meters, are drilled from surface using HQ sized core and this

initial hole (referred to as the “mother” hole) is drilled to the depth where the first directional hole

would be started. The directional hole (or “daughter” hole) is drilled using a smaller core size, NQ

and is deviated from the mother hole initially using a casing wedge which is oriented in the

direction of the mineralized target. Once the initial deflection has been achieved with the wedge,

the hole deviation can be controlled using a down hole directional motor which can change the dip

and azimuth of the hole by approximately plus or minus 1.5 degrees over a 10-metre run. The

direction of the hole can also be controlled by using various combinations of down hole stabilizers

and drill bits. The step out deeper drilling fences typically involve two mother holes with three

to four daughter holes from each of these. The deeper holes are surveyed, down hole with either

June 2019


a Reflex multi-shot or gyro survey instrument. The surveys are taken while the hole is being drilled

as well as every 10 to 15 meters from the bottom of the hole once it has been completed.

A summary of the exploration data used in the Mineral Resource models is given in Table 10-1.

Table 10-1 Exploration data used for the Mineral Resource models

Location Type Number of

Holes

Meterage

(m)

Wassa

RC 1,463 139,292

DD 892 276,852

GC (RC) 25,561 660,058

Wassa UG

DD 971 110,541

GC(Chan-Chips) 1,717 9,174

Hwini-Butre RC 3,165 75,384

DD 518 73,223

Benso

RC 465 33,276

DD 321 37,623

Geotech 14 1,637

GC (RC) 2,362 57,970

Chichiwelli RC 483 29,802

DD 23 3,692

All of the drillhole collars were surveyed using a Nikon Total Station (DTM-332) or Sokkia Total

Station by a GSR surveyor. Individual RC and DD holes have been identified and marked in the

field with Polyvinyl chloride (“PVC”) pipes. RAB drill holes have been also surveyed in the field

and identified and marked with wooden pegs.

10.1.2 Sampling

GSR follows a standardized approach to drilling and sampling on all its Ghanaian projects.

Sampling is typically carried out along the entire drilled length. For RC drilling, samples are

collected every 1 m. Where DD holes have been pre-collared using RC, the individual 1 m RC

samples are combined to produce 3 m composites which are then sent for analysis. Should any 3 m

composite sample return a significant gold grade assay, the individual 1 m samples are then sent

separately along with those from the immediately adjacent samples.

DD samples are collected, logged and split with a diamond rock saw in maximum 1.2 m lengths.

The core is cut according to mineralization, alteration or lithology. The core is split into two equal

parts along a median to the foliation plane using a core cutter. The sampling concept is to ensure

a representative sample of the core is assayed. The remaining half core is retained in the core tray,

for reference and additional sampling if required.

June 2019


RC sampling protocols were established in 2003. The composite length of 3 m has been established

to allow a minimum of at least two composites per drillhole intersection based on experience from

exploration drilling and mining. The hangingwall and footwall intersections can generally be easily

recognized in core from changes in pyrite content and style of quartz mineralization. The 3 m

composite sampling methodology is as follows:

• a sample of each drilled meter is collected by fitting a plastic bag on the lower rim of the

cyclone to prevent leakage of material;

• the bag is removed once the “blow-back” for the meter has been completed and prior to

the commencement of drilling the subsequent meter;

• both the large plastic sample bags and the smaller bags are clearly and accurately labelled

with indelible ink marker prior to the commencement of drilling. This is to limit error

and confusion of drilling depth while drilling is proceeding;

• 3 m composite samples are taken by shaking each of the 1 m samples (approximately 20

kg) and taking equal portions of the 3 consecutive samples into a single plastic bag to

form one composite sample (approximately 3 kg);

• the composite samples are taken using tube sampling, which uses a 50 mm diameter PVC

tube which has been cut at a low oblique angle at one end to produce a spear of

approximately 600 mm length;

• the technique assumes that a sample from the cyclone is stratified in reverse order to the

drilled interval. A representative section through the entire length of the collected sample

is considered to be representative of the entire drilled interval;

• the PVC tube is shuffled from the top to bottom of the sample, collecting material on the

way. The “shuffling” approach ensures sample accumulated in the tube does not just

push the remaining sample away; and

• the material in the tube is emptied into the appropriately labelled sample bag and in the

case of 3 m composite samples, stored separately from the 1 m samples.

The 1 m sample collection methodology is as follows:

• the 1 m re-sampling of selected mineralized composite zones using the 20 kg field

samples is undertaken with a single stage riffle splitter;

• the splitter is clean, dry, free of rust, and damage is used to reduce the 20 kg sample

weight to a 3 kg fraction for analysis;

• care is taken to ensure that the sample is not split when it is transferred to the splitter,

and is evenly spread across the riffles;

• when considered necessary, the sample is assisted through the splitter by tapping the

sides with a rubber mallet;

• excessively damp or wet samples are not put through the splitter, but tube-sampled or

grab-sampled in an appropriate manner. Alternatively, the sample is dried before

splitting. A common sense approach to wet sampling is adopted on a case by case basis;

June 2019


• similarly, clods of samples are not forced through the splitter, but apportioned manually

in a representative manner; and

• the splitter is thoroughly cleaned between each sample using a brush. Where possible,

the splitter is cleaned using an air gun attached to the drill rig compressor.

RAB samples are collected and bagged at 1 m intervals. As the samples are generally smaller in

size than the RC samples, 3 m composites are prepared by shaking the samples thoroughly to

homogenize the sample, before using the PVC tube to collect a portion of the three individual 1 m

samples. After positive results from the 3 m composites, the individual 1 m samples are split to

approximately 2 to 3 kg using the Jones riffle splitter and then submitted to the laboratory for

analysis.

10.2 Underground

Underground diamond drilling is performed using electric-hydraulic diamond drills utilizing the

underground mine’s 1,000 v power supply. All core drilled underground is either NQ (47.6 mm)

or NQ2 (50.6 mm) in core size. The final drilling density in the indicated resource area is

designed to be 12.5 m along strike and 10 m down dip. With the orebody generally striking

north-south, typical drilling azimuths range from 225o azimuth through to 315o azimuth, with the

majority of drill holes designed between 250o azimuth and 290o azimuth. Typical dips are in the

range of -50o to +50o. Drill hole lengths are generally in the 75 m – 160 m length range.

Downhole surveying is conducting using a Reflex multi-shot downhole surveying tool. When

collaring, a single survey is taken at 10 m depth. At 10 m depth, the drill hole orientation must

fall within ±2o azimuth and ±1.5o dip tolerance, when compared to design. For any hole where

the 10 m survey falls outside of tolerance, the geologist has the discretion to either (1) terminate

the drill hole and re-collar at the drilling company’s expense, or (2) to continue the hole. At the

completion of the drill hole, multi-shot surveys are collected at 15 m intervals on the way out.

All downhole surveys are collected by the underground mine geologists. The drilling crews do

not perform the surveys themselves.

Drill hole collar locations are captured by the underground mine surveying team. The surveyors

use a Leica TS15 total station to record the collar position in X, Y, Z location. The total station is

accurate to less than two seconds in azimuth. In cases where the mine surveyors cannot identify

the drill hole collar site, the designed collar coordinates are recorded in the databases.

June 2019


11 Sample Preparation, Analyses and Security

11.1 Sample Preparation

Sample preparation on site is restricted to core logging and splitting. The facilities consist of

enclosed core and coarse reject storage facilities, covered logging sheds and areas for the splitting

of RC and RAB samples. Sub-sampling of RC and RAB samples is carried out using a Jones Riffle

splitter.

11.2 Sample Despatch and Security

Samples are collated at the mine site after splitting and then transported to the primary laboratory

for the completion of the sample preparation and chemical analysis. Exploration samples are

trucked by road to the laboratories in Tarkwa.

Sample security involves two aspects, namely, maintaining the chain of custody of samples to

prevent inadvertent contamination or mixing of samples, and rendering active tampering of

samples as difficult as possible.

The transport of samples from site to the laboratory is by road using a truck dispatched from the

laboratory. As the samples are loaded, they are checked and the sample numbers are validated.

The sample dispatch forms are signed off by the driver and a company representative. The sample

dispatch dates are recorded in the sample database as well as the date when results are received.

No specific security safeguards have been put in place by GSR to maintain the chain of custody

during the transfer of core between drilling sites, the core library, and sample preparation and

assaying facilities. Core and rejects from the sample preparation are archived in secure facilities

at the core yard and remain available for future testing.

11.3 Laboratory Procedures

Sample assays are then performed at either SGS or TWL, both located in Tarkwa. GSR has used

both laboratories and regularly submits quality control samples to each for testing purposes. Both

laboratories are independent of GSR and are accredited for international certification for testing

and analysis.

The sample preparation and analysis processes at Wassa Site Laboratory (“WSL”), TWL and SGS

differ slightly. WSL has been used as the primary laboratory for 3 m composite and grade control

RC drill samples from July 2007 onwards. The laboratory had previously operated as a

metallurgical sample processing laboratory at the Wassa mine site.

The sample preparation and analysis process at WSL is as follows:

• sample reception, sorting, labelling and loading;

• dry entire sample (3 kg) at 110°C for between 4 and 8 hours;

• jaw crush entire sample to 3 mm, and secondary Keegor crusher to 1 mm;

• split 3 kg sample and pulverize for 3 to 8 minutes to 95% passing 75 µm;

June 2019


• sample homogenisation using a mat rolling technique, and sub-sample 1 kg into bulk

leach extractable gold (“BLEG”) roll bottle;

• bottle roll for 6 hours with LeachWellTM accelerant. Allow to settle for 30 to 60 minutes;

• filter 20 ml aliquot from bottle;

• di-isobutyl Ketone extraction and Atomic absorption spectroscopy (“AAS”)

determination of gold content; and

• 1 in 10 residue samples are retained for gold determination using fire assay.

TWL was the primary laboratory for samples until July 2007, when it was discontinued due to the

following issues:

• Contamination due to poor dust control in pulverizing area of the laboratory. Use of dust

attracting cloth gloves for sample handling. BLEG aliquot preparation area containing

dirt and liquids, which may result in sample cross-contamination.

• Large fluctuation in employee numbers (60 to 180), which resulted in a risk of training

and quality control issues when increasing employment numbers over a short period of

time.

• The use of a manual data tracking and capture system, which increased risk of data entry

errors. GSR considered this to be a sub-optimal process for a commercial laboratory.

• The sample preparation and analysis process used for all samples submitted to TWL is

illustrated in Figure 11-1.

June 2019


Figure 11-1 Transworld Laboratories sample processing flowsheet

The SGS laboratory in Tarkwa has been used for exploration samples since July 2007 with the

sample preparation and analysis process as follows:

• sample received, entered in LIMS, worksheets, printed and samples sorted;

TRANSWORLD LABORATORIES(GH) LTD.-

BLEG +Leachwell Sample Analysis Flow Sheet

Detection Limit 0.01 ppm Au

3-5 kg Sample

Sample Receival and sorting

Dry entire sample

at 110oC (12 hours)

Jaw crush entire sample

<6mm

Riffle split 3.0 to 4.0 Kg Retain residual split in

original receival bag.

If sample weight is greater than 5kg

Pulverise subsample cone splitting is recommended

<75um

Homogenise and weigh Retain residual pulp in

2.0 Kg into BLEG roll bottle pulp bag

Add: 30g Ca(OH)2

10ml of 200ppm CN solution(2g NaCN)

1000 ml water

1 LeachWell Tablet

Place on Bottle roller -

roll for 6 hours

Remove from roller and

allow to settle for 2 hours

Discard all Tails Filter 50ml sub sample Wash Tails of 10th sample Analysis for Gold by

into flask. Fire assay Method

Extract into 5ml of DIBK

Atomic

Absorption

Analysis

Data Processing

and Reporting

June 2019


• samples emptied into aluminium dishes;

• dry entire sample at between 105 and 110°C for 8 hours;

• jaw crush entire sample to 6 mm;

• split sample using a single stage riffle splitter, to result in a 1.5 kg sub-sample;

• pulverise sub-sample for 3 to 5 minutes, to give 90% passing 75 µm;

• sample homogenisation using a mat rolling technique, and put 1 kg of sample into the

BLEG roll bottle;

• the remainder of the sample is retained as pulp and crushed sample duplicates;

• bottle roll for 12 hours with LeachWellTM accelerant. Allow to settle for 2 hours;

• filter 50 ml of aliquot; and

• di-isobutyl Ketone and AAS for gold grade determination.

During the 2017 and 2018 drilling programs GSR discontinued using SGS laboratories and began

shipping samples to TWL. The Intertek lab sample flow sheet is shown in Figure 11-2.

The measures implemented by GSR are considered to be consistent with industry best practice.

June 2019


Figure 11-2 Intertek sample processing flowsheet

June 2019


11.4 Quality Control and Quality Assurance Procedures

Quality control measures are typically set in place to ensure the reliability and trustworthiness of

exploration data, and to ensure that it is of sufficient quality for inclusion in the subsequent Mineral

Resource estimates. Quality control measures include written field procedures and independent

verifications of aspects such as drilling, surveying, sampling and assaying, data management and

database integrity. Appropriate documentation of quality control measures and analysis of quality

control data are an integral component of a comprehensive quality assurance program and an

important safeguard of project data.

The field procedures implemented by GSR are comprehensive and cover all aspects of the data

collection process such as surveying, drilling, core and RC cuttings handling, description,

sampling and database creation and management. At Wassa, each task is conducted by

appropriately qualified personnel under the direct supervision of a qualified geologist. The

measures implemented by GSR are considered to be consistent with industry best practice.

The quality control employed by GSR to verify the results obtained from the laboratories takes the

form of the following types of check sample:

• Field duplicates to check sampling precision and deposit variability. Two separate

samples are collected at the drill site and bagged separately from which two individual

samples are produced. The results of these checks can be useful in highlighting natural

variability of the grade distribution.

• Pulp duplicates as a check of sampling precision and coarse gold in pulps. Two separate

pulp samples are prepared from a single coarse reject after sample splitting and on site

preparation. The results are useful in indicating problems with sample preparation and

splitting.

• Repeats as a check of analytical precision and coarse gold. Two separate aliquots are

prepared from separate samples taken from the original coarse reject and the two samples

are then checked against one another.

• Blanks for highlighting contamination problems and cross labelling when samples are

mislabelled in the laboratory.

• Standards as a check of analytical precision and accuracy.

11.5 Specific Gravity Data

SG determinations were carried out by GSR. SG is measured on representative core samples from

each drill run. This ensures representative SG data across all rock types irrespective of gold grade.

SG is measured at the core facility using a water immersion method. Each sample is weighed in

air, then coated in wax and weighed in air and immersed in water. Historically, a total of 606

determinations were collected on core samples.

The water immersion methodology is considered to provide accurate estimates of variations in

bulk SG throughout the Wassa gold deposits. After testing, each sample is carefully replaced at its

original location in the core box.

June 2019


Samples were selected from all the different lithologies intersected in the core of all the available

drill holes. The sampling procedure was guided by pit location, lithology, depth, quartz contents

(in oxide) and the oxidation state. A total of nineteen holes from Dead Man’s Hill, South East,

Starter, 419, 242, B-shoot and F-Shoot were selected with the results presented in Table 11-1.

Table 11-1 Specific gravity testing results

Material # Samples SG Value (g/cm3) Standard Error

Oxide 213 1.8 2%

Transition 42 2.19 3%

Fresh 327 2.7 1%

Quartz Vein 24 2.56 1%

Another 13 samples consisting of oxide (9), trans (1), fresh (2) and quartz (1) were sent to the

Western University College (WUC, Tarkwa) as independent checks. The average results were

1.76, 2.29, 2.73 and 2.59 g/cm3 respectively.

The SG determinations are considered accurate as the reconciliations between the mined tonnages

and those estimated from the resource models reconcile well.

A more recent SG study was implemented to see if the higher grade mineralization being mined

underground is heavier than waste rock and the lower grade material mined previously in the open

pits. A total of 40 samples were selected from four underground drill holes and were sent to

Intertek Laboratories for wax immersion SG determinations. The preliminary results from this

study indicate that the higher grade underground mineralization is heavier than the lower grade

open pit material. Gold mineralization at Wassa is directly related to the percentage of pyrite

associated with quartz veining; in general, the higher percentage of pyrite the higher the gold

grades. The underground mining exploits these higher grade areas of the mineralization with

associated higher percentages of sulfides which in turn accounts for the heavier mass of this

material. The results of this ongoing study are summarized in Table 11-2.

Table 11-2 Specific gravity from underground drill holes

Hole ID From To Length Grade g/t Au Avg Avg SG (g/cm3)

BS17-670-27 70.5 98.5 28.0 5.53 2.98

BS17-670-11 80.2 110.2 30.0 4.30 2.83

BS17-645-5 100.5 125.6 25.1 42.62 2.94

BS17-670-23 61.8 78.2 16.4 4.85 3.03

In June 2018, Golden Star started an in-house specific gravity measurement program. A total of

723 samples from surface drill core and 966 samples from underground core have been processed

to date (Table 11-3 and Table 11-4). These results have shown that the average SG for the

underground fresh ore is 2.8. The water displacement method to measure density is employed,

using paraffin sealed core samples.

June 2019


Table 11-3 Specific gravity from Wassa underground drill holes 2018

Rock type # determinations Avg S.G (g/cm3)

Banded Magnetic Mudstone 67 3.02

Diorite 725 2.83

Felsic Intrusive na na

Phyllite 67 2.74

Quartz vein 107 2.65

Total 966 2.81

Table 11-4 Specific gravity from Wassa Surface drill holes 2018


Banded magnetic Mudstone 32 2.70

Diorite 470 2.69

Felsic Intrusive 41 2.59

Phyllite 131 2.63

Quartz vein 49 2.57

Total 723 2.63

The rock types at Hwini Butre – Benso and Chichiwelli have a specific gravity of 2.6 to 2.7 g/cm3

applied, which has been used for tonnage calculations in the resource models. Specific Gravity

work conducted on the 2018 and 2019 drilling from Father Brown and Adiokrom has confirmed

this value and is summarized in Table 11-5.

Table 11-5 Specific gravity from Father Brown Surface drill holes 2018


Diorite 72 2.62

Gabbro 111 2.70

Quartz vein 5 2.56

Total 188 2.63

June 2019


12 Data Verification

12.1 Introduction

Core logging and sampling procedures adopted by GSR are considered to be in line with industry

standards. Consultants have been brought in over the years to assess and validate the logging

against the halved drill core and no major errors have been recorded.

GSR frequently sends “blind” test samples to the laboratory and monthly batch results are analysed

and any anomalous results are queried immediately. A small number of anomalous and/or poor

results have been noted over the years, but these have been identified and the reasons fall into two

main categories, namely:

• Mislabelling of individual samples, standards and blanks.

• Individual batch issues corresponding to changes in the laboratory setup or calibration;

in these cases, re-assay has been carried out.

12.2 Data verification by GSR

The field procedures implemented by GSR involve several steps designed to verify the collection

of exploration data and minimize the potential for inadvertent data entry errors. Data entry and

database management involves two steps punctuated by validation steps by the logging geologist.

Drill hole logs are captured directly into an SQL Acquire database via lap top computers which

are linked to the main database through a fiber optic line linking the core logging facility to the

main Wassa mine site office. Acquire has built in validation tools and draw down menus to

eliminate erroneous data entry during the logging process. Prior to importing the drill hole data

into the resource modeling software, the data is thoroughly checked.

Analytical data is also routinely checked for consistency by GSR personnel. Upon reception of

digital assay certificates, assay results, together with the control samples, are extracted from the

certificates and imported into the Acquire database. Failures and potential failures are examined

and, depending on the nature of the failure, re-assaying is requested from the primary laboratory.

Analysis of quality control data is documented, along with relevant comments or actions

undertaken to either investigate or mitigate problematic control samples.

12.3 Analytical QA/QC

12.3.1 Introduction

GSR relies partly on the internal analytical quality control measures implemented by SGS and

TWL but also implements external analytical quality control measures. These measures involve

using control samples, including blanks and certified reference materials (standards), in sample

batches submitted for assaying.

GSR has supplied QA/QC reports to various consultants over the numerous drilling campaigns

since 2004, and a summary of the historical and current QA/QC results is included here.

QA/QC data for all of 2014, 2015, 2016, 2017 and 2018 are evaluated in this document as they

have been included in the recent resource model update used for this study.

June 2019


12.3.2 Comparison of assay methodologies

In 2003, it was recognized that there was a need to implement an analytical method that could

reproduce assay results, as the conventional 50 g fire assay resulted in poor reproducibility between

field duplicates. This effect was also evident between pulp duplicates; although not as marked.

The conclusion of the analyses of the quality control data available then was that a component of

coarse gold present in the samples was contributing to poor reproducibility and that an analytical

process that makes use of significantly larger aliquots, such as LeachWell™ assays, should be

considered.

To address this, GSR now assays using a 1 kg BLEG assay, with a LeachWellTM accelerant. The

gold grade is determined using an AAS finish. Initially, the sample splitting was completed using

a rotary splitter and a 6 hour leach was used. Following analysis of the leach tailings, the leach

time has been extended from 6 to 12 hours. Due to time constraints, the use of the rotary splitter

has been discontinued and a Jones Riffle has been used to split sub-samples from the larger RC

drillhole samples. The difference between the fire assay and larger BLEG assays are illustrated in

Figure 12-1.

Figure 12-1 HARD plot comparing fire assay and BLEG for field duplicates

Figure 12-1 shows a significant improvement with respect to sample reproducibility between the

fire assay and BLEG methodologies. Using BLEG, 80% of pairs report Half Absolute Relative

Difference (“HARD”) precisions of less than 17%, compared to the 35% precision attributable to

the fire assay method. SRK recommend that GSR continue to monitor the reproducibility of the

sample grades from paired data analysis.

June 2019


12.3.3 Repeat (Coarse Reject) Duplicates 2011 to 2013

GSR no longer submits pulp samples for determining repeatability; but rather, submits coarse

reject samples (from the laboratory sample split). These coarse rejects are re-numbered and re-

submitted to the laboratory for repeat analysis. The coarse duplicates are intended to monitor the

sample preparation section of the laboratory. The majority of the drilling for the current resource

updates was conducted between 2011 and April 2018; therefore, the QA/QC data for this period

has been included in this document.

The HARD plot of all coarse rejects for 2011 is presented in Figure 12.2. The results of this HARD

analysis show that approximately 89% of the 369 coarse duplicate samples fall within

approximately 20% error and 76% fall within 10% error. This is acceptable for gold deposits of

this type.

Figure 12-2 HARD plot of all coarse rejects (2011) from SGS


analysis show that approximately 83% of the 2,173 coarse duplicate samples fall within

0

1

2

3

4

5

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 20% 40% 60% 80% 100%

Mea

n G

rad

e (g

/t)

HA

RD

Rank Percentile %

2011 REPEAT DUPLICATES HARD ANALYSIS N=369

HARD

Mean Grade

5 per. Mov. Avg. (Mean Grade)

June 2019


approximately 20% error and 60% fall within 10% error. This is acceptable for gold deposits of

this type.




approximately 20% error and 56% fall within 10% error. This is considered to be acceptable for a

gold deposit.

0

5

10

15

20

25

30

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 20% 40% 60% 80% 100%

Mea

n G

rad

e g/t

HA

RD

%

Rank Percentile %

WASSA COARSE REJECTS HARD ANALYSIS 2012-N=2,173

HARD

Mean Grade

5 per. Mov. Avg. (Mean Grade)

June 2019



12.3.4 QA/QC Data Summary between 2014 and early 2017

The analytical quality control data produced between 2014 and early 2017 is summarized in Table

12-1. The quality control data produced on this project represents approximately 16% of the total

number of samples.

Table 12-1 Summary of analytical quality control data from 2014 to early 2017

SGS (%) WGS (%) Total (%) Comment

Sample Count 61,943 96,596 158,539

Blanks 622 1.00% 6,159 6.38% 6,781 4.28% Coarse sand QC Samples 4,564 7.37% 4,302 4.45% 8,866 5.59%

ST07/9453 575 766 1,341 0.21 g/t Au ST14/9501 405 405 0.43 g/t Au ST16/9487 264 419 683 0.49 g/t Au ST626 664 664 0.51 g/t Au ST06/9481 89 280 369 1.02 g/t Au ST06/7384 167 167 1.08 g/t Au ST588 763 763 1.60 g/t Au ST39/6373 168 168 1.67 g/t Au ST602 324 324 1.91 g/t Au ST482 635 516 1,151 1.94 g/t Au ST575 476 476 2.43 g/t Au G914-2 14 14 2.45 g/t Au ST596 61 61 2.51 g/t Au ST37/6374 30 30 3.33 g/t Au ST43/7370 955 955 3.37 g/t Au G910-3 12 12 4.03 g/t Au ST48/8462 175 175 4.82 g/t Au

0

10

20

30

40

50

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 20% 40% 60% 80% 100%

Mea

n G

rad

e (g

/t)

HA

RD

Rank Percentile %

WASSA COARSE REJECTS HARD ANALYSIS 2013 N= 2,962

HARD

Mean Grade

5 per. Mov. Avg.

(Mean Grade)

June 2019


ST517 1108 1,108 5.23 g/t Au Grade Control Field Duplicates 6,567 10.60% 6,567 4.14%

Coarse Reject Duplicates 3,802 3.94% 3,802 2.40%

Total QC Samples 11,753 18.97% 14,263 14.77% 26,016 16.41%

12.3.5 Repeat (Coarse Reject) Duplicates 2014 to Dec 2018

As indicated in the previous section on field duplicates, Golden Star no longer submits laboratory

pulp duplicate samples for analysis. Instead, coarse reject samples from the SGS and TWL sample

split are re-numbered and re-submitted for repeat analyses and are intended to monitor the sample

preparation at the laboratory. Field duplicates are collected at the drill site and bagged separately,

from which two individual samples are produced.

Rank half absolute difference (HARD) plots of coarse reject duplicates processed by SGS and

TWL suggest that approximately 56 to 67% of gold assay samples have HARD below 10%. This

variance is typical of coarse reject duplicate pairs in a gold deposit, indicating that SGS and TWL

was able to reasonably reproduce this type of paired data. At (Wassa site Lab) WGS, HARD plots

suggest that approximately 46 to52% of the grade control field duplicate gold assay samples have

HARD below 10% which indicates that WGS was also able to reasonably reproduce the field

duplicate pairs. As expected, the variance of field duplicate sample pairs processed by WGS is

higher than the internal laboratory coarse duplicate paired data processed by SGS and TWL.

The HARD plot of all coarse rejects for 2014 is presented in Figure 12-5. The results of this HARD



gold deposit.

June 2019






gold deposit.

-100%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

100%

0.01 0.1 1 10 100

HR

D (

%)

Individual Mean (Au g/t)

Mean versus Half Relative Deviation Plot(SGS; Core Samples)

Au assay

0% Line

N = 2145 pairs

y = 1.4072xR² = 0.9193

0

5

10

15

20

0 5 10 15 20

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)

Original Assays (Au g/t)

Bias Chart Lab Pulp Duplicate Assay Pairs (0-20 g/t Au)(SGS; Core Samples)

2014 B-Shoot Lab Duplicates

+10%

-10%

N = 2145 pairs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

HA

RD

(%

)

Rank

Ranked Half Absolute Relative Deviation Plot(SGS; Core Samples)

Au assayN = 2145 pairs

56.1%

y = 1.4072xR² = 0.9193

0

100

200

300

400

500

600

700

800

0 100 200 300 400 500 600 700 800

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)


Bias Chart Lab Pulp Duplicate Assay Pairs (0-800 g/t Au)(SGS; Core Samples)


+10%

-10%

N = 2145 pairs

0.01

0.1

1

10

100

0.01 0.1 1 10 100

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)


Q-Q Plot Lab Pulp Duplicate Assay Pairs(SGS; Core Samples)

N = 2145 pairs

0%

1%

10%

100%

0.01 0.1 1 10 100

HA

RD

(%

)


Mean versus Half Absolute Relative Deviation Plot(SGS; Core Samples)

N = 2145 pairs

June 2019






gold deposit.

-100%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

100%

0.01 0.1 1 10 100

HR

D (

%)



Au assay

0% Line

N = 637 pairs

y = 0.8789xR² = 0.9465

0

5

10

15

20

0 5 10 15 20

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)


Bias Chart Coarse Reject Duplicate Assay Pairs (0-20 g/t Au)(SGS; Core Samples)


+10%

-10%

N = 637 pairs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

HA

RD

(%

)

Rank



66.6%

y = 0.8789xR² = 0.9465

0

20

40

60

80

100

120

0 20 40 60 80 100 120

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)




+10%

-10%

N = 637 pairs

0.01

0.1

1

10

100

0.01 0.1 1 10 100

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)


Q-Q Plot Lab Coarse Reject Duplicate Assay Pairs(SGS; Core Samples)

N = 637 pairs

0%

1%

10%

100%

0.01 0.1 1 10 100

HA

RD

(%

)



N = 637 pairs

June 2019






gold deposit.

-100%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

100%

0.01 0.1 1 10 100

HR

D (

%)



Au assay

0% Line

N = 332 pairs

y = 0.9321xR² = 0.8852

0

5

10

15

20

0 5 10 15 20

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)



2016-2017 B-Shoot Lab Duplicates

+10%

-10%

N = 332 pairs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

HA

RD

(%

)

Rank



59.3%

y = 0.9321xR² = 0.8852

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)



2016-2017 B-Shoot Lab Duplicates

+10%

-10%

N = 332 pairs

0.01

0.1

1

10

100

0.01 0.1 1 10 100

Co

ars

e R

eje

ct D

up

lic

ate

As

sa

ys

(A

u g

/t)


Q-Q Plot Coarse Reject Duplicate Assay Pairs(SGS; Core Samples)

N = 332 pairs

0%

1%

10%

100%

0.01 0.1 1 10 100

HA

RD

(%

)



N = 332 pairs

June 2019


Figure 12-8 HARD plot of all coarse rejects (2017) from SGS and Intertek

The HARD plot of all coarse rejects for the first quarter of 2018 is presented in Figure 12-9. The

results of this HARD analysis show that approximately 77% of the 2399 coarse duplicate samples

fall within approximately 20% error and 56% fall within 10% error. This is considered to be

acceptable for a gold deposit.

0

10

20

30

40

50

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 20% 40% 60% 80% 100%

Mea

n G

rad

e (g

/t)

HA

RD

Rank Percentile %

WASSA COARSE REJECTS HARD ANALYSIS 2017 N= 750

HARD

June 2019


Figure 12-9 HARD plot of all coarse rejects (2018) from Intertek

12.3.6 Certified Reference Material (“CRM”)

CRM material was introduced by GSR into the sample stream to monitor the accuracy, precision

and reproducibility of the assay results. CRM materials were sourced from Geostats Pty Ltd.

(“Geostats”), and from Gannet Holdings Pty Ltd. Although the CRM material could be easily

identified by the laboratory, the actual grade of the standard would be difficult to determine due to

the large number of different standards used. Standards in use between 2003 and 2018 are shown

in Table 12-2 to 12-8.

Laboratory performance has improved significantly in the last fifteen years and this can be seen in

the laboratory bias on a year to year basis.

A total of 16,100 standards were submitted to SGS between 2008 and 2017. Standards submitted

to SGS largely performed within expected ranges and mean grades are similar to expected values.

Results indicate that SGS reports both higher and lower than expected values, with some variation

to the detection limit; however, 96% or more of the determinations typically fell within +/–5% of

the expected value. Standards submitted to SGS from 2014 to 2017 performed much better with

100% of the determinations falling within +/–3% of the expected value and 75% falling within +/-

2% of the certified reference value.

A total of 4,320 standards were submitted to the Wassa site laboratory between 2014 and 2017.

Standards analyzed by Wassa site lab performed marginally worse with a number of individual

June 2019


samples beyond two standard deviations of the expected value. These results could possibly be

due to the mislabeling of samples. However, 100% or more of the determinations typically fell

within +/–2% of the expected value.

In 2018, GSR began using TWL located in Tarkwa. A total of 400 CRM were submitted in 2018,

with 89 percent or more of the determinations typically falling within +/–2% of the expected value.

In general, the performance of the standards inserted with samples submitted for assaying at SGS,

TWL and Wassa site laboratories is acceptable. The majority of the failures appear to be caused

by the mislabelling of samples.

