NI TECHNICAL REPORT NATOUGOU GOLD DEPOSIT PROJECT BURKINA FASO

Transcription

1 NI TECHNICAL REPORT NATOUGOU GOLD DEPOSIT PROJECT BURKINA FASO Prepared by Lycopodium Minerals Canada Ltd in accordance with the requirements of National Instrument , Standards of Disclosure for Mineral Project, of the Canadian Securities Administrators Qualified Persons: Neil Lincoln, P.Eng, VP Business Development & Studies, Lycopodium Minerals Canada Ltd. Marius Phillips, MAusIMM (CP), Principal Process Engineer, Lycopodium Minerals Canada Ltd. Glen Williamson, Principal Mining Engineer, AMC Consultants (Canada) Ltd John Graindorge, MAusIMM (CP), Principal Consultant Applied Geosciences, Snowden Mining Industry Consultants Jean-Sébastien Houle, ing, Project Director, Mining Environment, WSP Canada Timothy Rowles, MAusIMM (CP), Knight Piésold Consulting Submitted to: SEMAFO Inc 100 Boulevard Alexis-Nihon Saint-Laurent, Quebec H4M 2P3 Canada File Location: Rev: FINAL ISSUE NL IM NL REV NO. DATE REVIEW BY DESIGN APPROVED Lycopodium Minerals Canada, 5060 Spectrum Way, Suite 400, Mississauga, Ontario L4W 5N5 PROJECT APPROVED

2 NI TECHNICAL REPORT DATE & SIGNATURE PAGE Project Name: Title of Report: Location: Natougou Gold Deposit Project NI Technical Report Burkina Faso Effective Date of Report: 23 [SIGNED] DATE Neil Lincoln, P.Eng. (Lycopodium Minerals Canada Ltd.) 23 Marius Phillips, MAusIMM (CP) (Lycopodium Minerals Canada Ltd.) 23 Glen Williamson, ( AMC Consultants (Canada) Ltd) 23 John Graindorge, MAusIMM (CP) (Snowden Mining Industry Consultants) 23 Jean-Sébastien Houle, ing, (WSP Canada) 23 Timothy Rowles, MAusIMM (CP), Knight Piésold Consulting 23 Lycopodium Minerals Canada Ltd

3 NI TECHNICAL REPORT Table of Contents 1.0 SUMMARY Overview Accessibility, Climate, Local Resources, Infrastructure and Physiography Geology and Mineral Resources Exploration Drilling, Sampling and Assaying Quality Assurance and Quality Control Mineral Processing and Metallurgical Testwork Mineral Resource Estimate Mineral Reserves Mining Recovery Methods Infrastructure Environmental, Permitting and Social or Community Impact Operating and Capital Costs Economic Analysis INTRODUCTION Terms of Reference and Purpose of this Report Site Visits RELIANCE ON OTHER EXPERTS PROPERTY DESCRIPTION AND LOCATION Type of Mineral Tenure Boungou Permit ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY Access and Physiography Climate Infrastructure HISTORY GEOLOGICAL SETTING AND MINERALIZATION Regional geology Eburnian metamorphics and intrusives Tarkwaian Sequence Property Geology Natougou Geology Oxidation Lithology Alteration Mineralization DEPOSIT TYPE EXPLORATION 9.1 Lycopodium Minerals Canada Ltd

4 NI TECHNICAL REPORT 10.0 DRILLING Sampling techniques RC drillhole sampling Diamond drillhole sampling Drill sample recovery Logging Location data and survey methods Data Spacing and Distribution Location of Data Points Downhole Surveys Orientation of data in relation to geological structure Topography SAMPLE PREPARATION, ANALYSES AND SECURITY Sample Preparation Analytical techniques Primary Analysis Handheld XRF Analysis Quality of Assay Data and Laboratory Tests Internal Laboratory Quality Assurance and Quality Control Laboratory Inspections Assay Reports Certified Reference Materials Blanks Field Duplicates Umpire Laboratory Checks Sample Security Bulk Density Qualified Person s Opinion on Adequacy of Sampling DATA VERIFICATION RC versus Diamond Comparisons Assay Data Validation Qualified Person s Opinion MINERAL PROCESSING AND METALLURGICAL TESTING Introduction Geological Background Sample Selection for Detailed Testwork Program Samples for Detailed Comminution Testwork Samples for Detailed Metallurgical Testwork Samples for Variability Metallurgical Testwork Detailed Testwork Program Outline Comminution Testwork Metallurgical Testwork Borehole Bench Composites Metallurgical Testwork Master Composites Ore Mineralogy Head Analysis Gravity Concentration Cyanide Leach Grind / Recovery Testwork Flotation Grind Optimization Study Oxygen and Cyanide Investigations Lycopodium Minerals Canada Ltd

5 NI TECHNICAL REPORT Lead Nitrate Additions Leach Pulp Density Direct Leach vs Gravity Concentration Oxygen Optimization Testwork Residence Time Selection Bulk Leach Carbon Loading Kinetic Tests Carbon Equilibrium Loading Intensive Cyanidation of Gravity Concentrate Ancillary Testing Rheology Oxygen Uptake Testing Pre-leach and Tailings Thickening Testwork Tailings Filtration Testwork Variability Metallurgical Testwork Head Analysis Gravity / Leach Testwork Gravity / Leach Testwork Diagnostic Leach Metallurgical Testwork Summary Selected Treatment Route Key Process Design Parameters Metallurgical Recoveries Reagent Consumption Testwork Conclusions MINERAL RESOURCE ESTIMATES Geological Interpretation Drill sample analysis Sample Compositing Statistical Analysis Top Cut Analysis Unfolding Introduction Principles of Unfolding Creation of Strings for Unfolding Variography Gold In situ bulk density Block modelling and grade estimation Kriging neighbourhood analysis Volume model construction Block model coding Search Neighbourhood Parameters Grade estimation methodology Model validation Mineral Resource classification Mineral Resource reporting Cut-off grade Moisture Natougou Mineral Resource statement Grade-tonnage curve MINERAL RESERVE ESTIMATES 15.1 Lycopodium Minerals Canada Ltd

6 NI TECHNICAL REPORT 15.1 Introduction Method Mining Cut-off Grade Dilution and Mining Recovery Factors Mine Design Items Impacting on the Mineral Reserves MINING METHODS General description Pit dewatering Geotechnical parameters Rock strength Geological Input into Geotechnical Model Geotechnical Model Pit wall design Resource Model for Mining Mining Method Load and Haul Drill and blast Stockpile Rehandling Ancillary Equipment Mine Design Open Pit Optimization Pit design Layout of other mining related facilities Strategic mining schedule Detailed schedule Construction activity Schedule results Production Schedule Risks Equipment and Personnel Numbers Additional Information RECOVERY METHODS Process Design Selected Process Flowsheet Key Process Design Criteria Process and Plant Description Introduction Ore Receiving and Crushing Coarse Ore Storage Grinding and Classification Pebble Crushing Gravity Concentration Pre-Leach Thickening Leach Circuit Carbon Absorption Circuit Desorption and Carbon Regeneration Electrowinning and Gold Room Tailings Thickening Reagents Mixing and Storage Water Services Air and Oxygen Services Lycopodium Minerals Canada Ltd

7 NI TECHNICAL REPORT 17.3 Plant Consumption Water Consumption Energy Consumption Reagent and Consumable Consumption Plant Control System General Overview Control System Configuration and Communications Drive Controls PROJECT INFRASTRUCTURE Overall Site Roads Access to Site Project Site Roads Access Tracks Power Power Supply Electrical Distribution Electrical Buildings Transformers and Compounds Fuel Supply Potable Water Sewage and Solid Waste Management Sewage Treatment Solid Wastes Accommodation Camps Mine/Plant Site Facilities General Outside Plant Area Inside Plant Area Mine Services Process Plant Geotechnical Water Supply Infrastructure Water demands Decant from Tailings Storage Facility Pit dewatering Groundwater sources Surface water sources and storages Tailings Storage Facility Design objectives and site selection Facility design Construction Staging and Summarized quantities Surface Water & Sediment Management Anticipated water quality Management system design MARKET STUDIES AND CONTRACTS Market Studies Contracts ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Legal Framework and Permits to Obtain 20.1 Lycopodium Minerals Canada Ltd

8 NI TECHNICAL REPORT Policies and Strategies for Environmental Protection Legal Framework Mining Code Institutional Framework Permits to Obtain Baseline Studies Baseline Studies Conducted Description of the Main Environmental and Social Components Community Information and Consultation Program Project Impacts, Risk Analysis, Environmental and Social Management Plan Project Impacts Risk Analysis Environmental and Social Management Plan Resettlement People and Activities Affected by the Project Scope of Resettlement Resettlement Action Plan Budget Acid Rock Drainage Waste Rock and Construction Materials Tailings Waste Disposal and Sanitary Wastewater Solid Waste Hazardous Waste Sanitary Wastewater Closure, Decommissioning and Reclamation CAPITAL AND OPERATING COSTS Operating Costs Mining Operating Costs Plant and Administration Costs Power Operating Consumables Labour (Processing / Maintenance and Administration) General and Administration Cost (excluding G&A labour) Maintenance Exclusions Capital Cost Estimate Summary Estimating Currency and Base Date Mining Capital Expenditure Estimating Methodology Quantity Development Pricing Basis Field In-directs Engineering Services Owner s Costs Spares Duties, taxes and Insurances Contingency Deferred Capital Escalation Qualifications and Assumptions Lycopodium Minerals Canada Ltd

9 NI TECHNICAL REPORT Exclusions Deferred Capital Sustaining Capital Costs Project Implementation Strategy ECONOMIC ANALYSIS Introduction Taxes Value Added Tax Royalties Other Royalties / Agreements Revenue Deductions Project Financing Financial Model Inputs Key Economic Assumptions Production summaries Capital Cost Summary Operation Cost Summary Economic Results Cash Flow Sensitivities ADJACENT PROPERTIES OTHER RELEVANT DATA AND INFORMATION INTERPRETATION AND CONCLUSIONS Geology, Exploration and Mineral Resource Mining Metallurgy Environmental and Permitting Risks Resettlement Delays RECOMMENDATIONS Geology and Mineral Resources Mining Environmental and Permitting Overall Recommendation REFERENCES 27.1 Lycopodium Minerals Canada Ltd

10 NI TECHNICAL REPORT TABLES Table 1.1 Natougou Mineral Resource as at January 2016, reported above a 0.83 g/t Au cutoff grade Table 1.2 Mineral Reserves estimate as of 31 December Table 1.3 Mining Schedule Summary Table 1.4 Average LOM Operating Costs Table 1.5 Initial Capital Costs Table 1.6 Economic Analysis Summary Table 2.1 Persons Who Prepared this Technical Report Table 2.2 Persons Who Contributed to this Technical Report Table 4.1 Tapoa Permit Group Exploration Permits Table 4.2 Boungou permit boundary (UTM Clarke 1880 Datum: Adindan, Zone 31N) Table 5.1 Average monthly rainfall for Fada n Gourma Table 10.1 Tapoa Permit Group drillhole details Table 11.1 Innov-X Omega series elements and detection limits Table 11.2 CRM Details for Natougou Table CRM results Assayed at SGS Table 11.4 CRM results assayed at ALS Table 13.1 Major Comminution Domains Table 13.2 Comminution Testwork Summary Table 13.3 Bottle Roll Cyanide Leach Results for Borehole Bench Composites Table 13.4 Detailed Head Analysis of Master Composites Table 13.5 Summary of Grind Recovery Leach Results Table 13.6 Summary of Rougher Flotation Tests With and Without Gravity Table 13.7 Flotation Vs WOO Leach Financial Analysis Table 13.8 Grind Optimization Financial Analysis Table 13.9 Oxygen and Cyanide Investigation Test Matrix Table Residence Time Selection Gold Recoveries at Various Head Grades Table Residence Time Selection Financial Analysis Table Summary of Bulk Leach Test Results and Optimized Test Results Table Detailed Solution Analysis from Bulk Leaches Table Cyanide Speciation from Bulk Leaches Table MC1 Results of Sequential Triple Contact CIP Testwork Table MC2 Results of Sequential Triple Contact CIP Testwork Table Loaded Carbon Assays from Triple Contact Tests Table Carbon Equilibrium Loading Table GRG and Intensive Leach Recoveries Table egrg Concentrate Grades Table Slurry Rheology for Master Composites Table Outotec Thickener Test Results Table FLS Thickener Test Results Table Outotec Tailings Filtration Test Results P80 of 63 microns Table FLS Tailings Filtration Design Parameters Table Detailed Head Analysis of Variability Samples Section 1 & Section 2 (1, 3 to 11) Table Detailed Head Analysis of Variability Samples Section 2 (12 to 24, 27, 29) Table Detailed Head Analysis of Variability Samples Section 3 (1 to 15) Table Detailed Head Analysis of Variability Samples Section 4 (1 to 6, 8 to 13) Table Detailed Head Analysis of Variability Samples Section 5 (1 to 20) Table Detailed Head Analysis of Variability Samples Section 5 (21 to 29), Section Table Detailed Head Analysis of Variability Samples Section Table Bottle Roll Cyanide Leach Results for Section 1 Composites Lycopodium Minerals Canada Ltd

11 NI TECHNICAL REPORT Table Bottle Roll Cyanide Leach Results for Section 2 Composites Table Bottle Roll Cyanide Leach Results for Section 3 Composites Table Bottle Roll Cyanide Leach Results for Section 4 Composites Table Bottle Roll Cyanide Leach Results for Section 5 Composites (1 to 15) Table Bottle Roll Cyanide Leach Results for Section 5 Composites (16 to 29) Table Bottle Roll Cyanide Leach Results for Section 6 Composites Table Bottle Roll Cyanide Leach Results for Section 0 Composites (Oxide) Table Reagent Consumptions in Master Composite and Variability Tests Table Predicted Plant Reagent Consumptions Table 14.1 Summary Statistics for Gold Composites Table 14.2 Summary Statistics for bulk density samples Table 14.3 Relationship between the variogram modelling software coordinate fields and the unfold coordinate fields Table 14.4 Indicator variogram thresholds and variogram model mapping Table 14.5 Indicator variogram models for the main mineralized domain Table 14.6 Distribution tail modelling Table 14.7 Density back-transformed variogram model parameters Table 14.8 Block model prototype settings Table 14.9 MINZONE field coding Table OXIDE field coding Table HWZONE field coding Table ROCK field coding Table Search neighbourhood parameters Table Default gold grade values for un-estimated blocks Table Default density values for un-estimated blocks Table Statistical validation comparing input sample data against block model estimates Table Pit optimization parameters for Mineral Resource definition pit shell Table Resource classification model field values Table Table Natougou Mineral Resource as at January 2016, reported above a 0.83 g/t Au cutoff grade Grade-tonnage report at various cut-offs for Natougou resource estimate, as at January 2016 (exclusive of Mineral Reserves) Table 15.1 Mineral Reserves Estimate as of 31 December Table 15.2 Cut-off grade input parameters and calculations for Mineral Reserves Table 15.3 Mining Dilution and Recovery Parameters Table 16.1 Laboratory Rock Strength Tests Table 16.2 Typical strength estimates for rock mass zones Table 16.3 Slope design guidelines for the AMC proposed final pit Table 16.4 Drill and blast parameters Table 16.5 Open Pit High Grade Ore Cut-off Calculation Table 16.6 Open Pit Low Grade Ore Cut-off Calculation Table 16.7 Flitch Height Selection Schedule Results Table 16.8 Mining Dilution and Ore Loss Parameters Table 16.9 Pit Optimization Results Table Ore and waste tonnage within the three ultimate pit designs Table Out-of-pit waste dump storage capacity Table Waste dump rehabilitation parameters Table Strategic schedule - tonne and grade summary by year Table Grade Bin definition Table Low Grade and High Grade Cut-off Grades Table Schedule Milestones Table TSF Wall Construction Requirements Table Schedule summary Ex-Pit movement by year Lycopodium Minerals Canada Ltd

12 NI TECHNICAL REPORT Table Mill feed schedule by year Table Topsoil ex-pit and construction material by year Table Waste movement by destination by year Table Open pit equipment requirements at peak production Table Open pit workforce requirements Table 17.1 Summary of Key Process Design Criteria Table 17.2 Average Power Demand Summary Table 17.3 Annual Reagent and Major Consumable Consumption Table 18.1 Potable Water Usage Table 18.2 TSF Construction Staging and Major Quantities Table 20.1 List of Required Permits Table 20.2 List of environmental and social baseline studies conducted as part of the ESIA process Table 20.3 Resettlement Action Plan Budget (from SOCREGE, 2015) Table 20.4 Mine Closure Budget Table 21.1 Mining operating costs by year* Table 21.2 Mining operating cost by activity Table 21.3 Process Plant Operating Cost Summary Table 21.4 Process Plant Power Cost by Plant Area Table 21.5 Process Plant Consumables Cost by Plant Area Table 21.6 Plant General and Administration Summary Table 21.7 Plant Maintenance Materials Cost Table 21.8 Capital Estimate Summary by Area (4Q15, ±15%) Table 21.9 Capital Estimate Summary by Discipline (4Q15, ±15%) Table Foreign Currency Exposure Table Mining Capital Expenditures Table Derivation of Quantities Table Sources of Pricing Table Standard Direct Labour Gang Rates Table Deferred Capital Summary Table Sustaining Capital Schedule Table 22.1 Key Assumptions Used in the Financial Model Table 22.2 Economic Results Table 22.3 Cash Flow Table 22.4 Gold Price Sensitivity FIGURES Figure 4.1 Location map Figure 4.2 Tapoa Permit Group Figure 4.3 Boungou permit boundary Figure 5.1 Natougou access routes Figure 5.2 Climate information for Ouagadougou Figure 5.3 Photo overlooking the Natougou project area (photo taken March 2015, looking southeast) Figure 7.1 Overview of Regional Geology of Birimian Province Figure 7.2 Property Geological Map Tapoa Permits within Diapaga Belt Figure 7.3 Property Geological Map Boungou Permit Figure 7.4 Oxidation Types Figure 7.5 Photo of typical oxidation profile in core (BODD484) from Natougou Figure 7.6 Long section (looking northeast) showing oxidation impact on mineralization Figure 7.7 Lithology Types Figure 7.8 Boundou Shear Zone at Contact between Hangingwall and Footwall Volcanic Units Lycopodium Minerals Canada Ltd

13 NI TECHNICAL REPORT Figure 7.9 REE geochemical signatures for Natougou volcanic rocks Figure 7.10 Geochemical Signature (Y+Nb vs Rb) of Natougou Felsic Rocks Figure 7.11 REE geochemical signatures for Natougou Felsic Rocks Figure 7.12 Alteration types Figure 7.13 Core showing BPK and BBZ alteration Figure 7.14 Natougou deposit global geometry Figure 7.15 Cross-section (A-B; top, looking northwest) and long section (C-D; bottom, looking northeast) through Natougou deposit Figure 7.16 Cross-section Showing en Echelon Offset in Northwest Portion of Deposit Figure 7.17 Cross-section showing minor mineralization within the footwall below the main shear Figure 7.18 Photo of visible gold in core from Natougou Figure 9.1 Tapoa Permit Group exploration soil anomaly Figure 9.2 Tapoa Permit Group exploration rock chip samples >0.1 g/t Au Figure 10.1 Drillhole collar location plan for the Tapoa Permit Group Figure 10.2 Grade x thickness contour plot of Natougou deposit Figure 10.3 Histogram of core recoveries Figure 10.4 Histogram of RC drilling recoveries Figure 11.1 CRM Control Chart (z-score) Results Assayed at SGS Figure 11.2 Individual CRM Control Chart Results Assayed at SGS Figure 11.3 CRM Control Chart (z-score) results assayed at ALS Figure 11.4 Individual CRM Control Chart results assayed at ALS Figure 11.5 Coarse Blank Control Chart results assayed at SGS Figure 11.6 Coarse Blank Control Chart results assayed at ALS Figure 11.7 Original vs. Field Duplicate log QQ plot (SGS) Figure 11.8 Original vs. Field Duplicate Relative diff vs mean plot (SGS) Figure 11.9 Original vs. Field Duplicate Precision Plot (SGS) Figure Umpire (SMF BF) vs. Original Pulp (SGS) duplicate log QQ plot Figure Umpire (SMF BF) vs. Original Pulp (ALS) Duplicate log QQ plot Figure Umpire (SMF BF) vs. Original Pulp (SGS) Duplicate Precision Plot Figure Umpire (SMF BF) vs. Original Pulp (ALS) Duplicate precision plot Figure Bulk density measurement equipment (photo taken March 2015) Figure Bulk density vs depth below surface Figure 12.1 Drill section over the 20 m by 20 m grid showing RC holes, multi-purpose holes with core portion highlighted in grey and diamond holes Figure 12.2 Plan showing comparison area with equal RC (red) and diamond (blue) coverage... Figure QQ plot comparing diamond (X-axis) against RC (Y-axis) samples within the Comparison area (red line = mean; dashed blue lines are the 25 th, 50 th and 75 th percentiles) Figure 13.1 Cross Section of Natougou Deposit Figure 13.2 Location of Comminution Samples Figure 13.3 Spatial Location of Master Composite Samples Figure 13.4 Location of Variability Samples Figure 13.5 Mineral Abundance for Natougou Master Composites Figure 13.6 MC1 Grind Sensitivity Cyanide Leach Curves Figure 13.7 MC2 Grind Sensitivity Cyanide Leach Curves Figure 13.8 MC1 Grind Sensitivity Grade Recovery Curves Figure 13.9 MC2 Grind Sensitivity Grade Recovery Curves Figure MC1 Oxygen Investigation Tests with Cyanide Concentration 0.035% w/v Figure MC1 Oxygen Investigation Tests with Cyanide Concentration 0.05% w/v Figure MC2 Oxygen Investigation Tests with Cyanide Concentration 0.035% w/v Figure MC2 Oxygen Investigation Tests with Cyanide Concentration 0.05% w/v Figure MC1 Lead Nitrate Tests Lycopodium Minerals Canada Ltd

14 NI TECHNICAL REPORT Figure MC2 Lead Nitrate Tests Figure MC1 Oxygen Vs Air with Lead Nitrate Figure MC2 Oxygen Vs Air with Lead Nitrate Figure MC1 Leach Pulp Density Series Figure MC2 Leach Pulp Density Series Figure MC1 Tests with and without Gravity at Optimum Conditions Figure MC2 Tests with and without Gravity at Optimum Conditions Figure MC1 Tests with Varying DO Levels Figure MC2 Tests with Varying DO Levels Figure MC1 Bulk Leach and Optimized Leach Test Results Figure MC2 Bulk Leach and Optimized Leach Test Results Figure Rate Curves for Sequential Triple Contact CIP Testwork Figure Oxygen Uptake Rate Figure Gravity Recovery vs. Overall Gold Recovery Figure Selected Variability Leach Curves Arsenic, Sections 1 and Figure Selected Variability Leach Curves Arsenic, Sections 5 and Figure Selected Variability Leach Curves Copper, Sections 1 and Figure Selected Variability Leach Curves Copper, Sections 5 and Figure Selected Variability Leach Curves Mercury Figure Residue Grade Versus Head Grade Figure 14-1 Isometric view showing geological interpretation for the main mineralized structure Figure 14-2 Collar location plan for drilling at Natougou up to October Figure 14-3 Log histogram and log probability plot for gold for main mineralized domain (top) and minor footwall mineralized domain (bottom) Figure 14-4 Box and Whisker Plot of Density Measurements Figure 14-5 Schematic Illustration of the unfolding process in Datamine Figure 14-6 Plan View of the Unfold Control Strings Figure 14-7 Isometric view of the unfold strings looking from the south-southeast Figure 14-8 Comparison between the original data (left) and unfolded data (right) Figure 14-9 Median (50%) indicator variogram model Figure Example oblique section (looking northwest) showing MINZONE field coding Figure Example oblique section (looking northwest) showing OXIDE field coding Figure Example oblique section (looking northwest) showing HWZONE field coding Figure Example oblique section (looking northwest) showing ROCK field coding Figure Grade trend plots Au, main mineralized domain (source: Snowden) Figure Grade trend plots density, main mineralized domain (source: Snowden) Figure Example oblique section (looking northwest) showing block MIK gold estimates Figure Plan view showing Natougou Mineral Resource classification scheme Figure Grade-Tonnage Curve Figure 15.1 Final Pits Geometry Figure 16.1 Recommended pit slope design guidelines Figure 16.2 Ore vertical thickness histogram Figure 16.3 Pit Optimization Results Figure 16.4 Pit optimization sensitivity analysis Figure 16.5 Pit designs overview Figure 16.6 West pit design Stage Figure 16.7 West Pit Design Stage Figure 16.8 West Pit Design Stage Figure 16.9 West Pit Design Stage Figure West Pit Design Stage Figure West Pit Section A1 A Lycopodium Minerals Canada Ltd

15 NI TECHNICAL REPORT Figure West Pit Section B1 B Figure North East Pit - Stage Figure North East Pit - Stage Figure North East Section A1 A Figure North East Section B1 B Figure South East Pit Stage Figure South East Pit Stage Figure South East Pit Section A1 A Figure South East Pit Section B1 B Figure Pit contribution by year (tonnes mined) Figure Strategic mine plan mill feed Figure Strategic mine plan stockpile balance Figure Strategic mine plan recovered gold ounces Figure Site layout Figure Closing stockpile balance by period Figure Annual recovered gold ounces Figure Annual total ex-pit movement Figure Annual plant feed Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure 17.1 Overall Process Flow Diagram Figure 18.1 Overall Site Plan Figure 18.2 Water Balance Schematic Flow Diagram Figure 18.3 Location of Surface Water Sources and Storages Figure 18.4 Location of Tailings Storage Facility Figure 18.5 Layout of Underdrainage and Sub-liner Drainage Network Figure 18.6 Tailings Storage Facility Embankment Figure 18.7 Surface Water Management System Figure 20-1 Classified and protected areas near the project area Figure 22.1 NPV Sensitivity Analysis (Pre-tax) Figure 22.2 IRR Sensitivity Analysis (Pre-tax) Figure 23.1 Adjacent permit boundaries Lycopodium Minerals Canada Ltd

16 NI TECHNICAL REPORT Lycopodium Minerals Canada Ltd

17 NI TECHNICAL REPORT DATE & SIGNATURE PAGE Project Name: Title of Report: Location: Natougou Gold Deposit Project NI Technical Report Burkina Faso Effective Date of Report: 23 [SIGNED] DATE Neil Lincoln, P.Eng. (Lycopodium Minerals Canada Ltd.) 23 Marius Phillips, MAusIMM (CP) (Lycopodium Minerals Canada Ltd.) 23 Glen Williamson, ( AMC Consultants (Canada) Ltd) 23 John Graindorge, MAusIMM (CP) (Snowden Mining Industry Consultants) 23 Jean-Sébastien Houle, ing, (WSP Canada) 23 Timothy Rowles, MAusIMM (CP), Knight Piésold Consulting 23 Lycopodium Minerals Canada Ltd

18 NI TECHNICAL REPORT Table of Contents 1.0 SUMMARY Overview Accessibility, Climate, Local Resources, Infrastructure and Physiography Geology and Mineral Resources Exploration Drilling, Sampling and Assaying Quality Assurance and Quality Control Mineral Processing and Metallurgical Testwork Mineral Resource Estimate Mineral Reserves Mining Recovery Methods Infrastructure Environmental, Permitting and Social or Community Impact Operating and Capital Costs Economic Analysis INTRODUCTION Terms of Reference and Purpose of this Report Site Visits RELIANCE ON OTHER EXPERTS PROPERTY DESCRIPTION AND LOCATION Type of Mineral Tenure Boungou Permit ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY Access and Physiography Climate Infrastructure HISTORY GEOLOGICAL SETTING AND MINERALIZATION Regional geology Eburnian metamorphics and intrusives Tarkwaian Sequence Property Geology Natougou Geology Oxidation Lithology Alteration Mineralization DEPOSIT TYPE 8.1 Lycopodium Minerals Canada Ltd

19 NI TECHNICAL REPORT 9.0 EXPLORATION DRILLING Sampling techniques RC drillhole sampling Diamond drillhole sampling Drill sample recovery Logging Location data and survey methods Data Spacing and Distribution Location of Data Points Downhole Surveys Orientation of data in relation to geological structure Topography SAMPLE PREPARATION, ANALYSES AND SECURITY Sample Preparation Analytical techniques Primary Analysis Handheld XRF Analysis Quality of Assay Data and Laboratory Tests Internal Laboratory Quality Assurance and Quality Control Laboratory Inspections Assay Reports Certified Reference Materials Blanks Field Duplicates Umpire Laboratory Checks Sample Security Bulk Density Qualified Person s Opinion on Adequacy of Sampling DATA VERIFICATION RC versus Diamond Comparisons Assay Data Validation Qualified Person s Opinion MINERAL PROCESSING AND METALLURGICAL TESTING Introduction Geological Background Sample Selection for Detailed Testwork Program Samples for Detailed Comminution Testwork Samples for Detailed Metallurgical Testwork Samples for Variability Metallurgical Testwork Detailed Testwork Program Outline Comminution Testwork Metallurgical Testwork Borehole Bench Composites Metallurgical Testwork Master Composites Ore Mineralogy Head Analysis Gravity Concentration Cyanide Leach Grind / Recovery Testwork Flotation Lycopodium Minerals Canada Ltd

20 NI TECHNICAL REPORT Grind Optimization Study Oxygen and Cyanide Investigations Lead Nitrate Additions Leach Pulp Density Direct Leach vs Gravity Concentration Oxygen Optimization Testwork Residence Time Selection Bulk Leach Carbon Loading Kinetic Tests Carbon Equilibrium Loading Intensive Cyanidation of Gravity Concentrate Ancillary Testing Rheology Oxygen Uptake Testing Pre-leach and Tailings Thickening Testwork Tailings Filtration Testwork Variability Metallurgical Testwork Head Analysis Gravity / Leach Testwork Gravity / Leach Testwork Diagnostic Leach Metallurgical Testwork Summary Selected Treatment Route Key Process Design Parameters Metallurgical Recoveries Reagent Consumption Testwork Conclusions MINERAL RESOURCE ESTIMATES Geological Interpretation Drill sample analysis Sample Compositing Statistical Analysis Top Cut Analysis Unfolding Introduction Principles of Unfolding Creation of Strings for Unfolding Variography Gold In situ bulk density Block modelling and grade estimation Kriging neighbourhood analysis Volume model construction Block model coding Search Neighbourhood Parameters Grade estimation methodology Model validation Mineral Resource classification Mineral Resource reporting Cut-off grade Moisture Natougou Mineral Resource statement Grade-tonnage curve Lycopodium Minerals Canada Ltd

21 NI TECHNICAL REPORT 15.0 MINERAL RESERVE ESTIMATES Introduction Method Mining Cut-off Grade Dilution and Mining Recovery Factors Mine Design Items Impacting on the Mineral Reserves MINING METHODS General description Pit dewatering Geotechnical parameters Rock strength Geological Input into Geotechnical Model Geotechnical Model Pit wall design Resource Model for Mining Mining Method Load and Haul Drill and blast Stockpile Rehandling Ancillary Equipment Mine Design Open Pit Optimization Pit design Layout of other mining related facilities Strategic mining schedule Detailed schedule Construction activity Schedule results Production Schedule Risks Equipment and Personnel Numbers Additional Information RECOVERY METHODS Process Design Selected Process Flowsheet Key Process Design Criteria Process and Plant Description Introduction Ore Receiving and Crushing Coarse Ore Storage Grinding and Classification Pebble Crushing Gravity Concentration Pre-Leach Thickening Leach Circuit Carbon Absorption Circuit Desorption and Carbon Regeneration Electrowinning and Gold Room Tailings Thickening Reagents Mixing and Storage Water Services Lycopodium Minerals Canada Ltd

22 NI TECHNICAL REPORT Air and Oxygen Services Plant Consumption Water Consumption Energy Consumption Reagent and Consumable Consumption Plant Control System General Overview Control System Configuration and Communications Drive Controls PROJECT INFRASTRUCTURE Overall Site Roads Access to Site Project Site Roads Access Tracks Power Power Supply Electrical Distribution Electrical Buildings Transformers and Compounds Fuel Supply Potable Water Sewage and Solid Waste Management Sewage Treatment Solid Wastes Accommodation Camps Mine/Plant Site Facilities General Outside Plant Area Inside Plant Area Mine Services Process Plant Geotechnical Water Supply Infrastructure Water demands Decant from Tailings Storage Facility Pit dewatering Groundwater sources Surface water sources and storages Tailings Storage Facility Design objectives and site selection Facility design Construction Staging and Summarized quantities Surface Water & Sediment Management Anticipated water quality Management system design MARKET STUDIES AND CONTRACTS Market Studies Contracts ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT 20.1 Lycopodium Minerals Canada Ltd

23 NI TECHNICAL REPORT 20.1 Legal Framework and Permits to Obtain Policies and Strategies for Environmental Protection Legal Framework Mining Code Institutional Framework Permits to Obtain Baseline Studies Baseline Studies Conducted Description of the Main Environmental and Social Components Community Information and Consultation Program Project Impacts, Risk Analysis, Environmental and Social Management Plan Project Impacts Risk Analysis Environmental and Social Management Plan Resettlement People and Activities Affected by the Project Scope of Resettlement Resettlement Action Plan Budget Acid Rock Drainage Waste Rock and Construction Materials Tailings Waste Disposal and Sanitary Wastewater Solid Waste Hazardous Waste Sanitary Wastewater Closure, Decommissioning and Reclamation CAPITAL AND OPERATING COSTS Operating Costs Mining Operating Costs Plant and Administration Costs Power Operating Consumables Labour (Processing / Maintenance and Administration) General and Administration Cost (excluding G&A labour) Maintenance Exclusions Capital Cost Estimate Summary Estimating Currency and Base Date Mining Capital Expenditure Estimating Methodology Quantity Development Pricing Basis Field In-directs Engineering Services Owner s Costs Spares Duties, taxes and Insurances Contingency Deferred Capital Escalation Lycopodium Minerals Canada Ltd

24 NI TECHNICAL REPORT Qualifications and Assumptions Exclusions Deferred Capital Sustaining Capital Costs Project Implementation Strategy ECONOMIC ANALYSIS Introduction Taxes Value Added Tax Royalties Other Royalties / Agreements Revenue Deductions Project Financing Financial Model Inputs Key Economic Assumptions Production summaries Capital Cost Summary Operation Cost Summary Economic Results Cash Flow Sensitivities ADJACENT PROPERTIES OTHER RELEVANT DATA AND INFORMATION INTERPRETATION AND CONCLUSIONS Geology, Exploration and Mineral Resource Mining Metallurgy Environmental and Permitting Risks Resettlement Delays RECOMMENDATIONS Geology and Mineral Resources Mining Environmental and Permitting Overall Recommendation REFERENCES 27.1 Lycopodium Minerals Canada Ltd

25 NI TECHNICAL REPORT TABLES Table 1.1 Natougou Mineral Resource as at January 2016, reported above a 0.83 g/t Au cutoff grade Table 1.2 Mineral Reserves estimate as of 31 December Table 1.3 Mining Schedule Summary Table 1.4 Average LOM Operating Costs Table 1.5 Initial Capital Costs Table 1.6 Economic Analysis Summary Table 2.1 Persons Who Prepared this Technical Report Table 2.2 Persons Who Contributed to this Technical Report Table 4.1 Tapoa Permit Group Exploration Permits Table 4.2 Boungou permit boundary (UTM Clarke 1880 Datum: Adindan, Zone 31N) Table 5.1 Average monthly rainfall for Fada n Gourma Table 10.1 Tapoa Permit Group drillhole details Table 11.1 Innov-X Omega series elements and detection limits Table 11.2 CRM Details for Natougou Table CRM results Assayed at SGS Table 11.4 CRM results assayed at ALS Table 13.1 Major Comminution Domains Table 13.2 Comminution Testwork Summary Table 13.3 Bottle Roll Cyanide Leach Results for Borehole Bench Composites Table 13.4 Detailed Head Analysis of Master Composites Table 13.5 Summary of Grind Recovery Leach Results Table 13.6 Summary of Rougher Flotation Tests With and Without Gravity Table 13.7 Flotation Vs WOO Leach Financial Analysis Table 13.8 Grind Optimization Financial Analysis Table 13.9 Oxygen and Cyanide Investigation Test Matrix Table Residence Time Selection Gold Recoveries at Various Head Grades Table Residence Time Selection Financial Analysis Table Summary of Bulk Leach Test Results and Optimized Test Results Table Detailed Solution Analysis from Bulk Leaches Table Cyanide Speciation from Bulk Leaches Table MC1 Results of Sequential Triple Contact CIP Testwork Table MC2 Results of Sequential Triple Contact CIP Testwork Table Loaded Carbon Assays from Triple Contact Tests Table Carbon Equilibrium Loading Table GRG and Intensive Leach Recoveries Table egrg Concentrate Grades Table Slurry Rheology for Master Composites Table Outotec Thickener Test Results Table FLS Thickener Test Results Table Outotec Tailings Filtration Test Results P 80 of 63 microns Table FLS Tailings Filtration Design Parameters Table Detailed Head Analysis of Variability Samples Section 1 & Section 2 (1, 3 to 11) Table Detailed Head Analysis of Variability Samples Section 2 (12 to 24, 27, 29) Table Detailed Head Analysis of Variability Samples Section 3 (1 to 15) Table Detailed Head Analysis of Variability Samples Section 4 (1 to 6, 8 to 13) Table Detailed Head Analysis of Variability Samples Section 5 (1 to 20) Table Detailed Head Analysis of Variability Samples Section 5 (21 to 29), Section Table Detailed Head Analysis of Variability Samples Section Table Bottle Roll Cyanide Leach Results for Section 1 Composites Lycopodium Minerals Canada Ltd

26 NI TECHNICAL REPORT Table Bottle Roll Cyanide Leach Results for Section 2 Composites Table Bottle Roll Cyanide Leach Results for Section 3 Composites Table Bottle Roll Cyanide Leach Results for Section 4 Composites Table Bottle Roll Cyanide Leach Results for Section 5 Composites (1 to 15) Table Bottle Roll Cyanide Leach Results for Section 5 Composites (16 to 29) Table Bottle Roll Cyanide Leach Results for Section 6 Composites Table Bottle Roll Cyanide Leach Results for Section 0 Composites (Oxide) Table Reagent Consumptions in Master Composite and Variability Tests Table Predicted Plant Reagent Consumptions Table 14.1 Summary Statistics for Gold Composites Table 14.2 Summary Statistics for bulk density samples Table 14.3 Relationship between the variogram modelling software coordinate fields and the unfold coordinate fields Table 14.4 Indicator variogram thresholds and variogram model mapping Table 14.5 Indicator variogram models for the main mineralized domain Table 14.6 Distribution tail modelling Table 14.7 Density back-transformed variogram model parameters Table 14.8 Block model prototype settings Table 14.9 MINZONE field coding Table OXIDE field coding Table HWZONE field coding Table ROCK field coding Table Search neighbourhood parameters Table Default gold grade values for un-estimated blocks Table Default density values for un-estimated blocks Table Statistical validation comparing input sample data against block model estimates Table Pit optimization parameters for Mineral Resource definition pit shell Table Resource classification model field values Table Table Natougou Mineral Resource as at January 2016, reported above a 0.83 g/t Au cutoff grade Grade-tonnage report at various cut-offs for Natougou resource estimate, as at January 2016 (exclusive of Mineral Reserves) Table 15.1 Mineral Reserves Estimate as of 31 December Table 15.2 Cut-off grade input parameters and calculations for Mineral Reserves Table 15.3 Mining Dilution and Recovery Parameters Table 16.1 Laboratory Rock Strength Tests Table 16.2 Typical strength estimates for rock mass zones Table 16.3 Slope design guidelines for the AMC proposed final pit Table 16.4 Drill and blast parameters Table 16.5 Open Pit High Grade Ore Cut-off Calculation Table 16.6 Open Pit Low Grade Ore Cut-off Calculation Table 16.7 Flitch Height Selection Schedule Results Table 16.8 Mining Dilution and Ore Loss Parameters Table 16.9 Pit Optimization Results Table Ore and waste tonnage within the three ultimate pit designs Table Out-of-pit waste dump storage capacity Table Waste dump rehabilitation parameters Table Strategic schedule - tonne and grade summary by year Table Grade Bin definition Table Low Grade and High Grade Cut-off Grades Table Schedule Milestones Table TSF Wall Construction Requirements Lycopodium Minerals Canada Ltd

27 NI TECHNICAL REPORT Table Schedule summary Ex-Pit movement by year Table Mill feed schedule by year Table Topsoil ex-pit and construction material by year Table Waste movement by destination by year Table Open pit equipment requirements at peak production Table Open pit workforce requirements Table 17.1 Summary of Key Process Design Criteria Table 17.2 Average Power Demand Summary Table 17.3 Annual Reagent and Major Consumable Consumption Table 18.1 Potable Water Usage Table 18.2 TSF Construction Staging and Major Quantities Table 20.1 List of Required Permits Table 20.2 List of environmental and social baseline studies conducted as part of the ESIA process Table 20.3 Resettlement Action Plan Budget (from SOCREGE, 2015) Table 20.4 Mine Closure Budget Table 21.1 Mining operating costs by year* Table 21.2 Mining operating cost by activity Table 21.3 Process Plant Operating Cost Summary Table 21.4 Process Plant Power Cost by Plant Area Table 21.5 Process Plant Consumables Cost by Plant Area Table 21.6 Plant General and Administration Summary Table 21.7 Plant Maintenance Materials Cost Table 21.8 Capital Estimate Summary by Area (4Q15, ±15%) Table 21.9 Capital Estimate Summary by Discipline (4Q15, ±15%) Table Foreign Currency Exposure Table Mining Capital Expenditures Table Derivation of Quantities Table Sources of Pricing Table Standard Direct Labour Gang Rates Table Deferred Capital Summary Table Sustaining Capital Schedule Table 22.1 Key Assumptions Used in the Financial Model Table 22.2 Economic Results Table 22.3 Cash Flow Table 22.4 Gold Price Sensitivity FIGURES Figure 4.1 Location map Figure 4.2 Tapoa Permit Group Figure 4.3 Boungou permit boundary Figure 5.1 Natougou access routes Figure 5.2 Climate information for Ouagadougou Figure 5.3 Photo overlooking the Natougou project area (photo taken March 2015, looking southeast) Figure 7.1 Overview of Regional Geology of Birimian Province Figure 7.2 Property Geological Map Tapoa Permits within Diapaga Belt Figure 7.3 Property Geological Map Boungou Permit Figure 7.4 Oxidation Types Figure 7.5 Photo of typical oxidation profile in core (BODD484) from Natougou Figure 7.6 Long section (looking northeast) showing oxidation impact on mineralization Figure 7.7 Lithology Types Lycopodium Minerals Canada Ltd

28 NI TECHNICAL REPORT Figure 7.8 Boundou Shear Zone at Contact between Hangingwall and Footwall Volcanic Units Figure 7.9 REE geochemical signatures for Natougou volcanic rocks Figure 7.10 Geochemical Signature (Y+Nb vs Rb) of Natougou Felsic Rocks Figure 7.11 REE geochemical signatures for Natougou Felsic Rocks Figure 7.12 Alteration types Figure 7.13 Core showing BPK and BBZ alteration Figure 7.14 Natougou deposit global geometry Figure 7.15 Cross-section (A-B; top, looking northwest) and long section (C-D; bottom, looking northeast) through Natougou deposit Figure 7.16 Cross-section Showing en Echelon Offset in Northwest Portion of Deposit Figure 7.17 Cross-section showing minor mineralization within the footwall below the main shear Figure 7.18 Photo of visible gold in core from Natougou Figure 9.1 Tapoa Permit Group exploration soil anomaly Figure 9.2 Tapoa Permit Group exploration rock chip samples >0.1 g/t Au Figure 10.1 Drillhole collar location plan for the Tapoa Permit Group Figure 10.2 Grade x thickness contour plot of Natougou deposit Figure 10.3 Histogram of core recoveries Figure 10.4 Histogram of RC drilling recoveries Figure 11.1 CRM Control Chart (z-score) Results Assayed at SGS Figure 11.2 Individual CRM Control Chart Results Assayed at SGS Figure 11.3 CRM Control Chart (z-score) results assayed at ALS Figure 11.4 Individual CRM Control Chart results assayed at ALS Figure 11.5 Coarse Blank Control Chart results assayed at SGS Figure 11.6 Coarse Blank Control Chart results assayed at ALS Figure 11.7 Original vs. Field Duplicate log QQ plot (SGS) Figure 11.8 Original vs. Field Duplicate Relative diff vs mean plot (SGS) Figure 11.9 Original vs. Field Duplicate Precision Plot (SGS) Figure Umpire (SMF BF) vs. Original Pulp (SGS) duplicate log QQ plot Figure Umpire (SMF BF) vs. Original Pulp (ALS) Duplicate log QQ plot Figure Umpire (SMF BF) vs. Original Pulp (SGS) Duplicate Precision Plot Figure Umpire (SMF BF) vs. Original Pulp (ALS) Duplicate precision plot Figure Bulk density measurement equipment (photo taken March 2015) Figure Bulk density vs depth below surface Figure 12.1 Drill section over the 20 m by 20 m grid showing RC holes, multi-purpose holes with core portion highlighted in grey and diamond holes Figure 12.2 Plan showing comparison area with equal RC (red) and diamond (blue) coverage Figure 12.3 QQ plot comparing diamond (X-axis) against RC (Y-axis) samples within the Comparison area (red line = mean; dashed blue lines are the 25 th, 50 th and 75 th percentiles) Figure 13.1 Cross Section of Natougou Deposit Figure 13.2 Location of Comminution Samples Figure 13.3 Spatial Location of Master Composite Samples Figure 13.4 Location of Variability Samples Figure 13.5 Mineral Abundance for Natougou Master Composites Figure 13.6 MC1 Grind Sensitivity Cyanide Leach Curves Figure 13.7 MC2 Grind Sensitivity Cyanide Leach Curves Figure 13.8 MC1 Grind Sensitivity Grade Recovery Curves Figure 13.9 MC2 Grind Sensitivity Grade Recovery Curves Figure MC1 Oxygen Investigation Tests with Cyanide Concentration 0.035% w/v Figure MC1 Oxygen Investigation Tests with Cyanide Concentration 0.05% w/v Lycopodium Minerals Canada Ltd

29 NI TECHNICAL REPORT Figure MC2 Oxygen Investigation Tests with Cyanide Concentration 0.035% w/v Figure MC2 Oxygen Investigation Tests with Cyanide Concentration 0.05% w/v Figure MC1 Lead Nitrate Tests Figure MC2 Lead Nitrate Tests Figure MC1 Oxygen Vs Air with Lead Nitrate Figure MC2 Oxygen Vs Air with Lead Nitrate Figure MC1 Leach Pulp Density Series Figure MC2 Leach Pulp Density Series Figure MC1 Tests with and without Gravity at Optimum Conditions Figure MC2 Tests with and without Gravity at Optimum Conditions Figure MC1 Tests with Varying DO Levels Figure MC2 Tests with Varying DO Levels Figure MC1 Bulk Leach and Optimized Leach Test Results Figure MC2 Bulk Leach and Optimized Leach Test Results Figure Rate Curves for Sequential Triple Contact CIP Testwork Figure Oxygen Uptake Rate Figure Gravity Recovery vs. Overall Gold Recovery Figure Selected Variability Leach Curves Arsenic, Sections 1 and Figure Selected Variability Leach Curves Arsenic, Sections 5 and Figure Selected Variability Leach Curves Copper, Sections 1 and Figure Selected Variability Leach Curves Copper, Sections 5 and Figure Selected Variability Leach Curves Mercury Figure Residue Grade Versus Head Grade Figure 14-1 Isometric view showing geological interpretation for the main mineralized structure Figure 14-2 Collar location plan for drilling at Natougou up to October Figure 14-3 Log histogram and log probability plot for gold for main mineralized domain (top) and minor footwall mineralized domain (bottom) Figure 14-4 Box and Whisker Plot of Density Measurements Figure 14-5 Schematic Illustration of the unfolding process in Datamine Figure 14-6 Plan View of the Unfold Control Strings Figure 14-7 Isometric view of the unfold strings looking from the south-southeast Figure 14-8 Comparison between the original data (left) and unfolded data (right) Figure 14-9 Median (50%) indicator variogram model Figure Example oblique section (looking northwest) showing MINZONE field coding Figure Example oblique section (looking northwest) showing OXIDE field coding Figure Example oblique section (looking northwest) showing HWZONE field coding Figure Example oblique section (looking northwest) showing ROCK field coding Figure Grade trend plots Au, main mineralized domain (source: Snowden) Figure Grade trend plots density, main mineralized domain (source: Snowden) Figure Example oblique section (looking northwest) showing block MIK gold estimates Figure Plan view showing Natougou Mineral Resource classification scheme Figure Grade-Tonnage Curve Figure 15.1 Final Pits Geometry Figure 16.1 Recommended pit slope design guidelines Figure 16.2 Ore vertical thickness histogram Figure 16.3 Pit Optimization Results Figure 16.4 Pit optimization sensitivity analysis Figure 16.5 Pit designs overview Figure 16.6 West pit design Stage Figure 16.7 West Pit Design Stage Figure 16.8 West Pit Design Stage Lycopodium Minerals Canada Ltd

30 NI TECHNICAL REPORT Figure 16.9 West Pit Design Stage Figure West Pit Design Stage Figure West Pit Section A1 A Figure West Pit Section B1 B Figure North East Pit - Stage Figure North East Pit - Stage Figure North East Section A1 A Figure North East Section B1 B Figure South East Pit Stage Figure South East Pit Stage Figure South East Pit Section A1 A Figure South East Pit Section B1 B Figure Pit contribution by year (tonnes mined) Figure Strategic mine plan mill feed Figure Strategic mine plan stockpile balance Figure Strategic mine plan recovered gold ounces Figure Site layout Figure Closing stockpile balance by period Figure Annual recovered gold ounces Figure Annual total ex-pit movement Figure Annual plant feed Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure End of year face position Year Figure 17.1 Overall Process Flow Diagram Figure 18.1 Overall Site Plan Figure 18.2 Water Balance Schematic Flow Diagram Figure 18.3 Location of Surface Water Sources and Storages Figure 18.4 Location of Tailings Storage Facility Figure 18.5 Layout of Underdrainage and Sub-liner Drainage Network Figure 18.6 Tailings Storage Facility Embankment Figure 18.7 Surface Water Management System Figure 20-1 Classified and protected areas near the project area Figure 22.1 NPV Sensitivity Analysis (Pre-tax) Figure 22.2 IRR Sensitivity Analysis (Pre-tax) Figure 23.1 Adjacent permit boundaries Lycopodium Minerals Canada Ltd

31 NI TECHNICAL REPORT - SUMMARY Page SUMMARY 1.1 Overview The Tapoa Permit Group 1, which includes the Natougou gold deposit is located in Burkina Faso, West Africa. The Natougou project lies approximately 320 km east of Ouagadougou, the capital of Burkina Faso. SEMAFO Inc. (SEMAFO) acquired the Tapoa Permit Group from Orbis Gold Limited ( Orbis ). SEMAFO initiated a takeover of Orbis in December 2014 and Orbis was delisted from the Australian Securities Exchange ( ASX ) on 16 March Accessibility, Climate, Local Resources, Infrastructure and Physiography Access to the Natougou site is by means of Route Nationale RN04, an all-weather bitumen road from Ouagadougou, the capital of Burkina Faso, through Fada n Gourma to the Ougarou junction. From there, travel is via a laterite road to the property 60 km to the southeast. Fada n Gourma is the nearest town with basic hospital, hotel and limited resupply facilities. Any significant supplies will be sourced from Ouagadougou. The Natougou area is relatively flat and sits at an elevation of approximately 260 m above sea level. To the east and north of Natougou are mesas which rise approximately 10 m above the surrounding topography. A small hill is located in the very southern corner of the deposit. The land rises gently to the north, culminating in the height of land separating two watersheds. The main laterite access road into site is located along this ridge top. The road is reasonably well-drained and is accessible year-round to four-wheel drive vehicles. The climate of Burkina Faso is semi-arid, with a rainy season from May to September, and a hot dry season from February to April. Average temperatures range between 16 centigrade ( C) overnight in the cool season to over 40 C during the day in the hot season. Average annual rainfall is approximately 900 mm, although large inter-year variability is common. During October to April, the climate is heavily influenced by the dry, dust-laden northwest trade wind known as Harmattan, which blows from the Sahara Desert. 1 All the permits referred in this document are held by indirectly owned subsidiaries of SEMAFO Inc. The operating entity to be created for the Natougou Mine will be 90% indirectly owned by SEMAFO Inc. and 10% by the Government of Burkina Faso. Lycopodium Minerals Canada

32 NI TECHNICAL REPORT - SUMMARY Page Geology and Mineral Resources The Tapoa Permit Group is located within the Birimian Gold Province in West Africa. The Birimian Gold Province is a considered a world-class gold province and hosts most of the major gold deposits in West Africa, notably in Ghana, Ivory Coast, Mali, Senegal and Burkina Faso. Gold mineralization at the Natougou deposit is hosted within a flat lying shear zone which has a subtle anticlinal geometry. The apparent axis of the anticline strikes approximately 315, plunging gently to the northwest and each limb dips approximately 15. The mineralization is characterized by sheared amphibolite, quartz boudinage veining, biotiteplagioclase alteration and an increase in sulphide content. Sulphide assemblage comprises pyrrhotite, pyrite and minor arsenopyrite and chalcopyrite. Visible gold has been observed in core samples. The host lithology consists of mafic to intermediate volcanic/intrusive stratigraphy. 1.4 Exploration Exploration activities on the property have been ongoing since Regional soil sampling and rock chip sampling programs were commenced by Orbis in 2010 and permit scale mapping was conducted by Orbis geologists during the course of the 2014 field season and updated by SEMAFO in Shallow artisanal mining activity occurs across the Tapoa Permit Group and is used to identify targets for further exploration. Orbis defined a large-scale high-order (+50 ppb Au) gold-in-soil anomaly in the area surrounding the Natougou discovery. The soil anomaly, defined within a 6 km by 4 km survey area, includes multiple zones of higher-order anomalism that have received minimal exploration drilling to date. The higher order soil anomalies present as priority areas for follow-up exploration. 1.5 Drilling, Sampling and Assaying A total of 1,349 holes have been drilled on the property as of November 2015, using reverse circulation ( RC ) and diamond drilling techniques. SEMAFO drilled 624 holes in 2015; the remaining holes were drilled by Orbis between 2012 and 2014, with the bulk of the drillholes targeting the Natougou deposit. The dominant drillhole spacings at Natougou are 40 m along strike (315 ) by 40 m across strike (045 ) and 20 m along strike by 20 m across strike. The global recovery for drill core and RC chips at Natougou is acceptable, with an average recovery of 98% for drill core and 84% for the RC drilling. The majority of the mineralised samples are from diamond core drilling and as such, it is Snowden s opinion that the recoveries are acceptable for representative sampling and subsequent Mineral Resource estimation. Lycopodium Minerals Canada

33 NI TECHNICAL REPORT - SUMMARY Page 1.3 RC cuttings are collected from the cyclone at 1 m intervals. Each 1 m RC sample was reduced using a tiered riffle splitter to obtain a split of about 2 kg which was packed in a poly bag with a sample ticket included in each bag. Diamond core is half core sampled, with the core cut using a diamond saw. The majority of the diamond core is sampled based on the geological logging using a minimum sample length of 0.2 m and a maximum of 1.2 m. Samples were prepared either at the SGS laboratory in Ouagadougou or, since 2015, at the on-site sample preparation facility, which is managed and run by SGS. The sample preparation involves oven drying and crushing to 2 mm, followed by pulverising to 85% passing 75 µm. Gold content is determined by fire assaying, which has primarily been completed at the SGS laboratory in Ouagadougou. Based on the results of the quality control samples and multiple inspections, SGS has achieved reasonable precision and analytical accuracy. In the author s opinion, the drillhole data for the Natougou deposit is acceptable for use in resource estimation. 1.6 Quality Assurance and Quality Control Orbis implemented a quality assurance and quality control ( QAQC ) protocol throughout the majority of the initial drilling programs at Natougou, which included the addition of reference materials, field duplicates and blank samples to the sample batches. This practice was largely continued by SEMAFO in For each batch of samples, which typically comprised 74 samples, for the 2015 drilling, two reference material samples, two blanks and two pulp duplicates are inserted. However, Snowden Mining Industry Consultants ( Snowden ) notes that SEMAFO do not include field duplicate samples for RC drilling. Snowden recommends that SEMAFO reinstitute the use of field duplicates for RC drilling to monitor and assess the precision of the samples. Analysis of the reference materials shows the majority of results fall within the accepted control limits, suggesting that reasonable analytical accuracy has been achieved. Moreover, assays of the blank samples show no evidence of contamination during the laboratory sample preparation or assaying. For the field duplicates, overall the populations compare reasonably well, with some outliers typical of gold deposits. Snowden considers there is no evidence to suggest that the primary sample varies significantly from the duplicate sample and that SEMAFO have achieved reasonable precision during the sampling and assaying process. Snowden checked a random selection of assay certificates against the data within the database in 2013 and again in Zero discrepancies were identified from this validation. The assay certificates were sourced by Snowden directly from the SGS Ouagadougou laboratory. Lycopodium Minerals Canada

34 NI TECHNICAL REPORT - SUMMARY Page Mineral Processing and Metallurgical Testwork A detailed metallurgical testwork program was undertaken and was focussed on primary ore from the Natougou deposit. Quantities of oxide ore presented to the process plant are expected to be around 1% of reserves and as such, this ore type was not included in the master composite work. However, it was tested in the variability work. The detailed testwork was carried out from March 2013 to August 2015 under the direction of Lycopodium Minerals Canada Ltd (Lycopodium), with input from former property owner, Orbis and later SEMAFO, using HQ and PQ drill core recovered from both resource and metallurgical drilling campaigns. In general the Natougou primary ore is an abrasive, competent ore with above average comminution energy requirements. The ore has a high gravity recoverable gold content; leach kinetics are very slow when gravity is not included in the flowsheet. High dissolved oxygen levels and lead nitrate are required to achieve fast leach kinetics and adequate gold recovery. Anticipated lime consumption for primary ore is low to moderate, provided good quality water can be provided on site. Cyanide consumption is likely to be moderate. High lime consumption will be experienced if oxide ore forms part of the feed blend. The variability testwork showed that overall gold recoveries for the Natougou primary ore ranged from 84% to 99%. There was a distinct relationship between recovery in the gravity stage and overall recovery. LOM head grades for the process plant are expected to average 4.15 g/t with a gold recovery of 92.9% The results suggest that the residue grade is moderately correlated with the amount of coarse gold in the sample (measured by % gold in +75 micron fraction of the screen fire assay), arsenic head assay, and gold head assay. A constant tail relationship is not appropriate. With consideration of the parameters currently in the geological model, a relationship between the residue grade and the gold head assay was developed to produce the following predictive equation: Gold Residue (g/t Au) = *Gold Head Assay (g/t Au) For example for a gold head assay of 4.36 g/t Au, the gold residue grade would be 0.31 g/t Au. As silver residue grades are frequently at the assay detection limit, and no trend with head grade is apparent, it is recommended that a simple arithmetic average of all the silver recovery figures be used i.e. 67%. Lycopodium Minerals Canada

35 NI TECHNICAL REPORT - SUMMARY Page Mineral Resource Estimate Snowden has generated a resource block model for the Natougou deposit using geological and gold mineralization interpretations provided by SEMAFO. Multiple indicator kriging ( MIK ) with unfolding was used to estimate gold grades (Au) into a constrained block model reflecting the interpreted shear zone and surrounding host rocks. Bulk density measurements were collected from core using the Archimedes method. Whilst the samples are limited in terms of sample size (typically only 10 cm to 15 cm lengths are measured), a total of 23,843 measurements have been collected. A visual assessment of the density data shows areas of lower and higher density, with continuity between adjacent drillholes. As such, Snowden elected to estimate the bulk density using ordinary kriging (with unfolding) to ensure that the local variations in the density measurements are reflected in the resource model. The average bulk density for the fresh (i.e. unweathered) Natougou mineralization is 2.93 t/m 3. The Natougou Mineral Resource estimate has been classified as a combination of Measured, Indicated and Inferred Resources in accordance with CIM guidelines. The resource classification was based on a review of the underlying data quality along with an assessment of the level of confidence in the understanding of both the geological and grade continuity. The Mineral Resource has been limited to within a pit shell provided by SEMAFO, based on a gold price of USD 1,400/oz and pit optimization parameters developed for the Mineral Reserve estimate. The Natougou Mineral Resource comprises 2.64 Mt at 2.42 g/t Au of Measured and Indicated Resources and 2.68 Mt at 3.99 g/t Au of Inferred Resources, reported above a 0.83 g/t Au cut-off grade. The Mineral Resource statement is detailed in Table 1.1. Mineral Resources have been reported exclusive of Mineral Reserves. Lycopodium Minerals Canada

36 NI TECHNICAL REPORT - SUMMARY Page 1.6 Table 1.1 Natougou Mineral Resource as at January 2016, reported above a 0.83 g/t Au cut-off grade Classification Oxidation state Tonnes (Mt) Au g/t Ounces (koz) Measured Fresh Measured total Indicated Strongly oxidized Moderately oxidized Weakly oxidized Fresh Indicated total Measured + Indicated total Inferred Strongly oxidized Moderately oxidized Weakly oxidized Fresh Inferred total * Small discrepancies may occur due to rounding; Mineral Resources are reported exclusive of Mineral Reserves Snowden notes that the vast majority of the resources within the designed open pit have been converted into Mineral Reserves, as detailed in Section 15. Mineral Resources remaining within the USD 1,100/oz base case designed open pit comprise 653 kt at 1.75 g/t Au (37,000 oz of contained gold) of Measured and Indicated Resources (predominately Indicated) along with 3 kt at 1.76 g/t Au (150 oz) of Inferred Resources. The majority of the Mineral Resources within the designed pit, which have not been converted to Mineral Reserves, occur at the upper and lower edges of the mineralization. 1.9 Mineral Reserves To convert Mineral Resources to Mineral Reserves, modifying factors were applied including mining cut-off grades, mining dilution and mining recovery factors. Only Measured and Indicated Mineral Resources have been used to state Mineral Reserves. The Mineral Reserve estimate is based on a gold price of USD 1,100/oz. The Mineral Reserves are generated by reporting out of the feasibility study open pit designs above the cut-off grade. Mining recovery and dilution factors are applied to this quantity to estimate the Mineral Reserves. The Mineral Reserves use a cut-off grade of 0.92 g/t Au for weathered material and 1.07 g/t Au for fresh material. Lycopodium Minerals Canada

