GABRIEL RESOURCES LTD.

TECHNICAL REPORT ON THE ROSIA MONTANA GOLD PROJECT TRANSYLVANIA, ROMANIA

Brett L. Gossage, MAusIMM Coffey Mining Pty. Ltd. John M. Marek, P.E. Independent Mining Consultant, Inc. Christopher R. Lattanzi, P. Eng. Micon International Limited

Patrick G. Corser, P.E. MWH Americas, Inc. Stuart Smith, MAusIMM Aurifex Pty. Ltd.

March 4, 2009

Table of Contents
1.0 SUMMARY..........................................................................................................................................................1 1.1 MINERAL RESOURCES AND MINERAL RESERVES ...........................................................................2 1.2 UPDATED TECHNICAL AND ECONOMIC REVIEW ..............................................................................3 1.2.1 Methods of Production...............................................................................................................3 1.2.2 Project Facilities ........................................................................................................................7 1.2.3 Production Schedule .................................................................................................................9 1.2.4 Environmental and Socio-Economic Considerations .................................................................9 1.2.5 Permitting ................................................................................................................................11 1.2.6 Cost Structure .........................................................................................................................11 1.2.7 Project Economics...................................................................................................................13 1.3 CONCLUSIONS AND RECOMMENDATIONS......................................................................................14 INTRODUCTION AND TERMS OF REFERENCE ...........................................................................................16 RELIANCE ON OTHER EXPERTS ..................................................................................................................21 PROPERTY DESCRIPTION AND LOCATION.................................................................................................22 4.1 LAND TENURE .....................................................................................................................................23 4.2 MINERAL OWNERSHIP........................................................................................................................23 4.3 EXPLORATION CONCESSIONS..........................................................................................................23 4.4 EXPLOITATION CONCESSIONS .........................................................................................................24 ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY ..............25 5.1 ACCESS ................................................................................................................................................25 5.2 CLIMATE ...............................................................................................................................................25 5.3 INFRASTRUCTURE..............................................................................................................................25 5.4 PHYSIOGRAPHY ..................................................................................................................................25 HISTORY ..........................................................................................................................................................26 6.1 EXPLORATION AND MINING...............................................................................................................26 6.2 TECHNICAL AND ECONOMIC STUDIES.............................................................................................29 6.3 RESOURCE ESTIMATES .....................................................................................................................32 GEOLOGICAL SETTING .................................................................................................................................34 7.1 REGIONAL GEOLOGY .........................................................................................................................34 7.2 PROJECT GEOLOGY AND MINERALIZATION CONTROLS...............................................................36 7.2.1 Structure..................................................................................................................................39 7.2.2 Alteration .................................................................................................................................42 DEPOSIT TYPES..............................................................................................................................................43 PROJECT MINERALIZATION..........................................................................................................................44 9.1 MINERALIZED LITHOLOGIES..............................................................................................................44 9.2 VEINING ................................................................................................................................................45 EXPLORATION ................................................................................................................................................48 DRILLING .........................................................................................................................................................49 11.1 RC DRILLING ........................................................................................................................................49 11.2 DIAMOND DRILLING ............................................................................................................................49 11.3 CHANNEL SAMPLING ..........................................................................................................................50

2.0 3.0 4.0

5.0

6.0

7.0

8.0 9.0

10.0 11.0

11.4 12.0

SIGNIFICANT INTERSECTIONS ..........................................................................................................50

SAMPLING METHOD AND APPROACH.........................................................................................................60 12.1 INTRODUCTION ...................................................................................................................................60 12.2 CHANNEL SAMPLING ..........................................................................................................................60 12.3 RC DRILLING ........................................................................................................................................63 12.4 DIAMOND DRILLING ............................................................................................................................67 SAMPLE PREPARATION, ANALYSIS AND SECURITY.................................................................................70 DATA VERIFICATION ......................................................................................................................................73 14.1 LOGGING ..............................................................................................................................................73 14.2 QUALITY CONTROL PROCEDURES...................................................................................................73 14.3 STATISTICAL ANALYSIS OF ASSAY QUALITY CONTROL DATA .....................................................74 14.3.1 Introduction..............................................................................................................................74 14.3.2 Assay Accuracy .......................................................................................................................74 14.3.3 Assay Precision .......................................................................................................................76 14.4 GRID CONTROL AND SURVEY ...........................................................................................................90 14.4.1 Topography .............................................................................................................................90 14.4.2 Drill Hole Collar Locations .......................................................................................................90 14.4.3 Downhole Surveying................................................................................................................91 14.4.4 Underground Workings............................................................................................................91 14.4.5 Channel Locations...................................................................................................................92 14.5 BULK DENSITY.....................................................................................................................................92 14.6 VERIFICATION SAMPLING ..................................................................................................................93 ADJACENT PROPERTIES...............................................................................................................................97 MINERAL PROCESSING AND METALLURGICAL TESTING ........................................................................98 16.1 GENERAL..............................................................................................................................................98 16.2 THE RPA TECHNICAL REPORT ..........................................................................................................99 16.2.1 Samples and Flowsheet Development RPA Report .............................................................99 16.2.2 Ore Characterization RPA Report ........................................................................................99 16.2.3 Cyanidation Leaching and Grind Sensitivity RPA Report ...................................................101 16.2.4 Gravity Concentration and Sulphide Recovery RPA Report ...............................................102 16.2.5 Grinding Testwork RPA Report ..........................................................................................102 16.2.6 Variability Testing and Algorithm Development RPA Report ..............................................102 16.2.7 Cyanide Destruction [Detoxification] RPA Report ...............................................................103 16.2.8 Testwork to Develop Design Criteria .....................................................................................103 16.2.9 Flowsheet Options.................................................................................................................103 16.3 AUSENCO MANAGED TESTWORK AND FLOWSHEET MODIFICATIONS .....................................103 16.3.1 Introduction............................................................................................................................103 16.3.2 Samples - Ausenco ...............................................................................................................104 16.3.3 Program - Ausenco................................................................................................................104 16.3.4 Mineralogy and Diagnostic Leaching - Ausenco ....................................................................105 16.3.5 Cyanide Conditions - Ausenco ..............................................................................................105 16.3.6 Carbon Kinetics - Ausenco ....................................................................................................106 16.3.7 Gravity - Ausenco ..................................................................................................................107 16.3.8 Cyanide Detoxification - Ausenco ..........................................................................................107 16.3.9 Thickening and Agitation - Ausenco ......................................................................................108 16.3.10 Viscosity ................................................................................................................................108 16.3.11 Comparison with Previous Testwork......................................................................................108 16.3.12 ARD Testwork - Ausenco ......................................................................................................110

13.0 14.0

15.0 16.0

ii

16.4

16.5

16.6

16.7 16.8 16.9 17.0

NEWMONT TESTWORK PROGRAM .................................................................................................110 16.4.1 Background ...........................................................................................................................110 16.4.2 Program - Newmont ..............................................................................................................111 16.4.3 Gravity Evaluations Newmont.............................................................................................111 16.4.4 Cyanide Concentration - Newmont ........................................................................................112 16.4.5 Lead Nitrate - Newmont.........................................................................................................112 16.4.6 Additional Lime to the Leach to Increase the pH - Newmont .................................................112 16.4.7 Grind Sensitivity - Newmont ..................................................................................................112 16.4.8 Newmont Testwork Influence on Design ...............................................................................113 J. R. GOODE AND ASSOCIATES / RMGC TESTWORK PROGRAM - Phase 1..............................113 16.5.1 Background ...........................................................................................................................113 16.5.2 Grind Sensitivity and Gravity Recovery J. R. Goode Phase 1 .........................................114 16.5.3 Extended Leach Times J. R. Goode Phase 1..................................................................114 16.5.4 Cyanide Consumption J. R. Goode Phase 1 ...................................................................114 16.5.5 Gravity + CIL Leach of Gravity Tail versus Whole of Ore CIL Leaching Phase 1 ...............115 16.5.6 Bond Work Indices - Phase 1 ................................................................................................115 16.5.7 J. R. Goode Testwork Influence on Design Phase 1 ..........................................................115 J. R. GOODE AND ASSOCIATES / RMGC TESTWORK PROGRAM PHASE 2................................117 16.6.1 Background - Phase 2 ...........................................................................................................117 16.6.2 Gravity Separation Phase 2................................................................................................117 16.6.3 Cyanidation Testswork Phase 2 .........................................................................................117 16.6.4 Iron Chemisty Testwork .........................................................................................................118 16.6.5 J. R. Goode Testwork Influence on Design Phase 2 ..........................................................118 DESIGN CRITERIA .............................................................................................................................118 FLOWSHEET ......................................................................................................................................120 EXTRACTION ALGORITHMS .............................................................................................................120