Table 12-2 CRM for 2003 to 2007 (TWL)

Standard Certified Mean

(g/t Au)

Number Samples

Submitted

Mean Assay Grade

(g/t Au)

Laboratory Bias

(%)

Gannet A 0.22 196 0.22 0%

Gannet B 2.52 185 2.57 2%

Gannet C 3.46 21 3.53 2%

Gannet D 3.40 75 3.40 0%

Gannet E 2.36 77 2.45 4%

Gannet F 0.78 47 0.75 -4%

Gannet G 3.22 82 3.02 -6%

Gannet M 1.18 159 1.28 +2%

Gannet N 0.50 171 0.49 -2%

Table 12-3 Geostats CRM for 2008 to 2012 (SGS)


(g/t Au)

Number Samples

Submitted

Mean Assay Grade (g/t

Au)

Laboratory Bias

(%)

G901-10 0.48 82 0.51 6%

G305-3 0.71 14 0.66 -7%

G901-2 1.70 32 1.54 -9%

G906-4 1.90 137 1.99 5%

G999-4 2.30 36 2.40 4%

G302-2 2.44 70 2.50 2%

G901-1 2.50 38 2.38 -5%

G396-9 2.60 29 2.39 -8%

G900-7 3.19 193 3.22 1%

June 2019


Table 12-4 Gannet CRM for 2008 to 2012 (SGS)


(g/t Au)

Number Samples

Submitted

Mean Assay Grade

(g/t Au)

Laboratory

Bias (%)

ST 07/9453 0.21 476 0.21 2%

ST 14/9501 0.43 447 0.42 -3%

ST 16/9487 0.49 110 0.51 3%

ST 16/5357 0.52 654 0.52 0%

ST 486 0.57 124 0.54 -5%

ST 17/2290 0.78 14 0.79 2%

ST 481 1.02 32 1.05 3%

ST 06/5356 1.04 115 1.06 2%

ST322 1.04 18 1.07 3%

ST 06/7384 1.08 1881 1.04 -4%

ST 384 1.08 173 1.06 -2%

ST 39/6373 1.67 117 1.74 4%

ST 09/7382 1.93 205 1.87 -3%

ST 482 1.94 695 1.98 2%

ST 5355 2.37 145 2.39 1%

ST 05/9451 2.45 538 2.53 3%

ST 05/6372 2.46 168 2.44 -1%

ST 05/2297 2.56 78 2.49 -3%

ST 486 2.63 49 2.59 -5%

ST 10/9298 3.22 132 3.30 3%

ST 37/6374 3.33 129 3.08 -7%

ST 43/7370 3.37 834 3.33 1%

ST 5359 3.91 131 3.97 1%

ST 359 3.93 87 3.96 1%

ST 48/8462 4.82 508 4.89 1%

Table 12-5 Gannet CRM for 2013 (SGS)

Standard Certified Mean Number Samples

Submitted

Mean Assay Grade (g/t

Au)

Laboratory Bias

(%) (g/t Au)

ST07/9453 0.21 645 0.22 4%

ST14/9501 0.43 402 0.50 17%

ST06/7384 1.08 39 1.05 -3%

ST482 1.94 528 1.99 2%

ST05/6372 2.46 665 2.48 1%

ST37/6374 3.33 579 3.29 -1%

ST48/8462 4.82 187 4.89 1%

June 2019


Table 12-6 Gannet CRM for 2014 to 2017 (SGS)


Submitted

Mean Assay Grade

(g/t Au)

Laboratory Bias

(%) (g/t Au)

ST07/9453 0.21 575 0.21 0%

ST16/9487 0.49 264 0.49 0%

ST626 0.51 664 0.50 -2%

ST06/9481 1.02 89 1.03 1%

ST06/7384 1.08 167 1.05 -3%

ST602 1.91 324 1.97 3%

ST482 1.94 635 2.00 3%

ST575 2.43 476 2.44 0%

G914-2 2.45 14 2.46 0%

ST596 2.51 61 2.51 0%

G910-3 4.03 12 3.96 -2%

ST48/8462 4.82 175 4.91 2%

ST517 5.23 1108 5.20 -1%

Table 12-7 Gannet CRM for 2014 to 2017 (Wassa Site Lab)


Submitted

Mean Assay Grade

(g/t Au) Laboratory Bias (%)

(g/t Au)

ST07/9453 0.21 766 0.21 0%

ST14/9501 0.43 405 0.43 0%

ST16/9487 0.49 419 0.50 2%

ST06/9481 1.02 280 1.00 -2%

ST588 1.6 763 1.61 1%

ST482 1.94 516 1.95 1%

ST37/6374 3.33 30 3.31 -1%

ST43/7370 3.37 955 3.36 0%

ST39/6373 1.67 168 1.66 -1%

Table 12-8 Gannet CRM for 2018 (Intertek)


Submitted

Mean Assay Grade

(g/t Au)

Laboratory Bias

(%) (g/t Au)

G913-10 7.10 14 6.94 -2%

G915-3 9.22 23 8.98 -3%

G911-4 2.45 32 2.45 0%

G316-7 5.79 22 5.78 0%

G314-5 5.30 42 5.23 -1%

G314-3 6.68 7 6.59 -1%

ST588 1.6 69 1.60 0%

ST575 2.43 40 2.48 2%

ST37/6374 3.33 22 3.15 -6%

ST43/7370 3.37 34 3.29 -2%

ST73-8281 1.52 95 1.51 0%

June 2019


12.3.7 Blanks

Blank samples are routinely inserted into the sample stream to check for possible sample

contamination during the preparation and assaying process. Typically, blanks are inserted prior to

the delivery of samples for preparation and analyses. For RC samples, a blank is inserted before

the splitting process to monitor possible contamination occurring during the splitting of the original

sample collected from the drill cyclone.

The blank material used by Golden Star consisted of coarse sand. The samples sent to SGS’

laboratory consistently yielded values at or below the detection limit, with zero samples yielding

a value over 10 times the detection limit of gold. With no failures, the sample blanks performed

extremely well and indicate minimal, if any, sample contamination during processing.

Blank material processed at the Wassa site laboratory performed poorer, with numerous samples

yielding values close to, or above, 10 times the detection limit of gold. Over time, from 2014 to

2016, the blank samples’ performance noticeably declined. Further investigation of anomalously

high values indicates contamination in the sample preparation process in a number of cases. The

Wassa site laboratory was only used for the determination of open pit grade control samples, all

other samples are analysed using either SGS or TWL in Tarkwa. Golden Star will investigate the

cause of the elevated gold values causing blanks to fail at the site laboratory.

The blank assay data from 2011 to 2018 includes 3,877 assays, all assayed by SGS and later by

TWL. Summary statistics for the assays of blanks returned by the labs are shown in Table 12-9.

Table 12-9 Blank Sample Summary Statistics 2011 to Q1 2018

Sample type Year Count Minimum

(g/t)

Maximum

(g/t)

Median Mean

(g/t) (g/t)

Blanks 2011 278 0.01 0.01 0.01 0.01

Blanks 2012 194 0.01 0.27 0.01 0.01

Blanks 2013 210 0.01 0.11 0.01 0.01

Blanks 2014 56 0.01 0.07 0.01 0.02

Blanks 2015 69 0.01 0.03 0.01 0.01

Blanks 2016 553 0.01 0.03 0.01 0.01

Blanks 2017 930 0.01 0.04 0.01 0.01

Blanks 2017 498 0.01 0.03 0.01 0.01

Blanks 2018 1089 0.005 0.66 0.01 0.01

12.3.8 Umpire Laboratory Performance

Laboratory checks are performed occasionally to check on the reliability of the primary laboratory,

in this case SGS, Tarkwa. In 2013 and 2014, “round robin” sample check studies were conducted

using SGS, TWL and the Wassa site laboratory.

June 2019


In 2014, 252 quarter core samples were selected from drilling conducted between 2012 and

mid-2014. The intersections selected were HG intervals which averaged approximately 17 g/t Au.

As the coarse sample reject was no longer available for these intervals, a new sample was collected

by cutting the half core on file in the core yard. Golden Star has conducted similar quarter core

sampling studies on other deposits and repeatability of the original results is often not high due to

the change in sample size going to half the volume from the original sample. The Wassa quarter

core sampling study concluded the same as previous studies with good repeatability between the

original sample and its coarse sample reject and much poorer repeatability with the quarter core

sample. The average grade for both the original assay and the coarse sample reject duplicate

compare well at 17 g/t Au and the quarter core sample was less at 12 g/t Au. However, control

sample standards that were submitted with these sample batches consistently came up with a

negative bias, as seen in Table 12-10, so this can partially account for the lower average. The

HARD plots shown in Table 12-11 show the good correlation between the original assay value

and the coarse sample reject duplicate, but these do not repeat well when using the quarter core

samples analysed at TWL Laboratory. Although the negative lab bias and the smaller sample

volume attributes to poor repeatability, the Wassa deposit has a high nugget gold distribution

which alone will result in poor repeatability. The variability of the gold distribution was recognized

and GSR has put in sample protocols to help reduce the variability, i.e. larger sample volumes,

BLEG leach well analysis.

Table 12-10 Gannet CRM for Quarter Core Sample Analysis (Intertek)

Standard Certified Mean Number

Samples

Submitted

Mean Assay Grade (g/t Au) Laboratory Bias (%)

(g/t Au)

ST517 5.23 5 4.98 -5%

ST482 1.94 9 1.78 -8%

ST16/9487 0.49 14 0.46 -6%

Table 12-11 Summary HARD Plot Results for Quarter Core Sample Analysis

Laboratory No of

Samples

<10%

HARD

<15%

HARD

<20%

HARD

CORRELATION

COEFF ®

Orig SGS vrs Check SGS 252 65% 81% 90% 0.94

Orig SGS vrs Check Intertek 252 32% 45% 57% 0.60

SGS Check vrs Intertek Check 252 29% 44% 55% 0.45

In 2013, 120 RC samples were split into three samples which were sent to each of the laboratories

for gold analysis. The sample batches also contained control samples to monitor the precision of

the individual laboratories.

The three laboratories all performed well with the best correlation being between SGS and the

Wassa site laboratory. The HARD plots for the laboratory comparisons are shown below in Table

12-12.

June 2019


Table 12-12 Summary HARD Plots 2013 Round Robin Results

LAB <10% HARD <15% HARD <20% HARD CORRELATION COEFF

SGS vrs Wassa 65% 84% 92% 97%

SGS vrs Intertex 68% 84% 88% 97%

Wassa vrs Intertek 71% 84% 90% 98%

The high correlation at 20% HARD for all the labs demonstrates how the larger RC chip samples

give a better representation of grade. Approximately 90% of the 120 RC samples submitted for

this study show a 20% error, compared to the coarse reject core samples submitted in 2013 and

2014 which show a correlation of approximately 80% of the data set with 20% error.

In 2012, a “round robin” exercise was undertaken to have an independent check on the reliability

of Au assay results from the primary laboratory, SGS. A total of 10% of all assays from the 1 m

samples received each month were randomly picked from the data set. The data is grouped into

six separate ranges, namely 0.00 to 0.50 g/t, 0.50 to 0.90 g/t, 0.90 to 1.20 g/t, 1.20 to 2.00 g/t, 2.00

to 2.50 g/t and greater than 2.50 g/t. The selection in each range is manipulated until the 10% is

achieved with a bias towards the mineralized intervals.

Three samples, each weighing about 3 kg were prepared from each original sample bag using the

one stage riffle splitter. Four batches of 175 samples including duplicates and standards were

dispatched to SGS, WSL, TWL, and ALS Minerals in Ghana-Kumasi (“ALS”). All samples were

labelled with the same identification numbers. A total of 157 assays were returned by each

laboratory for analysis.

Statistical comparison of the data indicates that ALS returned lower grades and variance than SGS,

WSL and TWL. SGS and TWL correlated well with similar minimum and maximum grades, and

standard deviation population distribution. The descriptive statistics from the round robin exercise

are included in Table 12-13.

Table 12-13 Round-robin Descriptive Statistics

Laboratory Count Minimum (g/t) Maximum (g/t) Mean (g/t) Variance Std

Dev

SGS 157 0.01 12.0 1.33 1.75 1.32

WSL 157 0.01 8.9 1.09 1.47 1.21

TWL 157 0.01 11.68 1.15 1.68 1.30

ALS 157 0.01 9.32 1.02 1.31 1.15

In 2017, prior to switching from SGS to TWL in Tarkwa, GSR submitted 578 samples to both

laboratories, inclusive of CRM. Statistical comparison of the data indicates that TWL returned

slightly lower grades and variance than SGS. SGS and TWL correlated well with similar minimum

and maximum grades, and standard deviation population distribution. The descriptive statistics

from the round robin exercise are included in Table 12-14.

Table 12-14 Round-robin Descriptive Statistics 2017

Laboratory Count Minimum

(g/t)

Maximum

(g/t) Mean (g/t) Variance Std Dev

SGS 584 0.01 113.00 3.33 60.79 7.80

June 2019


TWL 584 0.01 109.20 3.21 55.81 7.47

When Comparing the results from the two laboratories the HARD analysis shows that

approximately 84% of the 584 repeat samples fall within approximately 20% error and 69% fall

within 10% error. This is a good correlation between the two laboratories and decision was made

to switch over from SGS to TWL. The HARD results for the comparison between the two

laboratories are shown in table 12-15.

Table 12-15 Summary HARD Plots 2017 Round Robin Results SGS - TWL

LAB <10% HARD <15% HARD <20%

HARD CORRELATION COEFF

SGS vrs TWL 69% 78% 84% 97%

June 2019


13 Mineral Processing and Metallurgical Testing

13.1 Historical Testing

On obtaining ownership of the Project in 2002, GSR commissioned a FS for a CIL operation with

the process engineering component undertaken by Metallurgical Process Development Pty Ltd.

(now known as “MDM”). The FS was completed in 2003. The metallurgical testwork conducted

in support of the MDM FS was conducted on samples from the Wassa area. Samples were

originally sent to SGS Lakefield in Johannesburg for both variability and bulk sample testwork.

Further variability testwork was conducted at AMMTEC in Perth.

A total of 24 variability samples were tested; 10 of fresh mineralized material, six of oxide, and 8

samples taken from the existing (now decommissioned and reclaimed) HL operation. Four bulk

samples were also tested, representing fresh, oxide, HL phase 1 and HL phase 2. The samples were

all taken from the Wassa Main area.

At a grind size of 75% -75 µm, and a 24-hour leach time, the fresh bulk sample achieved a leach

recovery of 92%. The Bond Ball Mill Work Index (“BWi”) for this sample was 14.8 kWh/t. Under

the same conditions, the oxide bulk sample achieved a leach recovery of 93%. The BWi for this

sample was reported as 8 kWh/t. Minor preg-robbing behaviour was noted, and gravity recovery

testwork indicated that plant recoveries of 30 to 40% could be expected from a gravity circuit.

13.2 Recent Metallurgical Testwork

13.2.1 Introduction

Within the framework of the NI 43-101, 2015 Wassa FS Report to evaluate the Wassa underground

mine further metallurgical testwork was completed. It is anticipated that the higher grade feed

would be blended with the open pit ore sources. The testwork evaluated the performance of future

potential feed material from underground mining using a series of samples taken from available

half core remaining from ore resource drilling and the physical characteristics and metallurgical

response from these were compared to those a reference sample of current plant feed.

At the time of testwork an exploration decline and bulk sample was obtained from underground,

which was expected to be reasonably representative of the remaining underground feed material.

The benefit of bulk sample treatment through the plant resulted in a reduced testwork program that

included a series of 6 variability and 4 crushability samples that were compared to a reference

sample taken from the current open pit ore feed.

The metallurgical testwork was undertaken by SGS in Cornwall, UK and the sample were

delivered and logged in around the middle of December 2014 with this initial phase of testwork

completed and the draft report issued in early April 2015.

13.2.2 Metallurgical Variability, Crushability and Reference Samples

For the purpose of the metallurgical program, the resources envisaged to be processed over the

underground LoM were differentiated spatially by GSR into six underground domains or zones:

June 2019


• Zone 1 Upper

• Zone 1 Lower

• Zone 2 Upper

• Zone 2 Lower

• Zone 3 Upper

• Zone 3 Lower

These nominal ore zones are depicted in Figure 13-1 with further details presented in Table 13-1.

Table 13-1 Ore zones represented by the variability samples

Northing Relative Level Tonnes Grade Contained

From To From To g/t Au Metal (oz) %

tonnes % Metal

Zone 1 upper 20200 19937.5 857 682 598,485 4.74 91,185 15% 14%

Zone 1 lower 20200 19937.5 682 607 707,388 6.78 154,089 18% 23%

Zone 2 upper 19937.5 19690 782 632 723,358 6.28 146,054 18% 22%

Zone 2 lower 19937.5 19690 632 507 537,555 4.32 74,696 14% 11%

Zone 3 upper 19690 19500 657 557 772,212 5.02 124,642 20% 19%

Zone 3 lower 19690 19500 557 482 613,447 4.20 82,797 16% 12%

Total Resource 3,952,446 5.30 673,462 100% 100%

Figure 13-1 View looking east-west of metallurgical sample locations

The metallurgical variability and crushability samples were selected based on available material

from the drilling programme to represent the six nominated zones. Six variability samples were

June 2019


selected, one for each zone from available HQ and NQ half cores remaining from the previous ore

resource assay sampling program. These core sections were further cut in half, with one section

used for the metallurgical testwork and the remaining quarter core sections retained on site for

reference. The six quarter core composite samples prepared for metallurgical investigations were

targeted to represent as far as possible the full width of the anticipated mining stopes across each

zone. Each sample of quarter cores weighed between 50 and 60 kg.

As quarter core samples were not suitable for the ore hardness (crushability) investigations due to

the limited physical size of each sample (samples with a minimum of 35 mm in two dimensions

are required), four full core samples (with a segment removed for assay purposes) from a recent

drilling programme were selected for the crushability tests. Each crushability sample consisted of

7 lengths of HQ drill core each approximately 200 mm in length. From these, three samples were

prepared for the UCS tests with the remaining core sections along with the remaining material

from the UCS test sample prepared for the Bond crushability index (low energy crushing) tests.

A single reference sample was also obtained by hand selection from the workings in the Starter pit

area at around the 910 m level. Around 100 kg of material was taken and this sample was used for

both metallurgical and crushability testwork.

Table 13-2 Summary and location of testwork samples

Sample Type Detail

Northing Easting Relative Level Number of Separate Sub-samples / Intersections

From To From To From To Average

m m m

Reference 20,419.6 20,396.4 40,003.9 39,973.8 910 910 910 6

Variability Z1U 19,971.8 20,042.9 40,112.9 39,983.6 828 682 763 6

Variability Z1L 19,946.8 19,987.8 39,994.1 39,911.5 678 615 664 7

Variability Z2U 19,770.0 19,846.4 40,084.4 39,930.3 753 653 713 5

Variability Z2L 19,700.2 19,757.1 40,079.2 39,930.9 602 530 575 6

Variability Z3U 19,531.0 19,576.1 40,023.4 39,978.8 602 562 585 4

Variability Z3L 19,497.1 19,565.1 40,040.3 39,945.0 555 510 533 5

Crushability 1 BSDD347MET 19,491.6 19,488.5 40,023.6 39,998.5 587 514 553 8

Crushability 2 WMET4 20,052.6 20,050.2 40,014.3 39,998.8 767 748 753 8



The locations of the reference, variability and crushability samples related to the resource / mining

blocks are presented in Table 13-2 along with the nominal ore zones selected. Some of the

crushability samples selected were adjacent to rather than completely within the representative ore

zone. Crushability 1 was from depth to the south of Zone 3 Lower while the other crushability

samples were from different depth within Zone 1 Upper and Zone 1 Lower. The reference sample

was taken from the current workings in the starter pit area above and to the north of Zone 1 Lower

at 910 mRL.

13.2.3 Details of Metallurgical Testwork

The metallurgical evaluation testwork programme included the following investigations:

• Scope of work for reference and variability samples:

o elemental scan: ICP multi-element analysis;

o analysis of sulphide and total sulphur;

June 2019


o analysis of carbonate and graphitic carbon;

o diagnostic leach (gold deportment tests);

o BWi; and

o Bond abrasion index (“Ai”).

• Standard flowsheet treatment tests – to confirm recoveries and reagent additions /

consumptions:

o grind calibration tests;

o gravity concentration;

o cyanide leaching of the gravity tails with pre-aeration; and

o settling tests.

• Scope of work for crushability and reference samples:

o unconfined compressive strength (“UCS”);

o Bond low impact crushing work index (“CWi”);

o BWi; and

o Ai.

13.3 Testwork Findings

13.3.1 Head Grade and Elemental / Chemical Analyses

The gold and silver head grades were determined by milling and screening at 106 µm with fire

assay of the two screen fractions. The results are summarized in Table 13-3.

Table 13-3 Screened head assay results

Sample

Overall Size fraction +106 micron

Size fraction -106 micron

Gold Distribution

Silver Distribution

Grade Grade Grade % % % %

g/t Au g/t Ag % g/t Au g/t Ag g/t Au g/t Ag +106 mic

-106 mic

+106 mic

-106 mic

Reference 1.53 0.10 1.88 11.32 0.20 1.14 0.10 13.91 86.09 3.68 96.32

Zone 1 Upper 6.51 0.43 2.36 28.29 1.60 7.03 0.40 10.28 89.72 8.83 91.17

Zone 1 Lower 7.99 0.63 2.26 42.29 4.20 7.31 0.60 11.95 88.05 15.00 85.00

Zone 2 Upper 5.11 0.36 1.26 17.26 1.00 4.38 0.30 4.25 95.75 3.52 96.48

Zone 2 Lower 4.64 0.21 2.35 9.94 0.80 4.52 0.20 5.03 94.97 8.77 91.23

Zone 3 Upper 4.07 0.45 1.57 9.45 0.60 4.42 0.50 3.65 96.35 2.09 97.91

Zone 3 Lower 5.26 0.55 2.14 25.30 2.80 5.29 0.50 10.31 89.69 10.93 89.07

In all samples, the gold and silver analyses in the coarse fraction (+106 µm) is higher than for the

finer fraction (-106 µm).

An ICP elemental scan was undertaken on the reference and variability samples; in addition, the

total carbon and organic carbon as well as the total sulphur and sulphide sulphur were analysed

using the Leco method. Results are presented in Table 13-4.

June 2019


Table 13-4 Elemental and chemical analysis results

Sample 1010A 2008A 3008A 4008A 5008A 6007A 7007A

(%) REF1 Z1U Z1L Z2U Z2L Z3U Z3L

Cu 0.003 0.019 0.011 0.01 0.01 0.008 0.01

Pb <0.001 0.002 <0.001 <0.001 <0.001 <0.001 0.002

Zn 0.006 0.009 0.01 0.008 0.009 0.008 0.007

As <0.001 0.001 0.003 0.001 0.001 0.001 <0.001

Cd <0.0001 0.0003 0.0003 0.0003 0.0002 0.0002 0.0002

Ni 0.002 0.004 0.004 0.002 0.002 0.005 0.003

Co <0.001 0.003 0.004 0.004 0.003 0.003 0.003

Mn 0.07 0.14 0.18 0.2 0.15 0.1 0.13

Bi <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

Sb <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

Hg <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

Te <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

Se <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

SiO2 78.46 74.96 65.39 66.51 59.42 65.39 57.55

Al 3.32 3.48 4.46 4.37 5.22 4.65 5.24

Fe 2.83 5.57 6.46 5.48 4.67 3.92 4.62

Mg 0.74 0.88 1.09 1.27 1.53 1.47 1.8

Cr 0.03 0.06 0.05 0.03 0.02 0.01 0.01

Ca 1.82 1.1 1.81 2.14 3.47 2.71 3.77

S 0.46 0.86 1.56 0.98 1.3 1.17 0.9

Na 0.92 0.96 1.46 1.93 1.98 1.57 2.16

K 1.36 1.7 1.79 1.57 1.38 2.11 1.6

% S (total) 0.46 0.86 1.56 0.98 1.3 1.17 0.9

% S (soluble) 0.02 0.03 0.04 0.04 0.04 0.03 0.03

% S (sulphide) 0.44 0.83 1.52 0.94 1.26 1.14 0.87

% C (total) 1.4 1.42 1.69 1.99 2.22 1.86 2.52

% C (organic) 0.03 0.02 0.03 0.02 0.03 0.02 0.02

% C (CO3) 1.37 1.4 1.66 1.97 2.19 1.84 2.5

As expected,the level of sulphide sulphur is higher in the higher grade variability samples than in

the reference sample. Similarly, the level of iron (Fe) and other base metals is higher; however,

the levels of the other base metals is still seen to be relatively low.

13.3.2 Diagnostic Leach

Diagnostic leaching is a method of quantifying the indicated deportment of gold in a sample and

the relative ease or difficulty with which the gold can be recovered. The sample is prepared by

grinding to a typical grind size likely to be employed (75% < 75 µm was selected) and is subject

to a cyanide leach to dissolve the free gold. The solids from the initial cyanide leach test are then

sequentially pre-treated with more aggressive acids to dissolve minerals that could be

encapsulating the residual gold and following each pre-treatment stage the sample is again treated

by cyanide leaching. As the level of sulphide minerals was indicated to be higher in the higher

grade underground material from the geological interpretation of the core samples and confirmed

from the elemental analyses presented in Table 13-4 the aim was to determine whether the

increased level of sulphide minerals was resulting in the samples being more refractory to

treatment for the recovery of gold.

In the diagnostic leach procedure, the samples are subject the following leach and pre-leach

treatments:

June 2019


• Direct cyanidation: recovers free and exposed gold.

• Hydrochloric acid pre-treatment: liberates gold encapsulated in carbonates, pyrrhotite,

galena and iron hydroxide minerals.

• Sulphuric acid (oxidative) pre-treatment: liberates gold encapsulated in sphalerite, labile

copper sulphate and labile base metal sulphide minerals.

• Nitric acid pre-treatment: liberates gold encapsulated in pyrite, arsenopyrite and

marcasite.

• Carbon combustion: burns off any organic carbon releasing gold that had previously been

adsorbed by the carbon and not therefore amenable to recovery by cyanide leaching.

Residual gold and silver present after the above tests represent gold encapsulated in silica and other

non-reactive gangue minerals.

Results of the diagnostic leach tests for gold are summarized in Table 13-5 and represented

graphically showing the deportment of gold in the samples in Figure 13-2.

Table 13-5 Summary of diagnostic leach results

Gold Deportment

Sample Reference

Ref 1 Z1U Z1L Z2U Z2L Z3U Z3L

% % % % % % %

Cyanide Soluble 91.90 96.82 97.05 93.13 86.92 89.50 85.34

In Carbonates / Pyrrhotite 1.38 0.88 1.10 1.70 8.83 2.37 2.99

In Sphalerite and Labile Sulphides 0.66 0.58 0.23 0.73 1.22 0.97 2.18

In Pyrite and Arsenopyrite 2.53 1.22 1.26 3.30 1.91 4.01 7.01

In Graphitic Carbon 0.59 0.27 0.10 0.35 0.38 0.45 0.40

Residual Gold 2.93 0.23 0.25 0.79 0.74 2.71 2.08

TOTAL 100.00 100.00 100.00 100.00 100.00 100.00 100.00

June 2019


Figure 13-2 Comparative indicated deportment of gold from diagnostic leach results

The recoveries are based on the back-calculated head grade from the gold recovered in the various

steps and the remaining gold in the final tails. A reconciliation of the back-calculated head grade

with the assay head grade is presented Table 13-2, which also presents the correlation between the

assay head grade and back-calculated head grade from the gravity / leach tests.

The results generally indicate that the mineralogy and metallurgy of the samples are somewhat

different with some samples appearing to have potentially more gold locked or associated with

different sulphide minerals and others less when compared to the reference sample. Two samples

(Z3U and Z3L) show potentially higher levels of gold encapsulated in pyrite while sample Z2L

shows higher levels of gold potentially associated with more reactive minerals such as pyrrhotite.

The results are not seen to be completely consistent with the gravity leach results discussed later.

Low levels of preg-robbing potential are indicated from the gold liberated in the burn off stage. It

should be noted that due to assay detection limits some of the lower deportments may be

marginally inaccurate. Given a detection limit of 0.01 g/t Au, measurements below this level were

assigned a nominal assay of 0.005 g/t Au; hence on the lower levels the deportment in these

fractions could be slightly overstated.

It was reported in the diagnostic leach tests during the hydrochloric acid digestion that a reasonably

vigorous reaction took place on the majority of the variability samples with the generation of green

91.9096.82 97.05

93.13

86.9289.50

85.34

50.00

60.00

70.00

80.00

90.00

100.00

110.00

Ref 1 Z1U Z1L Z2U Z2L Z3U Z3L

Pe

rce

nt

Sample ReferenceCyanide Soluble In Carbonates / Pyrrhotite

In Sphalerite and Labile Sulphides In Pyrite and Arsenopyrite

In Graphitic Carbon

June 2019


foam. This would tend to indicate a high level of carbonate and also acid soluble iron, possibly

pyrrhotite.

13.3.3 Crushability

Two separate tests were undertaken into the material strength and crushability by measuring the

uniaxial unconfined compressive strength (“UCS”) and the Bond crushing work index (“CWi”)

test, which indicates a material’s resistance to crushing. In the UCS test, a sample is prepared by

cutting to pre-set dimensions (re-coring) and this is then subject to a compressive load to measure

the strength at which the sample fails. The Bond CWi test, also known as the low impact energy

test, involves two swinging weighted pendulums which are allowed to fall and impact

simultaneously on the sample in order to measure from what height the pendulum needs to fall to

crush the sample. Both tests are undertaken on multiple individual samples; 3 prepared samples in

the case of the UCS tests and around 20 sample pieces for the Bond CWi test. The results of the

tests are presented in Table 13-6.

Table 13-6 Results of Crushability Tests: UCS and CWi

Sample average

density

Average

Depth UCS Result Mpa CWi (kwk/t) Depth

t/m3 RL m Average Max Min Average Std Dev m RL

Reference 2.67 910 59.5 73.7 41.8 9.8 1.6 910

Crushability 1 2.93 550 64.7 76.9 54.3 9.7 1.3 550

Crushability 2 2.87 753 53.9 94.4 31.1 11.1 1.2 753

Crushability 3 2.71 720 167.4 244.0 90.7 11.0 2.1 720

Crushability 4 2.84 699 82.4 90.0 68.9 12.3 2.9 699

The UCS test results are seen to be variable, with a relatively large variation between the maximum

and minimum measurements on the different samples which mainly appear to relate to the sample

tested rather than the depth of the material. Results were generally in the 30 to 95 MPa range,

indicating that the materials tested were medium strong to strong, although one sample

(Crushability 3) indicated to consist of quartzite (massive quartz vein), rather than schist identified

for the majority of the other samples tested, recorded a very strong measurement of around

240 MPa. The other sample of the same type of material measured 90 MPa, while a third sample

shattered during preparation and cutting and failed to produce the required test sample.

The CWi test results are in the easy to medium classification. Similar to the UCS results, the CWi

test results are also relatively variable with the reference sample (RL 910 m) generally indicating

results towards the lower end of those measures; however, no real correlation can be see between

the CWi results and relative level of the sample tested as shown in Figure 13-3.

June 2019


Figure 13-3 Variation of UCS and CWi result with depth (relative level)

13.3.4 Ball Milling Bond Work Index and Abrasion Index

For the 2003 FS into the treatment of the Wassa material by milling and CIL, testwork was

undertaken on representative samples of primary ore, oxides and spent HL material. The BWi for

the primary and oxide ores were reported to be in the region of 14.6 and 8 kWh/t, respectively.

More recent investigations have indicated that the BWi is generally noted to be increasing with

depth. Based on samples tested from three different drillholes from the Wassa starter pit area, SE

Area and MSN Area, BWi measurements, though somewhat inconsistent, appeared to indicate that

the BWi was increasing with depth.

From 2015, with fresh open pit ore feed, the unit power draw presented for the two ball mills is

shown to be between 14.5 and 16.5 kWh/t treated. This results in a calculated BWi of around 14 –

16 kWh/t, based on the reported mill feed and product sizes and power draw on the ball mills. An

allowance has been included in the calculations for mechanical and other losses between the drive

motor and mill. In recent years with the blend of underground and open pit the BWi continues to

remain within the 14 – 16 kWh/t range.

The findings of the BWi and Ai investigations from the 2015 tests are presented in Table 13-7 and

these are shown graphically as a function of the average depth of the sample in Figure 13-4 and

Figure 13-5, respectively.

The BWi tests were undertaken at a closing screen size of 106 µm to give a mill product of around

75-80% < 75 µm.

In summary, the findings of the latest testwork generally did not support the suspected increasing

BWi with further depth with the reference sample (910 m RL) showing the highest BWi reading.

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

0.0

50.0

100.0

150.0

200.0

250.0

300.0

500 550 600 650 700 750 800 850 900 950

WC

i kW

h/t

UC

S M

pa

Relative Level m

UCS Max UCS Min UCS Average CWi (kwk/t)

June 2019


Table 13-7 Results of 2015 BWi and Ai Tests

Sample Description BWi Ai Average

kWh/t RL m

Reference 15.7 0.394 910

Z1U Zone 1 Upper 15.3 0.33 763

Z1L Zone 1 Lower 14.7 0.276 664

Z2U Zone 2 Upper 14.9 0.228 713

Z2L Zone 2 Lower 14.5 0.175 575

Z3U Zone 3 Upper 14.4 0.229 585

Z3L Zone 3 Lower 13.9 0.152 533

Crushability 1 (347MET) 14 0.182 553

Crushability 2 (MET4) 15 0.205 753

Crushability 3 (MET5) 14.8 0.398 719

Crushability 4 (MET6) 14.8 0.326 700

Figure 13-4 2015 Ball Mill Bond Work Index against sample depth (relative level)

13.8

14

14.2

14.4

14.6

14.8

15

15.2

15.4

15.6

15.8

500 550 600 650 700 750 800 850 900 950

Bal

l Mill

BW

i kW

h/t

Relative Level metres

Bwi - Ref and Variability Bwi - Crushability

June 2019


Figure 13-5 2015 Abrasion Index against sample depth (relative level)

The abrasion index is a measure of the anticipated wear on components and consumables in the

comminution circuit and is applicable to both wear in the crushers and mills (grinding balls and

liners). The abrasion index is generally shown not to be increasing with depth and the trend appears

to be that the Ai measurement is falling slightly on the deeper samples. The Ai measured for the

reference sample is, with the exception of one sample (MET 5) representing the identified massive

quartz vein mineralization, higher than all the other variability and crushability samples tested.

This lower indicated abrasion index with depth may result in the reduced consumption of grinding

media and mill crusher liners as mining proceeds deeper into the underground mining areas. All

the samples fall into the slightly abrasive classification.

13.3.5 Gravity Gold and Leaching Tests

Gravity Tests

Gravity tests were undertaken by grinding a 1 kg sample to approximately 75% passing 75 µm and

then passing the sample through a Falcon centrifugal concentrator. The primary concentrate from

the Falcon was further processed on a Mozley shaking table, with the final concentrate weighed

and sent for assay. Tailings from the centrifugal concentrator and shaking table were subject to

cyanide leach tests.

The results of the gravity concentration tests are presented in Table 13-8.

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

500 550 600 650 700 750 800 850 900 950

Ab

rasi

on

Ind

ex A

i

Relative Level metres

Ai - Ref and Variability Ai - Crushability

June 2019


Table 13-8 Gravity Gold Recovery Test Results

Sample Ref Gravity Con Mass Assay

Metal Recovery to Gravity

Con

g Wt % Au (g/t) Ag (g/t) % Fe % S (total) Au % Ag %

Ref1 3.3 0.33 84.33 8.0 19.59 15.61 18.19 26.40

Z1U 2.1 0.21 322.6 18.8 37.28 21.86 10.41 9.18

Z1L 4.9 0.49 322.3 19.3 38.24 26.92 19.77 15.01

Z2U 2.5 0.25 324.3 26.3 37.05 31.09 15.87 18.26

Z2L 3 0.3 211.6 13.4 35.84 44.15 13.68 19.14

Z3U 2.7 0.27 199.2 14.8 34.31 38.32 13.21 8.88

Z3L 2.4 0.24 282.8 24.1 28.8 29.76 12.90 10.52

Gravity recoveries are seen to be lower than historically reported data. This is probably a function

of the laboratory tests which, for this stage of the investigation, were not optimized to maximize

gravity gold recovery. It can also be seen that the recovery from the reference sample is generally

higher than on the variability samples.

It was observed in all the gravity tests that the concentrates contained a magnetic component that

was readily picked up by a strong rare earth magnet, although not by a typical iron magnet. This

magnetic component was suspected to likely be pyrrhotite and this was reported by SGS to be

supported by the sulphur to iron ratios measured in the feed analyses.

Whole Ore Leach and CIL Evaluation Test

In order to investigate the effective leach parameters for the comparative leaching tests, a single

leach test was undertaken on the reference sample with and without carbon to confirm whether

any preg-robbing effect was evident. The results are presented in Table 13-9

Table 13-9 Whole Ore Leach and CIL test results

Solution (24h/48h) Solid tails Gold on Carbon Overall

Recovery

Back Calculated

Head Grade Au g/t Ag g/t Au g/t Ag g/t Au g/t Ag g/t Au % Au % Au g/t Ag g/t

Leach Test 1.13 0.08 0.105 0.05 1.55 0.15

Distribution % 93.23 67.03 6.77 32.97 93.23 67.03

CIL Test 0.14 0.01 0.1 0.05 93.4 12.7 1.21 0.19

Distribution % 14.27 6.49 8.29 26.41 77.44 67.09 91.71 73.58

The results generally indicated that no preg-robbing effect was evident with the recoveries without

carbon addition higher than those with carbon added to the leach (CIL test), although the gold

reconciliation was seen to be worse on the CIL test with a back-calculated gold head grade of

1.21 g/t Au compared to the screened analysis head grade and leach test back-calculated head

grade of 1.53 and 1.55 g/t Au respectively.