37 NI TECHNICAL REPORT - SUMMARY Page 1.7 Table 1.2 Mineral Reserves estimate as of 31 December 2015 tonnes Total Weathered Fresh Au grade Au metal tonnes Au grade Au metal tonnes Au grade Au metal Mt g/t Moz Mt g/t Moz Mt g/t Moz West Pit Proven Probable Total South Proven East Pit Probable Total North East Proven Pit Probable Total Total Proven Probable Total Notes: Cut-off grade of 0.92 g/t Au for weathered material and 1.07 g/t Au for fresh material based on total operating costs of 30 USD /t of ore, 96% Au metallurgical recovery for weathered ore and a variable recovery equation for fresh ore: recovery (%) = (1- ((Au* )/ Au )) *100, 1.49 USD/g Au treatment costs, royalty and penalties, and 99.93% payable gold Au price assumed is USD 1,100 per troy ounce Probable Mineral Reserves are based on Indicated Mineral Resources only Proven Mineral Reserves are based on Measured Mineral Resources only Tonnes and grades shown account for mining dilution at 18% and mining recovery at 95%. Dilution material averages 0.2 g/t Au The numbers may not add up due to rounding 1.10 Mining Mining uses a conventional open-pit mining method, with hydraulic excavators in backhoe configuration to mine the mineralized zone, and in face-shovel configuration to mine the majority of the waste. The majority of the rock requires blasting and only the softer material located within the top 5 m to 10 m of the deposit will be free digging and loaded directly by hydraulic excavators. The mining operations are anticipated to be contracted out with Birimian overseeing management and the provision of the mining technical services. The mine life is expected to last approximately 8 years excluding pre-stripping activities in 2017 and 2018; in 2026 the mill is fed solely from remaining stockpile. Over the life of mine, approximately 139 Mt of rock and topsoil are mined comprising of 9.6 Mt of ore and 130 Mt of waste at an average strip ratio of 13.6:1. A summary of the mining schedule is presented in Table 1.3. Lycopodium Minerals Canada

38 NI TECHNICAL REPORT - SUMMARY Page 1.8 Table 1.3 Mining Schedule Summary Mined Tonnes and Grades Total Total ore mined Mt Total Au grade mined g/t Total tonnes mined Mt Ex-pit strip ratio t:t Mill feed Total tonnes to mill Mt Total Au grade to mill g/t Value includes additional topsoil, without topsoil included the ex-pit strip ratio is Recovery Methods The metallurgical treatment route selected has been based on the results of the current testwork program and includes processing ore at 4,000 tonnes per day (tpd) via the following unit process operations: Single stage primary crushing with a jaw crusher to produce a crushed product size of 80% passing (P 80 ) of 133 mm. Mill feed surge/overflow bin that overflows to a 8,000 tonne stockpile to provide 48 hours of capacity. During extended periods of up to two days for primary crusher equipment maintenance, ore from the stockpile will be reclaimed by an excavator or dozer to feed the grinding circuit. The grinding circuit is a SATMC type, which consists of a closed circuit semi-autogenous (SAG) mill, pebble crusher for SAG mill discharge oversize and a closed circuit tower mill to produce a P 80 grind size of 63 µm. A gravity gold recovery circuit. Hydrocyclones are operated to achieve a cyclone overflow slurry density of 27% solids to promote better particle size separation efficiency. Subsequently, a pre-leach thickener is included to increase slurry density to the leach circuit, minimise leach tank volume requirements and reduce overall reagent consumption. Leach circuit with five tanks to achieve the required 36 hours of residence time at nominal plant throughput. Carbon-in-pulp (CIP) carousel circuit consisting of seven stages is a carbon adsorption circuit for recovery of gold dissolved in the leaching circuit. Lycopodium Minerals Canada

39 NI TECHNICAL REPORT - SUMMARY Page 1.9 AARL elution circuit with gold recovery to doré. The circuit includes an acid wash column to remove inorganic foulants from the carbon with hydrochloric acid. Carbon regeneration kiln to remove organic foulants from the carbon with heat. Tailings thickener to increase slurry density for water recovery prior to tailings discharge to the tailings storage facility. The processing facility also includes water, air and oxygen services (storage and distribution), and reagent and grinding media storage and usage Infrastructure The major infrastructure at the site to support a 4,000 tpd (1.34 Mtpa) mining and processing facility includes a 15.4MW onsite power generation via a hybrid heavy fuel oil and light fuel oil generators, electrical distribution, bulk fuel storage, tailings storage facility (TSF), water storage dams for water harvesting, sediment ponds, raw water storage facility, main access road, reagents and consumables storage, plant operations and maintenance buildings, administration building, medical facilities, warehousing main kitchen and dining room, engineering and exploration offices, accommodation camps with services for operations and maintenance personnel, and security. The TSF will have a capacity to store 10 Mt of tailings generated by the process plant is required for the life of the mine with tailings being produced at a rate of 1.34 Mtpa. The preferred site selected for the project is located 800 m to the north east of the process plant. The tailings storage facility will require a single embankment along the south and western extents of the facility with a total embankment length of 1665 m and with a maximum embankment height of 23.6 m at the south west corner. The eastern and northern margins of the facility are confined by a natural laterite ridge line and therefore no supporting embankment is required along these margins. The tailings beach surface at full capacity will cover an area of approximately 76.5 hectares. Tailings will be pumped to the TSF as a slurry at 62% to 65% solids and will be deposited sub-aerially to facilitate drying and consolidation of the tailings mass. Geochemical testing of the two composite tailings samples were conducted and found to be non-acid forming but were highly enriched in arsenic which was soluble under the ph conditions anticipated in the TSF. As a result of the high arsenic in the tailings solids and supernatant a robust seepage control system comprising an above liner underdrainage system, a geomembrane liner overlying a compacted in-situ low permeability sub-base and a sub liner seepage recovery drains have been included in the design. Lycopodium Minerals Canada

40 NI TECHNICAL REPORT - SUMMARY Page 1.10 The total water demand for the site was estimated at between 1.1 and 1.4 Mm 3 per year. The water demand for the process plant amounts to 0.75 Mm 3, which includes the process raw water requirement of 0.08 Mm 3 but excludes water in ore. Other water demands include a provision of between 0.2 Mm 3 and 0.6 Mm 3 for dust suppression and wash down water and 0.04 Mm 3 for potable water requirements. The demand will be met from TSF decant, pit dewatering (including precipitation on the pit area), runoff from the ROM pad and plant site and sediment impacted runoff collected in the sediment control ponds. The balance of the water demands will be made up of raw water harvested from the groundwater and the surface water sources. Raw water demands at the site will be met from two creeks which are located to the east and west of the process plant and water which will be harvested from the sediment ponds located around the site. An east water supply dam will be constructed approximately 1.5 km to the north east of the processing plant. The mean annual runoff at the dam site is estimated to be 0.95 Mm 3 from a catchment area of 1902 Ha. A west water supply sump will be constructed approximately 2.0 km to the west south west of the processing plant. The mean annual runoff at the dam site is estimated to be 1.15 Mm 3 from a catchment area of 2262 Ha. The flat topography at the dam sites would result in significant evaporation and seepage losses should this location be used to store water for extended periods of time and therefore a supplementary water storage facility, the raw water pond has been provided to store water more efficiently. Three major sediment ponds have been designed to capture runoff from the waste dumps with an upstream clean water diversion designed to carry non impacted water around the site Environmental, Permitting and Social or Community Impact Burkina Faso has a regulatory framework for environmental and social management. The relevant policies, laws and regulations of Burkina Faso were taken into account during the implementation of the Environmental and Social Impact Assessment (ESIA). The application for an Operating permit requires a Feasibility Study (FS) that must first be accepted by Ministère de l environnement et du développement durable 1 (MEDD). The FS must include an ESIA which in turn must include a Resettlement Action Plan ( RAP ) that has been accepted by all stakeholders. Once in production, a mining permit holder is required 2 to open under his name a fiduciary account named Fonds de préservation et de réhabilitation de l environnement minier 3 at the Banque Centrale des États de l Afrique de l ouest 4 ( BCEAO ). This account must be funded annually on January 1st by an amount equal to the total rehabilitation budget presented in the ESIA, divided by the number of years of production to cover the costs of mine reclamation, closure and rehabilitation. Both the ESIA and the PAR were near completion at the time of this Report and will be filed to the government of Burkina Faso in the second quarter of Ministry of Environment and Sustainable Development 2 Decree No /PRES/PM/MCE/MEF 3 Fund for the Preservation and the Rehabilitation of the Mining Environment 4 Central Bank of West African States Lycopodium Minerals Canada

41 NI TECHNICAL REPORT - SUMMARY Page 1.11 Many baseline studies have been conducted from 2013 and 2015 in order to fully document the sensitive environmental and social components of the project area. The stakeholder information and consultation process is an integral part of the ESIA. To date, SEMAFO has put in place mechanisms and communication tools so that all those involved in, or affected by, the project can freely express themselves. The information collected during these consultations has helped identify issues, risks, benefits, and opportunities in order for the project to avoid, minimize, or offset negative impacts and enhance the positive ones. The Project will have many impacts on the physical, biological, social and economic components of the Project s area. The impacts on physical environment are of medium importance, or less, given the mining operations have a zero water discharge and that the project is engineer to protect ground water from potential cyanide contamination. The most significant impact on social component is the resettlement of the population currently living on the project site. The economic impact of the project at local, regional, and national levels is positive. The project will provide jobs during construction, operation and closure phases of the project which will increase household incomes and improve living conditions. The revenues generated by the mining operation will also increase Burkina Faso s internal revenue through taxes and royalties charged by the local authorities. The Project will require the relocation of 165 concessions involving 900 inhabitants and compensation will be paid for 813 Ha of farmland. The current PAR budget is estimated at USD 7.4M, or approximately USD 8 M when including contingencies. Geochemical studies have been conducted to assess the potential for acid drainage (Acid Rock Drainage: ARD) and metal leaching (ML) of the waste rock and construction materials as well as CIP tailings and heap leach solids. It was found that the direct seepage form the waste dump should meet Burkina Faso Effluent Discharge Criteria for release to surface water. A Conceptual Closure and Rehabilitation Plan was developed including work to be conducted from the closure of the mine, at the end of operational activities, as well as progressive rehabilitation work. The estimated cost for the direct closing, decommissioning and restoration cost, engineering and post closure monitoring is estimated at USD 17.6 M. When including contingencies and salvage value recovery, the cost for mine closure and remediation is USD 17.2 M Operating and Capital Costs The average life of mine (LOM) operating costs are summarized in Table 1.4 and is based on the LOM operating strip ratio of 7.1:1 which excluded capitalized stripping costs. The operating costs do not include financing and leasing costs associated with the onsite power plant and bulk fuel storage. The operating costs are based on contract mining and include personnel costs, power costs based on fuel usage, reagent and consumable usage, maintenance costs, fuel usage, laboratory costs, and various administration costs covered in general and administration (G&A). Lycopodium Minerals Canada

42 NI TECHNICAL REPORT - SUMMARY Page 1.12 Table 1.4 Average LOM Operating Costs Area USD /t ore processed Mining $20.28 Processing $19.51 G&A $4.94 Total $44.73 The initial estimated capital cost to develop the Natougou project is USD 219 M and includes prestripping costs, contingency, taxes and duties and is summarized in Table 1.5. Table 1.5 Initial Capital Costs USD ( 000,000) Indirect Construction $13.6 Processing Plant $42.3 Reagents and Plant Services $13.7 Infrastructure $41.8 Owners costs $15.8 EPCM costs $15.9 Resettlement Action Plan $8.0 Initial Supplies Inventory $7.2 Plant & Infrastructure Subtotal $158.3 Pre-stripping $42.4 Contingency $18.7 Total $219.4 Deferred capital is estimated at USD 15.2M and sustaining and closure costs are estimated at USD 42.3M which include the estimated salvage value of the site in 2026 of USD 3.8M Economic Analysis The project financial model was compiled with inputs from various parties. The key inputs included mining costs, processing costs, G&A costs, project capital costs, sustaining capital costs and deferred capital costs. A pre and post tax model was developed and assumes contract mining, 100% ownership of the process plant and associated infrastructure, leasing to own power plant and bulk fuel storage, USD 1,100 /oz gold price, 5% discount rate for NPV calculations, no inflation or escalation is applied, and the latest Burkina Faso tax regulations are applied. Lycopodium Minerals Canada

43 NI TECHNICAL REPORT - SUMMARY Page 1.13 Project gold production averages 148,000 oz per year over the life of mine. During the first three years of production, average annual gold production is expected to be 226,000 oz with an average total cash cost 1 of USD 283/ oz and all-in sustaining cost 2 of USD 374/oz. Over the first three years, average head grade will be 5.72 g/t at a gold recovery of 94%. The results of the financial model are summarized in Table 1.6. Table 1.6 Economic Analysis Summary USD 1,100/oz After-tax operating cash flow ($M) $330 After-tax 5% NPV ($M) $262 After-tax IRR (%) 48 Payback period (years) Total cash cost represents the mining operation expenses and government royalties per once sold. 2 All-in sustaining cost represents the total cash cost plus sustainable capital expenditures and stripping costs per ounce. Lycopodium Minerals Canada

44 NI TECHNICAL REPORT - INTRODUCTION Page INTRODUCTION 2.1 Terms of Reference and Purpose of this Report This Technical Report has been prepared by Lycopodium Minerals Canada Ltd (Lycopodium) for Birimian Resources SARL (Birimian), in compliance with the disclosure requirements of the Canadian National Instrument (NI ) and in accordance with the requirements of Form F1. The Tapoa Permit Group, which includes the Natougou gold deposit is located in Burkina Faso, West Africa. The Natougou project lies approximately 320 km east of Ouagadougou, the capital of Burkina Faso. SEMAFO Inc. (SEMAFO) acquired the Tapoa Permit Group from Orbis Gold Limited ( Orbis ). SEMAFO initiated a takeover of Orbis in December 2014 and Orbis was delisted from the Australian Securities Exchange ( ASX ) on 16 March Subsequent to SEMAFO acquiring the Tapoa Permit Group, a definitive feasibility study (DFS) was undertaken and the results of the DFS are provided in this technical report. The individuals presented in Table 2.1, by virtue of their education, experience and professional association are considered Qualified Persons (QPs) as defined in NI for this report. The QPs meet the requirement of independence as defined in NI Section responsibilities are also listed in the table below. Lycopodium Minerals Canada

45 NI TECHNICAL REPORT - INTRODUCTION Page 2.2 Table 2.1 Persons Who Prepared this Technical Report Qualified Persons responsible for the preparation of this Technical Report Qualified person Neil Lincoln Position VP Business Development & Studies Employer Lycopodium Minerals Canada Ltd Independent of Birimian Yes Date of last site visit May Professional designation Report sections P.Eng 1,2,3, ,19, 21 (except for , ), 22, 24, 26.4, 27 John Graindorge Principal Geologist Snowden Mining Industry Consultants Yes March 2015 MAuslMM (CP) 4,5,6,7,8,9,10,11 12,14, 23, 25.1, 26.1 Marius Phillips Glen Williamson Jean- Sébastien Houle Timothy Rowles Principal Process Engineer Principal Mining Engineer Project Director, Mining Environment Regional Manager Lycopodium Minerals Canada Ltd AMC Consultants Pty Ltd WSP Canada Knight Piésold Consulting Yes No Visit MAusIMM (CP) Yes March 2014 MAusIMM (CP) 13, 17, , 16, , , 25.2, 26.2 Yes No Visit Eng 1.13, 20, (except for 20.5) 25.4, 25.5, 26.3 Yes 4-7 February 2014 MAusIMM (CP) ,20.5 Other experts upon whose contributions the Qualified Persons have relied are presented in Table 2.2. Unless otherwise stated, information and data contained in this report or used in its preparation have been provided by Birimian. Table 2.2 Persons Who Contributed to this Technical Report Expert Position Employer Independent of Birimian Date of last site visit Professional designation Report sections Patrick Moryoussef Vice-President, Mining Operations SEMAFO No May 2015 P. Eng 16, 21, 22 Sylvain Duchesne Vice-President, Construction and Engineering SEMAFO No May 2015 P. Eng 13, 17, 18, 21, 22 Michel Crevier Vice-President, Exploration and Mine Geology SEMAFO No March May 2015 P. Geo 4-12, Lycopodium Minerals Canada

46 NI TECHNICAL REPORT - INTRODUCTION Page 2.3 All measurement units used in this document are metric and all currencies are expressed in US dollars ( USD ), unless otherwise stated. Contained gold metal is expressed as Troy ounces ( oz ), where 1 oz = g. 2.2 Site Visits John Graindorge (Principal Consultant Applied Geosciences, Snowden) visited the Natougou project site between 13 March 2015 and 17 March 2015, where he observed the general site geology from outcrop and drill core, the core storage areas, sampling equipment and bulk density measurement equipment, along with the extents of artisanal mining activity in the area and the overall site layout. Neil Lincoln visited the site May 2015 where he observed the drill core at the core storage area, and walked the site of the proposed mine, process plant and infrastructure. Marius Phillips has not visited the site. Glen Williamson visited the Natougou gold project site on March for three days for the purpose of familiarizing himself with site conditions and reviewing the drill core to determine rock characteristics pertinent to estimating open-pit mineral reserves Jean-Sébastien Houle has not visited the site. Timothy Rowles (Regional Manager, Knight Piésold) visited the Natougou gold project site between the 4 February 2014 and 7 February 2014, and inspected the access road, proposed open pit, processing plant, water supply dams, regional water supply sources, tailings storage facility and mine infrastructure areas. Lycopodium Minerals Canada

47 NI TECHNICAL REPORT - RELIANCE ON OTHER EXPERTS Page RELIANCE ON OTHER EXPERTS Snowden has not reviewed the land tenure situation and has not independently verified the legal status or ownership of the properties or any agreements that pertain to the Tapoa Permit Group. Snowden has relied on information provided by Birimian relating to the legal status and ownership of the permits which comprise the Tapoa Permit Group. QP Jean-Sébastien Houle, Eng., relied upon Craig Wood, B. Sc. and Francis Barde, B.A., M. Env. of WSP for matters pertaining to environmental and permitting process as disclosed in Section \16.04\5047-STY-001_0 Lycopodium Minerals Canada

48 NI TECHNICAL REPORT - PROPERTY DESCRIPTION AND LOCATION Page PROPERTY DESCRIPTION AND LOCATION The Natougou gold deposit is part of the Tapoa Permit Group located in Burkina Faso, West Africa. The project lies approximately 320 km east of Ouagadougou, the capital of Burkina Faso in Figure 4.1. It is centred on UTM coordinates 326,500 me and 1,327,500 mn (WGS84z31). Figure 4.1 Location map Source: SEMAFO Snowden Mining Industry Consultants

49 NI TECHNICAL REPORT - PROPERTY DESCRIPTION AND LOCATION Page Type of Mineral Tenure The Mining Act of Burkina Faso (number /AN of 8 May 2003) provides the legal framework for exploration and mining activities within Burkina Faso. Exploration Permits up to 250 km 2 are granted for an initial period of three years, which may be renewed twice for periods of three years. At the second renewal, the permit area must be reduced by 25%, although an exemption from this requirement can be applied for at the time of the second renewal. Exploration activities are required to be reported annually to the Director General of Mines and Geology, along with the provisional program of work and budget expenses for the subsequent year. Exploitation (Mining) Permits, which are required for mining to commence, are granted for a term of 20 years (with unlimited five-year renewals) and the government carries the right for a 10% free carried interest (via a 10% equity interest in the local holding company). The application for an Exploitation Permit requires an Environmental and Social Impact Assessment ( ESIA ) and an environmental monitoring and management plan. The current Mining Act includes a sliding scale royalty to the government which is dependent on the gold price (3% to USD 1,000, 4% to USD 1,300 and 5% >USD 1,300). SEMAFO holds four contiguous Exploration Permits Dangou, Pambourou, Boungou and Bossoari, collectively known as the Tapoa Permit Group, covering approximately 773 km 2 within the Diapaga greenstone belt in Figure 4.2 in the southeast of Burkina Faso. Figure 4.2 Tapoa Permit Group Source: SEMAFO Snowden Mining Industry Consultants

50 NI TECHNICAL REPORT - PROPERTY DESCRIPTION AND LOCATION Page 4.3 Table 4.1 shows details for Birimian mineral tenements in Burkina Faso for the Tapoa Permit Group. Table 4.1 Tapoa Permit Group Exploration Permits Permit Name Grant Date First Renewal Date (Decree no.) Second renewal date (Decree no.) Final Permit Expiry Date Net Profit Interest Area (km 2 ) Boungou* 05/05/ /05/2012 (12-206) 05/05/2015 (approved) 05/05/2018 1% 250 Pambourou 28/09/2011 Bossoari 21/11/ /09/2014 (15-081) 21/11/2014 (15-083) 28/09/ /09/2020 1% /11/ /11/ % 39 Dangou** 16/12/ /12/2012 (14-018) 16/12/2015 (submitted) 16/12/2018 1% 250 * Includes Natougou deposit. **Second renewal submitted with a reduction exemption Boungou Permit The Boungou Exploration Permit, which hosts the Natougou deposit, covers an area of 250 km 2 and was originally granted on 5 May 2009 by Burkina Faso No. 2009/09 126/MCE/SG/DGMCC to Mr Saidou Pafadnam. In April 2010, Mr Pafadnam and MET BF Pty Ltd, a 100% subsidiary of Orbis, entered into a three-year option agreement. The original vendor of the permit (Mr Pafadnam) retains a 1% Net Profit Interest ( NPI ), payable upon any future gold sales under the terms of the option agreement (Convention d option pour permis de recherche Boungou) dated 21 April The permit was renewed for a further three years from May 2012 by Decree no. 2012/12 206/MCE/SG/DGMGC. On 21 January 2013, the Permit was transferred by Decree no. 2013/13 012/MME/SG/DGMG to Birimian Resources (100% SEMAFO). The permit has been granted until 5 May SEMAFO has applied for a renewal of the Boungou Exploration Permit for a three-year period (i.e. to 5 May 2018), along with exemption from the requirement for a 25% reduction in the permit area. Snowden understands that the second renewal application was submitted to the relevant authorities in Burkina Faso on 3 February 2015 and the renewal certificate was received by SEMAFO on 9 November The outline of the Boungou lease boundary is detailed in Table 4.2 and depicted in Figure 4.3. Snowden Mining Industry Consultants

51 NI TECHNICAL REPORT - PROPERTY DESCRIPTION AND LOCATION Page 4.4 Table 4.2 Boungou permit boundary (UTM Clarke 1880 Datum: Adindan, Zone 31N) Point (see Figure 4.3) Easting (me) Northing (mn) A 325,199 1,336,880 B 331,699 1,336,880 C 331,699 1,315,430 D 317,199 1,315,430 E 317,199 1,323,500 F 320,199 1,323,500 G 320,199 1,330,700 H 322,979 1,330,700 I 322,979 1,334,700 J 325,199 1,334,700 Snowden Mining Industry Consultants

52 NI TECHNICAL REPORT - PROPERTY DESCRIPTION AND LOCATION Page 4.5 Figure 4.3 Boungou permit boundary A B I J G H E F D C Source: SEMAFO (UTM - Clarke 1880 Datum: Adindan, Zone 31 North) Snowden Mining Industry Consultants

53 NI TECHNICAL REPORT - ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY Page ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY 5.1 Access and Physiography Access to Natougou is by means of Route Nationale RN04, an all-weather bitumen road from Ouagadougou, the capital of Burkina Faso, through Fada n Gourma to the Ougarou junction. From there, travel is via a laterite road to the property 60 km to the southeast. Fada n Gourma is the nearest town with basic hospital, hotel and limited resupply facilities. Any significant supplies must be sourced from Ouagadougou. The Natougou area is relatively flat and sits at an elevation of approximately 260 m above sea level. To the east and north of Natougou are mesas which rise approximately 10 m above the surrounding topography. A small hill is located in the very southern corner of the deposit. The land rises gently to the north, culminating in the height of land separating two watersheds. The main laterite access road into site is located along this ridge top. The road is reasonably well-drained and is accessible year-round to four-wheel drive vehicles. Numerous tracks allow for access to most places throughout the Natougou area in Figure 5.1. During the wet season (August to October), heavy rains may temporarily restrict vehicle movement in the immediate area of the deposit. Lycopodium Minerals Canada

54 NI TECHNICAL REPORT - ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY Page 5.2 Figure 5.1 Natougou access routes Source: SEMAFO Lycopodium Minerals Canada

55 NI TECHNICAL REPORT - ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY Page Climate The climate of Burkina Faso is semi-arid, with a rainy season from May to September, and a hot dry season from February to April. Average temperatures range between 16 centigrade ( C) overnight in the cool season to over 40 C during the day in the hot season. Average annual rainfall is approximately 900 mm, although large inter-year variability is common (Figure 5.2). During October to April, the climate is heavily influenced by the dry, dust-laden northwest trade wind known as Harmattan, which blows down from the Sahara Desert. Figure 5.2 Climate information for Ouagadougou Source: (accessed 12 Jan 2016) Lycopodium Minerals Canada

56 NI TECHNICAL REPORT - ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY Page 5.4 Monthly rainfall historic data has been obtained from the National Climatic Data Centre website for the Fada n Gourma weather station located approximately 100 km west from Natougou. Data covers the years 1920 to The average monthly rainfall data from this station is given in Table 5.1. August is on average the wettest month of the year with an average rainfall of mm, while December, January and February are the driest months with average rainfall close to 0 mm. Table 5.1 Average monthly rainfall for Fada n Gourma Month Average monthly rainfall (mm) January 0.1 February 0.5 March 6.7 April 26.0 May 75.3 June July August September October 29.8 November 1.6 December 0.9 Annual total Vegetation in the region is a mosaic of cultivated land and tropical acacia savannah. A photo overlooking the Natougou project area, taken during the March 2015 site visit, is presented in Figure 5.3 (the haze is dust from the Harmattan trade wind). Lycopodium Minerals Canada

57 NI TECHNICAL REPORT - ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY Page 5.5 Figure 5.3 Photo overlooking the Natougou project area (photo taken March 2015, looking southeast) Source: Snowden 5.3 Infrastructure There is no significant infrastructure near the site. Fada n Gourma is the nearest town with basic hospital, hotel and limited resupply facilities. Any significant supplies must be sourced from Ouagadougou. Lycopodium Minerals Canada

58 NI TECHNICAL REPORT - HISTORY Page HISTORY No exploration is known to have occurred on the Tapoa permits prior to 2010 when Orbis commenced soil and rock chip sampling. The soil and rock chip sampling was followed up in 2012 with a regional reverse circulation ( RC ) drilling program which resulted in the discovery of the Natougou gold deposit. Resource drilling commenced at Natougou in 2012 and culminated with an initial Mineral Resource estimate being completed by Snowden in August 2013 (Snowden, 2013), which was classified and reported in accordance with the 2004 edition of the Australasian Code for Reporting Exploration Results, Mineral Resources and Ore Reserves (the JORC Code ). Orbis completed further infill drilling at Natougou in 2014 and the Mineral Resource estimate was updated by Snowden in August 2014 (Snowden, 2014b) and was classified and reported in accordance with the 2012 edition of the JORC Code. Semafo initiated a takeover of Orbis in December 2014 and Orbis was delisted from the Australian Securities Exchange ( ASX ) on 16 March A further update of the resource was completed by Snowden in March 2015 for Birimian Resources (a wholly owned subsidiary of Semafo) reported in accordance with NI regulations. Between March 2015 and August 2015, Birimian completed an infill drilling program at Natougou aimed at upgrading the confidence in the resource estimate along with exploring targets proximal to the resource area. No modern production of gold has occurred within the Tapoa Permit Group. The central part of the Boungou permit has artisanal activity along the north to south trending drainage system. Extraction of gold by the local community from artisanal workings has occurred for an unknown period of time, with free gold recovered by gravity methods in gold pans or through simple sluicing methods. The vertical extent of the workings is unknown however they are thought to reach a maximum depth of approximately 20 m to 40 m, although the vast majority of the workings are less than 5 m deep. Snowden notes that the deeper workings are extremely localised and limited in extent. The total tonnage and grade of material extracted from artisanal workings at the Natougou deposit is unknown, however it is not considered to be material to the current Mineral Resource estimate. Snowden Mining Consultants

59 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page GEOLOGICAL SETTING AND MINERALIZATION 7.1 Regional geology The Tapoa Permit Group is located within the Birimian Gold Province in West Africa in Figure 7.1. The Birimian Gold Province is a world class gold province and hosts most of the major gold deposits in West Africa, notably in Ghana, Ivory Coast, Mali, Senegal and Burkina Faso. Figure 7.1 Overview of Regional Geology of Birimian Province Source: SEMAFO The Birimian is a Paleoproterozoic granite-greenstone province that developed during the Eburnean Orogeny (2195 Ma to 2067 million years ago (Ma); Pohl, 1988). Snowden Mining Consultants

60 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.2 The rocks of the Birimian are distinguished into two main groups (Lower Series and Upper Series), which despite the nomenclature, are regarded as being time equivalents (Appiah et al., 1991). The Lower Series consists of predominantly sediments, comprising black and grey phyllites, schists and meta-greywackes with subordinate volcanics, all of which have been metamorphosed to greenschist facies (Appiah et al., 1991; Dzigbodi-Adjimah, 1993). Phyllites and argillites are the most widespread lithology. The Upper Series consist of predominantly volcanic rocks, including andesitic tuffs and tholeiitic basaltic volcanics, with associated basic intrusives, and interbedded graphitic phyllites containing 1% to 2% pyrite. These volcanic dominated belts are generally parallel to the Natougou deposit and trend in a north-easterly direction. Individual belts are 15 km to 40 km in width and are spaced approximately 90 km apart Eburnian metamorphics and intrusives These rocks comprise much of the Lower to Middle Proterozoic of the Man Shield which is composed of granitoid rocks and high grade metamorphic equivalents (gneisses). These rocks yield Eburnian age datings between 2081±25 Ma and 1968±49 Ma, within a thermal event that started at 2127±65 Ma. The granitoids include quartz-diorite, tonalite, trondhjemite, adamellite, granodiorite and granite. The quartz-diorites are largely found cutting the greenstone accumulations, while the granites are developed within the sedimentary basins. The granitoids are divided into two types, namely, synorogenic foliated batholiths within the basin centres, and late orogenic, unfoliated intrusions within the Upper Series volcanic belts (Leube et al., 1990) Tarkwaian Sequence The Tarkwaian sequence is up to 2,600 m thick in the Tarkwa district. These sediments are present in all of the volcanic belts, being developed in the centres of each, with the exception of the Sefwi Belt where they occur on its eastern margin. No Tarkwaian rocks have been found within the main Birimian basins of the Lower Series. The Tarkwaian sediments overlie and truncate granitoids dated at 1890 Ma to 2061 Ma which cut the Birimian (Leube et al., 1990), but are older than a series of 1650 Ma mafic intrusives (Leube et al., 1990). The main Tarkwaian development over the Ashanti Belt is some 250 km long and averages about 16 km in width (Leube et al., 1990). Snowden Mining Consultants

61 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page Property Geology The Boungou permit, which contains the Natougou deposit, lies within the Diapaga greenstone belt, a northeast-southwest orientated belt that extends over 250 km in length and over 50 km in width in Figure 7.2. SEMAFO holds four contiguous permits, collectively known as the Tapoa Permit Group, covering approximately 70 km in strike length along the Diapaga belt. Permit scale mapping was conducted by Orbis geologists during the course of the 2014 field season, and updated by SEMAFO in 2015 in Figure 7.3. Laterite and alluvium cover extensive (lateral) portions of the permit however both are generally less than 10 m in thickness. Laterites cover the northern areas where mesas are more prevalent. Alluvium is more extensively developed in the southern half of the permit, associated with the larger drainages. The most significant rock type on the permit is a large mafic unit consisting predominantly of amphibolite with minor associated diorite and gabbro plutons, which occupies the central portion of the Boungou permit. The amphibolite hosts the mineralised shear at Natougou; the unit outcrops as small hills but can also be seen in valleys where small streams have cut through the cover sediments and exposed the bedrock. Although metamorphosed to an amphibolite, the presence of pillows and flow top breccia textures indicate that they are mafic volcanic flows. In addition, narrow volcaniclastic units are seen locally within this unit and seem to play a major role as host of the sub-horizontal shear zones due to the rheological differentiation. Along the western edge of the permit is a large granitoid batholith (tonalite, granodiorite and quartzdiorite). Lithic tuff, andesite and gabbro outcrop in the southeast corner of the permit along with minor meta-sediments (meta-sandstone and quartzite). Small intermediate to felsic intrusions (granodiorite, biotite-granite) occur in the southeastern part of the permit. Northwest-southeast oriented steeply dipping dolerite dykes intrude all the above lithologies at Natougou. The dolerites are fine to mediumgrained with abundant plagioclase and sparse hornblende and pyroxene phenocrysts. Snowden Mining Consultants

62 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.4 Figure 7.2 Property Geological Map Tapoa Permits within Diapaga Belt Source: SEMAFO Snowden Mining Consultants

63 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.5 Figure 7.3 Property Geological Map Boungou Permit Source: SEMAFO Snowden Mining Consultants

64 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page Natougou Geology SEMAFO re-logged all drillholes in 2015 in line with SEMAFO s lithological nomenclature and their assessment of the key features of the deposit. This re-logging was used by SEMAFO to establish a stratigraphic column for the project area. The following sections provide the details of the geological characteristics of the deposit Oxidation Four degrees of oxidation were identified by Orbis at Natougou in Figure 7.4, ranging from strongly oxidised through to fresh bedrock: Strongly oxidized material (SOX) has no remnant textures (saprolite zone). Moderately oxidized material (MOX) has observable textures and/or pieces of rock (saprock zone). Weakly oxidized rock (WOX) is relatively unaltered bedrock with oxidation only observed along fracture planes (transition between saprock and unoxidized bedrock). Fresh bedrock (FRS) has no observed oxidation on fractures. Figure 7.4 Oxidation Types Source: SEMAFO The base of complete oxidation is shallow at Natougou (Figure 7.5) with fresh rock encountered on average less than 10 m below surface. Snowden Mining Consultants

65 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.7 Figure 7.5 Photo of typical oxidation profile in core (BODD484) from Natougou Source: Snowden Only a very small portion of the mineralised shear is interpreted to be within the oxidized zones (Figure 7.6). This occurs along the very southeast edge of the deposit where the mineralized zone outcrops. Snowden Mining Consultants

66 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.8 Figure 7.6 Long section (looking northeast) showing oxidation impact on mineralization Source: Snowden Lithology The stratigraphy at Natougou is relatively simple and quite consistent from hole to hole. As shown in figure 7.7, the stratigraphy consists of two volcanic flows separated by a volcaniclastic unit. The footwall flow generally progresses upwards from a massive basalt flow to pillowed flows followed by flow breccia and volcaniclastics. The hangingwall is characterized by a medium grained volcanic flow (or sill). All these units are intruded by diorite and/or granodiorite sills, possibly originating from the felsic intrusion located immediately west of the deposit. Late dolerite dykes are also present and appear to be sub-vertical and strike northwest. Snowden Mining Consultants

67 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.9 Figure 7.7 Lithology Types Source: SEMAFO The Boungou Shear Zone, which hosts the main gold mineralization at Natougou, is located at the contact between the footwall and hangingwall volcanic units, where the volcanic flow top breccias have formed and the volcaniclastics deposited. The contact zone is thought to have served as an area of weakness, focussing the deformation. While the volcaniclastic units are not always present (although the intensity of the alteration can make it difficult to identify), the flow top breccias are interpreted to be ubiquitous across the deposit area. The Boungou Shear Zone is illustrated in Figure 7.8, which shows the shear zone (and associated gold mineralization) occurring between the upper volcanic units (V3B- GRE) and lower volcanic units (V3B-COU and V3B-BRE). Snowden Mining Consultants

68 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.10 Figure 7.8 Boundou Shear Zone at Contact between Hangingwall and Footwall Volcanic Units Source: SEMAFO Snowden Mining Consultants

69 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.11 The footwall and hangingwall basalts show similar geochemical signatures in the major oxides and the rare earth elements ( REE ) versus chondrite profiles (Figure 7.9, the lines on the graphs represent individual samples from Natougou). The breccias and volcaniclastics also show a similar trend, suggesting that the footwall volcanics were the likely source rock for the volcaniclastics. The altered breccias and volcaniclastics show an enrichment of the light REE typically characteristic of hydrothermal alteration effect. Previous exploration work used titanium (Ti) and zircon (Zr) profiles based on hand-held XRF data to locate the Boungou Shear Zone and separate the footwall and hangingwall volcanics. Both units were believed to be geochemically different. Recent work by SEMAFO suggests that proximal to the deposit, the difference is caused by the hydrothermal alteration which has preferentially altered the more porous volcaniclastics, breccias and pillowed flows in the footwall. Proximal exploration and deeper drilling below the Boungou Shear Zone shows Ti values within the footwall gradually increase to levels characteristic of the hangingwall values near the deposit. Similarly, Zr values slowly increase at depth towards hangingwall values. Additionally, the Zr values remain relatively high in holes where the Boundgou Shear Zone is less altered, particularly in the northwest part of the deposit. Snowden Mining Consultants

70 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.12 Figure 7.9 REE geochemical signatures for Natougou volcanic rocks Source: SEMAFO Snowden Mining Consultants

71 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.13 The felsic intrusive rocks at Natougou show geochemical signatures typical of Birimian rocks. The major, minor and REE values all point towards a tectonic arc setting following the mantle plume tholeiitic flows (Figure 7.10). In Figure 7.10, the red points correspond to intermediate intrusives, green to granodiorites and the blue points to rhyolites. For comparison, the shaded areas represent different intrusives identified at the Mana Property (also owned by SEMAFO). The grey area represents the early granodiorite intrusives associated with some of the Mana deposits (e.g. the Siou and Wona deposits). The brown area represents the typical arc related granite-granodiorite, while the orange area represents late dioritic and gabbroic intrusives. Figure 7.10 Geochemical Signature (Y+Nb vs Rb) of Natougou Felsic Rocks Source: SEMAFO The REE graphs (Figure 7.11) compare very well with felsic intrusives observed at the Mana deposit. As such, they are believed to have intruded the volcanics pre- to syn-deformation and premineralization. Snowden Mining Consultants

72 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.14 Figure 7.11 REE geochemical signatures for Natougou Felsic Rocks Source: SEMAFO Snowden Mining Consultants

73 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page Alteration Two types of alteration are significant with respect to the mineralised shear at Natougou. The two alteration types are locally termed the Boungou pink (BPK) and the Boungou bleached zone (BBZ), as shown in Figure 7.12 and Figure Figure 7.12 Alteration types Source: SEMAFO Figure 7.13 Core showing BPK and BBZ alteration Source: SEMAFO The BPK type is a pink-brown alteration consisting of fine-grained biotite-plagioclase and is generally present over the entire width of the mineralised shear zone. It is the schistosity developed by the alignment of the biotite that gives the mineralised zone the sheared texture. The BBZ type is a silica-sericite alteration generally developed as a discrete zone over widths of about 1 m within the BPK alteration. Variable amounts of quartz veining, often highly deformed (boudinaged or folded), can be observed within some of the BBZ. Small bands of silica-sericite are also interspersed within the wider zones of BPK alteration. Snowden Mining Consultants

74 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page Mineralization The main mineralized lode is interpreted as a flat-lying anticlinal shear that outcrops in the southeast and plunges gently to the northwest in Figure Figure 7.14 Natougou deposit global geometry Source: SEMAFO The mineralization has a strike length of approximately 2 km, striking towards a bearing of 315 and an across-strike length of approximately 1 km (towards 045 ). The mineralization is gently folded with the fold axis oriented along strike and the limbs dipping gently at approximately 15 in Figure Snowden Mining Consultants

75 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.17 Figure 7.15 Cross-section (A-B; top, looking northwest) and long section (C-D; bottom, looking northeast) through Natougou deposit Source: SEMAFO Mineralization is open across strike to the southwest as well as towards 315, although the shear zone is relatively deep in these directions. The average true thickness of the mineralization is approximately 4 m. Drilling to the west of the resource area up to the felsic intrusive suggests that the shear zone does not continue within the intrusive. The interpretation of the shear zone was therefore stopped at the intrusive/volcanic contact. Towards the northwest, an en echelon offset of the Natougou mineralization is noted in Figure The echelon happens near the apex of the fold, towards the west flank and is parallel to the axis. Although possibly caused by a fault, further examination of the alteration shows some difference from each side of the offset, which would not be expected with a late fault. In addition, detailed review of the core in the area did not show any evidence of faulting. Snowden Mining Consultants

76 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.18 Figure 7.16 Cross-section Showing en Echelon Offset in Northwest Portion of Deposit Source: SEMAFO Snowden Mining Consultants

77 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.19 Other minor footwall zones have been identified below the main shear zone. Although the footwall zones appear much less continuous and thinner than the main shear zone, some do contain significant gold mineralization. At the end of the 2015 delineation drilling program a short hole-extension program was completed to test the continuity of the footwall zone that was originally encountered in the central portions of the deposit by hole TPA0556, in which a secondary sheared and altered zone returned g/t Au over 2.69 m downhole. Nineteen holes were extended to reach the footwall zone, for a combined total of 371 m. Based on interpretation of the results, 13 additional holes (including the original discovery hole) intersected the footwall zone. The combined results suggests that the footwall zone is on average 1.5 m thick with local high grade results returned of up to g/t Au over 2.5 m downhole. The footwall zone is interpreted to be sub-parallel to and approximately 20 m below the main Boungou Shear. To date, the footwall zone has been traced over an area of approximately 300 m along strike by 90 m across strike, and remains open in all directions. The interpretation of this footwall zone was included in the current resource model in Figure Other zones of gold mineralization are believed to exist below the main shear and will be explored by SEMAFO further in Snowden Mining Consultants

78 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.20 Figure 7.17 Cross-section showing minor mineralization within the footwall below the main shear Source: SEMAFO Snowden Mining Consultants

79 NI TECHNICAL REPORT - GEOLOGICAL SETTING AND MINERALIZATION Page 7.21 Gold mineralization also occurs above the main mineralized lode in the north-western portion of the deposit and is referred to as the western hangingwall mineralized zone. This zone appears to be somewhat discontinuous and is currently poorly understood in terms of the nature and controls of this mineralization. The western hangingwall zone is included in the waste domain for the purposes of grade estimation and does not form part of the reported Mineral Resource. Arsenopyrite is almost invariably associated with the presence of gold in assayed samples. The percent arsenopyrite logged can be used as an initial identification of the mineralised lode. The trace element data from the handheld XRF in the form of elevated arsenic and sulphur values (As >500 ppm and S >5,000 ppm) corroborates the interpretation of the lode. Although not common, visible gold has been observed in core in some drillholes in Figure Figure 7.18 Photo of visible gold in core from Natougou Source: Snowden Snowden Mining Consultants

80 NI TECHNICAL REPORT - DEPOSIT TYPE Page DEPOSIT TYPE The Natougou deposit can be described as a West African shear zone hosted greenstone gold deposit. Gold mineralization is associated with biotite and silica-sericite alteration, along with disseminated sulphides such as pyrrhotite, pyrite and minor arsenopyrite and chalcopyrite with occasional free gold. The mineralization is structurally controlled and is hosted primarily within a large shear zone and its associated alteration. Snowden Mining Consultants

81 NI TECHNICAL REPORT - EXPLORATION Page EXPLORATION Regional soil sampling and rock chip sampling programs were commenced by Orbis in Permit scale mapping was conducted by Orbis geologists during the course of the 2014 field season and updated by SEMAFO in Results of the mapping are detailed in Section 7. Orbis defined a large-scale high order (+50 ppb Au) gold-in-soil anomaly in the area surrounding the Natougou discovery. The soil anomaly, defined within a 6 km by 4 km survey area, includes multiple zones of higher-order anomalism that have received minimal exploration drilling to date. The higher order soil anomalies present as priority areas for follow-up exploration. Figure 9.1 is a map showing the results of soil sampling indicating significant anomalies surrounding the Natougou deposit, probably in part related to drainage features. The results of rock chip samples (>0.1 g/t Au) across the Tapoa Permit Group are presented in Figure 9.2. The group of anomalous rock chip samples immediately to the north of the Natougou deposit coincide with the +50 ppb Au soil anomaly and are associated with extensive artisanal workings in the area. Additionally, Orbis completed 13 trenches between November 2014 and January 2015, with an average length of approximately 38 m. All of the trenches were within the Boungou permit. The trenches were hand dug to an approximate depth of 1.5 m and chip samples collected at 1 m intervals from the side wall close to the base of the trench. Nine of the trenches showed no significant intersections. The best results were obtained from trench BOTR006, which returned an intersection of 9 m at 9.43 g/t Au (horizontal width; not true width) based on a lower cut-off of 1 g/t Au (or 12 m at 7.15 g/t Au if a lower cut-off of 0.2 g/t Au is used). Further work is required to interpret the trench sampling results. Snowden Mining Industry Consultants

82 NI TECHNICAL REPORT - EXPLORATION Page 9.2 Figure 9.1 Tapoa Permit Group exploration soil anomaly Source: SEMAFO Snowden Mining Industry Consultants

83 NI TECHNICAL REPORT - EXPLORATION Page 9.3 Figure 9.2 Tapoa Permit Group exploration rock chip samples >0.1 g/t Au N Source: Snowden Snowden Mining Industry Consultants

84 NI TECHNICAL REPORT - DRILLING Page DRILLING Drilling at Natougou was performed by a combination of RC and diamond drilling. The diamond drillholes were pre-collared using RC drilling down to approximately 10 m above the interpreted top of mineralisation. Diamond tails (HQ diameter) were used to complete the holes. These are recorded as multi-purpose (MP) holes in the database. A limited number of diamond holes were cored from the surface, predominately on two 10 m by 10 m close spaced drilling panels. The total dataset comprises 1,349 drillholes, of which 1,198 were used for the Natougou resource estimate. SEMAFO drilled 624 holes in 2015; the remaining holes were drilled by Orbis between 2012 and Table 10.1 summarizes the drilling information for the Tapoa Permit Group and a collar location plan is provided in Figure Table 10.1 Tapoa Permit Group drillhole details Drill type Number of holes Pre-collar total length (m) Diamond tails total length (m) Average total depth (m) Total length (m) Diamond 174-9, ,029 Multi-purpose ,937 19, ,505 RC , ,712 Total 1,349 86,649 28, ,246 The 2015 drilling program by SEMAFO included proximal exploration drilling at Natougou. A total of 16,222 m of drilling was completed including a sterilization drilling program under the proposed locations of surface infrastructure. The highlight of this program was the discovery of the SW Extension Zone, located within the Boungou Shear Zone to the southwest of the known mineralization. The SW Extension Zone is essentially the down-dip extension of the main shear zone along the west flank of the anticlinal shape of the shear zone. Based on current drilling, the zone seems to form a northwest trending shoot of mineralization, sub-parallel to the two known zones of mineralization. The south-western portion of the Boungou Shear Zone appears to dip to the southwest until it reaches the intrusive body where the shear zone is interpreted to stop. Gold mineralization within the SW Extension is comparable to that observed within the main deposit area. Grades obtained also suggest similar gold concentrations with results of up to g/t Au over 10 m (TPA0406). The mineralization remains open to the northwest along the trend of the shoot. Further drilling is required within the SW Extension Zone to increase the confidence above an Inferred level. A contour plot of the grade times thickness (i.e. Au x m) from drilling at Natougou is provided in Figure Snowden Mining Consultants

85 NI TECHNICAL REPORT - DRILLING Page 10.2 Figure 10.1 Drillhole collar location plan for the Tapoa Permit Group N Source: SEMAFO Snowden Mining Consultants

86 NI TECHNICAL REPORT - DRILLING Page 10.3 Figure 10.2 Grade x thickness contour plot of Natougou deposit 10.2 Sampling techniques RC drillhole sampling SEMAFO 2015 Material from the 2015 RC drilling by SEMAFO was collected every 1 m into a poly bag (plastic bag) directly from the cyclone on the drill rig. The bags were pre-labelled with the hole ID, metre from and metre to. Historically, samples were weighed by Orbis to estimate recovery and three magnetic susceptibility readings were taken using a KT-10 magnetic susceptibility metre. This practice was abandoned by SEMAFO in Snowden Mining Consultants

87 NI TECHNICAL REPORT - DRILLING Page 10.4 For the 2015 RC drilling, approximately 30 kg to 40 kg of material was collected by the cyclone for each 1 m RC interval, which was reduced using a tiered riffle splitter to obtain a split of about 2 kg which was packed in a poly bag. Sample tickets were placed into each poly bag and the hole ID and sample depth recorded on the remaining ticket stub. The riffle splitter was cleaned after each sample with a brush. Sample bags were then transported to the on-site preparation laboratory for crushing and pulverizing and the sample pulps transported to the SGS laboratory in Ouagadougou for assaying. Quality control samples, including reference materials and blanks were also submitted with these samples. A second split of the same size was kept on site for reference and the rest of the RC sampled material discarded. A small sample of chips from each 1 m interval was removed with a sieve, washed and placed in labelled chip trays for logging and future reference. SEMAFO indicated that RC samples were collected dry 99% of the time. Orbis 2012 to 2014 For RC drilling conducted between 2012 and 2014 by Orbis, material from the RC drilling was collected every 1 m into a poly bag directly from the cyclone on the drill rig. The bags were pre-labelled with the hole ID, metre from and metre to. The sample was then weighed to estimate the drilling recovery. The magnetic susceptibility was also measured using a KT-10 magnetic susceptibility meter. Three readings were recorded in the field logging template and the average of the three used. A small sample of chips from each 1 m drilling run was removed with a sieve, washed and placed in labelled chip trays for future reference. All 1 m RC samples from the 2012 to 2014 drilling were composited to 4 m samples. A riffle splitter was used to split the 1 m samples to an approximate weight of 500 g resulting in an approximate 2 kg composite sample representing a 4 m interval of drilling. Sample tickets were placed into a poly bag and the hole ID and sample depth recorded on the remaining ticket stub. The riffle splitter was cleaned after each sample with a brush. Quality control samples, including reference materials, field duplicates and blanks were also submitted with these samples. The split 2 kg samples were placed in a poly bag and transported to camp to await shipment to the SGS laboratory in Ouagadougou. The original 1 m drill bags from each hole were transported directly after splitting and arranged in order by depth drilled at the bag farm. With the exception of initial RC drillholes, Orbis routinely collected a second 2 kg to 3 kg split of the 1 m samples for storage in an enclosed shed on site. Snowden Mining Consultants

88 NI TECHNICAL REPORT - DRILLING Page 10.5 The assay results from the 4 m composites were used by Orbis to select 1 m intervals for re-sampling. A threshold of 0.25 g/t Au was used to flag 4 m composite samples for the second phase of assaying. The 4 m composite samples either side of the identified mineralized section were also split to bracket potential mineralization. A riffle splitter was used to split the 1 m sample to an approximate 2 kg to 3 kg sample. This was placed in a pre-labelled polythene bag. A corresponding sample ticket was placed in each bag and the hole ID and depth recorded on the stub remaining in the ticket book. The riffle splitter was cleaned after each sample. Quality control samples were also submitted with these samples. Records of sampling show that 2012 to 2014 RC samples were collected dry 99% of the time Diamond drillhole sampling Diamond core samples were collected on 1 m intervals up to and including hole BODD089, thereafter all diamond sampling was either sampled on a maximum of 1.2 m intervals or to the lithological/alteration/mineralization boundaries, with a minimum sample length of 0.2 m. The core was cut in half lengthwise using a diamond saw and the sampled half core placed in a plastic bag and labelled with the hole ID and depth. A sample ticket labelled with the hole ID and depth was also placed in the bag. Quality control samples were also submitted with these samples Drill sample recovery Drilling recovery for diamond core has been reviewed to ensure appropriate core length was recovered. The global recovery for diamond drill core is acceptable, with an average recovery of 98% for drill core in Figure Snowden Mining Consultants

89 NI TECHNICAL REPORT - DRILLING Page 10.6 Figure 10.3 Histogram of core recoveries Source: SEMAFO Snowden understands that SEMAFO did not weigh the RC drill cuttings for the 2015 RC drilling. SEMAFO indicated that it does not consider it as an accurate measure to evaluate if appropriate material weight has been recovered. Whilst Snowden agrees that RC recovery estimates are not ideal due to assumptions relating to the drillhole volume and bulk density, Snowden recommends that SEMAFO monitor RC drilling recovery for future drilling to ensure that the sample quality is maintained. For reference, historical average recovery for RC drilling conducted by Orbis from 2012 to 2014 was estimated to be 84% in Figure Snowden Mining Consultants

90 NI TECHNICAL REPORT - DRILLING Page 10.7 Figure 10.4 Histogram of RC drilling recoveries Source: SEMAFO It is Snowden s opinion that the recoveries are acceptable for representative sampling and subsequent Mineral Resource estimation Logging Drill core and RC chips were logged according to lithology, alteration, sulphide percentage, vein composition and percentage, and structures. Prior to 2013, logging was hand written on a template and manually entered into the database; however from 2013 all logging was entered directly into Micromine Geobank Mobile using portable tablets. The logging geologist used a library of geology codes to log in order to minimize subjectivity. RC chips were logged at the drilling site, whilst the diamond core was transported to the shed in camp for detailed logging. A set of field photographs of the core trays were also taken at this time. Snowden Mining Consultants

91 NI TECHNICAL REPORT - DRILLING Page Location data and survey methods Data Spacing and Distribution The dominant drillhole spacing at the Natougou deposit is 40 m along strike (315 ) by 40 m across strike (045 ). Within the central-southern portion of the deposit, the drilling has been infilled to a nominal 20 m by 20 m grid. Additionally, two 50 m square areas within the 20 m drilling area have been further infilled to 10m by 10m to assess the short range grade continuity Location of Data Points All drillhole collars were re-surveyed in 2015 using a LEICA TCR 307 Total Station system. Data from the instrument was downloaded directly to a laptop and processed using Leica Survey Office software. Data was processed in COVADIS by the contractor. The Total Station System has a reported accuracy of 3.54 mm horizontally and vertically. The coordinate system basis used is WGS84 Zone 31N. It is Snowden s opinion that the collar coordinates have been accurately surveyed and are appropriate for use in Mineral Resource estimation Downhole Surveys The 2015 drillholes were all surveyed downhole using a Reflex GYRO electronic surveying tool. Both the azimuth and dip were recorded at 5m intervals (approximately) downhole. The Reflex GYRO system is not affected by magnetic interference from either the drill rods or magnetic minerals. The accuracy of the Reflex GYRO system is reportedly ±0.5 when measuring azimuth and ±0.2 when measuring dip angle. Pre-2015 drillholes were surveyed downhole using a Reflex EZ-shot electronic surveying tool. Both the azimuth and dip were recorded at 6 m intervals (approximately) downhole, within the PVC casing, and then at 30 m intervals until the bottom of hole was reached. Magnetic susceptibility readings were taken by the Reflex EZ-shot device to indicate if the azimuth readings were being materially distorted by magnetic minerals. No readings have been rejected based on anomalous magnetic susceptibility. Readings with apparent errors were re-taken. The Reflex EZ-shot camera has an accuracy of ±0.5 when measuring azimuth and ±0.2 when measuring dip angle Orientation of data in relation to geological structure As the deposit is predominantly flat lying, 54% of drillholes were drilled vertically. Approximately 37% of drillholes were drilled at declinations of between 75 and 85 and 5% between 65 and 75. Additionally, some holes were drilled at a declination of about 60 towards a direction of either 045 or 235 magnetic, depending on which side of the fold hinge the hole was drilled on. All drillholes were aligned to be near-perpendicular to the orientation of the mineralization. Snowden Mining Consultants

92 NI TECHNICAL REPORT - DRILLING Page Topography A topographical survey was completed on 18 November 2014 by Southern Mapping to produce a digital terrain model (DTM) of a portion of the Boungou Permit over an area that extends past the current resource estimate area in Figure The topographical survey used an aircraft mounted LiDAR system that scans the ground below with a 70 khz laser frequency rate, resulting in a dense DTM of the ground surface and objects above the ground. The survey was flown at a height of approximately 1,200 m. The survey was initially calculated on the ITRF2008 datum, UTM31N projection with ellipsoidal heights. Precise Point Positioning and GPS observations were all referenced to the ITRF2008 datum; therefore, no WGS84 datum adjustment was made to the data. The ellipsoidal heights were transformed to orthometric heights using the EGM2008 geoidal model. Figure 10.5 DTM produced from the 2014 LIDAR survey (black outline = mineralization extents; DTM coloured by 5 m intervals) Source: Snowden SEMAFO noted some minor discrepancies in elevation between the drillhole collar survey and the LIDAR survey. Consequently, SEMAFO elected to use the LIDAR coordinate as the collar elevation (i.e. Z coordinate) in the database for the current Mineral Resource estimate. Snowden does not believe that this discrepancy is material to the Mineral Resource estimate. Snowden Mining Consultants

93 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page SAMPLE PREPARATION, ANALYSES AND SECURITY 11.1 Sample Preparation Drill core and RC samples were submitted to three different laboratories in 2015 with the vast majority (89%) being assayed at the SGS laboratory in Ouagadougou. Approximately 6% of the 2015 samples were assayed at the SGS laboratory in Bamako, Mali and the remaining 5% were sent to the ALS laboratory in Ouagadougou. Samples from drilling conducted by Orbis between 2012 and 2014 were submitted predominately to the SGS laboratory in Ouagadougou. Both SGS laboratories were recognised in July 2015 by the South African National Accreditation System (SANAS) for meeting the requirements of the ISO/IEC standard. As approximately 95% of the samples were prepared and assayed by SGS, the following only describes the SGS sample preparation procedures. As of 2015, all drill core and RC drill cutting samples are prepared on site where SGS has set-up a mobile sample preparation facility; prior to this (i.e. for drilling conducted by Orbis), samples were prepared in full at the SGS Ouagadougou laboratory. Snowden notes that the on-site sample preparation procedure used for the 2015 samples is not materially different from earlier samples prepared in Ouagadougou. The 2015 on-site sample preparation procedure is as follows: Samples are first registered, ordered and then weighed. The data is directly captured in LIMS (Laboratory Information Management System). Samples are oven dried at a nominal 105 C for 3 to 6 hours depending on the material. Approximately 2 kg of sample is crushed to 80% passing 2 mm. - One in 20 samples is screened to ensure 80% passing 2 mm. The 2 kg sample is split in half using a rotary or riffle splitter. One half is used for pulverization and the other half stored as a coarse reject. - One preparation duplicate is selected every 50 samples. A 1kg split of the 2mm material is then pulverized to 85% passing 75 µm in a bowl and puck pulverizer. - One in 20 samples is screened to ensure 85% passing 75 µm. - One prep blank in 50 samples is taken and assayed for contamination \5047-STY-001_B Snowden Mining Industry Consultants

94 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page 11.2 A 250 g sub-sample is collected (by scooping) from the 1 kg split and conditioned for shipping to the Ouagadougou laboratory. - The remaining material (pulp reject) is returned to the original bag (or a plastic bag if the original is not suitable) and stored on-site. All preparation equipment is flushed with barren material prior to the commencement of the job. Cleaning of equipment (e.g. crushers and pulverisers) is by compressed air which is done between each sample. Snowden visited the SGS Ouagadougou laboratory in April 2014 and March 2015 as discussed in Section Analytical techniques Primary Analysis All samples submitted to SGS in Ouagadougou are analysed for gold using fire assay with the gold content determined by atomic absorption spectroscopy ( AAS ). All samples are assayed by fusing a 50 g sample with a litharge based flux followed by cupellation, dissolving the prill in aqua regia and determining the gold content by AAS. This technique has a detection limit of 0.01 ppm Au. It is Snowden s opinion that the analytical technique used for Natougou samples is appropriate Handheld XRF Analysis All pulps from primary analysis are collected and returned to the SEMAFO camp facilities. The pulps are then analysed by handheld XRF (Innov-X Omega series) for a suite of elements. The data is digitally downloaded from the machine and transferred to the database for storing with the primary analyses. The machine is calibrated by insertion of blanks and standards at positions in the sample sequence where primary blanks and standards were inserted. The calibration standards include NIST2702 and NIST2781, as well as SiO 2 as a blank. The elements recorded and their detection limits ( LOD ) are included in Table \5047-STY-001_B Snowden Mining Industry Consultants

95 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page 11.3 Table 11.1 Innov-X Omega series elements and detection limits Element OMEGA Series LOD Unit Element OMEGA Series LOD Unit Mg 0.5 % As 1 ppm Al 0.1 % Se 1 ppm Si 0.1 % Rb 1 ppm P 250 ppm Sr 1 ppm S 105 ppm Zr 1 ppm Cl 30 ppm Mo 0.5 ppm K 15 ppm Ag 4 ppm Ca 15 ppm Cd 4.5 ppm Ti 2.5 ppm Sn 7.5 ppm V 2.5 ppm Sb 8 ppm Cr 2.5 ppm W 8 ppm Mn 1.5 ppm Au 6 ppm Fe 1.5 ppm Hg 1 ppm Co 5 ppm Pb 2.5 ppm Ni 2.5 ppm Bi 1 ppm Cu 2.5 ppm Th 1 ppm Zn 2.5 ppm U 2 ppm SEMAFO recognizes the limitations of the handheld XRF data, which is primarily used to aid the geological interpretation and as an indicative assessment of the concentration of various elements within the rocks encountered at Natougou Quality of Assay Data and Laboratory Tests Internal Laboratory Quality Assurance and Quality Control SGS undertake an internal QAQC process involving standards, blanks and duplicates. Each analysis batch consists of 84 samples, of which 10 are quality control ( QC ) samples, comprising four reference materials, two duplicates (taken before crushing), two pulp duplicates, one blank (pulp) and one coarse blank. SGS also participate in regular round robin programs to monitor for bias. A minimum of 5% additional pulp check assays are performed on all batches (depending on the number of anomalies present within a given batch). SGS provide SEMAFO with a monthly report of results from the internal QAQC program. Snowden previously reviewed four of these reports and consider the results and the material within the reports informative and beneficial to SEMAFO \5047-STY-001_B Snowden Mining Industry Consultants