MINERAL RESOURCE AND MINERAL RESERVE ESTIMATES .................................................................124 17.1 DATABASE DEVELOPMENT..............................................................................................................124 17.1.1 Introduction............................................................................................................................124 17.1.2 Validation of Supplied Database............................................................................................124 17.1.3 Database Development .........................................................................................................125 17.2 GEOLOGY AND MINERALIZED DOMAIN MODELLING ....................................................................126 17.2.1 Introduction............................................................................................................................126 17.2.2 Lithological Model..................................................................................................................126 17.2.3 Alteration Model.....................................................................................................................128 17.2.4 Mineralization Modelling ........................................................................................................129 17.2.5 Hardness Modelling...............................................................................................................132 17.2.6 Validation of Geological Interpretation and Wireframe Models ..............................................134 17.3 STATISTICAL ANALYSIS ...................................................................................................................134 17.3.1 Introduction............................................................................................................................134 17.3.2 Data Coding...........................................................................................................................135 17.3.3 Compositing...........................................................................................................................135 17.3.4 Statistical Analysis of Composite Data - Gold and Silver.......................................................138 17.3.5 Outlier Analysis and Determination of Upper Cuts ................................................................140 17.3.6 Declustered Grade Statistics .................................................................................................141 17.3.7 Statistical Analysis of Composite Data - Sulphur...................................................................141 17.3.8 Correlation Analysis of Composite Gold and Silver Data.......................................................144 17.3.9 Statistical Analysis Bulk Density Data.................................................................................144 17.4 VARIOGRAPHY ..................................................................................................................................146 17.4.1 Introduction............................................................................................................................146 17.4.2 Variography Gold and Silver...............................................................................................147

iii

17.5

17.6

17.7 18.0

17.4.3 Variography Sulphur ...........................................................................................................147 17.4.4 Summary ...............................................................................................................................149 BLOCK MODEL DEVELOPMENT.......................................................................................................149 17.5.1 Introduction............................................................................................................................149 17.5.2 Block Construction Parameters .............................................................................................149 17.5.3 Hardness Coding...................................................................................................................150 17.5.4 Bulk Density Assignment .......................................................................................................152 17.5.5 Coding of Underground Workings .........................................................................................152 17.5.6 Block Model Validation ..........................................................................................................153 GRADE ESTIMATION .........................................................................................................................153 17.6.1 Introduction............................................................................................................................153 17.6.2 Ordinary Kriging - Gold and Silver .........................................................................................153 17.6.3 Comparative Estimates .........................................................................................................157 17.6.4 Sulphur Estimate ...................................................................................................................158 17.6.5 Mineralized Waste Dump Estimate........................................................................................158 17.6.6 Validation...............................................................................................................................158 17.6.7 Resource Reporting...............................................................................................................159 17.6.8 Comparative Estimates .........................................................................................................163 MINERAL RESERVES ........................................................................................................................167

OTHER RELEVANT DATA AND INFORMATION..........................................................................................169 18.1 MINING................................................................................................................................................171 18.1.1 Pit Design ..............................................................................................................................171 18.1.2 Mine Equipment.....................................................................................................................173 18.2 PROCESSING.....................................................................................................................................173 18.3 INFRASTRUCTURE AND ANCILLARY FACILITIES ..........................................................................174 18.4 ENVIRONMENTAL AND SOCIO-ECONOMIC CONSIDERATIONS...................................................175 18.4.1 Overview................................................................................................................................175 18.4.2 Environmental Considerations...............................................................................................176 18.4.3 Socio-Economic Considerations............................................................................................177 18.5 PERMITTING ......................................................................................................................................178 18.6 CAPITAL EXPENDITURE ...................................................................................................................180 18.6.1 Summary ...............................................................................................................................180 18.6.2 Basis for the Estimate............................................................................................................181 18.7 Operating Costs...................................................................................................................................183 18.8 PROJECT ECONOMICS.....................................................................................................................184 18.8.1 Base Case.............................................................................................................................184 18.8.2 Sensitivity Analysis ................................................................................................................185 INTERPRETATION AND CONCLUSIONS.....................................................................................................188 RECOMMENDATIONS...................................................................................................................................189 REFERENCES................................................................................................................................................190 CERTIFICATES ..............................................................................................................................................192

19.0 20.0 21.0 22.0

iv

List of Tables
Table 1.1 Grade Tonnage Distribution of Rosia Montana Resource Estimate.................................................................................... 2 Table 1.2 - RMGC Mineral Reserves - February 26, 2009 ..................................................................................................................... 3 Table 1.3 - Process Plant Production Schedule ..................................................................................................................................... 9 Table 1.4 - Estimated Capital Expenditures ......................................................................................................................................... 12 Table 1.5 - Estimated Average Cash Operating Costs......................................................................................................................... 12 Table 1.6 - Summary of Financial Analysis (life-of-mine) ..................................................................................................................... 14 Table 1.7 - Sensitivity of Gold Price ..................................................................................................................................................... 14 Table 2.1 - Currency Exchange Rates.................................................................................................................................................. 18 Table 2.2 - List of Abbreviations ........................................................................................................................................................... 20 Table 6.1 Rosia Montana Gold Project Project History ..................................................................................................................... 30 Table 6.2 Rosia Montana Grade Tonnage : Historic Resource Estimates ........................................................................................ 33 Table 9.1 Vein Orientations Grouped by Region ............................................................................................................................... 47 Table 11.1 - Exploration Drilling and Channel Sampling Summary...................................................................................................... 49 Table 11.2 Summary of Significant Intercepts 2006 Channel Sampling ........................................................................................ 50 Table 11.3 Summary of Significant Intercepts 2007 Channel Sampling ........................................................................................ 52 Table 11.4 Summary of Significant Intercepts 2008 Channel Sampling ........................................................................................ 53 Table 12.1 Channel Sample Weight Assessment ............................................................................................................................. 63 Table 12.2 Summary of Linear Correlation between Recovery and Gold/Silver Grades................................................................... 63 Table 12.3 Theoretical Sample Recoveries RC Drilling ................................................................................................................. 66 Table 12.4 Theoretical RC Sample Recovery versus Gold and Silver Grades ................................................................................. 66 Table 12.5 Igre Region RC Intervals Removed Due to Wet Drilling............................................................................................... 67 Table 12.6 Summary Calculated Diamond Drilling Recoveries......................................................................................................... 69 Table 14.1 Statistical Summary - SGS Gura Rosiei Laboratory - Gold Assays of Standards ........................................................... 77 Table 14.2 Statistical Summary - SGS Gura Rosiei Laboratory - Silver Assays of Standards.......................................................... 77 Table 14.3 Summary of Data Precision : Gura Rosiei Gold Fire Assay Data >0.15 g/t Au ............................................................... 79 Table 14.4 Summary of Data Precision : Precision: 2002 Drilling Program : Gura Rosiei Au Fire Assay Data ...................................... 79 Table 14.5 Summary of Data Precision : pre-June 2003 Period: Gura Rosiei Gold Fire Assay Data >0.1 g/t Au............................. 80 Table 14.6 Summary of Data Precision : post-June 2003 Period: Gura Rosiei Gold Fire Assay Data >0.1 g/t Au ........................... 80 Table 14.7 Summary of Data Precision : 2004 Period: Gura Rosiei Gold Fire Assay Data >0.1 g/t Au............................................ 80 Table 14.8 Summary of Data Precision : 2005 Period: Gura Rosiei Gold Fire Assay Data >0.1 g/t Au............................................ 80 Table 14.9 Summary of Precision and Relative Bias ........................................................................................................................ 82 Table 14.10 Summary of Precision and Relative Bias: 2001 to 2002 ............................................................................................... 83 Table 14.11 Summary of Precision and Relative Bias: 2003 to 2004 ............................................................................................... 84 Table 14.12 Summary of Data Precision : Silver Assay Data > 3.0 g/t Au ........................................................................................ 85 Table 14.13 Summary of Data Precision : 2002 period: Silver Assay Data >3.0 g/t Ag .................................................................... 86 Table 14.14 Summary of Data Precision : pre-June 2003 Period : Silver Assay Data >3 g/t Ag ...................................................... 86 Table 14.15 Summary of Data Precision : post-June 2003 period : Silver Assay Data >3 g/t Ag ..................................................... 86 Table 14.16 Summary of Data Precision : 2004 period : Silver Assay Data >3 g/t Ag ...................................................................... 87 Table 14.17 Summary of Data Precision : 2005 period : Silver Assay Data >3 g/t Ag ...................................................................... 87 Table 14.18 Summary of Precision and Relative Bias : Silver Assays >3.0 g/t Au ........................................................................... 88 Table 14.19 Summary of Precision and Relative Bias: 2001 2002 period: Silver Assays >3 g/t Ag .............................................. 89 Table 14.20 Summary of Precision and Relative Bias: 2003 to 2004: Silver Assays > 3.0 g/t Ag .................................................... 90 Table 14.21 Rosia Montana Drillhole Database : Drill Twins........................................................................................................... 95 Table 14.22 Summary Statistics - Data Pairs : Drilling vs Underground Channel Sampling............................................................ 96