Gravity Tails Leach Test Results

Leach tests were undertaken on the combined gravity tails from the centrifugal concentrator and

concentrate cleaning table. From the gravity tests and one of the diagnostic tests there was potential

that pyrrhotite could be present so the gravity tails samples were adjusted to pH 10.5 - 11 using

June 2019


lime and aerated until the pH and dissolved oxygen levels stabilized generally in line with the plant

practice of injecting oxygen into the transfer lime from milling to CIL. Pyrrhotite is highly reactive

and can result in high consumptions of oxygen and cyanide in leach if not preconditioned.

Leach tests were conducted for 48 hours with samples taken at 2, 4, 6, 24 and 48h and analysed

for gold and silver in solution. An initial cyanide level of 1 g/l was used and cyanide levels in

solution were maintained at >0.5 g/l by dosing of additional cyanide as required. The tails solids

were analysed for silver and gold. No lead nitrate was added in the leach tests.

Leach test results of the gravity tails are presented in Table 13-10.

Table 13-10 Leach test results and reagent consumptions

Sample

Reference

Gold Recovery % Assayed Tails

g/t Au

Consumption kg/t

24h 48h NaCN 24h NaCN 48h Lime as CaO

Ref1 77.22 88.69 0.09 0.43 1.31 0.88

Z1U 90.69 87.35 0.44 0.51 1.48 0.89

Z1L 86.72 87.64 0.68 0.40 1.15 0.75

Z2U 92.81 93.80 0.20 0.43 1.05 0.92

Z2L 87.81 88.06 0.42 0.15 0.91 0.88

Z3U 92.95 91.33 0.23 0.63 0.89 1.16

Z3L 94.57 93.25 0.18 0.63 1.01 1.11

It can be seen that in some tests, recoveries based on 48h leach solution analyses were lower than

for the those based on the 24h leach solution assays. This could be caused by analytical

discrepancies or errors based on solutions analysed or possibly some adsorption of dissolved gold

onto the fine milled solids. As no appreciable preg-robbing potential or effect was indicated in the

diagnostic leach and comparative leach and CIL tests, this is not considered to be a major concern

as any weakly adsorbed gold would be recovered on the plant due to the presence of activated

carbon in the leach circuit.

The leach curves on the gravity tails appear to be relatively consistent with the exception of that

for the reference samples which shows slower kinetic especially at 24h, although results in similar

overall recoveries at 48h.

June 2019


Figure 13-6 Leach recovery kinetic curves

Overall Gravity / Leach Recoveries

The overall recoveries from the gravity / leach testwork are presented in Table 13-11. These are

based on the maximum leach recovery at either 24 or 48h and on the back-calculated head grade

from the recovered gold and tailings assays.

Table 13-11 Overall gravity leach recoveries

Sample Reference Gold Recovery %

Gravity Leach Overall

Ref1 26.41 88.69 91.68

Z1U 16.38 90.69 92.22

Z1L 22.69 87.64 90.44

Z2U 20.19 93.80 95.05

Z2L 15.37 88.06 89.90

Z3U 16.91 92.95 94.15

Z3L 20.41 94.57 95.68

In the gravity / leach tests, poor reconciliations were achieved between the back-calculated head

grade and the assay head grades from the screened analyses on the master samples with the back-

calculated head grades consistently being considerable lower than the head assay results by as

much as 35% in two tests.

The comparison of the assay head compared to the back-calculated head grade for both the gravity

leach and diagnostic leach results are presented in Table 13-12.

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

100.00

0 5 10 15 20 25 30 35 40 45 50

Comparative Leach Curves

Ref1 Z1U Z1L Z2U Z2L Z3U Z3L

June 2019


Table 13-12 Reconciliation of assay and back-calculated head grades from testwork

Sample

Assay Head From Diagnostics From Gravity / Leach

Grade Grade Grade

g/t Au g/t

Ag g/t Au g/t Ag g/t Au g/t Ag

Reference 1.53 0.10 1.35 0.22 1.08 0.13

Z1U 6.51 0.43 6.74 0.89 4.18 0.31

Z1L 7.99 0.63 8.71 1.06 7.08 0.46

Z2U 5.11 0.36 5.10 0.62 4.03 0.38

Z2L 4.64 0.21 4.84 0.46 4.14 0.32

Z3U 4.07 0.45 4.12 0.33 3.20 0.30

Z3L 5.26 0.55 4.28 0.27 3.36 0.29

The correlations were better in the diagnostic leach tests compared to the gravity / leach tests with

both positive and negative discrepancies. Difference varied between -10% and +18% resulting in

an overall difference of only -2%.

13.3.6 Settling Tests

Comparative settling tests were undertaken on the reference sample and one selected variability

sample (Z1L). Initial scoping tests were undertaken using five different flocculants with the

settling tests undertaken using Nasaco anion flocculants N2132 and N2326. The results show very

similar settling performance on the reference samples and one variability sample selected.

The settling test results are presented in Table 13-13.

Table 13-13 Comparative settling test results

Sample Feed

Solids pH Flocculant

Flocculant

Dosage

Initial

Settling

Rate

Final

Solids

Content

Thickener

Underflow Unit

Area

% g/t m3/m2/day % m2/t/d

Reference

Test 1 9.43 10.5 N2132 50.04 1335.26 59.0 0.235

Reference

Test 2 10.08 10.5 N2326 46.62 2897.86 61.8 0.261

Z1L Test 1 9.04 10.5 N2132 52.21 2414.88 56.5 0.225

Z1L Test 2 9.13 10.6 N2326 51.69 2637.79 56.9 0.223

June 2019


14 Mineral Resources

14.1 Introduction

The Mineral Resource Statement presented herein represents an estimate for the Wassa Main

deposit and the satellite deposits Chichiwelli, Benso and Hwini-Butre. The Mineral Resource

Statement is presented in accordance with the guidelines of NI 43-101.

The GSR exploration team was responsible for a portion of the Wassa resource modelling exercise

which included all topographic surfaces, weathering surfaces and structural control lines. SRK

(UK) was commissioned for the modelling of the lithological and grade wireframes and SRK

(Canada) estimated gold grades for the Wassa Main deposit. The HBB resource modeling was

done by GSR geologists. The mineral resource classification and statement was conducted by GSR

under the supervision of S. Mitchel Wasel, a QP.

This section describes the Mineral Resource estimation methodology and summarizes the key

assumptions considered for the estimate. The Mineral Resource estimate reported herein is a

reasonable representation of the global gold Mineral Resource found at the Wassa Main and HBB

deposits given the current level of sampling. The Mineral Resources have been estimated in

conformity with generally accepted CIM “Estimation of Mineral Resource and Mineral Reserves

Best Practices” guidelines and are reported in accordance with NI 43-101. Mineral Resources are

not Mineral Reserves and do not have demonstrated economic viability. There is no certainty that

all or any part of the Mineral Resource will be converted into Mineral Reserve.

The databases used to estimate the Mineral Resources were audited internally by GSR. In the

opinion of the GSR QP, S. Mitchel Wasel, the current drilling information is sufficiently reliable

to interpret with confidence the boundaries for gold mineralization and that the assay data are

sufficiently reliable to support mineral resource estimation.

14.2 Resource Estimation Procedures

The resource evaluation methodology involved a database compilation and internal validation

exercise by GSR. At Wassa, GSR was responsible for structural control lines, topographic and

weathering surfaces, external consultants created lithological and grade wireframes with input

from GSR geologists. GSR provided SRK with borehole databases, structural control lines,

topographic surfaces and weathering surfaces. At HBB, GSR was responsible for the grade shell

interpretations, topographic and weathering surfaces and for the estimation of gold grades.

Prior to initiating the modelling and resource estimation process, SRK reviewed the databases for

the Wassa project.

After evaluating the available database, SRK proceeded with the grade wireframe modelling

(Wassa), the data conditioning (compositing and capping) for geostatistical analysis and

variography. At Wassa, the grade wireframe modelling was completed in Leapfrog Geo 4.4 under

following the guidelines that GSR and SRK have established together. The grade interpolation

methodology was discussed between GSR and SRK, it was decided to use Ordinary Kriging

(“OK”) with local varying angles and local variograms for the estimation of gold grades based on

the structural complexity and folded nature of the deposit.

June 2019


At HBB, after evaluating the available database, GSR proceeded with the wireframe modelling,

the data conditioning (compositing and capping) for geostatistical analysis and variography. The

vein modeling tool in Leapfrog Geo 4.4 was used for modelling HBB. The grade interpolation

methodology used is OK and it was performed in Leapfrog Geo with Edge 2.2.

The classification and preparation of the Mineral Resource Statement were conducted by GSR

under the supervision of GSR’s QP, Mr. S. Mitchel Wasel.

14.3 Resource Database

14.3.1 Wassa

The Wassa database is made up of four individual drillhole databases, namely: the GSR Wassa

exploration database, which contains exploration drilling conducted by GSR since 2002; the AW

exploration database, which contains historical exploration drill holes from SGL; the Satellite

grade control database; and the GSR grade control database. The Satellite grade control database

was not included in the Mineral Resource estimate as the blast holes samples are considered not

to be of a sufficient quality for inclusion into the Mineral Resource estimate.

A cut-off was applied to the GSR exploration database with only drillholes with a complete assay

table retained for the grade model and the subsequent Mineral Resource estimate.

Table 14-1 Wassa drill hole database as of December 2018

Location Type Number of Holes Meterage (m)

Wassa

RC 1,463 139,292

DD 892 276,852

GC (RC) 25,561 660,058

Wassa UG

DD 971 110,541

GC(Chan-Chips) 1,717 9,174

The borehole databases contain information including collar information, downhole deviation

surveys, gold assays, lithological descriptions, alteration, structural data, major structures and vein

descriptions.

GSR and SRK have performed validation routines to the resource database. Based on this

assessment, and the checks described in Section 12, it is the opinion of the QPs that the borehole

database is appropriate to form the basis of the Mineral Resource estimate.

14.3.2 Hwini-Butre

The Hwini-Butre database is made of DD holes and RC drilling data.

A cut-off was applied to the GSR exploration database with only drillholes with a complete assay

table retained for the grade model and the subsequent Mineral Resource estimate.

June 2019


Table 14-2 Hwini-Butre drill hole database as of December 2018


Hwini-Butre DD 518 73,223.44

RC 3165 75,384

The borehole databases contain information including collar information, downhole deviation

surveys, gold assays, lithological descriptions, alteration, structural data, major structures and vein

descriptions.

GSR has performed validation routines to the resource database. Based on this assessment, and the

checks described in Section 12, it is the opinion of the QPs that the borehole database is appropriate

to form the basis of the Mineral Resource estimate.

14.3.3 Benso

SRK was provided with a Gemcom project directory containing the SJR and GSR drilling data as

audited by GSR and the geological models subsequently produced by GSR including geological

wireframes, oxidation and topographic surfaces and Block Model parameters. Additional

information was provided as Excel spreadsheets documenting QA/QC data and results of density

determinations.

Table 14-3 Benso drill hole database as of December 2012


Benso

RC 465 33275.7

DD 321 37,622.50

Geotech 14 1,637.30

GC (RC) 2,362 57,970.00

14.3.4 Chichiwelli

SRK was provided with a Gemcom project directory containing the drilling data as audited by

GSR and the geological models subsequently produced by GSR including geological wireframes,

oxidation and topographic surfaces and Block Model parameters. Additional information was

provided as Excel spreadsheets documenting QA/QC data and results of density determinations.

Table 14-4 Chichiwelli drill hole database as of 2012


Chichiwelli

RC 483 29,802.20

DD 23 3,692.00

Geotech - -

GC (RC) - -

June 2019


As no mining has taken place at Chichiwelli, the topographic survey used for the 2010 Mineral

Resource statement remains valid. The Chichiwelli MRE also includes three small deposits located

in the Manso licence area. These deposits are Abada, Adoikrom South and C3PR. The techniques

used to estimate these deposits are consistent with those reported for Chichiwelli.

14.4 Grade Shell Modelling

14.4.1 Wassa

The grade shell model was generated using Leapfrog indicator interpolants, guided by an

anisotropic structural trend.

14.4.2 Wassa Mineralization Wireframe

The wireframe modelling was carried out by SRK using Leapfrog Geo 4.4 software. Mineralized

wireframes at Wassa are modelled using an indicator approach which uses a 0.4 g/t cut-off for the

LG envelopes and a 1.5 g/t cut off for HG. Visual inspection of assay data suggests that these

respective lower cut-off levels are reasonable to separate barren from auriferous sections

intersected by each borehole. Mineralized shells are created using this indicator approach

combined with structural trend surfaces created by the site geologists and reviewed by SRK.

14.4.3 Wassa Indicator Interpolants – Background

An indicator interpolant works in a similar way to a grade shell, but rather than interpolating the

raw grade, all data above the given indicator grade value is assigned a value of 1 and all data below

the indicator grade value are assigned a value of 0. A shell is then generated at a defined iso-value,

between 0 and 1. This helps to remove the impact of very HGs which can result in “blow-outs” or

unrealistic volumes that can result from standard grade shell modelling of highly skewed data

populations.

The indicator interpolant is influenced by an anisotropic structural trend, which is based on form

surfaces. The form surfaces represent vectors of grade continuity, where grade continuity is high

along the modelled form, and low across it. Due to the significantly deformed nature of the gold

mineralization, this type of 3-dimensional structural trend is vital to produce a geologically

realistic shape of the indicator interpolants.

14.4.4 Wassa Structural Trend

The structural geological influence on HG mineralization trends at Wassa has been studied by SRK

following an underground mapping exercise in late 2016 and applying observations from this, open

pit grade control data and the available downhole structural data to the wider drilling data set.

SRK, in conjunction with GSR, constructed a series of form surfaces to guide the structural trend

applied in the indicator interpolation, to reflect the structural geometry of the Au mineralization at

Wassa. These form surfaces represent the broad F4 folding event, a plunging synclinal feature

which affects grade distribution at the mine scale, with some subtly different internal orientations

attributed to the largest features associated with an earlier high strain folding event (F3).

June 2019


Figure 14-1 Example of Structural ‘Form’ Surfaces

(Oblique view dipping at 40° towards the SW of the trend surfaces shown relative to the drilling

results.)

A total of 102 structural ‘form’ surfaces have been used for the modelling of the 3 Wassa models.

A structural trend was designed using the form surfaces, described here above, as Trend Inputs.

The following parameters were used to define the Structural Trend for each model:

Trend Type Compatibility Trend Inputs Strength

Global

Mean

Trend

Strongest

Along Inputs Version 2

All 52 modelled

structural ‘form’

surfaces

7.0 to 10.0 N/A

An example of the resulting Structural Trend is presented in Figure 14-2.

June 2019


Figure 14-2 North-facing cross sections showing structural form (18950N and 19170N)

14.4.5 Wassa Indicator Interpolants

Prior to generating the indicator interpolant shells, the raw assay file was composited to 6 m, with

a minimum end composite length of 3 m. Indicator interpolants were defined at 0.4 g/t Au and 1.5

g/t Au cut-offs. The HG 1.5 g/t Au cut-off value was selected on the basis of a statistical and visual

evaluation of the grade distribution. The LG 0.4 g/t halo has been used historically at the operation

and is considered appropriate by site staff.

The indicator interpolants were restricted to be within a bounding box defined by the coordinates

provided in Table 14-5.

Table 14-5 Modelling extents

Axis Minimum extent Maximum extent

X 39000 E 41000 E

Y 18600 N 20600 N

Z -1000 Z 1200 Z

Table 14-3 summarizes the parameters that were applied to both the 1.5 g/t Au and 0.4 g/t Au

models:

June 2019


Table 14-6 Modelling parameters

Interpolant

Type Range Nugget

Iso-

Value Resolution

Volumes

Excluded

Spheroidal 100 50% 0.35 5m <5000m³

In order to better reflect the geometry of the Au mineralization at a local scale, and also to improve

continuity in areas of wider spaced drilling, SRK edited the indicator interpolant shells using

indicator polylines. Indicator polylines are digitised and editable strings that carry an associated

numeric value which is added to the assay data points on which the interpolant is based. In this

instance, indicator polylines with values of 1 (inside), 0 (outside) and indicator iso-value were

added to the interpolant. The “iso-value” indicator polylines allow the specific position of the outer

limit of the shell to be locally edited. This helped to influence continuity orientations at a smaller

scale, ensuring F3 continuity and geometry could be reflected in the resultant domain wireframes

in well drilled areas, and assisted in improving the continuity of the model in some of the more

sparely drilled areas.

14.4.6 Wassa Vein Modelling

Because of the wider spaced drilling in the deepest and most southerly portions of the Southern

long-range model, some of the thinner HG (>1.5 g/t) intersections, which could reasonably be

connected, are encapsulated by isolated shells with poor continuity. This is a function of the nature

of indicator interpolant modelling, which works best in densely drilled deposits and is not

particularly effective at connecting thin drillhole intersections over large distances.

For this reason, in some of the more sparsely drilled areas, the >1.5 g/t indicator interpolant shells

were replaced with Leapfrog “veins” where it was considered reasonable to connect some of the

thinner HG (>1.5 g/t) intersections. It should be stressed that, aside from these instances,

considering the distribution and style of mineralization and especially given the ability to manually

edit the shells, iso-surfaced indicator interpolants are still considered the most appropriate

modelling methodology for the >0.4 g/t mineralization and for the >1.5 g/t mineralization in areas

of thicker HG zones.

14.4.7 Wassa Final Model

The final model, displayed in Figure 14-3, constitutes the following:

• >0.4 g/t Au – the iso-surfaced indicator interpolant;

• >1.5 g/t Au – A combination of the iso-surfaced indicator interpolant and “vein” models,

whereby isolated discontinuous shells have been replaced with volumes modelled using

the Leapfrog vein modelling tool.

June 2019


Figure 14-3 Oblique view of final volumes looking east

(green = >0.4 g/t, red = >1.5 g/t south of 19550N)

Finally, the mineralized wireframes cover the region bounded by 39400 and 41000 easting, and

18600 and 21000 northing. Two resource wireframes were constructed by SRK (UK) with

guidance from Golden Star. These comprise a LG shell and HG shell, corresponding to a 0.4 g/t

gold and a 1.5 g/t gold threshold, respectively.

14.4.8 Hwini-Butre

Geology and mineralization domaining was undertaken by GSR.

The mineralization zones of Hwini-Butre are structurally controlled, with gold emplacement

related to the density of quartz veining and sulphide content.

The Mpohor complex exhibits the underlying north-south trends but also has extensive cross-

cutting features present, particularly in the north-west orientation. The Adoikrom and Father

Brown deposits occur in the south of the Mpohor complex and appear to be controlled by a series

June 2019


of shallow to moderately dipping shear structures with dips varying from 20° to the south,

steepening to 65° to the west.

Two estimation domains have been modelled at Hwini-Butre, as follows:

• Adoikrom; and

• Father Brown.

The modeling was done using the vein modeling tool in Leapfrog Geo 4.4 with a modeling COG

of 2.5 g/t.

Only DD and RC drilling has been used for the subsequent grade estimation. The resource

wireframes and drillholes are shown in Figure 14-4.

Figure 14-4 Mineral Resource wireframes and drill hole locations for the HwiniButre

14.4.9 Benso

Geology and mineralization domaining was undertaken by GSR.

June 2019


The mineralization zones of Benso are structurally controlled with gold emplacement related to

the density of quartz veining and sulphide content. The mineralization hosting structures generally

dip steeply to the west with foliation generally parallel to the bedding. The Subriso East deposit is

interpreted to dip less steeply to the west at approximately 50°. Oxidation associated with

weathering is variable but generally limited. The weathering forms a layer of lateritic clay rich

material grading into a soft saprolite. The vertical depth is generally 10 m or less but can reach

depths of 30 m in places.

Four estimation domains subdivided by oxidation state have been modelled for Benso, as follows:

• Subriso East (“SE”);

• Subriso West (“SW”);

• G-Zone; and

• I Zone.

The SE domain is physically separated from the others and strikes to the north with a dip to the

west of between 55-60°. The SW, G Zone and I Zone domains occur in sub-parallel structures and

strike to the north-west (320°) with a steep dip of 75-80° to the south-west. Because of this, it was

decided to treat the SE orebody as a separate for the purpose of grade interpolation.

Only DD and RC drilling has been used for the subsequent grade estimation. The resource

wireframes and drillholes are shown in Figure 14-5.

June 2019


Figure 14-5 Mineral Resource wireframes and drill hole locations for the Benso deposits

14.4.10 Chichiwelli

The mineralization zones of Chichiwelli are structurally controlled with gold emplacement related

to the density of quartz veining and sulphide content. The mineralization hosting structures

generally trend north-south and dip moderate-steeply to the east at 60°.

Two estimation domains have been modelled for Chichiwelli as follows:

• East Domain; and

• West Domain.

The East and West domains comprise some 10 individually separated wireframe solids which have

not been subdivided by oxidation.

Wireframes are based on a roughly 0.5 g/t Au grade value. In places composite grades fall below

this threshold value but have been included for the sake of maintaining continuity of the orebody

model. The style of mineralization seen at Chichiwelli is analogous to deposits observed elsewhere

in the Wassa region and, typically for shear zone hosted gold deposits, the mineralization grades

tend to pinch and swell within the defined mineralized bearing structures. The resource wireframes

and drillholes are shown in Figure 14-6.

June 2019


Figure 14-6 Mineral Resource wireframes and drillhole locations for Chichiwelli

14.5 Statistical Analysis and Variography

14.5.1 Wassa

Table 14-7 summarizes the gold statistics of the assays tagged by mineralized domains for both

long-range models.

Table 14-7 Summary Gold Statistics of Assays and Composites

For consistency with the April 2018 model, SRK chose to composite at 3.0-metre lengths within

the solid wireframes. SRK assessed the statistical sensitivities due to keeping all residual length

June 2019


composites or removing those composites smaller than 0.3 or 1.5 metre lengths. Results showed

no material relationship between the composite length and the grade. Further, the statistical

difference in the mean grade, standard deviation and the coefficient of variation was less than 2

percent when comparing all three scenarios. As with the composite length, SRK chose to maintain

consistency with the previous resource model and removed all composites smaller than 0.30

metres. Summary statistics for composite gold grades are also provided in Table 14-8.

In collaboration with Golden Star, SRK selected the capping value by comparing probability plots

of gold composites on a by-domain basis and plotting the mean grade and the number of affected

data by the chosen cap value (see Table 14-7). SRK chose to cap HG composites at 30 g/t gold

and LG composites at 15 g/t gold. These capping thresholds are the same as the April 2018 mineral

resource model. Table 14-8 compares the statistics for uncapped and capped composite gold

grades.

Table 14-8 Comparison of Uncapped and Capped Gold Composite Grades

June 2019


Figure 14-7 Probability Plot and Capping Sensitivity Plot

14.5.2 Wassa Local Variogram Models Long Range

The local estimation approach chosen for the Wassa Gold Mine required the specification of local

variogram models. SRK assessed and modelled local variograms for the HG and LG domains,

centered about each anchor point. Anchor point locations were reviewed by Golden Star prior to

finalization of their locations. Table 14-10 summarizes the anchor point locations and their local

orientations for variogram calculation and modelling. The modelled local variograms for these

anchor points are tabulated in Table 14-11. For the LG domain, SRK relied on variograms based

on the combined LG and HG capped composites due to the challenges of inferring reliable

variograms based solely on LG composites. One reason for the inference challenge may be related

to the spatial voids in the database where the HG domain resides. For anchor points 6, 8 and 13,

SRK used the HG domain variograms for the LG domain and adjusted the ranges wherever

possible to reflect the combined domain variograms.

For each domain (LG and HG), the local variogram parameters (Table 14-11) were then estimated

to the block model grid to be read into the grade estimation. In general, the local variograms should

be smoothly transitioning within the series. Abrupt changes in grade continuity, within a zone and

June 2019


between anchor point locations, were not expected. Highly localized changes were addressed by

the selection of anchor point locations. To ensure smoothness of the local variograms parameters

and consistency with the 2015 model, SRK used global kriging with a continuous spherical

variogram with ranges of 1,000 by 750 by 500 metres.

Table 14-9 Local variogram orientations and anchor point (AP) locations

June 2019


Table 14-10 Local variogram models by domain

14.5.3 Hwini-Butre

The statistics are based on composited assay values within the modelled wireframes; the data was

composited to full lengths within the mineralized zones. The statistics presented here are based on

drilling data that intersect the wireframes only.

The descriptive statistics for the individual modelled domains, are summarized in Table 14-11. For

all datasets, zero values were checked in the database, and were set to 0.005 g/t.

Table 14-11 Descriptive statistics for Hwini-Butre modelled domains (uncapped & capped)

Domain Oxidisation Count Minimum Maximum Mean Variance COV

Adoikrom Uncapped 953 0.005 106.71 6.01 67.39 1.37

Capped 953 0.005 25.00 5.06 23.02 0.95

Father Brown Zone

Uncapped 1287 0.005 154.40 8.90 180.70 1.5

Capped 1287 0.005 40.00 7.47 87.25 1.25

June 2019


HG capping was applied to the Hwini-Butre exploration drilling dataset, where extreme grades

occurred randomly rather than being a HG feature. The high caps were determined on the basis of

the shape of the tail of the log histogram and the log probability plots. Capping reduces the extreme

values to a nominated capped value, which affects the mean grades of the full-length composites.

A cap of 25 g/t has been applied to the Adoikrom dataset and a cap of 40 g/t has been applied to

the Father Brown Zone dataset.

Figure 14-8 Adoikrom Log Probability

June 2019


Figure 14-9 Adoikrom histogram

June 2019


Figure 14-10 Father Brown Log Probability

June 2019


Figure 14-11 Father Brown Histogram

Variography was undertaken on each of the modelled areas using only the composite data from

within each of the modelled domains. The variogram was rotated along strike and down dip of the

modeled domains. Radial variogram plots were used to determine the plunge. The total sill was set

to the variance of the composite population, the nugget was set to approximately 30% for

Adoikrom and 14% for Father Brown Zone of the total sill. 2 spherical models were used to assign

the range in the maximum (down dip), and intermediate (along strike) directions. Variography in

the minimum direction (thickness) was considered but has no influence due to the thinness of the

modeled domains relative to lengths along strike and dip. Variogram parameters derived from the

modelled variograms are summarized below in Table 14-12.

June 2019


Table 14-12 Variogram parameters for Hwini-Butre

14.5.4 Benso

The statistics are based on composited assay values within the wireframes modelled by GSR; the

data was composited to 2 m lengths within the mineralized zones, and composites of less than

1.50 m were removed.

The descriptive statistics for the individual modelled domains, split by oxidation state, are

summarized below in Table 14-13. The transition zone is relatively thin, and so has not been

analysed separately. For all datasets, zero values were checked in the database, and were set to

0.001 g/t.

Table 14-13 Descriptive statistics for Benso modelled domains (capped)

Domain Oxidisation Count Minimum Maximum Mean Variance COV

Subriso East Oxide 266 0.001 30.81 2.11 15.18 1.85

Fresh 649 0.001 51.58 2.54 25.49 1.99

Total 915 0.001 51.58 2.42 22.51 1.96

Subriso West Oxide 36 0.41 15.86 3.14 14.11 1.20

Fresh 571 0.001 223.83 3.88 147.39 3.13

Total 607 0.001 223.83 3.83 139.48 3.08

G Zone Oxide 44 0.001 21.15 2.76 18.69 1.57

Fresh 570 0.001 52.33 2.04 11.10 1.63

Total 614 0.001 52.33 2.09 11.64 1.63

I Zone Oxide 11 0.21 1.51 0.97 0.23 0.49

Fresh 86 0.11 18.18 2.72 10.96 1.22

Total 97 0.11 18.18 2.52 10.04 1.26

The four areas were combined into two areas for estimation purposes; namely Subriso East, and

Subriso West, G Zone and I Zone combined. The Subriso East domain is separated from the

Subriso West, G Zone and I Zone areas, and strikes roughly north-south, with a dip to the west of

between 55 and 60°. The Subriso West, G Zone and I Zone areas lie in sub-parallel structures,

striking roughly to the north-west (320°), with a steep dip of 75 to 80° towards the south-west. The

descriptive statistics for the two separate estimation domains are shown below in Table 14-14.

June 2019


Table 14-14 Descriptive statistics for simplified Hwini-Butre modelled domains (capped)

Domain Count Minimum Maximum Mean Variance COV

Subriso East

915 0.001 51.58 2.42 22.51 1.96

Subriso West, G Zone

and I Zone 1318 0.001 223.83 2.93 71.05 2.88

The statistical distributions for the two domains are relatively similar, with the histograms

indicating that the distribution is not normal, being highly negatively skewed. The log transformed

gold grade data demonstrates that there may be several populations within the distribution and that

the distribution approached log-normality.

HG caps were applied to the composite data as follows:

• Subriso East – 40g/t cap.

• Subriso West, G Zone, I Zone – 60g/t cap.

The estimation data sets noted above were used to derive variograms for estimation. In all cases,

the grade block model for each individual modelled solid was estimated using only the composites

inside that solid.

Variography was undertaken on the log transformed data, with a short lag, omnidirectional,

downhole variogram used to derive the nugget effect. Directional variograms were then calculated

within a rotated plane aligned with the strike and dip of the modelled solids. The variogram

parameters derived from the modelled variograms are shown below in Table 14-15. The

variograms were back transformed before being used in OK.

Table 14-15 Variogram parameters for the Benso zones

Parameter Subriso East Subriso West, G Zone and I

Zone

Co 0.25 0.19

C1 0.41 0.54

C2 0.34 0.27

a1 (strike) 20 20

a1 (dip) 15 8

a2 (strike) 50 50

a2 (dip) 40 30

14.5.5 Chichiwelli

The statistics are based on composited assay values domained within the mineralization

wireframes described previously, with sample data composited to 2 m lengths within the

mineralized zones.

The statistics presented here are based on all drilling data that intersect the wireframes. The

composites inside the modelled bodies were also split into oxidization states, but as there was little

June 2019


information for the transition zone, SRK combined the three oxidization states and used the

combined oxidations datasets throughout the statistical and geostatistical studies, and the

subsequent grade estimation.

The descriptive statistics for the two separate estimation domains are shown below in Table 14-16.

Table 14-16 Descriptive statistics for Chichiwelli modelled domains (capped)

Domain Count Minimum Maximum Mean Variance COV

East 418 0.001 41.10 1.75 17.64 2.41

West 559 0.001 46.30 1.69 10.14 1.89

HG capping was applied to both the East and West domains. The HG caps were determined on the

basis of the shape of the tail of the log histogram and the log probability plots. Capping reduces

the extreme values to a nominated capped value, which affects the mean grades of the two m

composites, as indicated by Table 14-17.

Table 14-17 Chichiwelli high grade capping

Domain Cap

Applied (g/t)

Mean Grade before Cap

(g/t)

Mean Grade after Cap

(g/t)

Percentage Difference

(%)

East 25 1.75 1.65 -6.06

West 15 1.69 1.59 -6.29

The estimation data sets noted above were used to derive variograms for estimation. In all cases,

the grade block model for each individual modelled solid was estimated using only the composites

inside that solid.

Variograms were modelled for the East and West domains separately. Variography was attempted

for the individual solids, but the resultant variograms were unable to be modelled. Raw

variography resulted in difficult to model variograms, and so a Gaussian transformation was

applied to the data. The first stage was to define the nugget effect from a short-lag omnidirectional

variogram, which is calculated along the drillhole, and then to model the variogram ranges from

directional variograms from along strike, down-dip and across dip directions. The directional

variograms are then back transformed into “raw” space and used for subsequent estimation. The

back transformed variograms and resultant variogram parameters are included in Table 14-18.

June 2019


Table 14-18 Variogram parameters for Chichiwelli zones

Parameter East West

Co 7.94 3.06

C1 3.60 1.59

Nugget Effect (%) 68.8 65.81

Range (m)

a1 (strike) 25 40

a1 (dip) 25 35

a1 (normal to strike) 8 4.7

14.6 Block Model and Grade Estimation

14.6.1 Wassa Block Model

A 3D block model including rock type, gold, percent mineralization, density and class was

constructed for each of the modeling areas, Wassa Short Range, Northern Long Range and

Southern Long Range. The selection of the block size was driven by the borehole spacing and

mainly by the geometry of the auriferous zones, but also based on mining parameters and in

accordance to previous resource estimate. The Long Range models block size was set at 10 x 10 x

3 m in the northing, easting and elevation directions, respectively along the mine grid. The block

model origins can be seen in Table 14-19.

Table 14-19 Block model parameters

Coordinate Origin Block Size (m) No. of Blocks

X 39,100 10.0 225

Y 17,700 10.0 360

Z 1,100 3.0 580

A percent block model was used to evaluate tonnages. Tonnage for each respective block was

obtained by weighting volumes corresponding to the interpreted auriferous zones and the

respective mean SG defined by weathering profile

The block model bulk density data was coded based on weathering surface which was built to

define oxide material from fresh material. The weathering surface defined the ‘top of fresh’

material; all blocks above the ‘top of fresh’ surface were designated as oxide and material below

the surface as ‘fresh’. The bulk density values assigned to the block model were based on series

of measurements made over the various exploration phases going back to the initial Golden Star

exploration program in 2002. The density values used for the tonnage estimate were provided by

GSR and are detailed below in Table 14-20.

June 2019


Table 14-20 Bulk density

Weathering Type Assigned Bulk Density Value (t/m3)

Oxidised 1.8

Fresh 2.8

14.6.2 Wassa Resource Estimation Methodology

SRK implemented the same methodology as for the April 2018 mineral resource model, using OK

with local varying angles and local variograms for the estimation of gold grades. The general steps

required to implement the approach are:

• Construct locally varying angles models for dip and dip direction.

• Calculate and model local variograms for each series and interpolate these local

variograms to construct a model of local variogram model parameters.

• Estimate gold grades using OK, calling upon the local models of dip, dip direction, and

variogram models.

• Check estimated model using qualitative and quantitative methods.

Table 14-21 summarizes the block model definition used for the model area using the mine grid.

No rotation was applied. Notably, the block model covers an area that is well beyond the extents

of the mineral resource domain wireframes.

The following sections summarize the method(s) used, assumptions made, and results obtained for

each of the four modelling steps.

Table 14-21 Block model definition using GEMS convention

Block Size

(metre)

Origin*

(metre)

Block

Count

X 10 39,100 225

Y 10 17,700 360

Z 3 1,100 580

* Coordinates relative to mine grid.

14.6.3 Wassa Local Angle Model

Local angles were derived from triangulated facets of the structural trend surfaces provided by

Golden Star and SRK (UK). This was achieved using CAE’s Datamine Studio 3, and an initial

angle data set for both dip and dip directions. As before, the structural trend surfaces were

generated using Leapfrog, and the mesh resolution provided a smooth variation of the dip and dip

direction angles.