96 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Laboratory Inspections A brief laboratory inspection of the SGS facilities in Ouagadougou was completed by Mr. John Graindorge (Snowden) and Ms. Ann Ledwidge (formerly Orbis) on 16 March 2015 with supervision by the Laboratory Manager, Helena Bouda. The sample preparation areas were organised and clean of excessive dust and many of the crushers included automated rotary splitting devices. The areas were clean and well maintained. The fire assay balances were clean and calibration records kept. All machines supported data capture directly into the Laboratory Information Management System ( LIMS ). The fire assay flux addition area used a semi-automated multi pot system with mixing of the sample+flux in a tumbler. The fire assay process (fusion and cupellation) was well organised and maintenance records are kept. Acid digestion rooms had clearly set out equipment locations and suitable safety systems. Spectra atomic absorption machines are utilised for analysis and have clear maintenance and calibration records kept. In the majority of cases, laboratory procedures were visible and easily available to staff. Snowden considers SGS Ouagadougou to be a suitable laboratory to be the primary analysis facility for samples. Snowden has not inspected the on-site sample preparation facility at Natougou which was used to process the 2015 drill samples. The results of the quality control data (discussed below) suggests that reasonable precision is being achieved Assay Reports All assay reports from the primary assay laboratory (SGS and ALS) are submitted to SEMAFO as digital data files and as PDF certificates. Between April and September of 2015, a total of 829 batches of samples (61,413 samples) were sent to SGS and 38 batches of samples (2,919 samples) were sent to ALS, for a grand total of 64,332 new samples added to the database since the previous resource estimate. An additional 415 quarter core duplicates were assayed at SGS along with 25 quarter core duplicates at ALS. Pulp duplicates from SGS and ALS were also assayed at SEMAFO s Burkina Faso mine laboratory, which acted as a check laboratory, with 806 and 247 pulp duplicates from SGS and ALS analyzed respectively Certified Reference Materials Between April 2015 and September 2015, fifteen different certified reference materials ( CRMs ) were inserted in the sampling sequence by SEMAFO (the same as used by Orbis from 2012 to 2014). The CRMs (pulps) were purchased from Ore Research and Exploration Pty Ltd and Rocklabs Limited and were inserted at a rate of one in every 20 samples. Details of the standards are given in Table The CRMs range from g/t Au up to g/t Au \5047-STY-001_B Snowden Mining Industry Consultants

97 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page 11.5 Table 11.2 CRM Details for Natougou CRM ID Source Expected Au (g/t Au) 1 standard deviation (g/t Au) 95% confidence level (g/t Au) STD12 OREAS STD16 OREAS STD17 OREAS STD18 OREAS STD62 OREAS STD204 OREAS OxH122 ROCKLABS OxH82 ROCKLABS Oxi121 ROCKLABS SF67 ROCKLABS SG66 ROCKLABS SH69 ROCKLABS Si54 ROCKLABS SJ80 ROCKLABS SK78 ROCKLABS The CRM results were analyzed by examining control charts and by assessing the location of the CRMs within the sample batch. Although the standards are not matrix matched, Snowden considers them appropriate for assessing laboratory analytical accuracy with respect to Natougou mineralization. CRM results 2015 In 2015, a total of 1,764 CRM samples were submitted with primary samples to the SGS laboratory in Ouagadougou. Initially, 79 (or 4.5%) failed the initial three standard deviation control limits and reruns were requested by SEMAFO. After re-assaying, only 19 failures remained, which equates to a failure rate of approximately 1.1%. As can be seen in Table 11.3 and Figure 11.1 and Figure 11.2, most of the failures were caused by one CRM (OREAS12a). Due to the excessive failures for this CRM, SEMAFO ceased using OREAS12a after 8 August \5047-STY-001_B Snowden Mining Industry Consultants

98 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page 11.6 Table CRM results Assayed at SGS Figure 11.1 CRM Control Chart (z-score) Results Assayed at SGS Source: SEMAFO \5047-STY-001_B Snowden Mining Industry Consultants

99 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page 11.7 Figure 11.2 Individual CRM Control Chart Results Assayed at SGS \5047-STY-001_B Snowden Mining Industry Consultants

100 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page 11.8 Source: SEMAFO \5047-STY-001_B Snowden Mining Industry Consultants

101 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page 11.9 A total of 78 CRM samples were submitted with primary samples to the ALS laboratory in Ouagadougou. Only one CRM result failed the three standard deviation control limits; a rerun of this batch returned results within the control limits. The results are summarised in Table 11.4 and shown in Figure 11.3 and Figure Table 11.4 CRM results assayed at ALS Figure 11.3 CRM Control Chart (z-score) results assayed at ALS Source: SEMAFO \5047-STY-001_B Snowden Mining Industry Consultants

102 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Figure 11.4 Individual CRM Control Chart results assayed at ALS Source: SEMAFO The results of the CRM assays for the 2015 drilling show reasonable analytical accuracy is being achieved by both SGS and ALS for all grade ranges, with no evidence for analytical bias. CRM results 2012 to 2014 Results from CRMs inserted into the sample batches by Orbis between 2012 and 2014 are detailed in the March 2015 NI Technical Report for the Tapoa Permit Group (Snowden, 2015). The results show that the majority of CRM assays were within three standard deviations of the expected value with minimal failures and no evidence for analytical bias Blanks Blanks results 2015 In 2015, 1,654 and 73 coarse blank samples were submitted with the primary diamond core samples to SGS and ALS respectively. Two blank samples were submitted per batch of 74 (SGS) and 78 (ALS) samples. The blank material is sourced from the Bobu quarry and comprises of non-mineralised sedimentary rock. The results are summarized in Figure 11.5 and Figure \5047-STY-001_B Snowden Mining Industry Consultants

103 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page There were only two samples that showed some evidence of possible sample contamination. Those batches were rerun and validated. The results from the blanks for the 2015 drilling data show no evidence for systematic contamination of samples at either SGS or ALS during sample preparation or assaying. Figure 11.5 Coarse Blank Control Chart results assayed at SGS Source: SEMAFO \5047-STY-001_B Snowden Mining Industry Consultants

104 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Figure 11.6 Coarse Blank Control Chart results assayed at ALS Source: SEMAFO Blanks results 2012 to 2014 Results from blanks inserted into the sample batches by Orbis between 2012 and 2014 are detailed in the March 2015 NI Technical Report for the Tapoa Permit Group (Snowden, 2015). Blanks were inserted at a ratio of one blank sample for every 50 samples. Only minimal failures were recorded and the results suggest that there was no significant contamination of samples during sample preparation and assaying Field Duplicates Field duplicate results 2015 In 2015, there were a total of 415 and 25 field duplicate samples (comprising quarter core duplicates for diamond core) submitted with the primary core samples to SGS and ALS respectively. No RC field duplicates were taken as SEMAFO indicated that they considered that the 40 m by 40 m and 20 m by 20 m drill programs were designed to test the main lode in its entirety by the more reliable core drilling sampling. As such, Snowden is unable to comment on the precision of the 2015 RC sample data. It is strongly recommended that SEMAFO include field duplicates in all future RC drilling to allow the precision of the sample data to be assessed \5047-STY-001_B Snowden Mining Industry Consultants

105 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page A quantile-quantile ( QQ ) plot is used to compare population distributions in order to assess potential bias. For results to be considered unbiased, they should plot along the 1:1 line. A QQ plot for the 2015 field duplicate results from the SGS laboratory in Figure 11.7 indicates no significant bias exists between the sample populations. The results from the samples assayed at ALS are not illustrated due to the limited number of pairs however the statistics indicate no significant bias. Figure 11.7 Original vs. Field Duplicate log QQ plot (SGS) Source: SEMAFO A "mean vs % difference plot" was used to assess the precision and bias of the field duplicates. The mean of duplicate pairs is calculated along with the relative percentage difference between the two values, calculated as the difference divided by the mean multiplied by 100. The mean is then plotted against the relative difference for each sample pair. This plot allows the precision and bias to be assessed in each grade range. Figure 11.8 indicates no significant bias exists between the sample populations at different grade ranges and that the precision is reasonable \5047-STY-001_B Snowden Mining Industry Consultants

106 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Figure 11.8 Original vs. Field Duplicate Relative diff vs mean plot (SGS) Source: SEMAFO Precision plots are used to assess the repeatability (i.e. precision) of the duplicate results by plotting the average of the duplicate results against the half absolute difference of the pair. The number of sample pairs plotting between 5%, 10%, 20% and 100% difference lines are assessed as a measure of the precision of the results. As the grade approaches the detection limit of the analytical technique the precision, expressed as a percentage, decreases. A line of significance is typically plotted on the graph to exclude any results close to the detection limit to account for this. Field duplicate results plotted on a precision plot show approximately 32% of the duplicate results plotting above the 20% difference line in Figure Field duplicate results are typically associated with a lower degree of precision compared to pulp duplicates especially for deposits were visible gold is present \5047-STY-001_B Snowden Mining Industry Consultants

107 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Figure 11.9 Original vs. Field Duplicate Precision Plot (SGS) Source: SEMAFO Overall the populations compare reasonably well, with some outliers at higher grades, typical of gold deposits containing visible gold particles. Snowden considers there is no evidence to suggest that the primary sample varies significantly from the duplicate sample and that SEMAFO have achieved reasonable precision during the sampling and assaying process for the diamond core samples. As no field duplicates were collected for the RC drilling, Snowden is unable to comment on the precision of this data and as such it is recommended that SEMAFO include field duplicates in all future RC drilling. Field duplicate results 2012 to 2014 Results from field duplicates for both RC and diamond core inserted into the sample batches by Orbis between 2012 and 2014, are detailed in the March 2015 NI Technical Report for the Tapoa Permit Group (Snowden, 2015). The results of field duplicates for the 2012 to 2014 drilling show reasonable precision, with approximately 25% of duplicate pairs having a difference (half absolute difference) of more than 20% on a precision plot (i.e. plotting above the 20% difference line), with no bias evident between the duplicate pairs Umpire Laboratory Checks A set of 806 and 247 pulp duplicate samples from SGS and ALS respectively were submitted to the SEMAFO mine laboratory at Mana (referred to as SMF BF in the figures below) as an umpire check of results from the primary laboratories. Basic statistics and log QQ plots are presented in Figure and Figure \5047-STY-001_B Snowden Mining Industry Consultants

108 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Basic statistics show no significant difference between the two populations and there is no bias evident from the QQ plots. Figure Umpire (SMF BF) vs. Original Pulp (SGS) duplicate log QQ plot Source: SEMAFO \5047-STY-001_B Snowden Mining Industry Consultants

109 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Figure Umpire (SMF BF) vs. Original Pulp (ALS) Duplicate log QQ plot Source: SEMAFO Precision plots of the results for both SGS and ALS indicate 30% to 35% of the results falling above the 10% precision line, at all grade ranges in Figure and Figure \5047-STY-001_B Snowden Mining Industry Consultants

110 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Figure Umpire (SMF BF) vs. Original Pulp (SGS) Duplicate Precision Plot Source: SEMAFO Figure Umpire (SMF BF) vs. Original Pulp (ALS) Duplicate precision plot Source: SEMAFO \5047-STY-001_B Snowden Mining Industry Consultants

111 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Overall the populations compare well and no bias is evident between the duplicate pairs. Snowden notes that the precision shown by the pulp duplicates has decreased slightly compared to pre-2015 umpire checks. This may be due to a combination of factors, including the addition of the on-site sample preparation facility at Natougou along with the use of the Mana mine site laboratory for the umpire checks. Whilst not ideal, the results indicate acceptable precision is being achieved at both primary laboratories (SGS and ALS, Ouagadougou). Snowden recommends that SEMAFO review the laboratory pulp sub-sampling procedures at the two primary laboratories and mine site laboratory, which may be contributing to the elevated imprecision shown by the umpire pulp duplicates. The 2012 to 2014 umpire check assays are discussed in detailed in the March 2015 NI Technical Report for the Tapoa Permit Group (Snowden, 2015). The results of the check assays show reasonable precision, with approximately 30% of duplicate pairs having a difference (half absolute difference) of more than 10% on a precision plot (i.e. plotting above the 20% difference line), with no bias evident between the duplicate pairs Sample Security RC and core samples are all crushed and pulverized on site by SGS personnel. Sample pulps are prepared for shipment and stored on site in a secure locked room until shipment to Ouagadougou. Transportation occurs on a regular basis with security guards. Personnel releasing the samples for shipment to the laboratory assume responsibility for the sample security and paperwork with recorded sample numbers accounted for prior to shipment to the laboratory. Both SGS and ALS laboratories in Ouagadougou check the received samples against the paperwork and sign-off on the receipt Bulk Density SEMAFO continued collecting bulk density measurements on site by using the Archimedes immersion technique (weight in air divided by the difference between the weight in air and the weight in water). The methodology, which remains unchanged from the previous estimate, is as follows: Select 10 cm to 15 cm whole core samples in fresh, altered and mineralized shear zones. Set scale to zero and place the sample directly on the scale and record the reading. Set the scale to zero and: - put the sample on the cradle and ensure sample is fully immersed in the water (the cradle should not touch the bottom or the sides of the water bucket) - wait until reading is stable and then record the reading. Calculate the bulk density (weight of sample in air divided by weight of sample in air minus weight of sample in water) \5047-STY-001_B Snowden Mining Industry Consultants

112 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page There has not been any special procedure (wax coating/plastic wrap, oven drying) given to the weathered and/or porous core samples. As such, SEMAFO recognizes the sample weight in water of the porous/friable samples will tend to be biased due to water entering the pore spaces, resulting in the bulk density being overstated for weathered/porous samples. Given the shallow nature of the weathering profile, SEMAFO and Snowden do not believe that the higher bulk density values within the weathered material will materially impact on the Mineral Resource global tonnage. The equipment in Figure 11.14, whilst not in use at the time of the site visit was found to be suitable. It is Snowden s opinion that the procedure for measuring bulk density at Natougou is appropriate for Mineral Resource estimation, however there is scope for improvement, including using larger sample sizes as discussed in Section 26. A plot of the bulk density measurements as a function of depth below surface for each rock type is shown in Figure Figure Bulk density measurement equipment (photo taken March 2015) Source: Snowden \5047-STY-001_B Snowden Mining Industry Consultants

113 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Figure Bulk density vs depth below surface Source: SEMAFO 11.6 Qualified Person s Opinion on Adequacy of Sampling The sampling practices and assaying practices used for the trench, rock chip and soil sampling programs are, in the author s opinion, adequate for the purposes of early exploration (i.e. to define areas of anomalous gold concentration for exploration targeting). While some minor errors are likely to be present in the geochemical assay data, the author believes these are minimal and not material to the assay data for the purposes of early exploration. The RC and diamond core drilling completed by Orbis between 2010 and 2014 included independent QC samples with the sample batches, the results of which show reasonable precision and accuracy have been achieved. The drilling conducted by SEMAFO in 2015 has similarly achieved reasonable precision and analytical accuracy, based on the results of the QC samples inserted by SEMAFO into the sample batches. Additionally, the diamond core drilling, which comprises the majority of mineralized intersections at Natougou, is achieving excellent core recovery. Moreover, infill drilling by SEMAFO in 2015 has intersected mineralized widths and grades in line with previous drilling conducted by Orbis \5047-STY-001_B Snowden Mining Industry Consultants

114 NI TECHNICAL REPORT - SAMPLE PREPARATION, ANALYSES AND SECURITY Page Assaying for gold has primarily been completed at the SGS laboratory in Ouagadougou, which, based on the results of the QC samples from the Orbis and SEMAFO drilling, along with multiple inspections, has achieved reasonable precision and analytical accuracy. ALS in Ouagadougou was also used by SEMAFO in 2015 for a relatively small number of samples, with similar levels of precision and analytical accuracy achieved. Handheld XRF analysis of sample pulps was completed on a number of additional elements to assist with the geological interpretation and as an aid for rock characterisation. It is the author s opinion that handheld XRF results are often somewhat erratic and should only be used as an indicative assessment of the geochemical properties of a sample. This limitation of the handheld XRF data is recognised by SEMAFO. In the author s opinion, the drillhole data for the Natougou deposit is reasonable for use in Mineral Resource estimation \5047-STY-001_B Snowden Mining Industry Consultants

115 NI TECHNICAL REPORT - DATA VERIFICATION Page DATA VERIFICATION 12.1 RC versus Diamond Comparisons Twinning of two drillholes was undertaken in 2012 and Results indicate there is reasonable repeatability of gold grades between the twinned holes. The absolute differences in grades is noted and attributed to the inherently variable nature of gold deposits. However, Snowden notes that at this stage, there is not enough twin-hole data to enable a statistical comparison to be completed. Drilling was completed to a nominal 20 m by 20 m spaced grid over central portion of the deposit. The area was previously drilled with diamond (DD) or multi-purpose drilling (MP: RC pre-collar with diamond tail) based on a nominal 40 m by 40 m grid spacing and was in-filled in 2015 using RC drilling to 20 m by 20 m. Figure 12.1 shows a section illustrating the continuity of gold grades and mineralised intercepts between the core (DD and MP) and RC drilling. Figure 12.1 Drill section over the 20 m by 20 m grid showing RC holes, multi-purpose holes with core portion highlighted in grey and diamond holes Source: SEMAFO Snowden Mining Consultants

116 NI TECHNICAL REPORT - DATA VERIFICATION Page 12.2 To assess the impact of the drilling method on the grade distribution of the mineralized samples, Snowden generated a QQ plot which compares the RC and diamond mineralized intersections. Multipurpose intersections are assumed to be diamond within the mineralization and have been combined with the diamond data. The comparison was limited to the 20 m by 20 m area where the coverage of RC and diamond is relatively equal. Two areas of close spaced diamond data were excluded to avoid any bias due to data clustering. The comparison area is shown in Figure Figure 12.2 Plan showing comparison area with equal RC (red) and diamond (blue) coverage N Source: Snowden A QQ plot in Figure 12.3 from the data within the comparison area shows that on average the RC drilling is lower grade than the diamond core drilling. The difference appears to vary somewhat from approximately 10% for samples above 3 g/t Au to as much as 50% for samples between 0.1 g/t Au and 3 g/t Au. Snowden acknowledges that ideally for this sort of comparison, only twinned holes should be used to minimise the impact of local geological variability, however there are not enough twinned holes at Natougou for a meaningful statistical analysis of twinned data to be completed. Some of the observed difference is likely due to the local geological variability. Additionally, as the RC samples are collected based on a nominal 1 m interval, due to the relatively narrow nature of the mineralisation dilution of samples at the upper and lower boundaries of the shear zone by un-mineralized footwall and hangingwall material is likely contributing to the observed difference. Snowden Mining Consultants

117 NI TECHNICAL REPORT - DATA VERIFICATION Page 12.3 There is a risk that in areas with significant RC drilling the Mineral Resource estimate will effectively be semi-diluted and the grade of the mineralization may be slightly underestimated. Approximately 15% of samples within the mineralization are from RC drilling. Snowden recommends that SEMAFO twin some holes so that a more robust statistical assessment of the two drilling methods can be conducted. Twin diamond holes should be drilled as close as possible to the original RC drillhole, ideally within 3 m. Figure 12.3 QQ plot comparing diamond (X-axis) against RC (Y-axis) samples within the Comparison area (red line = mean; dashed blue lines are the 25 th, 50 th and 75 th percentiles) QQ Plot DD,MP-RC (log-log) Data Set 1: AU DD,MPData Set 2: AU RC M AU DD,MP {1149 values} M Source: Snowden Snowden Mining Consultants

118 NI TECHNICAL REPORT - DATA VERIFICATION Page Assay Data Validation Assay data validation has been completed through the umpire and field duplicate sampling programs, along with inspections of the primary laboratory (SGS). Additionally, a random selection of assay certificates was checked by Snowden in 2013 and again in 2015, against the data within the database, with zero discrepancies identified from this validation. The assay certificates were sourced by Snowden directly from the SGS Ouagadougou laboratory. Snowden believes the assay data within the database is robust. Snowden has not conducted any independent sampling or assaying to verify the gold tenor of the samples. Given the results of the assay certificate checks and QAQC results, Snowden does not believe that independent sampling is required at this stage Qualified Person s Opinion The Snowden site visit in March 2015 included reviews of site, drilling, interpretation, sampling and analysis. All aspects of on-site data collection and management were reviewed. The primary analysis laboratory in Ouagadougou was inspected. Snowden considers that protocols and procedures for Natougou data are completed largely in line with industry best practice and are suitable for use in Mineral Resource estimation. Snowden Mining Consultants

119 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT MINERAL PROCESSING AND METALLURGICAL TESTING 13.1 Introduction The detailed metallurgical testwork program undertaken for the Definitive Feasibility Study (DFS) has focussed on primary ore from the Natougou Project. Quantities of oxide ore presented to the process plant are expected to be low (<5% of total resource) and as such, this ore type was not included in the master composite work. However, it was tested in the variability work. The detailed testwork was carried out from March 2013 to August 2015 under the direction of Lycopodium, with input from former property owner, Orbis and later Birimian, using HQ and PQ drill core recovered from both resource and metallurgical drilling campaigns. Testwork was performed by the following laboratories: ALS Metallurgy (ALS), Perth, Australia. ALS s responsibility included sample preparation, mineralogy, comminution testwork, gravity testwork, cyanide leaching, including grind size and reagent optimization, S.G. determinations, oxygen uptake and viscosity testing, carbon loading kinetics and variability testwork. JKTech Pty Ltd (JKTech), Queensland, Australia. JKTech provided interpretation of the results from the SAG Mill Comminution (SMC) tests conducted by ALS. Consep, Sydney, Australia gravity (egrg) testwork, modelling, Knelson concentrator sizing, intensive leach testwork. Outotec, Perth, Australia thickening and filtration testwork and equipment sizing. Subsequent to this work, Birimian organized for additional thickening and filtration testwork to be conducted at FLS, Brisbane, Australia, in December This section summarizes the metallurgical testwork and draws conclusions relevant for process plant design criteria. Lycopodium Minerals Canada

120 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT Geological Background Mineralization at the Natougou deposit is hosted within a flat-lying shear zone, the Boungou Shear Zone (BSZ), which has a subtle anticlinal geometry. The apparent axis of the anticline strikes approximately 315 degrees, and each limb dips approximately 15 degrees. Figure 13.1 shows a cross section of the deposit. The mineralization is characterized by sheared amphibolite, quartz boudinage veining, biotiteplagioclase alteration and an increase in sulphide content. Sulphide assemblage comprises pyrite, pyrrhotite and minor chalcopyrite and arsenopyrite. Visible gold has been observed in core samples. The host lithology consists of mafic to intermediate volcanic / intrusive stratigraphy. Mineralization at Natougou is associated with what is locally referred to as Boungou Pink Alteration (BPK). From field observations and petrography, this alteration appears to be biotite-plagioclase schist. Within the mineralized zone, there is also a second less common alteration referred to as Boungou Bleached Zone (BBZ), which is a sericite / silica alteration. Figure 13.1 Cross Section of Natougou Deposit Lycopodium Minerals Canada

121 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT Sample Selection for Detailed Testwork Program A total of 95 diamond drill core intercepts were transported from site by Orbis to ALS for metallurgical sample preparation. The drill core consisted of half and quarter HQ intervals. Samples for metallurgical testing were selected after crushing and assaying every drill core interval provided. Two master composites were formed using a selection of these intervals. Comminution samples were sourced from dedicated whole PQ core with the addition of two half HQ core intervals from resource drilling. Variability samples were provided by Birimian at a later date in the program and consisted predominantly of quarter HQ intervals. All samples tested were primary / fresh rock samples, with the exception of five oxide variability samples Samples for Detailed Comminution Testwork Drill core photos and logs indicated that the lithology did not vary significantly throughout the deposit. Therefore it was not necessary to test a large number of intervals or drill holes. From a comminution perspective two potential comminution domains were identified within the mineralized zone. In addition, as the mineralized zone is very thin, hanging wall and footwall samples were also tested. Table 13.1 lists the major comminution domains tested. Table 13.1 Major Comminution Domains Domain Lithology Alteration Description C1 BSZ / MAM BPK C2 BSZ / MAM BBZ Hangingwal l Massive mafic volcanic (amphibolite) with biotite / plagioclase alteration (Boungou pink alteration) Massive mafic volcanic (amphibolite) with sericite / silica alteration (Boungou bleached zone) MAM None Unaltered massive mafic volcanic (amphibolite) Footwall MAM None Unaltered massive mafic volcanic (amphibolite) The C1 domain is dominant, comprising 80 to 90% of the mineralized Boungou shear zone (BSZ). Mining dilution is predicted to be high (between 25 and 30%) due to the thinness of the mineralization. Therefore, samples of hanging wall and footwall were also tested. Hanging wall and footwall are predominantly unaltered amphibolites (MAM) with minor granodiorite (GGD). Initially these domains were tested separately to determine whether there was any significant difference in hardness and competency between them. Lycopodium Minerals Canada

122 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT 13.4 As the mineralized zone at Natougou is relatively thin, four dedicated PQ holes were drilled by Orbis to provide material for the comminution testwork. These holes were twinned with existing diamond drill holes from the resource drilling campaign. Areas in the orebody with thicker than average mineralized zones were deliberately targeted, in order to provide enough material to test the mineralized zone, hanging wall and footwall separately. An additional two half HQ intervals were also provided in order to achieve a better spatial coverage of the deposit. The distribution of these drill holes across the orebody is shown in Figure Figure 13.2 Location of Comminution Samples Lycopodium Minerals Canada

123 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT Samples for Detailed Metallurgical Testwork Drill holes for metallurgical testwork were selected by Lycopodium, in conjunction with Orbis, on the basis of: Spatial location across the orebody. Selecting intervals which encompassed material in the main mineralized zone, not the subparallel hanging wall structures which can also have elevated gold assays. Obtaining samples with a range of gold head grades, logged sulphides and veining to determine the effect of head grade and gold occurrence on overall gold extraction. Total mass of diamond core was limited as the resource drilling was in the relatively early stages at the time of composite formation. Each interval received was separately crushed and assayed as an individual borehole bench composite (BHBC). Geological interpretation of the drill core indicated that the lithology of intervals did not vary significantly, and from a metallurgical perspective, only one metallurgical domain could be characterized. As such, two master composites were formed which represented different time frames in the life of mine; Master Composite 1 (MC1) which represented material in the first one to two years and approximated the Concept Study Stage 1 pit, and Master Composite 2 (MC2) which represented material in the later years of the mine life. The master composites were formed based on the head assays received from the BHBC work, the geological logs and with consideration of the target head grades for each composite. The mass of each interval included in the master composites was based on a proportional weight to intercept length. Figure 13.3Figure 13.3 shows the spatial locations of the samples selected for MC1 and MC2. MC1 samples are shown in red, while MC2 samples are shown in yellow. Lycopodium Minerals Canada

124 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT 13.6 Figure 13.3 Spatial Location of Master Composite Samples Samples for Variability Metallurgical Testwork Towards the end of the program, Birimian dispatched a total of 106 samples to ALS in Perth, in two batches. The deposit was divided into seven sections for the purpose of characterizing the proposed pit. The first batch comprised samples from Sections 1, 3, 5 and 6, while the second batch comprised samples from Sections 0, 2, and 4. These samples represented a wide cross section of spatial locations in the deposit. The samples were quarter and half HQ core. All material was primary ore, with the exception of samples from Section 0. Figure 13.4 shows the location of the samples in a plan view across the deposit. Lycopodium Minerals Canada

125 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT 13.7 Figure 13.4 Location of Variability Samples The following colour coding applies for the variability samples; Section 0 Blue, Section 1 Green, Section 2 Yellow, Section 3 Red, Section 4 Pink, Section 5 Purple, Section 6 Turquoise. Lycopodium Minerals Canada

126 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT Detailed Testwork Program Outline Based on the results of the previous concept level testwork program carried out by Orbis in 2013, a conventional cyanidation treatment route, with and without gravity, was selected for the initial evaluation of samples from the Natougou. As the ore is predominantly primary material, flotation was also identified as a potential treatment route. A detailed testwork program has been undertaken for the DFS with the following objectives: Select the most suitable processing route. Determine the optimum plant operating parameters for the ore being processed. Evaluate the variability in metallurgical performance for the primary ore source. Obtain design data required for plant design Comminution Testwork A suite of comminution testwork was conducted to determine the variability in ore hardness and competency across the orebody, and provide parameters to allow design of the comminution circuit to be carried out. Comminution tests conducted included the following: Unconfined Compressive Strength (UCS). Crushing Work Index (CWI). Bond Rod Mill Work Index (RWI). Bond Ball Mill Work Index (BWI). SAG Mill Comminution test (SMC). Abrasion Index (Ai). All tests were conducted at ALS, with the results from the SMC tests being interpreted and ranked by JKTech. Salient information from the comminution testwork program is shown in Table Included in the test results is the concept study (historic) data for comparison. The data can be summarized as follows: Lycopodium Minerals Canada

127 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page DRAFT 13.9 Unconfined compressive strength (UCS) ranged from 57 MPa to 179 MPa, with the 85th percentile value at 120 MPa. The UCS values indicate the ore has medium competency and is suitable for primary crushing using a single toggle jaw crusher. Crushing work indices (CWI) ranged from 5.2 to 11.0 kwh/t, with the 85th percentile value at 10.2 kwh/t. Most samples tested showed low resistance to impact breakage. The abrasion indices (Ai) ranged from to 0.505, with the 85th percentile value at The abrasion index values indicate that the ore is generally abrasive but variable, but do not prelude the use of a multi-stage crushing circuit. Bond rod mill work indices ranged from 17.0 kwh/t to 25.9 kwh/t, with the 85th percentile value at 20.2 kwh/t. Bond ball mill work indices ranged from 13.2 kwh/t to 19.2 kwh/t (with a 106µm closing screen), with the 85th percentile at 17.8 kwh/t. The Bond rod and ball mill work indices are relatively high indicating a high grinding energy requirement. The A x b values ranged from 31.9 to 37.8 indicating that Natougou ore is highly competent. Ores which have relatively harder a x b values than the corresponding BWi values, as in this case, can exhibit higher energy inefficiency in SAG milling, suggesting that SAG milling will not be the most energy efficient comminution option. There was low variability between the two comminution domains within the mineralized zone, and as such these were grouped together for modelling purposes. The hanging wall and footwall samples tested were marginally harder and more competent than the samples from the mineralized zone but not significantly so. The CWi values reported are inconsistent with the BWi, RWi and A*b values generated from the testwork, and may be related to the orientation of the core during testing. For design purposes, a more conservative value of 20 kwh/t has been assumed, as this is more consistent with other comminution parameters and photos viewed of the core. Several repeat BWi tests were also conducted with a finer 75µm closing screen, as a later grind optimisation trade-off study indicated that economics favoured using a finer primary grind than had been adopted for the concept study design. These tests were used to adjust the design BWi to reflect the finer grind size. Lycopodium Minerals Canada

128 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table 13.2 Comminution Testwork Summary Parameter Unit Comp 1 Comp 2 Comp 3 Comp 4 Comp 5 Comp 6 Comp 7 Comp 8 Comp 9 Comp 10 Domain Hanging Wall C1 C2 C1 Foot Wall C1 C2 C2 C1 C2 75 μm kwh/t μm kwh/t RWi kwh/t Ai g CWi UCS (Average) SMC - A SMC - b SMC - Axb SG Parameter Unit Comp 11 Comp 12 Comp 13 Comp 14 Comp 15 Comp 16 Comp 17 Comp 18 Historic 85th % 1 Domain Hanging Wall C2 C1 Foot Wall Variability Variability C1 C μm kwh/t μm kwh/t RWi kwh/t Ai g CWi UCS (Average) SMC - A SMC - b SMC - Axb SG th Percentile is based on C2 and C1 samples only, i.e. excluding foot wall and hanging wall. 2 Axb is taken as the 15th percentile to reflect top 15th percentile of competency as per the other parameters shown at 85th percentile. 3 This data is not included in the comminution modeling or 85 th percentile values presented as these tests were conducted at a much later date at the request of Birimian and Metso Minerals. Lycopodium Minerals Canada

129 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Metallurgical Testwork Borehole Bench Composites A series of comprehensive head assays was conducted on each of the individual drill hole intervals, followed by a bottle roll of selected samples which represented a range of head grades, logged sulphide grades, logged veining and spatial locations. Ninety-five individual intervals, designated borehole bench composites (BHBC), were assayed for a total of 39 elements. The following important points were noted: Gold head assay duplicates varied considerably, suggesting the presence of spotty gold. Gold head assays ranged from 0.21 g/t Au to g/t Au, with a mean of 5.90 g/t Au. Screen fire assays indicated some very high assays for the +75µm fraction, indicating coarse grained gold in the ore. Some individual intervals assayed over 500 g/t in the +75µm fraction. Silver assays ranged from <0.3 g/t Ag to 6 g/t Ag. The gold to silver ratio was variable, but averaged 6.7. At no time did silver assay higher than gold. All organic carbon assays were low at 0.06% or less. Therefore preg-robbing should not be an issue. Arsenic assays varied from 10 ppm to 9450 ppm, with a mean of 2346 ppm. As arsenopyrite had been regularly noted in the drill logs, high assays were not unexpected. Sulphide sulphur ranged from 0.18% to 3.20%, reflecting the variation in logged sulphide values seen in the drill logs. Some elevated mercury values were reported, with the highest being 1.8 ppm Hg, from a sample assaying over 40 g/t Au. The mean of all samples was 0.3ppm Hg with the median at 0.2ppm. The high mercury assays were generally in samples with very high gold assay (>10 g/t Au). While it is expected that at the gold head grades likely to be treated in the plant mercury will be at the lower end of the assays reported, it indicated that mercury assay would need to be included in the carbon loading tests. Copper was generally low, ranging from 55 ppm to 325 ppm, with a mean of 114 ppm. At these levels copper is unlikely to adversely affect either cyanide consumption or carbon loading. Other base metals, such as nickel, lead and zinc, were similarly low and should not impact cyanide consumption, carbon loading or electrowinning. Lycopodium Minerals Canada

130 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Fifteen of the above samples were selected for cyanide leach testwork using conditions similar to those used in the scoping study. Bottle roll cyanidation of 1kg of leach feed was conducted under the following conditions: Primary grind P 80 of 75µm. Pulp density of 40% solids w/w in Perth tap water. Oxygen sparged. ph 10.0 to10.5 adjusted with commercial lime (60% available CaO). Initial cyanide concentration of 0.035% w/v. 48 hour leach duration with solution samples at 24 hours. Selected leach results for these 15 samples are shown in Table Table 13.3 Bottle Roll Cyanide Leach Results for Borehole Bench Composites Sample ID Drill Hole Interval % Au Recovery % Au Recovery 48 hours Residue Grade Consumption (kg/t) 24 hours g/t Au Lime NaCN BHBC #5 BODD023 (63-68m) BHBC #10 BODD044 (67-74m) BHBC #14 BODD052 (25-27m) BHBC #19 BODD074 (72-74m) BHBC #23 BODD078 (72-77m) BHBC #29 BODD086 (61-66m) BHBC #34 BODD093 (58-64m) BHBC #39 BODD099 (47-50m) BHBC #43 BODD103 (32-34m) BHBC #46 BODD107 (20-22m) BHBC #57 BODD135 (39-44m) BHBC #74 BODD035 (56-62m) BHBC #83 BODD010 (86-91m) BHBC #87 BODD041 (69-73m) BHBC #91 BODD108 ( m) Mean Median Lycopodium Minerals Canada

131 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Gold recoveries after 24 hours ranged from 84.5% to 94.0%, with an average of 89.6%. After 48 hours these increased to 87.1% to 96.6%, with an average of 92.2%. Lime consumption was low at an average of 0.57 kg/t, with cyanide consumption being low averaging 0.17 kg/t. A three stage diagnostic leach was conducted on the samples with low recovery, to identify the deportment of gold in the residue, and assist in targeting leach conditions to improve gold recovery. The diagnostic leach method employed was: A direct cyanide leach for 24 hours using 0.2% cyanide w/v. An aqua regia digest on the residue from above to determine gold locked in acid digestible minerals such as pyrite. A total fire assay on the residue from above to determine gold encapsulated in silicate minerals. The diagnostic leach indicated that 65% to 90% of the unrecovered gold was locked in sulphides; only 4% to 11% could be recovered with higher cyanide concentrations and increased contact time. Composites which had higher cyanide recoverable gold also had a higher proportion of gold locked in silicates. The diagnostic leach suggested that improvements in gold recovery may be achievable by utilising conditions that have been known to improve leaching of primary ores, such as lead nitrate additions, and pre-aeration or pre-oxygenation prior to cyanide leaching. In addition, as there was significant coarse gold identified via the screen fire assay, inclusion of a gravity circuit may also improve overall gold recovery Metallurgical Testwork Master Composites Ore Mineralogy A sample of each master composite was prepared by processing through a Knelson concentrator and hand panning to produce a gravity concentrate and a Knelson plus hand pan combined gravity tails sample for each composite. The gravity concentrates and combined gravity tails were submitted for mineralogical analysis using QEMSCAN (Quantitative Evaluation of Minerals by Scanning Electron Microscopy), optical observations (for coarse free gold) and X-ray diffraction (XRD) analysis. To characterize the gold mineralization, optical microscopy was carried out on the unmounted and screened (>75 m) portion of the concentrate samples, while QEMSCAN analysis was carried out on a single polished section of each fraction for both concentrate and tailings samples. Salient findings from the mineralogical analysis were as follows: Lycopodium Minerals Canada

132 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Feed (Combined Concentrate and Tailings) Pyrrhotite was the dominant sulphide mineral detected at approximately 3% of the sample mass. Arsenopyrite and pyrite were the next most common sulphides at approximately 1% each of the sample. Other trace sulphides detected included up to 0.2% chalcopyrite and even lesser amounts of sphalerite, pentlandite, galena, Ni-Co-Fe sulpharsenides and silver sulphides. At the sample P 80 of approximately 50µm, almost 90% of the combined sulphides were well liberated (>90% liberated). Amphibole, quartz, feldspars and biotite were the most common non-sulphide gangue minerals. Between 0.1 to 0.3% kaolinite was detected. Concentrate Native gold is the dominant gold mineral, with only a few grains of fine (<5µm) electrum detected. Twenty-four gold grains were found in the MC1 concentrate and fifteen gold grains in the MC2 concentrate using optical microscopy. Using optical microscopy, it was determined that a high proportion of the gold occurs as relatively coarse (>100µm), free gold grains, generally platy or cylindrical. The long axis typically ranges from 150µm to 250µm, while the short axis ranges from 50µm to 150µm. Using QEMSCAN, it was determined that the > 75% majority of the total gold mass in the sample examined was present as free gold grains. The remainder was less well liberated and mainly associated with arsenopyrite and pyrite. The mineral abundance graphs for both composites are shown in Figure Lycopodium Minerals Canada

133 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure 13.5 Mineral Abundance for Natougou Master Composites The mineralogy suggested that gravity, flotation and cyanide leaching would all need to be explored as part of the metallurgical test program. It was likely that higher recoveries would be expected from MC1 compared with MC2 due to the coarser gold grain sizes and better gold liberation in MC1. It was anticipated that the ore would have a high oxygen demand due to the presence of pyrrhotite and other sulphide minerals, and that lead nitrate may be required to mitigate the passivating effect of the sulphides Head Analysis Details of the multiple gold assays, and a selection of the multi-element head assays performed on the master composites, are presented in Table The following observations and conclusions can be drawn from the head assays: The average of the two head assays for MC1 was 5.57 g/t Au, while the average of the four head assays for MC2 was 3.40 g/t. The head assays for MC2 ranged from 1.93 g/t Au to 6.98 g/t Au. These results were higher than those expected using the weighted average of the borehole bench composite assays, and reflect the spottiness of the coarse gold at Natougou. The target grades based on anticipated ROM grades were 4.5 g/t Au for MC1 and 3.0 g/t Au for MC2. Lycopodium Minerals Canada

134 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The screen fire assays indicate a head grade of 7.05 g/t for MC1, and 3.00 g/t for MC2. The +75µm fractions assayed 105 g/t Au and 30.2 g/t Au respectively. The reproducibility of the testwork may be adversely affected by the spottiness of gold in the samples. The silver grade for MC1 was 0.6 g/t Ag, while for MC2 the average of two assays was 3.9 g/t Ag. The head assay for MC2 was unexpectedly high, based on both the borehole composite assays and the expected gold to silver ratio for Natougou of between 5:1 and 10:1. While it has been postulated that spottiness in the silver mineralisation may have caused this high silver assay, the reason has never been satisfactorily identified. In light of this, MC2 data, especially with regard to silver leaching and carbon loadings, should be treated with caution. Mercury levels are low and should not present an environmental or occupational health risk in the elution or electrowinning circuits. However, as some individual boreholes had elevated assays, personal monitoring devices should be considered for gold room workers. While arsenic assays are high in both composites (>2000 ppm), as this is associated with arsenopyrite, it is not anticipated that this will leach appreciably at the cyanide concentrations likely to be used in the plant. Nor would it be expected to load onto carbon. As such, while it must be considered in tailings storage facility design, it is not expected to affect gold plant design. Concentrations of base metals such as copper, lead, zinc and nickel are low and are unlikely to adversely affect cyanide consumption or carbon performance. Organic carbon levels are low and preg robbing should not be a problem. Lycopodium Minerals Canada

135 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table 13.4 Detailed Head Analysis of Master Composites Assay Units MC1 MC2 Ag g/t Ag g/t Al % As ppm Au g/t Au g/t Au g/t Au g/t Au (screen fire assay) g/t Ba ppm Be ppm <20 <20 Bi ppm <25 <25 C % C org % 0.03 <0.03 Ca % Cd ppm <20 <20 Co ppm Cr ppm Cu ppm Cu ppm Fe % Hg ppm <0.1 <0.1 K % Li ppm <20 <20 Mg % Mn ppm Mo ppm <20 <20 Na % Nb ppm <10 <10 Ni ppm P ppm Pb ppm Pb ppm - 55 Rb ppm S % S-2 % Sb ppm SiO2 % Sr ppm Te ppm Ti ppm U ppm <2 <2 V ppm W ppm Y ppm <100 <100 Zn ppm True SG Lycopodium Minerals Canada

136 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Gravity Concentration A one kg sample of each master composite was ground to either a P 80 of 212µm or a P 80 of 75µm, and passed through a 3 laboratory Knelson concentrator. The Knelson gravity concentrate was subject to mercury amalgamation. The mercury amalgam was assayed for gold, while each gravity product was assayed for gold and sulphur. While the mass recovery to a laboratory scale gravity concentrator is always higher than achieved in plant practice, the amalgamation stage allows an estimate of possible plant performance to be made. The test results showed that gravity concentration could potentially recover over 30% of the gold. Recovery of gold was higher at 75µm (between 38 and 45%) compared with 212µm, indicating that a finer grind would benefit the gravity recovery stage. Subsequent to the initial gravity / leach testwork at ALS, a 30 kg sample of each master composite was dispatched to Consep in Sydney for egrg testing. The aims of this testwork were to conclusively identify the gravity recoverable gold (GRG) content, predict the likely gold recovery in the gravity circuit and enable the most appropriate gravity circuit to be sized and specified. The egrg test involves three stages of grinding and Knelson concentration. Each time, the tailings from the previous stage is ground progressively finer before passing through the Knelson concentrator again. Size-by-size assays are conducted for the concentrates from each stage, along with the final tailings. Testing of Natougou ore identified that the GRG in the MC1 sample is moderately fine, with just over 30% being finer than 75µm and 17% being finer than 38µm. The GRG in the MC2 sample is finer than MC1, with over 40% being finer than 75µm and 22% being finer than 38µm. The egrg testwork showed that 65.6% of the gold in MC1 is potentially recoverable by gravity, along with 56.3% of the silver. For MC2, 61.9% of the gold and 30.0% of the silver is potentially recoverable by gravity. Subsequent to this testwork, a second sample of MC1 only was sent to Consep specifically to examine the deportment of copper, arsenic and mercury in the gravity and intensive leach stages, in addition to silver and gold. The intensive leach results from this work, along elemental deportment in the gravity stage, are reported in Section For this second MC1 sample, the GRG was determined to be 67.1% compared with 65.6% in the previous test. Consep reported that with the size distribution of the gold in this MC1 sample, it would be reasonable to expect that a well designed Knelson concentrator based gravity circuit could recover 75 to 85% of the GRG. Therefore, on this basis, a suitable plant design value for gravity circuit gold recovery would be approximately 50%. Lycopodium Minerals Canada

137 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Cyanide Leach Grind / Recovery Testwork Cyanide leach grind / recovery tests were performed on each master composite at P 80 values of 150, 106, 75 and 63 microns to determine the effect of grind size on gold recovery. These tests were carried out with and without gravity concentration to examine the impact of the gravity circuit on gold recovery at various grind sizes. Due the presence of oxygen consuming minerals in the ore, all tests were performed with oxygen rather than air. The cyanidation testwork was conducted in 3kg agitated vats under the following conditions: Gravity concentration (if applied) via Knelson concentrator, amalgamation of gravity concentrate, and leaching of gravity plus amalgam tailings. Leach pulp density of 40% solids w/w in Perth tap water. ph 10.0 to10.5 adjusted with commercial lime (60% available CaO). Initial cyanide concentration of 0.035% w/v, with residual cyanide levels maintained at or above 0.025% w/v. Dissolved oxygen levels of 20 ppm or greater. 48 hour leach duration. Lycopodium Minerals Canada

138 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The grind / recovery leach results are summarized in Table Table 13.5 Summary of Grind Recovery Leach Results Test No. Comp Grind Size P 80 (µm) Calc Head Grade g/t Au % Au Recovery % Au Recovery % Au Recovery Residue Grade % Ag Recovery Gravity 36 hours 48 hours g/t Au 48 hours Consumption (kg/t) Lime NaCN JS3362 MC N/A JS3363 MC N/A JS3364 MC N/A JS3365 MC N/A JS3366 MC N/A JS3367 MC N/A JS3368 MC N/A JS3369 MC N/A JS3428 MC JS3465 MC JS3466 MC JS3467 MC JS3468 MC JS3429 MC JS3469 MC JS3470 MC JS3471 MC JS3472 MC For tests without gravity, the results indicate extreme spottiness in residues, with no clear trend of lower residue grades and higher gold recoveries with a finer grind, as would normally be expected. Gold recoveries were much lower than seen in any of the bottle roll tests on the borehole bench composites. The leach kinetics for both gold and silver were extremely slow. As expected, cyanide and lime consumption was higher than seen in the bottle roll tests. For tests with gravity ahead of cyanide leach, much higher gold recoveries and lower residue grades were seen for all but one of the grind sizes compared with the leach only tests. Initially only two tests JS3428 and JS3429, were performed with gravity ahead of cyanide leach, using a grind size P 80 of 75µm, to confirm that recoveries of over 90% could be achieved for each composite. Several weeks later the entire range of grind sizes was tested to investigate grind sensitivity, including a repeat at 75µm. Lycopodium Minerals Canada

139 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page For MC1, the group of gravity leach tests conducted in parallel, JS3465 to JS3468, showed clear grind dependency, with the gold recovery increasing from 88.8% at 150µm to 94.6% at 63µm after 48 hours. Similarly, residue grades decreased from 0.71 g/t Au at 150µm to 0.33 g/t Au at 63µm after 48 hours. Silver recovery was variable, from 64% to 77%, but within the accuracy of the assays. Test JS3428, the initial test on MC1 at 75µm, achieved over 90% gold recovery. However, the residue was still quite spotty, with the three residues analysing 0.86, 0.52 and 0.49 g/t Au. This test had the highest calculated head grade of all the MC1 tests. This result serves to highlight the difficulty in selecting optimum conditions for the Natougou ore given the spottiness of the sample. This test has not been used in the grind optimisation analysis, discussed later in Section For MC2, the group of gravity leach tests conducted in parallel, JS3469 to JS3472, showed clear grind dependency, with the gold recovery increasing from 84.8% at 150µm to 90.2% at 63µm after 48 hours. Similarly, residue grades decreased from 0.44 g/t Au at 150µm to 0.26 g/t Au at 63µm after 48 hours. Silver recovery was variable, from 43.3% to 59.8%, but within the accuracy of the assays. Similarly Test JS3429, the initial test on MC2 at 75µm, achieved over 90% gold recovery. The residue was not spotty like in MC1, but nevertheless as the calculated head grade was the highest of all the MC2 tests and it did not fit in with the later grind series tests, it has not been used in the grind optimization analysis. Figure 13.6 and Figure 13.7 show the kinetic leach curves for both MC1 and MC2. For clarity, the outlying tests JS3428 and JS3429 have been omitted. Lycopodium Minerals Canada

140 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure 13.6 MC1 Grind Sensitivity Cyanide Leach Curves Figure 13.7 MC2 Grind Sensitivity Cyanide Leach Curves Lycopodium Minerals Canada

141 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The kinetic curves indicate: Higher overall gold recoveries with increasing fineness of grind. Improved leaching rates with increasing fineness of grind. Leaching is effectively completed in 36 hours for the finest grind sizes (P 80 of 63 and 75µm). Master Composite 2 has slower leach kinetics than Master Composite 1. In light of the above results, all further leach tests were preceded with a gravity concentration step, and gravity concentration considered essential for Natougou ore, when using a whole of ore leach flowsheet. A grind optimization study was undertaken to evaluate the effect of grind size on project economics. The evaluation compared gold revenue against operating and capital expenditure for the grind sizes considered and is detailed in Section Flotation Flotation was conducted using a rougher only flowsheet and simple baseline reagent regime, with the aim of comparing with the gravity plus whole of ore leach as presented in Section It was envisaged that the flotation concentrate could be reground, if required, and subject to a cyanide leach suitably tailored to a sulphide concentrate. Sighter flotation tests were conducted on each master composite. Five rougher concentrates were recovered over a period of fifteen minutes, with an upfront dose of copper sulphate and stage addition of potassium amyl xanthate (PAX) and frother (IF56). A natural ph was maintained throughout the tests. All sighter tests were performed at a P 80 of 75µm. Various reagents addition rates were tested. Excellent recovery of sulphides and gold was achieved with a very simple reagent scheme. The froth was stable and well mineralized. For all tests, over 99% of the sulphur was recovered, greater than 96% of the arsenic and between 87% and 93% of the gold. Very low mass pull to concentrate was achieved ranging from 5.6% to 6.1% of the feed mass. All further tests were carried out using 30 g/t of copper sulphate and 30 g/t of PAX. Following the sighter test results, investigation of gravity concentration to recover free gold ahead of flotation, was carried out. Table 13.6 presents these results along with the comparative sighter tests without gravity. All tests were carried out at a P 80 of 75µm. Lycopodium Minerals Canada

142 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table 13.6 Summary of Rougher Flotation Tests With and Without Gravity Test No. Comp Gravity Rec % Mass Pull to Grav + Flt Con% Gravity + Rougher Concentrates 1 to 5 Au Grade g/t Tails Au Rec % S Rec % As Rec % g/t Au JS3403 MC JS3425 MC JS3426 MC JS3427 MC Despite high recovery in the gravity stage, gold grades and recovery were similar to the previous tests, indicating that flotation was effectively recovering the gravity recoverable gold component. For MC1, the residue grade was actually lower in the test without gravity. A further series of tests was carried out to examine the grind sensitivity of the flotation performance. Due to the spottiness in residue assays seen in the earlier cyanide leach grind recovery testwork, it was elected to include a gravity recovery step prior to flotation. Primary grind sizes varied from a P 80 of 150µm to 63µm, with each feed charge subject to gravity separation prior to 15 minutes of rougher flotation with 30 g/t copper sulphate, 30 g/t PAX and frother as required. Results from the previous stage of testwork are included for comparison. The test results are shown in Figure 13.8 and Figure Figure 13.8 MC1 Grind Sensitivity Grade Recovery Curves Lycopodium Minerals Canada

143 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure 13.9 MC2 Grind Sensitivity Grade Recovery Curves The data shows that for MC1, there is little if any improvement in gold flotation recovery when decreasing the grind size from 106µm to 63µm. The tailings grade varied from 0.50 to 0.52 g/t Au which is within the likely reproducibility of the assay given the spottiness of the sample. However, Figure 13.8 shows that the gold grade of the concentrate increased and gave an improved grade recovery curve with a finer grind. The 63µm sample and one of the 75µm samples lay on the same grade-recovery curve which indicates that the reproducibility of these tests was not particularly good. For MC2, the gold recovery and tailings grade improved when grinding finer, down to 75µm. The 63µm test had an unusually low gravity recovery figure, and calculated head grade, which has impacted the overall gold recovery. However, the lowest tailings assay was recorded for this test. Higher concentrate grades were achieved when grinding finer. The two samples with a P 80 of 75µm lay on the same grade-recovery curve which indicates good reproducibility of these two tests. These tests also had the best grade-recovery curves. In light of the above results, data at a primary grind size P 80 of 75µm was used to compare the flotation and whole of ore leach routes. It was recognised that further improvements in gold recovery or flotation kinetics might be possible by the addition of more expensive gold specific specialty collectors in conjunction with PAX. However, it was also known that further improvements in cyanide leach gold recovery and/or reagent optimisation were also likely for the whole of ore leach flowsheet. Lycopodium Minerals Canada

144 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page In order to establish the preferred treatment route for Natougou primary ore, a trade-off study was conducted comparing whole of ore (WOO) leach and flotation, along with a two different tailings disposal options. Both options were sized to treat 2.0 Mtpa of ore per annum. The WOO flowsheet included gravity concentration, pre-leach thickening, CIL circuit and acid wash and elution of loaded carbon using AARL method. The flotation flowsheet included rougher flotation, rougher concentrate regrind, pre-leach (concentrate) thickening, cyanide leach of reground concentrate using CIL circuit, and acid wash and elution of loaded carbon using AARL method. The metallurgical testwork presented in Sections and was used as the basis for the trade-off analysis. It was assumed based on another project that the gold extraction in the flotation concentrate leach would be 95%, and the silver extraction would be 80%. The economic evaluation was based on the following inputs; mine life of 6.2 years, LOM head grade of 3.51 g/t Au and 0.60 g/t Ag, gold price of US$1,300/oz and silver price of US$20/oz, heavy fuel oil price of US$0.19/kWh. The average of MC1 and MC2 gold and silver extractions at a primary grind size P 80 of 75µm (91.1% gold and 70.0% silver for WOO leach; 85.6% gold and 67.8% silver using flotation concentrate leach) were used. A summary of the financial modelling is shown in Table Table 13.7 Flotation Vs WOO Leach Financial Analysis Parameter Unit Whole of Ore Flotation / Leach CAPEX US$M OPEX US$M NPV US$M IRR % Payback Years The whole of ore leach option gave a higher NPV and IRR than the flotation leach option. In addition, a whole of ore leach is likely to be a better process selection for the very small tonnage of oxide material which will also need to be treated through the plant. On this basis, the whole of ore leach was chosen for the remainder of the metallurgical testwork program Grind Optimization Study Following the selection of gravity concentration followed by cyanide leach as the preferred flowsheet, a grind optimization analysis was undertaken to determine the most economic grind size for the primary ore. Lycopodium Minerals Canada

145 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The comminution testwork presented in Section 13.4 was used as the basis for calculation of grinding energy and consumable costs. Cyanide leach test results in Section were used as the basis for the revenue calculation and the reagent consumption rates. The following inputs were provided by Orbis for the financial analysis; 2.0 M tpa plant throughput at 91.3% availability, gold prices of 1,300 USD/oz (base case), 1,100 USD/oz and 900 USD/oz. Table 13.8Table 13.8 shows the financial analysis for each of the grind sizes considered for both master composites. Table 13.8 Grind Optimization Financial Analysis Parameter Unit Master Composite 1 Master Composite 2 Grind Size µm Base Case Gold Price $/oz 1,300 1,300 1,300 1,300 1,300 1,300 1,300 1,300 Residue Grade g/t Au Revenue Increment $/t Operating Cost Increment $/t Net Revenue Increment 1,300 $,000/a 0 10,820 17,160 18, ,450 6,680 7,030 Net Revenue Increment 1,100 $,000/a 0 8,560 13,240 14, ,270 4,560 4,510 Net Revenue Increment 900 $,000/a 0 6,310 9,320 9, ,090 2,440 1,990 The grind size optimization work indicates that for MC1, a primary grind P 80 of 63µm offers the greatest revenue benefit, with continued revenue improvements even at a lower gold price of US$900/oz. For MC2, a primary grind size P 80 of 63µm is recommended at the base case gold price of US$1,300/oz. Should the prevailing gold price at the start of MC2 processing be less than US$1,300/oz, a 75µm grind size would yield the greatest revenue benefit. On this basis, a primary grind P 80 of 63µm was selected for the remainder of the testwork for both composites Oxygen and Cyanide Investigations A matrix of leach tests was conducted on both master composites, in an attempt to optimise leach conditions in terms of both gold recovery and reagent consumption. The cyanidation testwork was conducted using 3kg vat leaches using the following standard conditions: Primary grind P 80 of 63 microns. Lycopodium Minerals Canada

146 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Gravity concentration by Knelson concentrator, amalgamation and leaching of gravity plus amalgam tailings Leach pulp density of 40% solids w/w using Perth tap water. ph 10.0 to10.5 adjusted with commercial lime (60% available CaO). 48 hour leach duration. Oxygen and air addition was trialled, along with the effects of pre-oxygenation or pre-aeration for varying times, and two different cyanide concentrations. The matrix of varying test conditions is presented in Table Table 13.9 Oxygen and Cyanide Investigation Test Matrix Test No. Comp Pre-ox Time h Sparge Gas NaCN initial conc %w/v NaCN maint conc %w/v CT1031 / CT1039 MC1 / MC2 0 O >0.025 CT1032 / CT1040 MC1 / MC2 0 O >0.040 CT1033 / CT1041 MC1 / MC2 4 O >0.025 CT1034 / CT1042 MC1 / MC2 4 O >0.040 CT1035 / CT1043 MC1 / MC2 8 Air >0.025 CT1036 / CT1044 MC1 / MC2 8 Air >0.040 CT1037 / CT1045 MC1 / MC2 8 O >0.025 CT1038 / CT1046 MC1 / MC2 8 O >0.040 For the oxygen tests, the DO level was maintained at greater than 20 ppm throughout the pre-oxidation stage (if applicable) and the cyanide leach. For the air tests, the sample was sparged with air to maintain the DO level around 7 to 9 ppm in the pre-aeration stage and the cyanide leach. The leach results are illustrated graphically in Figure to Figure 13.13Figure Two tests JS3468 and JS3472, using the same conditions as CT1031 and CT1039 are also presented. All residue assays for gold were completed in triplicate. Lycopodium Minerals Canada

147 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure MC1 Oxygen Investigation Tests with Cyanide Concentration 0.035% w/v EXTRACTION (%) JS3468, O2, No pre-ox CT1031, O2, No pre-ox CT1033, O2, 4h pre-ox CT1035, Air, 8h pre-aeration CT1037, O2, 8h pre-ox TIME (Hours) Figure MC1 Oxygen Investigation Tests with Cyanide Concentration 0.05% w/v EXTRACTION (%) CT1032, O2, No pre-ox CT1034, O2, 4h pre-ox CT1036, Air, 8h pre-aeration CT1038, O2, 8h pre-ox TIME (Hours) Lycopodium Minerals Canada

148 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure MC2 Oxygen Investigation Tests with Cyanide Concentration 0.035% w/v EXTRACTION (%) JS3472, O2, No pre-ox CT1039, O2, No pre-ox CT1041, O2, 4h pre-ox CT1043, Air, 8h pre-aeration CT1045, O2, 8h pre-ox TIME (Hours) Figure MC2 Oxygen Investigation Tests with Cyanide Concentration 0.05% w/v EXTRACTION (%) CT1040, O2, No pre-ox CT1042, O2, 4h pre-ox CT1043, Air, 8h pre-aeration CT1045, O2, 8h pre-ox TIME (Hours) Lycopodium Minerals Canada

149 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The results indicate: Spottiness in residue assays and varying gravity recovery makes selection of optimum conditions difficult. Repeatability of tests is poor. For example, tests JS3468 and CT1031 used the same conditions but had residue grades of 0.33 and 0.42 g/t respectively. The lowest residue grades and highest gold recoveries were achieved in tests that used oxygen rather than air, for both composites. The addition of oxygen improved leach kinetics significantly compared to air during the first 24 hours of leaching. No consistent benefit was seen from using pre-oxygenation ahead of the cyanide leach in terms of residue grade, gold recovery or cyanide consumption. Kinetic curves with and without pre-oxygenation were generally very similar. Higher lime consumption was seen in tests which used pre-aeration or pre-oxygenation prior to leaching, indicating that sulphides were being converted to sulphates. However, this did not translate into lower residue grades. High cyanide concentrations resulted in higher cyanide consumption but did not achieve lower residue grades for the majority of conditions tested. As no clear benefit could be seen in including a pre-oxygenation stage, or using higher cyanide concentrations, all further tests were conducted with oxygen but without pre-oxygenation and using a starting concentration of 0.035% w/v cyanide, maintained at greater than 0.025% w/v throughout the leach Lead Nitrate Additions A series of tests were conducted to investigate whether lead nitrate additions could improve leach kinetics and/or overall gold recovery. Lead nitrate doses of 50 g/t, 100 g/t and 250 g/t were tested. The cyanidation testwork was conducted using 3kg vat leaches using the following conditions: Primary grind P 80 of 63 microns. Gravity concentration by Knelson concentrator, amalgamation and leaching of gravity plus amalgam tailings. Leach pulp density of 40% solids w/w using Perth tap water. Oxygen sparging to maintain a DO level of >20ppm. Lycopodium Minerals Canada

150 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page ph 10.0 to10.5 adjusted with commercial lime (60% available CaO). Initial cyanide concentration of 0.035% w/v, maintained at 0.025% w/v. A selection of the leach results are illustrated graphically in Figure and Figure Figure MC1 Lead Nitrate Tests EXTRACTION (%) JS3468, O2 sparged, No Pb(NO3)2 CT1048, O2 sparged, 50 g/t Pb(NO3)2 CT1050, O2 sparged, 100 g/t Pb(NO3)2 CT1052, O2 sparged, 250 g/t Pb(NO3) TIME (Hours) Lycopodium Minerals Canada

151 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure MC2 Lead Nitrate Tests EXTRACTION (%) CT1039, O2 sparged, No Pb(NO3)2 CT1055, O2 sparged, 50 g/t Pb(NO3)2 CT1057, O2 sparged, 100 g/t Pb(NO3)2 CT1059, O2 sparged, 250 g/t Pb(NO3) TIME (Hours) The results indicate: Improved kinetics for the first 12 hours of leaching for both composites when using lead nitrate. Lead nitrate addition rates of between 50 g/t and 100 g/t appear to be sufficient. The leach curves are very flat between 24 and 48 hours, indicating that the optimum residence time is likely to be less than 48 hours. Following these tests, and as a result of the improved kinetics exhibited when using lead nitrate, a further investigation of air versus oxygen was conducted, as it was possible that air and lead nitrate could achieve similar results to that of oxygen over 48 hours. The cyanidation testwork was conducted using 3kg vat leaches using the following conditions: Primary grind P 80 of 63 microns. Gravity concentration by Knelson concentrator, amalgamation and leaching of gravity plus amalgam tailings. Leach pulp density of 40% solids w/w using Perth tap water. Lycopodium Minerals Canada