Table 16.1 - Summary of Head Assays .............................................................................................................................................. 100 Table 16.2 - Diagnostic Leach Data % Gold Distribution ................................................................................................................... 101 Table 16-3 - Comparison of Heap Leach and Conventional Leaching............................................................................................... 101 Table 16.4 - Gold Recovery as a Function of Grind Size ................................................................................................................... 101 Table 16.5 Composite Details ......................................................................................................................................................... 104 Table 16.6 - Relative Mineralogical Abundances ............................................................................................................................... 105 Table 16.7 - Grind Sensitivity ............................................................................................................................................................. 106 Table 16.8 - Carbon Kinetic Parameters ............................................................................................................................................ 106 Table 16.9 - Gravity + CIL Leach versus CIL Leach........................................................................................................................... 107 Table 16.10 - Summary Thickening Results....................................................................................................................................... 108 Table 16.11 - Gravity Recovery as a Function of Grind Size.............................................................................................................. 112 Table 16.12 - Comparisons of Baseline Gravity + CIL Gold and Silver Extractions vs. Grind............................................................ 112 Table 16.13 - Comparison of Cirnic Gravity + CIL and Whole of Ore CIL .......................................................................................... 116 Table 16.14 - Comparison of Cetate Gravity + CIL and Whole of Ore CIL......................................................................................... 116 Table 16.15 Comparison of Whole of Ore Leaching and Gravity /Tail Leaching............................................................................. 117 Table 16.16 - Process Design Criteria................................................................................................................................................ 119 Table 17.1 Assay Database - Character Replacements ................................................................................................................. 125 Table 17.2 Exploration Drilling and Channel Sampling Summary................................................................................................... 125 Table 17.3 Hardness Model ............................................................................................................................................................ 134 Table 17.4 Domain Coding.............................................................................................................................................................. 136 Table 17.5 Summary Statistics - Gold ............................................................................................................................................. 138 Table 17.6 Summary Statistics - Silver............................................................................................................................................ 139 Table 17.7 Outlier Analysis - Sub-divided by Estimation Domain ................................................................................................... 142 Table 17.8 Comparison of Nave and Declustered Cut Composite Gold Datasets ......................................................................... 143 Table 17.9 Sulphur Data (%) - Descriptive Statistics....................................................................................................................... 144 Table 17.10 Bulk Density Data - Statistical Summary : Subdivided by Group Lithology ................................................................. 144 Table 17.11 Summary Variogram Models - Gold ............................................................................................................................ 148 Table 17.12 Summary Variogram Models - Silver........................................................................................................................... 148 Table 17.13 Summary Variogram Models - Sulphur ....................................................................................................................... 148 Table 17.14 Block Model Dimensions ............................................................................................................................................. 149 Table 17.15 Block Model Variables ................................................................................................................................................. 150 Table 17.16 Block Model Domain Coding ....................................................................................................................................... 151 Table 17.17 Block Model Bulk Density Assignments ...................................................................................................................... 152 Table 17.18 Summary of Coded Stope Volumes versus the Block Model Volumes ....................................................................... 152 Table 17.19 Sample Search Parameters - Ordinary Kriging Gold and Silver.................................................................................. 156 Table 17.20 Ordinary Kriging Estimation Domains.......................................................................................................................... 157 Table 17.21 Comparison of the Mean Block Model Grade : Gold ................................................................................................... 161 Table 17.22 Confidence Levels of Key Categorization Criteria ....................................................................................................... 161 Table 17.23 Grade Tonnage Distribution of Rosia Montana Resource Estimate............................................................................ 163 Table 17.24 Grade Tonnage Distribution Grouped by Region ........................................................................................................ 164 Table 17.25 Grade Tonnage Distribution : UC Check Estimate ...................................................................................................... 165 Table 17.26 Grade Tonnage Distribution : Aggressive Cuts ........................................................................................................... 165 Table 17.27 Grade Tonnage Distribution : Uncut Composites ......................................................................................................... 166 Table 17.28 Grade Tonnage Distribution : Inverse Distance............................................................................................................ 166 Table 17.29 Grade Tonnage Distribution : Nearest Neighbour ........................................................................................................ 167 Table 17.30 - RMGC Mineral Reserves - February 26, 2009 ............................................................................................................. 167 Table 18.1 Pit Design Criteria.......................................................................................................................................................... 171

vi

Table 18.2 Mine Plan Summary, by Pit ........................................................................................................................................... 172 Table 18.3 Mine Production Schedule............................................................................................................................................. 172 Table 18.4 Process Plant Production Schedule .............................................................................................................................. 174 Table 18.5 - Estimated Capital Expenditures ..................................................................................................................................... 180 Table 18.6 - Estimated Average Cash Operating Cost ...................................................................................................................... 183 Table 18.7 - Unit Cash Cost of Gold Production ................................................................................................................................ 184 Table 18.8 Summary of Financial Analysis (life-of-mine) ................................................................................................................ 185 Table 18.9 - Sensitivity to Gold Price ................................................................................................................................................. 185

vii

List of Figures
Figure 1.1 - Simplified Process Flowsheet ............................................................................................................................................. 6 Figure 1.2 Overall Site Plan................................................................................................................................................................. 8 Figure 4.1 Rosia Montana License Location ..................................................................................................................................... 22 Figure 7.1 Regional Geology of Rosia Montana Area ....................................................................................................................... 35 Figure 7.2 Rosia Montana Project Geology....................................................................................................................................... 37 Figure 7.3 Structural Interpretation of Rosia Montana....................................................................................................................... 40 Figure 7.4 Structural Summary Cetate and Cirnic .......................................................................................................................... 41 Figure 7.5 Section across Cirnic from the WNW to ESE................................................................................................................... 41 Figure 8.1 Electrum Rosia Montana Mine ...................................................................................................................................... 43 Figure 9.1 Vein Stereonets for Cetate and Cirnic.............................................................................................................................. 47 Figure 11.1 Drillhole Collar Locations................................................................................................................................................ 51 Figure 11.2 Under Ground Sample Locations Cetate-Gauri Area (Level +885m RL) .................................................................... 54 Figure 11.3 Under Ground Sample Locations Cetate-Gauri Area (Level +855m RL) .................................................................... 55 Figure 11.4 Under Ground Sample Locations Cetate-Gauri Area (Level +827m RL) .................................................................... 56 Figure 11.5 Under Ground Sample Locations Cetate-Gauri Area (Level +824m RL) .................................................................... 57 Figure 11.6 Under Ground Sample Locations Cetate-Gauri Area (Level +724m RL) .................................................................... 58 Figure 11.7 Under Ground Sample Locations Carnicel Area (Level +884m RL)............................................................................ 59 Figure 12.1 Surface and Underground Channel Sampling Locations ............................................................................................... 61 Figure 12.2 Underground Channel Sampling Process ...................................................................................................................... 62 Figure 12.3 Channel Sample Gold Assay Grades versus Sample Weights...................................................................................... 64 Figure 12.4 RC Sampling Process .................................................................................................................................................... 64 Figure 12.5 RC Sample Weight vs Gold Grade 5 m Composites................................................................................................... 67 Figure 12.6 Core Drilling and Sampling Process............................................................................................................................... 68 Figure 12.7 DDH Core Recovery versus Gold Grade........................................................................................................................ 69 Figure 13.1 Conventional Fire Assay Procedure............................................................................................................................... 71 Figure 13.2 Distribution of Total Sulphur Data .................................................................................................................................. 72 Figure 14.1 Perspective View of the Topographic Model .................................................................................................................. 91 Figure 16.1 - Simplified Process Flowsheet ....................................................................................................................................... 121 Figure 17.1 Estimation Regions with Drilling Data .......................................................................................................................... 127 Figure 17.2 Perspective View of the Geological Model ................................................................................................................... 128 Figure 17.3 Perspective View of the Silicic Alteration Zone Wireframes......................................................................................... 129 Figure 17.4 Gold 5 m Composites Indicator Coding : >=0.1 g/t ................................................................................................... 130 Figure 17.5 Gold 5 m Composites Indicator Coding : >=0.3 g/t ................................................................................................... 131 Figure 17.6 Gold 5 m Composites Indicator Coding : >=0.6 g/t ................................................................................................... 131 Figure 17.7 Gold 5 m Composites Indicator Coding : >=0.8 g/t ................................................................................................... 132 Figure 17.8 3D view of the Interpreted Mineralization Zones .......................................................................................................... 133 Figure 17.9 Channels Sample Along Veins..................................................................................................................................... 137 Figure 17.10 Log Probability Plot of Sample Intervals Drilling...................................................................................................... 137 Figure 17.11 Log Probability Plots of Composite Gold Data ........................................................................................................... 139 Figure 17.12 Log Probability Plots of Composite Silver Data.......................................................................................................... 140 Figure 17.13 Scatter Plot - Gold (g/t) versus Silver (g/t).................................................................................................................. 145 Figure 17.14 Scatter Plot - Gold (g/t) versus Silver (g/t).................................................................................................................. 145 Figure 17.15 Plan View of the Underground Workings ................................................................................................................... 153 Figure 17.16 Typical Cross Section Highlighting Blocks (pink) that Share Data ............................................................................. 157