The angle data set was then used to interpolate a block model of dip and dip directions, which was

later called upon for local estimation. The estimation of angles used inverse distance estimation

with a power of three, using an isotropic range of 500 metres with up to six conditioning angle

data.

June 2019


14.6.4 Wassa Grade Estimation Northern Long-Range Model

Golden Star advised SRK that reconciliation between the long and short-term models was poor,

with the former model showing consistently lower grades. Unlike the mineral resource models of

2017 and 2018, a hard boundary approach was used throughout the model, including in the planned

open pit mining areas. Specifically, only LG composites were used to inform the LG blocks, and

only HG composites were used to inform the HG blocks. Use of a hard boundary may mitigate

this impact of diluting grades across domain boundaries in the open pit portion of the model.

Using the models of local angles and local variograms, SRK performed the grade estimation using

a local OK methodology. The LG and HG estimation was based on the parameters summarized in

Table 14-24. Other than a change to the database and associated updates to the domain wireframes,

the composite length, block size and global variogram range remains largely unchanged from

previous models. For this reason, SRK chose to use estimation parameters that are consistent with

those used in 2017 and 2018. The first estimation run required a relatively high minimum number

of samples with the aim of obtaining a highly localized estimate. The second pass also required at

least two holes but was slightly more relaxed in the minimum number of samples required. The

third estimation pass reduced the minimum number of boreholes required but maintained a search

that is one-times the global variogram range. The fourth estimation run considered search ellipses

sized at least twice the variogram ranges, with the aim of estimating most of the blocks unvisited

by the first three passes. As the estimation considered a stationary search ellipsoid, these ranges

were selected to ensure that local estimation yielded estimates that conformed to the local

anisotropy.

Table 14-22 Northern model estimation parameters

14.6.5 Wassa Southern Long-Range Model

In April 2018, SRK noted that the north and south areas of the Wassa Gold model are distinct in

continuity of the resource domains and the availability of drill data for estimation. The northern

area is characterized by discontinuous LG and HG domains with dense drilling on 25-metre

sections with supplementary grade control data as close as 7.5-metre spacing. The southern area

was drilled nominally on 50-metre sections, with the region south of 19550N informed by

boreholes spaced at 100 to 200 metres apart. Since 2018, Golden Star has drilled more boreholes

in the southern extension, but there remains significantly wider spaced drilling relative to the

northern part of the deposit. As noted before, the wider spaced drilling in the south impacts the LG

and HG domains, allowing for greater continuity in both domains than was modelled in the

northern portion. Given the more continuous resource domains with less informing data, SRK

chose to impose additional controls on the HG composite, in addition to grade capping (see

Section3).

Pass Composites Maximum

Composites per Borehole*

Search Ellipse GSLIB

Min. Max Svx*

(metre) Svy*

(metre) Svz*

(metre) A1 A2 A3

1 9 16 8 50 50 50 270 -50 0 2 4 12 3 50 50 50 270 -50 0 3 3 16 90 90 50 270 -50 0 4 1 16 200 200 100 270 -50 0

June 2019


An extensive set of sensitivity models were performed for the southern extent of the grade model

in April 2018. These studies were not re-evaluated at this time. Instead, SRK implemented the

same HG treatment as before, specifically imposing a HG limited radii in this portion of the model.

This effectively permits composite grades above a certain threshold to influence a block estimate

if it resides within limited radii. All references to radii distances in this section are aligned with

the maximum, intermediate and minor axes, respectively.

A progressive capping scheme was adopted. In the HG domain, data were capped at 30 g/t gold

but composite grades greater than 15 g/t gold were only allowed to influence a local neighbourhood

of 30 metres by 30 metres by 15 metres. In the LG domain, composites were capped at 15 g/t, but

composite grades greater than 7.5 g/t gold were only allowed to influence a local neighbourhood

of 30 metres by 30 metres by 15 metres. These grade thresholds were chosen as half the cap value

within their respective domains. The 30 metres by 30 metres by 15 metres radii corresponds to

approximately 95 percent of the variogram value of the global variogram model.

SRK zeroed any blocks that were un-estimated. These tended to occur in the very far south (and

deeper) extents of the model and are a result of being too far from boreholes; there were less than

400 blocks across the various models and almost always in the LG domain. Otherwise, all other

HG and LG blocks were estimated. The LG and HG domain grades were then combined into a

single block grade based on a percentage weighted average of the estimated grade based on fill

volume of the respective mineral resource wireframes. These single block grades were used to

generate the swath plots.

14.6.6 Hwini-Butre

Two block models were produced for the whole Hwini-Butre area. No rotation was applied to the

models. Block sizes and sub-block sizes were chosen to reflect the geometry of the deposits. Grade

data for each of the modelled units was interpolated into the individual structures only. Block

model parameters for Hwini-Butre are summarized in Table 14-23 and Table Table 14-24.

Table 14-23 Adoikrom block model parameters

Coordinate Origin Boundary size Block Size (m) Sub-block count

X 175700 710 10 20

Y 32574.083 1210 10 10

Z 1084.75 820 10 10

Table 14-24 Father Brown Zone block model parameters

Coordinate Origin Boundary size Block Size (m) Sub-block count

X 175736.957 1250 10 20

Y 32417.052 1050 10 10

Z 1044.193 570 10 10

June 2019


Grade estimates for each of the mineralized zones were interpolated using OK. OK was carried

out in three passes for each mineralized zone, and the search parameters for the individual domains

are shown below in Table 14-25. The discretization grid was set at 2x2x1 (xyz) in all cases. The

search ellipsoids applied are similar to the variogram ranges for the first search. For the second

and third search, the search ellipses are at least twice the variogram ranges, with the aim of

estimating most of the blocks unvisited by the first pass.

Table 14-25 Hwini-Butre ellipsoid search neighbourhood parameters

Domain Search 1 Search 2 Search 3 Dip Dip

Azimuth Pitch

Adoikrom

X 50 100 200

65 270 70

Y 30 60 120

Z 30 100 200

Min. Samples

4 3 2

Max. Samples

20 20 20

Father Brown Zone

X 50 100 200

35 240 160

Y 30 60 120

Z 50 100 200

Min. Samples

4 3 2

Max. Samples

20 20 20

The density values used for the tonnage estimate were provided by GSR, and are detailed below

in Table 14-26.

Table 14-26 Hwini-Butre rock density

Oxidisation State Value (t/m3)

Fresh 2.7

14.6.7 Benso

A block model was produced for the whole Benso area. No rotation was applied to the model.

Block sizes were chosen to reflect the average spacing of drill lines along the strike. Grade data

for each of the modelled units was interpolated into the individual structures only, with soft

boundaries between oxidization states, and subsequently reported as oxide or fresh. Block model

parameters for Benso are summarized below in Table 14-27.

June 2019


Table 14-27 Benso block model parameters


X 173750 12.5 300

Y 56000 25 160

Z 1205 10 60

Block grades for each of the mineralized zones were estimated using OK. OK was carried out in

four passes for each mineralized zone, and the search parameters for the individual domains are

shown below in Table 14-28. The discretisation grid was set at 5x2x1 (xyz) in all cases. The search

ellipsoids are relatively large compared to the variogram ranges, but as there is quite a high data

density the blocks were usually estimated with data significantly closer than the edges of the

ellipsoid.

Table 14-28 Benso ellipsoid search neighbourhood parameters

Domain Search 1 Search 2

Subriso East X 100 200

Y 80 160

Z 20 40

Min. Samples 4 4

Max. Samples 36 36

Subriso West, X 100 200

G Zone and Y 80 160

I Zone Z 20 40

Min. Samples 4 4

Max. Samples 36 36

GSR modelled the oxidation surface to determine the boundary between oxide and fresh material.

No transition zone was modelled. The density values used for the tonnage estimate were provided

by GSR and are detailed below in Table 14-29.

Table 14-29 Benso rock density


Oxide 1.8

Fresh 2.7

14.6.8 Chichiwelli

A block model was produced for the whole Chichiwelli area. No rotation was applied to the model.

Block sizes were chosen to reflect the average spacing of drill lines along the strike. Grade data

for each of the modelled units was interpolated into the individual structures only, with soft

boundaries between oxidization states, and subsequently reported as oxide or fresh. Block model

parameters for Chichiwelli are summarized below in Table 14-30.

June 2019


Table 14-30 Chichiwelli block model parameters


X 631,093.64 12.5 100

Y 580,787.2 25 60

Z 1216 (max) 8 65

Block grades for each of the mineralized zones were estimated using OK. OK was carried out in

four passes for each mineralized zone, and the search parameters for the individual domains shown

below in Table 14-31. The discretisation grid was set at 5x2x1 (xyz) in all cases. The search

ellipsoids are relatively large compared to the variogram ranges, but as there is quite a high data

density, the blocks were usually estimated with data significantly closer than the edges of the

ellipsoid. Octants were used on the 1st and 2nd pass searches with three consecutive empty sectors,

however they were not applied on the 3rd search pass, hence the same number of minimum and

maximum samples for the 2nd and 3rd searches.

Table 14-31 Chichiwelli ellipsoid search neighbourhood parameters

Domain Search 1 Search 2 Search 3 Rotation

Parameters

East X 60 120 120 Azimuth: 20

Y 60 120 120 Dip: 60

Z 20 40 40

Min. Samples 3 3 3

Max. Samples 80 80 80

West, X 80 160 160 Azimuth: 20

Y 80 160 160 Dip: 60

Z 10 20 20

Min. Samples 3 3 3

Max. Samples 80 80 80

GSR modelled the oxidation surface to determine the boundary between oxide and fresh material.

No transition zone was modelled. The density values used for the tonnage estimate were provided

by GSR and are detailed below in Table 14-32.

Table 14-32 Chichiwelli rock density


Oxide 1.80

Fresh 2.68

June 2019


14.7 Model Validation and Sensitivity

14.7.1 Wassa

Block models are validated visually by comparing values of estimated block grades and nearby

drill hole composites on a section by section basis and using swath plots for the combined LG and

HG domains along northing, easting and a vertical swath. Sectional checks showed good

consistency between the informing data and local estimated blocks, as well as good conformity of

grade trends to the local folds in the mineralization. Swath plots generally showed that in areas of

abundant data, the model matches well with the composite grades in that moving average window.

Mismatches in the informing composites and the average block grades are attributed to those

regions of the model that are sparsely sampled, specifically in the southern extent of the

mineralized zone (Figure 14-12).

Figure 14-12 South-North swath plot comparing various estimation sensitivities

In all cases south of 19,550 North, the estimated grades are generally higher than the composite

grades in the southern portion of the model. This is attributed to the presence of fewer composites

with some very HG intersections, particularly between 18,900 and 19,000 north. The continuity of

the grade shells in the southern portion also contributes to their pronounced influence over larger

regions.

SRK anticipates that additional drilling in this area may greatly impact the continuity of the grade

domains and dampen the influence of these higher-grade intervals. If future interpretation and drill

results are similar to the northern portion of the Wassa mineral resource model, then this should

lead to lower average grades and potentially less volume.

Golden Star performed some reconciliation of this model with production and the 2015 mineral

resource model. The results of Golden Star’s review of the model were shared with SRK and

June 2019


suggest that this mineral resource update compares better against the grade control model than the

2015 model, while still reflecting the additional drilling and revised geology model.

14.7.2 Hwini-Butre

The block models were validated by comparing the block model mean grades with the declustered

composite mean grades and through validation slices through the block models.

The mean grades for each of the estimated block models were compared to the declustered mean

grade for the composite input data (Figure 14-13 and Figure 14-14). Each of the modelled zones

was compared separately. The differences between the declustered mean composite grades and the

block grades are relatively small, indicating that the model is similar to the input data on a global

scale.

The block model was also compared to the composite grades within defined sectional criteria in a

series of validation slices, the results of which are displayed on graphs to check for visual

discrepancies between grades along the defined coordinate line. The expected outcome of the

estimation process is to observe a relative smoothing of block model grades around the composite

values.

June 2019


Figure 14-13 Adoikrom swath plot

Figure 14-14 Father Brown Zone swath plot

The model validation for Hwini-Butre indicates that the estimation methodology has produced a

relatively robust model, on both the local and global scales. The validation plots for Adoikrom and

Father Brown Zone indicated that there are no obvious biases which have been introduced, and

that the grade distribution of the block model is relatively similar to that of the input data.

14.7.3 Benso




grade for the composite input data. Each of the modelled zones was compared separately. The

differences between the declustered mean composite grades and the block grades are relatively

small, indicating that the model is similar to the input data on a global scale.




June 2019



values.

Overall, the estimation of the Benso domains is robust and the results have been verified to a

reasonable degree of confidence. Globally, the block model average grade is relatively similar to

that of the declustered input data, indicating that no biases have been introduced.

The sectional validation slices show a reasonable correlation between the composite grades and

the block model grades and it appears that a reasonable degree of smoothing has taken place for

the majority of the domains.

14.7.4 Chichiwelli




grade for the composite input data. Each of the modelled zones was compared separately. The

differences between the declustered mean composite grades and the block grades are relatively

small with the largest differences up to 10% for a few of the less well samples domains, indicating

that the model is similar to the input data on a global scale.





values.

Overall, the estimation of the Chichiwell domains is robust and the results have been verified to a

reasonable degree of confidence. Globally, the block model average grade is relatively similar to

that of the de-clustered input data, indicating that no biases have been introduced.

The sectional validation slices show a reasonable correlation between the composite grades and

the block model grades and it appears that a reasonable degree of smoothing has taken place for

the majority of the domains.

14.8 Mineral Resource Classification

14.8.1 Introduction

Block model quantities and grade estimates for the Wassa HBB Project were classified according

to the CIM Definition Standards for Mineral Resources and Mineral Reserves (December 2005).

Mineral Resource classification is typically a subjective concept. Industry best practices suggest

that resource classification should consider the confidence in the geological continuity of the

mineralized structures, the quality and quantity of exploration data supporting the estimates and

the geostatistical confidence in the tonnage and grade estimates. Appropriate classification criteria

should aim at integrating all concepts to delineate regular areas at similar resource classification.

The geological modelling honours the current geological information and knowledge. The

location of the samples and the assay data are sufficiently reliable to support resource evaluation.

June 2019


The sampling information was acquired primarily by diamond core and RC drilling on sections

spaced at variable distances between the different deposit areas.

14.8.2 Wassa

A combination of quantitative and visual assessment was used to recommend a possible

classification scheme. The Measured category is constrained to the areas where grade control

drilling is available. This corresponds to an average distance of less than 10 metres to informing

composites. Given this setting, SRK independently assessed areas that may reasonably be

classified as Indicated. This corresponds to an average distance of less than 20 metres to informing

composites. Overall, the resulting blocks that satisfied these criteria showed good continuity for

the Indicated category. A similar strategy was used for Inferred category, requiring a minimum of

three boreholes found within a 70-metre radius, which is equivalent to a 100-metre borehole

spacing. Inferred blocks were supported by informing composites within 65 metres of the

estimated block.

Golden Star agreed with the suggested classification criteria. Smoothening of the classification

was performed by Golden Star, which SRK reviewed and found to be appropriate. The

classification was finalized by Golden Star.

14.8.3 Hwini-Butre

Classification for Hwini-Butre generally follows the same general principles as those applied at

Wassa. Classification has been assigned using a combination of drillhole spacing, geological and

wireframe confidence, as well as slope of regression values from the estimation process. The

classification was modelled visually by digitizing a wireframe in order to define contiguous zones

of confidence.

The Indicated wireframe was extended approximately half the drill hole spacing on section, as this

is where confidence in the geological interpretation was considered to reduce. Indicated Mineral

Resources have been defined in the areas of Hwini-Butre where drilling is sufficient to demonstrate

geological and grade continuity to a reasonable level, with a >0.6 slope of regression value used

as a rough guide. Inferred Mineral Resources have been defined in the remainder of Adoikrom and

Father Brown Zone.

The Resources were classified under the Guidelines of NI 43-101 and accompanying documents

43-101.F1 and 43-101.CP. A series of wireframes were digitised for Adoikrom and Father Brown

Zone, with the areas inside the modelled solids considered to be Indicated Mineral Resources, and

outside, Inferred Mineral Resources.

14.8.4 Benso

Classification for Benso generally follows the same general principles as those applied at Wassa

and Hwini-Butre. Classification has been carried out using a combination of drillhole spacing,

geological and wireframe confidence, and was modelled visually by digitizing a wireframe.

The Indicated wireframe was extended approximately half the drill hole spacing on section, as this

is where confidence in the geological interpretation was considered to reduce. Indicated Mineral

Resources have been defined in the Subriso East, Subriso West and G Zone areas of Benso where

drilling is sufficient to demonstrate geological and grade continuity to a reasonable level.

June 2019


Inferred Mineral Resources have been defined in some parts of Subriso East, Subriso West, G

Zone, and I Zone.

14.8.5 Chichiwelli

Classification for Chichiwelli generally follows the same general principles as those applied at

Wassa, Hwini-Butre and Benso, with classification carried out using a combination of drillhole

spacing, geological and wireframe confidence, and was modelled visually by digitizing a

wireframe.

The Resources were classified under the Guidelines of NI 43-101. Wireframes were digitised for

East Domain and West Domain, with the areas inside the modelled solids considered to be

Indicated Mineral Resources, and outside, Inferred Mineral Resources.

The majority of the Chichiwelli Mineral Resource is reported in the Indicated category. For the

three additional deposits covered by the Chichiwelli MRE, an Inferred classification has been

applied.

14.9 Mineral Resource Estimate

The following section presents the combined open pit and underground Mineral Resource estimate

for the Wassa Main and HBB deposits. Mineral Resources are reported inclusive of the material

which makes up the Mineral Reserve. The Mineral Resource Statement is presented in accordance

with the guidelines of NI 43-101.

GSR commissioned SRK to construct a mineral resource model with estimated gold grades for the

Wassa Main. The mineral resource classification and statement was conducted by GSR under the

supervision of S. Mitchel Wasel, a QP.

The “reasonable prospects for eventual economic extraction” requirement generally implies that

the quantity and grade estimates meet certain economic thresholds and that the Mineral Resources

are reported at an appropriate COG, taking into account extraction scenarios and processing

recoveries.

In order to determine the quantities of material offering “reasonable prospects for economic

extraction” by open pit mining, GSR used a pit optimizer and reasonable mining assumptions to

evaluate the proportions of the block model (Indicated and Inferred blocks) that could be

“reasonably expected” to be mined from an open pit.

The optimization parameters are based on actual costs from the operations. The reader is cautioned

that the results from the pit optimization are used solely for the purpose of testing the “reasonable

prospects for economic extraction” by an open pit and do not represent an attempt to estimate

Mineral Reserves.

GSR considers that the blocks located within the conceptual pit envelopes show “reasonable

prospects for economic extraction” and can be reported as a Mineral Resource.

Table 14-33 shows the combined Mineral Resource statement for the Wassa Main and HBB

deposits.

June 2019


Table 14-33 Mineral Resource estimate as of December 31, 2018

In declaring the Mineral Resources for the Wassa Main and HBB deposits, the following are noted:

• The identified Mineral Resources in the block model are classified according to the CIM

definitions for Measured, Indicated and Inferred categories and are constrained within a

Whittle pit shell using a gold price of US$1,450/oz and below December 2018

topographic surface. The Mineral Resources are reported in-situ without modifying

factors applied.

• The Wassa open pit Mineral Resource estimate is based on a COG of 0.4 g/t Au reported

within a conceptual Whittle shell. Pit optimization using industry standard software has

been undertaken on the Mineral Resource models using appropriate slope angles, process

recovery factors and costs.

• The Wassa underground portion of the Mineral Resource estimate is based on a COG of

2.1 g/t.

• The HBB Underground Mineral Resource has been estimated below the $1,450 per

ounce of gold pit shell using an economic gold grade cut-off of 3.2 g/t Au, which the

Company believes would be the lower COG for underground mining.

• The Mineral Resource models have been depleted using appropriate topographic

surveys.

• Block model tonnage and grade estimates were classified according to the CIM

Definition Standards for Mineral Resources and Mineral Reserves (December 2005).

The basis of the Mineral Resource classification included confidence in the geological

continuity of the mineralized structures, the quality and quantity of the exploration data

supporting the estimates, and the geostatistical confidence in the tonnage and grade

estimates.

• All figures are rounded to reflect the relative accuracy of the estimate.

• Mineral Resources are not Mineral Reserves and do not have demonstrated economic

viability.

June 2019


15 Mineral Reserves

15.1 Cut-off Grade Estimate

The following methodology and inputs were used to estimate the COG:

• Underground mining costs based on actuals for the 18 months prior to December 2018.

• Open pit mining costs based on actuals during 2017, adjusted to account for the larger

scale of the future Cut 3 mining.

• Processing based on historical actual costs.

• G&A based on historical actual costs.

• Process recovery based on historical actual costs.

• Government royalty of 5% on gross revenue.

• Stream payment of 8.4% on gross revenue.

Table 15-1 shows the COG estimates for the open pit and underground operations. The Marginal

and Break-even COGs presented are “plant-feed” estimates, hence they do not include dilution as

per in-situ cut-off estimates.

Table 15-1 Cut-off Grade estimate

15.2 Mineral Reserve Statement

The Wassa Mineral Reserves were estimated based on the Mineral Resources that are classified as

Measured and Indicated. The Reserves are summarized in Table 15-2.

June 2019


The Mineral Reserves have been prepared in accordance with CIM standard definitions for Proven

Mineral Reserves and Probable Mineral Reserves. The Indicated Mineral Resources reported

above include those Mineral Resources modified to estimate the Mineral Reserves.

The Mineral Reserves have been estimated using accepted industry practices for open pit and

underground mines, including the identification of the optimal final ore envelopes based on the

selected mining methods, appropriate modifying factors and COG estimates based on detailed cost

estimation. The identified ore bodies were subjected to detailed mine design, scheduling and the

development of a cash flow model incorporating the company’s technical and economic

projections for the mine for the duration of the LoM plan. A gold price of US$1,250/oz was used

for the Reserve estimation.

Any mineralization which occurs below the COG or is classified as an Inferred Mineral Resource

is not considered as Mineral Reserves and is treated as mineralized waste for the purposes of the

LoM plan. The Wassa Mineral Reserve Statement is as of 31 December 2018.

Table 15-2 Mineral reserve estimate as of December 31, 2018

Notes to Mineral Reserve Estimate:

• Mineral Reserve estimates reflect the Company’s reasonable expectation that all

necessary permits and approvals will be obtained and maintained. Mining dilution and

mining recovery vary by deposit and have been applied in estimating the mineral

reserves.

• Mineral Reserves are the economic portion of the Measured and Indicated Mineral

Resources. Mineral Reserve estimates include mining dilution at grades assumed to be

zero.

• The Mineral Reserve estimate was prepared under the supervision of Dr. Martin Raffield,

Chief Technical Officer for the Company and a QP.

• The Mineral Reserves at December 31, 2018 were estimated using a gold price

assumption of $1,250 per ounce.

• The slope angles of all pit designs are based on geotechnical criteria as established by

external consultants. The size and shape of the pit designs are guided by consideration

of the results from a pit optimization program.

• COGs have been estimated based on operating cost projections, mining dilution and

recovery, government royalty payment requirements and applicable metallurgical

June 2019


recovery. The marginal COG used for the open pit estimate is 0.7 g/t and the break-even

COG used for the underground estimate is 2.4 g/t.

June 2019


16 Mining Methods

16.1 Open Pit Mining

This section discusses the proposed pushback on the south side of the B Shoot pit named Cut 3.

The current LoM plan has Cut 3 commencing in 2023 towards the end of the mining of the

underground reserve.

16.1.1 Geotechnical Design

For the US$1,250 pit design, optimization angles of 45° and 52° were used for the oxide and fresh

rock masses respectively. Detailed engineering design was based on the following nominal bench

and berm configurations:

• Oxides: Bench height: 9 m, bench face angle: 60°, berm width: 4 m (Inter-ramp angle:

45°).

• Fresh Rock: Bench height: 18 m, bench face angle: 75°, berm width: 6 m (Inter-ramp

angle: 59°).

Utilisation of these parameters has resulted in a pit design with a maximum slope height of about

250 m at an overall slope angle (including ramps) of 50°. A maximum inter-ramp slope height of

140 m at an angle of 59° is achieved.

The overall final design slopes are stable against rock mass failure as the Wassa pit slopes are

formed in a strong, competent but well jointed rock mass.

16.1.2 Mining Method

A conventional approach to the pushback will be used employing excavators and trucks which are

considered typical for this type and style of gold mineralization. The mining will be carried out by

a contract mining company who will supply equipment, manpower and supervision services.

Drilling and blasting will be conducted over bench heights of 6 m and explosives delivered to the

hole by the manufacturer. Oxide or weathered material is generally only required to be lightly

blasted or in some areas can be excavated as ‘free dig’. Hydraulic excavators are used in

conjunction with conventional blasting practice, to mine a 2.5 or 3.0m flitch height. Broken rock

is loaded to 95t capacity off-highway haul trucks to a central stockpile or to the waste dump.

16.1.3 Mineral Inventory

Cut-off Grade Estimate

The COG for the Wassa open pit material is based on various estimates and assumptions,

including:

• gold price of US$1,250/oz;

• a Government gross revenue royalty of 5%;

• RGLD stream payment of 8.4% of gross revenue;

June 2019


• a process plant recovery for oxide and fresh material of 94%; processing costs are based

on US$15.0/t for material treated;

• the base mining cost (including grade control and haulage) is estimated to be US$3.35/t

for fresh material; and

• G&A cost of US$5.00/t processed.

The COG calculations are presented on Table 15-1 which shows a plant feed marginal COG of

0.7g/t and a break-even COG of 1.2g/t.

Pit Optimization

A pit optimization exercise was undertaken using Whittle Four-X software (“Whittle”) and

imported resource models with an estimate of the topographic surface as at December 31, 2017.

The imported block model is sized at 10 x 10 x 6 m blocks.

The Whittle pit optimization utilized various technical and economic assumptions obtained from

a combination of operating history, experience and company objectives with regards to gold price

and pit shell selection. The base case gold price for 2018 mine planning is US$1,250/oz. A

summary of the principal optimization parameters is given in Table 16-1.

Table 16-1 Wassa Pit Optimization Input Parameters

Parameter Unit Value

Revenue

Au price US$/oz 1,250

Government and stream royalty % 8.4+5

Mining Parameters and Costs – Wassa

Mining recovery % 95

Dilution % 10

Overall slope angle deg. 52

Reference Elevation m 154

Base Mining cost US$/t 2.75

Incremental Pit Depth Cost US$/t/m 0.003

Processing Parameters and Costs

Haul to Plant US$/t 0.48

Process plant recovery % 94

Process Cost US$/t 15.00

Other Costs

G&A Cost US$/t 5.00

Figure 16-1 shows the pit optimization results in terms of material movement, strip ratio and the

discounted cash flow for a range of gold prices. The best case NPV assumes optimized shells are

mined successively to the final shell, while the worst case NPV assumes that the final shell is

mined directly from top to bottom.

June 2019


Figure 16-1 Wassa Pit Optimization Results

16.1.4 Pit Design

The topography as of December 2017 was the starting point for the open pit design and is shown

in Figure 16-2.

Figure 16-2 Wassa Main Topography as of December, 2017

June 2019


The final pit design slopes and benches are based on the geotechnical parameters provided in Table

16-2.

Table 16-2 Wassa Open Pit Design Geotechnical Parameters

Parameter Unit Wassa Main

Oxide Slopes

Berm Width m 4

Bench Height m 9

Batter Angle deg 60

Overall Slope Angle deg 45

Inter-ramp Slope Angle deg 45

Fresh Slopes

Berm Width m 6

Bench Height m 18

Batter Angle deg 75

Overall Slope Angle deg 55

Inter-ramp Slope Angle deg 59

The pit ramps are designed at a width of 20 m for two-way haulage, reducing to 10 m for one-way

traffic in the last 20 m vertical and installed at a gradient of 10%. A minimum mining width of

30 m is utilized assuming the space required for a CAT 777 haul truck to perform a 3-point turn.

The results of the pit design have been utilized in conjunction with the latest block model,

topography and face positions to determine the contained mineable resource using the Measured

and Indicated Mineral Resource categories only.

Figure 16-3 shows a plan view of the topography of the current pits and outlines of the Cut 3 and

242 pushbacks.

Figure 16-3 Plan view of pits showing outlines of Cut 3 and 242 pushbacks

June 2019


Figure 16-4 shows the progression of open pit mining from current situation, to mining of Cut 3

and 242 pushbacks.

Figure 16-4 Pit design progression

Figure 16-5 to Figure 16-7 show sections across the B Shoot (Cut 3) and 242 pits including current

topography, final pit design and the block model. Note the location of HG underground targets

above 2.5g/t below the final bit bottom.

June 2019


Figure 16-5 Cross section locations

Figure 16-6 Section A-A' (Fig. 16-5) showing block model

June 2019


Figure 16-7 Section B-B' (Fig. 16-5) showing block model

Table 16-3 shows a comparison between the Whittle optimization results and the final reserve

design. Significantly more material was identified by Whittle than was incorporated into the

reserve design due to the following factors:

• optimization extended into HG crown pillars but was removed in final design in order to

keep underground and open pit separate;

• optimization identified small pockets of material outside of the main pit areas which overlie

access infrastructure and are thus not mineable; and

• complex design in pits to integrate current ramp accesses into pushback designs.

Table 16-3 Total material movement by stage for open pit mining

Optimization Pit Design

Ore Tonnes

[Mt]

Grade

[g/t]

Waste Tonnes

[Mt]

Ore Tonnes

[Mt]

Grade

[g/t]

Waste Tonnes

[Mt]

14.1 1.77 78.8 9.9 1.57 69.0

16.1.5 Pit Mining Schedule

The open pit has been scheduled (in conjunction with the underground) using Geovia MineSched.

An annual material movement of 12 to 15 Mtpa is sustained over the LoM. These rates have been

historically achieved on site.

The stages in order of priority are:

June 2019


• main pit, Cut 3 pushback; and

• 242 pit pushback.

Figure 16-8 shows ore tonnes, grade and waste tonnes by month and by pit and the sub-staging is

shown in Figure 16-9.

Based on historical performance at Wassa, the modifying factors applied to the in situ block model

are 10% external dilution at 0 g/t Au and 5% ore-loss.

The key parameters used in the open pit schedule were:

• mining commences in January 2023;

• mining rate of 35,000 to 40,000 tpd; and

• as Cut 3 tonnage drops off towards the bottom of the pushback, the available capacity

moves to the 242 pushback.

June 2019


Figure 16-8 Tonnes mined by month

June 2019


Figure 16-9 shows a plan and isometric view of the annual mining schedule

Figure 16-9 Open pit schedule plan and isometric views

16.1.6 Personnel

Currently, the open pit mining is on care and maintenance with only a minimal number of staff

focussed in this area. In 2023, when the pit mining restarts, a contractor will be engaged to carry

out the mining and the Company’s supervision and management structure will be reconstituted.

June 2019


16.1.7 Equipment Fleet

GSR has open pit mining equipment but much of this equipment is close to end of life. In 2023, a

contractor will be engaged who will bring a complete fleet of mining operations mobile equipment.

The equipment that is currently in operation or on care and maintenance is shown in Table 16-4.

Table 16-4 Current equipment fleet

Description Total

Excavators 3

Haul Trucks 12

Blasthole Drill Rigs 4

Dozers 5

Graders 3

Water Truck 1

Total 28

16.1.8 Mine Services

Waste Handling

Waste dumps are located as close to the final pit limit as possible and include design parameters

comprising a lift height of 10 m, berm width of 10 m and a batter angle of 37°. The final waste

dump will be re-profiled for closure to an overall slope angle of 22°.

The waste storage design shown in Figure 16-10 is incorporated into the mining schedule and has

sufficient capacity for the life of the mine.

June 2019


Figure 16-10 Wassa Main dump design in relation to final pit design

Pumping…

Groundwater does not appear to have posed a major risk to the project thus far and has been

managed without difficulty. Wassa maintains an array of dewatering bores that are monitored to

understand the hydrogeological conditions impacting on the open pit environments.

The pits are dewatered using in-pit pumps placed in sumps located in the lowest point of the pit.

Wet season arrangements include drainage diversion channels to manage surface water runoff

flows. Water is pumped to sediment settling ponds and released to the receiving environment in

compliance with relevant permits.

Groundwater inflows vary seasonally but, based on the current dewatering, there is a low

contribution to overall pit inflows. High groundwater inflows sometimes occur but are managed

using the current dewatering system.

Storm water runoff to the Starter pit catchment has been minimized through a combination of

catchment modifications (to reduce catchment size), fresh water diversions (to divert water away

from the Starter pit), and the establishment of a storm water collection sump below the

underground portal entrance. The design of the storm water collection sump has been

conservatively designed for a 1:100 year, 241 mm 24 hour duration event, over the Starter pit

June 2019


catchment of 20 ha. After backfilling of the pit to provide a platform for the portal, the storage

volume is 64,000 m3 which is 115 % of the capacity required to store the water from the 1:100

year event.

16.2 Underground Mining

16.2.1 Geotechnical Design

This section reviews the geotechnical design parameters reflecting the current practical situation

based on underground mapping and data collection since the inception of the underground mine

in 2015.

• Wall mapping was carried out in both permanent openings and ore drives to define the

structural discontinuities of the Wassa rock mass.

• Empirical support classification assessment was also carried out to determine the support

requirement for the permanent drives.

• Empirical stability graph assessments were carried out based on the rock mass

parameters from geotechnical characterization to determine the maximum stable spans

of the stopes.

• Numerical modelling software packages (Phase 2 and Examine 3D) are used to assess

the stability and stress distributions around the stope spans and the crown pillar.

Mapping Data

Based on the structural assessment of the geotechnical mapping carried out, the structures

presented in Table 16-5, were used to carry out the stope stability assessment.

Table 16-5 Joint Sets used for Stope Design

Rock Quality

Based on the structural assessment of the geotechnical mapping, the rockmass quality is considered

good using Barton’s classification and Geological Strength Index (GSI) rating. Table 16-6

represents the rock mass conditions of the Wassa geotechnical domains and it is used for the stope

stability assessment and ground support design.

Foliation 55 275 Tightly healed foliation planes

Set of Sub-horizontal north west trending joints

Set of tightly healed North East trending joints

Set of Steeply dipping north trending joints

Discontinuity

SetDip

Dip

DirectionComments

J3 15 276

J4 79 42

J1 45 11

J2 62 155 Set of tightly healed south east trending joints

June 2019


Table 16-6 Q Index (Q’) estimate

Support Requirement

Barton’s Q support classification system (Figure 16-11 Barton’s Q-Index Chart) has been

used to estimate the support requirement for the footwall and orebody extraction development.

The width of the footwall and orebody extraction development is 5.5 m and they are deemed to

have an Excavation Support Ratio of 1.6 defined as permanent mine openings by Barton.

The development excavations are plotted in red and are seen to be within the No Support Required

region of the chart. This conforms to observations which indicate very good rockmass with little

or no fallout and spalling. A standard pattern of bolts and mesh is applied to the roof and upper

walls of all development excavations notwithstanding the results of the analysis.