152 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Pre-aeration for four hours and air sparging to maintain a DO level of between 7 to 9ppm. 100 g/t of lead nitrate for MC1 and 50 g/t of lead nitrate for MC2. Initial cyanide concentration of 0.035% w/v, maintained at 0.025% w/v. ph 10.0 to10.5 adjusted with commercial lime (60% available CaO). Figure and Figure show a summary of the test results, along with the comparable tests with oxygen. Figure MC1 Oxygen Vs Air with Lead Nitrate EXTRACTION (%) CT1050, O2 CT1051, O2 CT1066, Pre-aeration + Air CT1067, Pre-aeration + Air TIME (Hours) Lycopodium Minerals Canada

153 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure MC2 Oxygen Vs Air with Lead Nitrate EXTRACTION (%) CT1054, O2 CT1055, O2 CT1068, Pre-aeration + Air CT1084, Pre-aeration + Air TIME (Hours) The results indicate: Much better reproducibility for this set of tests. Slower kinetics for the tests using air, along with higher residue grades. This was more marked for MC2 compared with MC1. Higher lime consumption for the tests using air, but lower cyanide consumption rates. Over 94% gold recovery could be achieved for MC1 and over 91% for MC2 in the oxygen tests. When using oxygen, residue grades of between 0.24 g/t gold and 0.31 g/t gold could be achieved from both composites. These tests show that even when using lead nitrate, oxygen is still preferred over air Leach Pulp Density The effect of leach pulp density was evaluated for each of the master composites. Parallel cyanidation tests were conducted at leach densities of 45% w/w solids and 50% w/w solids. The testwork from the previous lead nitrate series at 40% solids w/w is also included for comparison. Lycopodium Minerals Canada

154 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The cyanidation testwork was conducted using 3kg vat leaches using the following conditions: Primary grind P 80 of 63 microns. Gravity concentration by Knelson concentrator, amalgamation and leaching of gravity plus amalgam tailings. Sparging with oxygen to achieve dissolved oxygen levels of 20 ppm or greater. 100 g/t of lead nitrate for MC1 and 50 g/t of lead nitrate for MC2. Initial cyanide concentration of 0.035% w/v, maintained at 0.025% w/v. ph 10.0 to10.5 adjusted with commercial lime (60% available CaO). 48 hour leach duration. The leach results are illustrated graphically in Figure and Figure Figure MC1 Leach Pulp Density Series EXTRACTION (%) CT1050, 40% solids CT1051, 40% solids CT1069, 45% solids CT1071, 50% solids TIME (Hours) Lycopodium Minerals Canada

155 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure MC2 Leach Pulp Density Series EXTRACTION (%) CT1054, 40% solids CT1055, 40% solids CT1070, 45% solids CT1072, 50% solids TIME (Hours) The results indicate: Overall gold recovery and residue grade is not adversely affected by the leach density. Slightly slower kinetics were observed for the first 24 hours of the 50% w/w solids test for MC2. Cyanide usage is reduced by more than 20% when operating at 50% w/w solids compared to 40% w/w solids pulp density. Lime usage is also reduced by more than 20%, when operating at 50% w/w solids compared to 40% w/w solids pulp density. Viscosity testing (see Section ) at 50% w/w solids indicates that the viscosity of both master composites at 50% w/w solids in the adsorption circuit is not such that inter-tank screen throughput will be an issue. A pulp density of 50% w/w solids was selected for the remaining detailed testwork and variability testwork. Lycopodium Minerals Canada

156 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Direct Leach vs Gravity Concentration To evaluate the effect of not incorporating a gravity concentration stage prior to leaching using the optimum conditions, further additional tests were undertaken. The following conditions were used: Primary grind P 80 of 63 microns. Leach pulp density of 50% solids w/w in Perth tap water. Sparging with oxygen to achieve dissolved oxygen levels of 20 ppm or greater. 100 g/t of lead nitrate for MC1 and 50 g/t of lead nitrate for MC2. Initial cyanide concentration of 0.035% w/v, maintained at 0.025% w/v. ph 10.0 to10.5 adjusted with commercial lime (60% available CaO). 48 hour leach duration. Results of testwork are shown graphically in Figure and Figure Figure MC1 Tests with and without Gravity at Optimum Conditions EXTRACTION (%) CT1071, Gravity CT1080, No Gravity CT1082, No Gravity TIME (Hours) Lycopodium Minerals Canada

157 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure MC2 Tests with and without Gravity at Optimum Conditions EXTRACTION (%) CT1072, Gravity CT1081, No Gravity CT1083, No Gravity TIME (Hours) Results of testwork indicated: For MC1, after 48 hours of leaching, similar residue grades and gold recoveries were achieved for the tests with and without gravity. However, at 36 hours or less, gold recoveries were lower in tests that did not include gravity concentration For MC2, for the tests without gravity, the leach curve reached a plateau at just above 80% gold recovery, regardless of the leach time employed. This is in contrast to the test with gravity which achieved over 92% gold recovery after 36 hours. Cyanide consumption was lower in the tests that include gravity concentration. The testwork shows that gravity is an essential component of the Natougou flowsheet, and furthermore, consideration should be given to installing two centrifugal concentrators. This would assist in maintaining overall recovery should one unit be off line for maintenance Oxygen Optimization Testwork Further tests were carried out to examine the impact of higher leach temperatures and lower DO levels, which might conceivably be encountered in a plant situation. Lycopodium Minerals Canada

158 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page This led to a modification of the standard ALS leach procedure, whereby a water bath was used to heat the leach vessel to achieve a slurry temperature of 35 C, and oxygen was sparged at a much lower dosage rate than usual. Due to the elevated temperature, a second oxygen sparge had to be added partway through test CT1191 on MC1 to achieve above 15ppm DO. A test was also conducted on each composite using air and elevated slurry temperature. The following conditions were used: Primary grind P 80 of 63 microns. Gravity concentration by Knelson concentrator, amalgamation and leaching of gravity plus amalgam tailings. Leach pulp density of 50% solids w/w in Perth tap water. Slurry heated to a temperature of 35 C and maintained throughout the leach. 100 g/t of lead nitrate for MC1 and 50 g/t of lead nitrate for MC2. Initial cyanide concentration of 0.035% w/v, maintained at 0.025% w/v. ph 10.0 to10.5 adjusted with commercial lime (60% available CaO). 48 hour leach duration. Results of testwork are shown graphically in Figure and Figure temperature are also shown for reference. Tests at ambient Test CT1190 averaged 15ppm DO over the first 24 hours, while Test CT1191 averaged 13ppm DO (prior to the second sparge being added). Test CT1189 had a very high calculated head grade compared to previous samples, so the results from this test should be treated with some caution. Lycopodium Minerals Canada

159 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure MC1 Tests with Varying DO Levels EXTRACTION (%) CT Air CT ppm DO CT ppm DO CT ppm DO TIME (Hours) Figure MC2 Tests with Varying DO Levels EXTRACTION (%) CT Air CT ppm DO CT ppm DO CT ppm DO TIME (Hours) Lycopodium Minerals Canada

160 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Although residue grades for the air tests were comparable to the tests with oxygen after 48 hours, the kinetics of the leach were slower for both composites. The test results show that the leach kinetics are very sensitive to DO levels, and particularly DO levels early in the leach, which is not unexpected for a reactive sulphide ore. In light of this, the Natougou plant will incorporate external oxygen contactors on the first two leach tanks, and spargers rather than down the shaft oxygen on the following leach tanks to improve the oxygen mass transfer rates. The effect of temperature on the cyanide leach was essentially to lower the DO level, which resulted in slower leach kinetics as discussed above. Tests with higher DO levels generally had lower lime and cyanide consumptions than those test with lower DO levels, which further supports the use of oxygen and injection methods which maximize the oxygen mass transfer rates Residence Time Selection A residence time trade-off investigation was completed to determine the most economic cyanide leach time. Leach times of 24, 36 and 48 hours were considered. Optimum conditions for gold recovery were determined to be: A primary grind P 80 of 63 µm. Gravity recovery ahead of cyanide leach. Cyanide concentration of 0.035% v/v. Oxygen sparged. Lead nitrate additions. A number of tests were used to assess the effect of residence time on gold recovery. The tests chosen were conducted at or near to the optimum conditions, and encompassed the following: A primary grind P 80 of 63 µm. Gravity recovery ahead of cyanide leach. Cyanide concentration of 0.035% v/v. Oxygen sparged. Lead nitrate additions between 50 and 250 g/t. Lycopodium Minerals Canada

161 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Pulp densities ranging from 40 to 50% solids w/w. The following inputs were provided by Birimian for the financial analysis; 1.34 M tpa plant throughput at 92% availability (4000 tpd), 4.36 g/t Au head grade, 1,100 USD/oz gold price, 8 year mine life, 7% discount rate. A head grade versus residue grade relationship, using the average MC1 and MC2 test results, was used to predict the likely residue grade at the design head grade for each of the three residence times. Design gold recoveries were then calculated from the head and residue grades. Values used in the analysis are presented in Table Table Residence Time Selection Gold Recoveries at Various Head Grades Comp Calc Head Grade % Au Recovery % Au Recovery % Au Recovery g/t Au 24 hours 36 hours 48 hours MC MC Design Table shows the financial analysis summary, including capital costs, revenue, NPV and IRR. Incremental CAPEX, incremental revenue, NPV and IRR are calculated relative to the 24 hour leach. Table Residence Time Selection Financial Analysis Leach Time Installed CAPEX Incremental CAPEX Revenue Incremental Revenue NPV h M USD M USD M USD/y M USD/y M USD % IRR A nominal leach time of 36 hours was selected as the basis of further testwork and plant design on the following basis: The 36 hour leach has the best NPV and IRR on an installed capital versus revenue basis. The operating costs for the 36 hour leach are also marginally less than the 48 hour leach due to smaller tank sizes. Lycopodium Minerals Canada

162 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Bulk Leach A bulk leach (14 kg) at the target grind and leach conditions was conducted on each master composite to confirm likely leach extractions, reagent consumptions and provide leached slurry for the carbon adsorption testwork. The cyanidation testwork was conducted under the following conditions: Primary grind P 80 of 63 microns. Gravity concentration by Knelson concentrator, amalgamation and leaching of gravity plus amalgam tailings. Leach pulp density of 50% solids w/w in Perth tap water. ph adjusted with commercial lime (60% available CaO). Initial cyanide concentration of 0.035% w/v with residual cyanide levels maintained at 0.025% w/v. Target dissolved oxygen levels of 20ppm or greater, although only 18ppm was achieved in practice for MC1 and 13ppm for MC2. Lead nitrate additions of 100 g/t for MC1 and 50 g/t for MC2 36 hour leach duration. The results of the bulk leach test, along with comparable 3kg tests using the same reagent additions, but varying DO levels, are summarized in Table and illustrated graphically in Figure and Figure Lycopodium Minerals Canada

163 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Summary of Bulk Leach Test Results and Optimized Test Results Test No. Comp Sample Size Temp Average DO Level Calc Head Grade % Au Recovery % Au Recovery % Ag Recovery Consumption * (kg/t) kg C ppm g/t Au Gravity 36 hours 36 hours Lime NaCN CT1190 MC CT1191 MC CT1071 MC1 3 Ambient CT1208 MC1 14 Ambient CT1193 MC CT1194 MC CT1072 MC2 3 Ambient CT1209 MC2 14 Ambient * Consumption rates for 3kg tests based on 48 hour leach. The residue grade from the bulk leaches was 0.29 g/t Au for MC1 (based on average of four assays) and 0.25 g/t Au for MC2 (based on average of three assays). Figure MC1 Bulk Leach and Optimized Leach Test Results EXTRACTION (%) CT ppm DO CT ppm DO CT ppm DO CT Bulk Leach, 17ppm DO TIME (Hours) Lycopodium Minerals Canada

164 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure MC2 Bulk Leach and Optimized Leach Test Results EXTRACTION (%) CT ppm DO CT ppm DO CT ppm DO CT Bulk Leach, 13ppm DO TIME (Hours) The results indicate: The calculated head grade for the two bulk leaches is lower than the comparable 3kg tests. CT1208 in particular is very low. Repeat residue and solution assays were conducted, however they were almost identical. The gravity recovery for MC1 is low for the bulk leach. Repeat solution assays on the amalgam concentrate were very similar. The overall gold recoveries for the bulk leaches are within the bounds of the 3kg leaches when considering the average DO levels achieved. This is despite the low gravity recovery for the MC1 bulk leach. This indicates that 36 hours is an appropriate leach time. Lime and cyanide consumption rates for the bulk tests are similar to the 3kg tests conducted at ambient temperature. Lime consumption ranged from 0.24 to 0.27 kg/t, while cyanide consumption ranged from 0.58 kg/t to 0.98 kg/t. Cyanide consumption was moderate to high, especially for MC2. The leach solution from each bulk leach test was assayed for common base metals and deleterious elements such as mercury and arsenic. The results of these assays are presented in Table Lycopodium Minerals Canada

165 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Detailed Solution Analysis from Bulk Leaches Assay Units MC1 MC2 Ag (mg/l) Al (mg/l) As (mg/l) Au (mg/l) Ba (mg/l) < Bi (mg/l) <0.10 <0.10 Ca (mg/l) Cd (mg/l) <0.05 <0.05 Co (mg/l) Cr (mg/l) <0.10 <0.10 Cu (mg/l) Fe (mg/l) Hg (mg/l) < K (mg/l) Li (mg/l) <0.05 <0.05 Mg (mg/l) Mn (mg/l) <0.05 <0.05 Mo (mg/l) Na (mg/l) Ni (mg/l) P (mg/l) <1.0 <1.0 Pb (mg/l) <0.05 <0.05 Sr (mg/l) Ti (mg/l) <0.10 <0.10 V (mg/l) Y (mg/l) <0.01 <0.01 Zn (mg/l) Zr (mg/l) <0.05 <0.05 The head assay for MC2 was higher in silver, copper and other base metals than anticipated from the borehole bench composite assays, and in many respects is at the extreme of the borehole bench composites. As such, the silver and copper in solution for this composite should not be taken as being representative of a typical leach solution. Nevertheless, the solution assays for MC1 indicate that moderate to high copper concentrations, high arsenic concentrations and low to moderate nickel and zinc concentrations can be expected. Copper, nickel and zinc do have the ability to load onto carbon, and therefore compete with gold and silver for available sites. Lycopodium Minerals Canada

166 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Arsenic does not typically load onto carbon, so it is therefore an environmental issue rather than a metallurgical one. The tailings consultant, Knight Piésold, has indicated that as the arsenic content in the solids is high, the tailings storage facility will have to be lined regardless of whether the arsenic is in solution or as solids. Low mercury and lead assays were seen in the leach solution, which is advantageous for carbon loading, elution and electrowinning, and also workplace health and safety. The bulk leach test solution was also submitted for cyanide speciation. The results of this analysis are presented in Table Table Cyanide Speciation from Bulk Leaches Assay Units MC1 MC2 CN Free (mg/l) CN WAD (mg/l) CN Total (mg/l) CNO (mg/l) SCN (mg/l) Weak acid dissociable (WAD) cyanide includes the toxicologically important forms of cyanide, including free cyanide and weakly and moderately complexed metal cyanides, such as the zinc and nickel cyanide complexes and some 70% of the copper cyanide complexes. Total cyanide refers to the total cyanide ion, molecular hydrogen cyanide and most cyanide bound in metallic complexes and compounds that can be decomposed with a strong acid. It includes the relatively non-toxic iron cyanide complexes. It does not include the thiocyanate (SCN - ) or cyanate (CNO - ) ions. Total cyanide includes WAD cyanide as a sub-set and therefore includes free cyanide. The WAD, free and total cyanide readings are within the range of those seen in other testwork programs. The cyanide speciation shows that a significant portion of the cyanide is complexed as thiocyanate and is reflective of the high sulphide content of the ore and its reactivity Carbon Loading Kinetic Tests A triple contact sequential carbon loading test was conducted on the two master composite samples from the bulk leach tests to determine the Fleming k and n values for the adsorption circuit. Carbon was added to the first of three leached slurry samples and agitated for 2 hours, then separated and added to the second slurry sample and agitated for a further 2 hours, then separated and added to the third slurry sample and agitated for a further 20 hours (24 hours total contact time). Carbon concentrations were selected to target a gold in solution to gold loaded carbon ratio equivalent to 3,500 g Au/t C. Lycopodium Minerals Canada

167 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The results of the triple contact sequential carbon loading kinetic tests are summarized in Table and Table Table MC1 Results of Sequential Triple Contact CIP Testwork Cycle Solution Elaspsed Time Solution Carbon Carbon Final Volume Cycle Cumulative Gold Calc'd Calc'd Carbon Loading Loading Assay Incremental Cumulative Au Au Au Au ml Hours Hours ppm ppm ppm ppm Lycopodium Minerals Canada

168 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table MC2 Results of Sequential Triple Contact CIP Testwork Cycle Solution Elaspsed Time Solution Carbon Carbon Final Volume Cycle Cumulative Gold Calc'd Calc'd Carbon Loading Loading Assay Incremental Cumulative Au Au Au Au ml Hours Hours ppm ppm ppm ppm The loaded carbon assays match reasonably well with the calculated assays, although both are lower than anticipated. The solution assays are also higher than would usually be expected at the completion of this test. It was established that the leached samples sat for approximately a week between the leach test and the carbon adsorption work. This may have adversely impacted slurry chemistry as the slurry is very reactive, and cyanide and/or ph levels may have dropped to the point where copper cyanide species were formed which are more likely to load. The formation of Cu(CN) - 2 is favoured at ph values of less than 6 and at low cyanide concentrations. This species is more likely to 2-3- load onto activated carbon than Cu(CN) 3 and Cu(CN) 4 which are present at higher ph and cyanide concentrations. Cyanide and lime was not added prior to or during the adsorption test. The rate curves for these tests are represented in Figure Lycopodium Minerals Canada

169 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure Rate Curves for Sequential Triple Contact CIP Testwork 10,000 Gold Loading (g/t) 1,000 MC1 MC TIME (Hours) The Fleming constants that have been calculated from these graphs over the first 6 hours are shown below: MC1 MC2 Fleming rate constant k, h Fleming equilibrium constant n Cumulative carbon loading, g Au/t C 2,986 2,040 The Fleming k value is an empirical rate constant which is dependent on factors such as mixing efficiency and carbon sizing. The Fleming n value is constant related to the gold loading capacity of the carbon and is influenced by carbon activity, ph, ionic strength and the presence of competing ions such as copper cyanide complexes. Historically, work carried out by Lycopodium on Australian and overseas gold ores has resulted in Fleming constants ranging from: k : 70 to 260 n : 0.55 to 0.80 Clean primary ores tend to have k values of >200, while free milling oxide ores average approximately 150. Lycopodium Minerals Canada

170 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Despite the lower than anticipated overall gold loadings, the carbon loading kinetics are within the range of values observed in other operations. The modelling of the adsorption circuit for the process plant has been based on the MC1 values, due to the abnormal silver and copper content of MC2. Loaded carbon assays from these tests are presented in Table Table Loaded Carbon Assays from Triple Contact Tests Assay Units MC1 MC2 Au (ppm) Ag (ppm) As (ppm) Cu (ppm) Hg (ppm) Ni (ppm) Pb (ppm) <5 <5 Zn (ppm) The high copper assays of the loaded carbon strengthen the argument that solution chemistry was not optimised to prevent copper loading. Nevertheless, a cold cyanide wash has been incorporated into the elution circuit design, as this will assist in removing adsorbed copper along with other base metals and is a relatively inexpensive addition to the plant design. Despite the high arsenic solution assay, very little arsenic loaded as expected Carbon Equilibrium Loading Equilibrium carbon loading test work was carried out to determine the ultimate precious metal loading onto carbon. Carbon samples were contacted with the bulk leach slurry, with carbon and final solutions assayed for gold and silver. The carbon equilibrium loading empirical constants, derived from the test work, were then substituted into the following equilibrium loading equation, to determine the theoretical maximum carbon loading: Log (X/M) = m Log C + Log K Where: X/M = mg of gold adsorbed per gram of carbon at equilibrium, C = mg of gold remaining in solution, K, m = constants derived from test work. The results of the carbon equilibrium loading test work are summarized in Table Lycopodium Minerals Canada

171 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Carbon Equilibrium Loading Comp Gold Solution Tenor (ppm) MC MC The data from MC1 has been used in the carbon modelling work, which has been used to size the carbon adsorption and desorption (elution) circuits Intensive Cyanidation of Gravity Concentrate Subsequent to the initial egrg testwork that was performed by Consep, a second sample of MC1 was sent for testing, specifically to examine the deportment of gold, silver, copper, arsenic and mercury in the gravity and intensive leach stages. The concentrates produced from each pass through the Knelson concentrator were intensively leached. An elemental analysis of the leach solution was conducted to determine the concentration of gold, silver, copper, arsenic and mercury in solution. Results from this work are presented in Table Table GRG and Intensive Leach Recoveries Recovery, % Parameter Au Ag Cu As Hg GRG Intensive Leach Overall GRG + IL The intensive leach gives very high recoveries of precious metals, with 97.0% of the gold and 95.9% of the silver being recovered into solution. In common with the variability leach testwork, low and very low levels of copper and arsenic are leached into solution at 8.8% and 0.2% respectively. Mercury is also low at 27.6%. However, it should be noted that the test concentrate mass recovery is 1.4%, which is very much higher than what would typically be seen in a production scale gravity circuit. A typical gravity circuit will recover between 0.02 and 0.08% of the mill feed as a concentrate. A typical mass pull of 0.05% is 28 times less than the 1.4% mass pull of the egrg test. This has a significant influence on the ratio of gold to other metals in the gravity concentrate. Lycopodium Minerals Canada

172 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Based on past experience with Knelson concentrators, the gold grade of the concentrate in the plant is expected to be in the range of 5,000 to 10,000 g/t. In comparison, the gold grade of the first pass for the laboratory test was only 467 g/t Au. Table shows the concentrate grades achieved for each pass during the egrg test. Table egrg Concentrate Grades Concentrate Grade, ppm Parameter Au Ag Cu As Hg Pass , Pass , Pass , As the gold grade of the concentrate will be very much higher in the plant than indicated in Table 13.20, it is expected that the copper, arsenic and mercury grades in the plant will not exceed the values achieved in the tests. Although arsenic grades in the concentrate are high in the test, in a plant situation the mass pull to concentrate will be in the order of 30 times lower. In addition, as shown in Table 13.19, only a very small proportion, 0.2%, of the arsenic is cyanide soluble. The average arsenic concentration in solution was 34.2ppm which is similar to the pregnant solution analysis from the MC1 bulk cyanide leach. These levels are not expected to present process or HSE risks in the plant. Based on these test results, an intensive cyanidation module and electrowinning cell has been incorporated into process plant design to treat gravity concentrate Ancillary Testing Rheology Slurry viscosities were measured at 40, 45, and 50% w/w solids on a leach feed sample of each master composite, ground to a P 80 of 63 microns and adjusted to a ph of 10.5 with lime. It was anticipated that further tests would be carried out once thickener vendor data was received, and likely tailings pumping densities were established. This has not been conducted at the time of writing. The results are summarized in Table Lycopodium Minerals Canada

173 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Slurry Rheology for Master Composites Grind Size Comp P 80 µm Temp C MC1 63 MC2 63 Pulp Density % solids w/w Bohlin Visco 88 Shear Rate ( Sec -1 ) 38.9 cp 67.4 cp cp cp Yield stress values of 0 Pa were recorded for each of the master composites for all slurry densities tested. A slurry viscosity of less than 3,500 cps at a shear rate of 2.6 s -1 is considered acceptable for mixing / screening applications. A slurry viscosity of less than 100 cps at a shear rate of 119 s -1 and a yield stress of less than 30 Pa is considered acceptable for standard centrifugal pumping applications The results indicate the following: Agitation and flow through intertank screens will not be an issue at slurry densities of up to 50% solids w/w. Pumping of slurries at densities up to 50% w/w solids using centrifugal pumps will not be a problem Oxygen Uptake Testing The oxygen uptake test was conducted on feed sample of each master composite to determine the change in the rate of oxygen consumption of the slurry with time. No gravity step was included prior to the test, as it was deemed important not to introduce air into the sample through rehandling. The oxygen uptake test was conducted under the following conditions: A primary grind P 80 of 63 µm. 50% solids w/w Perth tap water. ph 10.5 initially and maintained greater than 10.0 adjusted with commercial lime (60% available CaO). Starting cyanide concentration of 0.035% w/v. Lead nitrate of 100 g/t for MC1 and 50 g/t for MC2. Lycopodium Minerals Canada

174 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Dissolved oxygen levels maintained at 20 ppm between uptake rate determination measurements. DO measurements at 0, 1, 2, 3, 4, 5, 6, 24 hours. The results are presented graphically in Figure Figure Oxygen Uptake Rate Oxygen Upteake Rate (mg/l/min) MC1 MC TIME (Hours) The results indicate the following: For MC1, the oxygen demand for the slurry was 0.12 mg/l/min after one hour and remained relatively constant thereafter up to 24 hours. For MC2, the oxygen demand for the slurry was 0.19 mg/l/min after one hour but decayed to approximately 0.10 mg/l/min after 3 hours before rising to approximately 0.16 mg/l/min after 6 hours. This oxygen consumption rate is considered to be very high. Oxygen addition methods such as external oxygen contactors and spargers will be required to achieve adequate oxygen concentrations in the slurry for efficient gold dissolution. This confirms the importance of high dissolved oxygen concentrations as established in the cyanide leach testwork. Lycopodium Minerals Canada

175 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Pre-leach and Tailings Thickening Testwork A 20 kg sample of each master composite sample was ground to a P 80 of 63 microns and adjusted to a ph of 10.5 by addition of commercial lime (60% available CaO). These samples were submitted for thickening testwork at Outotec. Testwork comprised flocculant screening tests, dynamic thickening tests, using a 99mm diameter Supaflo high rate thickener test rig, and basic rheology tests. The salient data from the testwork is summarized in Table Table Outotec Thickener Test Results Parameter Units MC1 MC2 MC1 MC2 Flux Rate t/m 2.h Flocculant Dose g/t Underflow Density % w/w Overflow Clarity ppm < <150 <150 Yield Stress Pa Thickener Diameter * m * At design case of 1.34 Mtpa The outcomes of testwork were: Based on static cylinder settling tests Magnafloc 333 flocculant was deemed to provide the best overall settling characteristics and was used for dynamic thickening testwork. The optimum thickener feed density was determined to be 16% w/w solids. Flux rates of up to 1.5 t/m 2.h were achieved at underflow densities of greater than 57% solids w/w. In a full scale thickener Outotec has indicated that densities 2 to 3% higher than those achieved in testwork should be attained. At lower flux rates of 0.5 t/m 2.h, densities of over than 62% solids w/w were achieved. Subsequent to this work at Outotec, Birimian organised for 20kg of MC1 in slurry form, at a nominal P 80 of 63 microns, to be sent to FLS for thickening testwork. This sample was subject to a series of static four litre cylinder tests, where the cylinders were fitted with slowly rotating pickets to simulate the action of a rake in a full scale thickener, as well as continuous fill deep tube settling tests. Basic rheology parameters were also determined. The ph of the slurry was maintained between 10.0 and 10.5 using sodium hydroxide pellets. The salient data from the testwork is summarized in Table Lycopodium Minerals Canada

176 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table FLS Thickener Test Results Parameter Units MC1 MC2 Duty Pre-leach Tailings Flux Rate t/m 2.h Flocculant Dose g/t Expected Underflow Density % w/w Overflow Clarity ppm <150 <150 Yield Stress Pa <11 <11 Thickener Diameter m On the basis of the testwork, and with allowance for design margin, a high rate thickener with a diameter of 14 m will be used for the pre-leach duty and a 16 m diameter thickener for the tailings duty Tailings Filtration Testwork Tailings filtration was considered as an option for the Natougou plant to reduce water consumption. The two master composite slurry samples that had been used for the thickener testwork were tested to determine pressure filter sizing parameters. A lithium tracer was used in the testwork to determine the wash efficiency that could be achieved if a wash step was incorporated for cyanide recovery. Results from the testwork are presented in Table Table Outotec Tailings Filtration Test Results P 80 of 63 microns Without Cake Wash With Cake Wash Parameter Units MC1 MC2 MC1 MC2 Filtration Rate (dry) kg/m 2.h Cake Moisture Content % w/w Av. Drying Air Consumption L/min Cake Thickness mm Cake Density kg/l Wash Water Consumption m 3 /t Wash Efficiency % Total Cycle Time min Lycopodium Minerals Canada

177 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The test results indicate that: The target moisture of less than 15% w/w can be achieved on both composites, when filtering without a wash step. When a wash step was included, the moisture increased to 12.8% for MC1 and 15.7% for MC2. Filtration rates of over 350 kg/m 2.h could be achieved for both composites without a cake wash. When a cake wash was included this dropped to 259 kg/m 2.h for MC1 and 197 kg/m 2.h for MC2. Wash efficiencies of over 99% could be achieved with 2BV of fresh water. Inclusion of a cake wash for cyanide recovery increased the cycle time and the cake moisture. FLS also conducted filtration testing on the MC1 sample that had been used for the thickening testwork, at a nominal P 80 of 63 microns. Both vacuum belt and pressure filtration testing was conducted. Table summarizes the sizing parameters generated from both the pressure and vacuum filtration testwork. Table FLS Tailings Filtration Design Parameters Parameter Units Values Feed Solids % w/w 65 Vacuum Filter Cake Formation Vacuum Pressure kpa -71 Cake Drying Vacuum Pressure kpa -68 Filtration Rate (dry) kg/m 2.h 1413 Cake Thickness mm 20 Cake Moisture Content % w/w 20 Form Time min 0.48 Dry Time min 0.57 Total Cycle Time min 1.07 Pressure Filter Filtration Rate (dry) kg/m 2.h 273 Cake Moisture Content % w/w 20 Total Cycle Time min 8.5 Filter Cycles per Hour Lycopodium Minerals Canada

178 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Variability Metallurgical Testwork Head Analysis A summary of the multiple gold assay and multi-element head assay performed on the variability composites are presented in Table to Table For brevity, not all the head assays conducted have been reported in these tables. Lycopodium Minerals Canada

179 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Detailed Head Analysis of Variability Samples Section 1 & Section 2 (1, 3 to 11) Assay Units Au 1 ppm Au 2 ppm Au 3 ppm Au : SFA ppm Ag ppm <0.30 < < <0.30 < As ppm 3,200 1,000 1, ,600 7, ,400 9,000 2,200 6,600 1,400 2,800 3,400 1,000 1,200 6,400 8,000 5,400 6,800 Ba ppm C total % C organic % 0.06 <0.03 < < <0.03 < <0.03 < Ca % Cd ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Co ppm Cr ppm < Cu ppm Fe % Hg ppm 0.2 < <0.1 < <0.1 < Mn ppm 1,500 1,500 1,400 1,400 1,500 1,300 1,600 1,500 1,200 1,700 1,500 1,600 1, ,700 1, ,100 1,200 1,100 Mo ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Nb ppm <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 20 <10 10 <10 <10 Ni ppm Pb ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 S total % S sulphide % Sb ppm SiO 2 (%) Sr ppm Te ppm < <0.2 < < <0.2 <0.2 <0.2 <0.2 <0.2 < < U ppm <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 Zn ppm Lycopodium Minerals Canada

180 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Detailed Head Analysis of Variability Samples Section 2 (12 to 24, 27, 29) Assay Units Au 1 ppm Au 2 ppm Au 3 ppm Au : SFA ppm Ag ppm <0.30 < < <0.30 <0.30 < < As ppm 390 3,600 1, ,400 3,400 1,120 5, , ,600 Ba ppm C % C org % < Ca % Cd ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Co ppm Cr ppm < Cu ppm Fe % Hg ppm < Mn ppm 1,600 1,400 1,500 1,900 1,100 1,500 1,200 1,800 1, ,500 1,600 1,500 1,700 1,300 Mo ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Nb ppm <10 <10 <10 10 <10 <10 <10 <10 <10 10 <10 <10 <10 <10 <10 Ni ppm Pb ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 S % S -2 % Sb ppm SiO 2 (%) Sr ppm Te ppm <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 < <0.2 <0.2 U ppm <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 Zn ppm Lycopodium Minerals Canada

181 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Detailed Head Analysis of Variability Samples Section 3 (1 to 15) Assay Units Au 1 ppm Au 2 ppm Au 3 ppm Au : SFA ppm Ag ppm < <0.30 < As ppm 1,800 2,000 1,000 2,200 3,600 5,600 4, ,800 1,400 2,000 3, Ba ppm C % C org % <0.03 < < <0.03 Ca % Cd ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Co ppm Cr ppm Cu ppm Fe % Hg ppm 0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 < <0.1 <0.1 <0.1 Mn ppm 1,100 1,200 1,500 1,100 1,200 1,300 1,200 1,300 1,600 1,400 1,100 1,600 1,600 1, Mo ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Nb ppm <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 Ni ppm Pb ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 S % S -2 % Sb ppm SiO 2 (%) Sr ppm Te ppm U ppm <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 Zn ppm Lycopodium Minerals Canada

182 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Detailed Head Analysis of Variability Samples Section 4 (1 to 6, 8 to 13) Assay Units Au 1 ppm Au 2 ppm Au 3 ppm Au : SFA ppm Ag ppm < As ppm 3,800 2,400 4,200 6, ,600 10,800 6,800 2, ,400 1,000 Ba ppm , C % C org % <0.03 <0.03 <0.03 <0.03 Ca % Cd ppm <20 <20 <20 <20 <20 <20 20 <20 <20 <20 <20 <20 Co ppm Cr ppm Cu ppm Fe % Hg ppm 0.1 < < Mn ppm 1,200 1,500 1,300 1,300 1,200 1, ,400 1,100 1,400 1,300 1,400 Mo ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Nb ppm <10 <10 <10 <10 10 <10 10 <10 <10 <10 <10 <10 Ni ppm < Pb ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 S % S -2 % Sb ppm SiO 2 (%) Sr ppm Te ppm <0.2 <0.2 <0.2 < <0.2 < <0.2 U ppm <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 Zn ppm Lycopodium Minerals Canada

183 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Detailed Head Analysis of Variability Samples Section 5 (1 to 20) Assay Units Au 1 ppm Au 2 ppm Au 3 ppm Au : SFA ppm Ag ppm <0.30 <0.30 <0.30 <0.30 <0.30 <0.30 <0.30 <0.30 < <0.30 <0.30 <0.30 <0.30 < As ppm 910 3,080 1, , ,380 1,570 1, ,580 1,980 3,230 Ba ppm C total % C organic % <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 Ca % Cd ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Co ppm Cr ppm Cu ppm Fe % Hg ppm < <0.1 < <0.1 <0.1 < <0.1 < Mn ppm 1,700 1,500 1,400 1,500 1, ,200 1,400 1,600 1,600 1,700 1,200 1,000 1,000 1, ,300 1,300 1,200 1,000 Mo ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Nb ppm <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 Ni ppm Pb ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 < <20 <20 <20 <20 <20 S total % S sulphide % Sb ppm < < < SiO 2 (%) Sr ppm Te ppm < < < U ppm <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 Zn ppm Lycopodium Minerals Canada

184 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Detailed Head Analysis of Variability Samples Section 5 (21 to 29), Section 6 Assay Units Au 1 ppm Au 2 ppm Au 3 ppm Au : SFA ppm Ag ppm <0.3 <0.3 <0.3 < <0.3 <0.3 < < <0.3 As ppm 1, , ,960 1, ,190 5,110 4,140 3,980 1,670 4,250 2,350 5,790 7,020 1,010 Ba ppm C total % C organic % <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 < <0.03 <0.03 <0.03 <0.03 <0.03 < <0.03 Ca % Cd ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Co ppm Cr ppm Cu ppm Fe % Hg ppm < < < <0.1 < <0.1 Mn ppm 1,400 1,100 1,400 1,200 1,000 1,100 1,000 1,600 1,500 1,300 1, ,100 1,500 1,800 1,400 1, ,500 Mo ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 Nb ppm <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 Ni ppm Pb ppm <20 <20 <20 <20 <20 <20 <20 <20 <20 <20 20 <20 <20 <20 <20 <20 <20 <20 <20 S total % S sulphide % Sb ppm 0.20 < SiO 2 (%) Sr ppm Te ppm < <0.2 <0.2 < <0.2 U ppm <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 Zn ppm Lycopodium Minerals Canada

185 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Detailed Head Analysis of Variability Samples Section 0 Assay Units Au 1 ppm Au 2 ppm Au 3 ppm Au : SFA ppm Ag ppm <0.30 < < As ppm 1, ,210 Ba ppm C total % 0.27 < < C organic % < Ca % Cd ppm <20 <20 <20 <20 <20 Co ppm Cr ppm <25 <25 <25 <25 <25 Cu ppm Fe % Hg ppm Mn ppm 1,500 1,100 1,400 2,000 1,600 Mo ppm <20 <20 <20 <20 <20 Nb ppm <10 <10 <10 <10 <10 Ni ppm <20 40 Pb ppm S total % <0.02 <0.02 < <0.02 S sulphide % <0.02 <0.02 <0.02 <0.02 <0.02 Sb ppm SiO 2 (%) Sr ppm Te ppm <0.2 <0.2 <0.2 < U ppm <2 <2 <2 <2 <2 Zn ppm Lycopodium Minerals Canada

186 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page The following comments can be made in reference to the primary ore samples (Sections 1 to 6): Average gold head grades range from 0.55 g/t to 24.6 g/t, offering a wide range over which to assess the effect of head grade on recovery. The duplicate gold assays vary, which indicates the presence of coarse grained gold. The screen fire assay values range from 0.59 g/t to 17.7 g/t. Considering the coarseness of the gold, there was generally a good match between the screen fire assays and the average of the fire assays. There were a few notable exceptions, namely samples 1.04, 4.13 and 5.29, which recorded much lower screen fire assays than the average of the three fire assays. Silver grades range from <0.30 g/t to 2.1 g /t. Gold to silver ratio averages approximately 10. At no time was the silver higher grade than gold. Mercury levels are generally low, averaging 0.10 ppm, and should not present an environmental or occupational health risk in carbon regeneration, elution or electrowinning. Arsenic levels are high, averaging over 2000 ppm across all the samples, which is similar to the BHBCs. There was a wide range in assay from 20 ppm to 10,800 ppm. Sulphide sulphur grades range from 0.16% to 3.22%, with an average of 1.14%. Concentrations of base metals such as copper, lead and zinc are generally low (<500 ppm total) and this should not result in increased cyanide consumption. However, the base metal levels are similar to those seen in MC1, and therefore cyanide concentration and ph will have to be monitored and controlled in the leach and adsorption circuits to avoid copper loading, given the reactive nature of the sulphides. Organic carbon levels are low and preg-robbing should not be a problem Gravity / Leach Testwork All variability composites were subjected to the following test conditions: Primary grind P 80 of 63 microns. Gravity concentration by Knelson concentrator, amalgamation and leaching of gravity plus amalgam tailings. Leach pulp density of 50% solids w/w in Perth tap water. ph adjusted with commercial lime (60% available CaO). Lycopodium Minerals Canada

187 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Initial cyanide concentration of 0.035% w/v, maintained at 0.025% w/v. Lead nitrate addition of 50 g/t. Dissolved oxygen levels of 20 ppm or greater. 36 hour leach duration. All tests were conducted in a bottle roll using a 1 kg sample. No specific changes were made for testing the small number of oxide samples received, as it is likely oxide ore will be blended. The gravity recovery and cyanidation leach results are summarized in Table to Table Table Bottle Roll Cyanide Leach Results for Section 1 Composites Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Mean Median Table Bottle Roll Cyanide Leach Results for Section 2 Composites Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Lycopodium Minerals Canada

188 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Mean Median Table Bottle Roll Cyanide Leach Results for Section 3 Composites Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Mean Median Lycopodium Minerals Canada

189 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Bottle Roll Cyanide Leach Results for Section 4 Composites Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Mean Median Table Bottle Roll Cyanide Leach Results for Section 5 Composites (1 to 15) Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Lycopodium Minerals Canada

190 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Bottle Roll Cyanide Leach Results for Section 5 Composites (16 to 29) Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Mean S Median S Table Bottle Roll Cyanide Leach Results for Section 6 Composites Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Mean Median Lycopodium Minerals Canada

191 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Bottle Roll Cyanide Leach Results for Section 0 Composites (Oxide) Sample ID % Au Recovery % Au Recovery % Au Recovery 36 hours Residue Grade Consumption (kg/t) Gravity 24 hours g/t Au Lime NaCN Mean Median For primary ore samples, the results of variability testing indicate that: Gold recovery in the gravity stage averaged 39% (range 13 to 85%). There was a distinct relationship between gold recovery in the gravity stage and overall gold recovery, as indicated in Figure The two master composites are in indicated in red. Figure Gravity Recovery vs. Overall Gold Recovery Gravity Recovery, % y = e x R² = Overall Gold Recovery, % The average residue grade for the variability samples was 0.28 g Au/t and ranged from 0.04 to 0.98 g Au/t. Overall gold (gravity + cyanide leach) recoveries varied from 83.8% to 99.1% with an average of 91.8%. Sections 3 and 5, with lower arsenic head grades, generally had higher overall recoveries than Sections 1, 2, 4 and 6, with higher arsenic head grades. Lycopodium Minerals Canada

192 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page An average cyanide consumption of 0.21 kg/t was observed for the variability samples with a range of 0.08 to 0.47 kg/t. There was quite a variation between samples throughout the deposit. The highest cyanide consumption rates were seen in samples from Section 6; this was the section of the deposit with the highest arsenic head grades. The average cyanide consumption for the 15 borehole bench composites was 0.17 g/t, so the variability testwork compares favourably. As the master composite testwork was conducted using a 3kg vat leach, these results are not directly comparable. Cyanide consumption from the optimised 3kg vat leach tests was 0.58 kg/t for MC1 and 0.82 kg/t for MC2. An average lime consumption (60% available CaO) of 0.40 kg/t was observed for the variability samples with a range of 0.19 kg/t to 0.71 kg/t. The average of the borehole bench composites was 0.57 kg/t. Leaching is rapid and substantially completed after 12 hours. A number of leach tests were also assayed for copper and arsenic to examine leach rates and final solution grades. A smaller number of repeat tests were also carried out by hand panning the Knelson concentrate rather than using amalgamation; so that mercury recovery could be investigated without the complication of potential mercury carry over from the amalgamation stage. The results from these tests are presented in Figure to Figure Figure Selected Variability Leach Curves Arsenic, Sections 1 and EXTRACTION (%) TIME (Hours) Lycopodium Minerals Canada

193 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure Selected Variability Leach Curves Arsenic, Sections 5 and EXTRACTION (%) TIME (Hours) Figure Selected Variability Leach Curves Copper, Sections 1 and EXTRACTION (%) TIME (Hours) Lycopodium Minerals Canada

194 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Figure Selected Variability Leach Curves Copper, Sections 5 and EXTRACTION (%) TIME (Hours) Figure Selected Variability Leach Curves Mercury EXTRACTION (%) TIME (Hours) Lycopodium Minerals Canada

195 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Arsenic extractions are all very low, at less than 0.3% after 36 hours. Arsenic solution assays ranged from 0.05 to 21.3 ppm. In comparison, the bulk leaches on MC1 and MC2 were 31.3 ppm and 31.7 ppm respectively, which are a similar order of magnitude. Copper extractions are low to moderate and generally within 6 to 12% after 36 hours. Sample 6.07 recorded a higher extraction of 15.4%, despite having a typical total copper head grade of 85ppm. Cyanide soluble copper head assays were not conducted. Solution assays of between 7.3 ppm and 21.6 ppm were recorded. MC1 bulk leach recorded a solution assay of 10.9 ppm which is within the range seen in the variability assays. MC2 is not considered representative in terms of copper head assay. Mercury extractions were very low, at less than 5% overall. The mercury in solution was also very low at approximately mg/l. Therefore specific mercury control measures are not considered necessary for the Natougou ore based on the bulk leach solution assays and these variability results. For the small number of oxide variability samples tested the results indicate that: Gold recovery in the gravity stage averaged 46% (range 34 to 46%). The average residue grade for the oxide samples was very low 0.07 g Au/t and ranged from 0.03 to 0.15 g Au/t. Overall gold (gravity + cyanide leach) recoveries varied from 94.6% to 98.2% with an average of 96.7%. An average cyanide consumption of 0.19 kg/t was observed for the oxide samples with a range of 0.15 to 0.23 kg/t. This is comparable with the primary ore samples. The lime consumption (60% available CaO) rate was extremely high, with an average of 6.08 kg/t, and a range between 3.14 and 8.53 kg/t. This is likely to be related to the high saprolite content, which can cause the ore to have naturally low ph. No further testwork was carried out targeting lower ph values in the leach to optimize lime (or caustic) consumption. Similarly to the primary ore samples, leaching is rapid and substantially complete after 12 hours. One sample gave a slightly elevated organic carbon assay at 0.15%. As no carbon testwork was performed, the likelihood of pre-robbing cannot be established. It should be noted that all of the variability testwork was conducted in Perth tap water, which is of good quality. Lycopodium Minerals Canada

196 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Gravity / Leach Testwork Diagnostic Leach A three stage diagnostic leach was conducted on selected samples with low recovery, to identify the deportment of gold in the residue. The diagnostic leach method employed was the same as had been previously used for the borehole bench composites. The diagnostic leach indicated that the majority (58 to 87%) of the unrecovered gold is locked in sulphides. Cyanide extractable gold ranged from 6 to 27%, indicating that with very aggressive leach conditions, longer leach times and/or finer grind, additional recovery is possible. In particular samples 3.12, 5.17 and 6.06 all had over 15% additional gold extracted. Silicate locked gold ranged from 4 to 25%. Mode of occurrence of gold in residue does not appear to be consistent within zones, for the samples tested Metallurgical Testwork Summary Selected Treatment Route The metallurgical treatment route selected has been based on the results of the current testwork program and can be summarized as follows: Single stage crushing. Milling circuit consisting of SAG mill, tower mill and pebble crusher. Use of a tower mill rather than a ball mill gives improved energy efficiency over a ball mill whilst maintaining the flexibility and expandability of a SAG mill based circuit. Pre-leach thickening. The ability to operate at low cyclone overflow densities in the grinding circuit, thereby improving classification efficiencies which potentially improves overall gold recoveries. Consistent and sufficiently high leach feed density to ensure adequate solids and carbon suspension in the adsorption circuit. Cyanide leaching using external oxygen contactors and oxygen spargers, in conjunction with lead nitrate additions. Kemix carousel carbon adsorption circuit, to maximise carbon loadings and minimise carbon inventory. Split AARL stripping circuit, with cold cyanide wash. Lycopodium Minerals Canada

197 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Treatment of pregnant solutions via electrowinning and smelting of sludge. Tailings thickening Key Process Design Parameters The key process design parameters derived from testwork, for the design blend, are as follows: Comminution Circuit Design Crushing Work Index * Bond Rod Mill Work Index Bond Ball Mill Work Index 20.0 kwh/t 21.9 kwh/t microns JKTech A x b 32.2 Abrasion Index Crushing Circuit P 80 Milling Circuit P mm 63 microns * A higher figure used for modelling that derived from testwork. Leach and CIP Design Residence Time Pulp Density 36 hours 50% w/w solids ph 10.5 Carbon loading kinetic parameters k 204 n Metallurgical Recoveries The variability testwork is presented in Section , and shows that overall gold recoveries for the Natougou primary ore ranged from 84 to 99%. There was a distinct relationship between recovery in the gravity stage and overall recovery. The results suggest that the residue grade is moderately correlated with the amount of coarse gold in the sample (measured by % gold in +75 micron fraction of the screen fire assay), arsenic head assay, and gold head assay. A constant tail relationship is not appropriate. Lycopodium Minerals Canada

198 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page With consideration of the parameters currently in the geological model, a relationship between the residue grade and the gold head assay has been developed and is illustrated in Figure Figure Residue Grade Versus Head Grade y = x R² = Residue Grade, g/t Au Head Grade (Screen Fire Assay) g/t Au So, the predictive equation is: Gold Residue (g/t Au) = *Gold Head Assay (g/t Au) R 2 = 0.34 So, for a gold head assay of 4.36 g/t Au, the gold residue grade would be 0.31 g/t Au. As silver residue grades are frequently at the assay detection limit, and no trend with head grade is apparent, it is recommended that a simple arithmetic average of all the silver recovery figures be used i.e. 66%. The combined gravity / leach recovery is a statement of the anticipated metallurgical extraction for the Natougou primary ore type but is not a statement of the anticipated plant performance. Consideration must be given to soluble losses, potential for short circuiting in the leach circuit and other associated plant problems that may impact on the overall plant gold and silver recovery Reagent Consumption Testwork The average lime additions, lead nitrate additions and cyanide consumptions from the optimized testwork conducted on the master composites, and bottle roll tests conducted on the variability composites are summarized in Table All master composite leach and variability testwork was conducted using Perth tap water. Lycopodium Minerals Canada

199 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Table Reagent Consumptions in Master Composite and Variability Tests Note: Sample / Zone Consumption, kg/t Lime 1 Cyanide 2 Lead Nitrate MC MC Section Section Section Section Section Section Section Variability Av (Primary Ore) Testwork conducted using commercial lime with 60% available CaO. 2. Master composite testwork 3kg vat leach; variability testwork 1kg bottle roll. Although it would be expected that the vat leaches would have higher cyanide consumption than the bottle roll tests due to the potential for HCN gas stripping in an open vessel, the magnitude of the difference is surprising. Anticipated Plant Reagent Requirements The anticipated plant lime and cyanide addition rates to the leach circuit differ from the testwork results due to the following: The testwork was conducted with commercial lime having a CaO availability of approximately 60%. The lime supplied is likely to have an available CaO content in excess of 75% and the required plant lime addition must be scaled back accordingly. The average of the lime addition from the variability testwork on primary ore has been used for conservatism. The cyanide requirements for the plant must make allowance for residual cyanide in the tail solution. For the purposes of calculating the anticipated cyanide addition to the circuit a residual cyanide level of 200 mg NaCN/L has been adopted. This is equivalent to 0.20 kg/t, and is in addition to the cyanide consumed during leaching. A 10% contingency over the bottle roll test cyanide consumption has also been allowed to take into account the open tanks in a plant situation. The unadjusted cyanide consumption from MC1 has been used for cyanide dosing circuit design purposes, which gives just over 25% design margin. Lycopodium Minerals Canada

200 NI TECHNICAL REPORT - MINERAL PROCESSING AND METALLURGICAL TESTING Page Predicted lime, sodium cyanide and lead nitrate consumptions for the selected treatment route are shown in Table 13.42, assuming that good quality water is used in the plant. Table Predicted Plant Reagent Consumptions Reagent Operating kg/t Design kg/t Lime Cyanide Lead Nitrate Note: 1. The testwork was conducted with commercial lime having a CaO availability of 60%. The lime supplied is likely to have an available CaO content in excess of 75% and the lime addition has been accordingly scaled back Testwork Conclusions The following conclusions can be drawn from the current and previous metallurgical and comminution testwork programs: The Natougou primary ore is an abrasive, competent ore with above average comminution energy requirements. The Natougou primary ore has a high gravity recoverable gold content; leach kinetics are very slow when gravity is not included in the flowsheet. High dissolved oxygen levels and lead nitrate are required to achieve fast leach kinetics and adequate gold recovery. High copper loading on carbon could be a problem if adequate free cyanide levels are not maintained and the ph is not controlled through the adsorption circuit. Anticipated lime consumption for primary ore is low to moderate, provided good quality water can be provided on site. Cyanide consumption is likely to be moderate. High lime consumption will be experienced if oxide ore forms part of the feed blend. LOM head grades for the process plant are expected to average 4.15 g/t with a gold recovery of 92.9%. Quantities of oxide ore are expected to be very low with only ± 1% of reserves. Lycopodium Minerals Canada

201 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page MINERAL RESOURCE ESTIMATES Mineral Resources reported in Section 1 were prepared by Mr. John Graindorge, Principal Consultant, a full-time employee of Snowden. Snowden is independent of SEMAFO. The resources were modelled using CAE Studio 3 (Datamine) and Snowden Supervisor software. Mineral Resources that are not Mineral Reserves do not have demonstrated economic viability. Snowden is unaware of any issues that could materially affect the Mineral Resources in a detrimental sense. These conclusions are based on the following: SEMAFO has an Exploration Permit (Boungou Permit) covering the Natougou deposit in good standing. Snowden is not aware of any outstanding legal issues relating to the project. There are no known marketing, political or taxation issues. There are no known infrastructure issues. SEMAFO is currently operating the Mana Mine in Burkina Faso Geological Interpretation SEMAFO provided Snowden with a geological interpretation of the mineralized structures as Datamine format wireframes and by flagging the mineralized intervals in the drillhole datasets in Microsoft Excel format. The interpretation was based on a variety of criteria such as the geological field logging of drill core and RC chips, gold assays, trace element data collected from pulps using an Innov-X Omega series handheld XRF and drill core photographs. The geological interpretation of the main lode is shown in Figure Core was oriented during drilling for all angled diamond drillholes. Alpha and beta measurements were collected on the shear fabric developed in the shear zone as well as on lithological contacts as part of the geological logging procedure. Measurements were plotted as apparent dips on drill sections to assist with interpretation. Oriented core in angled drillholes allowed SEMAFO geologists to plot strike and dip measurements for the mineralized lode. The plotting of apparent dips on cross-sections corresponded with the orientations of the interpreted mineralized lode. An offset to the main lode was interpreted in the northwest portion of the deposit. Snowden Mining Industry Consultants

202 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page 14.2 SEMAFO and Snowden recognize the presence of mineralization above the Boungou Shear Zone in the northwest region of the deposit, which is referred to as the western hangingwall mineralization. Snowden notes that this hangingwall mineralization carries significant uncertainty and further work will be required to adequately define the nature of this zone. The zone is bounded by the top contact of the main mineralized zone, and is bounded to the east by an interpreted feature that is evidenced by marked decreases in potassium and gold values. Given the uncertainty, no interpretation of the western hangingwall mineralization has been completed and this zone is included within the surrounding waste domain. As such, the western hangingwall mineralization does not form part of the Natougou Mineral Resource. The south-western margin of the mineralization is bounded by a granite contact, which was interpreted based on the drillhole logging. Figure 14-1 Isometric view showing geological interpretation for the main mineralized structure N Source: Snowden Snowden Mining Industry Consultants

203 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Drill sample analysis A collar location plan of drillholes, as of October 2015, used for the Natougou Mineral Resource estimate is presented in Figure Snowden Mining Industry Consultants

204 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page 14.4 Figure 14-2 Collar location plan for drilling at Natougou up to October 2015 Source: Snowden Snowden Mining Industry Consultants

205 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Sample Compositing The drillhole data was composited downhole prior to running the estimation process using a 1 m sample interval to minimize any bias due to sample length. The compositing was run within the attribute fields to ensure that no composite intervals crossed any lithological or grade boundaries. To allow for uneven sample lengths within each of the domains, the composite process was run using the variable sample length method. This adjusts the sample intervals, where necessary, to ensure all samples are included in the composite file (i.e. no residuals) while keeping the composite interval as close to the desired interval as possible. The compositing process was checked by: Comparing the lists of attribute field values in the raw and composite files; these should match. Comparing the sample length statistics in the raw and composite files; the two total length values should match and the mean composite interval should be one. No discrepancies were identified during the compositing process Statistical Analysis Statistical analysis was carried out on the composited dataset for gold and density. Gold Summary statistics of composites for gold for the two mineralized domains (the main Boungou Shear Zone and the minor footwall mineralization) are presented in Table 14.1 and log histograms and log probability plots are presented in Figure The statistics show that the Natougou mineralization has a strongly positively skewed gold grade distribution with a high coefficient of variation ( CV = ratio of the standard deviation to the mean). Due to the strongly skewed nature of the gold grades (CV>2) in both the main mineralized domain and the waste domain, Snowden elected to use multiple indicator kriging ( MIK ) to estimate the block gold grades. Snowden Mining Industry Consultants

206 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page 14.6 Table 14.1 Summary Statistics for Gold Composites Domain (MINZONE) Main Mineralization (1000) Minor footwall Mineralization (2000) Waste (0) Number of samples 5, ,196 Minimum Maximum Mean Standard deviation CV Variance Skewness Percentiles: 10% % % % % % % % % % % % Snowden Mining Industry Consultants

207 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page 14.7 Figure 14-3 Log histogram and log probability plot for gold for main mineralized domain (top) and minor footwall mineralized domain (bottom) Log Histogram for AU M M Points: 5993 Mean: 3.36 Std 9.55Dev: Variance: CV: Skewness: Maximum: 75%: % 0.87 (median): 25%: Minimum: AU M M 83 Points: Mean: Std Dev: Variance: CV: 2.08 Skewness: Maximum: 75%: % 1.28 (median): 25%: Minimum: Source: Snowden Snowden Mining Industry Consultants

208 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page 14.8 Density As discussed in Section 11.5, bulk density measurements from core were completed using the Archimedes method. Whilst the samples are limited in terms of sample size (typically only 10 cm to 15 cm lengths are measured), a total of 23,843 measurements have been collected. A visual assessment of the density data shows areas of lower and higher density, with continuity between adjacent drillholes. As such, Snowden elected to estimate the bulk density using ordinary kriging ( OK ), with unfolding, to ensure that the local variations in the core density measurements are reflected in the resource model. Summary statistics are provided in Table A box and whisker plot of the density samples coded by the mineralized domain and oxidation state is presented in Figure The statistics show that the density samples of the Natougou mineralization are near normally distributed. Table 14.2 Summary Statistics for bulk density samples 1 Statistic Waste Domain MINZONE=0 Main Mineralized Domain MINZONE=1000 Minor footwall Mineralization MINZONE=2000 SOX MOX WOX FR SOX MOX WOX FR FR Samples , , Minimum Maximum Mean Standard deviation SOX = strongly oxidized; MOX = moderately oxidised; WOX = weakly oxidised; FR = fresh Snowden Mining Industry Consultants

209 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page 14.9 Figure 14-4 Box and Whisker Plot of Density Measurements 1 40 (19606) Box-and-Whisker Plot Group SG [SG > 0] (558) 1040 (3121) 20 (423) (23) M M (25) 50 M 2040 (31) SG25-75 % M SG25-75 % M (54) M M M 1010 (2) 50 M SG25-75 % SG25-75 % 1.5 SG25-75 % 1.5 SG25-75 % SG25-75 % SG25-75 % 50 SG25-75 % Source: Snowden Top Cut Analysis No top cutting of composite gold grades was performed for the multiple indicator kriged grade estimates. 1 First four boxes show the SOX, MOX, WOX and FR samples from the waste domain. The fifth to eighth boxes show the same for the main mineralized domain. The last box is FR samples from the footwall mineralized domain. Snowden Mining Industry Consultants

210 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page A comparative ordinary kriged gold grade estimate was completed for validation and comparative purposes. Due to the skewed distribution and high CV, for the ordinary kriged estimate (only), a top cut of 30 g/t Au was applied. The top cut value was based on an assessment of the histogram and log probability plot for the main mineralized domain. A 30 g/t Au top cut value equates to approximately the 98 th percentile, with a total of 106 (approximately 1.8% of the distribution) composites affected. The application of the top cut reduces the mean grade of the main mineralized domain from 3.36 g/t Au to 2.84 g/t Au and reduces the CV from 2.84 to Unfolding Introduction Variograms were generated to assess the grade continuity of the various elements and as inputs to the kriging algorithm used to interpolate grades. Snowden Supervisor software was used to generate and model the variograms. Snowden elected to use the Datamine unfold process to address the impact of the variable dip on the modelling of variograms and the estimation of grades. Unfolding improves the grade estimation process as it transforms the sample coordinates into their original pre-folded state. This allows variogram analysis and grade estimation to be carried out using the pre-folding coordinates, which are then converted back to the folded (local) coordinate system. The unfolding process results in more samples being available for variogram modelling and grade estimation than would have been the case if standard resource estimation methods were used Principles of Unfolding The basics of the unfolding approach are shown in Figure 14-5, which shows a typical cross section with folded strata. The folded units (expressed as strings) are divided into consistent units by means of structural links; each of these units has similar dip, strike and true width characteristics. Structural links are also applied along strike and within the section plane. In areas of complex folding, more links and more structural zones may be required. The unfolding process then generates an unfolded coordinate system ( UCS ), comprising A, B and C coordinates. The A coordinate pertains to the distance across strike in the true thickness (width) direction, the B coordinate represents the down dip direction, and the C coordinate represents the along strike direction (between cross-sections). By making this coordinate transformation, both the informing sample data and the discretization points within the blocks to be estimated are converted into unfolded space. Thus the spatial continuity may be determined in the plane of the fold rather than using geometrical distances. Grade interpolation also takes place in this unfolded plane and the resultant grades are then mapped back into the real space block model. Snowden Mining Industry Consultants

211 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure 14-5 Schematic Illustration of the unfolding process in Datamine Source: Snowden Creation of Strings for Unfolding Control strings, defining the upper and lower contacts (hangingwall and footwall) of the deposit to be unfolded, are required for the unfolding process. The control strings were created on a sectional basis using a set of vertical sections orientated perpendicular to the strike of the stratigraphy around the curve of the fold. The sections are located approximately 40 m to 80 m a part in Figure The distance between the hangingwall and footwall strings was set to ensure that all of the mineralized horizons and the immediate surrounding host rocks were enclosed by the two sets of strings. These control strings define the UCSB (down dip) direction. Snowden Mining Industry Consultants

212 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure 14-6 Plan View of the Unfold Control Strings Source: Snowden Structural links (strings) were applied to the control strings as follows: Two point strings were snapped between the hangingwall and footwall strings within each section plane to define the true thickness orientation (UCSA). These strings were placed at the start and end of each pair of sectional strings Strings were snapped along the first and last points of each hangingwall and footwall string to define the along strike direction (UCSC). The final set of unfold strings are displayed in Figure The Datamine Unfold process uses the strings and de-surveyed drillhole data to produce an unfolded drillhole file with three new coordinate fields, labelled UCSA, UCSB and UCSC respectively. Snowden Mining Industry Consultants

213 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure 14-7 Isometric view of the unfold strings looking from the south-southeast N Source: Snowden The variograms were modelled using the unfolded coordinate fields instead of the original X, Y and Z fields. The three axes were assigned to the Cartesian grid used by Snowden Supervisor software as listed in Table Effectively the variograms are modelled within the dip plane. The results of the unfolding are shown in Figure Table 14.3 Relationship between the variogram modelling software coordinate fields and the unfold coordinate fields Cartesian (variogram) axis UCS axis Description X UCSB Down dip Y UCSC Along strike Z UCSA True thickness Snowden Mining Industry Consultants