viii

Figure 17.17 Au OK vs Au 10 m Composites : Easting................................................................................................................... 159 Figure 17.18 Au OK vs Au 10 m Composites : Northing ................................................................................................................. 160 Figure 17.19 Au OK vs Au 10 m Composites : Elevation ................................................................................................................ 160 Figure 18.1 Overall Site Plan........................................................................................................................................................... 170 Figure 18.2 - Project Sensitivities ....................................................................................................................................................... 186 Figure 18.3 - Operating Cost: Commodity Sensitivity......................................................................................................................... 186 Figure 18.4 - Operating Cost: Currency Sensitivity ............................................................................................................................ 186 Figure 18.5 - Cash Cost per Ounce Sensititity: Life-of-Mine .............................................................................................................. 187 Figure 18.6 - Capital to Completion Sensitivity - Foreign Exchange Rates........................................................................................ 187

ix

List of Appendices
Appendix 1 Appendix 2 Appendix 3 Appendix 4 Appendix 5 Appendix 6 Appendix 7 Appendix 8 Appendix 9 Appendix 10 Appendix 11 Appendix 12 RMGC Logging Scheme Due Diligence (2005) Sampling Procedure Sample Weight Analysis Statistical Quality Control Plots Stereo70 Grid Memo Description of Bulk Density Determination Methodology Twin Drill Hole Analysis Plots Channel versus Drilling Memo Listing of Drillholes and Channels used in the study Statistical Plots Gold and Silver Composite Data and Sulphur Sample Data Variogram Plots Stacked Transect Plots

1.0 SUMMARY
The Rosia Montana gold project is located in the Transylvania region of Romania and is held by Rosia Montana Gold Corporation S.A. (RMGC), in which Gabriel Resources Ltd., a Canadian corporation, holds an 80% interest. The Rosia Montana property has been mined for gold intermittently over a period of at least 2,000 years. Exploration conducted by RMGC since 1998, from both surface and underground, has delineated a significant gold deposit, with by-product silver, on the property. Feasibility studies have demonstrated the technical feasibility and economic viability of producing gold and silver from the Rosia Montana property, by conventional open pit mining, followed by conventional cyanide leaching processes, with or without an initial stage of gravity concentration, for the recovery of precious metals. In March, 2004, Roscoe Postle Associates Inc. (RPA) prepared a Technical Report entitled Technical Report on the Rosia Montana Project, Romania (the RPA report). That report described the status of the project as of January 1, 2004, including the results of a feasibility study completed in 2001, which concluded that the production of gold and silver from the property was economically viable, using conventional open pit mining, grinding and cyanide leaching of whole ore. In March, 2006, the authors of this present report issued a Technical Report entitled Technical Report on the Rosia Montana Gold Project, Transylvania, Romania (the 2006 Technical report). That report updated the status of the property to February, 2006 and, in particular, discussed the results of a feasibility study completed at that time (the 2006 feasibility study). The 2006 feasibility study again concluded that the Rosia Montana project was economically viable, using the same process of open pit mining and cyanidation. The present Technical Report describes work performed on the Rosia Montana project since February 2006, and discusses the status of the project at February, 2009. Other than the progressive relocation of the families who reside within the footprint of the project, only limited physical work, including geotechnical drilling and channel sampling, has been undertaken on the project since the 2006 Technical Report. The principal technical work performed has been a comprehensive review and updating of all capital expenditure and operating cost estimates for the project, leading to a re-examination of the design of the open pits and the process flowsheets. This work, which commenced in 2007 and has recently been completed, continues to confirm the technical feasibility and economic viability of the Rosia Montana project. A comprehensive Environmental Impact Assessment was completed early in 2006 and was submitted to the Romanian authorities for review. The review process was suspended by the Romanian government in September, 2007, and remains suspended at the date of this report. The timing at which construction will commence thus remains uncertain. In the interim, RMGC has placed orders totalling approximately $44 million for major equipment items with long lead times, including the primary crusher, the SAG mill and ball mills, and mill drive systems. A number of families also remain to be relocated before construction can commence. For much of 2007 and 2008, progress on the Rosia Montana project was blocked by political factors and a largely gridlocked governing environment. With the election of a new coalition government in Romania, which appears to favour resource development, it is RMGCs opinion that the prospects for 1

advancement of the project are more favourable now than at any point since April, 2007. RMGC expects that the environmental review process will recommence in 2009. The authors of this report have relied on that opinion. 1.1 MINERAL RESOURCES AND MINERAL RESERVES

The Mineral Resource estimate discussed herein is summarized in Table 1.1. The estimate is based on the results of all surface and underground exploration performed by RMGC from January, 1998 to May, 2005. The only additional exploration conducted on the property since May, 2005, has been limited underground channel sampling which has produced no results to suggest that the existing resources block model requires modification.
Table 1.1 Rosia Montana Gold Project Categorized Grade Tonnage of Accepted Ordinary Kriged Estimate 10 mE x 10 mN x 10 mRL, Cut Gold and Silver 5 m composites Gold Cutoff Grade 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 Contained Gold (Moz) 171.51 139.83 113.11 90.70 341.22 210.52 137.65 94.40 512.73 350.35 250.76 185.10 44.81 30.29 22.20 17.53 1.3 1.5 1.7 1.9 0.9 1.2 1.4 1.6 1.0 1.3 1.5 1.8 1.0 1.2 1.4 1.5 8 8 9 10 3 4 4 5 5 6 6 7 3 3 3 3 7.3 6.7 6.1 5.5 9.9 7.8 6.2 5.0 17.1 14.6 12.3 10.5 1.4 1.2 1.0 0.9 Contained Silver (Moz) 43.2 38.1 32.8 27.8 38.0 26.8 19.5 14.6 81.1 64.9 52.3 42.4 4.1 3.0 2.1 1.6

Mt

Au (g/t)

Ag (g/t)

Measured

Indicated

Measured and Indicated

Inferred

The Mineral Resources stated in Table 1.1 include the Mineral Reserves described below. The resource estimate summarized in Table 1.1 has been developed by conventional geological modelling and block modelling, following an extensive program of data verification and statistical analysis of individual and composited sample assay data. Grades have been interpolated into the block model by Ordinary Kriging. The geological model of the mineralization has been developed over a number of years, and is considered to be well established and of high confidence. Variographic studies, however, indicate that the mineralization is characterized by a high nugget effect, suggesting that the mineral resource estimate, while globally robust, is unlikely accurately to have predicted all local variations in grade. Strict grade control procedures will therefore be required during operation. It is unlikely, also, that all of the underground workings resulting from previous mining have yet been accurately located. Further work in this area is recommended prior to the commencement of open pit operations.