Figure 16-11 Barton’s Q-Index Chart

Modified Stability Number (N’) for Longitudinal and Transverse Stopes

The Q’ value derived from the geotechnical characterisation has been used in conjunction with the

stability graph parameters A, B and C to determine the Modified Stability Number (N’) for stope

June 2019


hangingwall, back and end walls. The stress parameter A has been estimated by calculating the

gravitational stress generated from the weight of the overburden rock above the mining. The

structural parameters B and C were derived from an assessment of the interaction of the dominant

joint sets with the stope boundaries.

Calculated N’ for the Q’ value derived from the rock mass characterisation for both the

longitudinal and transverse stopes are presented in Table 16-7.

Table 16-7 Modified stability number (N’) for longitudinal stope

Hydraulic Radius

Table 16-8 shows the hydraulic radius calculation for a stope that is 50m high by 25m wide with

an ore thickness of 30m. The orientation of the measurement axes is shown on Figure 16-12.

Table 16-8 Hydraulic radius of stope geometry

June 2019


Figure 16-12 Stope axes measurements

Figure 16-13 shows a plot of the stability graph for the stope dimensions described above with

all the faces of the stope plotting in the stable portion of the graph. This indicates that stopes of

50 m high by 25 m wide by 30 m ore thickness will be stable and this conclusion is confirmed by

field observations. The stability graph also indicates that the stope height could be increased to

75 m and possibly 100 m maintaining stability.

Figure 16-13 Stope stability graph

June 2019


Crown Pillar Stability Modeling

Phase 2 software was used to analyze crown and sill pillars that are of concern to underground

operations. Factor of safety of crown pillar between the lowest portion of B-shoot open pit and

underground stope 720N1 and in sill pillar left between 745S1 and 720N1 is 1.58 as shown in

Figure 16-14. This indicates that a stable crown pillar can be maintained between the open pit and

underground operations.

Figure 16-14 Phase 2 model: crown and sill pillar strength factor

16.2.2 Mine Design

Cut-off Grade Estimate

The COG for the Wassa underground material is based on various estimates and assumptions,

including:

• gold price of US$1,250/oz;

• a Government gross revenue royalty of 5%;

• a process plant recovery for oxide and fresh material of 95%; processing costs are based

on US$20.0 /t for material treated;

• the mining cost is estimated to be $48/t; and

June 2019


• G&A cost of US$9.00/t processed.

The COG calculations are presented on Table 15-1, which shows a plant feed break-even COG of

2.1 g/t.

Dilution and recovery

Dilution is estimated to be approximately 10% based on historical stope reconciliation. Mining

recovery is estimated to be 95% based on historical stope reconciliation. Note that the COGs

presented are for plant feed and that slightly higher in-situ COG of 2.4 g/t is used in the stope

design process which incorporates the dilution factor.

Stope Optimization

Stope dimensions were determined based on geotechnical recommendations and the mineralized

geometry. Stope optimizations were run using the Datamine Mining Shape Optimizer (“MSO”)

with an in-situ cut off-grade of 2.4 g/t Au.

The results of the stope optimizer are used to guide the detailed stope design process in terms of

identifying areas of consistent volume and grade. The stope optimizer results are not used to report

ore reserve estimates.

Current Mining As-Built

Figure 16-15 shows the as-built topography, open pit mining and underground development and

stoping in the B Shoot and F Shoot areas.

Figure 16-15 Long section looking east of open pit and underground as-built

June 2019


Access Development

Figure 16-16 shows a plan view of the as-built underground development and stoping with the

three surface portal locations marked by red triangles. The underground mine commenced with

the development of main access and ventilation return declines from the Starter Pit. More recently

a third surface portal was added in the B Shoot pit to provide increased intake ventilation capacity

and to reduce the tramming distance for waste haulage to surface.

Figure 16-16 Plan view of as-built development and stoping

June 2019


Figure 16-17 shows a view of the Starter Pit portal area including the sump and the ore stockpiles.

The direction that this view was taken from is shown in Figure 16-16.

Figure 16-17 Photograph of Starter Pit portal area

The main decline is designed with a cross-section of 5.8 m H x 5.5m W to allow truck passage

whilst under secondary ventilation conditions. To optimize the cost-per-tonne performance of the

project, large mechanised equipment is vital.

The decline will also act as the primary ventilation intake for the initial years of operation. The

Ghanaian mining regulations restrict air velocity to 6 m/s under these conditions. This size heading

will allow up 191 m3/sec of volume. Figure 16-18 shows a cross-section of the main decline with

the mine services (ventilation, piping, communications) with respect to a CAT AD55B

underground haul truck.

June 2019


Figure 16-18 Cross-section of main decline

The concepts guiding the mine design are:

• sub-level accesses on 25 m vertical centres aligned along the strike of the mineralization

to facilitate support service connections sub-vertically;

• all primary development in the footwall of the mineralization for long-term geotechnical

stability; and

• development of main ramps at a constant 15% (1:7; 8°) grade to maximize vertical gain

per metre developed and minimize final haul distance, which is the major operating cost

activity.

Figure 16-19 shows the planned development of the ore reserve down to 470L. Figure 16-20 shows

an isometric view of the same information looking to the north-east.

June 2019


Figure 16-19 As-built and planned ore reserve development (looking east)

Figure 16-20 Isometric view of as-built and planned development (looking NE)

June 2019


Stope Design

Design of the stopes to be mined was guided by the MSO stope optimization process. The MSO

generated shapes were further evaluated on the following parameters:

• That 25 m sub-levels would be the final spacing, and which stopes highlighted in the 10,

15 and 20 m high MSO runs might be able to be taken. For example, as blind up-hole

stopes of lesser height.

• Location to the final open pit base and walls. The evaluation considers crown pillar

requirements and the stability of final open pit walls.

• Isolated stoping areas are evaluated with consideration to the mine development costs to

produce from these areas.

Figure 16-21 and Figure 16-22 show the as-built and planned development and stoping in long

section and isometric views.

Figure 16-21 As-built and planned development and stoping (looking east)

June 2019


Figure 16-22 Isometric view of as-built and planned development and stoping (looking NE)

16.2.3 Paste Backfill System

A paste backfill system FS was completed in Q1 2018 and has transitioned into project and

construction execution phase. The paste backfill plant is scheduled for operation in Q3 2020 and

will supply 4,000 tpd of paste to the underground stopes. The use of paste backfill will allow the

operation to transition to a primary/secondary stoping system in contrast to the current system

which leaves 10 m pillars between the stoping panels. This change will increase ore recovery and

provide additional ore feed from each sub-level.

The FS was carried out by Outotec (Canada) Ltd which included paste characterization, strength

testing at different cement ratios, delivery pipe flow characterization, plant design, capital and

operating cost estimation. The FS evaluated the feasibility of using direct full stream tailings feed

from the process plant.

The findings and conclusion of the FS are presented below:

• The paste testwork for material characterization, rheology and strength met the acceptable

criterion.

June 2019


• Dewatering, including thickening and vacuum filtration, was achieved through proven unit

processes typical of most backfill plants.

• Strength requirements for typical primary stopes of 20m (L) x 20m (W) x 25m (H) in size

will require 4.5% cement to achieve the required strength of 270kPa. Secondary stopes

will require 3% cement to achieve 150kPa.

• Underground distribution was amenable through gravity (rather than by positive

displacement pumps) owing to the surface location of the paste backfill plant relative to

the underground stopes.

16.2.3.1 Process Overview

Tailings will be pumped from the end of the CIL plant either to the TSF or to the paste backfill

plant, with automated changeover and flushing processes. The tailings transfer will have

secondary containment along the approximate 3 km length of the services corridor connecting the

gold processing plant to the paste backfill plant. Figure 16-23 shows the relative locations of the

gold processing plant and the proposed paste backfill plant.

Figure 16-23 Process plant, tailings transfer and paste backfill plant locations

June 2019


The tailings feed to the paste backfill plant will directly enter a thickener. Thickened tailings

underflow will be pumped into a large agitated storage tank that is sized to provide extended plant

operation in scenarios where there is a lower feed to the paste backfill plant. A dedicated

flocculation system will provide flocculant to both the thickener and filters within the paste backfill

plant.

The paste backfill plant will be a continuous process when operating. Disc filtration will be used

to produce filter cake, which will be added to a continuous mixer. Dry cement and thickened

tailings that bypass filtration will also be added to the mixer to achieve the required density and

binder content. The process flow diagram for the paste backfill plant is shown in Figure 16-24.

The cemented paste will be delivered to the stopes via gravity through a cased borehole and an

underground pipe network to each stope a required. Excess water from dewatering will be pumped

either to water ponds adjacent to the CIL plant, or to the TSF, depending on site water management

and water quality.

The paste backfill plant will be automated and equipped with a Human Machine Interface system

which allows for plant control and will also give the operators easy access to pressure values from

sensors installed on the distribution system. Live and trended data on these underground sensors

are a critical tool for operation of the plant.

In additional to instrumentation underground, the paste backfill plant will be connected to the gold

processing plant, as this connectivity is required so that the tailings management system (pumping

to either the TSF or to the backfill thickener) can be observed and controlled from both sites. As

this connection is required, the control room at the CIL plant would have the ability to observe the

backfill plant operation and any alarms.

June 2019


Figure 16-24: Paste Backfill Plant Process Flow Diagram

June 2019


16.2.4 Mining Schedule

Figure 16-25 shows the underground mining schedule for the ore reserve from 2019 to 2024.

During this period, 7.5Mt of ore will be mined at a grade of 3.95 g/t for 949,000 oz contained.

Figure 16-25 Underground mining schedule

16.2.5 Underground Personnel

The total workforce for Wassa by year and by department is shown in Table 16-9 for the duration

of the mining of the underground and open pit reserves.

Table 16-9 Total Workforce by LoM Year

June 2019


16.2.6 Equipment Fleet

The underground mine equipment fleet is presented in Table 16-10.

Table 16-10 Equipment fleet

Equipment Fleet 2019 2020 2021 2022 2023 2024

Twin boom jumbo 4 5 5 5 4 3

LHD 5 5 5 5 5 4

40T truck 7 7 6 5 4 4

60t truck 1 3 4 4 4 4

Longhole drill 3 3 3 3 3 2

Bulk explosives loader 1 1 1 1 1 1

Scaler 1 2 2 2 2 1

Integrated tool carrier 3 3 3 3 3 2

Telehandler 1 1 1 1 1 1

Grader 2 2 2 2 2 1

Light vehicles 14 14 14 14 12 8

16.2.7 Ventilation Design

Basis of design

Ventilation design analysis using VentSim software is conducted regularly to determine primary

and secondary fan requirements. The principal objectives of a ventilation design are:

• to remove the diesel exhaust fumes from mechanised mobile equipment;

• to remove blasting fumes from the workings and provide for a reasonable re-entry period;

and

• to maintain working conditions in the mine in accordance with mine regulations.

Ghanaian Mining regulations stipulate the following:

• maximum velocity of 6 m/s in travelling roadways;

• for diesel engine equipment, not less than 0.06 m3/kW/s;

• minimum velocity of 0.2 m/s in headings and 0.1 m/s in large openings;

• 32.5ºC wet bulb is the maximum that men are allowed to work in; and

• CO2 must be continuously monitored in return airways and information transmitted to

surface.

Current System and LoM Ventilation Modification Design

The mine is currently developed from a portal via two declines (main decline and ventilation

decline). The primary fresh air flow is drawn down the mine through the main decline and return

air is exhausted parallel through the ventilation decline by 4 x 90 kw exhaust fans.

June 2019


This system is used to ventilate active working levels from 820 Level to 645 Level by the aid of

secondary fans. Return air raises of 25m high x 4m x 4m are developed to connect each level and

finally connects the ventilation decline at 795 Level.

Panel 1 Ventilation System

The current ventilation system will be used until Q3 2019 when a primary exhaust raise of 4.1 m

diameter is raisebored from 695 Level underground to 844 Level in the open pit.

The exhaust raise will be fitted with two exhaust fans underground with a system duty point of

160 m3/s at a total fan pressure of 2.0kPa, requiring shaft electrical power of 560kW. This will run

simultaneously with the current 4 x 90kw fans running at 158.8m3/s with a fan pressure of 1.6 kPa,

which was be upgraded to 132 kW in Q4 2018.

Panel 2 And 3 Ventilation System

A series of 4 m x 4 m wide inter-level fresh air raises will be excavated from 670 Level to 570

Level. These raises will act as bypasses for the ramp to decrease the ramp air velocity.

As mining progresses deeper, a booster fan of maximum fan system duty point of 320 m3/s at a

total fan pressure of 2.6 kPa will be installed on 645 return air way drive which will be connected

to the vent raises below 645 Level.

Ventilation Quantities

Considering that the removal of diesel fumes is generally the primary concern in trackless mining

operations, the calculation is based on a diesel dilution rate of 0.06 m3/s/kW of diesel engine

power. Using the entire fleet with their modelled availability and utilisation gives a realistic steady

state estimate of the total diesel fleet in operation. When calculating secondary air requirements

such as an ore-drive where an LHD will be the largest engine operating, then 100% of the rated

power is used for sizing requirements.

Airflow Simulation Inputs and Results

The following key system inputs were used in the simulation:

• All headings assumed as design without over-break.

• Secondary airflow to mining area per the schedule.

• All headings and drop raises friction factor of 0.0115 kg/m3.

• The primary raise-bore was given a friction factor of 0.0049 kg/m3.

• Shock losses are automatically modelled for all development.

• A minimum airflow for a mining/development area was elevated to 35 m3/s with an

average value of 40 m3/s was used to make allowances for duct leakage, flow through,

and parallel drives.

• Any bulkheads required have a resistance of 250 Ns2/m3, 50 Ns2/m3 and stope curtains

of 0.2 Ns2/m3.

Estimated airflow requirements on Table 16-11 take into account the fleet schedule and mining

schedule.

June 2019


Table 16-11 Estimated Ventilation Requirement

16.3 Combined Underground and Open Pit Mining Schedule

Table 16-12 and Figure 16-26 show the combined open pit and underground mining schedules

from 2018 to 2025.

Table 16-12 Open pit and underground production schedule

June 2019


Figure 16-26 Open pit and underground production schedule

June 2019


17 Recovery Methods

17.1 Flow Sheet Description

Gold recovery is achieved at Wassa through the use of conventional CIL technology, although the

plant itself contains a few atypical features due to its history and development. The Wassa

processing operation was originally started in 1998 and incorporated HL technology to recover

gold from the mined and prepared ore. This involved crushing, screening and agglomeration of the

mined feed material before being placed on HL pads and irrigated with a weak cyanide solution to

recover the gold. The solution was processed through carbon columns, stripped from the loaded

carbon and smelted through to gold doré bars.

Forecast recoveries using HL processing were not achieved and the HL plant was closed in 2001.

Following a FS completed in 2003, the plant was subsequently restarted by GSR in 2005 using

crushing, milling and CIL. The CIL plant was designed to process 3.5 Mtpa from a feed blend

comprising 45% fresh ore, 25% oxidised ore and 30% reclaimed spent HL material. Spent HL

material reclaimed from the pads was added to the mill feed via a scrubber until this material was

depleted in 2014. Following this time, mill feed comprised fresh ore from the open pit until 2015,

where underground material was blended with the open pit ore.

The plant flowsheet has transitioned from the historical HL processing and currently consists of

the following operations:

• A four-stage fine crushing circuit is employed incorporating an open circuit primary jaw

crusher followed by secondary, tertiary and quaternary cone crushers with the secondary

and tertiary crushers operated in closed circuit with sizing screens. A single secondary,

two tertiary and four quaternary crushers give a nominal crushed product size from the

crushing circuit of 80% <8 mm.

• Two independent milling circuits, each comprising a 5.03 m diameter x 6.7 m long ball

mill fitted with 3 MW motors feeding individual clusters of classifying cyclones.

Reported mill product size is around 70% <75 µm.

• Two separate gravity gold recovery circuits using 48” Knelson centrifugal concentrators

process a portion of the classifying cyclone feed in each mill circuit.

• The gravity concentrate from the Knelson concentrators is retreated using a Gemini

shaking table to produce a HG gold concentrate for direct smelting, while the tails from

the centrifugal concentrators and shaking table are returned to the milling circuits.

• Classifying cyclones and pre-leach thickener. The thickener underflow feeds a transfer

vessel together with the secondary cyclone underflow where cyanide is added before the

slurry is transferred to the CIL circuit. Oxygen is injected into the transfer line after the

transfer pumps.

• A counter current CIL circuit. consisting of six stages of agitated vessel each of 2500 m3,

gives an overall residence time of 18-20 hours at a 7,4000 tpd mill capacity. Hydrogen

peroxide is added periodically to CIL tank 1 to maintain the dissolved oxygen level.

Activated carbon is retained in each tank using interstage basket screens and is moved

June 2019


counter-current to the slurry flow using submerged vertical spindle pumps in each tank.

Loaded carbon is recovered from the first CIL stage.

• Loaded carbon is acid washed and then stripped of gold using caustic soda in an 11.5 t

pressure Zadra elution system with the gold electrowon onto steel mesh before smelting.

• Eluted carbon is thermally regenerated and returned to the last stage of the CIL circuit.

• The gravity gold concentrate and electrowon gold are smelted separately to produce gold

doré bars.

• Additional supporting facilities include:

• Two, 2.1 tpd capacity pressure swing absorption oxygen plant located in the milling

area of; and

• emergency diesel powered generators.

The key plant design and operating parameters are shown in Table 17-1 and a schematic flowsheet

for the Wassa plant is presented In Table 17-1. The schematic incorporates the new densifying

cyclone and thickening circuit currently being installed.

The Wassa process operation achieved compliance with the International Cyanide Management

Code in early 2010 and was recertified in 2017.

Table 17-1 Key Plant Design and Operating Parameters

Blended Feed Primary Ore Feed (Project Design) (Current Operations)

Nominal throughput Mtpa 3.5 2.65

Crushing Circuit Product % passing 80%<6 mm 80%< 8mm

Crushing Circuit Utilisation % 85 75

Plant Design Availability % 93 92

Mill product grind % passing 80%<85 micron 70%<75 micron

CIL Feed Density

Design / Current % Solids 44 (cyclone overflow) 40 (CIL tanks - measured)

With new thickener % Solids 44-46

CIL Retention Time (calculated) h (total) 19-21 23

With new thickener 25

June 2019


Figure 17-1 Current Wassa plant flowsheet

June 2019


17.1.1 Plant Accounting

Plant throughput is reported based on the belt weighers installed on the conveyors feeding the two

ball mills from the crushed ore stockpile. There is also a belt weigher installed on the crushing

circuit product to the crushed ore stockpile.

Plant performance and accounting is assessed based on samples of feed and tailings taken

automatically using inline slurry samplers, which are composited into 12 h shift accounting

samples. The feed sample is taken after the milling and gravity circuit before transfer to the CIL

circuit and the gold recovered by gravity and smelted separately is added to calculate the plant

feed grade. The feed and tail slurry samples are analysed using bottle roll laboratory tests to assess

the BLEG tests.

The slurry samples are filtered and washed and the solids are pulverised to nominally 95% <75 µm

before being subject to BLEG bottle roll extraction. The BLEG tests are run for 8 h at high cyanide

concentration and the solutions from the filtration of the slurry samples and from the BLEG tests

are analysed by gold extraction into an organic phase and then measured by AAS. Extended BLEG

tests are also undertaken to confirm that all the recoverable gold has been extracted during the

standard BLEG leach period. The BLEG tails are periodically fired assayed to determine residual

gold in the samples not recovered in the BLEG tests (gold potentially locked in silica, pyrite or

other sulphide minerals) and a BLEG factor is determined to be used in assessment of the total

gold in the plant tails to determine the overall plant gold recoveries.

It was reported that previous attempts to use fire assay for solid sample analysis for gold used for

plant accounting have been unsuccessful due to the limited size of the sample analysed and the

quantity of relatively coarse gold suspected to be reporting in the plant feed. As such, the BLEG

analysis procedure is seen to be more dependable. BLEG, however, will only determine the amount

of cyanide recoverable gold present in each sample and does not allow for gold locked in sulphides

or other gangue minerals to measure the total gold present in the samples. A general BLEG factor

is therefore applied to determine the plant tails and overall gold recovery, although a factor is

reportedly not applied to the feed grade.

The gold recovered by gravity is smelted separately and this is added to the gold in the mill product

sample to determine the gold grade in the feed. A sample is taken of crushed ore from the feed to

the ball mills and this is used as a check measurement on the plant feed grade although is not used

for accounting purposes.

Reconciliation is undertaken on a monthly basis between the gold produced and the gold present

in the feed and tails. This also considers the changing gold inventory on the plant from month start

to month end. Based on the reconciliation the reported head grade is adjusted to correlate with the

monthly gold production.

17.2 Historical Plant Production

Production statistics for the CIL operation since 2007 are shown in Table 17-2. Monthly operating

data from the Wassa plant during 2014, while the plant has been processing mainly Wassa primary

open pit material constituting 85 to 100% of the total plant feed, shows that gold recoveries have

ranged, on a monthly basis, from 90.4 to 94.2%, averaging 92.7%, for head grades varying between

1.07 and 1.83 g/t Au (average 1.41 g/t Au).

June 2019


Table 17-2 Overview of Historic Plant Performance

Item Unit 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Ore - Open Pit kt 2,824 2,845 2,506 2,434 2,391 2,499 2,549 2,533 2,495 2,567 1,926 526

g/t Au 1.27 1.53 2.78 2.36 2.04 2.09 2.4 1.45 1.46 1.30 1.27 0.77

Ore - Underground kt 56 691 1,075

g/t Au 2.51 3.03 4.17

Heap Leach kt 928 324 147 214 188 7.7 146 96.4

g/t Au 0.64 0.3 0.73 0.59 0.39 0.24 0.3 0.3

Total Feed kt 3,752 3,187 2,653 2,648 2,579 2,507 2,695 2,629 2,495 2,623 2,617 1,600

g/t Au 1.17 1.4 2.67 2.22 1.92 2.09 2.29 1.41 1.46 1.32 1.73 3.06

Recovery

Gravity % 19.3 19.1 22.7 26.5 27 49.6 54.4 35.4 22.6 24.0 22.1 27.0

CIL % 72.8 74.8 72.4 68.3 67.3 45 45.6 57.3 70.8 69.6 71.7 73.0

Total % 92.1 93.9 95.1 94.8 94.3 94.6 94.5 92.7 93.4 93.6 93.8 95.7

Au Produced oz 126,059 125,427 223,848 183,931 160,616 160,917 183,788 112,836 108,266 104,381 137,234 149,697

June 2019


17.3 Future Plant Production

17.3.1 Future production

The process plant plan over the LoM has underground ore feed from 2019 to 2024. From 2023

until the end of the mining of the Reserve in 2028, open pit ore is available.

The current installed process plant capacity is 7,400 tpd. From 2019 to 2023, with predominantly

underground ore feed to the process plant, the throughput averages 4,000 tpd. The current milling

circuit comprises two identical parallel mills. Overall, mill utilisation will be managed in line with

throughput to obtain the grind size and the remaining downstream process will continue to operate

unchanged. This will result in an increase in residence time in leach that will translate to improved

recoveries as well as opportunities to optimize the grind size through the crushing and

comminution circuits.

From 2024 onwards, process plant throughput increases to 6,300 tpd with the re-introduction of

open pit ore.

June 2019


18 Infrastructure

18.1 Site Layout

The current site layout for the Wassa mine is provided below in Figure 18-1 and shows the existing

location for the following mining areas and major infrastructure:

• main roads, towns and power lines;

• open pits and waste storage areas;

• processing facilities;

• TSFs; and

• site accommodation.

Infrastructure specific to the underground mine comprises an electrical substation, office and

workshop areas as shown in Figure 18-2.

18.2 Electrical Infrastructure

18.2.1 Surface

Grid power from the national power supplier GridCo comes from a 161 kV line to local substation

where power is transformed down through a 33 MVA transformer to 34.5 kV. Two feeders

provide electricity to the GSR Wassa Mine substation at 34.5 kV.

The GSR substation comprises a duty 16 MVA transformer and a parallel, standby 18 MVA unit.

The substation reduces voltage to 6.6 kV and is reticulated across site, including the primary load

being the process plant. The existing full load power draw is around 12 MVA.

A dedicated switch at the GridCo substation and an underground cable feeds the underground

electrical substation at 34.5 kV. The underground substation provides medium voltage (MV)

distribution to the Starter Pit portal and underground workings. The substation comprises

switchgear, a 5 MVA transformer that steps power down from 34.5 kV to 6.6 kV for distribution,

and for backup, two 2000 kVA 400 V diesel generators and step-up transformer (400V/6.6 kV)

combination. The location of the underground electrical substation is shown in Figure 18-2.

From the underground electrical substation, power is distributed at 6.6 kV to end users

underground, at workshops and offices, and locally stepped down to use as required at 1000 V,

415 V and 240 V. Spare switches in the underground electrical substation are available for future

requirements.

18.2.2 Underground

The underground electrical system has been designed and installed according to Ghanaian mining

regulations and to efficient mining standards and will have high availability, medium utilisation

and low operating maintenance. The primary high voltage is 6.6 kV for reticulation and 1 kV for

low voltage.

June 2019


Figure 18-1 Wassa site layout

June 2019


Figure 18-2 UG Electrical substation, office and workshop area

Distribution underground is from the underground electrical substation. Two 95 mm2 XLPE

cables feed a ring main unit at the Starter Pit portal area, where power feeds surface infrastructure

locally and also continues the primary MV feed underground. A 1,500 kVA mini-substation is

located at the Starter Pit portal area stepping power down to 1000 V for equipment and 240 V for

lighting and small power.

From the ring main unit at the Starter Pit portal, electrical is fed underground via two 70 mm2 HV

SWA XLPE PVC (high voltage, steel wire armoured, cross linked polyethylene, poly vinyl

chloride) cables to a number of 1,500 kVA mini-substations. The cable route from the surface

Starter Pit portal area is via the main decline from the portal and the return airway for an 800 m

length to the first 1,500 kVA mini-sub-station at 820 mRL and then a combination of bore holes

and horizontal transfers to the bottom of the mine. Each mini-substation contains a 6.6 kV/1000V

transformer and a 6-way distribution board servicing the equipment load locally.

The key electrical loads installed are primarily for the ventilation fans, air compressors, dewatering

pumps and the electric-hydraulic drills. The mining sequence by panel typically comprises 5-6

sub-levels developed from top to bottom. The sub-levels will be in the fully (electrically) loaded

scenario during development, reducing in load as development progresses towards the base of the

panel and prior to commencement of the bottom up stoping sequence. As such, the operating power

load profile will quickly rise to 100 % before dropping off to 20-40% and then back to 100% over

a 2-year period (or panel life cycle). This results in a much higher installed electrical capacity than

the typical operating draw at any one time.

June 2019


One 1,500 kVA substation (with 6.6 kV to 1 kV transformer) is installed every 3-4 sub-levels for

local distribution and will continue as development and mining progresses at depth. These

substations contain a transformer and 6-way distribution board. The current and future mini-

substation locations are shown in Figure 18-3.

Figure 18-3 Current and future primary reticulation installations

The primary voltage for all underground fixed and mobile equipment is 1 kV, allowing longer

cable runs before voltage drop limits use. The low voltage is taken off the mini-substations through

a distribution board (including switches/circuit breakers) enabling feed to 6 end users (other

boards, equipment starters, etc). Depending on the end user and operating load, cable distribution

from the mini-substation locally will be 70 mm2, 35 mm2 or 10 mm2 cable. Trailing cables are

used from each jumbo starter in the footwall drives to reach the faces of the ore and are 35 mm2

(241.1 class) in size.

The current installed load is 4.2 MVA, drawing a maximum of 3.6 MVA in operation. Additional

loads are planned, primarily with ventilation, pumping and paste backfill, that will increase to a

maximum operating draw of 5.9 MVA by the end of 2020, beyond the current installed capacity

of the underground electrical reticulation. An electrical upgrade is in progress that will provide an

additional 8 MVA capacity and will be reticulated underground via a borehole installation at 11kV.

The underground operating loads are presented in Table 18-1.

June 2019


Table 18-1 Current and future underground loads

Equipment details

2019 Current

2020 2021 2022

Operating load kW

Operating load kW

Operating load kW

Operating load kW

Ventilation - primary 618 1178 1178 3378

Ventilation - sublevel 802 932 1007 727

Compressed air 180 270 270 270

Pumping - primary 367 999 966 1013

Pumping - sublevel 65 90 90 90

Drilling - production 481 540 540 540

Drilling - definition and exploration 332 548 692 692

Paste backfill 0 0 1125 1125

KW TOTAL 2845 4556 5867 7835

KVA TOTAL 3556 5696 7334 9794

18.3 Mine Services

18.3.1 Compressed Air

As a result of the mechanised nature of the operation, there will be limited requirements for

compressed air. The main uses include:

• use with ANFO charging kettles;

• occasional hand-held raise driving;

• connection to mine refuge chambers; and

• usage in the future underground maintenance workshop.

The compressed air system comprises 2 x 90 kW compressors located on surface at the Starter Pit

portal. Compressed air is distributed underground via a 110 mm poly pipe down the main decline.

Due to pressure drop along the reticulation and incremental increases in duty an additional

compressor is planned.

18.3.2 Service Water

A 30,000 liter water tank is installed above the portal area to supply the underground mine with

service water for drilling, dust suppression and general use. The service water tank is filled using

the 90 kW Flygt pump that is permanently installed in the Starter pit sump. Service water is

reticulated throughout the mine by 110 mm HDPE lines installed in the primary headings and

reducing to 63 mm HDPE for supply to end use locations.

June 2019


18.4 Dewatering

18.4.1 Underground dewatering

The underground mine dewatering system is designed and installed to remove both ground water

and service water (collectively called mine water), including up to 10% by volume silt or sand

sized particles. Initial settling occurs underground in sumps excavated on production levels as

mine development continues. The dewatering system comprises sub-level pumps that are located

on the operating areas that direct water to the main pump system that dewaters to surface.

The sublevel pump system is located on the production and development levels. These are electric

Flygt pumps varying in size and typically 37kW for the production levels and 18 kW for the face

pump supporting the decline development. These pumps direct water to the main dewatering

system.

The main dewatering system currently utilizes a number of pumps that transfer water several levels

150 mm steel rising main pipelines to the Starter Pit sump at the portal. The current system

installation is shown in Figure 18-4 and typically dewaters to the starter Pit sump at a rate of 35

L/s and, where required, up to 65 L/s.

Figure 18-4 Current dewatering long section (excluding F-Shoot for clarity)

A permanent pump station will be installed at the 620 mRL level. The purpose of this pump station

is to handle the majority of the dewatering from underground and pump directly to the surface via

a borehole and be directed to existing settling and discharge routes. Settling of solids will be done

underground via settling sumps close to the pump station. The design of the pump station will

utilize multistage pumps due to a combination of both total dynamic head and flow rate required.

A borehole will be constructed connecting the pump station to surface and a 200 mm NB steel

rising main installed.

June 2019


As the mine progresses at depth beyond the 620 mRL level, additional pumps will be utilized and

directed to the 620 mRL pump station. These pumps will be similar to existing pumps with

sublevel 37 kW Flygt pump and for main dewatering at depth a 90 kW Flygt pump and, if required,

supported by a 55 kW Mono pump. A reduced dewatering system above the 620 mRL pump

station will remain in place to intersect inflow at higher levels and dewater to the Starter Pit sump

on surface. The final dewatering system is shown in Figure 18-5.

Figure 18-5 Final dewatering long section

18.4.2 Pit dewatering

The water inflow into the pits is a combination of rainfall and groundwater, with rainfall being the

predominant inflow. The design of the storm water collection sump for each pit has been

conservatively designed for a 1:100 year, 241 mm 24 hour duration event. Catchment

modifications have also been completed to reduce inflows and capacity requirements for in pit

sumps. The catchment for each pit is shown in Figure 18-6. The dewatering discharge from the

pits is reused (at process plant, dust suppression, etc.) and directed to existing settling and drainage

systems for release.

The Starter Pit via the portal has direct connectivity to the underground. A storm water collection

sump below the underground portal entrance is installed and is 115% of the capacity required to

store the water from the 1:100 year event. To ensure operating sump levels are minimized (and

available storm water volume maximized) a 90 kW Flygt submersible pump (45 L/s) and 160 mm

HDPA pipe is installed. A larger Pioneer diesel pump (165L/s) and two 160 mm HDPE pipes

installed as a standby pump and operates in rainfall events that are beyond the capacity of the

submersible pump.

Like the Starter Pit, the B-Shoot pit will have direct connectivity with the underground via the

Portal 3 and the future vent raises and rock pass at the 844 mRL level. A storm water collection

June 2019


sump below the underground portal entrance is installed and is 200% of the capacity required to

store the water from the 1:100 year event. A Pioneer diesel pump (150 L/s) and two 160 mm

HDPE pipes installed as the duty with a future 90 kW Flygt submersible pump (30 L/s) are to be

installed such that a similar operating strategy to that of Starter Pit can be employed.

The 242 Pit and the South East Pit have significantly larger storage capacities in a 1:100 year event

– 360% and 240% respectively. These pits have permanent diesel pumps (Pioneer and Sykes

pumps) and pipe systems installed at the sumps. In addition to the installed pumps across all pits,

there are a further two standby pumps available for use; one Sykes diesel pump with a duty of

150L/s and one Pioneer diesel pump with a duty of 65 L/s.

Figure 18-6 Pit catchments

18.5 Workshops

The main surface workshop services primarily the open pit and surface fleet. The workshop is

equipped with overhead cranes, services, welding bay, tool storage and offices. The main diesel

fuel storage for the site is located in this area.

An underground specific workshop nearby the UG office is currently the primary location for

underground fleet maintenance. This workshop is fitted out with maintenance facilities, offices,

welding bay, oil-water separator, compressed air, electrical (1000 V) test facilities for the jumbos

June 2019


and tools storage. Some daily servicing of the underground fleet takes place at the Starter Pit portal

and this is equipped with a 1000 V test panel.

As the mine develops at depth, an underground workshop will be installed. The proposed

workshop will be at the 595 mRL sublevel and, once installed, will become the main location for

primary maintenance and servicing and will comprise similar facilities as the surface workshop.

The underground workshop will provide an effective area for quality maintenance and will reduce

the need and time for the underground fleet to travel to surface for maintenance.

18.6 Waste Disposal

The current and historical waste dumps are located adjacent to the Wassa Pit complex, with Waste

Dump 1 and the 419 Dump directly south of the Wassa Main Pit. Waste from the open pit and

underground operations have been hauled to the existing waste dump locations shown in Figure

18-7. The 419 Dump is currently active, contains 15.5 Mt and is at 1010 mRL on the eastern

section and 1040 mRL in elevation on the south-western section.