214 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure 14-8 Comparison between the original data (left) and unfolded data (right) 1 Source: Snowden 14.4 Variography The variograms were modelled for gold and density using Snowden Supervisor software in unfolded space Gold Indicator variograms were generated for the main mineralized domain (MINZONE=1000). Due to the low number of samples within the minor footwall mineralization (MINZONE=2000), variograms were not modelled for this domain. The indicator variogram models from the main mineralized domain were applied to the minor footwall mineralization and waste domains, with the threshold grades adjusted for each domain distribution in Table Top row = plan; bottom row = vertical section across the fold hinge Snowden Mining Industry Consultants

215 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Table 14.4 Indicator variogram thresholds and variogram model mapping Threshold percentile Main mineralized domain (MINZONE=1000) Au (g/t) Variogram reference number Minor footwall mineralized domain (MINZONE=2000) Au (g/t) Variogram reference number Waste domain (MINZONE=0) Au (g/t) Variogram reference number 10% % % % % % % % % % % % Variograms were modelled based on the following general approach: The drillhole composites were unfolded and modelled using the UCS coordinate fields. All variograms were standardized to a sill of one. The nugget effect was modelled from the true downhole variogram. Variograms were modelled using two nested, spherical structures. The variograms were evaluated using indicator variograms for a total of 12 grade thresholds. Variograms were only modelled for the main mineralized domain. The maximum and intermediate directions of continuity were generally aligned with the overall strike (UCSC) and down dip (UCSB) directions respectively. The minor direction of continuity was aligned in the true thickness direction (UCSA). The variogram directions did not change for different thresholds (i.e. no rotating anisotropy). The indicator variogram models for the main mineralized domain are summarized in Table 14.5 and the median indicator (i.e. 50 th percentile) variogram model is shown in Figure Snowden Mining Industry Consultants

216 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Table 14.5 Indicator variogram models for the main mineralized domain Element MINZONE Threshold Variogram Rotation angles (unfolded) 1 st Spherical variogram structure 2 nd Spherical variogram structure reference Nugget Range Range Range Range Range Range Percentile Au (g/t) number UCSA UCSB UCSA Sill Sill UCSA UCSB UCSC UCSA UCSB UCSC Au % Au % Au % Au % Au % Au % Au % Au % Au % Au % Au % Au % \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

217 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure 14-9 Median (50%) indicator variogram model Gamma Pair Counts (0.250) % (0.87) Indicator Variogram for AU Downhole La % (0.87) Indicator Variogram for AU Direction 1: 00--> La Sph( 0.24, 10 ) Sph( 0.26, 500 ) Sph( 0.36, 4 ) Sph( 0.34, 50 ) N( 0.4 ) N( 0.4 ) Sample Separation (m) Sample Separation (m) % (0.87) Indicator Variogram for AU Direction 2: 00--> La % (0.87) Indicator Variogram for AU Direction 3: 90--> La Sph( 0.26, 100 ) Sph( 0.26, 6 ) Sph( 0.34, 30 ) Sph( 0.34, 4 ) N( 0.4 ) 0.4N( 0.4 ) Source: Snowden Sample Separation (m) Sample Separation (m) 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

218 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Upper and lower tail modelling The upper and lower tails of the gold grade distributions, above the 99 th percentile and below the 10 th percentile, were modelled using a hyperbolic and power model respectively. Grades were modelled up to the maximum composite grade for the domains, as per Table The upper and lower tail model parameters are summarized in Table 14.6 for each domain. Table 14.6 Distribution tail modelling Domain (MINZONE) Tail Model type Model parameter 0 Lower Power Upper Hyperbolic Lower Power Upper Hyperbolic Lower Power Upper Hyperbolic In situ bulk density Variograms for density were generated using the UCS coordinates, for the main mineralized domain (MINZONE=1000) and waste domain (MINZONE=0). Due to the low number of density samples within the minor footwall mineralization (MINZONE=2000), the variogram from the main mineralized domain were used. A normal scores transform was used, with the model sill values back-transformed. The variogram models are presented in Table Similar to the gold grade continuity, the maximum and intermediate directions of continuity were generally aligned with the overall strike (UCSC) and down dip (UCSB) directions respectively. The minor direction of continuity was aligned in the true thickness direction (UCSA) \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

219 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Table 14.7 Density back-transformed variogram model parameters Element MINZONE Rotation angles (unfolded) 1 st Spherical variogram structure 2 nd Spherical variogram structure 3 rd Spherical variogram structure Nugget Range Range Range Range Range Range Range Range Range UCSA UCSB UCSA Sill Sill Sill UCSA UCSB UCSC UCSA UCSB UCSC UCSA UCSB UCSC Density Density \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

220 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Block modelling and grade estimation Kriging neighbourhood analysis A kriging neighbourhood analysis ( KNA ) was performed using Snowden Supervisor software to optimize various kriging parameters, based on the median (50 th percentile) indicator variogram for gold. The KNA assesses the impact of the kriging parameters on the kriging efficiency and slope of regression statistics. The main aim of a KNA is to assess the level of conditional bias (i.e. degree of over-smoothing) induced by various kriging parameters such as the parent block size, number of informing samples and search ellipse radii. Snowden used the results of the KNA to verify the choice of parent block size, number of informing samples and the search ellipse radii. Based on the KNA results, along with consideration of the geometry of the mineralization, the following parameters were selected: Parent block size of 20 me by 20 mn by 1 mrl. A minimum of 10 samples and maximum of 24 samples for the initial search pass. Search ellipse radii of 5 m in the true thickness direction (UCSA) by 60 m along strike (UCSC) by 40 m down dip (UCSB) for the initial search pass Volume model construction The block model extents, along with parent and sub-cell sizes are listed in Table Table 14.8 Block model prototype settings Model setting X Origin Y Origin Z Origin Maximum X Maximum Y Maximum Z Parent cell size X Parent cell size Y Parent cell size Z Minimum cell size X Minimum cell size Y Minimum cell size Z Value 325,000 me 1,326,200 mn 0 mrl 328,000 me 1,329,000 mn 325 mrl 20 m 20 m 1 m 5 m 5 m 0.5 m 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

221 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Block model coding The block model was coded based on the wireframe surfaces of the gold mineralization and oxidation zones, along with the topographic surface and surfaces defining the western hangingwall mineralization. Mineralization The gold mineralization was coded using a field called MINZONE. Field codes are summarized in Table An example cross-section showing the MINZONE field coding is presented in Figure Table 14.9 MINZONE field coding Field Value Description MINZONE 0 Un-mineralized MINZONE 1000 Main mineralized zone MINZONE 2000 Minor footwall mineralization Figure Example oblique section (looking northwest) showing MINZONE field coding Source: Snowden 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

222 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Oxidation The oxidation zone was coded using a field called OXIDE. Field codes are summarized in Table An example cross-section showing the OXIDE field coding is presented in Figure Table OXIDE field coding Field Value Description OXIDE 10 Strongly oxidized zone OXIDE 20 Moderately oxidized zone OXIDE 30 Weakly oxidized zone OXIDE 40 Fresh Figure Example oblique section (looking northwest) showing OXIDE field coding Source: Snowden Western Hangingwall Mineralization The western hangingwall mineralization was coded in the block model primarily for internal reporting purposes, using a field called HWZONE. Field codes are summarized in Table An examplecross section showing the HWZONE field coding is presented in Figure \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

223 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Table HWZONE field coding Field Value Description HWZONE 0 Not within hangingwall mineralized zone HWZONE 1 Hangingwall mineralized zone Figure Example oblique section (looking northwest) showing HWZONE field coding Source: Snowden Granite Contact The granite contact which occurs on the south-western margin of the mineralisation was coded using a field called ROCK. The mineralization is interpreted to terminate against the granite contact. Field codes are summarized in Table An example-cross section showing the ROCK field coding is presented in Figure Table ROCK field coding Field Value Description ROCK 0 Undefined ROCK Western granite 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

224 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure Example oblique section (looking northwest) showing ROCK field coding Source: Snowden Search Neighbourhood Parameters A three-pass search strategy was utilized for all estimates with the same search neighbourhood parameters applied to all domains. Details of the estimation search neighbourhood parameters are presented in Table The number of samples per drillhole was limited to three to ensure that a reasonable number of holes were used to estimate each block (at least three drillholes for the first search pass for gold and two for density). Given the low variance of the density composite data, the minimum and maximum number of samples was reduced to five and 15 respectively, for the initial search. This was done to ensure that the local variations in density were reflected in the block estimates. In Snowden s opinion, using more samples would likely result in over-smoothing the density estimate \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

225 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Table Search neighbourhood parameters Estimate Rotation about UCSA Initial search pass Second search pass Third search pass Axis lengths (m) No. of samples Axis lengths (m) No. of samples Axis lengths (m) No. of samples UCSA UCSB UCSC Min. Max. UCSA UCSB UCSC Min. Max. UCSA UCSB UCSC Min. Max. Au Density \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

226 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Grade estimation methodology Due to the strongly skewed nature of the gold grades (CV>2) in both the main mineralized domain and the waste domain, Snowden elected to use MIK to estimate the block gold grades. The MIK estimate was compiled using a total of 12 grade thresholds, based on the population deciles (10%, 20%,, 80%, 90%) with additional thresholds at the 95%, 97.5% and 99% included to model the higher grade portion of the distribution. Datamine software was used to: Unfold the sample data. Estimate the gold grade (using unfolding) of the MINZONE domains using MIK using the AU field in the drillhole file. The results were written to a field called MIKAU. The MINZONE field was used to constrain the MIK Au grade estimation with hard boundaries between all domains. Estimate the bulk density (using unfolding) of the MINZONE domains using ordinary kriging using the SG field in the drillhole file. The results were written to a field called DENSITY. The MINZONE field was used to constrain the density estimation with hard boundaries between all domains. The estimated domains and the percentile bins used for each MIK estimate are listed in Table 14.4 above. All estimates were parent cell estimates. The POSTIK process in the GSLIB suite of software was used for post-processing the MIK output from Datamine to enable order relation corrections to be applied (see Deutsch and Journal, 1992, p77 for details) and to allow the skewed tails of the gold grade populations to be modelled and used as part of the estimation process. The upper tail was modelled with a hyperbolic function between the last indicator cut-off and the maximum composite grade, while the lower tail was modelled with a power function below the first indicator cut-off. The final MIK product was an e-type estimate (i.e. average grade of the block based on the MIK probability estimates) which was subsequently imported back into Datamine. No change of support was applied to the MIK estimates as only an e-type estimate was produced. Where no gold or density estimate could be made due to sparse data, a default value was applied, as per Table and Table The median value for each domain was used for the default gold grade due to the skewed nature of the grade distributions, whereas the mean was used for the density default values \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

227 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Table Default gold grade values for un-estimated blocks MINZONE Default MIKAU value (g/t Au) Table Default density values for un-estimated blocks MINZONE OXIDE Default DESNITY value (t/m 3 ) 0, 1000, , 1000, , , Model validation The estimates were validated using: A visual comparison of the block grade estimates and the drillhole composite data. Generation of east-west, north-south and horizontal trend plots of the estimates, declustered and naïve composite grades, along with the number of composite samples available. A global comparison of the average composite (naïve and declustered) and estimated grades. An ordinary kriged gold estimate (top cut to 30 g/t Au) as a comparison to the MIK gold estimates. Trend plots of the average naïve and declustered composite and model grades for gold and density are presented in Figure and Figure respectively. For Figure 14-14, the MIK Au estimate (black line) should be compared to the declustered composite data (blue line), while the OK Au estimate should be compared to the top-cut and declustered composite data (dashed purple line). A summary of the grade comparison statistics from the estimation domains is presented in Table An example cross-section is shown in Figure \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

228 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure Grade trend plots Au, main mineralized domain (source: Snowden) Naive Mean Declustered Mean Declustered Top-Cut (30.000) Sample Count Data: AU [ntholes_fcu.csv 2] Filters: MINZONE = 1000 Estimate Mean MIKAU [nt1511v2.dm] Filters: MINZONE = 1000, SV_AU >= 1 Density: DENSITY Estimate Mean AU_OK [nt1511v2.dm] Filters: MINZONE = 1000, SV_AU >= 1 Density: DENSITY Validation Trend Plot MINZONE = 1000, 40m X Slice Centroid (mx) Validation Trend Plot MINZONE = 1000, 40m Y Naive Mean Declustered Mean Declustered Top-Cut (30.000) Sample Count Data: AU [ntholes_fcu.csv 2] Filters: MINZONE = Estimate Mean MIKAU [nt1511v2.dm] Filters: MINZONE = 1000, SV_AU >= 1 Density: DENSITY Estimate Mean AU_OK [nt1511v2.dm] Filters: MINZONE = 1000, SV_AU >= 1 Density: DENSITY Slice Centroid (my) Validation Trend Plot MINZONE = 1000, 5m Z Naive Mean Declustered Mean Declustered Top-Cut (30.000) Sample Count Data: AU [ntholes_fcu.csv 2] Filters: MINZONE = 1000 Estimate Mean MIKAU [nt1511v2.dm] Filters: MINZONE = 1000, SV_AU >= 1 Density: DENSITY Estimate Mean AU_OK [nt1511v2.dm] Filters: MINZONE = 1000, SV_AU >= 1 Density: DENSITY Slice Centroid (mz) 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

229 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure Grade trend plots density, main mineralized domain (source: Snowden) 3.10 Naive Mean Declustered Mean Sample Count Data: SG [>= 0 ] [ntholes_fcu.dm] Filters: MINZONE = 1000 Validation Trend Plot MINZONE = 1000, 40m X Estimate Mean DENSITY [nt1511v1.dm] Filters: MINZONE = 1000, SV_DEN > 0 Density: Slice Centroid (mx) 3.05 Validation Trend Plot MINZONE = 1000, 40m Y Naive Mean Declustered Mean Sample Count Data: SG [>= 0 ] [ntholes_fcu.dm] Filters: MINZONE = 1000 Estimate Mean DENSITY [nt1511v1.dm] Filters: MINZONE = 1000, SV_DEN > 0 Density: Slice Centroid (my) 3.0 Validation Trend Plot MINZONE = 1000, 3m Z Naive Mean Declustered Mean Sample Count Data: SG [>= 0 ] [ntholes_fcu.dm] Filters: MINZONE = 1000 Estimate Mean DENSITY [nt1511v1.dm] Filters: MINZONE = 1000, SV_DEN > 0 Density: Slice Centroid (mz) 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

230 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Table Statistical validation comparing input sample data against block model estimates Grade MINZONE Statistic Naïve Sample data Declustered Block Model Block vs. naïve Difference Block vs. declustered Mean % -9% Au (MIK) 0 Au (MIK) 1000 Au (MIK) 2000 Density 0 Density 1000 Density 2000 Variance Maximum Minimum Mean % -9% Variance Maximum Minimum Mean % 22% Variance Maximum Minimum Mean % 0% Variance Maximum Minimum Mean % 0% Variance Maximum Minimum Mean % 0% Variance Maximum Minimum \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

231 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure Example oblique section (looking northwest) showing block MIK gold estimates Source: Snowden The conclusions from the model validation work are: Visual comparison of the model grades and the corresponding drillhole grades shows a good correlation. A comparison of the global drillhole and model domain grades for the MIK gold estimate within the main mineralized domain shows that the estimates are within 10% which is considered a good outcome. With the exception of poorly sampled regions, the grade trend plots show a good correlation between the patterns in the model cell grades (MIK) compared with the drillhole grades. The choice of MIK to estimate gold grades is validated by the top cut ordinary kriged estimate, which, while reasonable, does not appear to adequately reproduce the higher grade regions of the mineralization. The validation of the block density estimates shows a good correlation between the block estimates and the input sample data \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

232 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page A global change of support shows that the grade-tonnage curve from the block model matches reasonably well with the theoretical grade-tonnage curve based on the sample data at block support Mineral Resource classification The Natougou Mineral Resource estimate has been classified as a combination of Measured, Indicated and Inferred Resources in accordance with CIM guidelines. The Mineral Resource has been limited to within a pit shell provided by SEMAFO, based on a gold price of USD 1,400/oz and pit optimization parameters established as part of the Mineral Reserve estimation. The classification was developed based on an assessment of the following criteria: Nature and quality of the drilling and sampling methods. Drilling density. Confidence in the understanding of the underlying geological and grade continuity. Analysis of the QAQC data. A review of the drillhole database and the company s sampling and logging protocols. Confidence in the estimate of the mineralized volume. The results of the model validation. Results of a pit optimization completed by SEMAFO using a gold price of USD 1,400/oz. The pit shell parameters are detailed in Table \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

233 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Table Pit optimization parameters for Mineral Resource definition pit shell Parameter Value Physical inputs Mining dilution 18% Mining recovery 95% Mining slopes (overall) 60 Processing recovery 92% Financial inputs (US$) Gold price 1,400/oz Mining cost (waste) 2.33/t Mining cost (ore) 3.52/t Rehabilitation 0.10/t Processing cost 22.27/t General and administration costs 5.44/t Selling cost 0.16/g Royalties 5% Discount rate 0% The resource classification scheme adopted by Snowden for the Natougou Mineral Resource estimate is outlined as follows: Where the drilling density was approximately 20 me by 20 mn (or less), mineralization within the main mineralized domain (i.e. MINZONE=1000) was classified as a Measured Resource. Where the drilling density was approximately 40 me by 40 mn (or less), mineralization within the main mineralized domain or minor footwall mineralized domain (i.e. MINZONE=1000 or 2000) was classified as an Indicated Resource. Where the drilling density was greater than 40 me by 40 mn, the mineralization was classified as an Inferred Resource. The edge of the interpreted mineralization, where the model was deemed to be largely extrapolated beyond the data, and areas of mineralization outside the $1,400/oz pit shell, were not classified and as such do not form part of the stated Mineral Resource. 1 The economic and technical parameters used to generate the resource pit shell differ slightly from those established for the Mineral Reserves. The resource pit shell was generated purely for reporting purposes to define the open-pitable extents of the mineralisation. The minor variation in pit optimisation parameters is not considered material with respect to the Mineral Resource estimate \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

234 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page The western hangingwall mineralization (HWZONE=1) remains unclassified due to the unconstrained nature of the block grade estimates in this area, and as such does not form part of the stated Mineral Resource. The classification was recorded in the model using a field called RESCAT, which is described in Table The Natougou Mineral Resource classification scheme is depicted in Figure Table Resource classification model field values RESCAT Description 0 Not classified (waste) 1 Measured 2 Indicated 3 Inferred 4 Unclassified mineralization Figure Plan view showing Natougou Mineral Resource classification scheme Source: Snowden 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

235 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Mineral Resource reporting Cut-off grade The Natougou Mineral Resource has been reported above a 0.83 g/t Au cut-off grade. The cut-off grade is based on the optimization completed by SEMAFO using the parameters defined in Table Snowden notes that the Mineral Resource is relatively insensitive to the reporting cut-off grade at cutoffs below 1.0 g/t Au, in terms of contained metal Moisture All Mineral Resources have been reported on a dry tonnage basis Natougou Mineral Resource statement The Natougou Mineral Resource comprises 2.64 Mt at 2.42 g/t Au of Measured and Indicated Resources and 2.68 Mt at 3.99 g/t Au of Inferred Resources, reported above a 0.83 g/t Au cut-off grade. The Mineral Resource statement is detailed in Table Mineral Resources have been reported exclusive of Mineral Reserves. Table Natougou Mineral Resource as at January 2016, reported above a 0.83 g/t Au cut-off grade Classification Oxidation State Tonnes (Mt) Au g/t Ounces (koz) Measured Fresh Measured total Indicated Strongly oxidized Moderately oxidized Weakly oxidized Fresh Indicated total Measured + Indicated total Inferred Strongly oxidized Moderately oxidized Weakly oxidized Fresh Inferred total * Small discrepancies may occur due to rounding; Mineral Resources are reported exclusive of Mineral Reserves 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

236 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Snowden notes that the vast majority of the mineralization within the reserve pit has been converted into Mineral Reserves, as detailed in Section 15. Mineral Resources remaining outside of the converted Mineral Reserve within the designed open pit comprise 653 kt at 1.75 g/t Au (for 37,000 oz of contained gold) of Measured and Indicated Resources along with 3 kt at 1.76 g/t Au (150 oz) of Inferred Resources. The majority of the Mineral Resource remaining within the designed pit occurs at the upper and lower edges of the mineralization Grade-tonnage curve The Natougou resource estimate has been reported (exclusive of Mineral Reserves) at various cut-off grades, as detailed in Table A grade-tonnage curve is shown in Figure Table Grade-tonnage report at various cut-offs for Natougou resource estimate, as at January 2016 (exclusive of Mineral Reserves) Cut-off g/t Au Tonnes t Measured Indicated Inferred Au Ounces Tonnes Au Ounces Tonnes Au g/t oz t g/t oz t g/t Ounces oz 0 114, ,200 3,138, ,400 2,744, , , ,200 3,130, ,400 2,744, , , ,200 3,094, ,200 2,743, , , ,200 3,031, ,700 2,741, , , ,200 2,966, ,000 2,737, , , ,100 2,892, ,900 2,730, , , ,000 2,812, ,500 2,717, , , ,800 2,703, ,200 2,711, , , ,600 2,598, ,700 2,692, , , ,500 2,564, ,800 2,683, , , ,400 2,476, ,300 2,662, , , ,100 2,323, ,700 2,607, , , ,000 1,727, ,200 2,261, , , ,500 1,310, ,900 1,913, , , , , ,700 1,549, , , , , ,600 1,304, , , , , , , , , , , , , , , , , , , \16.04\5047-STY-001_0 Snowden Mining Industry Consultants

237 NI TECHNICAL REPORT - MINERAL RESOURCE ESTIMATES Page Figure Grade-Tonnage Curve 5047-\16.04\5047-STY-001_0 Snowden Mining Industry Consultants

238 NI TECHNICAL REPORT - MINERAL RESERVE ESTIMATES Page MINERAL RESERVE ESTIMATES 15.1 Introduction Mineral Reserves for the Natougou project were developed as part of a feasibility study concluded in February Mining at Natougou will be with a conventional open pit mining method, utilizing hydraulic excavators and 63 t capacity off-highway dump trucks. The majority of the excavated material will require drilling and blasting, and only the softer weathered material will be loaded directly by hydraulic excavators. The Mineral Reserves represent the economically mineable part of Measured and Indicated Mineral Resources and are presented in Table The Mineral Reserves are additional to the Mineral Resources. Table 15.1 Mineral Reserves Estimate as of 31 December 2015 West Pit South East Pit North East Pit Total tonnes Total Weathered Fresh Au grade Au metal tonnes Au grade Au metal tonnes Au grade Au metal Mt g/t Moz Mt g/t Moz Mt g/t Moz Proven Probable Total Proven Probable Total Proven Probable Total Proven Probable Total Notes: Cut-off grade of 0.92 g/t Au for weathered material and 1.07 g/t Au for fresh material based on total operating costs of 30 US$/t of ore, 96% Au metallurgical recovery for weathered ore and a variable recovery equation for fresh ore: recovery (%) = ( 1- ((Au* )/ Au )) *100, 1.49 USD /g Au treatment costs, royalty and penalties, and 99.93% payable gold Au price assumed is USD 1,100 per troy ounce Probable Mineral Reserves are based on Indicated Mineral Resources only Proven Mineral Reserves are based on Measured Mineral Resources only Tonnes and grades shown account for mining dilution at 18% and mining recovery at 95%. Dilution material averages 0.2 g/t Au The numbers may not add up due to rounding \5047-STY-001_B AMC Consultants

239 NI TECHNICAL REPORT - MINERAL RESERVE ESTIMATES Page Method To convert Mineral Resources to Mineral Reserves, the following process was applied to Measured and Indicated Mineral Resources, and is described in more detail in Section 16: Determination of cut-off grades based on metal prices and costs, processing costs and metallurgical recoveries, incremental ore mining cost, and general and administration costs. Selection of an appropriate mining dilution and mining recovery method to convert the resource block model into a mining block model. Definition of ultimate and interim pit shells using the Lerchs-Grossman algorithm as part of the Whittle 4X software. Design of ultimate and staged pit designs based on the selected pit shells, geotechnical and operational mining considerations. Derivation of a mine plan based on the pit designs, mining equipment operational constraints and productive capacities by material type to achieve mill feed targets and associated material movement. Derivation of mining costs based on the mine plan. Economic evaluation of the mine plan as presented in Section Mining Cut-off Grade Two gold cut-off grades were calculated based on the different mining costs and processing recoveries applicable to the weathered and fresh ore. The cut-off grades have been calculated at 0.92 g/t of gold and 1.07 g/t of gold for weathered and fresh ore respectively. AMC notes that the cut-off grades calculated follow SEMAFO s cut-off grade policy and represent the cut-off grades for low grade ore rehandled from stockpiles at treatment time. AMC is of the opinion that SEMAFO s cut-off grade policy is reasonable. Royalty payment is included in the treatment costs. The cut-off grades presented in Table 15.2 represent the mine diluted recovered grades \5047-STY-001_B AMC Consultants

240 NI TECHNICAL REPORT - MINERAL RESERVE ESTIMATES Page 15.3 Table 15.2 Cut-off grade input parameters and calculations for Mineral Reserves Incremental Ore Mining Costs Processing Parameters Net Revenue Gold Low grade cut-off grade Parameter Run of mine re-handle Processing Cost (including Tailings Storage Facility and sustaining capital) Technical Services and Mine Supervision GA Total Incremental Ore Mining Costs Unit US$/t Ore mined US$/t Ore mined US$/t Ore mined US$/t Ore mined US$/t Ore mined Weathered Open pit Fresh Yearly production tonnage Mt Gold Metallurgical Recovery % 96 (1-((Au x )/Au)) x 100 Gold Price US$/oz Au 1,100 1,100 Treatment costs U Payable Gold % 99.93% 99.93% Net Revenue Gold US$/oz Au 1, ,052.9 US$/g Au Au cut-off grade g/t Dilution and Mining Recovery Factors In the process of estimating the Mineral Reserves, dilution and mining recovery factors were applied to the Mineral Resource using the regularization process described in Section 16. The dilution and mining recovery factors are tabulated in Table The dilution grade averages 0.2 g/t Au. Table 15.3 Mining Dilution and Recovery Parameters Open pit Mining Value Recovery (%) 95 Dilution (%) Mine Design Three distinct pits were designed utilizing Whittle 4X optimization shells as a guide. The West pit has been subdivided into six stages, the North pit into two stages, and the South pit into two stages (refer to Section 16) \5047-STY-001_B AMC Consultants

241 NI TECHNICAL REPORT - MINERAL RESERVE ESTIMATES Page 15.4 Figure 15.1 presents the final three pits geometry used to report the Mineral Reserve. Figure 15.1 Final Pits Geometry 15.6 Items Impacting on the Mineral Reserves The Competent Person is not aware of any mining, metallurgical, infrastructure, permitting, or other factors that may materially affect the current Mineral Reserves \5047-STY-001_B AMC Consultants

242 NI TECHNICAL REPORT - MINING METHODS Page MINING METHODS 16.1 General description Natougou will be mined using a conventional open pit approach of drilling and blasting ore and waste rock, with material mined by hydraulic excavators loading into off-highway rear dump haul trucks. Near surface rock is highly weathered and may be free-dig. Ore will be hauled to a run-of-mine (ROM) ore pad located adjacent to the ore processing plant or dumped directly into the primary crusher. Waste will be hauled to ex-pit waste rock dumps. Waste rock will be backfilled into the mined-out pit void once mining has progressed to enable sufficient pit void to be available. The deposit will be mined from three distinct pits, approximately 580 m to 1,800 m long, 300 m to 500 m wide, 50 m to 80 m deep and approximately 50 m to 100 m apart. Haulage ramps will be 20 m wide at a 10% gradient, with ore and waste rock trucks using the same haul ramps Pit dewatering Dewatering of the pit area is required to maintain: Access to the mine face. Dry mining conditions for drill/blast and load/haul operations. Dry pit wall conditions for stability. Sources of water reporting to the pit area are: Groundwater. Rainfall. Specific geological structures, such as permeable faults and fracture zones intersecting the fresh basement rock and that are intersected by the pit, may yield sufficient water to be drained via a bore and pump. Golders (2015) hydrogeology report states that such structures may be possible but are unproven to date, and therefore no costs have been estimated for such situations. Pumps designed for the intense wet season storms are likely to have more than sufficient capacity for such zones if the water from these zones is allowed to drain into the pit. Drainage of such fractured zones by bores will more likely be required for stability reasons. The maximum groundwater inflow rate predicted is 2.75 ML/day, equating to 32 L/sec. This quantity is considerably less than the extreme daily rain event inflows that need to be planned for, and therefore will easily be able to be handled by the planned pumps. AMC Consultants

243 NI TECHNICAL REPORT - MINING METHODS Page 16.2 Following drainage of the high level aquifers, the groundwater seepage into the pit reduces to the order of 0.1 ML/day (equating to approximately 1 L/sec). It is expected that these levels of seepage are likely to evaporate prior to the water reaching the pit floor. For the potential maximum daily extreme rainfall value of 73 mm/day that will be planned for expected to occur once every 20 months (i.e. approximately once every two years) the resulting potential daily maximum water inflow into the pits is 82 ML. Surface water will be directed around the pit by a series of perimeters bunds and channels. This water would be directed to settlement ponds prior to release. Pumping of the in-pit water would be conducted with mobile pumps placed in temporary sumps at the lowest point of each pit. The water would be collected in more permanent sumps near the pit base before being pumped to the collection sump outside the pit perimeter. A maximum of six sump pumps will be required in 2021 when six mining areas are mined concurrently. Vertical lift pumps will move water from the permanent sumps to the collection sumps outside the pit perimeter bund. The basis of the selection of the strategy for pit dewatering is that the pit can be kept dry under average wet season conditions, and can recover from a major rain event (85th percentile daily maximum extreme rainfall) in a reasonable time without major loss of pit operating time (three days). Two aspects are therefore examined: extreme event (storm) management, and average annual water movements. The final pits and final potential drainage area for rainfall into the pit have been calculated based on the maximum footprint of the pits, expanded by 50 m. This area is approximately 1.5 km2. A schedule of annual potential drainage area for rainfall into the pit has been determined from the mining schedule. For the purposes of determining an approximate pumping cost (purchase and operating), a standardtype vertical lift pit pump has been selected specifically a Sykes XH150 pump, capable of pumping at 100 litres per second and with a lift height of 100 m. At maximum, six pumps will be required in Year 3 of operation when approximately 886 ML of water is projected to be pumped out (groundwater and surface water combined). AMC Consultants

244 NI TECHNICAL REPORT - MINING METHODS Page Geotechnical parameters Golder Associates (January 2015) conducted a geotechnical feasibility study of the Natougou Gold Project. The project is located in the Pretozoic age Diapaga greenstone belt within the Birimian Gold Province of West Africa. Mineralization is hosted within the Boungou Shear Zone, which is a shallow, relatively flat-lying healed shear zone that has a gentle anticlinal form with an axis that trends and plunges slightly to the northwest. The predominant rock type of the deposit is amphibolite with minor occurrences of diorite, granodiorite and rhyolite. Two alteration types, Boungou Pink Alteration and Boungou Bleached Zone, are present in the deposit. A northwest striking fault with southwest dip is reportedly present in the northwest portion of the deposit. Geotechnical data acquisition on the property took numerous forms: Geotechnical logging of core from nine newly drilled geotechnical boreholes. Geotechnical logging of core from six previously drilled resource boreholes. Laboratory-conducted strength tests on core samples obtained from the newly drilled geotechnical boreholes. Down-the-hole core orientations were conducted in some of the boreholes, and this information was used to produce stereoplots of discontinuity orientation to assist in kinematic analyses of pit slope angles. In addition to the geotechnical data acquisition program, hydrological data was gathered from six newly installed electronic data loggers recording water levels within standpipe piezometers constructed in boreholes Rock strength Core samples obtained from the nine geotechnical boreholes were sent to RockLab (South Africa) for Uniaxial Compressive and Direct Shear strength testing. The results of these tests are presented in Table Table 16.1 Laboratory Rock Strength Tests Rock Type Mean UCS (MPa) Peak Friction Angle ( ) Amphibolite (MAM) 198 Amphibolite Pink Alteration (BPK) Amphibolite Bleached Zone (BBZ) 137 Boungou Shear Zone (BSZ) Granodiorite (GGD) Diorite (GDI) 1, Rhyolite (VRH) 112 AMC Consultants

245 NI TECHNICAL REPORT - MINING METHODS Page 16.4 Based on these tests, a peak friction angle of 40 was adopted for pit wall stability assessment Geological Input into Geotechnical Model Geological model surfaces for the development of a geotechnical model included: Surface topography. Top of moderately oxidized rock. Top of weakly oxidized rock. Top of fresh rock. Top of BSZ. Base of BSZ. The majority of the deposit is expected to be composed of amphibolite with minor occurrences of rhyolite, granodiorite and diorite. The minor rocks are presumed to occur as intrusions. The top and base of the BSZ indicate that the BSZ is apparently offset by a northwest-striking, steeply-dipping fault in the northwest of the deposit. The presence, orientation and condition of any other faults or shear zones within the deposit are currently not known Geotechnical Model A geotechnical model was developed in order to determine slope angles, bench heights and berm/batter widths. The approach adopted to develop the model involved: Dividing the rock mass into sub-horizontal zones of differing rock mass strength based on the degree of weathering and field estimates of intact rock strength (from the geotechnical core logging programme). Assuming that there are no adversely located and orientated major geologic structures that may have an adverse effect on slope stability. Identifying geotechnical domains based on the orientation of geologic discontinuities. Precambrian crystalline bedrock comprising predominantly amphibolite is encountered across most of the project area. In many places, the uppermost section of the crystalline bedrock complex has been altered by deep weathering processes to form a distinct horizon of laterite that transitions to strongly weathered saprolite at depth. Underlying the saprolite is a zone of moderately to slightly weathered rocks termed saprock. Within the study area the saprolite and saprock are thought to form a composite aquifer in which most of the groundwater is contained. AMC Consultants

246 NI TECHNICAL REPORT - MINING METHODS Page 16.5 Field rock strength estimates across the deposit are identified as: Soil and completely weathered rock (R0 to R1): composed of clay, silty clay and gravelly clay with a typical thickness of 5 m to 10 m and a mean thickness of ~4 m. Highly to moderately weathered rock (R2 to R3): typically amphibolite with a thickness of 0 m to 20 m and a mean thickness of ~7 m. Moderately to slightly weathered rock (R4): typically amphibolite with a thickness of 0 m to 20 m and a mean thickness of ~11 m. Slightly weathered to fresh rock (R5 or R6): typically amphibolite with a top of fresh rock located about 5 m to 30 m below surface and with a mean depth of ~22 m to the top of the fresh rock. The rock mass strength for these zones was estimated using the Hoek-Brown failure criterion, with the intact rock strength derived from the RockLab test and field strength estimate results. Values of the Hoek-Brown material parameter, mi, were obtained from RocLab (RocScience software) output using field-determined values of the rock mass characterization (Geological Strength Index, GSI). These results are tabulated in Table Table 16.2 Typical strength estimates for rock mass zones Zone UCS (MPa) GSI mi Cohesion (kpa) R0 to R1 (soil or completely weathered) R2 to R3 (highly to moderately weathered) R4 (moderately to slightly weathered) R5, R6 (fresh rock) Angle of Friction ( ) (10 100) 75 (40 200) 185 ( ) 40 (30 50) 50 (40 80) 70 (40 90) 26 (25 29) 26 (25 29) 26 (25 29) 700 (200 1,100) 40 (35 51) 1,400 (900 3,000) 55 (48 61) 4,000 (1,600 10,000) 63 (58 66) Pit wall design The stability of the batters excavated in rock was assessed by kinematic analyses of the potential for planar, wedge and direct toppling failures. Flexural toppling was not considered as this failure mode requires the presence of persistent, steeply-dipping, closely-spaced parallel discontinuities such as bedding partings. The foliation partings observed on site do not appear to be steeply-dipping nor are they closely spaced. The potential for large-scale flexural toppling to develop about a joint set is considered to be low due to the expected wide spacing and assumed low persistence of the joints. AMC Consultants

247 NI TECHNICAL REPORT - MINING METHODS Page 16.6 Based on this stability assessment, and assuming a 20% probability of failure, the recommended slope design is presented in Table AMC considers a 20% probability of failure acceptable. Figure 16.1 presents the geotechnical sectors based on a preliminary pit designed in Table 16.3 Slope design guidelines for the AMC proposed final pit Material Batter Height (m) Batter Angle ( ) Berm width (m) Inter-ramp Angle ( ) Soil and extremely weathered rock 5 10 (1) Highly weathered rock Moderately weathered to fresh rock wall segments E3, E6 and W4 (2) Moderately weathered to fresh rock wall segments E1, E4 and W2 (2) Moderately weathered to fresh rock wall segments E2, E5, E7, W1 and W3 (2) The total thickness of soil and extremely weathered rock is expected to vary between 5 m and 10 m across the site. Refer to Figure 16.1 for pit wall segment locations Figure 16.1 Recommended pit slope design guidelines AMC Consultants

248 NI TECHNICAL REPORT - MINING METHODS Page 16.7 AMC notes that based on the sensitivity analysis undertaken on slope angle, and presented in Figure 16.1, the project economics are not materially sensitive to variations in slope angle Resource Model for Mining The Resource model nt1511v1_min.dm used as basis for mine planning was provided by Snowden as a sub-blocked model with parent cell size of 20 m in X (east) direction by 20 m in Y (north) direction by 1 m in Z direction, as described in Section The Resource model was regularized to produce a mining model suitable for mine planning with block dimension of 5 m in X (east) direction by 5 m in Y (north) direction by 2 m in Z direction Mining Method Load and Haul Mining will take place on 6 m benches. A backhoe configuration (PC1250 Komatsu excavators) will be used to mine in ore zones, ore/waste contact zones and free-dig the weathered waste. The weathered material represents approximately 7% of the total waste material. Face shovels (PC2000 Komatsu front shovels) will be used in areas of bulk waste rock. Waste rock can be mined by hydraulic excavators configured as front loading shovels or backhoes on bench heights to suit, but ore mining will be done using a backhoe configuration on 2 m benches. The average productivity of the PC1250 excavators in blasted ore is expected to be approximately 955 t/hr and 733 t/hr when free-digging oxide waste. The productivity of the PC2000 in fresh waste is expected to reach 1,641 t/hr. Dozing of the ore will be required to minimize dilution when mining to the footwall of the mineralization and in shallow dipping sections of the deposit. Mining of the ore should only take place during dayshifts so that the geology department can guide excavations to minimize dilution and mining loss. Hauling of ore and waste will be undertaken by 63 t off-highway trucks (Komatsu HD605) Drill and blast Although there may be some minor volumes of highly weathered material near surface that can be excavated without drill and blast (free-dig by hydraulic excavators), the majority of the deposit will require drill and blast. Blast hole drill rigs are expected to be down-hole hammer rigs and will drill on 6 m-high benches. Trials will determine whether drill and blast on 10 m-high benches in the bulk waste rock zones is beneficial. Production blast holes will be vertical and a 6 m bench will need single-pass drill holes 7+ m long, to account for sub-drill. Blast hole patterns are presented in Table AMC Consultants

249 NI TECHNICAL REPORT - MINING METHODS Page 16.8 Table 16.4 Drill and blast parameters Design Parameters Unit Oxide Waste Fresh Waste Fresh Ore Bench height (m) Blast hole diameter (mm) Bulk Explosive (type) HANFO HANFO HANFO Blast hole burden (m) Blast hole spacing (m) Explosive density (g/cc) Estimated Powder Factor (kg/m3) Despite the potential for free-digging in the oxide waste zone, AMC has assumed that up to 25% of the oxide waste could require blasting due to variability in the oxide to fresh contact zone definition. Explosives will be stored on site. An explosive magazine, ANFO storage facility and emulsion tanks will be supplied by SEMAFO. High shock-energy explosives, such as heavy ANFO (HANFO), will be used to fragment the highstrength rock. Monitoring of material movement within a blast will be done using locatable reference points, ranging from graduated poly tubing inserted down a blast hole to electronic monitoring devices. This will allow predictive modelling of the blast movement to be developed and allow a predicted location for the postblast position of the ore-waste rock interface. To manage and control ore movement during blasting, controlled blasting techniques will be adopted. It will be critical to avoid cratering during ore blasts to avoid excessive movement of ore. Generous stemming zones will be required and the minimum drilling diameter of the D25KS drills of 127 mm will be the largest that could be contemplated for ore blasts. Blasting against final walls is likely to require specialist techniques including trim blasting and pre-split of final pit walls. Pre-splitting will utilize small diameter holes and will need to be drilled at -10 degrees (back under the drill) and be 12 m to 13 m long on 1 m centres, every 12 vertical metres. Decoupled or decked explosive charges will be used for pre-split blasting along all final walls in fresh rock. Trial blasts during the operational phase will determine the final pre-splitting requirements. Trim blasting is expected to be required to buffer final walls against production blasts. However, trials of production and trim blasting will be completed to determine the need for additional blasting measures. AMC Consultants

250 NI TECHNICAL REPORT - MINING METHODS Page 16.9 In addition, there will be a specialized contracted-out RC drill rig, under the supervision of mine geologists, drilling 10 m to 15 m grade control holes, depending on the thickness of the ore zone. Any low-grade zones identified will be sampled by blast holes, with low-grade and high-grade parcels blocked-out, blasted using controlled blasting techniques, blast movement modelled and tracked, and selectively mined. Mining will progress in horizontal benches until the bench above the modelled ore-waste rock contact. At this point, RC grade-control drilling will confirm the location of the ore-waste rock boundary. Modelling of blast movements will allow the ore-waste rock post-blast position to be predicted. A buffer zone of dilution (0.5 m) will be added to allow for the inaccuracies in the system. The dipping nature of the deposit means that vertical distance to the top of ore will vary within a blast, from a full bench height down to zero. Blast boundaries will be cut short, if required, to preserve the ore/waste interface. Dozers will face up any waste rock blocks that are insufficient height for an excavator. Ore and waste rock will be drilled and blasted together to assist in blast movement control and ensure adequate explosive distribution in the ore Stockpile Rehandling During dayshifts, the majority of the ore will be dumped directly into the crusher or hauled onto the ROM stockpile for blending and storage for use during night shifts. Ore placed on the ROM pad will be loaded into the crusher using a front-end loader (WA 600) as required to meet grade-blending requirements at the mill during day shifts; on night shifts, the entirety of the ore will be sourced from the ROM pad and loaded with the front-end loader. The ROM pad will have a capacity of approximately 150 kt of ore, segregated into six fingers refer to Figure Ancillary Equipment AMC has determined ancillary equipment numbers as a relation to the total production fleet numbers and based on experience. Activities considered when estimating the auxiliary equipment fleet include: Clean-up of digging and drilling areas. Waste dump maintenance. Dust suppression and maintenance of haul roads. Servicing of the production fleet. Topsoil clearing. AMC Consultants

251 NI TECHNICAL REPORT - MINING METHODS Page The support fleet consists of two D8 dozers that will be used for clearing topsoil, forming waste dumps and stockpiles and creating the surface roads to the waste dumps and open pits at the start of the operation. During the production phase, the D8 dozers will provide support in the pits and at the waste dumps. In the rehabilitation stage, the D8 dozers are required to rehabilitate the open pits and disturbed areas, and re-shape the out-of-pit dumps. In addition, one wheel dozer (CAT 834) will be used to maintain the roads and loading areas around the pit and waste dumps. One grader (CAT16) is planned for road maintenance, blast pattern levelling and re-contouring of the rehabilitated areas. One water truck (45,000 L) will be used during the dry season, or as required during the wet season, for dust prevention. Equipment numbers reflect operating requirements; the mining contractor may keep additional pieces of key ancillary mining equipment on site to ensure that adequate backup is available Mine Design Open Pit Optimization Input Parameters AMC estimated ore and waste mining costs for weathered and fresh rock based on high-level assumptions and previous work conducted by Orbis, assuming an owner-operated operation. Mineralized material incremental mining cost was then calculated for input into the cut-off grade calculation. Processing parameters, including costs and metallurgical recoveries were provided by Lycopodium. General and administration (GA) costs, metal price and selling costs were provided by SEMAFO. In line with SEMAFO s cut-off grade policy at their Mana operation, AMC calculated two economic cutoff grades: a low grade cut-off grade and a high grade cut-off grade both varying by ore type. The high grade cut-off, shown in Table 16.5 along with underlying parameters, categorizes ore that is mined during the life of mine. The low grade cut-off, similarly presented in Table 16.6, defines the gold grade at which ore is stockpiled for reclamation at the end of mine life. AMC Consultants

252 NI TECHNICAL REPORT - MINING METHODS Page Table 16.5 Open Pit High Grade Ore Cut-off Calculation Ore Mining Costs Waste Mining Costs Ore Costs Processing Parameters Net Revenue Gold High grade cut-off grade Unit Weathered Open pit Ore Mining Costs (DB, Load, Haul, Support) US$/t Ore mined Grade Control US$/t Ore mined Run of mine re-handle US$/t Ore mined Sustaining capex US$/t Ore mined Total Ore Mining Costs US$/t Ore mined Waste Mining Costs (Drill, Blast, Load, Haul, Support) Waste rehabilitation US$/t waste mined US$/t waste mined Fresh Total Waste Mining Costs US$/t mined Open pit incremental mining costs US$/t Ore mined Processing Cost (including Tailings Storage Facility and sustaining capital) US$/t Ore mined Technical Services and Mine Supervision US$/t Ore mined G&A US$/t Ore mined Total Ore Costs US$/t Ore mined Yearly production tonnage Mt (1-((Au x Gold Metallurgical Recovery % )/Au)) x 100 Gold Price US$/oz 1,100 1,100 Treatment costs, royalty US$/g Payable Gold % 99.93% 99.93% Net Revenue Gold US$/oz 1, ,052.9 US$/g Au cut-off grade g/t AMC Consultants

253 NI TECHNICAL REPORT - MINING METHODS Page Table 16.6 Open Pit Low Grade Ore Cut-off Calculation Parameter Ore Costs Processing Parameters Net Revenue Gold Low grade cut-off grade Unit Open pit Weathered Fresh Open pit incremental mining costs US$/t Ore mined Processing Cost (including Tailings Storage Facility and sustaining capital) US$/t Ore mined Technical Services and Mine Supervision US$/t Ore mined G&A US$/t Ore mined Total Ore Costs US$/t Ore mined Yearly production tonnage Mt (1-((Au x Gold Metallurgical Recovery % )/Au)) x 100 Gold Price US$/oz Au 1,100 1,100 Treatment costs, royalty US$/g Au Payable Gold % 99.93% 99.93% Net Revenue Gold US$/oz Au 1, ,052.9 US$/g Au Au cut-off grade g/t Dilution and mining recovery estimation methodology The mineralization is not visually distinguishable from the surrounding waste and varies in thickness between 0.5 m and 10 m with a mean of 2.9 m (refer to Figure 16.2). Figure 16.2 Ore vertical thickness histogram Histogram of vertical thickness Frequency 16% 14% 12% 10% 8% 6% 4% 2% 0% Vertical thickness (m) AMC Consultants

254 NI TECHNICAL REPORT - MINING METHODS Page These characteristics led AMC to evaluate the projected dilution and mining recovery using a block model regularization method. To select the most economic flitch height, AMC evaluated three mining flitch heights suited to the mill production capacity and possible mining equipment: 1.5 m, 2 m and 2.5 m. To represent the size of the anticipated digging equipment and the proposed grade control drilling spacing, three regularized mining models were produced using a block size of 5 m in the X and Y direction and the selected flitch heights as Z dimension. Ore mining costs were adjusted by flitch height to reflect the mining equipment size and selectivity capabilities. Pit optimization and high level schedules were then produced using Whittle software. The indicative discounted cash flow generated was used to compare the benefits on the project economics of higher selectivity and increased ounces recovered against higher ore mining costs (refer to Table 16.7). Table 16.7 Flitch Height Selection Schedule Results Flitch height (m) Dilution (%) Ore loss (%) Ore tonnes (Mt) Mill Feed Au grade (g/t) Waste tonnes (Mt) Mining costs ($/t) Revenue (MUS$) Mining CAPEX (MUS$) Indicative Discounted Cashflow (MUS$) , , , Although the 1.5 m flitch height presents the highest indicative discounted cashflow, the 2 m flitch height was selected as being more practical considering operators skill levels experienced by SEMAFO in Burkina Faso. To reflect the anticipated mining method, a further step in the modelling of ore loss and dilution was introduced after the pit designs were generated. The majority of the ore will be mined to a gently dipping footwall (10 dip on average). The vertical extent of the ore will be delimited by grade control drilling to an accuracy of approximately 0.5 m. A surface representing the vertical limit of the bottom of the ore zone will be produced by site geologists. This surface will be used to adjust production hole lengths to minimize the amount of material blasted in the footwall and therefore minimize dilution. This method of minimizing dilution was modelled by incorporating a maximum footwall dilution of 0.5 m at the bottom of the pit designs and regularizing the Resource block model to the selected 2 m flitch height. This refinement in the modelling resulted in an overall dilution of 18% and ore loss of 5% based on the final pit designs. Dilution grade averages 0.2 g/t Au. Mining dilution and ore loss parameters are shown in Table Table 16.8 Mining Dilution and Ore Loss Parameters Open pit Mining Value Ore loss (%) 5 Dilution (%) 18 AMC Consultants

255 NI TECHNICAL REPORT - MINING METHODS Page The practical requirements to achieve the projected dilution include: Dedicated and experienced grade controllers and samplers. Fast turn-around for grade control assays. Mining of ore only on day shifts. Reverse circulation (RC) drilling for grade control. Blast control and monitoring of blast movement. Dozing of the footwall ore waste contact to maximize ore recovery in selected areas. Optimization results The sizes and shapes of the ultimate pits were obtained using the Lerchs & Grossmann (LG) algorithm as implemented in the Whittle 4X software. Slope angles for the pit optimization were based on R-Rev0 Pit Geotech Feasibility Study.pdf, Golder Associates, January 2015, and modified to take into account the placement of haulage roads. Ore was defined in Whittle 4X using the low grade cut-off grades presented in Table 16.5 of 0.92 g/t for oxide and 0.96 g/t for fresh ore, but all economic inputs follow Table As presented in Table 16.9 and Figure 16.3, pit shell 30 displays a suitable combination of project life and indicative cash flow and was selected as the ultimate shell. Pit shells 5, 12, 17 and 23 were selected as indicative pushbacks to guide the ultimate pit and phase design process. AMC Consultants

256 NI TECHNICAL REPORT - MINING METHODS Page Table 16.9 Pit Optimization Results Pit Shell Revenue Factor Total Ore Waste Total Rock Strip Ratio Processing Cost Mining Cost Revenue Total Cost Undiscounted Cash flow Discounted Best Cash Flow Discounted Worst Cash Flow Tonnes In-situ Au Rec Payable Tonnes Tonnes Au (Mt) (g/t) oz (Mt) (Mt) W:O ($M) ($M) ($M) ($M) ($M) ($M) ($M) , , , , , , , , , , , , , , , , , , , , , , ,017, , ,036, , ,062, , ,090, , ,099, , ,140, , ,162, , AMC Consultants

257 NI TECHNICAL REPORT - MINING METHODS Page Figure 16.3 Pit Optimization Results A sensitivity of the selected open pit optimization was conducted at ±10%, ±20% for processing costs, mining costs, dilution and metal price. Overall slope angle sensitivity of ±2 was also evaluated. The results are shown in Figure The pit size, ore tonnes and cash flow are most sensitive to metal price, followed by mining costs, processing costs, dilution and pit slope angle. AMC Consultants

258 NI TECHNICAL REPORT - MINING METHODS Page Figure 16.4 Pit optimization sensitivity analysis Pit design Pits were designed based on the selected pit optimizations. Three pits have been designed, namely the West pit, the North East pit and the South East pit. The West pit is the largest of the three pits measuring approximately 1700 m in length, 400 m in width and 100 m at its deepest point. The North East pit is 1100 m long, 300 m wide and 100 m deep. The South East pit is the smallest pit measuring 500 m long by 300 m wide by 95 m deep. Stage designs, final designs and representative sections through the main pits are presented in Figure 16.5 to Figure AMC notes that the pit floor is designed to follow the bottom of the mineralization. AMC Consultants

259 NI TECHNICAL REPORT - MINING METHODS Page Figure 16.5 Pit designs overview Figure 16.6 West pit design Stage 1 AMC Consultants

260 NI TECHNICAL REPORT - MINING METHODS Page Figure 16.7 West Pit Design Stage 2 Figure 16.8 West Pit Design Stage 3 AMC Consultants

261 NI TECHNICAL REPORT - MINING METHODS Page Figure 16.9 West Pit Design Stage 4 Figure West Pit Design Stage 5 AMC Consultants

262 NI TECHNICAL REPORT - MINING METHODS Page Figure West Pit Section A1 A2 Figure West Pit Section B1 B2 AMC Consultants

263 NI TECHNICAL REPORT - MINING METHODS Page Figure North East Pit - Stage 1 Figure North East Pit - Stage 2 AMC Consultants

264 NI TECHNICAL REPORT - MINING METHODS Page Figure North East Section A1 A2 Figure North East Section B1 B2 AMC Consultants

265 NI TECHNICAL REPORT - MINING METHODS Page Figure South East Pit Stage 1 Figure South East Pit Stage 2 AMC Consultants

266 NI TECHNICAL REPORT - MINING METHODS Page Figure South East Pit Section A1 A2 Figure South East Pit Section B1 B2 AMC Consultants

267 NI TECHNICAL REPORT - MINING METHODS Page A summary of the tonnes of ore and waste contained within the pit designs are presented in Table Based on the stated quantities and a yearly milling rate of 1.34 Mtpa of ore, the mining operations are expected to last for approximately 8 years, including the production ramp up at the commencement of the mine. The milling operation will extend for additional time at the end of mine life for processing of stockpiled low grade ore. Table Ore and waste tonnage within the three ultimate pit designs Total Ore Waste Total Strip Ratio tonnes tonnes tonnes Mt Mt Mt W:O North East Final South East Final West Pit Grand total Note: The numbers may not add up due to rounding 1 Excludes topsoil 16.7 Layout of other mining related facilities The general site layout is presented in Figure Two major out-of-pit dumps, the East dump and the West dump will store waste mined from the three pits to the estimated volume presented in Table The East dump is approximately 1500 m long by 350 m wide by 25 m high. The West dump is approximately 1900 m long by 770 m wide and 38 m high. In-pit dumps will be created in order to reduce hauling costs as the out-of-pit dumps grow in size and the mined-out voids can be safely backfilled. It is estimated that approximately 55% of the mined-out waste will be backfilled into the pits while the rest is stockpiled onto the out-of-pit dumps. Table Out-of-pit waste dump storage capacity Waste storage capacity M m³ West dump 25.2 East dump 6.4 Total 31.6 AMC Consultants

268 NI TECHNICAL REPORT - MINING METHODS Page For mine closure the waste dumps will be rehabilitated by re-shaping the dumps to a final slope, placing a layer of top soil and revegetating. Waste dump rehabilitation parameters are presented in Table Table Waste dump rehabilitation parameters Lift Height (m) 15 Berm between lifts (m) 8 Rill angle (degrees) 20 Overall slope as-built (degrees) 17.5 Approximately 30 cm of top soil will be removed from the surface of the footprint of the pits, mine roads and out-of-pit dumps and stored into approximately 2 m high stockpiles for future rehabilitation. Top soil stockpiles will be built to the west and east of the West and East dumps and be used to cover the waste dumps. The ROM stockpile has been designed to accommodate up to 150 kt of ore as contingency during the wet season to maintain feed in the event of flooding of the pits (equivalent to 41 days of milling). The ROM stockpile will also be used throughout the year to allow for blending of the different grades of ore to provide a constant feed grade to the plant for sustained periods of time which will assist in maximizing metallurgical recovery Strategic mining schedule The strategic mine plan was developed using Minemax Scheduler. The main inputs to the strategic mine plan include the mining model, the ten design phases waste dumping envelopes representing the maximum extent of the out-of-pit waste dumps, and in-pit waste dumps presented in Figure The strategic mine plan took into consideration the possibility to backfill mined-out phases in order to reduce haulage costs by considering haulage distances and time required to haul a block to all possible waste dumping destinations. The following main scheduling constraints were used in deriving the strategic mine plan: Achieve planned mill feed tonnage on a yearly basis. Pre-strip enough overburden and waste to enable continuous mill feed to the crusher. Produce a smooth total material movement and haul truck requirement. Limit vertical advance to eight 6 m benches per year per stage. AMC Consultants

269 NI TECHNICAL REPORT - MINING METHODS Page A series of strategic mine plans were developed to test project NPV on a pre-tax basis against different gold ounce production profiles and increasing plant throughput from 4,000 tpd to both 4,500 tpd and 5,000 tpd from Year 3 of production. Producing a levelled gold ounces profile and maximizing NPV for the project was achieved by applying a varying cut-off grade and optimizing waste placement on a period by period basis. The adoption of an elevated cut-off strategy is associated with higher total material movement, but results in a higher NPV compared to a strategy of feeding the mill at a constant cut-off grade. The cut-off grades used to segregate ore into grade bins (GB) are presented below: GB0 : 0.92 g/t 0.99 g/t GB1 : 0.99 g/t - 2 g/t GB2 : 2 g/t 2.5 g/t GB3 : 2.5 g/t - 3 g/t GB4 : 3 g/t - 5 g/t GB5 : > 5 g/t A strategic mine plan resulting in a production of approximately 225 koz of gold per year in the first years of production was selected for the Study as it produced a value close to the highest value case and was in line with SEMAFO s corporate guidelines. The strategic mining schedule was produced in yearly increments for the life-of-mine and is presented in Figure Pre-stripping will commence with the West starter pit (PITWA1), which displays the highest value per tonne of ore. Approximately 16 Mt of material are mined during the pre-stripping period (2018). Total material movement peaks at 21 Mt from 2019 to Mining starts in the South East pit from 2019 and the North pit in All pits are exhausted by Figure Pit contribution by year (tonnes mined) AMC Consultants

270 NI TECHNICAL REPORT - MINING METHODS Page Table Strategic schedule - tonne and grade summary by year Total Mill Feed Mt Waste Mined Mt Mill feed grade g/t Au Recovered grade g/t Au Au ounces produced koz 1, Total Mined Mt As presented in Figure 16.22, targeted mill feed is achieved on a yearly basis from 2020 to In 2026, mill feed mainly consists of low grade ore reclaimed from the long term stockpiles and stops after five months of production. Figure and Figure show the strategic mine plan stockpile balance and recovered gold ounces respectively. Figure Strategic mine plan mill feed Figure Strategic mine plan stockpile balance AMC Consultants

271 NI TECHNICAL REPORT - MINING METHODS Page Figure Strategic mine plan recovered gold ounces 16.9 Detailed schedule The detailed schedule was created in Geovia s MineSched scheduling software. This schedule used the strategic schedule as guide for cutback sequencing, total material movement, recovered gold ounces profile and waste dumping strategy. The detailed schedule was generated in quarterly periods for the pre-stripping phase and first three years of production, and in yearly increments thereafter. Three distinct pits were scheduled; the West pit (consisting of six stages), the North East pit (two stages) and the South East (two stages). Two out-of-pit waste dumps were scheduled, the west and east dumps, which contained three access ramps each. Eight in-pit waste dumps were also used in the schedule to reduce haul distance: in-pit waste dumps IPD1 to IPD5 lie in the West pit, IPD7 and IPD8 lie in the North East pit and IPD9 lies in the South East pit. Figure presents the naming convention followed for all pits and dumps. AMC Consultants

272 NI TECHNICAL REPORT - MINING METHODS Page Figure Site layout N 0 TSSP ODE TSSP The main schedule input parameters are listed below: Scheduling period in quarters initially, then yearly from year Pre-strip occurs in 2018; 2019 is the first year of plant feed. Target Mtpa plant feed with the following process plant production ramp-up: - 50% capacity in Year 1, Month 1-75% capacity in Year 1, Month 2-100% capacity in Year 1, Month 3 Target 225 koz recovered gold for first three years of plant feed. Wet season occurs in Quarter 3 of each calendar year where mining productivity is reduced to 75% of nominal capacity, based on SEMAFO s experience at the Mana mine. Haulage of ore and waste using Komatsu HD605 with nominal capacity of 63 t. Material loading conducted with PC1250 excavators in ore and PC2000 in waste, with additional capacity provided by a WA600 front-end-loader. AMC Consultants

273 NI TECHNICAL REPORT - MINING METHODS Page Schedule in 6 m bench height (ore will be mined in 3 m x 2 m flitches). Schedule a maximum 10 benches per pit per annum. Allow a minimum of 200 m of on-bench space before the mining of the next bench can be started. The detailed schedule aimed at: Following the strategic schedule. Minimizing pre-strip. Achieving a levelled total material movement and mining equipment numbers. Backfilling the pits where possible. Delaying low grade material fed to the plant until the end of mine life. Topsoil has been defined as the top 0.3 m layer of material present within the footprint of the pits, mining roads, and out-of-pit waste dumps. Weathered material corresponds to the top 5 m to 10 m of material below the top soil. Fresh material is defined as material below the weathered zone. Five ore grade bins were defined as presented in Table These grade bins were used to help achieve the required recovered gold ounces profile. In practice three separate long-term stockpiles will be required. Grade Bin 1 represents the low grade material (LG), which is generally stockpiled until the end of mine life to help achieve the target recovered gold ounces. Grade Bins 2 to 5 define high grade ore (HG). AMC Consultants

274 NI TECHNICAL REPORT - MINING METHODS Page Table Grade Bin definition Grade Bin Lower COG (Au g/t) Upper COG (Au g/t) (oxide) 1.07 (fresh) The economic COGs and the marginal COGs are presented in Table Table Low Grade and High Grade Cut-off Grades Weathering Low grade COG (Au g/t) High grade COG (Au g/t) Weathered Fresh The topsoil has been scheduled as needed and occurs in the pre-strip phase and A total of 1.3 Mt (683,000 m 3 ) will be removed to topsoil stockpiles behind the dumps (away from the pits). AMC Consultants