The geological block model, together with a series of technical and economic criteria, has been used as the basis for selecting the ultimate limits of four open pits (Cetate, Cirnic, Orlea and Jig) by conventional floating cone techniques. All material within the ultimate pit shells which, on the basis of economic criteria, is scheduled to be processed for the production of gold and silver, is considered a Mineral Reserve. The estimated Proven and Probable Mineral Reserve for the Rosia Montana project is 214.9 million tonnes at average grades of 1.46 grams of gold per tonne and 6.88 grams of silver per tonne, as summarized in Table 1.2.
Table 1.2 RMGC Mineral Reserves - February 26, 2009
Category Cutoff Net/hour ($ x 1000) Proven Probable Total $2.24 to $22.00 $2.24 to $22.00 Ore (kt) 112,455 102,476 214,931 Direct Plant Feed Ore Gold (g/t) 1.630 1.270 1.46 Silver (g/t) 9.01 4.55 6.88 Recovered Grade (g/t Au) 1.290 1.000 1.15 Recovered Grade (g/t Ag) 5.54 2.69 4.18 Contained Gold (Moz) 5.9 4.2 10.1 Contained Silver (Moz) 32.6 15.0 47.6 Recovered Recovered Gold Silver (Moz) (Moz) 4.6 3.3 7.9 20.0 8.9 28.9

The Mineral Resources and Mineral Reserves reported herein have been estimated in accordance with the Canadian Institute of Mining, Metallurgy and Petroleum standards and definitions. 1.2 UPDATED TECHNICAL AND ECONOMIC REVIEW

The updated technical and economic review recently completed for the Rosia Montana project is based on processing of the Mineral Reserves at a rate of between 13 million and 15.5 million tonnes per year, providing a productive life of approximately 16 years. Approximately 29 million tonnes of lower grade material mined in the early years of production will be stockpiled for processing in the final years of the operating life. Total life-of-mine production is estimated at 7.9 million ounces of gold and 29 million ounces of silver. 1.2.1 Methods of Production

Mining from the Cetate, Cirnic, Orlea and Jig open pits will be undertaken using shovels and trucks. The total material contained in the design pits is approximately 472 million tonnes, comprising 215 million tonnes of ore and 257 million tonnes of waste, at an average waste ratio of 1.2 tonnes of waste per tonne of ore. For production scheduling, the Cetate, Cirnic and Orlea pits were divided into a series of sequential mining phases, designed to maximize the present value of the project, while honouring practical operational constraints. The small Jig pit was scheduled as a single phase. The open pits at Rosia Montana will be mined in benches 10 metres high, using the following equipment and procedures: Drilling will be performed with conventional rotary blast hole drills utilizing 25 cm diameter bits. Three such drills will be required throughout most of the mine life. Blasting will be performed with conventional ANFO explosives and ANFO-slurry blends. 3

Blast hole cuttings will be assayed for grade control purposes. geologic data will also be recorded.

Hardness information and

Ore and waste will be loaded by 19 m3 hydraulic shovels, of which three will be required throughout most of the mine life. Ore and waste will be hauled by 146-t capacity rear-dump trucks. Ore will be delivered to the primary crusher. Mine waste will be delivered principally to the Cetate and Cirnic waste storage areas, although approximately 49 million tonnes of waste rock are scheduled to be used in construction of the tailings impoundment while, later in the mine life, a significant quantity of waste will be used to completely backfill the Jig pit and partially backfill the Orlea and Cirnic pits. A maximum fleet of 21 haul trucks will be required in production years of 8 through 13. The major mining equipment will be supported by bulldozers, graders, front-end loaders, maintenance equipment and miscellaneous mobile units.

Mine equipment requirements were estimated in accordance with standard industry procedures, based on the productivity, availability and utilization of each unit. Haul truck requirements were estimated from detailed haulage profiles for each time period, for each material type and for each destination. Truck productivity was calculated using haul time simulations over the measured haul profiles. The selection of the processing flowsheet for the Rosia Montana project has been based on extensive metallurgical testwork, as discussed in Section 16 of this report. The process is entirely conventional and consists of crushing, grinding, gravity concentration, cyanide leaching in a carbon-in-leach (CIL) circuit, and recovery of gold and silver by adsorption on to carbon, electrowinning and smelting. A simplified process flowsheet is shown in Figure 1.1. Tailings slurry from the processing plant will be treated in the cyanide detoxification circuit and pumped for permanent storage in the tailings management facility (TMF). The TMF, which will ultimately cover an area of approximately 360 ha, has sufficient capacity to store all tailings produced during the life of the mine, together with run-off resulting from two Probable Maximum Precipitation events. Reclaim water from the tailings will be re-cycled back to the processing plant for reuse. The area of the TMF impoundment is underlain by a layer of colluvium which, when compacted, provides a liner of sufficient impermeability to protect groundwater from contamination. At present, it is not planned to install any additional geotextile components. Tailings will be impounded behind an engineered dam, which will be constructed in stages to an ultimate height of approximately 180 metres. The dam has been designed to withstand the Maximum Credible Earthquake for this region of Romania. A secondary containment dam will be constructed, downstream of the TMF, to collect any seepage from the TMF. Water retained by the secondary dam will be pumped back to the TMF. Both the tailings dam and the secondary containment dam have been designed to comply with Romanian standards, and with the guidelines of the International Commission on Large Dams and the Canadian Dam Association. The overall design of the TMF is based on

Figure 1.1 Simplified Process Flowsheet (Page 1 of 2)

Figure 1.1 Simplified Process Flowsheet (Page 2 of 2)

considerable field investigations of the geological, geotechnical and hydrogeological characteristics of the site. 1.2.2 Project Facilities

The principal productive facilities at the Rosia Montana project will be the open pits, the major mining equipment and the processing plant, all as described above. Additional project infrastructure and ancillary facilities are described in the following paragraphs. The Rosia Montana project is generally well served by existing infrastructure. Electric power will be supplied from an existing 110 kV transmission line which passes through the project site. It has sufficient capacity to supply the peak project demand of 55 MW. Some relocation of the existing line and interconnection facilities, including a substation, will be required. Fresh water requirements will be met from a pumping station located on the Aries river, near the village of Campeni. Water will be delivered to the project site through a buried pipeline 12.7 km long. Required road works include a new access road to the project site, along the left bank of Rosia Montana creek, and a by-pass road east of the Corna creek. Ancillary facilities to be constructed at the site include a mine workshop for maintenance of all mobile equipment, a tank farm for fuel storage, an explosives storage magazine, fire protection systems, communications systems and office, laboratory and warehousing facilities. Extensive water management facilities will also be provided, principal among which are the Cetate water management dam and a state-of-the-art water treatment plant. Historical mining operations in the Rosia Montana area continue to produce acid rock drainage (ARD) which presently flows, untreated, into local watercourses. Additional ARD is expected to be produced as a result of the proposed Rosia Montana project. The management and treatment of ARD, therefore, will be an important component of the project. Interceptor ditches, seepage collection ponds and secondary dams will be constructed at various locations across the site. These facilities will collect contaminated seepage and runoff, both from previous mining activities and from the new project. Water collected in these secondary facilities will be delivered directly to the ARD plant or to the Cetate dam, where it will be stored for subsequent processing through the water treatment plant, prior to discharge. These water management facilities have been designed not only to minimize the environmental impact of the Rosia Montana project, but also to improve significantly the quality of effluent produced from previous mining activities. Figure 1.2 is a general site plan showing the locations of the open pits and the other facilities planned to be constructed at Rosia Montana.

Figure 1.2 Overall Site Plan

1.2.3

Production Schedule

The design annual schedule of process plant throughput and precious metal production is summarized in Table 1.3.

Table 1.3 Process Plant Production Schedule Year 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Total Throughput (thousand t) 11,383 12,952 13,057 13,480 13,432 13,535 14,179 13,984 14,867 15,382 15,358 14,223 14,117 14,268 13,536 7,178 214,931 Feed Grade (g/t) Gold Silver 1.868 12.70 2.030 12.65 2.113 11.14 1.805 7.82 1.693 8.61 1.599 8.64 1.474 9.84 1.385 5.65 1.243 2.93 1.082 3.38 1.432 2.75 1.383 3.63 1.417 2.50 0.989 5.50 0.892 8.01 0.892 8.01 1.459 6.88 Recovery (%) Gold Silver 83.1 65.6 80.4 61.0 79.3 60.2 77.5 61.0 78.6 61.3 79.2 61.2 76.7 59.6 79.2 60.0 82.9 58.0 81.8 58.9 82.5 54.5 79.3 57.9 74.5 57.6 71.7 61.0 70.3 61.4 70.3 61.4 78.7% 60.8% Recovered Metal (thousand oz) Gold 567.8 679.7 703.6 606.3 574.4 551.4 515.6 493.2 492.8 437.7 583.6 501.6 479.3 325.2 272.9 144.7 7,929.8 Silver 3,048.2 3,212.5 2,816.8 2,067.3 2,280.2 2,302.0 2,671.4 1,524.2 812.6 984.2 740.7 960.3 653.6 1,540.0 2,138.6 1,134.1 28,886.7