June 2019


Figure 18-7 Current waste dump locations and volumes

Table 18-2 Waste stockpile current capacities

Waste dump Current capacity - Mt

SAK Dump 12.3

Waste Dump 1 14.9

Mid-East 2 2.0

Waste Dump 2 7.0

419 Dump 15.5

DMH Waste Records not available

June 2019


The waste dumps have been designed with 10 m bench heights and 10 m berm widths. The dump

designs consider operational and rehabilitation phases. For operations, the dumping design has

bench slope angles of 37° (angle of repose) while the rehabilitation design has a bench slope angles

of 25°, resulting in overall batter slope angle of 22° after the final rehabilitation of the dump. The

waste dump slope design is shown in Figure 18-8.

The construction of the waste dumps contain the following features:

• Adequate drainage to ensure that any discharge from the waste dump is contained for

settlement and/or monitoring, to enable compliance with the EPA effluent discharge limits.

• The top surface of the dump, and any berms partway up the dump slopes, will be

constructed to shed water away from the surface of the dump.

• Water collecting drains will be constructed around the perimeter of the dump to route

discharges and runoffs into settlement and monitoring ponds.

Figure 18-8 Waste dump slope designs for operations and rehabilitation

The active 419 waste dump will continue as the main location for waste disposal from the

underground and surface operations when Wassa Main Pit Cut 3 commences. Waste rock from

the underground operation will be hauled through the existing haulage routes to the 419 waste

dump. The waste dump construction will continue to follow the existing slope designs and the

final LoM elevation for the 419 waste dump is planned at 1090 mRL as shown in Figure 18-9.

At the final expected height of 1090 mRL, the waste dump will contain in excess of 53 Mt and

will be able accommodate waste rock from the Wassa operations for the LoM.

June 2019


Figure 18-9 419 Waste dump location and elevation at LoM

18.7 Tailings Storage Facilities

18.7.1 Introduction

TSF 1 is located northwest of the plant site at the head of a southerly draining valley, immediately

adjacent to the historical leach pad area. Ground levels in the valley range from 1000 mRL on the

valley floor to more than 1060 mRL on the surrounding hills. The TSF 1 is a cross valley

impoundment created by the construction of a Main Embankment in the south with confining

Saddle Embankments at the north of the facility. Natural ridges provide containment at the east

and west of the storage area. Access to TSF 1 is via an unpaved access road west of the plant site

area. The catchment area of TSF 1 is estimated to be about 140 Ha, of which about 124 Ha will be

covered with tailings at the end of the facility design life. Deposition into TSF 1 will cease in 2019

with paddock deposition being completed to achieve the closure land form. Re-vegetation trials

commenced in 2017 towards the next land use.

TSF 2 is located in the valley system that trends eastwards to the immediate north of TSF 1 and

lies about 2.5 km from the CIL processing plant and some 1.25 km downstream of the TSF 1

Saddle Dam 5. The 260 Ha TSF 2 footprint (of which some 72 Ha have been developed to date)

is sited within a 340 Ha Project area (TSF 2 footprint plus nominal buffer).

An aerial view of the current TSF are shown in Figure 18-10.

June 2019


Figure 18-10 GSWL TSF 1, TSF 1 extension and TSF 2 Cell 1 (March 2018)

18.7.2 TSF History Overview

TSF1 was commissioned in August 2004 to meet the tailings storage needs for the original LoM.

Since starter embankment construction, the TSF’s embankments have undergone the originally

designed and permitted embankment raises, resulting in a TSF elevation of 1039 mRL.

As part of planning for the TSF 2 project, a number of alternatives were considered and, from

these, four feasible alternatives were identified: the construction of a new TSF at two different

locations; increasing the elevation of the existing facility to 1049.5 mRL; and increasing the

elevation of the existing facility whilst advancing plans for a new TSF. After a thorough study of

the alternatives, GSWL committed to constructing a new TSF, termed TSF 2. GSWL additionally

assessed alternatives relating to the scope of the RAP for the project, determining that the scope

of the RAP would encompass the entire Togbekrom community.

In compliance with the requirements of the EPA’s Environmental Assessments Regulations, 1999

(L.I. 1652), GSWL registered a new TSF project with the EPA in May 2010 and obtained

authorization to proceed to permitting in July 2010. GSWL submitted an Environmental Scoping

Report to the EPA in March 2011 and subsequently submitted an EIS for the construction and

operation of the proposed TSF 2. The EIS was approved by the EPA in April 2013 (EPA/EIA/383),

June 2019


and conditions of the EIA permit led to GSWL re-designing the TSF 2 facility to accommodate a

geomembrane liner.

While conducting the impact assessments and the preparation of the EIS, GSWL sought

permission to raise the TSF 1 by an additional 5 m and for continued deposition into the facility

between the period of August 2011 and May 2015. All the embankments were subsequently

constructed to the permitted elevation of 1039 mRL.

In March 2015, GSWL obtained an EPA permit to expand the TSF 1 into the then disused phase

1 HL pad area that was located directly to the east of TSF 1. The extension of the TSF 1 into this

16.2 Ha brownfields area provided an additional estimated storage capacity of some 2.09 Mt of

tailings using conventional deposition from embankment spigotting, and in excess of 2.17 Mt of

capacity, primarily through paddock deposition (spigotting from day walls) across the entire

TSF 1, to achieve the optimal drainage design, ahead of TSF 1 closure.

As a result of the delayed commencement of the TSF 2 project to enable the re-design of the facility

to accommodate geomembrane line, GSWL, in compliance with the requirements of the EPA

Environmental Assessment Regulations, 1999 (L.I. 1652) and Section 3.7 of the EPA Permit

(EPA/EIA/383), applied to the EPA in July 2014 for the renewal of the TSF 2 permit. Following

advice from the EPA in January 2015, GSWL then updated the EIS for TSF 2 with permit issuance

in January 2016 and an effective date of November 2015 (EPA/EIA/442). The development of

TSF 2 necessitated the resettlement of some 105 households within the Togbekrom and

surrounding hamlets to New Ateiku, approximately 10 km away. All the project affected people

were successfully relocated to their new homes in Q1 of 2013.

The TSF 2 has a current designed storage capacity of some 41 Mt of tailings, which under an

annual throughput of 2.7 Mtpa, would provide storage of tailings for some 15 years. The TSF 2

will be constructed in three cells in the following order: Cell 1, Cell 2 and Cell 3, and staged from

Stage 1 to 11.

At the time of permit renewal, the TSF 2 design had been revised to a cellular arrangement with

lining of the entire basin with HDPE geomembrane. However, in February 2016, the Mines

Inspectorate Division of the Minerals Commission directed that, as per the Minerals and Mining

Regulations, 2012 (LI 2182), the TSF 2 design be constructed with a clay liner. As GSWL was

well advanced in regards to TSF 2 construction preparations at that stage, following a series of

meetings with the Chief Inspector of Mines and formal submissions in that regard, GSWL was

given dispensation for the HDPE lining of TSF 2 Cell 1, with all future cells and stage raises to

incorporate a compacted soil liner. The construction of TSF 2 Cell 1 subsequently commenced in

July 2016. The verbal approval from the Inspectorate Division for the commencement of

deposition into TSF 2 was given in February 2017 with EPA approval in April 2017 and deposition

commencement in May 2017.

The re-design of TSF 2 with compacted soil liner was submitted to the Minerals Commission

Mines Inspectorate Division in January 2017. In July 2017, the Mines Inspectorate Division had

completed their review and recommended the design to the Chief Inspector of Mines for approval.

In 2017, in recognition of the lead time for permitting, GSWL commenced an EIA to support the

submission of a Supplementary EIS for the TSF 2 Cell 2 construction. The TSF 2 Cell 2

Supplementary EIS was submitted to the EPA in October 2018 and reflected the modified

compacted soil liner design.

June 2019


Figure 18-11 TSF1 and TSF2 layout as per Knight Piésold report

June 2019


18.8 Tailings Storage Facility 2

18.8.1 Geotechnical characterization

A detailed geotechnical investigation comprising sub-soil, in-situ and laboratory testing of soils of

the TSF 2 basin was carried out by Knight Piésold Consulting Ltd. (Knight Piésold 2011) using

test pitting, cable percussion drilling, standard penetration testing, permeability testing, moisture

content, grading, Atterberg Limits tests, consolidation tests, triaxial testing on undisturbed soil

samples and falling head permeability tests. The results established the soil profile of the basin,

the strength of the foundation soils, and the permeability of the different soil types, to inform in

the design of the TSF embankments, base and environmental protection features.

The TSF 2 basin is characterized by a rugged and dissected ground profile that defines the soil

profiles in the area according to topographical location. Two main soil types are found in the TSF

footprint. They are alluvial soils formed from the deposition of eroded materials from the

surrounding hills, and residual soils formed in-situ from the chemical weathering of the underlying

base rocks. Soils can be classified as either: high ground and side-slope soils that are found along

slopes and crests of hills, plateau and other high ground that characterizes the TSF footprint; or as

basin valley and embankment foundation soils that dominate the valleys and low-lying areas.

Guelph permeability tests conducted on nearby surface soils in the valley floor indicated that in

some areas the soils have very low permeability (lower than 1.0 x 10-8 m/s). In-situ falling head

permeability tests showed that the residual soils, at depths greater than 1.0 m, have a relatively

high permeability. Laboratory falling head permeability tests corroborated the field studies and

showed that in the valley floor, very low permeability strata exists to around 1.0 m depth.

18.8.2 TSF 2 Design

The TSF 2 design comprises three cells separated by embankments, a temporary embankment and

a series of perimeter saddle dams, providing primary containment to ensure that tailings are

contained within the valley basin. Other key environmental protection features include a

combination of geomembrane and/or compacted soil liner, as well as spillway, decant barge,

secondary confinement, ground water drains, and basin under-drains which are incorporated in the

design to enable efficient and appropriate water management for the TSF.

The TSF 2 design was based on an annual throughput of 2.7 Mt. The facility is designed for a

storm capacity of a 1:100 year, 24 hour duration event with allowance for wave run-up and no

flow through the spillway; and to safe discharge the flow of a 1:1000 year, 24 hour duration storm

event.

The design of the TSF 2 meets the requirements of the Minerals and Mining (Health, Safety and

Technical) Regulations, 2012 (L.I. 2182) and takes due consideration of the recommendations of

the International Committee on Large Dams (“ICOLD”) (ICOLD various), the Australian

Committee on Large Dams (1999) and the Canadian Dam Association guidelines (2007).

TSF 2 is being constructed in stages and stage storage capacities are presented in Table 18-3.

Alternative stage raises are under evaluation to facilitate annual raising during suitable

construction weather conditions (Knight Piésold 2017).

June 2019


18.8.3 Stability Analysis

Stability analyses were conducted for static and seismic loading conditions and static post

liquefaction conditions for critical embankments and stages using SLOPE/W® and the

Morgenstern-Price method of analysis, which considers force and moments equilibrium of circular

slips.

A conservative peak seismic design horizontal ground acceleration of 0.1 g, obtained from

“Seismicity of Southern Ghana: Causes, Engineering Implications and Mitigation Strategies” by

N.K. Kumapley (1996), was employed in the pseudo static analyses.

For the stability analyses on the upstream slopes, the worst case scenario was considered, where

no tailings are present in front of each embankment stage. For the stability analyses of the

downstream slopes, the worst case scenario was also considered, where the TSF was full to

capacity in front of each stage raise, i.e. 1 m below crest. Modelling scenarios assessed drained

and undrained conditions and modelled for worst case phreatic conditions – making the analysis

highly conservative in nature.

The minimum Factor of Safety (“FOS”) values calculated for all conditions on both the

downstream and upstream slopes were found to meet, and in some conditions exceed the Minerals

and Mining (Health, Safety and Technical) Regulations, 2012 (L.I. 2182) requirements for factors

of safety.

Stability of the facility was also assessed under the condition where, following the design seismic

event, the tailings may be subjected to liquefaction. Seismic stability assessment of the various

embankments was conducted in the undrained condition for upstream failure and static drained

condition for downstream failure. Tailings were modelled with a residual post-liquefied undrained

strength but with no earthquake loading. The minimum FOS values calculated for the post-

liquefied condition of the downstream and upstream slopes meet and, in some conditions exceed,

the regulatory requirements.

June 2019


Table 18-3 TSF stage storage capacities

Stage Cell Crest Level

(m RL)

Beach Level (m RL)

Embankment Height (m)

In-storage Density (t/m³)

Storage capacity

(Mt)

Cum. Capacity

(Mt)

Duration of Stage

Deposition (months)*

Cumulative Period of

Deposition (months)

Rate of Rise (m/month)

Rate of Rise

(m/year)

Inundation Area (ha)

1 1 1011.5 1010.5 18.5 1.10 3.24 3.24 14.4 14.4 1.3 15.0 34.6

2 1 1018.5 1017.5 25.5 1.10 3.63 6.87 16.1 30.5 0.4 5.2 46.0

3 1 1023.0 1022.0 30.0 1.10 3.14 10.01 14.0 44.5 0.3 3.9 61.5

4 2 1010.0 1009.0 18.5 1.10 3.34 13.35 14.8 59.3 1.2 14.6 37.8

5 3 1001.0 1000.0 14.0 1.10 3.01 16.36 13.4 72.7 1.0 12.1 36.5

6 3 1007.8 1006.8 20.8 1.10 2.9 19.26 12.9 85.6 0.5 6.3 37.3

7 2+3 1012.5 1011.5 25.5 1.25 3.35 22.61 14.9 100.5 0.3 3.8 71.0

8 2+3 1015.0 1014.0 28.0 1.29 3.34 25.94 14.8 115.3 0.2 2.0 98.8

9 2+3 1017.4 1016.4 30.4 1.33 3.52 29.46 15.6 131.0 0.2 1.8 110.5

10 2+3 1020.0 1019.0 33.0 1.37 3.64 33.10 16.2 147.2 0.2 1.9 111.2

11** 2+3 1023.0 1022.0 37.0 1.40 7.86 40.96 34.9 182.1 0.1 1.0 133.5

* Duration is based on design throughput of 2.7 Mtpa and design densities Alternative stage raise options are under evaluation to facilitate annual dry season raising (Knight Piésold 2017).

June 2019


18.8.4 Water Balance

Water balance modelling was completed using GoldSim® software using probabilistic Monte

Carlo procedures. The modelling accounts for a full range of possible climatic conditions.

The TSF 2 water balance model illustrated that, for the first 7 years of operation of the TSF 2, the

facility operates under a net negative water balance. Additional raw water for processing will be

obtained from TSF 1 or mine dewatering.

Water balance modelling and sensitivity analysis showed the importance of diverting water away

from Cell 3 in the later stages of TSF 2. The analysis showed that run-off diversion is an

appropriate method for management of potential surplus water conditions.

18.8.5 Seepage Modelling

Seepage analyses were performed to estimate the potential steady-state seepage rates, distribution

of total head throughout the embankment and its foundation, and the phreatic surface in the

embankment using the finite element software SEEP/W ® 2007.

Modelling was conducted for the embankment at Stage 3, where the driving head is considered to

be the most critical. The pond was modelled at a conservative minimum distance of 100 m from

the embankment upstream face. The downstream face and the ground surface extending beyond

the downstream toe were modelled as potential seepage boundaries. The liner on the basin and the

slope areas of the facility was included in the model.

The results suggest that 0.00214 L/s (2.14e-006 m3/s) and 0.00052 l/s (5.2e-007 m3/s) of seepage

flow may be recorded per meter run of wall for the first and second scenarios respectively. This

equates to a maximum flow rate of 15 m3/hr for the full length of the Cell 1 secondary confinement

channel.

June 2019


19 Market Studies and Contracts

19.1 Market Studies

Gold is a freely traded commodity on the world market. The World Gold Council, in its gold

demand trends report (WGC 2015), stated that the global gold market is in an overall demand-

supply balance. Top-line demand was broadly neutral despite substantial underlying differences

across geographies and sectors among jewellery, technology, investment, central banks, and

institutions. Total global supply was little changed year-on-year as lower recycling offset growth

in mine supply.

For the Wassa Mine, all gold produced is shipped to a South African gold refinery in accordance

with a long-term sales contract currently in place for GSR. The gold is shipped in the form of doré

bars, which average approximately 90% gold by weight with the remaining portion being silver

and other metals. The sale price is based on the London p.m. fix on the day of the shipment to the

refinery.

19.2 Contracts

The following contracts will be part of the rehabilitation, mine development, and operations of the

Wassa Mine:

• Long-term doré bar sales contract is in-place with a South Africa gold refinery

• A general mining explosive supply agreement is in-place with AEL Mining Services, a

South Africa based mining explosive supplier, for GSR’s Ghana mine operations.

Explosive price is subject to monthly adjustments based on raw material cost changes.

June 2019


20 Environmental Studies, Permitting and Social or Community Impact

20.1 Relevant Legislation and Required Approvals

Act 703 is the governing legislation for Ghana’s minerals and mining sector. It requires that mines

obtain environmental approvals from relevant environmental agencies as outlined in Table 19-1.

Ghanaian environmental legislation is well developed and is enforced by the EPA.

20.1.1 Environmental Assessment Requirements

The overarching Act that regulates the environmental regime of Ghana is the EPA Act, 1994

(“Act 490”). The main legal framework used by the EPA for regulating and monitoring mineral

operations is the Environmental Assessment Regulations, Legal Instrument 1652 of 1999

(LI 1652). These regulations cover requirements for environmental permitting, EIA, the

production of preliminary environmental reports (“PERs”) and subsequent EIS, environmental

certificates, EMPs and reclamation bonding.

The EPA grants environmental approval to projects, in the form of an Environmental Permit. The

decision on whether or not to grant the permit is based on the findings of an EIA, which also covers

social aspects and is documented in an EIS. For a mine, an EIS must include a reclamation plan

(Regulation 14 of LI 1652) and a provisional EMP. The EIS may be subject to a public exhibition

period and public hearing before formal review by the EPA. Responses of regulators and

community obtained through these processes are redirected to the proponent for incorporation into

the Final EIS, before an Environmental Permit is granted.

Within 24 months of receipt of an Environmental Permit, mines are required to obtain an

Environmental Certificate from the EPA (Regulation 22 of LI 1652). The Environmental

Certificate is a follow-up mechanism that confirms the commencement of operations; acquisition

of other permits and approvals, where applicable; compliance with mitigation commitments made

in the EIS or EMP; and submission of annual environmental reports to the EPA.

An EMP must be submitted within 18 months of commencement of operations and must be

approved by the EPA. A provisional EMP is typically provided in an EIS, with the expectation

that the new project EMP would be incorporated into the mine’s overarching EMP when it is

updated. Mines are required to update their EMPs every three years and must submit the updated

EMPs to the EPA for approval (Regulation 24 of LI 1652).

All mines in Ghana are required to have a reclamation plan (Regulation 14 of LI 1652). In addition,

mining operations must submit annual environmental reports (Regulation 25 of LI 1652) and

monthly environmental returns of the environmental parameters monitored to EPA. Comments are

also expected in cases where monitored values exceed limits and, as appropriate, a project is to

provide the measures to prevent further occurrences.

June 2019


Table 20-1 Primary Environmental Approvals Required for Mining Operations

Regulatory institution Approvals that have to be obtained Reporting, inspections and

enforcement

Environmental

Protection Agency

Established under the

Environmental Protection

Agency Act, 1994 (Act

490), the EPA is

responsible for among

other things, the

enforcement of

environmental regulations.

Environmental Permit

In accordance with Section 18 of the Mining Act, 2006 (Act 703), and the

Environmental Assessment Regulations, 1999 (LI 1652), of the EPA, a

holder of a mineral right requires an Environmental Permit from the EPA

in order to undertake any mineral operations.

Approved Environmental Management Plan

An EMP must be submitted within 18 months of commencement of

operations and updated every three years (Regulation 24 of LI 1652).

Environmental Certificate

This must be obtained from the EPA within 24 months of commencement

of an approved undertaking (Regulation 22 of LI 1652).

Approved reclamation plan

Mine closure and decommissioning plans have to be prepared and

approved by the EPA (Regulation 14 of LI 1652).

Reclamation bond

Mines must post a reclamation bond based on an approved reclamation

plan (Regulation 22 of LI 1652).

Reporting

Mines must submit monthly

returns and annual environmental

reports to the EPA.

Inspections

The EPA undertakes regular

inspections to ensure that mineral

right holders are compliant with

permit conditions and the

environmental laws generally.

Enforcement

The EPA is empowered to

suspend, cancel or revoke an

Environmental Permit or

certificate and/or even prosecute

offenders when there is a breach.

Minerals Commission

and Mines Inspectorate

Division


Minerals and Mining Act,

2006 (Act 703), the

Minerals Commission

administrate mineral rights

in trust for the people of

Ghana.

Exploration and mining operating plans

A holder of a licence shall not commence operations unless an Operating

Permit is issued by the Inspectorate Division for the operations.

Modifications to operating plans are required to be approved by the Chief

Inspector of Mines.

Emergency response plan

An emergency response plan must be submitted to the Inspectorate

Division for approval.

Resettlement plan

LI 2175 defines specific requirements for compensation and resettlement,

including approval of resettlement plans by the district planning

authority.

Closure Plan

Regulations 273 to 277 provided detailed requirements for closure

requirements and plans.

Other

An array of other permits and licences (e.g. explosives) are required to be

obtained in support of mining operations, which incorporate

environmental and social requirements.

Reporting

Mines must submit monthly and

quarterly returns.

Inspections

The Mines Inspectorate

undertakes regular inspections to

ensure that mineral right holders

are compliant with regulations

and laws generally.

Enforcement

Regulations 21 and 22 allow the

Mines Inspectorate to issue

improvement and/or prohibition

notices for contraventions of the

Regulations.

Water Resources

Commission (“WRC”)


Water Resources

Commission Act, 1996

(Act 522), the WRC is

responsible for the

regulation and

management of the use of

water resources.

Approvals for water usage

Under Section 17 of the Mining Act, 2006 (Act 703), a holder of a

mineral right may obtain, divert, impound, convey and use water from a

watercourse or underground reservoir on the land of the subject of the

mineral right, subject to obtaining the requisite approvals under Act 522.

The Water Use Regulations, 2001 (LI 1692), regulate and monitor the use

of water.

Reporting

Holders of water use permits

must submit quarterly and annual

reports to the Water Resources

Commission.

Inspection

The WRC has power to inspect

works and ascertain the amount

of water abstracted.

Enforcement

Both Act 522 and L.I. 1692

prescribe sanctions for breaches.

Forestry Commission and

Forestry Services

Division

In accordance with Section 18 of the Mining Act, 2006 (Act 703), a

holder of a mining right must obtain necessary approvals from the

Forestry Commission.

June 2019


Guidelines and standards relevant to the mining industry have been made under Act 490. These

include the Mining and Environmental Guidelines (1994), which provide guidance on the contents

of EIS and EMP reports and of reclamation plans. They also include guidelines on EIA procedures,

effluent and emission standards, ambient quality and noise levels and economic instruments.

The EPA conducts routine monitoring of environmental parameters for mining operations and the

results obtained are cross-checked with the monthly return values submitted by operations and

compared relevant standards.

The EPA is empowered to suspend, cancel, or revoke Environmental Permits where the holder is

in breach of LI 1652, the permit conditions or the mitigation commitments in the EMP.

20.1.2 Minerals and Mining Requirements

Act 703 establishes laws on the process for obtaining mineral rights, and the administration and

management of these rights and for the protection of the environment. Supporting Act 703 are the

Minerals and Mining Regulations, 2012. These cover general aspects (LI 2173), matters relating

to compensation and resettlement (LI 2175), explosives (LI 2177), support services (LI 2174), and

health, safety and technical requirements (LI 2182). The regulations listed below have particular

relevance to environmental and social management:

• Minerals and Mining (Health, Safety and Technical) Regulations 2012 (LI 2182) – these

regulations define requirements for approval of mine closure plans, hazard classes for

TSF, and set requirements for embankment design, factors of safety, impoundments,

freeboard, discharge systems, safety arrangements, monitoring, planning, auditing and

closure.

• Mining General Regulations 2012 (LI 2173) – these promote preferential employment

of Ghanaians and preferential procurement of goods and services from Ghanaian service

providers. Mines are required to prepare localisation plans to achieve this and to submit

frequent reports (monthly, six-monthly and annual reports) that provide information on

Ghanaian and expatriate staff numbers as well as information on payments of salaries

and wages, royalty and corporate tax.

• Mines (Support Services) Regulations, 2012 (LI 2174) – these extend the requirement to

preferentially employ Ghanaians to providers of services to mines.

• Mines (Compensation & Resettlement) Regulations, 2012 (LI 2175) – these require that

displaced people are resettled to suitable alternative land and that their livelihoods and

living standards are improved. The resettlement plan must be approved by the district

planning authority and then given effect by the Minister responsible for Mines.

GSWL has a localisation plan that has been approved by the Minerals Commission that covers

expatriate staff and is in full compliance with the regulation requirements.

GSR is listed on the Ghana stock exchange and continues to submit its annual financial reports as

required by the law.

20.1.3 Water Resources Legislation Requirements

The Water Resources Commission Act, 1996 (Act 552) establishes the Water Resources

Commission (“WRC”) and sets requirements regulating the use of water resources. The Water

June 2019


Use Regulations, 2001 (LI 1692), and Drilling Licence and Groundwater Development

Regulations, 2006 (LI 1827), complement the Act by specifying: the requirements for obtaining

permits for water use, water rights, and priorities for water use; and water drilling licences, and

well construction requirements; respectively.

A summary of environmental approvals held by GSWL is provided in Table 20-2.

20.1.4 Overview of Permitting of Existing Operations

Environmental approval for development of the Wassa operations, including the original extent of

the Main pits, was obtained based on an EIS developed by SGS for SGL in 1998. The Main pits

complex comprised the interconnected South East, 242, F and B Shoots, South and Main South,

and 419 pits.

In September 2002, Golden Star purchased the fixed assets of the project and the Wassa operations

recommenced under WGL, with 90% ownership by Golden Star 90% and 10% ownership by the

Government of Ghana.

In 2004, the operations were expanded and converted to a conventional CIL process via the WGL

Wassa EIS. The South Akyempim pits were permitted in 2006. Expansions to the Main pits

complex (with cutbacks to 242, South and Main South, F and B-shoots) were later permitted in

2010 through the subsequent GSWL (Wassa) Pits Expansion EIS.

In late 2005, Golden Star acquired SJR (Ghana) Limited and with it, the Hwini-Butre and Benso

properties. These previously operated satellite projects were expanded in 2007 with the Hwini-

Butre and Benso (HBB) EIS. The G-Zone waste rock dump (Benso) was later permitted for

expansion via EIS in 2010.

The original TSF was permitted in 2004 as part of the original WGL Wassa EIS. Stage raises to

1035.5 mRL, 1037 mRL and 1039 mRL were permitted in 2011, 2012 and 2013, respectively. In

2015, an extension to TSF 1 was permitted and, in the same year, TSF 2 was also permitted. TSF

2 was subsequently re-permitted in 2016 following the facility redesign.

Underground exploration was permitted in 2015 and, in 2017, an expansion of the main Wassa

operations to incorporate underground mining and pit and waste dump expansions occurred.

June 2019


Table 20-2 Environmental Approvals Obtained for the Wassa Mine

Approval Permit No. Date of Issue Expiry Date Comments

Environmental Protection Agency

Approval of the Satellite Goldfields Limited Wassa project EIS

N/A 1998 There are no formal approval documents on record

EIA and EMP for Exploration in Subri River Forest Reserve

N/A 2004 There are no formal approval documents on record

Environmental Permit for the Wassa Power Project

Form D (0010335) 07/05/2004 N/A

Based on Volta River Authority Wexford Power Project 161 kV Power Transmission Line Bogoso to Akyempim Environmental Scoping Report (2003)

Environmental Permit to pursue operations

EPA / EIA/112 18/03/2004 N/A Based on Wexford Goldfields Limited Wassa project EIS (2004)

Hwini-Butre Permit

EPA/EIA/175 24/02/2006

N/A

St Jude Resources (Ghana) Limited based on Hwini-Butre EIS and Subriso EIS

Benso Subriso Permit

Detox Plant and Discharge to Kubekro Creek Approval

Letter 23/12/2005 N/A

South Akyempim Environmental Permit

EPA/EIA/190 02/06/2006 N/A Based on EIS on South Akyempim Project (2005)

Hwini-Butre/Benso Project Environmental Permit

EPA/EIA/247 02/10/2007 N/A Based on the Hwini-Butre and Benso EIS (2005)

Wassa Pits Expansion Project Environmental Permit

EPA/EIA/322 20/12/2010 N/A Based on Wassa Pits Expansion EIS (2010)

G-Zone Waste Rock Dump Environmental Permit

EPA/EIA/323 13/12/2010 N/A Based on Supplementary EIS for G-Zone Waste Dump (2010)

Environmental Certificate EPA/EMP/093 15/04/2011 N/A 2014-2017 renewal processed.

2018-2020 EMP submitted for renewal

Reclamation Bond 22/07/2011 N/A Applied for renewal

TSF 1 embankment raise to 1035.5 mRL

Letter 4/08/2011 N/A

TSF 1 embankment raise to 1037 mRL

Letter 9/05/2012 N/A

Environmental Permit for Mineral Exploration (Manso)

EPA/PR/PN/770 4/09/2012 3/09/2014 New permit not presently required

TSF 2 Permit EPA/EIA/383 5/04/2013 4/10/2014 Based on corresponding EIS (2013)

TSF 1 embankment raise to 1039 mRL

Letter 12/04/2013 N/A

Father Brown/Dabokrom Supplementary EIS

Letter Invoiced 14/01/2014

Based on Father Brown/Dabokrom Impact Prediction Study (2012)

TSF 1 extension Environmental Permit

EPA/EIA/419 13/03/2015 N/A Based on TSF 1 extension EIS (2014)

TSF 2 (re-design) Environmental Permit

EPA/EIA/442 25/11/2015 N/A Based on TSF 2 EIS (2015)

Wassa Underground Exploration Permit

EPA/PR/PN/929 3/07/2015 4/07/2017 Transition to EPA/EIA/508

Wassa Expansion Project Environmental Permit

EPAEPA/EIA/508 30/10/2017 N/A Based on Wassa Expansion EIS (2016)

Water Resources Commission

Permission to divert Adehesu creek at South Akyempim

N/A 06/12/2006 N/A

Water Use Permit Diversion of Ben and Subri Streams

N/A 27/03/2008 N/A

June 2019


20.1.5 Environmental Certificate and EMP for the Overall Operations

GSWL (then WGL) received its first Environmental Certificate for the period 21/09/2006 to

20/09/2009. On a routine basis since that time, GSWL has continued to submit its thre-year EMP

as required by regulations for Environmental Certificate permitting. The most recent certificate

renewal process was initiated with the submission of a new EMP to the EPA in December 2017.

Following review by the EPA, the Environmental Certificate was invoiced in June 2018. The EMP

has been finalized and resubmitted for Environmental Certificate issuance.

The Environmental Certificate and the EMP are for the overall Golden Star Wassa operations;

incorporating the Wassa operation, the suspended Hwini-Butre and Benso operations, and all

associated infrastructure, including the Hwini-Butre Benso access road. The existing

Environmental Certificate remains in force until such time as the new Environmental Certificate

is issued.

20.1.6 Notable Conditions of Approval

The Environmental Permit and the EIS’ for operations require compliance with applicable

legislation and that the Company must post a reclamation bond within one year of commencement

of operations. GSWL posted its initial reclamation bond in November 2004. The bond is updated

periodically to reflect approval of new/expansion projects. As at the end of 2018, the GSWL bond

was US$9,572,231.

The mining leases also contain conditions relevant to environmental management. The Wassa

Mining Lease (LVB 7618/94), Benso Mining Lease (LVB26871/07), and Hwini-Butre Mining

Lease (LVB1714/08) stipulate conditions for the encroachment of mining activities on community

infrastructure, the disturbance of vegetation, the conservation of resources, reclamation of land

and prevention of water pollution.

Approval Permit No. Date of Issue Expiry Date Comments

Water Use Permit (Akyempim) GSWLID134/1/17 01/01/2017 31/12/2019

Water Use Permit (dewater Wassa Main and Starter)

GSWLID134/2/17 01/01/2017 31/12/2019

Water Use Permit (bores and 242)

GSWLID212/17 01/01/2017 31/12/2019

Water Use Permit (C Zone fish cages)

GSWLID455/17 27/06/2017 26/06/2020

Water Use Permit (Mpohor) GSWLID212/19 01/01/2019 31/12/2021

Water Use Permit (Benso) GSWLID193/19 01/01/2019 31/12/2021

District Assembly

Togbekrom Resettlement Plan WEDA/DEV 15 9/01/2013 N/A Wassa East District Assembly

Awunakrom Resettlement Plan AWDA/DEV 21 4/03/2013 N/A Ahanta West District Assembly

June 2019


20.2 International Requirements

20.2.1 Environment and Conservation

The Government of Ghana is a party to a number of international treaties relating to the

environment, notably:

• Ramsar Convention on Wetlands of International Importance - there are five designated

Ramsar sites along the coast of Ghana, although none are in the project area.

• Convention of International Trade in Endangered Species.

• United Nations Framework Convention on Climate Change.

Ghana has more than 1,000 IUCN-management protected areas, including 317 forest reserves

(EarthTrends 2003). There are two forest reserves near the project area; the Bonsa River Forest

Reserve and the Subri River Forest Reserve. Approximately 12 km of the Hwini-Butre Benso

access road traverses the Subri River Forest Reserve.

20.2.2 Human Rights

In 2005, GSR, with the full support of its Board of Directors wrote to the UN Secretary General

as a statement of commitment to adoption of the United Nations Global Compact. GSR’s 2018

Corporate Responsibility Report is its 13th report on progress of implementation of the UN Global

Compact, and GSR continues to integrate the UN Global Compact principles into its business

activities (www.unglobalcompact.org). Through its annual public Corporate Responsibility Report

(formerly Sustainable Development Report), GSR details ways in which the company is

contributing to advance Ghana’s performance in regard to the Millennium Development Goals. In

2018 Golden Star officially reported in accordance with the Global Reporting Initiative standards.

20.2.3 Anti-Corruption

The Government of Ghana was designated as Extractive Industries Transparency Initiative

compliant in 2010. In support of this, GSR publicly reports, on an annual basis, payments it makes

to the Government of Ghana. As at the end of 2018, GSR businesses have made significant

contributions to the people of Ghana through Government payments:

• GSWL Life to date: Over US$222 million; and

• In 2018, the Office of the Administrator of Stool Lands, Traditional Authorities, Stool

Lands, and District Assemblies expected royalty distributions from the GSWL

operations of over US$1.0 million.

GSR, being registered in the US and Canada, is subject to the US Dodd–Frank Wall Street Reform

and Consumer Protection Act, the US Corruption of Foreign Officials Act and the Canadian

Corruption of Foreign Public Officials Act. Internal GSR policies address these items for GSR

management.

20.2.4 Voluntary Codes

GSR has adopted a number of voluntary international codes and standards of practice pertaining

to corporate responsibility at the Wassa operations:

http://www.unglobalcompact.org/

June 2019


• cyanide management – the GSWL operations have been in full certification to the

International Cyanide Management Code since 2010;

• TSFs – current TSF1 and TSF2 designs align with the requirements of the ICOLD;

• gold mining and processing – as a member of the World Gold Council, Golden Star

ascribes to the Responsible Gold Standard; and

• resettlement, land acquisition, and compensation – since 2009, Golden Star has ensured

all resettlement projects conform to the International Finance Corporation’s

Performance Standard 5 on Land Acquisition and Involuntary Resettlement.