275 NI TECHNICAL REPORT - MINING METHODS Page Some important milestones of the mining schedule are presented in Table Table Schedule Milestones Date Milestone Mine haul roads construction starts Clearing of WA1 and ES1 cutbacks and west waste dump 2018 Q1 Topsoil clearing under west waste dump Top soil clearing and pre-strip starts in WA1 and ES1 ROM pad construction starts TSF wall construction starts First ore from ES1 mined 2018 Q4 Topsoil clearing under east waste dump First ore from WA1 mined Plant feed starts 2019 Q1 More waste placed onto TSF wall Topsoil and waste mining in WA2 and WA5 starts 2019 Q2 Plant feed at full capacity 2019 Q3 Topsoil and waste mining in WA6 starts 2019 Q4 ES1 complete 2021 Q1 Topsoil and waste mining in WA3, WA4, EN1 starts 2021 Q2 WA1 complete WA5 complete 2021 Q4 In-pit dumping starts WA2 and WA6 complete 2022 WA3 and WA4 complete 2023 Topsoil and waste mining in EN2 and ES2 starts 2024 EN1 complete 2025 ES2 and EN2 complete 2026 Plant feed completed using low grade stockpiles Rehabilitation commenced and completed Construction activity During the life of the mine, waste materials are diverted to the TSF wall construction as seen in Table AMC Consultants

276 NI TECHNICAL REPORT - MINING METHODS Page Table TSF Wall Construction Requirements Period Engineered waste rock required ( 000 m3) Scheduled waste rock ( 000 m3) The pre-stripping period is scheduled to start after the wet season. During this period, approximately 1,045,000 m3 of waste is used for the ROM pad construction. Haul roads will be made by clearing topsoil and sheeting with coarse road base sourced from the pit. Three long-term stockpiles will be created near the ROM pad with a capacity of 1.3 Mt for the largest and two additional 0.3 Mt stockpiles. Safety bunds will be required around all open pits and around the base of waste dumps to prevent access. Surface water drainage channels will be constructed for the starter pits and, where appropriate, for the final pits Schedule results The main outcomes of the detailed schedule are presented below: The detailed schedule needs to move more waste than the strategic schedule during the prestripping period to meet plant feed requirements on a quarter by quarter basis. Mining operations will take eight years, including the pre-strip period. Processing will last eight years, the last year of which will be stockpile re-handle only. It is expected that rehabilitation will take place throughout the mine life when the chance becomes available and during Year 8. No topsoil re-handle has been scheduled, but this movement has been included in the final cost model and truck equipment number calculations. Total annual ex-pit material movement peaks at approximately 21 Mtpa. AMC Consultants

277 NI TECHNICAL REPORT - MINING METHODS Page Recovered gold ounces profile follows the strategic schedule at approximately 225 koz for the first three years of mill production. Recovered gold ounces have been calculated using metallurgical recoveries based on a period by period mill feed grade. In-pit dumping starts in the last quarter of Year A maximum long term stockpile capacity of 1.67 Mt is required as seen in Figure The majority of this stockpile is made up of the lowest grade bin material, which is being stored for plant feed at the end of the mine life. It has been assumed that there will be no detrimental effect to the metallurgical recovery of the stockpiled ore. Additional work during the detailed planning phase will result in a smoothed production profile. Figure Closing stockpile balance by period Scen 52b - Stockpile Capacity Stockpile Cpacity (kt) 1,800 1,600 1,400 1,200 1, Q1 2018Q2 2018Q3 2018Q4 2019Q1 2019Q2 2019Q3 2019Q4 2020Q1 2020Q2 2020Q3 2020Q4 2021Q1 2021Q2 2021Q3 2021Q Bin1 Bin2 Bin 3 Bin 4 Bin 5 AMC Consultants

278 NI TECHNICAL REPORT - MINING METHODS Page Table Schedule summary Ex-Pit movement by year Total Tonnages ex-pit LG ore ex-pit kt ,379 LG Au grade ex-pit g/t HG ore ex-pit kt 40 1,265 1,308 1, ,188 HG Au grade ex-pit g/t Total ore ex-pit kt 73 1,714 1,712 2,121 1,054 1,262 1, ,567 Total Au grade ex-pit g/t Waste expit (excl topsoil) kt 17,235 19,341 19,750 18,550 19,454 16,061 15,462 1, ,806 Topsoil ex-pit kt Topsoil under dumps kt ,296 Total tonnes ex-pit kt 17,510 21,198 21,484 20,869 20,508 17,441 16,641 2, ,057 Total tonnes ex-pit + under dumps kt 18,182 21,822 21,484 20,869 20,508 17,441 16,641 2, ,353 Waste - oxide kt 2,429 1, , , ,179 Waste - fresh kt 14,807 17,498 19,229 15,965 19,179 14,574 15,422 1, ,627 Waste - topsoil (pits + dumps) kt ,979 Ex-pit strip ratio t:t Mining rate t/day 49,815 59,787 58,859 57,176 56,188 47,782 45,591 6,591 47,724 Volume ex-pit Volume LG ex-pit kbcm Volume HG ex-pit kbcm ,450 Volume total ore ex-pit kbcm ,267 Volume waste whole site kbcm 7,038 7,209 6,795 6,634 6,749 5,729 5, ,058 AMC Consultants

279 NI TECHNICAL REPORT - MINING METHODS Page Table Mill feed schedule by year Total Mill feed LG tonnes to mill kt ,379 LG Au grade to mill g/t HG tonnes to mill kt 1,075 1,283 1,185 1,336 1, ,188 HG Au grade to mill g/t Total tonnes to mill kt 1,256 1,343 1,343 1,343 1,347 1,343 1, ,567 Total Au grade to mill g/t Total recovered Au grade g/t Total recovered Au metal kg 6,996 7,032 7,107 4,498 4,893 3,360 2, ,856 Total recovered Au metal koz ,185 Mill feed breakdown Oxide ore tonnes to mill kt Oxide ore grade to mill g/t Oxide ore recovery % Fresh ore tonnes to mill kt 1,229 1,337 1,334 1,343 1,342 1,327 1, ,474 Fresh ore grade to mill g/t Fresh ore recovery % Measured Resource Category kt ,584 Indicated Resource Category kt ,144 1,284 1,312 1,297 1, ,983 Table Topsoil ex-pit and construction material by year Total Topsoil from under dumps kt ,296 Topsoil from pits kt Tonnes sent to ROM pad construction kt 2, ,581 Tonnes sent to TSF wall kt , , ,927 Volume sent to ROM pad construction kt 1, ,045 Volune sent to TSF wall kt ,881 AMC Consultants

280 NI TECHNICAL REPORT - MINING METHODS Page Table Waste movement by destination by year Total Topsoil stockpiles Mt West dump Mt East dump Mt IPD1 Mt IPD2 Mt IPD3 Mt IPD5 Mt IPD8 Mt IPD9 Mt Rompad construction Mt TSF wall Mt Total Mt Figure Annual recovered gold ounces 250 Scen 52b - Recovered ounces Recovered Gold (koz) AMC Consultants

281 NI TECHNICAL REPORT - MINING METHODS Page Figure Annual total ex-pit movement 25 Scen 52b - Total Ex-pit movement Ex-pit Tonnes (Mt) Figure Annual plant feed Plant Feed (Mt) Scen 52b - Plant feed End-of-year progression plots are provided in Appendix. AMC Consultants

282 NI TECHNICAL REPORT - MINING METHODS Page Production Schedule Risks The production schedule risks include: Inability to meet the first quarter plant feed At the end of the pre-strip year, a total of 72 kt has been stockpiled. The Quarter 1 plant feed requirement is 251 kt. Most of this plant feed will be direct ex-pit feed. AMC expects that there will be some floor stock readily available in the pits, in particular within pit ES1. AMC suggests that the first quarter is scheduled in monthly or even weekly increments to ensure adequate plant feed. Inability to achieve the target recovered gold ounces profile. Most of the high grade ore is extracted from pit WA1, which has the highest strip ratio. In order to minimize the pre-strip, pit ES1 is mined at the same time, however this pit has a lower average grade. The recovered gold ounces profile on a quarter by quarter basis is not smooth, but the annual recovered gold ounces profile is achieved. Good reconciliation data should be maintained to track the performance of the diluted mining model Experienced excavator operators and grade control teams will be required in order to minimize the dilution of the ore Inability to mine the target material during wet season. An approximate 25% reduction in ex-pit mining rate has been scheduled. Generally, most ore will tend to lie on the bottom benches of the pit. Good de-watering and bench management will be required to ensure that ore is exposed above potential flooded areas. Stockpile balances show that there will be sufficient material in stockpiles should de-watering be an issue, but most of this stockpile material will be of lower grade and the recovered gold ounces profile may not be met. In-pit dumping may not be able to start or occur as predicted. This will mean that more waste will be hauled to out-of-pit waste dumps. As out-ofpit waste dump space is generally not regarded as an issue, this should not be of high significance. AMC Consultants

283 NI TECHNICAL REPORT - MINING METHODS Page Equipment and Personnel Numbers Contractor personnel numbers were estimated for all mining department personnel, including management and supervision, technical staff, operators and maintenance working on site. No allowance has been made for additional personnel in other departments that may be required to support the mining operations. Instead an allowance of 0.5% has been added to the total contract labour costs to represent the additional charge. The open pit operators and maintenance personnel are planned to work a four panel, two shift roster. Features of this roster are: 24 hours per day, 365 days per year, two shifts per day 12 hours per shift 7 days on, 7 days off basis 15 days annual leave The above roster and mining schedules were used to derive peak equipment requirements. To estimate the haulage fleet, haul profiles were measured from each destination to each source on a period by period basis. The open pit primary equipment requirement at peak production is summarized in Table Table Open pit equipment requirements at peak production Equipment Main activity Number Komatsu PC 1250 (Excavator) Komatsu PC 2000 (Face shovel) Komatsu HD 605 (63 t haul truck) Ore mining (day shift), waste mining (night shift) Waste mining 2 Hauling of ore and waste 13 CAT D8 (dozer) Ancillary work 2 CAT 834 (wheel dozer) Ancillary work 1 CAT 16 (grader) Ancillary work 1 Water truck (45,000 L) Dust suppression 1 Atlas Copco DM25 (Production drill) Production drilling 6 Atlas CAT MD 5150C (presplit drill) Pre-split drilling and ore contact drilling 1 WA 600 (Wheel loader) ROM loader, waste loading support 1 2 AMC Consultants

284 NI TECHNICAL REPORT - MINING METHODS Page Total operator numbers required were calculated for the number of machines on site at any given time, taking into account rosters, sick and vacation leave. Equipment such as trucks, excavators, drills and dozers were considered to be manned at all times. The number of maintenance personnel was derived using a maintenance labour of 0.6 to 0.8 for the main pieces of mining equipment. The factor is based on the number of hours of maintenance to the number of hours of operation, applied to each piece of equipment. The factor is also based on manufacturer information and a maintenance efficiency ratio to reflect the skill level and experience of the workforce in the country. The total personnel required was estimated based on the production throughput of the operation and the equipment numbers. SEMAFO s expatriate personnel, contractor management and senior staff were assumed to work on a 35 days on and 28 days off fly-in fly-out basis. National technical staff will be working on a 10 days on, 5 days off rotation. Based on SEMAFO s experience in the country and availability of skilled labour, the number of expatriates in the owner s team has been kept to six persons in supervisory positions. Mining production and maintenance positions are expected to be filled by the local workforce. AMC assumed that the local workforce will be accommodated in nearby villages. Management staff, technical staff and expatriates will be accommodated in the camp built at the mine site. The yearly requirement for the open pit workforce peaks at 314 persons and is summarized in Table AMC Consultants

285 NI TECHNICAL REPORT - MINING METHODS Page Table Open pit workforce requirements Personnel Numbers Fixed - Production Contractor / Owner Roster Expat or National Mine Supervisor - Shift Owner 7d on/7d off N Mine Supervisor - Shift Contractor 7d on/7d off N Open Pit Manager Contractor 35d on / 28d off E Clerk Contractor 7d on/7d off N Training & OH&S Coordinator Contractor 10d on / 5d off N Mine Production L/Hand Contractor 7d on/7d off N General labourer Contractor 7d on/7d off N Dewatering Crew Contractor 7d on/7d off N Sub Total Fixed - Maintenance Maintenance Manager Owner 35d on / 28d off E Assistant Owner 7d on/7d off N Crane Operator Owner 10d on / 5d off N Auto-electrician Owner 7d on/7d off N Mechanic - Shift Owner 7d on/7d off N Welder Owner 7d on/7d off N Maintenance supervisor Owner 10d on / 5d off N Maintenance Manager Contractor 10d on / 5d off E Clerk Contractor 10d on / 5d off N Maintenance Superintendent Contractor 10d on / 5d off N Maintenance Warehouse Supervisor Contractor 10d on / 5d off N Store-person Contractor 10d on / 5d off N Maintenance Planner Contractor 10d on / 5d off N HV Electrician Contractor 10d on / 5d off N Maintenance supervisor Contractor 10d on / 5d off N AMC Consultants

286 NI TECHNICAL REPORT - MINING METHODS Page Personnel Numbers Contractor / Owner Roster Expat or National Tyre Fitter Contractor 7d on/7d off N Welder Contractor 7d on/7d off N Sub Total Fixed - Technical Services Chief Mining Engineer Owner 35d on / 28d off E Senior production superintendent Owner 35d on / 28d off E Mine Surveyor Owner 10d on / 5d off N National surveyor Owner 7d on/7d off N Surveyor - Technicians Owner 7d on/7d off N Mining technician Owner 7d on/7d off N Sub Total Fixed - Geology Chief Geologist Owner 35d on / 28d off E Chief Geostatitian Owner 35d on / 28d off E Senior Mine Geologist Owner 10d on / 5d off N Mine Geologist Owner 10d on / 5d off N Geology technicians Owner 7d on/7d off N ROM pad dispatcher Owner 7d on/7d off N Pit dispatcher Owner 7d on/7d off N Geotechnicians Owner 7d on/7d off N Sub Total Mining Operations - Variable Numbers National Truck Operator Contractor 7d on/7d off N National Excavator operator Contractor 7d on/7d off N National Ancillary Operator Contractor 7d on/7d off N National Drill Operator Contractor 7d on/7d off N National Fitter Contractor 7d on/7d off N Shot Crew Contractor 7d on/7d off N AMC Consultants

287 NI TECHNICAL REPORT - MINING METHODS Page Personnel Numbers Contractor / Owner Roster Expat or National Sub Total Total Personnel - Expat E Total Personnel - National N Total Personnel AMC Consultants

288 NI TECHNICAL REPORT - MINING METHODS Page Additional Information Figure End of year face position Year 2018 N ODE TSSP3 275mRL TSSP4 ROM AMC Consultants

289 NI TECHNICAL REPORT - MINING METHODS Page Figure End of year face position Year 2019 N ODE 275mRL TSSP4 ROM AMC Consultants

290 NI TECHNICAL REPORT - MINING METHODS Page Figure End of year face position Year 2020 N ODE 275mRL TSSP4 ROM AMC Consultants

291 NI TECHNICAL REPORT - MINING METHODS Page Figure End of year face position Year 2021 N ODE TSSP4 ROM AMC Consultants

292 NI TECHNICAL REPORT - MINING METHODS Page Figure End of year face position Year 2022 N ODE TSSP4 ROM AMC Consultants

293 NI TECHNICAL REPORT - MINING METHODS Page Figure End of year face position Year 2023 N ODE ROM AMC Consultants

294 NI TECHNICAL REPORT - MINING METHODS Page Figure End of year face position Year 2024 N ODE ROM AMC Consultants

295 NI TECHNICAL REPORT - MINING METHODS Page Figure End of year face position Year 2025 N ODE ROM AMC Consultants

296 NI TECHNICAL REPORT - RECOVERY METHODS Page RECOVERY METHODS 17.1 Process Design The process plant design for the Project is based on a robust metallurgical flowsheet designed for optimum recovery with minimum operating costs. The flowsheet is based on well proven unit operations in the industry. The key criteria for equipment selection are suitability for duty, reliability and ease of maintenance. The plant layout provides ease of access to all equipment for operating and maintenance requirements whilst maintaining a layout that will facilitate construction progress in multiple areas concurrently. The key project design criteria for the plant are: Nominal throughput of 4,000 tpd ore. Process plant availability of 92% supported by the design of the crushing plant, standby equipment in critical areas and on-site hybrid HFO/diesel generator power supply. Sufficient automated plant control to minimize the need for continuous operator interface and allow manual override and control if and when required. Study design documents have been prepared incorporating engineering design criteria and key metallurgical design criteria derived from the results of the metallurgical testwork Selected Process Flowsheet The treatment plant design incorporates the following unit process operations: Single stage primary crushing with a jaw crusher to produce a crushed product size of 80% passing (P 80 ) of 133 mm. Mill feed surge/overflow bin that overflows to a 8,000 tonne stockpile to provide 48 hours of capacity. During extended periods of up to two days for primary crusher equipment maintenance, ore from the stockpile will be reclaimed by an excavator or dozer to feed the grinding circuit. The grinding circuit is a SATMC type, which consists of a closed circuit semi-autogenous (SAG) mill, pebble crusher for SAG mill discharge oversize and a closed circuit tower mill to produce a P 80 grind size of 63 µm. A gravity gold recovery circuit. Lycopodium Minerals Canada

297 NI TECHNICAL REPORT - RECOVERY METHODS Page 17.2 Hydrocyclones are operated to achieve a hydrocyclone overflow slurry density of 27% solids to promote better particle size separation efficiency. Subsequently, a pre-leach thickener is included to increase slurry density to the leach circuit, minimise leach tank volume requirements and reduce overall reagent consumption. Leach circuit with five tanks to achieve the required 36 hours of residence time at nominal plant throughput. Carbon-in-pulp (CIP) carousel circuit consisting of seven stages is a carbon adsorption circuit for recovery of gold dissolved in the leaching circuit. AARL elution circuit with gold recovery to doré. The circuit includes an acid wash column to remove inorganic foulants from the carbon with hydrochloric acid. Carbon regeneration kiln to remove organic foulants from the carbon with heat. Tailings thickener to increase slurry density for water recovery prior to tailings discharge to the tailings storage facility. An overall process flow diagram depicting the unit operations incorporated in the selected process flowsheet is presented in Figure The key issues considered in process and equipment selection are outlined in the following section. Lycopodium Minerals Canada

298 NI TECHNICAL REPORT - RECOVERY METHODS Page 17.3 Figure 17.1 Overall Process Flow Diagram Lycopodium Minerals Canada

299 NI TECHNICAL REPORT - RECOVERY METHODS Page Key Process Design Criteria The key process design criteria listed in Table 17.1 form the basis of the detailed process design criteria and mechanical equipment list. Table 17.1 Summary of Key Process Design Criteria Units Design Source* Plant Throughput tpd 4000 Client Head Grade g Au/t 4.15 Client Overall Gold Recovery1 % 92.9 Calc from Testwork Crushing Plant Availability % 75.0 Lycopodium Plant Availability % 92.0 Client Crushing Work Index (CWi) kwh/t 10.0 Testwork Bond Rod Mill Work Index (RWi) kwh/t 21.9 Testwork Bond Ball Mill Work Index (BWi) kwh/t 18.9 Testwork SMC Axb Consultant Bond Abrasion Index (Ai) incl HW & FW dilution Consultant Grind Size µm 63 Testwork Gravity Recovery Au % 46.4 Testwork Leach Thickener Solids Loading t/m2.h 1.5 Testwork Leach Circuit Residence Time hrs 36 Testwork Leach Slurry Density % solids (w/w) 50 Testwork Number of Leach Tanks 5 Lycopodium Number of Adsorption Tanks (Stages) 7 Lycopodium Sodium Cyanide Addition kg NaCN/t ore 0.45 Testwork Lead Nitrate Addition Kg PbNO 3/t ore 0.05 Testwork Dissolved Oxygen Level in Leach ppm >20 Testwork Hydrate Lime Addition 3 kg/t 0.32 Testwork Elution Circuit Type AARL Agreed Elution Circuit Size t 2.5 Agreed Frequency of Elution strips / week 7.1 Lycopodium Tailings Thickener Solids Loading t/m2.h Testwork Tailings Discharge Slurry Density % solids (w/w) 62-65% Testwork 1. At design head grade of 4.1 g/t Au. 2. Design A x b value derived from the 85 th percentile ranking of specific energies determined for each individual ore type. 3. Lime addition based on 75% CaO. Lycopodium Minerals Canada

300 NI TECHNICAL REPORT - RECOVERY METHODS Page Process and Plant Description Introduction The Natougou mineralization is predominantly hosted in quartz veins and a significant quantity of the gold occurs as visible, free gold. Sulphide minerals comprise pyrrhotite, arsenopyrite, pyrite, and minor chalcopyrite. The Natougou quartz vein deposit is considered free milling. A significant component of the gold is amenable to gravity recovery and consequently the proposed process facility will consist of the following process areas: Primary crushing and coarse ore storage. Grinding, utilizing a SAG Mill, Tower Mill and Pebble Crusher (SATMC) circuit. Leach and CIP Carousel circuit. Gold recovery and carbon handling circuit (consisting of a cold acid wash followed by a split, Anglo American (AARL) elution circuit and horizontal carbon regeneration kiln). High density tailings handling and disposal. The Natougou project process plant will be designed to process 4,000 tpd (1.34 Mtpa) with an average gold head grade of 4.15 g/t. The primary crushing circuit will operate for 365 days per annum, for a nominal 24 hrs per day. On this basis, and at a design operating availability of 75%, the crushing circuit will operate for a nominal 6,570 hrs per annum, i.e. 204 t/h. Downstream of the crushing circuit, the comminution, leach, adsorption and tails thickening circuits will operate for 365 days per annum, for a nominal 24 hrs per day. On this basis, and at a design operating availability of 92%, these circuits will operate for a nominal 8,059 hrs per annum, i.e. 167 t/h. The gold leach circuit will be comprised of five tanks, providing a nominal residence time of 36 hrs. The CIP carousel adsorption circuit will be comprised of seven adsorption contactors and will operate with a carbon concentration of 50 g/l. The adsorption contactors can operate on a one or two day cycle. The gold recovery and carbon handling circuit will operate on a batch basis and 2.5 tonnes of carbon could be moved twice a day. Lycopodium Minerals Canada

301 NI TECHNICAL REPORT - RECOVERY METHODS Page Ore Receiving and Crushing Run-of-mine (ROM) ore from the open pit will be transported to the plant by 65 t capacity rear dump trucks. The trucks will tip directly into the ROM bin however allowance will be made for a ROM stockpile. The ROM stockpile will be primarily utilized for emergency storage and ore blending. ROM ore will be reclaimed, from the stockpile, to the ROM bin by a front-end loader. A static grizzly (800 mm), mounted above the ROM bin, will prevent the ingress of oversize material. A rock breaker will be installed to assist in breaking down oversize material, retained on the static grizzly. Ore will be withdrawn from the ROM bin, by a variable speed apron feeder, directly to a vibrating grizzly (75 mm aperture size). Oversize from the grizzly will report directly to the jaw crusher, which will operate in open circuit. Crushed ore from the crusher discharges, together with undersize from the grizzly, directly onto the primary crusher discharge conveyor, which will convey the crusher product to the mill feed bin. The crushed ore product conveyor will be fitted with a weightometer, to monitor the crushing area throughput and for controlling the apron feeder variable speed drive. The crushing circuit will be serviced by a single dust collection system, comprised of a series of extraction hoods, ducting and a bag house. Dust collected from this system will be discharged onto the crusher product conveyor. A static tramp metal magnet will be installed at the discharge end of the primary crusher discharge conveyor. On a periodic basis, tramp metal, removed by the magnet, will be manually removed. Any spillage generated, within the crushing area, will be manually recovered and transported to the mill feed bin. Auxiliary equipment for the crushing circuit will include: Crushing area control room and rock breaker operator cabin. Primary crusher maintenance hoist. Primary crusher grease pump. Primary crusher area camera. ROM bin camera. Lycopodium Minerals Canada

302 NI TECHNICAL REPORT - RECOVERY METHODS Page Coarse Ore Storage The mill feed bin will have a live capacity of approximately 55 t (equivalent to 20 minutes plant feed at 167 t/h). The mill feed bin includes an overflow facility, with excess crushed ore conveyed to the crushed ore stockpile. The crushed ore stockpile will have a capacity of 8,000 tonnes (equivalent to 48 hour plant feed at 167 t/h). Crushed ore will be reclaimed from the stockpile, to the mill feed bin, via a front end loader. Crushed ore will be withdrawn from the mill feed bin, by a variable speed apron feeder. The feeder will discharge onto the SAG mill feed conveyor, which will convey the crushed ore to the SAG mill feed chute. The SAG mill feed conveyor will be fitted with a weightometer, used for controlling the speed of the apron feeder and for mass accounting of feed presented to the grinding circuit. Hydrated lime will be added directly to the SAG mill feed conveyor, via the lime variable speed rotary feeder. The hydrated lime storage silo will have a storage capacity of 60 t, equivalent to 47 days storage at 1.34 Mtpa. Any spillage generated, within the reclaim area, will be manually recovered and transported to the mill feed bin. Supernatant liquor will be pumped to the SAG mill discharge screen, via the mill feed area sump pump Auxiliary equipment for the reclaim area will include: Trolley hoist for apron feeder maintenance. Crushed ore stockpile dust suppression sprays. Apron feeder discharge camera. Lime silo dust collector. Lime hoist, lifting frame. Weightometer calibration chain Grinding and Classification The grinding circuit will be a SATMC circuit, comprised of a single, variable speed, SAG mill and a single fixed speed tower mill. The SAG mill will operate in closed circuit with a pebble crusher, whilst the tower mill will operate in closed circuit with hydrocyclones. The product particle size exiting the grinding circuit (hydrocyclone overflow) will contain 80% passing 63µm material. To meet the design throughput, and achieve the required leach product size, a 6.10m diameter x 3.72m EGL SAG mill (20ft x 12.2ft; 2.30 MW) and a 3.3 MW tower mill will be required. Lycopodium Minerals Canada

303 NI TECHNICAL REPORT - RECOVERY METHODS Page 17.8 Crushed ore, reclaimed from the mill feed bin, will be conveyed to the SAG mill feed chute via the SAG mill feed conveyor. Process water will be added to the SAG mill feed chute, to control the in-mill pulp density. The SAG mill will be fitted with discharge grates which will allow slurry to pass through the mill and will also relieve the mill of pebble build-up. The SAG mill product will discharge to a single deck vibrating screen, for size classification. Grinding media (125 mm balls) will be added to the SAG mill via the SAG mill feed chute, utilizing the mill area gantry crane, kibble and feed chute. SAG mill discharge screen oversize will be conveyed to a pebble crushing circuit. Undersize from the discharge screen will gravitate to the SAG mill discharge hopper, where it will combine with discharge from the tower mill, prior to being pumped to the classification hydrocyclone cluster by one of two (duty/standby) variable speed hydrocyclone feed pumps. The classification hydrocyclone cluster overflow will gravitate, via a trash screen, to the pre-leach thickener feed distribution box. The trash screen undersize will constitute the pre-leach thickener feed whilst trash screen oversize will be discharged to a trash bunker. Underflow slurry, from the classification hydrocyclone underflow launder, will be split into three streams with one stream returning to the tower mill, via the tower mill feed port. The remaining two streams will each feed a dedicated gravity circuit screen. Tower mill product will discharge to the SAG mill discharge hopper. Grinding media (50 mm balls) will be added to the tower mill via the ball feed port, utilizing the mill area gantry crane, kibble and feed chute. Spillage within the grinding circuit will be managed, utilizing a dedicated drive-in sump and sump pump. The drive in sumps will be sized to allow for two emergency stops of the SAG mill. Drive in sump supernatant liquor will be returned to the SAG mill discharge hopper with solids being recovered by front-end loader and returned to the stockpile. Within the tower mill circuit dedicated sump pumps will collect spillage and wash down generated in the circuit and will transfer the slurry to the SAG mill discharge hopper. Auxiliary equipment within the grinding area will include: SAG mill and Tower mill drive lubrication system. SAG mill liner handler and relining monorail. Mill area gantry crane. Metallurgical sampler on hydrocyclone overflow, for metallurgical balancing and accounting. Lycopodium Minerals Canada

304 NI TECHNICAL REPORT - RECOVERY METHODS Page Pebble Crushing Oversize from the SAG mill discharge screen will be conveyed to the pebble crusher feed bin, via two belt conveyors. A self cleaning tramp metal magnet will be mounted above the pebble recycle conveyor, to remove any scrap metal and steel media which could potentially damage the pebble crushers. Upstream of the tramp metal magnet, the pebbles will pass under a metal detector, prior to discharging into the pebble crusher feed bin. The feed bin will provide surge capacity ahead of the pebble crusher and allows a controlled feed to be presented to the crusher. Should the pebble crusher be out of service, a diverter gate ahead of the pebble bin will allow pebbles to bypass the pebble bin and crusher directly to the pebble recycle conveyor. Similarly, should the metal detector detect tramp metal (not removed by the cross-belt magnet), the diverter gate ahead of the pebble bin will automatically allow pebbles to bypass the pebble bin and crusher, directly to the pebble recycle conveyor. Pebbles will be withdrawn from the pebble crusher feed bin, by a retractable variable speed belt feeder. The pebble crusher will discharge crushed pebbles directly onto the pebble recycle conveyor which in turn will return the crushed pebbles to the SAG mill feed conveyor. A weightometer, installed on the pebble return conveyor, will monitor the pebble crushing circuit throughput. Spillage within the pebble crushing circuit will be managed, utilizing a dedicated drive-in sump and sump pump. Drive in sump supernatant liquor will be returned to the SAG mill discharge screen with solids being recovered by front-end loader and returned to the stockpile. Auxiliary equipment within the pebble crushing area will include: Pebble crusher lubrication unit and hydraulic power pack. Pebble crusher maintenance hoists Gravity Concentration Underflow from the hydrocyclone cluster will be split into three streams by the hydrocyclone underflow launder, with approximately 50% of the underflow constituting the feed to the gravity circuit (25% to each concentrator). The remaining 50% will gravitate to the Tower mill feed port. The gravity circuit will consist of two gravity screens and two variable speed centrifugal concentrators. Hydrocyclone underflow will discharge onto the vibrating, single deck gravity feed preparation screen. The gravity feed preparation screen will be fitted with a 2mm aperture screen panel, with -2mm undersize material constituting the feed to the gravity concentrators. Gravity screen oversize material (+2mm) will be returned to the Tower mill top feed port. Lycopodium Minerals Canada

305 NI TECHNICAL REPORT - RECOVERY METHODS Page Gravity screen undersize will constitute the feed to the gravity concentrator. Each gravity concentrator will have a capacity of 150 t/h and operate as a batch process with nominally 24 cycles per day. Concentrate from the gravity concentrators will feed the Intensive Cyanidation Reactor (ICR) circuit. Tails from the gravity concentrators will return to the Tower mill top feed port. Concentrate from the gravity concentrators will be discharged to the ICR gravity concentrate storage cone. Prior to commencing intense cyanidation, concentrate held in the storage cone will be fluidized, with water, to remove fines from the system. Vessel overflow, carrying the fines will gravitate to the gravity area sump pump, from where it will be returned to the SAG mill discharge hopper. The ICR leach solution (2% NaCN, 1% NaOH and Leach Aid) will be made up within the agitated reaction vessel feed tank. From the feed tank the leach solution will be pumped though the reaction vessel, with the leach solution circulating through the reaction vessel for approximately 20 hrs. Upon completion of the leach cycle, solution from the reaction vessel will be drained to the reaction vessel feed tank. The residue within the reaction vessel will be washed, with wash water recovered to the reaction vessel feed tank. Residue solids will be fluidized and pumped to the SAG mill discharge hopper. The ICR pregnant eluate will be pumped from the reaction vessel feed tank to the ICR pregnant eluate tank, located in the gold room. The ICR area will be serviced by a gold trap and dedicated sump pump. Any spillage within this area will be pumped back to the SAG mill discharge hopper. Auxiliary equipment within the gravity recovery area will include: Gravity concentrator maintenance hoists. ICR leach aid screw feeder Pre-Leach Thickening Trash screen undersize will gravitate directly to the pre-leach thickener feed box, where flocculant will be added to aid with particle settling. Overflow from a 14m diameter pre-leach thickener will gravitate to the process water tank. Underflow from the pre-leach thickener, at 50% solids, will be pumped by dedicated thickener underflow pumps, to the leach circuit feed distribution box. The pre-leach thickener area will be serviced by a dedicated sump pump. Spillage and wash down collected by the sump pump will be returned to the thickener feed box. Auxiliary equipment within the pre-leach thickener area will include: Pre-leach thickener flocculant mixer. Leach feed samplers. Lycopodium Minerals Canada

306 NI TECHNICAL REPORT - RECOVERY METHODS Page Leach Circuit Pre-leach thickener underflow will be pumped to the leach feed distribution box. The slurry from the leach feed distribution box will gravitate to the first leach tank. If the first leach tank is offline, the slurry will be diverted to the second leach tank, via an internal dart plug distribution system. The leach circuit will consist of five, mechanically agitated, leach tanks (13 m diameter x 13.5m high) operating in series. This equates to a nominal residence time of 36 hrs at a feed rate of 167 tph. Each leach tank will have a live volume of 1,725 m³. Cyanide, for gold dissolution, will be added to the leach circuit by dedicated cyanide dosing pumps. The primary cyanide dosing point will be the leach feed distribution box, with further addition points located down the leach train. The operating ph of the leach circuit will be maintained above 10.5 to maintain the protective alkalinity of the circuit and prevent the loss of cyanide to gaseous hydrogen cyanide. Protective alkalinity will be maintained by the addition of hydrated lime to the SAG mill feed conveyor. To aid with gold dissolution, oxygen will be added to the leach circuit. Oxygen will be supplied from the onsite oxygen plant, with oxygen being injected into each leach tank, via dedicated oxygen spargers. To sustain the leach dissolved oxygen levels, external oxygen contactors will be installed on the first two leach tanks. Lead nitrate will be added to the first two leach tanks. Following dissolution, the solubilized gold will be recovered by carbon adsorption, within the dedicated carbon-in-pulp (CIP) carousel Pumpcell circuit. Leach circuit tails slurry will gravitate to the carousel circuit feed distribution launder. Should a leach tank be off-line for maintenance, it will be possible to bypass any of the leach tanks. The ability to bypass tanks will be made possible by the installation of two pneumatic gates located within the leach inter-stage launders. One gate will divert slurry to the following leach tank while the second gate will allow slurry diversion to the subsequent leach tank. The leach circuit will be serviced by two sump pumps. All sump pumps will return spillage to the leach feed distribution box. Auxiliary equipment within the leach area will include: Cyanide analyser. Hydrogen Cyanide (HCN) monitor. Control and titration room. Lycopodium Minerals Canada

307 NI TECHNICAL REPORT - RECOVERY METHODS Page Carbon Absorption Circuit The slurry from the leach circuit will gravitate to the CIP plant feed launder. The feed launder will distribute the slurry to the first tank, within the carousel adsorption sequence. The CIP circuit will consist of seven, mechanically agitated, tanks operating in series. This equates to a residence time of 1.3 hrs at a feed rate of 167 t/h. Each tank will have a live volume of 50 m³. The tanks will operate with a carbon concentration of 50 g/l. The adsorption tanks will operate on a two day cycle with carbon recovered to the acid wash circuit every second day. Activated carbon will be retained in each of the tanks, by an inter-tank screen, which is integral to the tank agitator mechanism. The inter-tank screen will be a stainless steel wedge wire cylinder equipped with an internal agitator and external rotating wiper blade mechanism to prevent screen blinding. The CIP circuit operates as a carousel, with carbon retained within the tanks and the slurry advanced counter current to the carbon adsorption stage. Slurry will nominally gravitate from Tank 1 through to Tank 7, with the gravitational flow between tanks induced by the pumping action of the tank agitator/screen mechanism. Periodically (1 hr every second day), a complete batch of loaded carbon, from the first contactor, will be pumped by the loaded carbon recovery Pump to the loaded carbon screen, where it will be washed with spray water to remove excess slurry. The excess slurry (screen underflow) will gravity return to the CIP feed launder whilst the loaded carbon will gravitate to the acid wash column. Regenerated carbon (or fresh carbon) will be hydraulically added to the CIP circuit, from the carbon regeneration circuit. The regenerated carbon (or fresh carbon) will be pumped, to the CIP circuit, via the carbon sizing screen. The sizing screen will remove excess water and carbon fines. The dewatered carbon will discharge into the last adsorption tank, with excess water and carbon fines gravitating to the carbon safety screen. Slurry discharging the last adsorption tank will gravitate to the adsorption tails hopper, from where it will be pumped to a carbon safety screen via the carbon safety screen feed box. The carbon safety screen will capture and recover any carbon exiting the adsorption circuit. The safety screen oversize will report to a fine carbon bin while the undersize will gravitate to the tails thickener feed box. A sampler, installed on the carbon safety screen feed will periodically collect a sample of the adsorption tail stream. This sample will be used for circuit monitoring and for metal accounting. The adsorption circuit will be serviced by a dedicated sump pump. The sump pumps will return spillage to the CIP feed distribution launder. Auxiliary equipment within the Pumpcell circuit area will include: Maintenance gantry crane. Pumpcell mechanism maintenance frame. Lycopodium Minerals Canada

308 NI TECHNICAL REPORT - RECOVERY METHODS Page Spare Pumpcell mechanism. High pressure cleaner Desorption and Carbon Regeneration The desorption circuit will consist of separate acid wash and elution columns. A cold acid wash will be utilized. Following acid wash, gold will be eluted from the carbon, utilizing a split Anglo American (AARL) elution process. With the CIP circuit operating on a two day cycle, one elution will be completed every second day. The elution circuit will however be sized to complete two strips per 24 hour period. At a total carbon (gold + silver) loading of 7,318 g/t the required daily carbon movement equates to 2.5 tonnes. Acid Wash The cold acid wash sequence will be required to remove accumulated, calcified scale, from the carbon surface. The acid wash column fill sequence will be initiated by pumping carbon, from the first adsorption tank, into the Acid Wash Column via the Loaded Carbon Recovery Screen. Carbon will gravitate from the Loaded Carbon Recovery Screen directly into the Acid Wash Column. Once the Acid Wash Column is filled to the required level, the carbon fill sequence will be stopped. The acid wash cycle will utilize a 3% w/v hydrochloric acid solution. This dilute acid will be prepared, by the addition of raw water and neat (32%) hydrochloric acid, into the Hydrochloric Acid Dilution Tank. The acid wash sequence will involve the injection of the dilute acid solution into the column, by the Hydrochloric Acid Dosing Pump, via the feed manifold located beneath the column. Once the required amount of acid has been added to the column, the Hydrochloric Acid Dosing Pump will be stopped and the carbon will be allowed to soak for a period of one hour. Upon completion of the acid soak, the acid rinse cycle will be initiated by pumping water through the column, to displace the spent acid solution to the tailings thickener. Acid rinse water will be sourced from the Transfer Water Tank and pumped through the column by the Transfer Water Pump. During the rinse cycle, four bed volumes (4 BV) of water, at 2 BV/h, will be pumped through the column. The first 2 BV will include a caustic injection, to neutralize the acid waste, whilst the last 2 BV are comprised of a fresh water rinse only. Acid waste and displaced solution from both the acid rinse and wash steps will pass through the Acid Wash Discharge Strainer before discharging to the Tails Thickener Feed Box. The sequence will conclude with carbon being hydraulically transferred to the Elution Column. Water, for carbon transfer between the acid wash and elution columns, will be supplied from the Transfer Water Tank via the Transfer Water Pump. Lycopodium Minerals Canada

309 NI TECHNICAL REPORT - RECOVERY METHODS Page Elution The elution sequence will commence with the injection of a set volume of water into the column, via the Lean Eluate Tank, along with the simultaneous injection of cyanide and caustic solution. A set amount of cyanide and sodium hydroxide (caustic) will be added to achieve a 2% w/w NaOH and 2% w/w NaCN solution. Both reagent additions will be automatically stopped once the prescribed volume has been added. The pre soak period will then commence. During this period, the caustic solution will be circulated through the column and be pre heated to 95 C via a diesel fired elution heater. Upon completion of the pre soak period, the elution sequence will commence and gold will be stripped from the carbon. The split AARL process will use the last four bed volumes of low grade (lean) eluate from the previous elution as the first eluant for the current elution. Lean eluate, from the Lean Eluate Tank will be pumped, by the Lean Eluate Pump, through the Recovery Heat Exchanger, picking up residual heat from the eluate exiting the elution column. The pre-heated, incoming eluate, will then pass through the Primary Heat Exchanger to elevate the eluate temperature to 125 C prior to entering the base of the column. A diesel fired elution heater will provide the heat to the Primary Heat Exchanger. A temperature probe will monitor the temperature of eluate exiting the column and will be used to control the heater output. Eluate will flow up and out of the top of the column, passing through the Recovery Heat Exchanger via the Elution Discharge Strainers. Initially, eluate emerging from the heat exchanger will be directed to the Pregnant Eluate Tank. The elution sequence will progress for a fixed period of time, allowing the 4 BV of lean eluate to pass through the column at 2 BV/h. Once the lean eluate storage volume is exhausted, new incoming strip water solution (6 BV) will be sourced from the Strip Water Tank. This strip solution will also pass through the recovery and primary heat exchangers, to be heated to 125 C, prior to entering the base of the column. However, on discharging from the top of the column, only 2 BV of this eluate will report to the pregnant eluate tank. At this point the elution heater will be switched off, with the last 4BV of eluate utilised to cool down the elution column and its contents, to less than 100 C. The last 4 BV of eluate will be directed to the Lean Eluate Tank for re-use in subsequent elution sequences. Upon completion of the cool down sequence, the carbon will be hydraulically transferred to the carbon regeneration kiln de-watering screen, via the Transfer Water Pump. The acid wash area and elution area will be serviced by the Acid Wash and Elution Area Sump Pumps respectively. Acid wash area spillage will be pumped to the Tailings Thickener. Elution area spillage will be pumped to the leach circuit feed distribution box. Auxiliary equipment within the Acid Wash and Elution circuits will include: Elution Heater Expansion Tank and Thermal Oil Drum Pump. Elution Heater Pump. Strip Solution Heater. Lycopodium Minerals Canada

310 NI TECHNICAL REPORT - RECOVERY METHODS Page Acid wash and elution column discharge strainers. Carbon Regeneration After elution, the carbon will be hydraulically transferred from the Elution Column to the carbon regeneration circuit by pressurizing the column with transfer water. The carbon and transfer water will be directed to the carbon dewatering screen, allowing excess water to be removed prior to the carbon discharging into the Kiln Feed Hopper. Dewatering screen undersize will gravitate to the elution area transfer water tank. Carbon will be withdrawn from the Kiln Feed Hopper, via the Kiln Screw Feeder, and discharged directly to the Carbon Regeneration Kiln, at a rate of 124 kg/h. Within the diesel fired, horizontal rotary kiln, the carbon will be heated to 700 C, to remove volatile organic foulants from the carbon surface, thereby restoring the carbon activity. Re-activated carbon exiting the kiln will discharge directly to the carbon quench tank, where it will be rapidly cooled. From the quench tank carbon will be pumped, by the regenerated carbon transfer pump to the carbon sizing screen. Sizing screen oversize will gravitate to the CIP adsorption circuit. Sizing screen undersize will gravitate to the CIP tails hopper. Fresh carbon will also be added to the CIP circuit via the fresh carbon conditioning tank. The carbon regeneration area will be serviced by dedicated sump pump. Any spillage generated within this area will be pumped to the carbon safety screen. Auxiliary equipment within the Carbon Regeneration circuit will include: Kiln Exhaust Fan. Kiln Air blower. Carbon Addition Hoist. Carbon Bag Lifting Frame Electrowinning and Gold Room Soluble gold and silver recovery, from the pregnant eluate, will be achieved by electrowinning onto stainless steel cathodes. The electrowinning circuit will consist of two electrowinning cells, each containing 22 cathodes and 23 anodes. A dedicated rectifier, per electrowinning cell, will supply the necessary current, to electroplate the gold and silver onto the cathode. Lycopodium Minerals Canada

311 NI TECHNICAL REPORT - RECOVERY METHODS Page Once sufficient pregnant eluate is available, within the Pregnant Eluate Tank, the electrowinning sequence will be initiated by starting the Electrowinning Feed Pump. The flow of pregnant eluate to the cell will be manually controlled, to sustain the desired linear velocity. During the electrowinning cycle, the electrowinning cell discharge will be continuously returned to the pregnant eluate tank, via gravity. Once the target barren solution grades have been achieved, the electrowinning cycle is complete, and barren eluate will discharge to the Pregnant Eluate Tank. Barren eluate, from the Pregnant Eluate Tank will be returned to the leach circuit, via the Barren Eluate Pump. To conserve water, the options will exist to re-use barren eluate as strip solution, with the ICR leach or for reagent make-up. Pregnant eluate from the intensive cyanidation reactor will be stored within a dedicated ICR Pregnant Eluate Tank. Once sufficient pregnant eluate is available, within the ICR Pregnant Eluate Tank, the electrowinning sequence will be initiated by starting the ICR Electrowinning Feed Pump. The flow of pregnant eluate, to the dedicated ICR electrowinning cell, will be manually controlled, to sustain the desired linear velocity. During the electrowinning cycle, the electrowinning cell discharge will be continuously returned to the ICR Pregnant Eluate Tank, via gravity. Once the target barren solution grades have been achieved, the electrowinning cycle is complete, and barren eluate will discharge to the ICR Pregnant Eluate Tank. Barren eluate, from the ICR Pregnant Eluate Tank will be returned to the leach circuit, via the ICR Barren Eluate Pump. Upon completion of electrowinning, gold sludge on the plated cathodes will be washed off the cathodes, with a high pressure cathode washer. The gold bearing sludge will be recovered to a sludge hopper, from where it will be filtered, via a pressure filter. To assist with metal accounting, dedicated sludge hoppers will be provided for the ICR electrowinning and leach electrowinning sludge. The gold bearing filter cake will be thermally dried in an electric drying oven. Dried filter cake will be mixed with a prescribed flux mixture (silica, nitre and borax), prior to being charged into the diesel fired gold furnace. The fluxes added react with base metal oxides to form a slag, whilst the gold and silver remains as a molten metal. The molten metal will be poured into moulds, to form doré ingots, which will be cleaned, assayed, stamped and stored in a secure vault ready for dispatch. The slag produced will periodically be returned to the grinding circuit, via the SAG mill. The gold room and electrowinning area will be serviced by a gold trap and dedicated gold room area sump pump. Any spillage within this area will be pumped back to the leach circuit. Auxiliary equipment for the Electrowinning and Gold Room circuit will include: Furnace dust collector and extraction fan. Electrowinning cell fume hood and extraction fan. Gold room hoist. Lycopodium Minerals Canada

312 NI TECHNICAL REPORT - RECOVERY METHODS Page High pressure cathode cleaner. Smelting furnace cascade trolley and slag cart. Doré moulds and doré wash table. Flux bin, platform scale, flux mixing table. Doré balance. Doré vault Tailings Thickening Slurry from the CIP circuit will be pumped to a carbon safety screen via the carbon safety screen feed box. The carbon safety screen will capture and recover any carbon exiting the adsorption circuit. The safety screen oversize will report to a fine carbon bin while the undersize will gravitate to the tails thickener feed box. A sampler, installed on the carbon safety screen feed will periodically collect a sample of the adsorption tail stream. This sample will be used for circuit monitoring and for metal accounting. Flocculant will be added to the 16 m diameter tails thickener to enhance the settling properties of the solids. Overflow from the tails thickener will gravitate to the process water tank. Tails Thickener underflow, at a solids content of 65% solids, will be pumped to the Tailings Storage Facility (TSF). The tails thickening and disposal circuit will be serviced by a dedicated sump pump. Any spillage collected within this area will be returned to the tails thickener feed box. Auxiliary equipment for the tails thickening and filtration circuit will include: Tails thickener hydraulic drive. Tails thickener flocculant mixer. Final tails sampler Reagents Mixing and Storage The major reagents utilized within the process plant will include: Hydrated lime (Ca(OH) 2 ) for ph control. Lycopodium Minerals Canada

313 NI TECHNICAL REPORT - RECOVERY METHODS Page Sodium cyanide (NaCN) for gold dissolution and desorption. Lead nitrate (PbNO 3 ) for gold dissolution (leach accelerant). Caustic soda (NaOH) for neutralization and desorption. Hydrochloric acid (HCl) for carbon acid washing. Flocculant for thickening. Antiscalant to reduce fouling in the process water distribution, carbon wash and stripping circuit. Fluxes for smelting charge preparation. Hydrated Lime Powder, hydrated lime will be delivered to site in bulk tankers or 1,000 kg bulk bags. Bulk tankers will off-load directly to the lime storage silo. The bulk bags will be lifted, by the lime hoist, to the lime storage silo. The lime storage silo will have a storage capacity of 47 days or 60 tonne. Hydrated lime will be withdrawn from the storage bin, by a variable speed rotary feeder, and discharge directly to the SAG mill feed conveyor. The hydrated lime area will be serviced by a single sump pump. Any spillage generated within the lime mixing area will be pumped to the final tails thickener. Sodium Cyanide (NaCN) Sodium cyanide will be delivered to site in 1,000 kg bulk bags or in bulk iso-tainers. The bulk bag will be lifted, by the cyanide hoist, to the bulk bag splitter mounted above the cyanide mixing tank. Solid cyanide will be released, from the bulk bag, by the bag splitter. Process water will be added to the mixing tank to achieve a solution with the desired cyanide concentration (20% w/w). The mixing tank will be mechanically agitated to assist with cyanide dissolution. A standpipe, connected directly to the mixing tank, provides a storage buffer for subsequent make-up cycles. Cyanide solution will be distributed, via a cyanide ring main, to the leach circuit. A dedicated metering pump will deliver cyanide solution, to the elution column. Sodium hydroxide will be added to the cyanide mixing tank, to maintain the ph above 10.5 during the mixing process. Control of the ph above 10.5 is required to prevent the formation of gaseous hydrogen cyanide (HCN) during the mixing process. Lycopodium Minerals Canada

314 NI TECHNICAL REPORT - RECOVERY METHODS Page The sodium cyanide and sodium hydroxide mixing and storage areas will be serviced by a common sump pump. Any spillage generated within this area will be pumped to the leach circuit feed distribution box. Sodium Hydroxide (NaOH) Sodium hydroxide (caustic) will be delivered to site in 1,000 kg bulk bags. The bulk bag will be lifted, by the caustic hoist, to the bulk bag splitter mounted above the caustic mixing tank. Caustic will be released, from the bulk bag, by the bag splitter. Raw water will be added to the mixing tank to achieve a solution with the desired caustic concentration (20% w/v). The mixing tank will be mechanically agitated to assist with caustic dissolution. Dedicated metering pumps will deliver caustic solution to the intensive cyanidation reactor, acid wash column, elution column and pregnant eluate tank. Hydrochloric Acid (HCl) Hydrochloric acid (32% w/w) will be delivered to site in 1,000 L bulk containers (IBC). The acid transfer pump will transfer acid, from the bulk containers, to the acid storage tank. From the acid storage tank, the hydrochloric acid dosing pump will transfer the requisite amount of acid, to the hydrochloric acid dilution tank. The hydrochloric acid storage area will be serviced by an air operated sump pump. Any spillage generated within this area will be pumped to the tailings thickener. Lead Nitrate Lead nitrate will be delivered to site in 1,000 kg bulk bags. The bulk bag will be lifted, by the caustic hoist, to the bulk bag splitter mounted above the lead nitrate mixing tank. Lead nitrate will be released, from the bulk bag, by the bag splitter. Raw water will be added to the mixing tank to achieve a solution with the desired caustic concentration (20% w/v). The mixing tank will be mechanically agitated to assist with lead nitrate dissolution. A standpipe, connected directly to the mixing tank, provides a storage buffer for subsequent make-up cycles. Dedicated metering pumps will deliver lead nitrate solution to the leach feed distribution box. The lead nitrate mixing and storage area will be serviced by a dedicated sump pump. Any spillage generated within this area will be pumped to the leach feed distribution box. Lycopodium Minerals Canada

315 NI TECHNICAL REPORT - RECOVERY METHODS Page Flocculant Flocculant powder will be delivered to site in 25 kg bags and mixed in a proprietary mixing system, comprised of a bulk dry hopper, screw feeder, flocculant blower, mixing tank and storage tank. The Flocculant Plant will mix flocculant powder with raw water to achieve the required storage concentration (0.25% w/w). Flocculant will be withdrawn from the storage hopper by the flocculant screw feeder. The screw feeder will convey flocculant to the flocculant eductor, from which the flocculant powder will be pneumatically conveyed, to the flocculant mixer, by the flocculant blower. Raw water will be added to the mixer, to hydrate the flocculant powder, prior to discharging into the agitated flocculant mixing tank. Upon completion of the mixing cycle, the flocculant will be transferred to the flocculant storage tank, by the flocculant transfer pump. From the holding tank, flocculant will be distributed to both the pre-leach and tails thickeners (via in-line mixers) by the respective flocculant dosing pumps. The flocculant area will be serviced by a sump pump. Any spillage generated within this area will be pumped to the tails thickener. Antiscalant Antiscalant will be delivered to the plant in bulk containers (IBC). Metering pumps will distribute antiscalant, directly from the IBC, to the elution and process water circuits. Fluxes The following fluxes, will be delivered to the plant in 25 kg bags, and used in the gold room; Borax (Na 2 B 4 O 2 ), Sodium Nitrate (NaNO 3 ), Sodium Carbonate (Na 2 CO 3 ) and Silica (SiO 2 ) Water Services The process plant will utilise raw water, treated water and process water. Any shortfall of process water will be made up, preferentially, from water contained within the site water storage pond (WSF). If insufficient water is available within the process water pond, raw water will be utilised for process make-up water. Lycopodium Minerals Canada

316 NI TECHNICAL REPORT - RECOVERY METHODS Page Run-Off Water, Ponds and Water Management Dedicated ponds will be provided to manage site water and environmental impacts. Water from the open pit de-watering station will be collected within in the water storage facility. Waste dump run-off (via waste rock dump toe drains) will be collected within in dedicated sediment ponds. From the sediment pond, water will be pumped to the water storage pond (WSF). From the WSF, water will be pumped to the raw water storage tank. Any shortfall of water will be supplemented by water from the bore fields. Process Water Process water will predominantly consist of pre-leach thickener overflow, tails thickener overflow and TSF return water. Process water will be stored in a process water pond. From the process water pond, water will be transferred to the process water tank, which will provide and intermediate storage buffer. From the process water tank, process water will be reticulated throughput via a ring-main and dedicated the process water pumps, with off-takes supplied for the predominant user points, namely: Grinding. Gravity circuit. Leaching and CIP circuit. Tails thickening. Reagent mixing. Raw Water and Fire Water Raw water for the process plant and mining operation will be sourced from the WSF and a dedicated bore field. Raw water from the various sources will be stored within the process water tank. Water, from the plant raw water tank, will be pumped to the process water tank, to supplement the water make-up requirements. Raw water from storage tank will be reticulated throughput the plant by dedicated raw water pumps, to the predominant user points, namely: Grinding circuit cooling water. Acid wash and elution. Carbon regeneration. Reagent make-up. Lycopodium Minerals Canada

317 NI TECHNICAL REPORT - RECOVERY METHODS Page Gland seal water. Firewater will be supplied from the plant raw water storage tank, via a dedicated suction manifold. The firewater system will comprise: An electrical jockey pump. An electrical firewater pump. A diesel standby firewater pump. The firewater system pressure will be maintained, by the jockey water pump. An electric fire water pump will automatically start on a drop in line pressure. The diesel fire water pump will automatically start if the line pressure continues to drop below the target supply pressure or during a power failure. Gland water will be supplied from the raw water storage tank. A dedicated high pressure gland water system (pumps) will supply the gland water requirements, for the tails disposal pumps. A dedicated low pressure gland water system (pumps) will supply gland water to the remainder of the process plant gland water duties. Raw Water Treatment Raw water, for potable water generation will be subjected to water treatment by reverse osmosis (RO), for calcium, magnesium and chloride removal. Treated water, from the RO plant will be stored within a dedicated potable water storage. From the potable water storage tank, potable water will be distributed for human consumption and to the safety showers and eye wash stations. Raw water, for use within the process plant (cooling water circuit, acid wash circuit and elution circuit) will be subjected to water treatment by filtration. Treated water, from the water filtration plant will be stored within a dedicated filtered water storage tank respectively. From the filtered water storage tank, water will be distributed to the acid wash and elution circuits and SAG mill cooling water circuit. A closed circuit cooling water circuit is utilized, with cooling water being returned to the filtered water storage tank Air and Oxygen Services Plant air at 700 kpag will be provided by two high pressure air compressors, operating in a lead-lag configuration. The entire high pressure air supply will be dried. Dried air will be distributed to the required plant areas, via two dedicated air receivers servicing the grinding and balance of plant areas respectively. A dedicated duty/standby high pressure air compressor will supply high pressure air for the tails filtration plant. Lycopodium Minerals Canada

318 NI TECHNICAL REPORT - RECOVERY METHODS Page Oxygen, for use within the leach circuit, will be supplied from a dedicated VSA oxygen plant Plant Consumption Water Consumption A water balance for the process plant has been completed. Water from the pre-leach thickener and tailings thickener overflow streams are recycled within the process plant to reduce the external water requirement. Approximately 23 m 3 /h of decant return water is recycled from the TSF to the process plant. Another approximately 42 m 3 /h of raw water is required to make-up the water consumption for the process plant. The fresh water will be sourced from the WSF. At the process plant, the raw water tank will provide a combined fresh water and fire water reserve. The required fire water capacity will be stored below the level of the suction piping for the fresh water pumps to ensure it is always available Energy Consumption The power demand for the plant, along with the rest of the site and camp, will be provided by on-site hybrid power generation using heavy fuel oil (HFO) and diesel fuel. The average power demand is summarized in Table 17.2 and utilized for the operating cost estimate. The average power demand does not reflect the instantaneous power demand for equipment start-up and power plant capacity sizing. Lycopodium Minerals Canada

319 NI TECHNICAL REPORT - RECOVERY METHODS Page Table 17.2 Average Power Demand Summary Plant Areas Installed Power (kw) Average Continuous Draw (kw) Total Annual Power Consumption (kw / year) Area Feed Preparation ,497,960 Area Milling (excl Mills) 1, ,337,560 Area SAG Mill (2300 kw Motor) 2,371 1,745 15,286,200 Area 130 Tower Mill (3356 kw Motor) 3,570 2,628 23,018,652 Area Pebble Crushing ,520 Area Screening & Tailings ,305,240 Areas 161/162/163 - Leaching & CIP ,810,600 Area Elution ,760 Area Gold Room ,840 Area Reagents ,400 Area Plant Water Services ,156,320 Area Raw Water ,680 Area Air Services ,520 Area Oxygen Plant ,040 Area Fuel Dispensing ,840 Area TSF ,520 Area Outside Plant Buildings ,600 Area Inside Plant Buildings ,600 Area 370/380 - Camp and Camp Services ,440 Area Mine Facilities ,880 Total 11,561 6,405 56,105,172 Lycopodium Minerals Canada

320 NI TECHNICAL REPORT - RECOVERY METHODS Page Reagent and Consumable Consumption Reagent storage, mixing and pumping facilities will be provided for all reagents for the process plant. Table 17.3 provides a summary of reagents and consumables that will be used at the process plant at the design consumption rate for a plant throughput of 4,000 tpd. Table 17.3 Annual Reagent and Major Consumable Consumption Reagent / Consumable Annual Consumption Jaw Crusher Liners (fixed and swing jaw) 8 sets Pebble Crusher (mantle and concave) 13 sets SAG Mill Liners 0.5 sets Vertical Mill Liners 1.5 sets SAG Mill Grinding Media 500 t Vertical Mill Grinding Media 915 t Hydrated Lime 403 t Sodium Cyanide 604 t Lead Nitrate 134 t Activated Carbon 40 t Sodium Hydroxide (Caustic) 32 t Hydrochloric Acid 150 t Flocculant 54 t Borax 5.7 t Sodium Nitrate (Nitre) 0.8 t Sodium Carbonate (Soda Ash) 0.8 t Silica Sand 2.4 t Smelting Furnace Crucibles 5.2 units Diesel Fuel (plant usage only) 905 m Plant Control System General Overview The following provides a broad overview of the control strategy that will be employed for the plant. The general control philosophy for the plant will be one with a moderate level of automation and remote control facilities, to allow process critical functions to be carried out with minimal operator intervention. Instrumentation will be provided within the plant to measure and control key process parameters. Lycopodium Minerals Canada

321 NI TECHNICAL REPORT - RECOVERY METHODS Page The main control room, located in the Mill Office, will house two PC based operator interface terminals (OIT) and a single server. These workstations will act as the control system supervisory control and data acquisition (SCADA) terminals. The control room is intended to provide a central area from where the plant is operated and monitored and from which the regulatory control loops can be monitored and adjusted. All key process and maintenance parameters will be available for trending and alarming on the process control system (PCS). Two additional OITs will be provided for data logging and engineering / programming functions. A field touch panel will be installed in the feed preparation area to allow local operator control of the crushing plant to facilitate ease of operation for rock breaking and stockpiling. A second field touch panel will be installed in the elution area to allow local operator control of the elution sequence. A third field touch panel will be supplied for the milling and gravity circuit area. The process control system that will be used for the plant will be a programmable logic controller (PLC) and SCADA based system. The PCS will control the process interlocks and PID control loops for nonpackaged equipment. Control loop set-point changes for non-packaged equipment will be made at the OIT. In general, the plant process drives will report their ready, run and start pushbutton status to the PCS and will be displayed on the OIT. Local control stations will be located in the field in proximity to the relevant drives. These will, as a minimum, contain start and latch-off-stop (LOS) pushbuttons which will be hard-wired to the drive starter. Plant drives will predominantly be started by the control room operator, after inspection of equipment by an operator in the field. The OITs will allow drives to be selected to Auto, Local, Remote, Maintenance or Out-of-Service modes via the drive control popup. Statutory interlocks such as emergency stops and thermal protection will be hardwired and will apply in all modes of operation. All PLC generated process interlocks will apply in Auto, Local and Remote modes. Process interlocks will be disabled or bypassed in Maintenance mode with the exception of critical interlocks such as lubrication systems on the mill. Local selection will allow each drive to be operated by the operator in the field via the local start pushbutton which is connected to a PLC input. Remote selection will allow the equipment to be started from the control room via the drive control popup. Maintenance selection will allow each drive to be operated by maintenance personnel in the field via the local start pushbutton which is connected to a PLC input. A PLC output will be wired to each drive starter circuit for starting and stopping drives. Status indication of process interlocks as well as the selected mode of operation will be displayed on the OIT. Lycopodium Minerals Canada