1.2.4

Environmental and Socio-Economic Considerations

An EIA for the Rosia Montana project, prepared in accordance with international best practice, was filed with the Romanian government in May, 2006. In May, 2007, in response to the EIA public consultation process which took place during the summer of 2006, RMGC responded to 5,610 questions and 93 statements received by the Romanian government and judged by it to merit a response. As required under Romanian law, the Romanian government then set up the Technical Assessment Committee (TAC), comprised of officials from the various Ministries of the Romanian government involved in the permitting process to review the project, the EIA and RMGCs responses to the questions asked during the public consultation. The TAC held meetings on July 10, July 19 and August 9, 2007 which covered the first four chapters of the EIA representing the bulk of the EIA. The TAC meetings to date have been constructive with a thorough technical analysis of the project. Separately, RMGC participated in intergovernmental meetings between the Romanian and Hungarian governments on July 30 and 31, 2007, as required under the Espoo Convention. On September 12, 2007, RMGC received a letter from the Ministry of Environment and Sustainable Development (MESD) indicating that the review process for the EIA had been suspended. The MESD based its action on a court challenge by an NGO, opposing the project, regarding the validity of an urbanism certificate wholly unrelated to the EIA review process. RMGCs position, supported by local counsel, is that an urbanism certificate is not required for the TAC review process. Administrative complaints filed by RMGC failed to resolve the suspension issue and, as a result, RMGC filed a lawsuit against the MESD, as well as the Minister of the MESD personally and the State Secretary, in November, 2007. The lawsuits are ongoing and RMGC expects the court to rule 9

sometime during 2009. Notwithstanding the ongoing litigation relating to actions taken by a prior Romanian government, RMGC continues to pursue all options to re-start the review of its EIA with representatives of the new government of Romania, which was elected at the end of 2008 and officially took office in January, 2009. RMGC expects that the TAC process will be re-started in 2009 and that the EIA for the project will be given the fair and transparent review it is entitled to under Romanian law. Baseline and impact studies have indicated that the principal environmental considerations for the Rosia Montana project are ARD, the preservation of culturally important and archaeological structures or artifacts, and the effective reclamation and closure of the site at the end of its operating life. As discussed above, extensive measures are planned for the collection, control and treatment of ARD. The principal areas of cultural and archaeological interest are: The existence of buildings of cultural significance within the Rosia Montana valley. The existence of mine working, artifacts, and traces of habitation and infrastructure which date from Roman times.

The majority of the buildings of cultural significance are contained in relatively close proximity to each other, in a location which has been designated as the Protected Area. RMGC has undertaken not to allow mining activities to encroach upon the Protected Area or the buffer zone which surrounds it. Other buildings of cultural significance, outside of the limits of the Protected Area, will be preserved to the extent possible. RMGC has funded a major program of archaeological research in the Rosia Montana area, with expenditures currently in excess of $10 million. Underground openings dating from Roman times have been made accessible and are being mapped by independent teams of French, German and Romanian archaeologists. Artifacts discovered both underground and on surface have been collected in a temporary facility, open to the public, pending the planned construction of a more formal museum. Funerary monuments discovered on surface are being preserved. The effect of these measures to preserve sites of cultural and archaeological significance has been to limit the design open pits to a materially smaller surface area than would have been dictated solely by economic analysis. Closure plans for mining operations are required under Romanian law. developed by RMGC and its consultants, and is included in the EIA. The principal social-economic impacts of the Rosia Montana project will be: The creation of employment in the local area, with multiplier effects at a regional level. The need to relocate families from areas which will be disturbed by the project. A closure plan has been

The Rosia Montana area has a long history of mining, but is currently in a relatively depressed state, with reported levels of unemployment in excess of 70%. Increased employment and economic

10

opportunity in the area remain firmly tied to mining development. In this context, it is considered that the Rosia Montana project will have a beneficial impact. RMGC has developed, and is implementing, a Resettlement and Relocation Action Plan, governing its actions with respect to the relocation of families or individuals. A number of families have already been relocated, and are being monitored by RMGC, which reports that all are satisfied with their new circumstances. RMGC reports that, in general, the necessary relocations are proceeding well, although, in situations such as this, it is not uncommon for a few individuals or families to resist relocation. RMGC has established a community relations office within the Protected Area in the village of Rosia Montana. This office is open to any members of the public who wish to discuss the project, or any aspect of it. The office also contains a well laid out display of Roman artifacts recovered from the area. 1.2.5 Permitting

As is common to most countries, Romania requires that a significant number of permits, licenses, authorizations and endorsements be obtained before the construction and/or operation of a mining project can commence. Several permits have already been granted, but a significant number remain to be obtained. The issuance of several permits, including those permits necessary to commence construction, is contingent upon approval of the EIA. In the absence of any other extraordinary events, legal or otherwise, RMGC anticipates that it would take at least 6 months from the re-start of the permitting process to complete the EIA approval process; complete the purchase of the outstanding properties; receive all other permits and approvals, including initial construction permits; complete the control estimate for the project, and complete the required financing. Ultimately, RMGCs ability to obtain construction permits for the mine and plant is predicated on securing 100 percent of the surface rights in the industrial zone. It is has been well documented in prior public disclosure that the project has been the target of opposition of several NGOs. These NGOs have attempted to use use the courts as a tactic for delaying the permitting process for the project. While the majority of these lawsuits have been frivolous and abusive, some of them have resulted in the loss of certain permits issued to RMGC, most notably Archaeological Discharge Certificate #4. However, in most cases involving key permits, approvals or processes relating to the project, the efficacy of the actions of the relevant Romanian government agencies has been upheld by the courts. RMGC will continue to defend all permits granted to it by the Romanian government. Similarly, RMGC expects that the Romanian government agencies charged with issuing permits and approvals will do so in a manner consistent with all applicable legal provisions. 1.2.6 Cost Structure

The estimated capital expenditures and direct operating costs associated with the development and operation of the Rosia Montana project are summarized in Tables 1.4 and 1.5, respectively. All costs are expressed in constant US dollars of fourth quarter 2008 value. The cost estimates make no allowance for the payment of value-added tax. The unit mining cost of $2.88 per tonne processed, as shown in Table 1.5, is equivalent to $1.31 per tonne of material mined.

11

Table 1.4 Estimated Capital Expenditures Area Initial Capital ($ million) 89 115 99 137 158 203 73 Sustaining Capital ($ million) 89 13 152 33 45 34 366 Total Capital ($ million) 179 128 251 170 158 248 107 1,242

Mining Processing TMF Infrastructure, Utilities EPCM(1) and Indirect Costs Owners Cost Contingency

(1)

Total Capital 876 EPCM: Engineering, procurement and construction management

Table 1.5 Estimated Average Cash Operating Cost Component Mining Processing G & A and off-site Total Unit Cost ($/tonne processed) 2.88 8.23 1.46 12.57

With the exception of surface rights to be acquired after December 31, 2008, the base date for the capital expenditure summarized in Table 1.4 is a notional three months prior to the commencement of construction, a date which is presently indeterminate. This approach has led to the exclusion from the estimate of the following items: All costs for surface rights previously acquired and for construction of the Alba Iulia resettlement site, estimated at $90 million. All payments, including future payments, on the items of long-lead-time equipment already ordered, estimated at $44 million. The cost of engineering activities undertaken prior to March, 2009, estimated at $13 million.

Capital expenditures were estimated in accordance with procedures in common use within the mining industry. In general: Budgetary quotations were obtained for all major equipment units. The cost of minor units of equipment was obtained from data on file, or by factoring. In some cases, budgetary quotations obtained in mid-2007 have been escalated to fourth quarter 2008. Construction labour rates have been based on surveys conducted in Romania by RMGC in 2007, escalated to fourth quarter 2008. Labour productivity has been adjusted downward from North American and Australian standards, to account for local Romanian conditions. 12

Quantities of earthwork, concrete, structural steel and like items have been based on quantity take-offs from general arrangement drawings, single line electrical diagrams, and piping and instrumentation diagrams. Unit prices of bulk materials were based on budgetary quotations obtained from potential Romanian suppliers. The major components of EPCM and indirect costs were estimated from first principles. Minor components were factored. Contingency allowances were allocated on the basis of the adjudged quality of the underlying estimate.

Operating costs have been estimated in accordance with standard industry procedures. Labour costs have been based on the project manning charts, and a survey of salaries and wages conducted by RMGC in 2007, and escalated to fourth quarter 2008. Provision has been made for a small contingent of expatriate managers. The cost of process reagents and consumables has been based on consumption levels determined through metallurgical testwork, coupled with budgetary price quotations obtained from Romanian and international suppliers. Mine consumables have been based on manufacturers recommendations, in-house data and representative unit prices. Maintenance costs have either been factored as a percentage of capital cost, or have been based on manufacturers recommendations and in-house data. The cost of diesel fuel, delivered to site, has been estimated by RMGC at $0.89/L. The cost of electric power has been estimated at $0.083/kWh, based on an analysis conducted by a consultant. The cost of sodium cyanide has been estimated at 1,600 Euros per tonne. The average cash cost of producing gold, after deducting silver credits but including royalty and other taxes, is estimated to average $272 per ounce over the first five years of production and $335 per ounce over the life of the mine. 1.2.7 Project Economics

The overall economics of the Rosia Montana project have been evaluated by conventional discounted cash flow techniques, based on the production schedules, capital expenditures and operating costs discussed in this report, together with the following additional parameters: The metal prices used for the base case economic analysis are $750 per ounce for gold and $10.50 per ounce for silver. The analysis is based on 100% equity financing, with no debt component. A tax rate of 16% has been applied, based on information provided by RMGC. Provision has also been made for the net smelter return royalty of 4% payable to the government of Romania. A provision of $128 million has been included for final reclamation and closure.