As GSR has adopted these voluntary standards and codes, a key component of GSR’s corporate

assurance includes independent review, audit and/or validation of conformance to the principles

ascribed herein.

20.3 Environmental and Social Setting

20.3.1 Biophysical setting

The concession area falls within the wet semi-equatorial climatic zone of Ghana. It is characterized

by an annual double maxima rainfall pattern occurring in the months of May to July and from

September to October. The average annual rainfall measures at the nearest meteorological station

(Ateiku) is 1,996 ± 293 mm. The average annual rainfall measured at the Wassa weather station is

about 1,750 mm / year.

20.3.2 Hydrology

The Wassa operations fall within Pra River basin, one of the two major rivers draining the south-

western parts of Ghana. The Pra Basin is located in south central Ghana (Figure 20-1) and is an

extensive basin with several river systems criss-crossing its entire surface.

The topography of the Pra basin ranges between sea level and an elevation of 800 m above mean

sea level. The highest elevations in the area are located in the northern sections and the fringes of

the eastern parts of the basin where elevations of up to 800 m above sea level are common. The

southern sections are relatively flat to slightly undulating, although there are a few peaks in the

central regions. The nature and orientation of the highlands determine the direction of the general

drainage network in the entire basin.

The Wassa mining lease area is drained by tributaries of the Pra, namely the Toe to the far south,

Kubekro to the east and the Petetwum to the north. The Petetwum River flows directly into the Pra

River and is drained by the Petetwum, Nankadam, and Kumue streams. The Subiri River, locally

known as Subri, which drains the western end of the concession, is a tributary of the Bonsa.

June 2019


Figure 20-1 Map of the Pra Basin showing the approximate location of the Project site

20.3.3 Hydrogeology

Hydrogeological Regime

In 1995, SGL commissioned Minerex Environmental Limited to conduct a detailed and extensive

hydrogeological assessment, utilizing over 200 boreholes as part of the specialist baseline studies

for the then proposed Wassa Gold Project.

June 2019


The study found that, generally, the groundwater gradient dips steeply off the plateau areas

following, but not as steep as, the topography. Water levels fluctuate seasonally by only 1 to 2 m

and in nine months of monthly sampling (wet and dry season), were slightly higher in November

(gradient 0.048) than in February (gradient 0.043), with water levels falling in March, but rising

in May through July (MEL, 1996 c).

From a hydrogeological perspective, distinct lithological units are apparent with an upper oxidized

zone and a lower fresh rock zone (Table 20-3).

Table 20-3 Baseline Study Identified Hydrogeological Units (MEL, 1996c)

Unit no. Hydrogeological unit Weathered state EC (µS/cm)

A Phreatic aquifer Oxidized 1 to 5

B Confined aquifer in valleys and unconfined on plateaus Unoxidized 0.01

C Quartz veins Unoxidized 1 to 5

The upper aquifer was found to be generally phreatic and the principal groundwater flow occurs

where vein quartz occurs more abundantly.

The lower aquifer is within the unoxidized bedrock. It is unconfined in topographically elevated

areas and semi-confined in the valleys where there is a vertical upward head gradient. The recharge

for this aquifer is on the topographic ridges local to the area where a downward vertical head

gradient exists. The groundwater has a higher mineralization in the confined zones with the

presence of H2S and slightly higher iron and manganese concentrations than the groundwater in

the saprolite (MEL 1996 a, MEL 1996 c).

The confined aquifer may be very static with low throughput and not currently discharging to any

zone in large quantities. The hydrochemistry may indicate a very long residence time and no direct

discharge point, only small dispersed seepages through the aquitard zones (MEL 1996 c).

Analysis of the very low frequency geophysics data for conductive zones (MEL, 1996 b) indicated

that the potential for significant water makes was generally low, and in some valley areas, confined

groundwater was discharging into swamps. The study reports (MEL 1996 a, b and c) present

detailed groundwater piezometric contours, as well as a refined stratigraphic and hydrogeological

model.

The study found that, while the groundwater hydrochemistry could not be clearly sub-divided into

groups, a correlation did appear to exist between the confined nature of groundwater in the

boreholes and the hydrochemistry. Groundwater in the valley areas had a higher calcium

bicarbonate signature than the groundwater from more elevated plateau areas which had a neutral

ionic signature and low ionic strength, indicating that the groundwater resident in the aquifer

longer has become more saturated with respect to calcium carbonate.

The Wassa Main and underground mine area is underlain by the Birimian system, which is known

to yield relatively substantial amounts of groundwater, particularly where the rocks are highly

weathered, fractured and/or inter-bedded with quartz veins. To build on the earlier work of MEL

a detailed hydrogeological assessment was conducted in 2015 to understand the hydrogeological

conditions expected to be encountered at depth with the deepening of the Wassa Main pits and

underground mine establishment.

June 2019


Hydraulic (packer) testing was conducted in August and September 2015, in purpose-selected

exploration holes, to assess the hydraulic conductivity of the various geological units and features

encountered within the study area. Holes were selected to:

• intercept the mining zones, including at depth and expanse;

• test geological formations and known fault structures; and

• enable drill rig accessibility.

Given the depth of the underground mine, and associated geological units, packer tests were

performed using a Standard Wireline Packer System equipped with a Straddle Packer System to

allow specific intervals of each hole to be isolated for permeability testing using both single and

double packer configurations.

The general lithology proved very impermeable (Table 20-4), with most tests recording extremely

low K values of between 4 x 10-5 m/day and 9 x 10-4 m/day, with an average of 3 x 10-4 m/day, to

be used as background for the conceptual and numerical hydrogeological model. This value is

indicative of all hard rock below 160 mbgl depth, i.e. the fresh rock below the saprock.

Table 20-4 Interpreted Packer test results

Hole ID K Interval (m) Hole ID K Interval (m) m/day Top Bottom Thickness m/day Top Bottom Thickness

BSDD308M 9E-05 553 684 131 BSDD139 1E-04 196 460 264

1E-03 523 553 30 BSDD317 1E-04 292 608.2 316.2

1E-04 493 523 30 BSDD185 7E-05 162 621.1 459.1

BSDD315M 2E-04 649 883 234 BSDD325 4E-05 252 618 366

BSDD155A 1E-04 366.5 463.1 96.6 BSDD257 5E-04 330 523.2 193.2

2E-02 342.4 366.5 24.1 BSDD323 2E-04 429 530 101

1E-04 276.5 342.4 65.9 4E-03 387 429 42

BSDD109 4E-04 191.1 323.6 132.5 2E-04 327 387 60

BSDD296 2E-04 488 576.1 88.1 BSDD291 9E-04 463 569 106

3E-03 473 488 15 2E-02 433 463 30

2E-04 245 473 228 9E-04 253 433 180

Note: Zones of higher permeability are highlighted in italics.

Thin zones of higher permeability were found to be associated with faults/quartz veins that yielded

K values of 1 x 10-3 m/day to 4 x 10-2 m/day. These zones are assigned a K value of 2 x 10-2 m/day

in the conceptual and hydrogeological model.

The narrow, higher permeability zones are found along discreet zones associated with fracturing

and faulting. These are isolated and generally will form a very small percentage of the overall rock

mass and will only cause localized higher inflow in the underground workings (Figure 20-2). The

data from the testing has been interpreted and incorporated into the numerical groundwater

modelling.

Ground water level elevation contours were recorded and demonstrate that generally the

groundwater flows in a south-westerly direction following the major topographical features. In

June 2019


close proximity to the active open pits, the prevailing hydraulic head is towards the open pit in

response to the active pit dewatering (Figure 20-3).

June 2019


Figure 20-2 Conceptual Groundwater Model

June 2019


Figure 20-3 Groundwater flow direction

June 2019


Groundwater Modelling

Numerical groundwater modelling using Feflow Version 6.2 was conducted to assist in

characterising the groundwater flow regime and to aid in forecasting the impacts of mine

dewatering and contaminant migration on the receiving environment through the LoM and post

mine closure.

Model boundaries were identified to reflect the geometry of the groundwater system. As there is a

good correlation between surface topography and depth to groundwater, surface drainage

catchment watersheds were selected as boundaries, for a total model area of some 358 km2 (Figure

20-4).

Figure 20-4 Model boundaries and 3D model

June 2019


Key findings of the modelling as it relates to the Wassa underground and Main pits are:

• The drawdown (dewatering cone) from the operations is not expected to have any effect

on existing (community) groundwater boreholes.

• The average water level is 8.8 mbgl. The piezometric head and topographical elevation

display a strong positive correlation (0.87) reflecting that groundwater flow directions

mimic surface topography.

• Groundwater in both the shallow weathered zone and deeper bedrock aquifers flows

from the elevated areas towards the rivers, following the topography. The regional flow

direction is from north-west to south-east.

• In base case conditions, the underground will likely produce an average of 1149 m3/d

at the end of the 2021, and the open pit will likely produce 400 m3/d at the end of 2024.

During connection with the higher permeability zones, peaks of groundwater inflows

in the underground could reach 2000 m3/d at certain periods. These volumes are well

within the current WRC permitted abstraction.

• Contaminant modelling shows that the sulphate concentrations will not exceed

270 mg/l, with resulting low impact. The potential contaminant plume is controlled by

the cone of depression, and slow groundwater movement (recovery) conditions. Thus

the receiving environment will not receive underground mine leachate in the recovered

state. Additionally, no decant is expected to occur.

• Leachate from mine waste rock dumps is similarly controlled by the cone of depression,

and is not expected to impact on the receiving environment.

• The model shows that, at closure, recovery of the groundwater table (dewatering cone)

will occur to some 68% of the pre-mining level as measured in 1996. The groundwater

is predicted to stabilize at 106 meters above mean sea level (mamsl), against a pre-

mining level of 141 mamsl (MEL, 1996).

• The initial groundwater water recovery is rapid, with 75% of the expected Main pit

groundwater level recovery (80 mamsl) occurring within 8 years of the cessation of

dewatering (Figure 20-5). Ninety percent of the recovery will occur within 16 years

and the final ten percent of the recovery will take a further 24 years, or 40 years since

the cessation of dewatering.

June 2019


Figure 20-5 Modelled groundwater level recovery (Wassa Main)

June 2019


Figure 20-6 Dewatering cone at life of mine

June 2019


20.3.4 Geochemistry

Historically, the geochemistry of the rocks intersected in the Wassa Main pits complex has

consistently shown that the rock lithologies, ore and waste, are not acid generating (“NAG”).

The potential for acid rock drainage (“ARD”) was originally assessed during the Wassa Mine

environmental impact assessment (SGS, 1998). It was observed that waste rock was of lower

sulphide sulphur content than ore, thus the analysis focused on ore samples. Analysis found that

ore had a low acid producing and high neutralization potentials and the ARD potential was defined

as very low.

Later in 2002, a due diligence study by SGS (2002) found no evidence of development of an acidic

condition in pit waters. Analysis during operations showed the pH to consistently range from pH

6.5 – 8.0. The study observed that the host rock exhibits carbonate concentrations that buffer any

ARD potential.

In 2004, 48 waste samples, from a full spatial distribution of the pits, were analyzed for

geochemical characteristics including X-ray diffraction, X-ray fluorescence (whole rock analysis),

leach extraction tests and acid base accounting. The results (2004 Wassa EIS), found that the risk

of ARD at the Wassa Mine is very low, and that very low concentrations of minerals are leached,

even under active leaching conditions.

In addition to the studies conducted for the Wassa EIS, Golden Star continues to assess the

geochemical characteristics of the ore, waste and tailings materials annually. This assessment

continues to demonstrate the low potential for acid generation.

As part of the impact assessment process for the Wassa expansion project, 220 m of diamond-

drilled exploration core was selected from future mining areas to represent the proposed spatial

distribution, depth, and rock units likely to be intersected in future open pit and underground mines,

for examination and logging of key geochemical and hydrogeological features. This program was

carried out in order to quantify the ARD and metal leaching potential associated with the rock

units.

Acid base accounting

Acid base accounting analyses were conducted on 156 samples of various pit, waste rock and

underground core from the proposed underground mine.

The sulphur content of rock materials from the pits (mean= 0.17 ± 0.29%, with BMU mean=0.02

± 0.01%, basalt mean= 0.26 ± 0.4%) and dumps (BMU mean =0.08 ± 0.07%, basalt mean= 0.19

± 0.17%) is highly variable. The sulphur content in samples from the underground mine area varies

between 0.005-0.57%, with felsite averaging 0.11 ± 64% and phyllites 0.26 ± 0.22%.

The acid potential (“AP”) of the different rock types from the pits (0.3-61, mean=5.2 kg CaCO3

eqv/t), waste rock (0.3-17, mean=4.6 kg CaCO3 eqv/t) and underground mine area (0.16–18,

mean=6.7 kg CaCO3 eqv/t) is generally low.

The neutralization potential (“Bulk NP”) of rock samples from pits (17-263, mean=97 kg CaCO3

eqv/t), waste rock (12-321, mean=92 kg CaCO3 eqv/t) and underground (25-211, mean=111 kg

CaCO3 eqv/t) is generally very high.

June 2019


The carbonate neutralization potential (“NP”) is generally higher than Bulk NP indicating that

ankerite represents a significant proportion of total carbonates in the different rock types at Wassa.

However, ankerite has limited neutralizing capacity under oxidizing field conditions, as ferrous

iron is an extra source of acidity due to the strong hydrolysis of the ferrous iron in solution (Blowell

et al., 2000). The paste pH (8.3-10.1) was generally alkaline in all rock units indicating availability

of excess buffering capacity to neutralize acidity formed from the initial oxidation of sulphides

during the testing procedure. There is generally sufficient reactive NP in the rock materials with

Bulk NP exceeding AP in all the samples. This is also indicated by the generally high positive net

neutralization potentials of rock samples from the pits (16-263, mean=91 kg CaCO3 eqv/t), waste

rock (11-321, mean=87 kg CaCO3 eqv/t) and underground (19-206, mean=105 kg CaCO3 eqv/t).

Classification of ARD potential shows that all the rock samples from the pits, waste rock, and

underground are not potentially acid generating (Figure 20-7) following the guidelines of Morin

and Hutt (2007) and MEND (2009).

Using the alternative classification method of Price et al. (1997) and Soregoli and Lawrence

(1997), results also show that all the rock samples from underground, and the majority of waste

rock, and pit rock samples have no acid generating potential (Figure 20-8). Exceptions were two

diorite samples, and a sample each of phyllite and basalt, which had a low acid generating

potential.

June 2019


Figure 20-7 Paste pH vs NPR for pit, waste and underground samples

Figure 20-8 NPR vs %S for pit, waste and underground samples

June 2019


Geo-environmental Context

The shallower gold resources at Wassa are extracted by open pit mining using excavators and

trucks. Some of the pits have been mined out, while others are in various stages of backfilling with

waste rock. Waste rock not used for backfilling of pits is generally stockpiled in waste rock dumps

adjacent to or west of the main pit complex (Figure 20-9).

Figure 20-9 Conceptual Geo-environmental model (E-W cross section)

The deeper gold resources are extracted from the underground mine, accessed via decline, at times

concurrently to the open pit operations. The pit is expected to intersect early upper longitudinal

stoping areas during the cut 3 operations (Figure 20-10).

Hydrogeological assessment has shown that inflow of groundwater will occur along discreet zones

of faulting and fracturing with potential permeability of up to 120 L/min (2 L/s) measured.

Hydraulic testing of the underground area to depths of 800 m below surface has shown that

generally the formation is not water bearing with very low permeability overall.

Storm water and groundwater inflow into mining areas is managed by dewatering from sumps.

Sumps are established underground on production levels as mining progresses. The water from the

sumps is pumped to sediment settling ponds before being discharged into the environment.

June 2019


Figure 20-10 Conceptual Geo-environmental model (N-S layout cross section)

June 2019


Underground Mine Drainage Quality

Given the geo-environmental context, assessments were carried out to understand the quality of

leachate/water expected to be dewatered from the underground mine. Synthetic precipitation

leaching procedure and net acid generation leach tests were carried out on samples from the

underground mine area, in order to obtain indications of the potential drainage quality from the

underground mine.

As leachate generated by NAG leach tests represents complete and instantaneous oxidation and

leaching of all reactive minerals, these tests assess the maximum (worst case) quality of drainage

from the underground mine. Under field conditions, sulphide oxidation and release of elements

will occur gradually and as such, concentrations in mine drainage are expected to be lower than

NAG leachate chemistry at any given time (INAP, 2010). The results indicated that none of the

measured constituents would exceed the water quality guidelines in the underground mine

drainage. The underground mine drainage was predicted to be generally neutral to alkaline with

low concentrations of TDS, sulphate and metals (Figure 20-11). This has been validated by routine

underground mine water quality sampling (Section 20.3.5).

Figure 20-11 Ficklin diagram showing composition of underground mine leachate

20.3.5 Water Quality

MEL (1996c) found that, while the groundwater hydrochemistry could not be clearly sub-divided

into groups, a correlation did appear to exist between the confined nature of groundwater in the

June 2019


boreholes and the hydrochemistry. Groundwater in the valley areas had a higher calcium

bicarbonate signature than the groundwater from more elevated plateau areas that had a neutral

ionic signature and low ionic strength, indicating that the groundwater resident in the aquifer

longer has become more saturated with respect to calcium carbonate.

GSWL currently maintains an extensive water quality monitoring program for both surface and

ground water (GSWL Annual Environmental Report 2018). Sites have been routinely sampled

since 2003 with external laboratory analysis conducted since 2012. The parameters analyzed are

compared to the WRC’s Raw Water Criteria and Guidelines for Domestic Water Use, as well as

the EPA’s sector specific effluent quality guidelines for discharges into natural water bodies (EPA

guidelines).

GSWL’s interpretation of the data is that both surface water and groundwater quality has remained

consistent with the findings of the Wassa Gold Project Environmental Baseline Study (SGS 1996),

the EIS (SGS 1998) and associated specialist studies (MEL 1996a, b and c) throughout operations.

Groundwater in the area generally ranges from slightly acidic to basic in nature, reflecting the

nature of the soils, as well as the lack of connection between the aquifers. Studies have shown that

the shallow groundwater is often acidic (Geosystems 2013 and 2015), while the water quality in

deeper bores reflects the greater saturation of neutralizing minerals resulting from the more

confined nature, and associated longer residence time, of the deeper aquifer.

The nitrate and nitrite concentrations are low, as is the phosphorus concentration of groundwater,

reflecting the low contents in the rocks from which the soils develop, and to a greater extent the

intense leaching to which they have been subjected.

Surface water in the vicinity of the main pits, on average, conforms to the EPA Effluent Quality

Guidelines. Occasional peaks in suspended sediment and nitrogen from the operations are removed

by mine dewatering treatment processes. Elevated levels of iron are seen to reflect the baseline

conditions and rock geochemistry.

20.3.6 Air quality

Routine air quality monitoring in the operational area typically takes the form of monthly 24-hour

assessments of total suspended particulate, particulate matter, depositional dust, nitrogen oxide

and nitrogen dioxide. Prevailing air quality is also monitored at communities nearest to the

operational area. Except during the Harmattan, air quality generally exhibits low levels of

particulates, reflecting the largely rural nature of the area. Sources are mostly anthropogenic,

resulting from domestic activities such as open fire cooking, gardening and human movement.

Recent impact assessment studies for the Wassa expansion (underground mine, and pit and dump

expansion) incorporated predictive modelling using the AERMOD dispersion model to determine

the potential impacts of the expansion on air quality. The model predicts ground level

concentrations and deposition rates of the modelled emissions using a regional mesoscale

meteorological dataset (MM5) over the modelling domains.

The study found that even under predicted worst case conditions and without mitigations, ground

level concentrations of the key emissions at the nearest sensitive receptors met the majority of the

applicable Regulations. The predicted concentrations illustrate that mine derived emissions alone

are predicted to be within EPA ambient air quality guidelines at all times. It was only under

cumulative conditions of worst case weather (single worst 24 hours), and seasonal Harmattan

June 2019


peaks, that PM10 and TSP ground level concentrations exceed the 24 hour guideline. With

mitigation, the mine’s contribution to this cumulative level would be reduced. These findings have

been validated by routine air quality monitoring of the operations.

20.3.7 Noise and Vibration

Noise monitoring is undertaken routinely at Akyempim and Kubekro, which are the nearest

communities. The available data confirms that the noise emanations are predominantly local

anthropogenic sources and were not observed to have resulted from activities at Wassa.

The recent impact assessments for the Wassa expansion assessed sound propagation using the

ISO 9613 compliant software package CadnaA, to assess sound propagation under a variety of

meteorological conditions. Meteorological data derived from the MM5 data was utilized. A

baseline noise model, developed to enable comparison of mine development scenarios with the

existing conditions, confirmed the validity of the model.

The predictive noise modelling assessment found that even under worst case conditions, and

without mitigation, noise levels at receptor sites would be within the EPA guidelines for ambient

noise under all modelled scenarios. Routine noise monitoring by GSWL continues to demonstrate

conformance to the EPA noise guidelines, in respect of the operations, with no noise exceedances

recorded over the last several years.

Impact assessment also incorporated modelling to predict impacts associated with blast induced

vibration. Modelling utilized the United States Bureau of Mines (USBM 1980) ground vibration

propagation equations, and the ICI formula for estimation of air blast overpressure (ICI 1995). The

predictive study of blast induced vibration found that even under worst case modelled scenarios,

the levels predicted for ground vibration and air blast overpressure at nearest receptors would meet

the Minerals and Mining (Explosives) Regulations (LI 2177).

The impact assessment findings have been validated by routine blast monitoring that has

demonstrated conformance to the regulatory limits.

20.3.8 Biodiversity

The Wassa expansion project infrastructure and operations will occur entirely within the existing

Wassa Main pits excavations and previously compensated areas, so it is unlikely that any further

impact to flora and fauna will result from the new project.

The concession is in the transitional area between moist, semi-deciduous forest and wet rainforest

zones. Prior to mining, during baseline studies in 1996, the natural vegetation was observed to be

degraded by earlier logging and farming activities. It comprised broken forest, secondary forest

and upland regrowth, and valley bottom swamps. No endangered plant species were recorded in

the field surveys (SGS 1996).

Across much of West Africa, the status of vegetation has changed considerably with time in

response to conversion of forest lands to agriculture and other land uses. Review of the 1996 floral

species list in the present day shows that of the identified species, three are now classified as

vulnerable (IUCN 2016) including Mitragyna stipulosa, Turraeanthus africanus and Guarea

cedrata. Three other species were only identified to Genus level and may also have modified

conservation status in the present day, including Terminalia, Entandrophragma and Pterocarpus-

sp. Of the species of conservation significance, Golden Star actively propagates a number of these

June 2019


species and utilizes them in the mine site revegetation, and in the last four years alone has

propagated in excess of 13,000 seedlings of these species.

In 2010, a baseline biodiversity survey was conducted in the vicinity of a number of the previously

assessed areas to consider locations for future waste rock dump expansions. During this time, the

vegetation was primarily a combination of secondary forest and farmland. In the surveyed

quadrats, 214 plant species were identified. The survey identified a number of quadrats that, at the

time, hosted high quality unprotected remnant forest stands and, as a result, Golden Star

determined to avoid these areas in the design of future waste dump expansions. Likewise, the

proposed Wassa Expansion project has also been designed to specifically avoid these areas,

although it is conceivable that since 2010, with the ongoing development of farmlands in the area,

these stands of vegetation may have been converted to agriculture.

In 2012, a large scale baseline biodiversity assessment was conducted in areas previously

undisturbed by mining activity as part of the environmental impact assessments for the then

proposed TSF 2 project. The study, focussing on a valley to the north of TSF 1, covered an area

of predominantly cultivated and cleared farming lands (81%) with isolated remnants of former

forest in partly cleared (9%), uncultivated (0.3%) and unclassified land (8%).

The study identified 70 floral species from the project area footprint and surrounds, of which one

individual of Tieghemella heckelii is listed as endangered (IUCN 2016), and eight timber tree

species as vulnerable as a result of overexploitation for timber products and clearing of land for a

variety of uses (IUCN, 2016). Of the species of conservation significance, Golden Star has

propagated over 17,800 seedlings of these species in the last four years for use in mine site

reclamation. The TSF 2 EIAs (Geosystems 2013 and 2016) observed that this mitigation should

not only reverse the impact, but also act to improve the local conservation status of these species.

There are two forest reserves in the vicinity of the Wassa operational area; the Bonsa River Forest

Reserve and the Subri River Forest Reserve. The Benso site is located approximately 17 km to the

west of the southern portion of the Subri River Forest Reserve, with approximately 12 km of the

Hwini-Butre Benso access road traversing the Subri River Forest Reserve. The most important

portion of the Subri River Forest Reserve, namely a Globally Significant Biodiversity Area, is not

impacted by the Hwini-Butre Benso access road.

The Subri River Forest Reserve covers an area of approximately 590 km2. It is an actively managed

reserve and is currently being logged on a 40-year cycle. Approximately 2,590 Ha of the reserve

is being used for silvicultural research. The reserve forms part of the watershed between the Bonsa

and Pra Rivers and is traversed by their tributaries, resulting in extensive areas of swampy

vegetation.

The 1996 baseline study found no species of small mammal, bats, birds, herpetofauna, or

amphibians of outstanding conservation merit. Of the large mammals, several species were

reported as being of conservation significance, and it was observed that, given the prevailing high

hunting pressures and impacts of logging activities, it was necessary to traverse more than 10 km

into the Forest Reserve to observe any of these species. Ongoing intense hunting pressure, logging

and agricultural conversion have continued since 1996.

As at the present day, the conservation status of few have changed. Kinixys homeana (Hinge-back

tortoise) is now vulnerable, and Scotonycteris ophiodon (Pohle’s Fruit Bat) is near threatened,

whilst the large mammal species identified as being of conservation significance remain as such,

June 2019


and continue to be afforded some degree of protection under the Ghana Wildlife Conservation

Regulations, 1971 (LI 685). Of the birds, Necrosyrtes monachus (Hooded Vulture) has been up-

listed to critically endangered owing to indiscriminate poisoning, trade for traditional medicine,

hunting and persecution (IUCN 2016), and Psittacus erithacus (Grey Parrot) has been classified

as Vulnerable owing to international pet trade in the species (IUCN 2016). Neither species has

been identified in biodiversity assessments since the 1996 baseline.

The fauna survey conducted as part of the 2010 biodiversity assessment of the then proposed waste

rock dump expansion locations identified 7 large mammals, 9 small mammals, 55 birds, 34

butterflies, and 5 amphibians, of which two species Phataginus tricuspis (African White-bellied

Pangolin) and Anomalurus pelii (Pel’s Flying Squirrel) were of conservation significance, both

being classified as Near Threatened. The Pangolin is now classified as Vulnerable, whilst the

Flying Squirrel has been reclassified as Data Deficient (IUCN 2016). None of the other species

identified have since been classified as being of conservation significance.

The 2012 impact assessments for the then proposed TSF 2 project, included baseline fauna surveys

for both terrestrial and aquatic species and ecosystems. Of the 341 Ha of the wider TSF 2 project

area, less than 0.3% of the land was uncultivated, and reflecting the indiscriminate hunting and

clearing of forest for agricultural purposes (Geosystems 2013), the fauna of the project area was

relatively impoverished. The study found no species of Lepidoptera, amphibians, reptiles, birds or

aquatic species listed as being of conservation significance according to the IUCN. Of the mammal

species identified, a single African White-bellied Pangolin was identified in the wider project area.

20.3.9 Social setting

5Administrative Setting, Nearest Settlements and Land Ownership

The Wassa Mine is in a rural setting and there are no major urban settlements within 30 km of the

operations. It is in the Wassa East District (previously in the Mpohor Wassa East District) of the

Western Region of Ghana, 62 km north of the district capital of Daboase and 40 km east of Bogoso.

Cape Coast is approximately 90 km to the south.

The villages nearest the mine are listed below. The villages of Akyempim, Akyempim New Site

(formally Akosombo that was resettled by the Company) and Kubekro are the closest communities

to the Wassa operational site. The Togbekrom community was resettled to create space for

construction of TSF 2.

Table 20-5 Overview of Local Communities

Community Divisional Area Estimated* population (SGS

1996) Population (WEDA 2013)

Akyempim Mamponso 2,500 2,533

Akosombo Mamponso N/A 166

Kubekro Anyinabrem 300 335

Nsadweso Anyinabrem 2,400 1,541

Togbekrom Anyinabrem NM 674

NM= not measured in survey, * = as estimated by traditional leaders

The District Assembly is the supreme organ charged with the administration and supervision of

the district development activities and the District Chief Executive is the most senior government

official.

June 2019


The project is located within the Wassa Fiase Traditional Area, with its Paramountcy at Tarkwa.

Within the traditional structure, the Paramount Chief (Omanhene) is the head and exercises

traditional control over the divisional and sub-divisional chiefs (Adikro) of communities

(traditional towns and villages).

The 1992 Constitution of Ghana provides for three categories of land ownership or land holding:

customary (stool/skin) lands (78 %); state lands (or public land) (20 %); and vested lands (or public

land) (2 %). Customary lands are managed by traditional authorities in accordance with customary

laws. The State exerts considerable control over the administration of customary lands. Access to

land is available only under leasehold.

As the lands within the Wassa concession are mineralized, they are state-owned with the mineral

rights granted to GSWL under the Minerals and Mining Act, 703, 2006. The Constitution of Ghana

(1992), State Lands Act (1962), Minerals and Mining Act 703 (2006), Minerals and Mining

(Compensation and Resettlement) Regulations (2012), Mining and Environmental Guidelines,

Environmental Protection Agency Act 490 (1994) and Environmental Assessment Regulations

(1999) each have provisions pertaining to land access, including land acquisition, land and farm

compensation and resettlement.

All land affected by the planned expansion activities are traditionally in the ownership of the

Mamponso Stool of the Wassa Fiase Traditional Area. The relationship between the Divisional

Stool Chiefs and the inhabitants is based on tenancy. The tenants typically pay annual rent by

means of a portion of their annual crop returns.

Land Use, Livelihoods, Health and Education

The Wassa Mining Lease (LVB 87618/94) area is 5,289 Ha and, as at December 2018,

approximately 595 Ha of disturbance had occurred in the area from GSWL activities. GSWL has,

however, provided compensation for a total of 1,293.63 Ha of land disturbed by infrastructure

development (including TSF 2) and operational activities and for buffer areas.

Prior to development of the mine, the main land use in the 52.89 km2 concession area was found

to be farming. Cocoa was the main crop. Other major crops cultivated in the concession were oil

palm, maize, intercropped with cassava, and plantain. In addition, there were compound farms

surrounding villages and hamlets with crops such as coconut, cocoyam, avocado pear, citrus,

mango, maize and cassava in mixtures. There were no commercial plantations within the

concession and commercial logging was almost entirely restricted to the portion of the Subri River

Forest Reserve.

Most people were found to be dependent upon crop farming for their livelihood. Crop farming

was also the principle source of employment. Farming in the concession was dominated by migrant

farmers from other regions of Ghana, using land owned by indigenous families on a leasehold

basis.

Livelihoods of people in district are still based on agriculture and about two thirds of economically

active people are employed in the agricultural sector. About one quarter of people are employed

in the mining/quarrying, manufacturing and wholesale/retail sectors.

In the district, more than half of the homes are constructed from mud/earth. Roofing materials are

generally metal sheet. In the district, most people obtain water for drinking and domestic use from

boreholes or rivers/streams. About one fifth of households obtain water from pipes outside

June 2019


dwellings and public stand pipes. Less than half of the households in the region use mains

electricity for lighting and most use wood or charcoal for cooking. Most households use public

toilets or pit latrines. Most households also do not have access to formal waste disposal facilities.

Much waste is dumped in public open space and liquid wastes are generally disposed of in

compounds/street or gutters.

About a third of the population over 12 years of age own mobile phones and very few households

have landline telephones.

Malaria is a common illness experienced by the catchment communities and remains a serious

public health concern nationally. It is regarded as a leading cause of morbidity and mortality,

especially among pregnant women and children under five years (NDPC & UNDP 2010). Other

common ailments in the area are respiratory tract infections and diarrhoea.

Literacy levels in the Western Region were 58.2 % in 2008, with a bias toward males (68 %), and

remained largely unchanged. Attendance of primary and middle/ junior school is higher in the

district than in the region and in 2010 only 25% of people in the district over 11 years were not

literate. In recent years, there has been an increase in female attendance at the primary and junior

school levels, but about 10% less females than males complete school (NDPC & UNDP 2010).

20.4 Environmental and Social Management

20.4.1 Golden Star Corporate Commitment

GSR has policies pertaining to the environment, community relations and human rights,

community development and support, and health, safety, and wellbeing. In support of these

policies, GSR demonstrates its management commitment through provision of appropriate and

dedicated specialist human resources in the disciplines of environment, safety, health, community

affairs and resettlement. In 2018, GSWL employed 71 dedicated personnel in the disciplines of

environment, communities, safety, health, and security, representing some 10% of the total

employees. Environmental expenditure in 2018 represented almost 1.2 % of total operating

expenditure.

GSR supports achievement of its corporate policies by providing training and development for its

workforce with over 100 000 personnel hours committed to personal development training at the

Wassa operations in 2018.

20.4.2 Social Investment

Golden Star Development Foundation

The primary vehicle for GSR’s social investments is the community-led Golden Star Development

Foundation, which is funded annually with US$1/oz Au produced and 0.1% of pre-tax profit.

Under the foundation umbrella, GSWL works with local Community Mine Consultative

Committees (“CMCC”), government bodies, and third-party non-governmental organisations

(among others) to strategize and implement a variety of community development projects and

programs.

In 2018, GSR contributed almost US$ 0.15 M to the foundation, bringing contributions to date to

over US$ 3.68 M.

June 2019


Golden Star Oil Palm Plantation

Golden Star Oil Palm Plantation (“GSOPP”) is a community-based oil palm plantation company

established in 2006 as a non-profit subsidiary of GSR. The program adopts the small-holder

concept of sustainable agribusiness, which addresses environmental, food access, and community

concerns. Initially, development is sponsored by GSR as part of its local economic development

program. The plantations are later able to become self-supporting and the small-holder farmers

pay back the start-up loans to GSOPP to allow for further development. GSR commits US$1/oz

Au produced to the program, resulting in over US$ 6.6 M in funding as at year end 2018. To date,

GSOPP has established 1,133 Ha of plantations. In 2018, GSOPP produced and sold over 10,000

tonnes of oil palm fruit. In 2018, GSOPP was expanded into the parts of TSF 1 that had reached

closure elevation.

Capacity Building and Livelihoods Enhancement

Employment, particularly for the youth, continues to be of the foremost concern to GSR’s

catchment communities. Education and training initiatives are extended to our community out-

reach programs, with a view of imparting lasting educational benefits to stakeholder communities.