322 NI TECHNICAL REPORT - RECOVERY METHODS Page Vendor supplied packages will use vendor standard control systems as required throughout the Project. Vendor packages will generally be operated locally with limited control or set-point changes from the PCS system. General equipment fault alarms from each vendor package will be monitored by the PCS system and displayed on the OIT. Fault diagnostics and troubleshooting of vendor packages will be performed locally. Vendor package control will be implemented for the following equipment: Pre-leach thickener rake mechanism. Gravity concentrators. Intensive cyanidation reactor. CIP carousel system. Carbon regeneration kiln. Flocculant mixing system. Water treatment plant. Compressed air system. Instrument air dryers. Oxygen plant. Tailings thickener rake mechanism. The use of actuated isolation or control valves will be implemented around the plant for automatic control loops or sequencing as part of the plant control or the elution sequence. All actuated valves and control valves will be operated from the OITs with remote position indication available. Automatic control valves will be controlled by PID loops within the PCS. The PCS will perform all digital and analogue control functions, including PID control, for all nonpackaged plant. Faceplates on the PCS displays will facilitate the entry of set-points, readout of process variables (PVs) and controlled variables (CVs) and entry of the three PID parameters (Proportional, Integral and Derivative). The majority of equipment interlocks will be software configurable. However, selected drives will be hard wired to provide the required level of personal safety protection e.g. the emergency stop buttons associated with each and every motor and the pull wire switches associated with conveyors. Lycopodium Minerals Canada

323 NI TECHNICAL REPORT - RECOVERY METHODS Page All alarm and trip circuits from field or local panel mounted contacts will be based on fail-safe activation. Alarm and trip contacts will open on abnormal or fault condition. If equipment shutdown occurs due to loss of mains power supply, the equipment will return to a de-energised state and will not automatically restart upon restoration of power. Sequential group starts and sequential group stops will not be incorporated for non-packaged plant equipment, with the exception of the elution circuit. However, in any process, critical safety and equipment protection interlocks will cause a cascade stop in the event of interlocked downstream equipment stopping (e.g. trip of SAG mill feed conveyor will result in stop of apron feeder). Standard vendor packages may include automatic sequence start / stop controls within the vendor package only Control System Configuration and Communications The process control system will be distributed throughout the plant, with a PLC installed in each of the following locations: Feed preparation. Grinding / screening / gravity circuit. Leaching / CIP. Elution / reagents / gold room Water services. East water dam. West water dam. East sediment pond. West sediment pond. Water storage facility. Tailings storage facility. The process plant PLCs will be interlinked via fibre optic cables and will all report back to the main control room. All field instrument and controls will be cabled back to their relevant switchroom utilizing field marshalling panels where appropriate. Owing to the site topography and location of waste dumps, remote pumping stations will utilise fibre optic communications installed as optical fibre ground wire (OPGW) onto the HV power lines. Lycopodium Minerals Canada

324 NI TECHNICAL REPORT - RECOVERY METHODS Page Drive Controls Drives will be powered from starters installed in a Motor Control Centre (MCC) switchboard located in the electrical substation. Each drive MCC will present a Running indication and a Fault alarm to the PLC system and will have provision for a PLC output contact for Process Interlocks. Variable Speed Drive (VSD) units will be of a Variable Voltage Variable Frequency (VVVF) type utilising Pulse Width Modulated (PWM) technology. The drive will be mounted in a free standing cubicle. The drive will be provided with an integral control panel for programming and operation at the VVVF unit for commissioning and emergency running. The starting of conveyors and rotating equipment, such as mills, will be preceded with a start siren. Interlocks will not be provided to stop large loads starting simultaneously. De-contactor connectors will be adopted for sump pumps. Sump pumps will have low current trip relays installed in the MCC. Lycopodium Minerals Canada

325 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page PROJECT INFRASTRUCTURE 18.1 Overall Site The overall site plan (shown in Figure 18.1) includes the open pit mine, waste dumps process plant, tailings storage facility (TSF), water storage facility (WSF), west and east water supply dams, sediment ponds, storage areas, buildings, power plant, bulk fuel storage, accommodation camps and main access road. The process plant, associated buildings, power plant and bulk fuel depot are located south of the open pit mine. The TSF is located east of the open pit. The accommodation camps are located close to the southern perimeter of the process plant for ease of personnel access. The main access road approaches the site from the east. The site as a whole will be fenced to clearly delineate the mine area, prevent animal access and deter access by unauthorized persons. Road access into the fenced area will be through a manned security checkpoint. Security fencing will surround the accommodation camp and general site infrastructure. High security fencing will surround the process plant. Lycopodium Minerals Canada

326 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page 18.2 Figure 18.1 Overall Site Plan Lycopodium Minerals Canada

327 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page Roads Access to Site Materials and consumables will be transported to site via a new laterite mine access road. The new access road will link the project site with the existing laterite road that connects to Route Nationale RN04, an all-weather bitumen road from Ouagadougou. Culverts will be installed as appropriate on creek crossings. Fill material for the road will be obtained from excavation alongside the road where possible. The main access road will be designed for a 60 km/h speed limit. Construction of the access road will commence as part of project early works and is expected to be fully completed prior to the 2017 wet season to enable continued all weather access to site for construction Project Site Roads Plant internal roads will provide access between the administration area, process plant facilities, bulk fuel storage, power plant, mine services area, and accommodation camp. These roads will generally be 6 m wide and will be constructed flush with bulk earthworks pads to ensure that storm water sheet flow is achieved across the site, thereby avoiding the need for deep surface drains and culvert crossings within the plant area Access Tracks Several new tracks will be constructed to access infrastructure such as the TSF, sediment ponds, and west and east water supply dams, and water bore pumps remote from the plant site. The access tracks will be cleared and graded natural earth tracks. Exact routes will be determined during construction of the Project to best fit local terrain and vegetation density Power Power Supply Due to the remote location of the Natougou site, power will be provided from a hybrid heavy fuel oil (HFO) and diesel fuelled power station located adjacent to the process plant. The power plant will supply 6.6 kv to the processing plant from which power will be distributed. Power will also be distributed to infrastructure outside the process plant via overhead pole lines. The power plant has been sized at 11.6 MW connected load to accommodate a peak load of 9 MW, and average running load of 6.4 MW with the following configuration: 2 x 3.8MW medium speed HFO units (CAT 8CM32C or equivalent machines). 5 x 1.6MW high speed diesel units (CAT 3516B or equivalent machines). Lycopodium Minerals Canada

328 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page 18.4 The two HFO units will satisfy the average running load and the diesel units will supplement the power required during start-up, peak load and when one of the HFO unit is offline for maintenance. One of the diesel units will be used for emergency backup. The SAG and tower mills at the process plant are the largest loads. The SAG mill has been specified with a variable speed drive and the tower mill with a fixed speed drive and LRS to provide a 'soft start' capability to reduce the load surge during start-up and minimize the need for 'spinning reserve' in the power station Electrical Distribution The electrical system is based on 6.6 kv distribution and 400 V, 50 Hz working voltage. The 6.6 kv feeder from the power plant will feed the site distribution 6.6 kv switchboard. For the process plant the 6.6 kv supply will be stepped down from 6.6 kv to 400 V at each switchroom using separate 6.6 kv / 400 V distribution transformers fed from the HV distribution board. The following switchrooms will be provided in the plant: Feed preparation area. Mill area. Leach, desorption, reagents and air services area. Water services area. Switchrooms will house 400 V motor control centres (MCCs), area VVVF drives, plant control system cabinets, plant lighting transformers, various distribution boards and UPS power distribution. 6.6 kv overhead power lines will provide power to various remote facilities (TSF pumps, bore pumps, WSF pump, water storage supply dam pumps, etc.). Pole mounted transformers will step down the voltage at each location and supply an outdoor 400 V switchboard local to each equipment area. The staff camp power will be supplied from a local MCC/transformer fed from the 6.6 kv overhead line Electrical Buildings Electrical buildings will be pre-fabricated flat pack panel buildings to minimize installation time on site. Buildings will be installed on a structural framework over 2 m above ground level to allow for bottom entry of cables into electrical cabinets. The electrical buildings will be installed with air-conditioners and suitably sealed to prevent ingress of dust. Lycopodium Minerals Canada

329 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page Transformers and Compounds All the 6.6 kv / 400 V distribution transformers will be of ONAN cooling configuration and vector group Dyn11. Fire rated concrete walls will be constructed around the pad mounted transformers Fuel Supply Bulk fuel supply will be provided by an onsite fuel storage facility and will store HFO and diesel for the power plant, mine trucks, light vehicles and users at the process plant. Day storage tanks are provided at the power plant and in the process plant. The bulk fuel storage facility will include approximately 30 days storage of HFO and diesel and includes 1,600 m 3 for diesel and 800 m 3 for HFO. Diesel fuel dispensing is also provided for the mine trucks and light vehicles Potable Water The potable water demand for the Project has been calculated on a per capita usage basis and is summarised in Table Table 18.1 Potable Water Usage Area Usage (L/person/day) No of Personnel m 3 /day (including contingency) Camps Plant/Mine Total A common potable water system will be provided for the accommodation camps and the process plant usage and will be located at the staff camp and distributed to the various users. A vendor package modular potable water treatment plant including filtration, ultra-violet sterilisation and chlorination will be installed. Water will be delivered via a reticulation system using a constant pressure variable flow pump system. The pump skid will include a UV disinfection unit to provide additional security against contamination. Lycopodium Minerals Canada

330 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page Sewage and Solid Waste Management Sewage Treatment Effluent from all water fixtures in the process plant, mine services area, accommodation camps and administration areas will be pumped to a common sewage treatment plant vendor package located near the WSF. Treated effluent will be discharged to the WSF. Treatment plant sludge will be suitable for direct landfill burial Solid Wastes Wastes will be sorted and reused or recycled as much as possible. Waste lubricating oils and general non-hazardous solid wastes will be incinerated via on onsite incinerator. Dangerous or hazardous waste will be collected and stored briefly before being transferred to a suitable permitted facility, either on-site or off-site depending on the specific materials and requirements Accommodation Camps Staff and technician accommodation camps will be provided located to the south of the process plant and will provide accommodation for International Expatriate, senior staff and African national staff not originating from the local area. The staff camp will consist of the following: 13 x blocks consisting of 4 single rooms with en-suite bathrooms. 2 x blocks consisting of 2 single rooms with en-suite bathrooms. Dining/recreation room. Fitness centre. Laundry. Office and storage The technician s camp will consist of the following: 8 x blocks consisting of 4 double rooms. Common ablution block. Mess hall. Lycopodium Minerals Canada

331 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page 18.7 Laundry. Soccer field. A 150 person construction camp will be built for the peak load workforce levels. This camp will use the common facilities in the staff camp. There will be a variety of room types for senior staff and trades level staff Mine/Plant Site Facilities General Site buildings will be 'fit for purpose' industrial type structures. Workshops, warehouses and reagent storage sheds will be constructed of a concrete slab on ground with structural steel frame and metal cladding. Offices and amenity buildings will be erected on site and consist of concrete blockwork or brickwork Outside Plant Area The following building and facilities will be provided outside the plant area: Laboratory (dry and wet). Main warehouse and office. Outdoor storage area for consumables. Administration building. Mining/engineering office (for Birimian staff). Medical centre. Main mess and kitchen. Exploration office. CRS building. Long term reagents storage. Main gatehouse. Lycopodium Minerals Canada

332 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page Inside Plant Area The following building and facilities will be provided inside the plant area: Plant office with control room. Plant workshop and maintenance. Plant workshop yard. Plant mess. Male and female ablutions. Short term reagent storage. Plant gatehouse Mine Services Mine services will be provided by the mining contractor and is anticipated to include the following buildings and facilities: Mine truck workshop. Mine warehouse. Mine office. Truck-wash down facility Process Plant Geotechnical A geotechnical investigation was conducted to assess the founding conditions at the proposed process plant location. This investigation included drilling of rotary cored boreholes, test pit excavation, in-situ testing and laboratory testing. The ground conditions generally comprised a thin topsoil cover overlying residual soils and extremely weathered rock, grading into distinctly weathered, slightly weathered and fresh rock. The residual deposits were typically encountered as silt, clay and gravel in varying proportions. The extremely weathered rock was encountered as saprolite in the boreholes, becoming distinctly weathered, slightly weathered and fresh amphibolite rock with depth. The depth of weathering at the plant site was relatively shallow (Knight Piésold, 2016). Lycopodium Minerals Canada

333 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page Water Supply Infrastructure Water demands A water balance model was prepared to estimate the demand for raw water on site, considering the process water demand, losses and gains from the tailings storage facility, pit dewatering, dust suppression, potable water demands and runoff from the ROM pad and plant site. Utilisation of ground water resources from boreholes were incorporated into the model. Any potential shortfall was modelled to be supplied from the WSF which will be fed from two water harvesting facilities located to the east and the west of the process plant. A detailed water balance modelling section is incorporated in Section 6.0 of the Tailings Storage Facility and Site Water Management Report prepared by Knight Piésold (Knight Piésold, 2016). The total water demand for the site was estimated at between 1.1 and 1.4 Mm 3 per year. The water demand for the process plant amounts to 0.75 Mm 3, which includes the process raw water requirement of 0.08 Mm 3 but excludes water in ore. Other water demands include a provision of between 0.2 Mm 3 and 0.5 Mm 3 for dust suppression and wash down water and 0.04 Mm 3 for potable water requirements. The demand will be met from TSF decant, pit dewatering (including precipitation on the pit area), runoff from the ROM pad and plant site and sediment impacted runoff collected in the sediment control ponds. The balance of the water demands will be made up of raw water harvested from the groundwater and the surface water sources. The site has a negative water balance during early years of operation but as the pit and tailings storage facility extents increase, the amount of rainfall runoff captured at these locations increases and the project transitions to a positive water balance reducing the reliance on water harvesting from the surface water sources. The critical years for the project are the year prior to commissioning the plant and the first year of production when the pit area is relatively small and significant surface water harvesting is required to fill the WSF to provide an inventory of water for processing requirements. The water balance was run utilizing 27 years of daily rainfall records sourced from Fado N Gourma and also run using synthetic average, extreme wet and dry rainfall sequences. The water balance model indicates that the proposed water supply system provides sufficient water to meet the process water requirements under climatic conditions observed in all historical years, synthetic average and extreme dry conditions up to 1 in 100 year average recurrence interval. The water balance model also indicates that the proposed water supply system provides sufficient water for dust suppression requirements under most conditions. However when the model was run under the synthetic 1 in 100 year dry conditions a shortfall for dust suppression occurs in the first year of operation (13,000 m 3 ). Additionally when the model was run using historical records a shortfall for dust suppression occurs in 4 out of the 27 years modelled ranging from between 4000 m 3 and 225,000 m 3. In the years where the shortfalls occurred rainfall runoff was estimated in the model to be less than 1.5% of rainfall. The shortfalls can be mitigated by utilising a commercial dust suppression product to reduce water application requirements on the haul roads, access roads and around the infrastructure areas. Lycopodium Minerals Canada

334 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page Sufficient time will be available to instigate these changes to dust suppression practices on site as the WSF volume at the end of the previous wet season can be used to determine the requirement to switch to dust suppression additives or to conduct an investigation to source additional groundwater supply. The water balance schematic flow diagram is provided in Figure Decant from Tailings Storage Facility The water balance modelling indicates that for tailings pumped to the TSF at the design solids (62% to 64%), the pond on the TSF would increase during the wet season and reduce to the minimum pre-set level during the following dry season. Recovery from the TSF decant would gradually increase to supply up to 73% of the process water demand on an annual basis, or up to 0.5 Mm 3 per year in the later years of operation. The maximum pond volume that would develop under average conditions would be 0.3 Mm 3, which would increase to 0.5 Mm 3 in the event of a 1 in 100 year wet year occurring in the final year of operations. The TSF pond volume recovers to minimum levels within one year of a 1 in 100 year average recurrence interval wet year. Lycopodium Minerals Canada

335 NI TECHNICAL REPORT PROJECT INFRASTRUCTURE Page Figure 18.2 Water Balance Schematic Flow Diagram Lycopodium Minerals Canada

336 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page Pit dewatering The ground water inflow to the pit is expected to be very low and contribute insignificant water volumes. However as the pit has a large aerial extent of approximately 100 hectares, precipitation will accumulate in the pit during the wet season which will be pumped to the raw water pond for use in the process circuit. Under average conditions the pit dewatering is estimated to be approximately 0.58 Mm 3 when the pit reaches its full extent increasing to approximately 0.86 Mm 3 under 1 in 100 year wet conditions Groundwater sources Golder Associates conducted two regional ground water investigations (Report Number R-Rev1 & Rev0) to identify and quantify the volume of water which can be sustainably extracted from ground water resource areas in proximity to the project. Three ground water resource areas were identified which have been estimated to supply a combined flow of 0.06 Mm3/year over the life of the project. This water will preferentially be directed to meeting potable water supplies with any surplus directed to meet process plant raw water requirements Surface water sources and storages The balance of the water demands at site will be met from two creeks which are located to the east and west of the process plant and water which will be harvested from the sediment ponds located around the site. The location of the surface water sources and storages are provided on Figure 18.3 with details of these facilities provided in the following sections. Lycopodium Minerals Canada

337 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page Figure 18.3 Location of Surface Water Sources and Storages East Water Supply Dam The East Water Supply Dam will be constructed approximately 1.5 km to the north east of the processing plant. The mean annual runoff at the dam site is estimated to be 0.95 Mm 3 from a catchment area of 1902 ha. Of the total runoff estimated at this dam, a minimum social / environmental release rate of 25% of inflow has been set for the facility capping the amount of water that can be harvested from the facility to an estimated 0.71 Mm 3 per year under average conditions. The flat topography at the dam site would result in significant evaporation and seepage losses should this location be used to store water for extended periods of time and therefore a supplementary water storage facility, the WSF has been provided to store water more efficiently. The east water supply dam has been designed as a cross valley embankment of 360 m in length with a maximum embankment height of 8.4 m (with a 3m depth allowance for spillway flow) to provide a storage capacity of 0.51 Mm 3. The embankment will be constructed as a low permeability embankment with a filter drain within the downstream portion of the embankment and an erosion resistant rock fill facing on the upstream and downstream batter slopes. Lycopodium Minerals Canada

338 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page The facility is equipped with a spillway with the capacity to pass the peak flow from a 1 in 200 year ARI critical duration storm event and a water recovery decant tower to allow water to be pumped to the Raw water pond at a rate of 400 m 3 /hr. West Water Supply Sump The West Water Supply Sump will be constructed approximately 2.0 km to the west south west of the processing plant. The mean annual runoff at the dam site is estimated to be 1.15 Mm 3 from a catchment area of 2262 ha. Of the total runoff estimated at this dam a minimum social / environmental release rate of 25% of inflow has been set for the facility capping the amount of water that can be harvested from the facility to an estimated 0.86 Mm 3 per year under average conditions. The West Water Supply Sump has been designed as a below ground sump to reduce the potential for flooding of agricultural land on the flood plain of the creek. The sump will be 6 m deep with crest dimension of 180 m by 100 m with a capacity of 100,000 m 3. A small elevated access road from the valley flank to the sump has been designed to allow all weather access to the sump. The facility is equipped with a 40 m wide spillway to pass small flows equivalent to the capacity of the defined creek channel downstream of the facility. Larger flows from rare events will result in flooding of the floodplain within the valley and the water level on the flood plain will exceed the crest of the facility i.e. the facility will be fully inundated. Water will be recovered from the facility and pumped to the Raw Water Pond for use in the process plant with the pump designed with a peak pumping capacity of 1000 m 3 /hr. Sediment Ponds The East and West sediment ponds have been included as potential water supply sources in the water balance model. The northern sediment pond is not equipped with a pump or water recovery infrastructure as the discharge from the pond can be harvested from the West Water Supply Sump which is located downstream of this sediment pond. The East Sediment Pond will have a capacity of 100,000 m 3 with a 480 m long embankment with a maximum design height of 4 m. The facility has been designed with a spillway capable of passing the peak flow from a 1 in 100 year ARI critical duration storm event. Water will be recovered from this facility at a rate up to 150 m 3 /hour and pumped to the WSF when water is required by the project; alternately, it will be allowed to overflow to the natural creek. The West Sediment Pond will have a capacity of 75,000 m 3 with a 470 m long embankment with a maximum design height of 3.5 m. The facility has been designed with a spillway capable of passing the peak flow from a 1 in 100 year ARI critical duration storm event. Water will be recovered from this facility at a rate up to 150 m 3 /hour and pumped to the WSF when water is required by the project; alternately, it will be allowed to overflow to the natural creek. Lycopodium Minerals Canada

339 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page Water Storage Facility (WSF) The Water Storage Facilities (WSF) will be the primary water storage facility on site. It has been designed with a capacity of 1.15 Mm 3 and will receive and store water from the east and west water supply dams/sumps, the sediment ponds and pit dewatering. The facility has been designed as a balance cut to fill facility with the 1500 m long embankment varying in height from 5 m to 10 m and the excavation depth varying from 0.3 m to 7 m. The facility will be constructed as a single zone low permeability embankment with the upstream batter and basin lined with a 1 mm HDPE liner to minimise seepage losses from the facility and reduce the potential for wetting up of the embankment. Water will be recovered from the facility via a decant tower which will be accessed via a causeway constructed from the north western corner of the facility adjacent to the process plant. As the WSF has no external catchment and all inputs to the facility are pumped, the pond is not equipped with a spillway rather the maximum operating level shall be set at 750 mm below the defined crest RL of the embankment which will provide sufficient storage capacity to accommodate a 24 hour probable maximum precipitation event without overtopping Tailings Storage Facility Design objectives and site selection A tailings storage facility (TSF) with a capacity to store 10Mt of tailings generated by the ore processing operations is required for the life of the project with tailings being produced at a rate of 1.34 Mtpa. The tailings storage facility design is described in detail in the Tailings Storage Facility and Site Water Management Report prepared by Knight Piésold (Knight Piésold, 2016). The principal objectives for the design of the TSF for the Natougou Gold Project are as follows: To produce a design complies with relevant international regulations, guidelines and standards using best available technology. Permanent, secure and total containment of all tailings within an engineered disposal facility. Achieve a high density, consolidated tailings mass by employing controlled subaerial deposition of tailings. Control, collect and remove free draining liquids from the tailings during operation for recycling as process water to the maximum possible extent. Maintain excess storage capacity within the tailings storage facility to contain the design storm allowance. Lycopodium Minerals Canada

340 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page Provide an engineered structure to control discharges from the storage due to extreme events greater than the design storm allowance. To reduce seepage from the facility during operation and on closure. Cost-efficient utilisation of available material for embankment construction. Allow for rapid, effective and stable rehabilitation. Provide monitoring features for all aspects of the facility and associated works to ensure adopted performance standard can be measured and achieved. Several phases of site selection studies were undertaken assessing a total of eleven sites with the potential to store tailings generated from the project. The sites evaluated were located within a radius of 11 km from the processing plant and included facilities utilized available topography and standalone paddock facilities. The preferred site selected for the project is located 800 m to the north east of the processing plant. The location of the preferred site is provided in Figure Figure 18.4 Location of Tailings Storage Facility Lycopodium Minerals Canada

341 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page Facility design The tailings storage facility will require a single embankment along the south and western extents of the facility with a total embankment length of 1665 m and with a maximum embankment height of 23.6 m at the south west corner. The eastern and northern margins of the facility are confined by a natural laterite ridge line and therefore no supporting embankment is required along these margins. The tailings beach surface at full capacity will cover an area of approximately 76.5 hectares. A geotechnical investigation was conducted at the TSF site which included geotechnical drilling, test pit excavation, in-situ testing and laboratory testing of the foundation and construction materials. Full details of the geotechnical investigation are included in Knight Piésold Geotechnical Investigation report of the revised TSF location (Knight Piésold, ). The investigation revealed that the TSF site in underlain by high strength amphibolite rock with overlying residual soils of depth varying from 0.5 m in the centre of the basin to 16.8 m below the laterite ridge line. The residual soils were generally firm to stiff sandy and gravelly clays and silts. No groundwater was encountered during the investigation. The site is considered to have a low seismic hazard rating with peak ground acceleration with an average recurrence interval of 1 in 475 years of 0.02g based on the Global Seismic Hazard Map (GSHAP, 1999). Tailings will be pumped to the TSF as a slurry at 62% to 64% solids and will be deposited sub-aerially to facilitate drying and consolidation of the tailings mass. Two tailings samples, one representing the first 12 months of production and the second representing the remainder of the life of mine production were provided to Knight Piésold for testing. The grading and Atterberg tests indicate that the tailings can be classified as non-plastic silt with sand. The sample consisted of approximately 80% silt, 14% fine sand and 6% clay. The tailings tests indicated that the tailings are expected to settle and release water rapidly. It is expected that in the TSF a tailings dry density of approximately 1.27 t/m 3 will be achieved initially, increasing to 1.52 t/m 3 to give an overall average density of 1.48 t/m 3 at the end of operations. Water release from the tailings is expected to be moderate at between 30% and 37% of the water in slurry. Geochemical testing of the two composite tailings samples was also conducted. The testing comprised acid-base accounting, multi-element analysis, short term leach and kinetic testing. Complete results for the geochemical testing program are provided in Knight Piésold s Waste Rock and Tailings Geochemical Characterization Report (Knight Piésold, 2015). The tailings were found to be non-acid forming but were highly enriched in arsenic which was soluble under the ph conditions anticipated in the TSF. The arsenic concentration in the kinetic tests peaked at 2.3 mg/l. As a result of the high arsenic in the tailings solids and supernatant a robust seepage control system comprising an above liner underdrainage system, a geomembrane liner overlying a compacted in-situ low permeability subbase and sub liner seepage recovery drains have been included in the design. The layout of the underdrainage system and sub-liner drainage network is provided in Figure Lycopodium Minerals Canada

342 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page Figure 18.5 Layout of Underdrainage and Sub-liner Drainage Network Lycopodium Minerals Canada

343 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page A dam consequence category assessment has been conducted for the TSF based on the ANCOLD Guidelines (ANCOLD, May 2012 & ANCOLD October 2012) to assign a dam failure consequence category and an environmental spill consequence category. The assessment classified the TSF as a High B facility based predominantly on the adverse geochemistry of the tailings and the potential impact on the downstream environment should a dam failure event occur. Based on the consequence category the TSF is designed with a wet season storage allowance of a 1 in 100 year average recurrence interval wet season plus an extreme storm allowance of a 1 in 100 year average recurrence interval 72 hour storm event. Spillways have been provided in the design of the facility for each stage of construction with the spillways designed to pass the peak flow generated from the critical duration probable maximum precipitation event. The embankments of the facility have been designed to allow for deployment of a geomembrane liner over the full basin area and on the upstream face of the embankment and therefore the upstream embankment batter slopes has been maintained at a low angle of 18.4 (1V:3H). The natural ridge line which provides the containment along the northern and eastern margins of the facility was found to have a natural slope which would be too steep to allow the slope to be prepared for liner deployment and therefore a wedge of fill has been designed to form a fill platform along these margins so that the liner can be installed. The embankment will be raised in five stages over the life of the project by downstream construction methods with the embankment comprising a downstream structural zone constructed of competent ROM waste rock, a transition zone comprising a gravelly weathered waste rock and a low permeability embankment zone below the liner constructed from residual soil won from the open pit and local borrow areas. The downstream batter slope of the embankment will be constructed at a slope of 1V:2H during operation and flattened so that the final batter slopes will be 1V:3H on closure. The embankment geometry is provided in Figure Figure 18.6 Tailings Storage Facility Embankment Lycopodium Minerals Canada

344 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page Geotechnical analysis of the facility has been conducted to assess the geotechnical stability of the facility and the seepage rates from the facility. The stability assessment indicates that the embankment has a factor of safety in excess of 1.5 under both drained and undrained conditions with the post seismic stability of the facility also over 1.5 assuming full liquefaction of the stored tailings. The seepage analysis of the facility indicates that the embankment will be fully drained with a seepage rate from the facility of approximately 2.2 m 3 per hectare per day at the final stage of operations. The tailings will be deposited off the embankment at the southern and western margins of the facility and from the fill platforms along the northern and eastern margins of the facility. This will position the supernatant pond within the centre of the facility and away from the embankments. A decant structure will be constructed at the centre of the facility comprising a recovery tower and access causeway to allow supernatant water to be recovered from the facility and recycled to the process plant. The design of the facility includes for the installation of standpipe piezometers, shallow and deep monitoring bores and survey pins to allow the performance of the facility to be monitored. An operating manual will be prepared for the facility and regular audits conducted by a specialist tailings design company to ensure that the facility is being operated in accordance with the design intent and the facility is performing as intended. The geochemistry of the tailings which will be deposited in the facility requires a robust closure cover to minimise rainfall infiltration into the facility and reduce the potential for seepage from the facility post closure. The closure design therefore includes a geomembrane liner placed over the tailings overlain by a low permeability soil layer to further reduce infiltration and protect the liner. This liner system will be covered by a 1m layer of waste rock and topsoiled to promote vegetation establishment. On closure a closure spillway will be constructed to drain the top surface of the facility with the spillway designed to pass the peak flow generated from the critical duration probable maximum precipitation event Construction Staging and Summarized quantities The timing for the construction stages and the major quantities which will be required to construct the facility are provided in Table These quantities have formed the basis of the capital and operating costs provided in Section 21 of this report. Lycopodium Minerals Canada

345 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page Table 18.2 TSF Construction Staging and Major Quantities Stage Construction Date Structural Embankment Fill / Cover (m 3 ) Transitional Embankment Fill (m 3 ) Low Permeability Fill / Cover (m 3 ) Geomembrane Liner (m 2 ) Stage 1 May 2017 to Apr ,000 22,000 30, ,000 Stage 2 Jan 2019 to May ,000 12,000 14, ,000 Stage 3 Jan 2020 to May ,000 18,000 19, ,000 Stage 4 Jan 2022 to May ,000 11,000 13,000 64,000 Stage 5 Jan 2024 to May ,000 6,000 6,000 27,000 Closure May 2026 to Aug , , ,000 LOM Totals* 2,732,000 66, ,000 1,615,000 * Note totals do not add up due to rounding of staged quantities Surface Water & Sediment Management The surface water management at the site incorporates the establishment of suitable surface water control measures in order to reduce impacts to downstream environments for all aspects of the Project, from initial development through to completion of rehabilitation. The Surface Water Management designs are described in detail in the Tailings Storage Facility and Site Water Management Report prepared by Knight Piésold (Knight Piésold, 2016) Anticipated water quality A geochemical assessment has been conducted for the waste rock which will be generated as part of mining operations at the project to assess the chemistry of the runoff and seepage waters from the waste dumps. Complete results for the geochemical testing program are provided in Knight Piésold s Waste Rock and Tailings Geochemical Characterization Report (Knight Piésold, 2015). The testing comprised acid-base accounting, multi-element analysis, short term leach and kinetic testing with a total of 130 samples of waste rock tested. The geochemical testing indicates that the waste rock at the project has a low potential to generate acid with 97% of the samples tested found to be non-acid generating. The waste rock was found to be variably enriched in arsenic with approximately one third of the samples considered to have elevated concentrations of arsenic. The short term leach testing, multi-element analysis and kinetic testing of the waste rock indicate that the arsenic concentration in the undiluted runoff or seepage water from majority of the waste rock (~70%) will meet WHO drinking water standard (WHO 2011), Burkina Faso drinking water standard and the Burkina Faso standard for discharge of waste water to surface water. The undiluted runoff or seepage water from approximately 30% of the waste is unlikely to directly meet drinking water standards but the majority should meet the Burkina Faso standard for discharge of waste water to surface water. Lycopodium Minerals Canada

346 NI TECHNICAL REPORT - PROJECT INFRASTRUCTURE Page A surface water and sediment management system was therefore designed for the project to capture surface water running off the waste dumps to remove sediment and to allow the water quality to be tested and either recycled to the process plant or released should it meet release guidelines when the water is not required for process water make-up Management system design Three major sediment ponds have been designed to capture runoff from the waste dumps with an upstream clean water diversion designed to carry non impacted water around the site. The sediment ponds and surface water diversion are shown in Figure Figure 18.7 Surface Water Management System The sediment ponds have been designed to remove sand and coarse silt sized particles from the peak flows generated under a 1 in 10 year average recurrence interval event. Each sediment pond is equipped with a spillway capable of passing the peak flow from a 1 in 100 year average recurrence interval. The clean water diversion has been designed to pass the peak flow from a 1 in 100 year average recurrence interval event with the sediment impacted drains designed to pass the peak flow from a 1 in 50 year average recurrence interval event. Lycopodium Minerals Canada

347 NI TECHNICAL REPORT - MARKET STUDIES AND CONTRACTS Page MARKET STUDIES AND CONTRACTS 19.1 Market Studies No formal market studies have been undertaken. A gold price of USD 1,100/oz from Birimian has been used for the Mineral Reserve estimate and the economic analysis. Natougou will produce gold doré which is readily marketable on an 'ex-works' or 'delivered' basis to a number of refineries in Europe and Africa. There are no indications of the presence of penalty elements that may impact the price or render the product unsalable. SEMAFO has contracts in place for sale of gold from its producing mine, Mana, in Burkina Faso and it has been assumed that Natougou doré will attract similar terms Contracts There are no material contracts or agreements in place as of the effective date of this report. Refining contracts are typically put in place with well recognized international refineries and sales are made based on spot gold prices. These contracts typically include fees for transportation of the product from the site, insurance, assaying, refining and an allowance for metal losses during refining. The ability to get a contract in place for the sale of doré prior to start of production is not seen as a risk to the Project. Lycopodium Minerals Canada

348 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT The NGP is administratively part of the Partiaga department in the province of Tapoa which is part of the Eastern Region of Burkina Faso. The project site is located approximately 320 km east of Ouagadougou, the country s capital, and is easily accessible by road throughout most of the year. The approach developed by Birimian throughout the various environmental and social studies conducted from June 2015 to April 2016 for the Environmental and Social Impact Assessment ( ESIA ) took into consideration the social and environmental concerns of all interested parties. These concerns were then integrated into the initial stages of the project design. This approach maximized the project s integration into the environment and has minimized its negative impacts to increase the environmental and social acceptability of the project. In addition, this approach has allowed a better understanding of the social aspects arising from the resettlement of households due to the presence of the mine facilities Legal Framework and Permits to Obtain Burkina Faso has a regulatory framework for environmental and social management. The relevant Burkina Faso policies, laws and regulations pertaining to mine development were taken into account for the ESIA Policies and Strategies for Environmental Protection Since the early 1990s, Burkina Faso has developed numerous policies and strategies for the management of their natural resources. A declaration of Mining Policy was formulated in 1995 that highlighted the importance of the private sector as an engine of economic development. Other policies on environmental protection include the following: Strategy for Accelerated Growth and Sustainable Development (SCADD: Stratégie de Croissance accélérée et de Développement durable); Government Program for Emerging and Sustainable Development (PAGEDD: Programme d Action du Gouvernement pour l Émergence et le Développement durable, ); Rural Development Strategy (SDR : Stratégie de développement rural 2015); National Policy on Environmental Matters (PNE: Politique nationale en Matière d Environnement); WSP Canada

349 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page 20.2 Environmental Plan for Sustainable Development Program (PEDD: Plan d Environnement pour le développement durable); National Policy on Rural Land (PNSFMR: Politique Nationale de Sécurisation Foncière en Milieu Rural); and National Action Program for Adaptation to Climate Variability and Change (PANA: Programme d action national d adaptation à la variabilité et aux changements climatiques) Legal Framework The Burkina Faso legal framework with respect to environmental and social aspects related to economic activities is supported by a number of laws and decrees: Environmental Code (Code de l environnement). Mining Code (Code minier). Forest Code (Code forestier). Public Health Code (Code de la santé publique). General Local Authorities Code (Code général des collectivités territoriales). Act on Rural Land Tenure (Régime foncier rural). Act on Agrarian and Land Reorganization (Reorganisation agraire et fonciere). Law on Water Management (Loi d'orientation relative à la gestion de l'eau). Act on Pastoralism (Loi d orientation relative au pastoralisme). Other relevant regulations include: Decree No /PRES/PM/MCE/MECV/MATD dated 26 December 2007 on specific environmental regulations for the exercise of mining in Burkina Faso. Decree No /PRES/PM/MAHRH/MECV/MRA dated 6 December 2006 on the protection of aquatic ecosystems. Decree No /PRES/PM/MAHRH/MECV/MPAD/MFB/MS dated 6 December 2006 determining the perimeters of protection for water bodies and streams. WSP Canada

350 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page 20.3 Decree No /PRES/PM/MEE dated 17 July 2001 on the scope, content and procedure for Environmental Impact Assessment Study and Environmental Impact Instruction. Decree No /PRES/PM/MEE dated 7 May 2001 on setting standards for discharges of pollutants into the air, water, and soil Mining Code The Mining Code (Law N /CNT pertaining to the Mining Code of Burkina Faso) is administered by the Ministère de l Énergie, des Mines et des Carrières1 ( MME ) and provides the legal framework for the mining industry in the country. The state owns title to all mineral rights and these rights are acquired through a map-based system by direct application to the MME. The Mining Code guarantees a stable fiscal regime for the life of any mine developed. It also guarantees stabilization of the financial and customs regulations and rates during the operational period to reflect the rates in place at the date of signing. The Mining Code also states that no new taxes can be imposed with the exception of mining duties, taxes and royalties. However, the title holder can also benefit from any reductions of tax rates during the life of the operating license. There are three types of mining permits and three types of authorizations: Mining permits: - Research Permit. - Industrial Operating Permit. - Semi Mechanized Small-Scale Operating Permit. Authorizations: - Prospecting Authorization. - Traditional Artisanal Mining Authorization. - Quarrying Authorization. Details pertaining to the above permits can be found in Section herein. 1 Ministry of Energy, Mines and quarries WSP Canada

351 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page 20.4 Adopted in June 2015 by the National Transitional Council, the new mining code complies with Economic Community of West African States (ECOWAS) and the community mining policies of the West African Economic and Monetary Union (UEMOA from its French designation, Union économique et monétaire ouest-africaine) by integrating a number of provisions for a better contribution of mining projects to the country's economic development. In particular, the new Mining Code establishes the creation of a mining fund for local development to fund municipal and regional development plans, to which the mining license holders will contribute up to 1% of their monthly turnover or the value of extracted products during the month. Like the preceding Mining code, the new code requires mining projects to conduct an Environmental and Social Impact Assessment (ESIA) and to meet the Environmental Code requirements. Other relevant changes in the new mining code include: the removal of the mining agreement for the research phase; application of a specific transaction tax on mining titles; the possibility of suspending or removing mining titles or authorizations without notice when required by public order. The new mining code provides for the adoption of decrees relating to the implementation of various provisions of this code. No such decree has been adopted up to and including the date of this Report Institutional Framework The main institutional stakeholders for environmental matters include: Ministry of Environment, the green economy and climate change (MEEVCC: Ministère de l Environnement, de l économie verte et du changement climatique). Ministry of Energy, Mines and quarries(mme: Ministère de l Énergie, des Mines et des Carrières). Office of Mines and Geology (BUMIGEB: Bureau des Mines et de la Géologie du Burkina). Chamber of Mines of Burkina Faso (CMB: Chambre des mines du Burkina). National Commission of Mines (Commission nationale des Mines). National Council for the Environment and Sustainable Development (CONEDD: Conseil national pour l Environnement et le Développement durable). National Bureau of Environmental Assessment (BUNEE: Bureau national des Évaluations Environnementales): this organization is part of MEEVCC and has the mandate to promote, regulate, and manage the environmental assessment process of the country. BUNEE holds sessions to review the terms of reference submitted by the project promoter. It formulates an opinion on the admissibility of studies and makes recommendations to MEEVCC on the environmental acceptability of projects. WSP Canada

352 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page 20.5 Technical Committee on Environmental Assessments (COTEVE: Comité technique sur les Évaluations Environnementales): this organization was created by Decree No /MECV/CAB in 19 May 2006 establishing the powers, composition, and functioning of COTEVE. COTEVE is the technical and scientific framework to examine and analyse research reports and notices of environmental impacts presented by the project promoters to MEEVCC. Other Ministries and Departments involved: - Department of Infrastructures (Ministère des Infrastructures). - Ministry of Territorial Administration, Decentralization and Homeland Security (Ministère de l'administration territoriale, de la Décentralisation et de la Sécurité). - Department of Health (Ministère de la Santé). - Department of Agriculture and Food Security (Ministère de l Agriculture et de la Sécurité Alimentaire). - Department of, Hydraulic Resources and Sanitation (Ministère des Ressources Hydrauliques et de l Assainissement). - Department of Animal and Fishery Resources (Ministère des Ressources Animales et Halieutiques). - Ministry of Women, National Solidarity and Family (MFSNF: Ministère de la Femme, de la Solidarité Nationale et de la Famille) Permits to Obtain The application for an Operating permit requires a Feasibility Study (FS) that must first be accepted by MEEVCC. The FS must include an ESIA, which must include a Resettlement Action Plan (RAP) that has been accepted by all stakeholders. Once in production, a mining permit holder is required to open under his name, a Trust account named Fonds de préservation et de réhabilitation de l environnement minier1 at the Banque Centrale des États de l Afrique de l ouest 2 ( BCEAO ). This account must be funded annually on January 1 by an amount equal to the total rehabilitation budget presented in the ESIA, divided by the number of years of expected production to cover the costs of mine reclamation, closure and rehabilitation. The required permits and administrative procedures are presented in Table Fund for the Preservation and the Rehabilitation of the Mining Environment (decree N /PRES/PMJMCE/MEF). Central Bank of West African States WSP Canada

353 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page 20.6 Table 20.1 List of Required Permits Permit / Authorization Main Requirements Timeframe Cost Environmental Compliance Certificate Permit for Industrial Exploitation Delivered by the MEEVCC on the basis of a compliant ESIA Report: Submission of the ESIA Terms of Reference (ToR) to the BUNEE Validation of the ToR Submission of the draft version of the ESIA to the BUNEE Public Inquiry COTEVE Session Final ESIA Report approved by the BUNEE Request to the MME specifying the type of industrial license being sought. References of the exploration license under which or from which the request is being made. The mineral substances or for whom or which the license is sought. Definition of coordinates of the perimeter and requested area. Location perimeter license requested on a topographic map at 1:200,000. Detailed plan at an appropriate scale where the coordinates of the perimeter requested are attached to outstanding points invariable ground and well defined. Detailed statement showing the results of research conducted and evidence of expenditures incurred during the last period of license validity. FS and Development Plan and operation of the deposit, including an ESIA and public inquiry results. Surface plan specifying land reserved for the operation and installation of industrial units. Commitment of the applicant, in the case of a large mining operations, the allocation of 10% of the share capital of the company operating free of all charges, to the benefit of the state. Mining Convention that the applicant intends to sign with the State; agreement consistent with the mining contract and attached to the operating license required. Before the expiry of the Exploration License 90 days before expiry of the Exploration Permit Cost for review and validation of the ToR: 500,000 FCFA. Fees for the public inquiry and COTEVE session: Approximately 3 million FCFA. Cost for file processing-project value of 50 billion FCFA and more: Flat Fee: 25 million FCFA. Proportional rights: 0.02% of the total investment cost = 26.3 million FCFA (assuming a project cost of about billion FCFA). Costs for the National Commission of Mines session: about 2 million FCFA. Fixed duties on mineral titles: 5,000,000 FCFA. Proportional rights (area taxes): First five years: 7,500,000 FCFA/ km²/y. Year 6 to10: 10 million FCFA/ km²/y. From the 11thyear: 15 million FCFA/ km²/y. Cost of proportional royalty on gold production: 5% of turnover if gold is USD 1,300 and higher, 4% if USD 1,000-1,300 and 3% if less than USD 1,000. WSP Canada

354 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page 20.7 Permit / Authorization Main Requirements Timeframe Cost Authorization for the Management of raw water Authorization for the Collection of raw water for civil work Authorization for Road infrastructures Authorization for Hydraulic work or dam Application to the Directorate of Legislation and regulations of the Ministry in charge of water. No license or authorization required (besides the Mining Licence) However, tax must be paid. Authorization from the Ministry of Transport on the basis of technical studies. Authorization from the Ministry in charge of water on the basis of technical studies. 30 days after the quarter of the levy Tax of raw water for mining and industrial purposes: 200 FCFA /m³ a FCFA /m 3 for every m 3 backfillplaced. 20FCFA /m³ for every m³ of concrete poured. --- Depending on the infrastructure and technical studies that have been conducted. --- Depending on the work and technical studies that have been conducted. Notes: a) Currently there is a Water Tax in Burkina Faso, however the Chamber of Mines, the Ministry of mines, the Ministry of Finances, and the Ministry of Water have not yet arrived at a consensus as to how to apply this to the Mining Sector. As a result mining companies have negotiated an exoneration of the water tax until the construction costs of the water reservoir are covered. For the Feasibility study, this water tax has been left out of the operating costs, as Birimian is exempt because there is no source of water to pump. WSP Canada

355 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Baseline Studies Baseline Studies Conducted A number of baseline studies have been conducted from 2013 and 2015 in order to fully document the sensitive environmental and social components of the project area. These studies are listed below in Table 20.2 including when the study was done and the leading consultant. Table 20.2 List of environmental and social baseline studies conducted as part of the ESIA process Study Time Frame Lead Consultant Physical environment Waste rock and tailings geochemical characterisation (7 different studies) Knight Piésold Soils December 2014 INGRID Baseline design climatology assessment May, June 2014 Knight Piésold Preliminary air quality baseline study July 2014 INGRID Groundwater August 2014 and March 2015 Golder Associates Groundwater levels and additional sampling June, July 2015 WSP Surface water study (flow measurements) August 2015 WSP Air quality baseline study June, July 2015 WSP Soils October 2015 SOCREGE Biological environment Preliminary Fauna baseline study January, June and November 2014 INGRID Preliminary Flora baseline study December 2013 and August 2014 INGRID Preliminary aquatic ecosystems baseline study November 2014 INGRID Inventory of trees within the project footprint June and July 2015 SOCREGE Ecosystem baseline study July 2015 SOCREGE Aquatic Biodiversity (Flora) October 2015 SOCREGE Human Environment Preliminary archaeological survey 2014 SOCREGE Preliminary social baseline study February 2014 INGRID Demographic and socioeconomic surveys June 2015 SOCREGE Inventory of farm plots, houses and other community and households assets June and July 2015 SOCREGE Archaeological survey June and July 2015 SOCREGE Social baseline study August 2015 SOCREGE Complementary socioeconomic surveys August 2015 SOCREGE WSP Canada

356 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Description of the Main Environmental and Social Components Physical and Biological Environment The project site is under a tropical climate and is influenced by the climate variability of the Sahelian zone. The rainy season occurs from April to October, with more than 20 mm of rain monthly, representing seven months of the rainy season. The dry season runs from November to February. Monthly average temperatures range from 22.1 C and 35.2 C, with a maximum peak of 40.0 C in April and an average minimum of 17.6 C in January. Drought episodes have occurred in the last ten years, sometimes with negligible monthly cumulative rainfall. However, flooding, due to its random nature, occurs unexpectedly. The study area has a dense river system with numerous gullies, ravines and small intermittent streams. Cultivated lowlands are observed along the major streams. During the rainy season, much of the project area is under cultivation, while during the dry season, bare soils are observed. The permanent vegetation is sparse. Similar to most water catchments in the Sahelian zone, the flows in the Dianangou River catchment, where the study area is located, are intermittent and most of the runoff is occurring immediately after rain events (flash floods). Five soil classes characterize the project area, namely the raw mineral soils, the poorly evolved soils, the brown soils, the iron and manganese sesquioxides soils and the waterlogged soils. These different classes of soil all have a medium fertility level, except for the poorly evolved soils, which have a low level of fertility. There are seven units of land types, primarily shrub savanna (60.6%), rain-fed crop (28.5%), tree savanna (6.6%) and grassland savanna (3.6%). Other types (<1%) of land include Ripicolous formation, Barren soil and orchard and other forest plantations. The vegetation in the project area is characterized by an agrosystem mainly dominated by Vitellaria paradoxa (Shea) and associated with a much degraded savanna in areas suitable for agriculture. Species of the genus Combretum dominate this formation. Degraded forest galleries are found in the streams riperian areas. Overall, a strong anthropogenic pressure is observed associated with the use of the land by various activities (agriculture, livestock, artisanal mining, charcoal production and controlled bushfires). To this must be added the increased demand for firewood and other forest products caused by the population increase. These factors have strongly and negatively influenced the forest resources, causing habitat fragmentation and rarefaction or loss of flora species used by local populations to meet their basic needs. Thirty-seven threatened plant species have been identified in the project area and its surroundings. WSP Canada

357 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Most of the woody species identified during the inventories on the entire ecological study limit are considered of Low Concerned by UICN (Union internationale pour la conservation de la nature 1 ). Nonetheless, few species have been identified as Vulnerable in terms of conservation. It is important to note that the Project area covers about 58 km², is located inside the study limit which covers an area ten times greater. In terms of wildlife, all mammals observed in the project area are common species with the ability to tolerate a high level of human activity. A number of classified and protected areas are located close to the project area, the closest ones being the Tapoa Boopo pastoral zone (8km), the Ougarou hunting concession (9km), the Pagou Tandougou hunting concession (12 km) and the Leopaldi community hunting area (11km). Important wildlife conservation areas are also located nearby, namely the W National Park (70 km east from project area) and the Arli National Park Arli (20 km south of the project area). This special geographical situation suggests that during periods of annual migrations, some species of mammals frequenting these national parks may pass through the study area.. Figure 20-1 below locates the project area in relation to the nearest protected areas. 1 International Union for Conservation of Nature WSP Canada

358 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Figure 20-1 Classified and protected areas near the project area Human Environment The project site directly affects parts of the communities of Boungou (administrative village since 2012), and Natongou and Tambouangou (two hamlets officially administered by the village of Tatiangou). Surrounding communities are located: to the north, the village of Toabili (or Tawori, in Moore), where Birimian s exploration camp is located; to the northwest, the village of Kodjini; to the west, the hamlet of Takondibagou, pertaining to the village of Samtangou; to the south, by the village of Tatiangou. The organizational structure of the local communities is typical to the Gourmantché people, the majority ethnic group. It is made of castes and is very hierarchical, with centralized power. Gourmantché are a people with strong animistic beliefs, where geomancy is practiced. The population of the project area is comprised of locals, that are Gourmantché, and migrants predominantly composed of Mossi and Fulani. Three main languages that are spoken: gourmantchéma, Mooré and Fulfuldé. WSP Canada

359 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page On the customary level, the area is governed by the customary chief of Partiaga who still has considerable power of influence over the project area and throughout the province of Tapoa. Just like in other regions of Burkina Faso, women in the project area are in a situation of high vulnerability marked by a low rate of schooling and literacy, high fertility (average of six children per women), an uneven gender distribution of domestic tasks and agricultural work in winter, and unequal access to resources and means of production. Agriculture is the main productive activity in the project area. The agricultural system is extensive and traditional, with rudimentary methods. Fields and fallows occupy most of the project area. Maize, red sorghum, okra and sorrel are cultivated around the concessions in the rainy season. Further from concessions, in larger plots known as bush fields, sorghum, millet, cowpeas, peanuts and sesame are also cultivated. The bush is also used as a "reserve" for the supply of firewood. Fallows and the few unexploited areas within the project area are used for animal grazing and herding in order to reduce conflicts between farmers and herders. However, these conflicts are common and often originate when agricultural producers use cattle tracks to grow crops. This phenomenon has increased in recent years with the introduction of cash crops, including cotton and sesame. Since 1996, an artisanal gold mining site was established in Boungou, in the neighbourhood known as Boungou Woula, or Bougou Site. In the years following its creation, the site has seen a large influx of miners from other areas of Burkina Faso and neighbouring countries, reaching close to 2,000 people. Today the Boungou Site is primarily made up of households that have settled, that practice artisanal mining but also agriculture, animal husbandry, trade, etc. and whose children attend school there. The population of the Bougou Site, which is located outside the project s footprint, has been estimated to be 900 people in All villages and hamlets in the project area have sacred places dedicated to the gods that the inhabitants are strongly attached to. In particular, a number of sacred trees are found in Boungou for the protection and prosperity of the village. Many households also have an idol in the courtyard of their concession. The dead are usually buried near the concessions, but each village has at least one cemetery. Churches and mosques can be found in all the communities of the study area. Finally, the project area, like the entire Tapoa province, has been subject to the following security challenges: road attacks (the section of the RR28 between Ougarou and Tawori is the most crimeprone area of the province); armed assaults on concessions are regularly reported; and an increased trend towards heavily armed bandits. SEMAFO has adapted its security measures to the new context by heightening its security escorts to and from the project area, by bolstering its gate and on-site security presence, in addition to carrying out spot checks in the neighbourhood. WSP Canada

360 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Community Information and Consultation Program The stakeholder information and consultation process is an integral part of the ESIA. To date, Birimian has put in place mechanisms and communication tools so that all those involved in, or affected by, the project can freely express themselves. The information collected during these consultations has helped identify issues, risks, benefits, and opportunities in order for the project to avoid, minimize, or offset negative impacts and enhance the positive ones. As part of the ESIA and RAP development, information about the project was transmitted by information sheets and meetings with administrative and local authorities, technical services, as well as representatives of the surrounding villages, including women groups. Once the Project will be accepted and permitted, a Stakeholder Engagement Plan (SEP) will be developed and implemented. Since the acquisition of the project, many initiatives have been undertaken by Birimian to inform and consult with affected communities as part of the project preparation. These initiatives include: Establishment of a permanent team for environmental and community relations. Establishment of a grievance mechanism procedure. Participatory identification of potential community investment projects (School, donations, supplies, sports activities, water wells, solar panels, toilettes, road maintenance) through community meetings conducted by the SEMAFO Foundation. Several ad hoc meetings with authorities and other stakeholders. Local committee to inform stakeholders of project progress including discussion with existing artisanal miners. A Provincial Compensation and Resettlement Committee of the people affected by the project was set up and held its first meeting in October This Advisory Committee will be mobilized regularly throughout the RAP development and implementation phase to ensure the coordination of all parties involved. Throughout the ESIA process, the main concerns raised by the affected communities and other stakeholders included: The need for sustained livelihood assistance for displaced families during the transition and restauration period. The preservation of the agricultural lands along the Boumbouanga River, located west of the proposed open pit, which represents the main production area for the Boungou population. WSP Canada

361 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page The impact of dams on the availability of water for downstream agricultural activities. The possibility for people and livestock to access and use water reservoirs that will be created for the mine. Respect for graves and sacred sites, some of which could be affected by the mine. Loss of access to gold artisanal mining areas, an important income-generating activity for all the communities in the area. The prevention and fight against banditry, which already affects the area and could increase with mining operations. Support for community development, including access to water and basic health services. Birimian took these concerns expressed by stakeholders into account and took measures to optimize the project design in order to avoid any of these constraints. These actions led to a more balanced approach between the financial objectives of Birimian and the preservation and conservation of the environmental and social components of the population Project Impacts, Risk Analysis, Environmental and Social Management Plan Project Impacts The methodology used to identify and analyze the environmental impacts is based on an approach recognized by international funding agencies. This approach identifies the direct interactions between the project activities considered impact sources and the physical, biological, and human components. These interactions are customized according to project-specific phases (construction, operation and closure). All interactions identified are then analyzed on the basis of three criteria (intensity, extent and duration) to obtain a global indicator, the importance of the impact. Three levels of the importance of the impact were determined: minor, medium, and major. It is noted that the NGP will result in several negative impacts and some of the major impacts expected include the following: Surface water and groundwater. Noise levels. Social and economic aspects regarding resettlement. WSP Canada

362 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Cultural heritage. Most of the impacts on the physical environment are of medium importance given the disturbances on air, soil, surface, and ground water during the construction and operational phases. The mining installation and operations were designed for zero water discharge which is a clear advantage. Protection of the natural groundwater from process water containing cyanide has been properly planned and designed. Impacts on the biological components are mostly minor since these components are poorly represented in the project area. Impacts on the human components have an importance ranging from minor to major depending on the issues being raised. The most significant impact caused by the project will be the resettlement of the population currently living on the project site. The economic impact of the project at local, regional, and national levels is positive. Beginning from the construction phase, direct and indirect jobs will be created, resulting in tangible economic benefits for both local and regional communities. The project will create hundreds of skilled and unskilled direct and indirect jobs, most of them awarded to Burkinabe workers. This job creation will increase household incomes and improve living conditions. In addition, the procurement of goods and services required for the construction, operation, and closure of the mine will bring significant economic benefits to local and regional businesses, the majority in terms of supplying food and/or various products. The revenues generated by the mining operation will increase Burkina Faso s internal revenue through taxes and royalties charged by the local authorities. These revenues should have a beneficial impact at both the local and regional levels through increased investments in social and health services, and educational facilities. In addition, Birimian supports a number of social programs for the displaced households, and in a broader context, the local and regional communities Risk Analysis A Preliminary Risk Analysis was conducted to assess the environmental risks of the Project. Similar to any other heavy industrial activities, the Project will involuntary undergo critical situations like spills, breakages, etc., that may generate a direct impact on the environment. The causes and consequences of each of these situations were determined and existing preventive and emergency measures to implement were identified. The criteria considered for this risk assessment takes into account the severity of events, the consequences, and the likelihood of an occurrence. An analysis of the Project s facilities and consumables to be used on the mine site revealed a number of situations involving risks. The main environmental risks associated with the Project are as follows: Fire. Explosion. WSP Canada

363 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Degradation of walls and ramps in the pits and waste dump areas, berms and retention structures. Spills or leaks of hazardous materials. Toxic emissions. Natural disasters. Insurrection of the population. In order to minimize the level of risk related to the environmental aspects, health and safety and security measures have been identified. In addition, an Emergency Response Plan ( ERP ) will be implemented at the earliest stages of the operational phase of the project. The main objective will be management of these risks that cannot be completely eliminated by the protection measures already in place and will plan the appropriate emergency response should an accident occur. The intent of the ERP is to define emergency situations that could reasonably occur, and the measures of prevention, preparedness, response, and repairs associated with them, including staff training Environmental and Social Management Plan The Environmental and Social Management Plan ( ESMP ) presents all the environmental and social management measures to be implemented as part of the Project as well as all the operational aspects. The ESMP covers all project phases and allows avoiding, minimizing, enhancing, or compensating the various anticipated impacts by reducing them to an acceptable level for all stakeholders. The ESMP identifies the objectives to comply with Burkina Faso regulations and the international good practices in the mining sector. The ESMP also includes environmental monitoring programs and environmental and social follow-up, providing the basis for assessing the effectiveness of the management measures to be implemented by Birimian. The ESMP includes a number of measures to strengthen the capacity of the stakeholders concerned by the application of environmental and social management measures. Management measures are to be implemented at the earliest stages of the construction phase of the mine. Some measures will last throughout the operations at the mine site and others will last beyond the closure and rehabilitation phase of the Project. The planned management measures for the physical, biological, and human components at the construction phase include the following: Protection of soils. Control of runoff water, restrictions during heavy rain periods, respecting buffer zones along watercourses, etc. Implementation of restrictions regarding tree cutting, limits for working areas, etc. WSP Canada

364 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Reduction of noise and dust emissions. Control of traffic speed, access roads, the use and maintenance of equipment (fuel and lubricant tanks, vehicles and motorized equipment, etc.). Management of human resources, logistics, mobilization and demobilization of personnel and contractors. Management of the project-induced population influx into the area, including people seeking jobs and commercial opportunities related to construction and operation of the mine. Maximization of job opportunities for the local workforce and local suppliers of goods and services as well as for women s benefits and management of unrealistic expectations. Population and workers awareness to the risks of transmitting HIV/AIDS and other endemic diseases/ Protection and, where unavoidable, relocation of worship and sacred sites. Some measures implemented during previous project phases concerning soil, surface water, ground water, ambient noise, population and social cohesion, economy, and infrastructure, etc. will be maintained during the operational phase. Additional measures will include the following: Control of the mine tailings site in compliance with applicable regulations and requirements. Management of waste rock dumps and progressive re-vegetation to minimize wind erosion. Management of water, hazardous materials, wastes, traffic, maintenance of vehicles, etc. Mining will be carried out according to good practices and with specific attention to Occupational Health and Safety. Finally, various management measures are planned for the closure phase and include the following: Without compromising the integrity and security of the site and the people, some structures will be maintained. Some unnecessary infrastructure and facilities will be dismantled. Site rehabilitation and re-vegetation. Restoration of livelihood conditions for neighbouring populations. WSP Canada

365 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Access roads and other infrastructures of interest for the communities built for the project may be left in place for their use at the end of mine life. Restricted areas may be defined within the permit to protect the environment, the natural habitat, archaeological sites and/or public interest infrastructures. A monitoring program will be implemented during the construction phase that will be conducted by Birimian on an ongoing basis. The program will ensure compliance with the commitments agreed to as part of the ESIA and environmental obligations, as well as compliance with the proposed management measures and with laws, regulations and other environmental considerations included in the contractors technical specifications. These measures to be implemented will be included in the contractors technical specifications according to their respective activities. The environmental and social monitoring program to be implemented will: Monitor changes for certain sensitive environmental components. Compare current conditions with pre-project initial conditions to identify trends or impacts that may result from project activities or natural events. The main components of the environment that will be monitored as part of this program includes: Surface and ground water quality. Ambient air quality. Ambient noise. Status of the flora and effectiveness of re-vegetation. Fauna. Local and regional economy. Community health and security. Communities perception and degree of satisfaction with regards to the project Resettlement People and Activities Affected by the Project The physical displacement of a large number of people (165 households or approximately 895 people) from three communities and the expropriation of a large area of agricultural land (approximately 813 ha) represents an important activity that will require immediate and focused effort. WSP Canada

366 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Scope of Resettlement The population resettlement requirements are based on the following assumptions: The area impacted is based on the site layout with a 50 m buffer zone to be expropriated around the mine site (fenced area). Census data and property inventories as of September The replacement of all private houses (with improved housing), all public infrastructures, plus financial compensation for granaries, sheds, ovens, parks, shops, roosts and similar small infrastructures. For the farmlands, in-kind (grains) and cash compensation over 5 years, based on the value of a basket of harvested products, as established from the average yields and market values compiled by the local branch of the Ministry of Agriculture for the area, and the surface area of the parcel, as recorded during the field study Resettlement Action Plan Budget The budget for the RAP is estimation to approximately USD 7.4M as presented in Table This cost with USD 600,000 contingency was included in the financial analysis of the project. Table 20.3 Resettlement Action Plan Budget (from SOCREGE, 2015) DESCRIPTION COST (USD) Compensations Private housings 3,304,858 Related infrastructures 31,718 Public Infrastructures 187,652 Commercial infrastructures 5,702 Farmland 338,779 Crops 2,905,227 Pasture / grassland 19,240 Trees 85,985 Sacred sites and graves 5,167 Sub-total - Compensations costs 6,884,329 Moving allowance - private housings 46,750 Moving allowance - Commercial activities 2,840 Economic measures for resettlement and income restoration 192,875 RAP implementation 278,333 TOTAL 7,405,127 WSP Canada