Under the estimates and assumptions used for the base case analysis, the Rosia Montana project would be expected to generate an undiscounted, life-of-mine cash flow, after tax, of $1.66 billion, a net present value of $997 million at a discount rate of 5% per year, and an after tax internal rate of return of 20.4% per year. The payback period is estimated at 3.5 years. 13

Table 1.6 presents a summary of the life-of-mine results of the base case financial evaluation of the Rosia Montana project, and a comparison of these results with the estimates contained in the 2006 Technical Report. It is evident that, in common with mining projects throughout the world, the Rosia Montana project has experienced significant escalation in both capital expenditures and operating costs over the last three years. On the other hand, the price of gold has also increased, with the result that the overall economics of the project today are more favourable then they were in early 2006.
Table 1.6 Summary of Financial Analysis (life-of-mine) Item Gold produced Silver produced Average gold production Average silver production Cash cost Pre-production capital Sustaining capital Closure cost Undiscounted cash flow, after tax NPV after tax, 5% discount IRR after tax Payback Units Moz Moz koz/y koz/y $/oz $M $M $M $M $M %/y y This Report 7.93 28.89 511 1,860 335 876 366 128 1,662 997 20.4 3.5 2006 Report 7.94 28.89 509 1,852 237 638 208 70 1,022 498 17.6 3.8

The economics of the Rosia Montana project are most sensitive to variations in gold price and mined grade. They are less sensitive to variations in capital expenditure or operating cost. Table 1.7 summarizes the sensitivity to variations in the base case gold price of $750/oz.
Table 1.7 Sensitivity to Gold Price Gold price ($/oz) Undiscounted NPV ($ million) Internal rate of return (%/yr) Payback period (years) 600 688 10.9 5.2 750 1,662 20.4 3.5 900 2,621 28.0 2.7

1.3

CONCLUSIONS AND RECOMMENDATIONS

The exploration activities undertaken by RMGC since 1998 have delineated a significant gold deposit, with by-product silver, on the Rosia Montana property. Updated estimates of capital expenditure and operating costs, recently completed, have confirmed the technical feasibility and economic viability of producing an estimated 7.9 million ounces of gold and 29 million ounces of silver from the property, over an operating life of approximately 16 years, from an estimated Proven and Probable Mineral Reserve of 215 million tonnes at an average grade of 1.46 g/t Au and 6.9 g/t Ag. This reserve is contained in four open pits which will be mined conventionally by shovels and trucks. The process plant feed will be ground to 80% minus 150 m and gold and silver will be recovered as dor bars by conventional gravity concentration, CIL processing, electrowinning and smelting techniques. 14

A comprehensive Environmental Impact Assessment was completed early in 2006 and was submitted to the Romanian authorities for review. The review process was suspended by the Romanian government in September, 2007, and remains suspended at the date of this report. The timing at which construction will commence thus remains uncertain. In the interim, RMGC has placed orders totalling approximately $44 million for major equipment items with long lead times, including the primary crusher, the SAG mill and ball mills, and mill drive systems. A number of families also remain to be relocated before construction can commence. For much of 2007 and 2008, progress on the Rosia Montana project was blocked by political factors and a largely gridlocked governing environment. With the election of a new coalition government in Romania, which appears to favour resource development, it is RMGCs opinion that the prospects for advancement of the project are more favourable now than at any point since April, 2007. RMGC is hopeful that the project will be evaluated in timely fashion, on its merits and in a fair and transparent manner. The geological model of the mineralization has been developed over a number of years, and is considered to be well established and of high confidence. Variographic studies, however, indicate that the mineralization is characterized by a high nugget effect, suggesting that the mineral resource estimate, while globally robust, is unlikely accurately to have predicted all local variations in grade. Strict grade control procedures will therefore be required during operation. It is unlikely, also, that all of the underground workings resulting from previous mining have yet been accurately located. Further work in this area is recommended prior to the commencement of open pit operations. Since, with the exception of continuous elution, only conventional, well-proven processes will be used at Rosia Montana, the overall level of technical risk is considered to lie within the range of normal mining risk. On the basis of the discussion contained within the body of this report, it is concluded that the Rosia Montana project is both technically feasible and economically viable, and that the main challenge to be overcome before the project can be brought to fruition lies in the area of permitting. While RMGC is considered to have appropriate plans and strategies in place to deal with this challenge, the outcome of the permitting process is not fully within its control. Accordingly, it is recommended that RMGC maintain its focus on the entire permitting process. It is recommended, also, that the resettlement and relocation process be advanced to the extent consistent with maintaining the support of the local community for the project.

15

2.0 INTRODUCTION AND TERMS OF REFERENCE

Coffey Mining Pty. Ltd. (Coffey Mining) (formerly RSG Global Pty. Ltd.), Independent Mining Consultants, Inc. (IMC), Metifex Pty. Ltd. (Metifex), MWH Americas, Inc. (MWH), Aurifex Pty. Ltd. (Aurifex) and Micon International Limited (Micon) have been retained by Rosia Montana Gold Corporation S.A. (RMGC) to prepare a Technical Report on the status of the Rosia Montana gold project. The Rosia Montana property is located in the Transylvania region of Romania and is held by RMGC, in which Gabriel Resources Ltd. (Gabriel), a Canadian corporation, owns an 80% interest. Exploration from both surface and underground has delineated a significant gold deposit on the property. In March, 2004, Roscoe Postle Associates Inc. (RPA) prepared a Technical Report entitled Technical Report on the Rosia Montana Project, Romania (the RPA report). That report described the status of the project as of January 1, 2004, including the results of a feasibility study completed in 2001, which concluded that the production of gold and silver from the property was economically viable, using conventional open pit mining, grinding and cyanide leaching of whole ore. In March, 2006, the authors of this present report issued a Technical Report entitled Technical Report on the Rosia Montana Gold Project, Transylvania, Romania (the 2006 Technical report). That report updated the status of the property to February, 2006 and, in particular, discussed the results of a feasibility study completed at that time (the 2006 feasibility study). The 2006 feasibility study again concluded that the Rosia Montana project was economically viable, using the same process of open pit mining and cyanidation. The present Technical Report describes work performed on the Rosia Montana project since February 2006, and discusses the status of the project at February, 2009. Other than the progressive relocation of the families who reside within the footprint of the project, only limited physical work, including geotechnical drilling and channel sampling, has been undertaken on the project since the 2006 Technical Report. The principal technical work performed has been a comprehensive review and updating of all capital expenditure and operating cost estimates for the project, leading to a re-examination of the design of the open pits and the process flowsheets. This work commenced in 2007 and has recently been completed. During the course of these studies, RMGC also revised its approach to implementation of the project. Initially, it had been contemplated that the engineering, procurement and construction management (EPCM) would be conducted under a single blanket contract. Since earthworks, including construction of the tailings impoundment, constitute a large portion of the pre-production construction program, it has now been decided that, once approval to proceed is received from the Romanian authorities, the EPCM function would be split into two contracts: one solely for earthworks and the other for all remaining project facilities. The work of the two contractors would be coordinated by an expanded Owners team. In the opinion of the Qualified Person responsible for the estimation of resources on the Rosia Montana property, the channel sampling undertaken since the 2006 Technical Report has produced no results which suggest that the existing resource block model requires modification. Thus, the resource estimate upon which this report is based remains the same as that discussed in the 2006 Technical Report. Accordingly, much of this report is reproduced or summarized from the 2006 Technical Report, with only 16

those changes necessary to bring the information up to date. The new information contained in this report with respect to updated estimates of capital expenditures, operating costs and overall project economics is discussed in Section 18, Other Relevant Data and Information. Shortly after publication of the 2006 Technical Report an Environmental Impact Assessment (EIA) was completed and submitted to the Romanian authorities for review. The review process was then suspended by the Romanian goverment in September, 2007, and remains suspended at the date of this report. The timing at which project construction can commence is therefore uncertain. The Qualified Persons responsible for this Technical Report are listed below. All have visited the Rosia Montana property, although no visit has been made by any of the authors since 2006, as no physical activity which would influence the findings of this report has been performed on the property during the past three years. Brett L. Gossage, MAusIMM, of Coffey Mining, is responsible for the mineral resource estimate discussed in the 2006 Technical Report and reproduced in this report. John M. Marek, P.E., of IMC, is responsible for the design of the open pits, the scheduling of production from within those open pits, the selection of the mine equipment fleet, and the estimation of mine capital expenditures and operating costs. Stuart J. Smith, B.App. Sc., MAusIMM, of Aurifex, is responsible for the discussion of metallurgical testwork, the review and modification of the process flowsheet, and the estimation of process operating costs. Patrick G. Corser, P.E., of MWH, is responsible for the design of all earthworks and certain infrastructure, including the embankment for the tailings management facility (TMF), all water impoundments and all roads, and for the estimation of capital expenditures for these components of the project. Christopher R. Lattanzi, P.Eng., of Micon, has reviewed and taken responsibility for the capital expenditure estimates prepared by Metifex for all project facilities other than mining facilities and earthworks; for the estimates prepared by RMGC of the Owners component of project capital expenditures and of the general and administration component of operating costs, and for the financial analysis of the project prepared by RMGC.