The Golden Star Skills Training and Employability Program (“GSSTEP”) provides training to

young people in practical and technical skills in sectors unrelated to mining, contributing to the

diversification of the local economy’s employment base. This program has also been integrated

into many of the negotiated resettlement agreements that conform to the IFC Performance Standard

5 on involuntary resettlement.

Inaugurated in 2009, as at the end of 2018, 14 GSSTEP programs had been run, providing skills

training to over 600 youth in masonry, commercial cookery, carpentry, mobile phone repairs,

building electrical, beads and jewellery making, hair dressing, local fabric bags and sandal making,

and other trades.

In 2013, under the umbrella of GSSTEP, GSWL initiated a pilot community youth apprenticeship

program (“CYAP”), which offered selected local residents a one-year attachment within the

Company. The pilot project enrolled 44 young people from 15 catchment communities in

disciplines ranging from welding and drill rig maintenance, to fixed plant, heavy equipment, and

pump operations. As a result of CYAP, local graduates will be better positioned to fill skilled

employment vacancies within the company to further boost local hiring.

GSR also provides scholarships for needy students attending secondary school. Since 2008, the

company has provided scholarships for over 892 children. A further 3,000 registered dependents

of employees are also supported through educational subsidies on an annual basis.

20.4.3 Corporate Responsibility

In accordance with its commitment to the UN Global Compact, GSR supports and respects

internationally proclaimed human rights within their sphere of influence. As per GSR’s policies

on Community Relations and Human Rights, GSR works to create a culture that makes the

protection of human rights an integral part of the short and long-term operations, including the

performance management systems.

June 2019


GSR periodically conducts human rights reviews with its top suppliers with results reported to the

GSR Corporate Responsibility Committee. This provides further assurance that GSR is not

complicit in any human rights abuses – directly or indirectly.

Building on training covering human rights matters for our Human Resources personnel – and

later our wider workforce – GSR developed a similar program covering matters related to

harassment and discrimination awareness and prevention.

Golden Star continues to implement a number of major safety improvement programs to further

embed safety management into its operations. These programs were further embedded into our

operations including:

• Risk management system enhancements.

• Crisis and emergency management system upgrades and training.

• Safety culture surveys.

• Safety leadership training.

GSR is dedicated to engaging in accurate, transparent, and timely two-way consultation with local

stakeholders in order to communicate on the business and address the needs of local partners.

Regular dialogue with stakeholders – including but not limited to public meetings, open houses,

and sensitization forums – is central to understanding key issues and concerns related to the

operations, and, in turn, helps to realize sustainable solutions suitable to the stakeholders.

GSR assumed the role as a catalyst for sustainable economic development in the communities in

which operations are situated. Doing so enhances relationships with partners by maximizing the

benefits that accrue to the stakeholder communities. Accordingly, GSR makes regular investments

in local communities that go beyond traditional philanthropy, namely by adopting a strategic

approach to social investment. This helps to create lasting, meaningful benefits for local

communities and contributes to a positive long-term legacy surrounding the operations.

In the area of security and human rights, in 2014, GSR commenced a program of training and

awareness with its security personnel, and military personnel, in the Voluntary Principles on

Security and Human Rights. As at the end of 2018, over 740 security personnel had ascribed to

the principles through this program with the Voluntary Principles now a standard part of induction

for new personnel to the security team.

20.4.4 Environmental and Social Management System

For existing operations, environmental management is addressed through an Environmental and

Social Management System (“EMS”) developed along the lines of an ISO 14001 EMS. This

allows the operation to provide a program addressing the legal and corporate needs for monitoring

and reporting. The EMP and the associated Environmental Certificate provide the legal framework

for GSWL environmental management, whilst EIS and associated Environmental Permits provide

the legal framework for project developments.

Community management at GSWL is carried out by the Environment and Social Responsibility

Department. GSWL has established a series of CMCCs within the local stakeholder communities.

An Apex CMCC collects the recommendations and then makes them to the corporate and company

entities (such as the Golden Star Development Foundation) on behalf of the three functional areas

June 2019


(Wassa, Hwini-Butre and Benso). This aims to ensure that full representation across the GSWL

operations occurs without interference from GSWL.

The CMCCs are responsible for selecting development projects and assisting the operations

understanding of community concerns and needs. Development opportunities for the stakeholder

communities are funded by the Golden Star Development Foundation, and through direct mine-

funding.

GSWL maintains a grievance mechanism enabling catchment communities to document concerns

and grievances for investigation / action. The mechanism is well publicized by GSWL and used

actively by the community and other stakeholders. Details of registered grievances and their

resolution are recorded and reported internally and to the regulators.

20.5 Environmental and Social Issues

This section highlights environmental and social issues that could affect permitting, operations or

maintenance of approvals, issues that are of concern to local stakeholder communities and/or

issues with management costs that may affect the value of the assets. Environmental and social

impacts that can be managed readily without remarkable cost are not discussed here.

20.5.1 Community Expectations and Sensitivities

Employment

The main socioeconomic concern for most stakeholders is employment. The local community

around the Wassa operation see working at the mine as a preferred occupation. The extension of

the mine life with the development of the underground mine is expected to receive local support.

Employment levels for the Wassa underground mine have yet to be developed; however,

community expectations will be managed via the normal community consultative methods.

Although GSR is unable to employ all the people seeking work, there is a local hiring policy in

place that provides affirmative action for the employment of local stakeholder communities. All

vacant positions are advertised locally first and then nationally. Local people are used exclusively

for unskilled positions, and as much as possible for all other positions within the operation. The

project will draw most of its required workforce from within the Western Region. The Wassa

operation has started a local training program along the lines of an apprenticeship where people

from the local community are offered the opportunity to train in work areas where the mine may

need workers in the future. These programs are complemented by an array of other alternative

livelihoods initiatives.

Access to land, noise and blasting effects

Other community concerns include access to land, and noise and blasting effects. Routine

environmental monitoring continues to demonstrate high levels of conformance to regulatory

standards for water, air, noise and vibration.

20.5.2 Resettlement and compensation

Where physical, social and/or economic displacement is anticipated, GSWL applies the

requirements of the International Finance Corporation, Performance Standard 5 for land

acquisition and involuntary resettlement. Should any compensation be required for future

June 2019


operations this will be undertaken in accordance with applicable laws, and in accordance with

GSWL Farm Compensation and Land Acquisition procedures. Life of mine plans for mining,

tailings and waste rock storage are fully accommodated by existing permits and compensated land

areas.

20.5.3 Unauthorized Small-Scale Mining

Galamsey is the local name for unauthorized or illegal mining. It is often associated with

environmental degradation, safety hazards, and general community and social concerns. GSR has

reported that galamsey in the area of the project has little potential to affect the operations. The

main project site is well secured, with other infrastructure located between the Wassa main pits

complex and the nearest community. In general, the removal of unauthorized persons from the

wider project area has posed no difficulty, with persons moving on as requested. As the

underground mine entrance is located within the existing open pits complex, unauthorized small-

scale mining is not expected to adversely affect the operations.

20.5.4 Process Water Balance and Discharges to the Environment

The water management at the GSWL site has been such that discharges to the receiving

environment from the TSF have not been required since 2010. The Wassa operation has an

approved detoxification plant to treat cyanide in supernatant waters that is available should a

discharge be required. However, the water balance model for the current configuration of the site

indicates that under normal conditions discharges should not be required.

Normal mining operations continue with the operational requirement for the installation of sumps

and the removal of rainfall and groundwater that enters the mining areas. The management of this

water is to pump the water to sedimentation structures and then release the water to the receiving

environment. This is carried out in compliance with the permits. To improve the overall

management of surface run-off from the mining areas, five sedimentation structures are employed.

These are primarily to remove suspended solids from the run-off water that may be elevated during

storm events.

The dewatering water abstracted from the underground operations is treated for off-site release, as

required, via the existing systems of treatment and discharge. The natural stream and creek systems

contain seasonal flood flows, and the additional dewatering volumes are within current permitted

abstraction volumes.

20.5.5 Geochemistry

Studies undertaken to date indicate that ore and waste rock have low potential for acid generation

(Section 20.3.4). Exploration drilling data are reviewed to confirm whether lithologies, weathering

and alteration - geochemical controls – are expected to significantly differ from those that are

currently apparent.

20.5.6 Legacy Issues

When GSR (through GSWL) took over the WGL operation, most of the infrastructure was in place.

Since that time, the former HL area has been encompassed within the TSF 1 footprint. The

establishment of the Reclamation Security Agreement and associated bond with the EPA addresses

June 2019


security for reclamation and closure. There are no other legacy issues associated with the GSWL

site.

20.6 Closure Planning and Cost Estimate

Environmental Permits, and approval of Mining Operating Plans are required, and have been

obtained, for the operations. The EIS’ submitted to the EPA to obtain the various Environmental

Permits contain provisional closure plans and cost estimates. The annual Mining Operating Plans

also contain details relating to closure and reclamation.

The rehabilitation and closure of the existing operations (e.g., processing plant, TSF, pits, waste

dumps, and transportation corridor) are covered under the EMP, Reclamation Security Agreement,

and associated bond. With each successive expansion or new development, GSWL develops a

conceptual closure plan that is incorporated into the applicable EIS.

The rehabilitation and closure of the existing operations (e.g. processing plant, TSF, pits and waste

dumps) are covered under existing GSWL Asset Retirement Obligations.

For the Wassa expansion a closure cost estimate was incorporated in the project EIS and is updated

through annual updates of the asset retirement obligations based on the following principles:

• No allowance for scrap value.

• Progressive closure integrated with on-going operations.

• Costs based on a mix of current contractor rates and work being undertaken directly by

the operation.

• No provision for ongoing treatment of water. The underground mine will be allowed to

flood to the natural level, which is the current closure plan for the underground

operations.

• Community post-closure issues will not be included.

GSR expects the EPA to request updates to the reclamation bond as new expansions are approved,

through the Reclamation Security Agreement. The current bond over the combined Wassa, Hwini-

Butre and Benso concessions of GSWL is US$9.6 million.

June 2019


21 Capital and Operating Costs

21.1 Capital Costs

Table 21-1 presents the capital expenditure schedule for the Wassa Mine.

Sustaining capital includes the following items:

• UG primary access development – main access decline and associated ventilation and

materials handling infrastructure.

• UG mobile equipment – fleet increases to meet the daily tonnage demand of 4,000tpd

during the later part of 2020; and equipment replacement.

• UG pumping – primary pumping infrastructure.

• Plant tailings – cost of tailings dam lifts over the LoM.

Development capital includes the following items:

• Raisebore holes for ventilation and waste backfill delivery.

• Paste backfill plant construction.

• Contractor mobilization and pre-strip for Cut 3 and 242 mining.

Table 21-1 Capital cost schedule

21.2 Operating Costs

Table 21-2 shows the annual total operating costs for the Wassa Mine.

Underground stoping and development costs are estimated based on historical costs. Pastefill costs

are estimated based on the FS work carried out by Outotec.

Open pit mining costs are estimated based on historical experience of GSR with contract mining

companies in Ghana adjusted for the size of the Cut 3 and 242 operation.

June 2019


Processing costs are based on historical experience. The unit cost estimate is adjusted based on the

following plant throughput rates:

• 3,000-4,000 tpd $25/t

• 4,000-6,000 tpd $20/t

• 6,000-8,000 tpd $15/t

G&A costs are estimated based on historic performance and varying from $12 Mpa when

underground mining is the primary production method, rising to $14 Mpa during open pit

operations.

Closure costs are estimated based on our total rehabilitation requirements at Wassa.

Table 21-2 Operating costs

June 2019


22 Economic Analysis

22.1 Inputs and Assumptions

Inputs to the cash flow model include:

• Total LoM production of 18,607 kt at an average grade of 2.46 g/t containing 1.5 million

ounces of gold.

• Underground mining production of 7,482 kt at an average grade of 3.94 g/t containing

949 thousand ounces of gold.

• Open pit mining production of 9,920 kt at an average grade of 1.57 g/t containing 500

thousand ounces of gold.

• Stockpile processing of 1,205 kt at an average grade of 0.63 g/t containing 24 thousand

ounces of gold.

• A metallurgical process recovery of 95% yielding 1.4 million recovered ounces.

• Revenue based on a gold price of US$1,300/ounce.

• A mineral royalty of 5% of gross revenue.

• A stream payment equivalent to 8.4% of gross revenue.

• Total underground development capital costs estimated at $50 million.

• Total open pit development capital costs estimated at $109 million

• Total underground sustaining capital costs estimated at $65 million.

• Total open pit sustaining capital costs estimated at $32 million.

• All expenditures on the project prior to January 2019 are considered as sunk costs.

22.2 Taxes and Royalties

Golden Star holds a 90% interest in the Wassa Mine with the Government of Ghana holding a 10%

ownership interest. The Government of Ghana receives a gross revenue Mineral Royalty of 5% on

all gold production. RGLD received a stream payment equivalent to 8.4% of gross revenue.

The corporate tax rate on mining companies in Ghana is 35%.

22.3 Cash Flow Model and Project Economic Results

The Wassa Mine annual economic model is shown in Table 22-1.

The following post-tax economic indicators were calculated:

• Free cash flow $218 million

• NPV at 5% discount rate $175 million

• LoM Cash operating cost $671/oz

• LoM mine-site AISC $814/oz

June 2019


Table 22-1 Economic Model

June 2019


22.4 Sensitivity Analysis

The base case results are based on:

• Gold price $1,300/ounce

• Plant feed gold grade 2.46 g/t gold

• Total capital cost $255 million

• LoM total operating cost $1,046 million

A sensitivity analysis has been prepared varying these four inputs. Table 22-2 shows the impact

of varying input values on the base case pre-tax economic indicator – NPV5% in millions of dollars.

Figure 22-1 presents these sensitivities in graphical format.

Of these parameters, the project economics are most sensitive to gold price and gold grade

followed closely by operating cost. The operation is least sensitive to capital expenditure changes.

Table 22-2 NPV5% Sensitivity

Variable -30% -20% -10% Base 10% 20% 30%

Capex 219 204 190 175 161 146 131

Opex 339 285 230 175 120 64 8

Gold price and grade -85 15 96 175 254 332 419

Figure 22-1 NPV5% Sensitivity

-200

-100

0

100

200

300

400

500

-30% -20% -10% Base 10% 20% 30%

NP

V 5

% (

$M

)

Capex Opex Gold price and grade

June 2019


23 Adjacent Properties There is no relevant data regarding adjacent properties.

June 2019


24 Other Relevant Data and Information There is no other relevant data available.

June 2019


25 Conclusions and Recommendations

Geology and Mineral Resources

The mineral resources were estimated using block models. The composite grades are capped where

this is deemed necessary, after statistical analysis. OK is used to estimate the block grades. The

search ellipsoids are orientated to reflect the general strike and dip of the modelled mineralization.

Block model tonnage and grade estimates are classified according to the CIM Definition Standards

for Mineral Resources and Mineral Reserves (May 2014) by S.Mitchel Wasel, MAusIMM(CP),

of GSR, qualified person (not independent) as this term is defined in NI 43-101. The basis of the

mineral resource classification includes confidence in the geological continuity of the mineralized

structures, the quality and quantity of the exploration data supporting the estimates, and the

geostatistical confidence in the tonnage and grade estimates. Three-dimensional solids are

modelled reflecting areas with the highest confidence, which are classified as Indicated mineral

resources.

Further definition drilling and development could convert some of the existing underground

Inferred mineral resources to Indicated category. This will be a benefit for extending the mine life.

The underground deposit remains open for possible expansion at depth below the current planned

mine bottom, especially following the HG down plunge trend, which could increase the project

mineral resource base.

The Wassa Underground could contain opportunities to expand and/or extend underground

production in the future.

Risks:

• The complex geometries of the HG gold mineralization at Wassa requires tight spaced

drilling prior to mining. Failure to create drill access to delineate the stoping areas ahead

of mining would result in mining outside the ore zone, thus mining waste or LG material.

Recommended work programs:

• Exploration drilling continues at surface to define additional resources and to upgrade

Inferred resources to Indicated status. The drilling program includes surface drilling to

convert B shoot Inferred to Indicated resources. Some exploration drilling is also ongoing

at HBB. Underground, a hanging wall drive has been developed in 2018 and this is

providing the platforms from which the underground grade control drilling is conducted.

• There are some parameters that should be examined further as they could have a positive

effect on the global mineral resources. The first parameter is the shape of the long-range

wireframe compared to the short-range wireframe. The long-range wireframe shows a

smaller volume than the short-range wireframe for the same areas of the mine. Therefore,

the long-range model underestimates the tonnes for these areas. The other parameter that

requires further investigation is the HG capping used during block model grade estimation.

Model to mill reconciliations are reporting a higher grade to the plant than that which has

June 2019


been estimated by both the short range and long-range models. GSR will look at the HG

capping to determine what cap should be used for the HG to produce a block model grade

estimate which reconciles closer to the actual grades being achieved by the mill. Using a

higher grade cap in the resource estimation process will add more tonnes and metal to the

resources and subsequent reserve base. Both of these revisions cost very little money and

effort and could result in a significant increase in the overall ounces at Wassa.

Mineral Reserves

The reserve estimation process is carried out using state-of-the-art open pit and underground

optimization process, computer aided mine design and scheduling processes and are reported using

NI 43-101 standards by M. Raffield a non-independent, qualified person.

Mining

The underground mine reached commercial production in January 2017 and, since that time, has

shown consistent improvement in ore tonnage generation capacity. By the end of 2018, it was

producing at a rate of 3,500 tpd, with plans to increase to close to 4,000 tpd in 2020.

Geotechnical conditions in the mine are very good with consistently low rates of unplanned

dilution being achieved in the stopes and development.

Wassa has a long history of successful open pit operation and the mining of Cut 3 and the 242 pits

from 2020 are expected to be straightforward.

Risks:

• Flooding of the underground operation – significant effort has been expended to understand

and mitigate this risk. The pit catchment areas have been reduced through earthworks and

drainage diversions around the pit areas to ensure minimal water ingress into the pits from

the surrounding areas. The sump capacities in the pits below the holings into the

underground have been designed to contain 100-year storm events over 24 hour periods.

The pumping systems from the pits and underground are in the process of being upgraded

to ensure enough capacity is available to dewater the sumps and the underground mine

during and following such rain events.

Recommended work programs:

• Optimization of the stope and development designs for the reserve below the current

mining areas to ensure efficient flow of equipment, ventilation and ore/waste rock.

• Design and implementation of a paste backfill system which will provide support to mined

out stopes and enable an increased extraction ratio through the reduction of sill and rib

pillar requirements.

Metallurgy and Mineral Processing

Historical experience with both open pit and underground sourced ore from Wassa indicates that

the ore is free milling with a high recovery through the gravity and CIL circuit. The operating

methodology in 2019 is to run the plant at 3,500 tpd (capacity is 7,400 tpd) based primarily on

underground feed. In 2024, the plant will increase to 6,500 tpd as the ore from the open pit mining

becomes available.

June 2019


Infrastructure

The surface infrastructure is in good operational condition and has a long history of good operation

in supporting the open pit operations. The new underground infrastructure has operated well for

the past three years.

The new TSF is operating well and no issues are expected with the subsequent embankment raises.

The design TSF capacity exceeds the life of mine storage requirements.

Environmental and Social Management

The Wassa operations including underground mine and development, pit and waste dumps are

largely contained within existing footprints or compensated buffer areas.

The primary environmental approvals for the Wassa operations are the EPA Environmental

Permits and Certificate, and Mine Operating Plan with the Minerals Commission. The Mine

Operating Plan for 2019 has been submitted and approved. All required EPA Environmental

Permits have been invoiced and the most recent three-year EMP has been submitted as per

Regulations, reviewed by the EPA and is pending Environmental Certificate issuance.

GSR has an environmental and social management system developed along the lines of an

ISO 14001 management system. The management is carried out by Environmental and

Community Relations specialists. GSR has also established a series of Community Mine

Consultative Committees for on-going engagement of local communities.

Dedicated studies for the Wassa operations demonstrate no or low potential for acid drainage

generation and, overall, the geochemical impact of mining the Wassa underground is expected to

continue to be low. Mine water discharges consistently achieve EPA discharge criteria.

Capital and Operating Costs

Total capital of $255 million is comprised of $143 million of development capital (including $104

million of pre-stripping activities), $94 million of sustaining capital, $6 million contingency and

$12 million exploration.

Mine operating costs include:

• $37/t-ore underground stoping cost;

• $6.00/t-ore paste backfill cost;

• $3,400/m underground development cost;

• $3.35/t open pit mining costs;

• $15/t to $25/t processing costs depending on throughput; and

• $5/t to $9/t G&A costs depending on throughput.

Economics

The mine has been evaluated on a discounted cash flow basis. The cash flow analysis was prepared

on a constant 2019 US dollar basis. No inflation or escalation of revenue or costs has been

incorporated.

June 2019


Using a long-term gold price forecast of $1,300/oz, the post-tax free cash flow is $218 million

and post-tax NPV5% is $175 million.

Life of mine cash operating cost is estimated at $671 per ounce and mine-site all-in sustaining

cost at $814.

The NPV5% is most sensitive to changes in gold price, plant head grade and operating cost and

least sensitive to capital cost.

June 2019


26 References Allibone A., McCuaig T.C., Harris D., Etheridge M., Munroe S., Byrne D.; 2002a; Structural

Controls on Gold Mineralization at the Ashanti Gold Deposit, Obuasi, Ghana; Society

of Economic Geologists, Special Publication, Vol. 9, pp. 65–93.

Allibone A., Teasdale J., Cameron G., Etheridge M., Uttley P., Soboh A., Appiah-Kubi J., Adanu

A., Arthur R., Mamphey J., Odoom B., Zuta J., Tsikata A., Pataye F., Famiyeh S., Lamb

E.; 2002b; Timing and Structural Controls on Gold Mineralization at the Bogoso Gold

Mine, Ghana, West Africa; Economic Geology, Vol. 97, pp. 949-969.

Apau D.; 2012; Structural Evolution of the Wassa Gold Deposit, South Central Ghana; Implication

for Mineralization; University of Ghana, 112p (Unpublished).

Bonhomme M.;1962; Contribution à l’étude géochronologique de la plate-forme de l’Ouest

Africain; Ann. Fac. Sci. Univ. Clermont-Ferrand, Géol. Minéral., No. 5, 62 p.

Bourassa Y.; 2003; Geology of the Wassa Mine Report; Golden Star Resources Ltd, 32p

(Unpublished).

Bowell R.J., Foster R.P., Stanley C.J.; 1990; Telluride mineralization at Ashanti gold mine, Ghana;

Mineralogical Magazine, Vol. 54, pp. 617-627.

Crow A.T.; 1952; The rocks of the Sekondi Series of the Gold Coast; Gold Coast Geological

Survey, Bulletin No. 18, 68 p.

Dampare S.B., Shibata T., Asiedu D.K., Osae S., Banoeng-Yakubo B.; 2008; Geochemistry of

Paleoproterozoic metavolcanic rocks from the southern Ashanti volcanic belt, Ghana:

Petrogenetic and tectonic setting implications; Precambrian Research, Vol. 162, pp.

403-423.

Davis D.W., Hirdes W., Schaltegger U., Nunoo E.A.; 1994; U-Pb age constraints on deposition

and provenance of Birimian and gold-bearing Tarkwaian sediments in Ghana, West

Africa; Precambrian Research, Vol. 67, pp. 89-107.

Agyei Duodu J., Loh G.K., Boamah K.O., Baba M., Hirdes W., Toloczyki M., Davis D.W.; 2009;

Geological Map of Ghana 1 : 1,000,000; Geological Survey of Ghana, Map 1:1M.

Eisenlohr B.N.; 1992b; Conflicting evidence on the timing of mesothermal and paleoplacer gold

mineralization in early Proterozoic rocks from southwest Ghana, West Africa;

Mineralium Deposita, Vol. 27, pp. 23-29.

June 2019


Eisenlohr B.N., Hirdes W.; 1992a; The structural development of the early Proterozoic Birimian

and Tarkwaian rocks of southwest Ghana, West Africa; Journal of African Earth

Sciences, Vol. 14, No. 3, pp. 313-325.

Feybesse JL., Billa M., Guerrot C., Duguey E., Lescuyer JL., Milési JP., Bouchot V.; 2006; The

paleoproterozoic Ghanaian province: Geodynamic model and ore controls, including

regional stress modeling; Precambrian Research, Vol. 149, pp. 149-196.

Geosystems, 2013. Golden Star (Wassa) Limited; Tailings Storage Facility (TSF) 2 Project

Environmental Impact Statement, February 2013.

Ghana Statistical Service, 2012. Housing and Population Census. Summary Report of Final

Results. May, 2012.

Golden Star Resources, 2015. Wassa Expansion Project Environmental Scoping Report.

Golder Associates (2016). Wassa Expansion Project Environmental Impact Statement. November

2016.

Goodfellow W.D,; 2007; Mineral Deposits of Canada, A Synthesis of Major Deposit Types,

District Metallogeny, the Evolution of Geological Provinces and Exploration Methods;

Geological Association of Canada, Mineral Deposits Division, Special Publication

No.5.

Hall, J.B. and M.D. Swaine (1981). Distribution and Ecology of Vascular Plants in a Tropical Rain

Forest. W. Junk, The Hague.

Hirdes W., Davis D.W.; 1998; First U-Pb zircon age of extrusive volcanism in the Birimian

Supergroup of Ghana/West Africa; Journal of African Earth Sciences, Vol. 27, No. 2,

pp. 291-294.

Hirdes W., Davis D.W., Eisenlohr B.N.; 1992; Reassessment of Proterozoic granitoid ages in

Ghana on the basis of U/Pb zircon and monazite dating; Precambrian Research, Vol. 56,

pp. 89-96.

Hutchinson D.J and Diedrichs M.S. 1996. Cablebolting in Underground Mines, Vancouver BiTech

Publishers, pp406

Imperial Chemical Industries, 1971. Handbook of Blasting Tables. ICI (1971).

INAP. 2010. Global Acid Rock Drainage Guide (the GARD Guide), Version 0.8. The International

Network for Acid Prevention, http://www.gardguide.com.

June 2019


International Union for Conservation of Nature and Natural Resources, 2016. www.iucnredlist.org

(IUCN, 2016).

Isaaks E.; 2013; Grade Estimation for the Wassa Resource Model; Independent Mineral

Consultant, 21p (Unpublished).

John T., Klemb R., Hirdes W., Loh G.; 1999; The metamorphic evolution of the Paleoproterozoic

(Birimian) volcanic Ashanti belt (Ghana, West Africa); Precambrian Research, Vol. 98,

pp. 11-30.

Junner N.R.; 1935; Gold in the Gold Coast; Gold Coast Geological Survey, Memoir No. 4, 76 p.

Junner N.R.; 1940; Geology of the Gold Coast and Western Togoland; Gold Coast Geological

Survey, Memoir No. 11, 40 p.

Kesse G.O.; 1985; The Mineral and Rock Resources of Ghana; A.A. Balkema Publishers, Book,

610 p.

Kitson, A.E.; 1928; Provisional geological map of the Gold Coast and Western Togoland, with

brief descriptive notes thereon; Gold Coast Geological Survey; Bulletin No. 2.

Knight Piésold; 2011; Wassa Tailings Storage Facility 2, Site Investigation Factual Report,

December 2011.

Knight Piésold; 2012; Golden Star Wassa Limited, Wassa Gold Mine, TSF Detailed Design

Report.

Knight Piésold; 2015; Golden Star Wassa Limited, Wassa Gold Mine, TSF Detailed Design

Report.

Knight Piésold; 2017; Conceptual Level Altenrative Staging Design of TSF 2, for Annualised

Construction with Compacted Soil Liner (CSL); Memorandum; October 2017.

Ledru P., Johan V., Milési J.P., Tegyey M.; 1994; Markers of the last stages of the

Palaeoproterozoic collision: evidence for a 2 Ga continent involving circum-South

Atlantic provinces; Precambrian Research, Vol. 69, pp. 169-191.

Leube A., Hirdes W., Mauer R., Kesse G.O.; 1990; The Early Proterozoic Birimian Supergroup

of Ghana and Some Aspects of its Associated Gold Mineralization; Precambrian

Research, Vol. 46, pp. 139-165.

Loh G., Hirdes W., Anani C., Davis D.W., Vetter U.; 1999; Explanatory Notes for the Geological

Map of Southwest Ghana 1 : 100,000; Geologisches Jahrbuch, Reihe B, Heft 93, 150 p.

http://www.iucnredlist.org/

June 2019


MEL (1996 a). Satellite Goldfields Limited, Wassa Hydrogeological Assessment, Progress Report

for Work During First Quarter 1996. MEL Report 1031R087.

MEL (1996 b). Reinterpretation of VLF Data to Locate Zones of Preferential Groundwater Flow.

MEL Report 1031R138.

MEL (1996 c). Satellite Goldfields Limited, Wassa Hydrogeological Assessment, Progress Report

for Work During Third Quarter 1996. Minerex Environmental Limited, December 1996.

MEL Report 1031R156.

MEND, 2009. Prediction Manual for Drainage Chemistry from Sulphidic Geologic Materials.

Report prepared by W.A. Price, CANMET, British Columbia, for the MEND Program.

Milesi J.P., Ledru P., Ankrah P., Johan V., Marcoux E., Vinchon Ch.; 1991; The metallogenic

relationship between Birimian and Tarkwaian gold deposits in Ghana; Mineralium

Deposita, Vol. 26, pp. 228-238.

Milési JP., Ledru P., Feybesse JL., Dommanget A., Marcoux E.; 1992; Early Proterozoic ore

deposits and tectonics of the Birimian orogenic belt, West Africa; Precambrian

Research, Vol. 58, pp. 305-344.

Morin, K., Hutt, N. 2007. Morrison Project - Prediction of Metal Leaching and Acid Rock

Drainage, Phase 1. Minesite Drainage Assessment Group, 588p.

Mumin A.H., Fleet M.E.; 1995; Evolution of gold mineralization in the Ashanti Gold Belt, Ghana:

evidence from carbonate compositions and parageneses; Mineralogy and Petrology,

Vol. 55, pp. 265-280.

Mumin A.H., Fleet M.E., Chryssoulis S.L.; 1994; "Gold mineralization in As-rich mesothermal

gold ores of the Bogosu-Prestea mining distric of the Ashanti Gold Belt, Ghana:

remobilization of ""invisible"" gold"; Mineralium Deposita, Vol. 29, pp. 445-460.

Nicholas D.E. (1981). Method Selection – A Numerical Approach. Design and Operation of

Caving and Sub-Level Stoping Mines. New York. AIME Chapter 4.

Oberthür T., Vetter U., Davis D.W., Amanor J.A.; 1998; Age constraints on gold mineralization

and Paleoproterozoic crustal evolution in the Ashanti belt of southern Ghana;

Precambrian Research, Vol. 89, pp. 129-143.

Oberthür T., Vetter U., Schmidt Mumm A., Weiser T., Amanor J.A., Gyapong W.A., Kumi R.,

Blenkinsop T.G.; 1994; The Ashanti Gold Mine at Obuasi, Ghana: Mineralogical,

Geochemical, Stable Isotope and Fluid Inclusion Studies on the Metallogenesis of the

Deposit; Geologisches Jahrbuch, D 100, pp. 31-129.

June 2019


Oberthür T., Weiser T., Amanor J.A., Chryssoulis S.L.; 1997; Mineralogical siting and distribution

of gold in quartz veins and sulfide ores of the Ashanti mine and other deposits in the

Ashanti belt of Ghana: genetic implications; Mineralium Deposita, Vol. 32, pp. 2-15.

Perrouty S., Aillères L., Jessell M.W., Baratoux L., Bourassa Y., Crawford B.,; 2012; Revised

Eburnean geodynamic evolution of the gold-rich southern Ashanti Belt, Ghana, with

new field and geophysical evidence of pre-Tarkwaian deformations, Vol. 204-205, pp.

12-39.

Perrouty S., Jessell M.W., Aillères L., Apau D., Velasquez G., Siebenaller L., Miller J., Bourassa

Y., Beziat d., Baratoux L., 2012; Tectonic Context of Eoeburnean Gold Mineralization

in Wassa mine, Southwest Ghana, unpublished.

Pigois JP., Groves D.I., Fletcher I.R., McNaughton N.J., Snee L.W.; 2003; Age constraints on

Tarkwaian palaeoplacer and lode-gold formation in the Tarkwa-Damang district, SW

Ghana; Mineralium Deposita, Vol. 38, pp. 695-714.

Price W.A., Morin K., Hutt N., 1997. Guidelines for prediction of acid rock drainage and metal

leaching for mines in British Columbia: Part II. Recommended procedures for static and

kinetic tests. In: Proceedings of the Fourth International Conference on Acid Rock

Drainage. Vancouver, B.C. Canada, 1, pp. 15–30.

Reclamation Security Agreement between Wexford Goldfields Limited and Environmental

Protection Agency, 14 November 2004.

SGS, 1996. Satellite Goldfields Limited, Wassa Gold Project, Environmental Baseline Study. SGS

Laboratory Services (Ghana) Limited, November 1996.

SGS Laboratory Services Ghana Limited, 1998. Wassa Project Environmental Impact Statement

for Satellite Goldfields Limited.

SGS, 2002. Wassa Environmental due diligence audit. SGS Laboratory Services (Ghana) Limited.

Soregaroli, B.A., Lawrence, R.W., 1997. Waste Rock Characterization at Dublin Gulch: A Case

Study. Proceedings of the Fourth International Conference on Acid Rock Drainage,

Vancouver, B.C. Canada, p631-645.

SRK Consulting (UK) Limited 2013. NI 43-101 Technical Report on Mineral Resources and

Mineral Reserves Golden Star Resources Ltd, Wassa Gold Mine, Ghana Effective Date

31st December 2012.

June 2019


Tunks A.J., Selley D., Rogers J.R., Brabham G.; 2004; Vein mineralization at the Damang Gold

Mine, Ghana: controls on mineralization; Journal of Structural Geology, Vol. 26, pp.

1257-1273.

NDPC & UNDP (2010). 2008 Ghana Millenium Development Goals Report. April 2010. National

Development Planning Commission (NDPC) / Government of Ghana and the United

Nations Development Programme (UNDP) Ghana.

United States Bureau of Mining, 1980. RI 8507 Structure Response and Damage Produced by

Ground Vibration from Surface Mine Blasting. DE Siskind, M.S. Stagg, J. W. Kopp and

C. H. Dowding.

Wexford Goldfields Limited (WGL), 2004. Environmental Impact Statement for the Wassa

Project.

Whitelaw O.A.L.; 1929; The Geological and Mining Features of the Tarkwa-Abosso Goldfield;

Gold Coast Geological Survey, Memoir No. 1, 46 p.

June 2019


27 Date and Signatures The effective date of this technical report is December 31, 2018.

[“Signed and sealed”]

Martin Raffield, PhD, PEng


Mitch Wasel, MAusIMM (CP)


Philipa Varris, MAusIMM (CP)

Dated June 20, 2019

royalgold.compage ii june 2019 ni 43-101 technical report on resources and reserves, golden star...

Documents