367 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Acid Rock Drainage Samples of waste rock and tailings were subjected to laboratory geochemical tests to assess their potential to generate acid rock drainage ( ARD ) and leachable metals ( ML ). The results of the ARD assessment were compared to the evaluation criteria presented in the Global Acid Rock Drainage ( GARD ) Guide (INAP, 2009), a reference document on best practices related to mine waste characterization and ARD prediction, prevention and mitigation measures. Results of the metal leaching tests and process water chemistry were compared to the applicable effluent discharge guideline values specified in Burkina Faso Effluent Discharge Criteria ( EDC ), Article 10 of Decree No /PRES/PM/MEE dated 7 May 2001 on setting standards for discharges of pollutants into the air, water, and soil, and World Health Organization drinking water guidelines Waste Rock and Construction Materials The ARD/ML assessment of the waste rock included samples of all the major lithologies located in the open pit with samples number scaled based on the relative proportion of the lithologies in the pit. A total of 130 samples of waste rock were collected from exploration boreholes within the footprint of the pit. The samples collected were from a wide geographic spread over the pit and included both fresh and oxidized rock. Higher risk samples with elevated sulfur and arsenic from close to ore zone were also targeted in the sampling. The results of this ARD/ML program to date are presented below. 97% of the waste rock samples were found to be non-acid forming or acid consuming. Samples classified as PAF or uncertain had total sulfur and sulfide sulfur contents of approximately 1% to 1.5%. Based on the sulfur assay database, only 0.6% of the waste has sulfur grades above 1% therefore it is expected that the volume of waste rock which will be potentially acid forming is likely to be below 1%. Based on the testing results, circum neutral to weakly alkaline seepage flows are expected. Therefore, any acidity produced by limited zones of PAF rock should be adequately buffered. WSP Canada

368 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page The waste rock samples were generally found to have a small number of elevated elements and low levels of element enrichment, with arsenic the most commonly enriched metalloid. The initial distilled water extract testing indicated that arsenic is readily soluble in water. It also indicated a reasonable trend between total arsenic and dissolved arsenic. Kinetic cells were established to further examine the arsenic leaching potential. Three cells were established according to total arsenic concentrations, i.e. non enriched to slightly enriched representing approximately 75% of the waste rock, significantly enriched representing approximately 12.5% of the waste rock and highly enriched waste representing approximately 12.5% of the waste rock. One hundred and twenty waste samples were used to form the cells according to their arsenic concentrations to take into consideration variability with the samples. The kinetic cells confirmed that low arsenic waste is unlikely to release arsenic above drinking water or surface water guideline, moderate arsenic waste is likely to release arsenic above drinking water guidelines but below surface water guidelines and high arsenic waste is likely to exceed drinking water guidelines consistently with short term exceedances of the surface water guideline. Overall it was estimated that direct seepage from the waste dump would not be suitable for direct use as drinking water but should meet Burkina Faso Effluent Discharge Criteria for release to surface water. Sediment ponds have been provided in the design layout to capture and allow for testing of water running off the waste dump prior to release to the environment Tailings Two master composite tailings samples were tested, one representing the first year production at higher gold grades and the second sample representing the remainder of the life of operations. The tailings sample were found to be non-acid forming with the long term tailings pore water quality is expected to be circum neutral to weakly alkaline. The tailings had a moderate level of elemental enrichments, with arsenic and sulfur found to be highly enriched and selenium found to be significantly enriched in both samples and variable enrichment in molybdenum, bismuth and antimony. Distilled water extract and kinetic testing indicates that the supernatant will have significantly elevated concentrations of arsenic. The humidity cell testing of tailings recorded arsenic in the leachate above drinking and surface water guidelines throughout, peaking at 2.3 mg/l after the first week before reaching steady state release at around 0.5 mg/l from Week 15 onwards. This indicates that the tailings will likely release high concentrations of arsenic within a short period following deposition, but that long term release rates will also be elevated. The tailings supernatant will also be elevated in cyanide although no testing of residual cyanide concentration in the tailings supernatant was conducted. The design of the Natougou Tailings Storage Facility ( TSF ) has accounted for the arsenic leaching potential and the contained cyanide and includes for a basal HDPE liner, underdrainage systems and seepage recovery system below the liner to reduce the potential for release of arsenic to the environment. Since the TSF will be essentially be zero discharge, there should be no issues with effluent quality under dry conditions. During extreme rainfall events, current design is 1/100 year event, if the TSF storage capacity is exceeded and an effluent discharge is required, there is the possibility to exceed both the Total CN and arsenic effluent discharge criteria. WSP Canada

369 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Waste Disposal and Sanitary Wastewater Solid Waste Solid wastes generally include bags, pallets, empty drums, worn out parts, liners and other supply packaging. Suppliers will be invited to take back used materials. For all non-recyclable wastes, a solid waste disposal site is to be created at a suitably enclosed area, restricted to prevent animal access, and located to avoid contamination of water and vegetation. This site is to be operated as a landfill site. No onsite waste incineration is planned with the exception of the biomedical waste and all packaging associated with cyanide. Domestic waste (glass, metal, paper and plastics) will be separated and stored in special containers for recycling by qualified companies in Ouagadougou wherever possible or where these services are available Hazardous Waste Hazardous wastes, which will primarily include waste oils, packaging for process reagents and laboratory chemicals, will be disposed of in a safe and environmentally sound manner. Waste oils will be recycled by the supplier, while most reagents and chemicals that require disposal will be disposed of tailings storage area. Empty sodium cyanide boxes and inner bags will be incinerated on site to avoid any possible uses by the population. Spills of hazardous materials on site will be given the highest operating priority and will generally include the excavation of contaminated soils, neutralization of the affected site, disposal and/or neutralization of the impacted soils on site. The mining equipment on site will be immediately available for use in the event of a spill. Biomedical waste will be safely stored at the site clinic and then incinerated on site to be safely disposed of Sanitary Wastewater At the mining camp, a sewage piping network will be provided to collect sewage from the various lavatories, showers and laundry facilities then sent to a modular bacterial digester wastewater treatment system to process both black and grey waters. This unit will be designed to avoid contamination of the ground water or existing watercourses. In the process area, satellite workshops and small office buildings, a standard septic tank and soakaway drain-field system will be installed for each building. WSP Canada

370 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Closure, Decommissioning and Reclamation The NGP closure, decommissioning and reclamation costs are estimated to USD 17.2 M. These costs were included in the financial analysis for these closure activities and the calculation of the cost for the Reclamation Trust to BCEAO. The mine closure budget is presented in Table Table 20.4 Mine Closure Budget 1.0 Closing and restoration direct costs estimation Amount USD 1.1 Buildings and infrastructure dismantling a $2,000, Restoration of the footprint of buildings, ore pad and access road $661, Securing pits $184, Restoration of waste rock piles $1,713, Restoration of the tailing management facility $12,280, Restoring sedimentation ponds $179,210 Subtotal 1.0 (direct costs) $17,018, Closing and restoration indirect costs estimation 2.1 Engineering c Plans, specifications and supervision (± 2.5% of direct costs) $425,471 Subtotal 2.1 $425, Post-operative and post-restoration monitoring d Annual monitoring of the integrity and stability of structures over 5 years $11,535 Annual agricultural monitoring over 5 years $9,903 Environmental monitoring at a frequency of 6 times per year for 5 years $109,570 Subtotal 2.2 $131,008 Subtotal 2.0 (indirect costs) $556, TOTAL - Closing and restoration direct and indirect costs Subtotal (Without contingency) $17,575,331 Contingency $3,515,066 Salvage value b $3,841,897 TOTAL $17,248,500 Notes: a) Salvage value for the equipment and structures proposed by from A.M. King Industries Inc. b) The cost of this package is estimated by Knight Piésold, c) The engineering costs uses a percentage derived from past projects in Burkina Faso d) The cost estimation assumes that 50% of the monitoring will be carried out by the owner and are thus included G&A costs. WSP Canada

371 NI TECHNICAL REPORT - ENVIRONMENTAL STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT Page Some of the items presented in the cost estimation above are based on a Conceptual Closure and Rehabilitation Plan and bear an accuracy of up to +/- 30%. The Closure and Rehabilitation Plan and its costs estimates will have to be better detailed prior to provide the Reclamation Trust to BCEAO (see section ). It is the QP opinion that this level of accuracy on the closure cost will have negligible effect on the global project economics. This Conceptual Closure and Rehabilitation Plan include work to be conducted from the closure of the mine, at the end of operational activities, as well as progressive rehabilitation work. The goal is to return the site to a satisfactory state as quickly as possible in terms of: Reducing the risks for health and safety. Controlling erosion. Limiting maintenance and monitoring. Developing a compatible profile with the future uses of the site, primarily for the plant site. The main objectives of the Closure and Rehabilitation Plan include restoring ecosystems and take-over and recovery of land uses. This plan includes: Dismantling and removal of plant equipment, machinery and infrastructure. Progressive rehabilitation to allow rapid recovery of the vegetation cover and early recovery of the ecosystem. Sustainability of rehabilitation work and control of water and wind erosion; Take-over and recovery of land uses. Maximization of material and equipment recovery. Site rehabilitation as part of a participatory approach involving interested communities. Implementation of a post-closure monitoring program. All structures that can be used by communities will be maintained, with the exception of the facilities that may constitute a risk to people or the environment. WSP Canada

372 NI TECHNICAL REPORT - REFERENCES Page CAPITAL AND OPERATING COSTS 21.1 Operating Costs Mining Operating Costs Operating costs were generated from first principles based on the following operating assumptions: Contract open pit mining. Owner management and technical services. Costs were benchmarked against operations of similar size in Africa and SEMAFO s existing operation in Burkina Faso. Costs are shown in United States dollars (USD). Escalation has not been considered in the estimate (no inflation is applied); costs are in real dollars. No contingency has been allocated to the open pit operating costs. A contingency of 15% has been applied to the mining infrastructure. Repair and maintenance costs for the mining equipment were sourced from Komatsu s distributor BIA Overseas S.A. in Belgium and Caterpillar (CAT) equipment distributor JA Delmas in France. Explosive costs were sourced from AEL Mining Services (AEL) in Burkina Faso. AEL also provided the costs for operating the blasting activity as a turnkey operation, including the provision of an experienced blast crew, explosive truck and blasting related equipment and tools. Equipment quotations were received for all major mining equipment. As a contractor operation was estimated, equipment purchase costs were translated into leasing costs and used to calculate the contractor mining costs using an interest rate of 5.5%. To simulate the fact that the mining contractor would not be purchasing new equipment, but would likely mobilize used equipment, only 75% of the capital value was used to calculate leasing costs. Importation tax on parts of 8.5% and freight costs of 7.5% for all parts were added to the repair and maintenance costs provided by the equipment manufacturers. Salary information was provided by SEMAFO based on its operating experience at the SEMAFO s operations in Burkina Faso. RC grade control related costs including assay costs of USD 5.2/sample and all-inclusive drilling costs of USD 0.75/t of ore were also provided by SEMAFO. AMC believes these costs to be reasonable. Lycopodium Minerals Canada

373 NI TECHNICAL REPORT - REFERENCES Page 21.2 AMC assumed that Birimian would be providing fuel and lubricant at cost to the mining contractor. Fuel cost of USD 1.08/L was used after first gold is poured. During the pre-stripping period, Birimian will be exempt from paying value added tax (VAT) and duties and, as such, a fuel price of USD 0.92/L was used. AMC applied an 8% mark-up to the equipment, consumables and labour costs in the cost model, except for the AEL costs, which were provided inclusive of mark-up. An additional 0.5% mark-up was applied to the labour costs to model off-site costs for support staff not based on site. Supplementary information was estimated by AMC based on its internal database for similar types of equipment or activities. The cost estimate was based on the final detailed schedule presented in Section 16. The open pit contractor operating costs estimate covers the following activities: Grade control. Drill, blast, load, haul and dump of ore to stockpile and waste to the waste dumps. Backfilling of the open pits. Auxiliary operations such as clearing of the open pits and waste dump footprint area, mine haul road construction, maintenance of benches and waste dumps, dewatering. Installation and maintenance of a mining workshop and wash bay. Maintenance of the mine fleet. Open pit contractor management and supervision. Mining and maintenance personnel. Consumables including fuel, parts, explosives, etc. Equipment ownership costs expressed as a lease cost per operating hour. Installation and maintenance of a mining contractor camp. Costs associated with operating the office for the technical services team and contractors are included in the G&A costs as presented in Section 22. As presented in Table 21.1 and Table 21.2 the mining cost per tonne of material mined within the pits over the life of mine averages 2.47 USD/t. Lycopodium Minerals Canada

374 NI TECHNICAL REPORT - REFERENCES Page 21.3 Table 21.1 Mining operating costs by year* Mining Costs Unit Total Load $'000 25, ,546 4,376 4,415 4,256 3,666 3, Haul $'000 52, ,184 9,289 9,437 9,043 7,518 7,004 1,368 0 Drill & Blast $'000 99, ,152 18,109 16,049 17,620 13,931 14,188 2,038 0 Pre-Split $'000 9, ,505 1,889 1,198 1, , Grade Control $'000 7, ,438 1,434 1, , Auxiliary Equipment $'000 28, ,183 4,125 4,116 4,766 4,546 4,583 1, ROM Rehandle $'000 9, ,190 1,273 1,273 1,273 1,276 1,273 1, Long term Stockpile Rehandle $'000 6, Clearing and Grubbing $' Dewatering $'000 2, Geology $'000 8, ,263 1,263 1,263 1,263 1,263 1, Management, Supervisory and Fixed Roles $'000 35, ,473 5,288 5,288 5,288 5,288 5,288 2, Miscellaneous Operational Overheads $'000 3, Contracted Services $'000 11, ,689 1,689 1,689 1,689 1,689 1, Total $' , ,671 50,618 48,721 49,655 42,889 42,760 13,079 1,633 Average unit cost $/t mined *Costs for 2017 and 2018 are capitalized. Lycopodium Minerals Canada

375 NI TECHNICAL REPORT - REFERENCES Page 21.4 Table 21.2 Mining operating cost by activity Mining Activity - $/t mined Unit Cost Load $/t 0.21 Haul $/t 0.44 Drill & Blast $/t 0.82 Pre-Split $/t 0.08 Grade Control $/t 0.07 Auxiliary Equipment $/t 0.23 ROM Rehandle $/t 0.07 Long term Stockpile Rehandle $/t 0.05 Dewatering $/t 0.02 Geology $/t 0.07 Management, Supervisory and Fixed Roles $/t 0.29 Miscellaneous Operational Overheads $/t 0.03 Contracted Services $/t 0.09 TOTAL $/t Plant and Administration Costs Process plant and administration operating costs have been developed by Lycopodium and Birimian based on a treatment rate of 4,000 tpd of ore with the plant operating 24 h/d, 365 d/y with a 92% plant utilization (nominal 8,060 h/y) and a P 80 grind size of 63 µm. The operating cost estimate has been compiled from a variety of sources and is based on whole of ore treatment and a head grade of 4.2 g/t Au ore. Operating costs are presented in United States Dollars (USD) and are based on pricing from SEMAFO s existing operation in Burkina Faso and pricing obtained during the third quarter of 2015, to an accuracy of ±15%. The process plant operating costs for the facilities are summarized in Table Lycopodium Minerals Canada

376 NI TECHNICAL REPORT - REFERENCES Page 21.5 Table 21.3 Process Plant Operating Cost Summary Cost Centre USD M/y USD /t Ore Operating Consumables $6,679,657 $4.99 Plant Maintenance $2,204,101 $1.64 Laboratory (Plant) $112,632 $0.08 Power (fuel only) 1 $9,525,031 $7.09 Power Plant Costs 2 $1,979,915 $1.57 TPA Exoneration ($243,560) ($0.10) Labour (Plant Operations & Maintenance) $5,688,778 $4.24 Subtotal Process Plant $25,946,544 $19.51 General & Administration $4,208,241 $3.13 Labour (Administration) $2,429,875 $1.81 Subtotal - General & Administration $6,638,116 $4.94 Total $32,584,660 $ Power cost is derived based on fuel price and fuel usage at the power plant. Associated maintenance costs are covered separately. 2 Includes maintenance costs Power The power requirements for the process plant were based on the mechanical equipment list and adjusted for equipment load factor and utilization. A power unit price of USD 0.17/kWh was calculated based on a hybrid HFO/diesel fuelled on site power station and supported by vendor quotation and stated fuel consumption data, but excludes maintenance and leasing costs and is captured as a separate cost item. A summary of the power cost for the plant by plant area is tabulated below in Table Table 21.4 Process Plant Power Cost by Plant Area Plant Area Total Power (USD /t) Feed Preparation $0.19 Milling $5.38 Leaching & CIP $0.48 Elution / Goldroom $0.18 Tails Handling & TSF $0.17 Reagents $0.04 Plant & Raw Water, Plant, Air & Oxygen Services $0.34 Mining Services $0.10 All buildings, camps $0.21 Total $7.09 Lycopodium Minerals Canada

377 NI TECHNICAL REPORT - REFERENCES Page Operating Consumables The consumables consumption requirements for the Natougou process plant were based on crushing and grinding data, testwork consumption rates and industry standards. Pricing for reagents and consumables were based on cost information from SEMAFO s existing operations in Burkina Faso and budget quotations received from suppliers and adjusted to a DAP (Delivered at Place) price and include customs and duties where applicable. The diesel cost was supplied by Birimian and diesel consumption for the plant mobile equipment was estimated based on the amount of mobile equipment required for the plant and mine. The consumables cost by plant area is summarized below in Table Table 21.5 Process Plant Consumables Cost by Plant Area Plant Area USD$ /t ore Feed Preparation $0.13 Milling $1.18 Leaching & CIP $0.18 Elution / Goldroom $0.01 Reagents $2.39 Water Treatment & Cleaning $0.03 Fuel (plant and mobile equipment) $1.07 Total $ Labour (Processing / Maintenance and Administration) Labour rates and costs were provided by Birimian and were based on a developed organizational chart of the Natougou operations and cost information from SEMAFO s existing operations in Burkina Faso. The labour rates are based on a skill level and consist of a base salary and the required overhead allowances General and Administration Cost (excluding G&A labour) General and Administration costs were developed jointly by Lycopodium and Birimian and are based on information on SEMAFO s existing operations in Burkina Faso and similar West African operations in the region and are summarized in Table Lycopodium Minerals Canada

378 NI TECHNICAL REPORT - REFERENCES Page 21.7 Table 21.6 Plant General and Administration Summary General and Administration USD /y Site Office Costs (telecom, licenses, stationery, supplies) $707,200 Insurances $662,000 Financial Costs (banking charges, legal fees, etc) $91,640 Transportation $1,582,000 Recruiting/Relocation Costs $21,225 Training $38,216 Security Contract $870,000 Miscellaneous Costs $235,960 Total $4,208, Maintenance The maintenance cost for the processing plant was factored from the equipment supply capital cost and is summarized in Table Allowances for plant mobile equipment, and general maintenance have been made. Table 21.7 Plant Maintenance Materials Cost Annual Maintenance Cost USD Annual Maintenance Cost USD /t Total $2,204,101 $1.64/t ore Exclusions The operating costs stated above makes no allowance for the following: All SEMAFO head office / corporate costs. All sunk costs. All project financing costs. Any impact of foreign exchange rate fluctuations. Any escalation from the date of the estimate. Any contingency allowance. Lycopodium Minerals Canada

379 NI TECHNICAL REPORT - REFERENCES Page 21.8 Tailings storage costs, including future lifts and rehabilitation. This is captured in Section 21.4 Sustaining capital. Government monitoring / compliance costs Capital Cost Estimate The overall capital cost estimate was compiled by Lycopodium and is presented here in summary format. The capital cost estimate reflects the Project scope as described in this report. Mine capital costs (developed by AMC) are included in the estimate tables below. Knight Piésold provided quantities and rates for the TSF, east and west water storage dams, water storage facility and sediment ponds and are included in the estimate tables below. All costs are expressed in USD unless otherwise stated and based on Q pricing. The estimate is deemed to have an accuracy of ±15%. The various elements of the Project estimate have been subject to internal peer review by Lycopodium and have been reviewed with SEMAFO for scope and accuracy Summary The capital estimate is summarized in Table 21.8 by area and Table 21.9 by discipline. The preproduction capital cost is estimated at USD million. The estimate Work Breakdown Structure (WBS) is based on the standard Lycopodium Minerals WBS for gold projects. Table 21.8 Capital Estimate Summary by Area (4Q15, ±15%) USD ( 000,000) Indirect Construction $13.6 Processing Plant $42.3 Reagents and Plant Services $13.7 Infrastructure $41.8 Owner costs $15.8 EPCM costs $15.9 Resettlement Action Plan $8.0 Initial Supplies Inventory $7.2 Plant & Infrastructure Subtotal $158.3 Pre-stripping $42.4 Contingency $18.7 Total $219.4 Lycopodium Minerals Canada

380 NI TECHNICAL REPORT - REFERENCES Page 21.9 Table 21.9 Capital Estimate Summary by Discipline (4Q15, ±15%) Discipline Supply Cost USD ( 000) Freight Cost USD ( 000) Installation Cost USD ( 000) Contingency USD ( 000) Taxes/Duties/ Inspection USD ( 000) Total USD ( 000) General $8,351 $947 $1,957 $1,212 $459 $12,926 Earthworks $3,516 $863 $16,352 $3,134 $10 $23,875 Concrete $3,289 - $3,539 $1,089 $24 $7,941 Steelwork $2,156 $869 $3,112 $893 $119 $7,148 Platework $1,082 $508 $324 $306 $60 $2,280 Field Erected Tankage $721 $411 $2,308 $450 $40 $3,929 Mechanical $22,770 $2,982 $2,496 $2,501 $1,252 $32,002 Bulk Fuel Facility 1 $784 $54 $135 $67 - $1,039 Plant Piping $4,214 $548 $1,054 $1,163 $232 $7,210 Overland Piping $416 $50 $446 $178 $23 $1,112 Buried Piping $857 $111 $334 $261 $47 $1,611 Electrical $5,102 $272 $3,711 $1,374 $281 $10,740 Power Generation 1 $3,708 $281 $702 $351 - $5,041 Instrumentation & Control $2,871 $177 $1,334 $629 $158 $5,168 Buildings & Architecture $3,452 $160 $2,387 $800 $190 $6,988 Mining $1,625 $588 $43,555 $553 $22 $46,344 Owners Cost $22,343 - $2,220 $1,746 $317 $26,627 EPCM Costs $7,243 - $7,301 $1,454 - $15,998 Geotech EPCM Costs $173 - $1,109 $128 - $1,410 Grand Total $94,673 $8,822 $94,374 $18,289 $3,233 $219,391 1 Costs associated with initial costs during construction. Balance of costs are covered in deferred capital costs Estimating Currency and Base Date The estimate is expressed in USD based on prices and market conditions current at fourth quarter 2015 (4Q 2015). The following exchange rates have been used in the compilation of the estimate: USD 1.00 = AUD 1.37 USD 1.00 = EUR 0.92 USD 1.00 = CAD 1.39 USD 1.00 = GBP 0.68 USD 1.00 = ZAR 15.5 USD 1.00 = XOF 600 Lycopodium Minerals Canada

381 NI TECHNICAL REPORT - REFERENCES Page Foreign currency exposure is shown in Table below. Table Foreign Currency Exposure Currency Percentage of Capital Estimate USD$ ~86% EUR ~11% AUD <3% GBP <1% ZAR <1% The estimate capital cost accuracy as defined by the scope in this document and attached appendices is as follows: ±15% Plant Capital Costs. ±15% Infrastructure Capital Costs Mining Capital Expenditure The open pit operations are planned to be contracted out. All mining equipment, workshop and contractor camp facilities are assumed to be provided by the open pit contractor; they have been included in the operating costs as a leasing cost and are presented in Section SEMAFO will be installing its own explosive magazine and storage facility, the cost of which is captured in Table Establishment costs and mobilization costs have been included as capital expenditure. Costs incurred during pre-stripping until first ore is milled in January 2019 have been capitalized. These costs are presented in Table A contingency of 15% has been applied to the mining infrastructure including the workshop and camp. No contingency has been applied to the pre-strip amount. Table Mining Capital Expenditures Mining Capital Expenditures, USD ( 000,000) Establishment, mobilization 2.5 Pre-stripping, mine road construction 42.4 Total Mining CAPEX 44.9 Lycopodium Minerals Canada

382 NI TECHNICAL REPORT - REFERENCES Page Estimating Methodology General arrangement drawings and a layout 3D model have been produced with sufficient detail to permit the assessment of the engineering quantities for earthworks, concrete, steelwork, mechanical and electrical for the crushing plant, processing plant, conveying systems and infrastructure. The layouts and model have been based on recently completed facility designs, modified construction and as-built drawings of past project facilities, as well as initial concept drawings and computer modelling. Unit rates that reflect the current market conditions have been established for bulk materials, capital equipment and labour via an extensive Budget Quotation Request (BQR) process. Labour rates from the market have been benchmarked against in-house labour rates and indirect cost modelling to ensure adherence suitability to the current projects market. The rates used in the estimate have been reviewed and deemed to reflect the current market conditions. Budget pricing for equipment and infrastructure facilities was obtained from suitable suppliers and contractors Quantity Development The Project works were quantified to represent the defined scope of work and to enable the application of rates to determine costs. Allowances for earthworks compaction, waste, rolling margin and the like are included in the build-up of unit costs. Quantity information was derived from a combination of sources and categorised to reflect the maturity of design information as follows: Study Engineering: includes quantities derived from concept or preliminary engineering for the purpose of the study. Includes equipment lists and redline mark-ups of previous project drawings / data by engineering. Estimated: Includes quantities derived from sketches or redline mark-ups of previous project drawings / data by estimating or similar project go-bys. Factored: Quantities derived from percentages applied as a factor derived from previous projects. The derivation of quantities is provided in Table 21.12, weighted by value of the direct permanent works (i.e. excluding temporary works, construction services, commissioning assistance, engineering costs, escalation and contingency). Lycopodium Minerals Canada

383 NI TECHNICAL REPORT - REFERENCES Page Table Derivation of Quantities Classification Quantity Unit Study Engineering % Estimated % Factored % Concrete m Structural Steel t Platework t Field Erected Tanks t Mechanical Equipment lot Piping - Plant lot Piping - Overland km E & I Plant lot Pricing Basis Estimate pricing was derived from a combination of the following sources: Budget Quotation: Budget pricing solicited specifically for the study or project estimate. Database: Historical database pricing that is less than six months old. Estimated: Historical database pricing older than six months, escalated to the current estimate base date. Factored: Factored from costs with a basis. Table summarizes the source of pricing by major commodity, weighted by value of the direct permanent works (excluding temporary works, construction services, commissioning assistance, EPCM costs and contingency), including supply and installation. Table Sources of Pricing Classification Budget Quotation % Database % Estimated % Factored % Concrete Structural Steel Platework Mechanical Equipment Piping - Plant Piping - Overland E & I Equipment Lycopodium Minerals Canada

384 NI TECHNICAL REPORT - REFERENCES Page Pricing has been categorized by the following cost elements, as applicable, for the development of each estimate item. Bulk Materials This component covers all other materials, normally purchased in bulk form, for installation on the project. Costs include the purchase price ex-works, any off-site fabrication, and transport to site (unless otherwise stated), and over-supply for anticipated wastage. Plant Equipment This component represents prefabricated, pre-assembled, off-the-shelf types of mechanical or electrical equipment item. Pricing is inclusive of all costs necessary to purchase the goods ex-works; generally excluding delivery to site (unless otherwise stated) but including operating and maintenance manuals. Vendor representation and commissioning spares have been allowed for separately in the estimate. Installation This component represents the cost to install the plant equipment and bulk materials on site or to perform site activities. Installation costs are further divided between direct labour, equipment and construction indirect costs. The labour component reflects the cost of the direct workforce required to construct the Project scope. The labour cost is the product of the estimated work hours spent on site multiplied by the cost of labour, inclusive of overtime premiums, statutory overheads, and payroll burden. The equipment component reflects the cost of the construction equipment and running costs required to construct the Project. The equipment cost also includes cranes, vehicles, small tools, consumables, and PPE. Construction indirect costs encompass the remaining cost of installation and include items such as offsite management, onsite staff and supervision above trade level, crane drivers, equipment and labour mobilization and demobilization, Rest and Recreation (R&R), meals and accommodation costs. The labour gang rates, and equipment costs estimated for each major trade commodity are shown in Table Lycopodium Minerals Canada

385 NI TECHNICAL REPORT - REFERENCES Page Table Standard Direct Labour Gang Rates Item Direct Labour USD/h Equipment USD/h Total Hourly Rate USD/h Concrete Installation SMP (excl Mills) Mill Erection Field Erected Tanks Building Installation Electrical & Instrumentation Generally, for bulk commodities, lower to mid range vendor and contractor rates were used in the compilation of the estimate. Freight The freight estimate was derived from the following: The assessment of freight tonnes or bulk volume and number of containers from the quantities derived in the capital cost estimate and vendor supplied shipping lists and advice. The most likely country of origin. The application of rates solicited from the market specifically for the project that includes preshipment, ocean freight, and post-shipment overland freight. By calculation as a percentage of supply costs for items where freight tonnes and bulk volumes are yet to be determined. By benchmarking against similar projects located in similar regions Field In-directs Project construction offices and establishment, communications, computers, IT services, servers, telephones and temporary fuel facilities are included in the capital estimate. Construction services such as power, water, fuel and consumables are included. Construction indirect costs for all direct labour is included in the installation rates for all works in the capital estimate. This is inclusive of meals, PPE, travel and clothing. Earthworks rates are inclusive of fuel, maintenance and running costs of machinery. Construction equipment and project cranes are included in the capital cost estimate. Lycopodium Minerals Canada

386 NI TECHNICAL REPORT - REFERENCES Page Engineering Services The estimate for engineering design services is based on an assessment of the work content using the drawing and equipment lists prepared for the study as a guide. Design man-hours will be estimated on a deliverable basis and benchmarked against previous Lycopodium experience for similar sized projects. Duration based costs such as project management, construction and commissioning support will be estimated using the project implementation schedule and previous project experience as a guide Owner s Costs The Owner s costs include: Pre-production costs. First fills (grinding media, lubricants, fuel, and reagents). Opening stocks. Plant mobile equipment (the majority of this will be handed over from the construction team s mobile fleet following plant handover). Consumable, insurance and commissioning spares. Vendor representative and training costs for the process plant. Maintenance tools and equipment. Crop compensation and resettlement costs. Office equipment and furniture Spares Spares have been calculated as a percentage of Mechanical & Electrical Equipment supply Duties, taxes and Insurances Government Duties and Taxes The capital estimate includes an estimate of all government taxes and duties. Lycopodium Minerals Canada

387 NI TECHNICAL REPORT - REFERENCES Page Project Insurances Project insurances have been excluded from the estimate Contingency An amount of contingency has been provided in the estimate to cover anticipated variances between the specific items allowed in the estimate and the final total installed project cost. The contingency does not cover scope changes, design growth, etc., or the listed qualifications and exclusions. Contingency has been applied to the estimate on a line-by-line basis as a deterministic allowance by assessing the level of confidence in each of the defining inputs to the item cost, these being engineering, estimate basis and vendor or contractor information, and then applying an appropriate weighting to each of the three inputs. It should be noted that contingency is not a function of the specified estimate accuracy and should be measured against the project total that includes contingency. The resultant contingency for the project is 9.2% before taxes, duties and inspection Deferred Capital Deferred capital costs are covered in Section Deferred capital costs are not included in the Capital Cost Estimate detailed in Table Escalation There is no allowance for project escalation in the Capital Cost Estimate Qualifications and Assumptions The capital estimate is qualified by the following assumptions: The base date for the bulk of pricing for the estimate is fourth quarter 2015 (4Q 2015). Prices of materials and equipment with an imported content have been converted to USD at the rates of exchange stated previously in this document. All pricing received has been entered into the estimate utilizing native currencies wherever possible. The bulk earthworks commodity rates that include imported material are based on the assumption that suitable construction / fill materials will be available from borrow pits within 2 km of the work fronts. Engineering quantities and rates for the TSF, access road and pipeline corridor, pit diversion channel, sediment ponds, WSF, and west and east water supply dams have been provided by Knight Piésold. Sub-consultant design costs related to these work scopes for project implementation have also been included. Lycopodium Minerals Canada

388 NI TECHNICAL REPORT - REFERENCES Page There is no allowance for unforeseen blasting in the bulk earthworks cost estimates, given the results of the site geotechnical investigation. The estimate allows for supply of structural steel and minor platework from South East Asia. It has been assumed mobile equipment purchased and used by the owner s construction team will be handed over to the owner s operations team upon completion of construction. No allowance for additional mobile equipment has been made for operations Exclusions The following items are specifically excluded from the capital cost estimate: Permits and licences. Project sunk costs. Exchange rate variations Deferred Capital Project deferred capital costs are summarized in Table Deferred capital includes payments for the bulk fuel facility and power plant lease and finance costs following first gold pour. Table Deferred Capital Summary Main Area USD ( 000,000) Bulk Fuel Facility $4.2 Power Plant $12.4 Total $ Sustaining Capital Costs Project sustaining capital costs include subsequent TSF stage raises over the life of mine, plant annual capital expenditures, mine closure costs, salvage value and rehabilitation costs. These costs are not included in the pre-production capital cost summary covered in Section Lycopodium Minerals Canada

389 NI TECHNICAL REPORT - REFERENCES Page The sustaining capital schedule over the life of mine is estimated as shown in Table Table Sustaining Capital Schedule Year Grand Total USD ( 000) 2019 $6, $6, $1, $5, $1, $3, $1, $ $9, $5,675 Grand Total $42, Project Implementation Strategy The implementation strategy for the Project is based on an engineering, procurement and construction management (EPCM) implementation approach and horizontal discipline based contract packaging. Horizontal packages are for the earthworks, building works, concrete works, field erected tankage, structural, mechanical and piping installation, electrical and instrumentation supply and installation. An experienced engineering firm will be engaged to provide engineering and procurement (EP) services for the development of the process plant and the associated infrastructure. An experienced engineering firm will also be engaged to provide construction management (CM) services as part of an integrated team with Birimian for the development of the process plant and the associated infrastructure. Lycopodium Minerals Canada

390 NI TECHNICAL REPORT - ECONOMIC ANALYSIS Page ECONOMIC ANALYSIS 22.1 Introduction A financial model for evaluating the Project was developed jointly by SEMAFO and Lycopodium with input provided by the following: AMC - Responsible for the open pit design, mine plan and production schedule, capital and operating costs related to the mine. Lycopodium - Responsible for the process plant design, process selection, capital and operating costs for the process plant and infrastructure. SEMAFO Responsible for application of new Burkina Faso tax legislation, providing the market assumptions including the gold price projections and all the remaining information needed to complete the financial analysis. All costs are constant in USD with no provision for inflation escalation. The annual cash flow projections were estimated over the Project s production life based on production schedule, sales revenue, production costs, capital expenditures and corporate costs (taxation, royalties, etc.). The financial indicators examined included pre and post-tax cash flow, net present value (NPV) at 5% discount rate, internal rate of return (IRR) and payback period Taxes The tax planning for the Project was developed by SEMAFO taking into consideration the new mining tax laws applied to capital costs, operating costs, sales of gold and profits. The following assumptions have been applied when calculating corporate tax payable: 27.5% corporate tax rate (for the purpose of the financial model, this is applied to the Project cash flows only). 1% development tax. Tax payable periodically (assumed quarterly). Project assets are depreciated over their useful life, according to Burkina Faso Taxation legislation. Lycopodium Minerals Canada

391 NI TECHNICAL REPORT - ECONOMIC ANALYSIS Page Value Added Tax A Value Added Tax (also known as TVA - Taxe sur la valeur ajoutée) at a rate of 18% is levied by the Burkina Faso Government on purchases by individuals and corporations on non-exempt goods and services. Businesses can claim back TVA on most business inputs except for fuel. It is assumed all of the product sales will be to overseas customers, so TVA is not applicable Royalties Burkina Faso Royalties apply to the Project and the rate applied is based on the gold price. For the Project, a 4% royalty rate has been applied Other Royalties / Agreements No royalties or payments other than those described have been included within the financial analysis Revenue Deductions The cost of shipping, insurance and refining is a revenue deduction for the purpose of determining net revenue Project Financing No assumptions have been made about the Project financing in the financial model Financial Model Inputs Key Economic Assumptions Lycopodium Minerals Canada

392 NI TECHNICAL REPORT - ECONOMIC ANALYSIS Page 22.3 Table 22.1 summarizes the key assumptions used in the financial model. Table 22.1 Key Assumptions Used in the Financial Model Assumption Value Commodity Prices Gold price 1,100 USD/oz Silver price USD/oz Exchange Rates USD:CAD 1.39 USD:AUD 1.37 USD:EURO 0.92 USD:ZAR USD:GBP 0.67 USD:CFA 600 Price inflation N/A Escalation N/A Discount Rate 5% Royalties 4% Taxes Corporate Tax Rate 27.5% TVA 18% WHT 10% Production summaries The production schedule for the Project is described in Section Capital Cost Summary The capital costs for the Project are as described in Section Operation Cost Summary The operating costs for the Project are as described in Section Economic Results The pre and post-tax NPV, IRR and payback results for the Project based on the assumptions above are summarized in Table Lycopodium Minerals Canada

393 NI TECHNICAL REPORT - ECONOMIC ANALYSIS Page 22.4 Table 22.2 Economic Results Pre-Tax Post-Tax NPV at 5% discount rate USD 343 M USD 262 M IRR 55% 48% Payback Cash Flow The life of mine schedule, ore processing schedule, gold recovery, gold revenues, operating costs, initial capital, deferred capital, sustaining capital, taxation, calculated NPV and IRR are shown in Table Lycopodium Minerals Canada

394 NI TECHNICAL REPORT - ECONOMIC ANALYSIS Page 22.5 Table 22.3 Cash Flow Total or Average Construction Year Year Year Year Year Year Year Year LOM Period MINE SCHEDULE Waste Mined (t) 67,840,054 17,020,156 12,508,621 8,614,138 4,665,929 11,570,114 11,507,941 1,953,156 Capitalized Stripping Activity (t) 61,945,571 18,109,796 3,087,493 7,262,607 10,134,443 14,789,050 4,608,056 3,954,126 - Ore Mined (t) 9,567,241 72,610 1,714,444 1,712,439 2,120,648 1,053,503 1,262,403 1,178, ,482 Total Strip Ratio Total Mined (t) 139,352,866 18,182,406 21,822,093 21,483,667 20,869,229 20,508,481 17,440,573 16,640,778 2,405,638 Ore Grade (g/t) PROCESSING SCHEDULE Ore Processed 9,567,241 1,256,010 1,343,200 1,343,200 1,343,200 1,346,880 1,343,200 1,343, ,351 Head Grade (g/t) Recovery (%) 92.9% 93.9% 93.7% 93.7% 92.4% 92.7% 91.2% 90.0% 87.4% Gold - Recovered (oz) 1,184, , , , , , ,030 84,807 10,676 REVENUES (in $000s) Revenues Gold Sales 1,305, , , , , , ,105 93,560 11,794 Cost of Production (427,912) (61,768) (58,181) (40,992) (52,783) (67,179) (68,846) (64,883) (13,280) OTHER OPERATING COSTS Royalties (52,215) (9,907) (9,959) (10,065) (6,374) (6,932) (4,764) (3,742) (472) Selling Costs (3,155) (585) (589) (597) (390) (422) (299) (242) (31) Other (19,406) (3,394) (3,214) (3,042) (1,918) (2,625) (2,180) (1,925) (1,109) TAXES (97,411) - - (36,788) (36,027) (8,048) (12,875) (3,673) - Working Capital - (6,729) (7,708) (16,721) 5,979 1,362 3, ,301 3,118 Initial Capex (169,795) (169,795) Initial Supplies Inventory - (7,218) 3,609 3,609 Sustaining Capex (25,043) (6,690) (5,919) (1,500) (5,121) (1,500) (2,813) (1,500) - Capitalized Stripping Activity (147,477) (42,378) (7,028) (17,112) (23,660) (35,807) (11,332) (10,160) - - Deferred Capex (15,249) (3,539) (5,257) (5,455) (998) Rehabilitation & Closure Costs (17,249) (257) (257) (257) (257) (257) (257) (257) (15,450) NET CASH FLOW 330,477 (219,391) 147, , ,547 25,655 76,376 20,309 38,248 (11,821) Total Cash Cost /oz ,284 All-in Sustaining Cost /oz ,284 Lycopodium Minerals Canada

395 NI TECHNICAL REPORT - ECONOMIC ANALYSIS Page Sensitivities The capital cost, operating cost, gold price and head grade were varied by ±5% up to ±20% to determine the sensitivity of the project. The revenue from the varying head grade was determined using the gold recovery formula in Section 13. The resulting pre-tax NPV and %IRR was plotted to show the project sensitivities in Figure 22.1 and Figure The project s economics are sensitive to head grade and gold price. Further, the project sensitivity was performed using a USD 100 variation from the base case gold price as illustrated in Table At a gold price of USD 1,200/oz, the project after-tax NPV increased by 27% to USD 334 million and the after-tax IRR increased to 58%. Table 22.4 Gold Price Sensitivity $1,000 oz Gold Base Case USD 1,100 oz Gold USD 1,200 oz Gold After-tax operating cash flow (USD M) $256 $330 $413 After-tax 5% NPV (USD M) $199 $262 $334 After-tax IRR (%) Payback period (years) Lycopodium Minerals Canada

396 NI TECHNICAL REPORT - ECONOMIC ANALYSIS Page 22.7 Figure 22.1 NPV Sensitivity Analysis (Pre-tax) Discounted Cash Flow (USD '000,000) $600 $550 $500 $450 $400 $350 $300 $250 $200 Natougou DFS Sensitivity -20% -15% -10% -5% 0% +5% +10% +15% +20% % Change from Base Case CAPEX OPEX Gold Price Grade Figure 22.2 IRR Sensitivity Analysis (Pre-tax) 90% Natougou DFS Sensitivity 80% 70% %IRR 60% 50% 40% 30% -20% -15% -10% -5% 0% +5% +10% +15% +20% % Change from Base Case CAPEX Gold Price Grade OPEX Lycopodium Minerals Canada

397 NI TECHNICAL REPORT - ADJACENT PROPERTIES Page ADJACENT PROPERTIES The Tapoa Permit Group is surrounded by other registered permits as shown in Figure All but two permits contiguous to the Tapoa Permit Group are privately owned. The two southwest contiguous permits are believed to be held under Cluff Mining (now Amara Mining plc); however, no additional information was obtained from public records for these permits. Figure 23.1 Adjacent permit boundaries Source: SEMAFO 5047-\24.0\5047-STY-001_0 Snowden Mining Consultants

398 NI TECHNICAL REPORT - OTHER RELEVANT DATA AND INFORMATION Page OTHER RELEVANT DATA AND INFORMATION As far as Lycopodium, Snowden, Knight Piésold, WSP and AMC Consultants are aware, there is no other relevant data or information pertaining to this Technical Report \24.0\5047-STY-001_0 Lycopodium Minerals Canada

399 NI TECHNICAL REPORT - INTERPRETATION AND CONCLUSIONS Page INTERPRETATION AND CONCLUSIONS 25.1 Geology, Exploration and Mineral Resource The Tapoa Permit Group, including the Natougou project, is located in a well-known gold province and the property has a history of artisanal mining. Results from the exploration rock chip and soil sampling programs conducted by Orbis between 2010 and 2014 across the Tapoa Permit Group, indicate that there are multiple targets which warrant further exploration to assess the potential of additional gold mineralisation in the area. The Natougou deposit represents a significant gold discovery in the area. The project has been drilled using diamond core and RC drilling techniques down to a nominal spacing of 20 m by 20 m in a significant portion of the deposit area. The author is satisfied that the drill sample database and geological interpretations are sufficient to enable the estimation of Mineral Resources and sample security procedures provide confidence in the integrity of the samples and assay results. The geological interpretation carried out by Birimian has considered all material items and represents an accurate reflection of the current geological understanding. Alternative interpretations are unlikely to materially impact on the Mineral Resource estimate. Twin drillhole results indicate reasonable repeatability of gold grades on the property, although only minimal twin holes have been completed to date. SEMAFO included RC drilling in the infill drilling conducted at the Natougou deposit to bring the drill spacing down to a nominal 20 m along strike by 20 m across strike. RC drilling at Natougou typically shows poorer sample recovery than the diamond core drilling, which generally has excellent recovery. To assess the impact of the drilling method on the grade distribution of the mineralized samples, Snowden compared RC and DD intersections within the 20 m by 20 m area where the coverage of RC and DD is relatively equal. The comparison shows that on average the RC drilling is lower grade than the diamond core drilling. The difference appears the vary somewhat from approximately 10% for samples above 3 g/t Au to as much as 50% for samples between 0.1 g/t Au and 3 g/t Au. Some of the observed difference is likely due to the local geological variability. Additionally, as the RC samples are collected based on a nominal 1 m interval, due to the relatively narrow nature of the mineralization, dilution of samples at the upper and lower boundaries of the shear zone by un-mineralized footwall and hangingwall material is likely contributing to the observed difference. There is a risk that in areas with significant RC drilling the Mineral Resource estimate will effectively be semi-diluted and the grade of the mineralization may be slightly underestimated. It is recommended that SEMAFO twin some holes so that a more robust statistical assessment of the two drilling methods can be conducted. Accepted estimation methods have been used to generate a 3D block model of gold values. In Snowden s opinion, the use of MIK estimation technique with unfolding is appropriate for the highly skewed gold grade distribution and folded nature of the deposit. The estimate has been classified with respect to CIM guidelines with the resources classified at a Measured, Indicated and Inferred status, according to the geological confidence and sample spacing that currently define the deposit \24.0\5047-STY-001_0 Lycopodium Minerals Canada

400 NI TECHNICAL REPORT - INTERPRETATION AND CONCLUSIONS Page Mining Open pit mining will use conventional open pit truck and excavator method with contractor mining. Mining activities that will occur as part of the sequence are: grade control drilling, drill-and-blast, and load-and-haul. Grade control will be performed using reverse circulation (RC) drilling and sampling methods, on a 0.5 m sampling interval and a 10 m X (east) x 10 m Y (north) grid. Grade control will play an important role in the mining process to ensure that the mineralization of interest is adequately defined prior to mining, and to help minimize dilution. Selectivity of the mining equipment at the face and good blasting practices will also be required to minimize dilution and ore loss. Blasting will require particular attention during operations to ensure adequate fragmentation and tracking of the blast movements, as the ore is not visually distinguishable from the waste rock. Due to the inherent hardness of the rock and flat dip of the mineralization, flexibility will be required in the blast designs, especially holes length, to ensure appropriate explosive distribution within the ore. The first three years of production are key to the project, when approximately 225 koz of gold per year are produced. The high grades contained within the initial stage of the West pit are the main contributors to the high ounces produced at the start of the mine life. The relatively high strip ratio associated with this pit necessitates the pre-stripping of approximately 18 Mt of material in the twelve months prior to first gold pour. Opportunities for optimizing the size of the starter pit and reducing prestripping requirements, while maintaining the ounce production profile, should be investigated further during detailed engineering. The current waste disposal strategy relies on backfilling the mined-out void to minimize haulage costs, as soon as practical and safe to do so. This strategy is cost-effective but will need to be weighed against any ore sterilization potential it represents. The possibility of mining pushbacks, additional to the ones defined in this study, should be reviewed based on updated metal price and other relevant information prior to making a final decision on backfilling the pits Metallurgy The following conclusions can be drawn from the current and previous metallurgical and comminution testwork programs: The Natougou primary ore is an abrasive, competent ore with above average comminution energy requirements. The Natougou primary ore has a high gravity recoverable gold content; leach kinetics are very slow when gravity is not included in the flowsheet. High dissolved oxygen levels and lead nitrate are required to achieve fast leach kinetics and adequate gold recovery \24.0\5047-STY-001_0 Lycopodium Minerals Canada

401 NI TECHNICAL REPORT - INTERPRETATION AND CONCLUSIONS Page 25.3 Anticipated lime consumption for primary ore is low to moderate, provided good quality water can be provided on site. Cyanide consumption is likely to be moderate. High lime consumption will be experienced if oxide ore forms part of the feed blend. LOM head grades for the process plant are expected to average 4.15 g/t with a gold recovery of 92.8%. Quantities of oxide ore are expected to be very low, at less than 1% of tonnes scheduled for processing Environmental and Permitting The Mining Code of Burkina Faso guarantees a stable fiscal regime for the life of any mine developed. It also guarantees stabilization of financial and customs regulations and rates during the period of operation to reflect the rates in place at the date of the approved Operating Permit. The Mining Code also states that no new taxes can be imposed with the exception of mining duties, taxes and royalties. The application for an Operating Permit requires that a Feasibility Study must first be accepted by the Ministry of Environment and Sustainable Development. The Feasibility Study submitted must include an Environmental and Social Impact Assessment (ESIA) which in turn must include a Resettlement Action Plan (RAP) that has been accepted by all stakeholders. Both the ESIA and RAP are expected to be completed imminently and plans are to append those documents to this Feasibility Study for submission to the government in the coming months as part of the application for an Operating Permit. No other permitting constraints have been identified. Geochemical studies have been conducted to assess the potential for acid drainage (Acid Rock Drainage: ARD) and metal leaching (ML) of the waste rock and construction materials as well as CIP tailings and heap leach solids. It was found that the direct seepage form the waste dump should meet Burkina Faso Effluent Discharge Criteria for release to surface water. Most of the woody species identified during the inventories on the entire study limit which is more than ten times larger than the project area are considered of Low Concerned by UICN (Union Internationale pour la Conservation de la Nature 1 ) and very few species have been identified as Vulnerable. All mammals observed in the project area are common species with the ability to tolerate a high level of human activity. The resettlement of a large number of people (165 households or approximately 895 people), and a large area of agricultural land (approximately 813 ha) represents an important activity that will require immediate and focused effort. 1 International Union for Conservation of Nature 5047-\24.0\5047-STY-001_0 Lycopodium Minerals Canada

402 NI TECHNICAL REPORT - INTERPRETATION AND CONCLUSIONS Page 25.4 Since June 2015, Biriman has had ongoing formal public community engagement meetings through a Provincial Compensation and Resettlement Committee approved by government. Through these meetings the project has been introduced to the communities and opportunities for concerns and questions regarding the project impacts and resettlement plans to be tabled and addressed. This process is ongoing. It is important to note that SEMAFO have direct experience with resettlement on a large scale at previous project in Burkina Faso Risks Resettlement Delays It is important to note that delays to the relocation program early on could result in delays to the West Catchment Dyke and East Catchment Dyke completion and missing the rain-season window in time to fill the those ponds with water. This would effectively postpone start-up until the next rain season (one year). This potential problem in water supply could be mitigated by tapping in a river that has been identified outside of the watershed of the Project Site \24.0\5047-STY-001_0 Lycopodium Minerals Canada

403 NI TECHNICAL REPORT - RECOMMENDATIONS Page RECOMMENDATIONS 26.1 Geology and Mineral Resources The following recommendations are made with respect to ongoing work at the Natougou gold project: Snowden recommends that SEMAFO incorporate field duplicate samples at a nominal rate of around 1:20 as part of the QAQC protocols for RC drilling. This will allow an assessment to be made of the sample precision for the RC drilling and facilitate ongoing QC monitoring. Moreover, this will be especially important if RC drilling is contemplated for grade control purposes during open pit mining. To date, only two twin diamond holes have been drilled to validate the RC drilling results, which constitute approximately 15% of the mineralized samples. Comparisons between RC and DD drilling within an area of relatively equal coverage suggest some differences do occur, with the RC drilling lower grade on average. If further RC drilling is to be completed, it is recommended that SEMAFO increase the number of twin holes so that a more robust statistical assessment of the two drilling methods can be conducted. Twin diamond holes should be drilled as close as possible to the original RC drillhole, ideally within 3 m. Ideally larger intervals of core should be used to measure the bulk density to provide a more representative measurement. Additionally, wax coating of oxidised core samples should be done to provide accurate bulk density measurements within the oxidised zones. Snowden recommends that SEMAFO generate a geological interpretation of the western hangingwall mineralisation to constrain the gold grade estimates in this area, with the ultimate aim of potentially including this material in future Mineral Resource estimates Mining AMC suggests that the following recommendations are followed in developing operational mine plans: Review the mine designs to minimize waste mining at the footwall contact. Issue request for quotes to mining contractors to confirm mining costs and understand contractor s equipment availability and size, which may impact on the mine design, especially ramp width and bench height. Review the size of the starter pit within the West pit to evaluate whether pre-strip costs could be reduced without impacting on gold ounces produced in early project life. Conduct drill and blast tests to ascertain the possibility to drill and blast larger bench heights in waste and thereby reduce costs \16.04\5047-STY-001_0 Lycopodium Minerals Canada

404 NI TECHNICAL REPORT - RECOMMENDATIONS Page Environmental and Permitting Given the importance to the construction schedule and required seasonal timings it is recommended that the ESIA report be submitted as soon as possible to the regulatory authorities and participate in both the COTEVE evaluations and public hearings. Resettlement efforts should continue before the mining permit is granted to allow for resettlement activities to begin as soon as possible Overall Recommendation It is recommended that Birimian advance the implementation of the project as soon as possible. Sufficient metallurgical testwork and trade off studies have completed to confidentially establish the process design criteria. Capital and operating costs have been estimated to a level of accuracy consistent with a definitive feasibility study. It is recommended that Birimian file the environmental study impact assessment as soon as possible with the government of Burkina Faso \16.04\5047-STY-001_0 Lycopodium Minerals Canada

405 NI TECHNICAL REPORT - REFERENCES Page REFERENCES ANCOLD - Australian National Committee on Large Dams, Guidelines on Tailings Dams, Planning, Design, Construction, Operation and Closure, May 2012 ANCOLD - Australian National Committee on Large Dams, Guidelines on the Consequence Categories for Dams, October 2012 Appiah, H., and Norman, D.I The Geology of the Prestea and Ashanti Goldfields: A Comparative Study in: Ladeira E A (Ed.), Brazil Gold 91 Balkema, Rotterdam. pp C14757 Gravity Recoverable Gold Test Work Report, Consep, 11/12/14. C15554 Gravity Recoverable Gold and Intensive Cyanidation Test Work Report, Consep, 30/08/15. Dzigbodi-Adjimah, K Geology and geochemical patterns if the Birimian gold deposits, Ghana, West Africa, Journal of Geochemical Exploration vol. 47 issue 1-3 April, P Golder Associates, 2015, Geotechnical feasibility study of open pit walls for Natougou gold project, Burkina Faso, Africa, interim report, 13pp. Golder Associates, 2015, Interim Report on Geotechnical Feasibility Study for Open Pit Walls for Natougou Gold Project, Burkina Faso, Africa, (Golder report No R-Rev0). Golder Associates, Stage II Groundwater Study, Feasibility Study, Natougou Gold Project, Burkina Faso, Africa Report reference # R-Rev1, March Government of Burkina Faso, Burkina Faso Environmental Law, Chapter III: Water Quality Standards. Translated version (not dated). GSHAP Global Seismic Hazard Assessment Program Global Seismic Hazard Map, 1999 Institut de Gestion des Risques Miniers et du Développement (INGRID), État initial de la flore dans la zone, d étude écologique de Natougou, December 2013 and August 2014 INGRID État initial de la faune dans la zone, d etude ecologique de Natougou, January, June and November INGRID État initial de l écologie aquatique dans la zone, d etude du projet Natougou, November INGRID Revue non structurée de l'état initial du milieu social, projet Natougou Gold Février \16.04\5047-STY-001_0 Lycopodium Minerals Canada

406 NI TECHNICAL REPORT - REFERENCES Page 27.2 INGRID Étude pédologique, projet aurifere de Natougou, December Knight Piésold Consulting, Natougou Gold Project, Baselin Design Climatology Technical Memo Reference PE401-76/2-A djtm M14013, June 2014 KOTE, L. and K. LEONCE Prospection archéologique sur le périmètre minier de Boungou, Rapport d étude d impact. August Knight Piésold Consulting, Natougou Gold Project, Waste Rock and Tailings Geochemical Characterisation Report Report Reference BR /11, June 2015 Knight Piésold Consulting 1, Natougou Gold Project, Definitive Feasibility Study Report, Tailings Storage Facility and Site Water Management Report Reference BR /13, January 2016 Knight Piésold Consulting 2, Natougou Gold Project, Geotechnical Investigation Report Revised Plant site and TSF Location Report Reference BR /14, January 2016 Leach Residence Time Selection, Lycopodium, May Leube, A., Hirdes, W., Mauer, R., and Kesse, G.O The Early Proterozoic Birimian Supergroup of Ghana and some aspects of its Associated Gold Mineralisation, Precambrian Research, 46: Metallurgical Testwork conducted upon Composites from the Natougou Project, Report No. A15733, ALS Metallurgy, February Metallurgical Testwork conducted upon Composites from the Natougou Project, Report No. A16522, ALS Metallurgy, August Metallurgical Testwork conducted upon Composites from the Natougou Project, Report No. A16543, ALS Metallurgy, November Natougou Comminution Circuit Design, Report No. 7583, Orway Mineral Consultants, November Natougou Detailed Feasibility Study Process Flowsheet Comparisons Business Case Analysis (Flotation versus Whole of Ore Leach), Lycopodium, June Natougou Grind Size Optimization, Lycopodium, September Pohl, W., and Gunther, M Preliminary results of fluid inclusions research on tin, tungsten and gold deposits of Rwanda, Central Africa, IGCP Project No 255 Newsletter1, 19-23, TU Braunschweig MRAC Tervuren \16.04\5047-STY-001_0 Lycopodium Minerals Canada

407 NI TECHNICAL REPORT - REFERENCES Page 27.3 Snowden, Natougou gold deposit, Burkina Faso, Mineral Resource Estimate, unpublished internal report prepared by Snowden Mining Industry Consultants for Orbis Gold Limited, project number AU4122, August Snowden, 2014a. Independent Technical Report, Valuation of Orbis Gold Ltd Mineral Assets, report prepared by Snowden Mining Industry Consultants for Orbis Gold Limited, project number AU4523, December Snowden, 2014b. Natougou gold deposit, Burkina Faso, Mineral Resource estimate, unpublished internal report prepared by Snowden Mining Industry Consultants for Orbis Gold Limited, project number AU4450, August Snowden, Tapoa Permit Group, Natougou Gold Deposit, NI Technical Report prepared by Snowden Mining Industry Consultants for SEMAFO Inc., project number AU4582, March S2574 Thickening Report - Part A and Part B, Outotec, 18/06/15. Report of Investigation into the Thickening and Filtration of Natougou Gold Pre-Leach and Tailings for Birimian Resources SARL, FLS, December Tailings Trade-off Study, Lycopodium, September WHO World Health Organization, Guidelines for Drinking Water Quality, 4th Edition, T2 T3 Natougou External Test Report Rev 00, Filtration Report, Outotec, 30/06/ REP-001_B, Metallurgical Testwork Report, Lycopodium, November \16.04\5047-STY-001_0 Lycopodium Minerals Canada

408 Certificate of Qualified Person I, Neil Lincoln, of Mississauga, Ontario, Canada, do hereby certify that as the author of NI Technical Report, Natougou Gold Deposit Project, Burkina Faso dated 23 that: 1. I am employed as the VP Business Development and Studies with Lycopodium Minerals Canada Ltd, 5060 Spectrum Way, Suite 400, Mississauga, ON, Canada. 2. I graduated from the University of the Witwatersrand, South Africa, in 1994 with a Bachelor of Science in Metallurgy and Materials Engineering (Minerals Process Engineering) degree. 3. I am a professional engineer in good standing with the Professional Engineers Ontario (PEO) in Canada (no ). 4. I have practiced my profession continuously as a metallurgist for 22 years and with Lycopodium Minerals since I am responsible for compiling the overall technical report and responsible for sections 1, 2, 3, , 19, 21 (except for and ), 22, 24, 26.4, and I have read the definition of qualified person set out in National Instrument (NI ) and certify that, by reason of my education, affiliation with a professional association (as defined in NI ) and past relevant work experience, I fulfill the requirements to be a qualified person for the purpose of NI I visited the site during May 2015, and inspected the proposed open pit, process plant, water supply dams, tailings storage facility and mine infrastructure areas. 8. I am independent of SEMAFO in accordance with the application of Section 1.5 of National Instrument I have not had any prior involvement with the Property. 10. I have read National Instrument and Form F1 and the Technical Report has been prepared in compliance with same. 11. At the effective date of the Technical Report, to the best of my knowledge, information and belief, the Technical Report contains all scientific and technical information that is required to be disclosed to make the Technical Report not misleading. Dated at Mississauga, Ontario, Canada, this 23rd day of. [signed] Neil Lincoln, P.Eng Lycopodium Minerals Canada Ltd 5060 Spectrum Way, Suite 400, Mississauga, ON, L4W 5N5, Canada

409 Certificate of Qualified Person I, Marius Ward Phillips, of Queensland, Brisbane, Australia, do hereby certify that as the author of NI Technical Report, Natougou Gold Deposit Project, Burkina Faso dated 23 that: 1. I am employed as the Senior Process Consultant with Lycopodium Minerals Pty Ltd, 163 Leichhardt Street, Spring Hill, Queensland, Australia. 2. I graduated from the Technikon of the Witwatersrand, South Africa, in 1997 with a National Higher Diploma in Extractive Metallurgy. 3. I am a chartered professional engineer in good standing with the Australian Institute of Mining and Metallurgy in Australia (no ). 4. I have practiced my profession continuously as a metallurgist for 24 years and with Lycopodium Minerals since I am responsible for compiling the overall technical report and responsible for sections 13, 17 and I have read the definition of qualified person set out in National Instrument (NI ) and certify that, by reason of my education, affiliation with a professional association (as defined in NI ) and past relevant work experience, I fulfill the requirements to be a qualified person for the purpose of NI I have not visited the site. 8. I am independent of SEMAFO in accordance with the application of Section 1.5 of National Instrument I have not had any prior involvement with the Property. 10. I have read National Instrument and Form F1 and the Technical Report has been prepared in compliance with same. 11. At the effective date of the Technical Report, to the best of my knowledge, information and belief, the Technical Report contains all scientific and technical information that is required to be disclosed to make the Technical Report not misleading. Dated at Spring Hill, Queensland, Brisbane, Australia, this 23rd day of. [signed] Marius Phillips, (MAusIMM, CP) Lycopodium Minerals Pty Ltd 163 Leichhardt, Spring Hill, Brisbane, QLD, 4004, Australia

410

411