Unless specifically noted to the contrary, all currency amounts in this report are expressed in United States Dollars (US$) or Romanian new lei (RON). All cost estimates are expressed in US$ of fourth quarter 2008 value. Quantities are generally stated in SI units, including metric tons (tonnes, t), kilograms (kg) or grams (g) for weight; kilometres (km), metres (m), centimetres (cm) and millimetres (mm) for distance; square kilometres (km2) or hectares (ha) for area; and grams per tonne (g/t) for gold and silver grades (g/t Au, g/t Ag). Precious metal grades may also be expressed in parts per million (ppm) or in parts per billion (ppb), and quantities may be reported in troy ounces (oz). The estimates of capital expenditure and operating cost discussed in this report were initially prepared in terms of the currency in which the expenditures are expected to be incurred. These are principally RON, Euros and US dollars, with lesser amounts in Australian and Canadian dollars. All cost estimates 17

were then converted to US dollars at the rates of exchange shown in Table 2.1. The exchange rates used are based on 90-day trailing averages as of January 27, 2009.
Table 2.1 Currency Exchange Rates Currency US dollars Romanian lei Euros Australian dollars Canadian dollars Exchange Rate (Units per US$) 1.00 3.00 0.76 1.50 1.23

Throughout this report, references are made to Appendices. These Appendices have not been filed with the electronic version of the report, because of their bulk. Printed copies of the Appendices are available for inspection at the offices of Gabriel Resources Ltd., Suite 1501, 110 Yonge Street, Toronto, Ontario, Canada, M5C 1T4. Table 2.2 is a list of the abbreviations used in this report.
Table 2.2 List of Abbreviations Term acid rock drainage Ammtec Limited antimony approximately argon arsenic atomic absorbtion spectrophotometry Aurifex Pty. Ltd. Ausenco International Limited Bechtel Australia Pty. Ltd. calcium oxide Canadian Institute of Mining, Metallurgy and Petroleum carbon carbon dioxide carbon-in-leach carbon-in-pulp Coffey Mining Pty. Ltd. centimetres coefficient of variation concentration of hydrogen ion copper cubic metres cubic metres per hour degrees degrees celsius diamond drill hole dollars east engineering, procurement and construction management Environmental Financial Guarantee Environmental Impact Assessment expected value European Union Gabriel Resources Ltd. Abbreviation ARD Ammtec Sb Ar As AAS Aurifex Ausenco Bechtel CaO CIM C CO2 CIL CIP Coffey Mining cm CV pH Cu m m/h C DDH $ E EPCM EFG EIA EV EU Gabriel

18

Geostats Pty. Ltd. gold g grams per tonne grams of gold per tonne grams of silver per tonne GRD Minproc Ltd. greater than half absolute relative differences half relative difference hectare hour Independent Mining Consultants inductively coupled plasma mass spectrometry internal rate of return International Cyanide Management Code iron Joint Ore Reserve Committee joint venture J.R. Goode and Associates kilogram(s) kilograms per tonne kilometre(s) kilowatt(s) kilowatt hours kilowatt hours per tonne Knight Piesold Limited less than litre megapascalls Memorandum of Understanding mercury Metifex Pty. Ltd metre(s) metres above sea level metres per hour Micon International Limited micron(s) milligram(s) milligrams per litre millimetres million million ounces million tonnes million tonnes per year million years Ministry of Environmental Sustainable Development Minvest S.A. MWH Americas Inc. National Agency for Mineral Resources National Instrument 43-101 nearest neighbour net present value Newmont Limited Non-Government Organization ordinary kriging Organization for Economic Co-operation and Development parts per billion parts per million percent percent by weight Pincock Allen and Holt potassium quality assurance / quality control Quantile Quantile quartz-illite-pyrite reduced level Regia Autonoma a Cuprului Deva

Geostats Au g g/t g/t Au g/t Ag Minproc > HARD HRD ha h IMC ICP-MS IRR ICMC Fe JORC JV J.R. Goode kg kg/t km kW kWh kWh/t Knight Piesold < L MPa MOU Hg Metifex m masl m/h Micon m mg mg/l mm M Moz Mt Mt/y Ma MESD Minvest MWH NAMR NI 43-101 NN NPV Newmont NGO OK OECD ppb ppm % wt% PAH K QAQC Q-Q QIP RL Regia Deva

19

relative standard deviation Resettlement and Relocation Action Plan reverse circulation rock quality designation Romanian lei Romanian lei per square kilometre Roscoe Postle Associates Inc. Rosia Montana Gold Corporation RSG Global Pty. Ltd. S.C. Minexfor S.A. semi-autogenous grinding silica silicic alteration silver size at which 80% of particles are smaller SNC-Lavalin Engineers and Constructors sodium chloride sodium cyanide square kilometre(s) square metre(s) sulphur sulphur dioxide Systme International d Units tailings management facility Technical Assessment Committee thousand three-dimensional tonne(s) tonnes per cubic metre tonnes per day tonnes per hour tonnes per square metre per hour total cyanide troy ounce(s) underground uniform conditioning United States United States dollar(s) Volt(s) Washington Group International Inc. weak acid dissoluble cyanide year

RSD RRAP RC RQD RON RON/km RPA RMGC RSG Minexfor SAG SiO2 SIK Ag P80 SNC NaCI NaCN km m S SO2 SI TMF TAC k 3D t t/m t/d t/h t/m/h CNTOT oz UG UC US US$ V WGI CNwad y

20

3.0 RELIANCE ON OTHER EXPERTS

The authors of this report are not qualified to comment on any legal issues associated with the Rosia Montana project, including title and tenure to the property, and the length of time which may be required to obtain the permits necessary to commence construction. The authors have relied on the representations and judgments of RMGC in regard to these matters. The sections of this report dealing with environmental and socio-economic considerations (Section 18.4) and permitting (Section 18.5) have been prepared by RMGC and have been relied upon by the authors. The capital expenditure estimates for all project facilities, other than mining facilities and earthworks, have been prepared by Metifex of Brisbane, Australia, under the direction of Seit Meka. Mr. Meka is a member of the Institute of Engineers of Australia, but is not a Qualified Person as defined by Canadian National Instrument 43-101 (NI 43-101). Christopher R. Lattanzi, P.Eng., has reviewed the work performed by Metifex and is the Qualified Person responsible for this work. The capital expenditure estimates developed by Metifex are discussed in Section 18.6. RMGC has estimated the Owners component of project capital expenditure discussed in Section 18.6, and the general and administration component of operating cost discussed in Section 18.7. RMGC has also prepared the financial analysis of the project discussed in Section 18.8. Christopher Lattanzi has reviewed the estimates and analyses prepared by RMGC and is the independent Qualified Person responsible for this work.

21

4.0 PROPERTY DESCRIPTION AND LOCATION

The Rosia Montana project is located within The Golden Quadrilateral which is situated within the Apuseni and Metaliferi Mountains of Transylvania, Romania and covers an area of approximately 900 km2 immediately to the north of the city of Deva. The Golden Quadrilateral is an historic mining centre, which has been mined intermittently since Geto-Dacian (pre-Roman) times over a period of at least 2,000 years. The district reached maximum development and peak production during the period of the Austro-Hungarian Empire (end of the 17thcentury to 1918). Figure 4.1 shows the location of the project.
Figure 4.1 Rosia Montana License Location

The Rosia Montana project is situated in west-central Romania near the village of Rosia Montana in Alba County within the Rosia Montana mining district. It is located immediately northeast of the town of Abrud, approximately 45 km northwest of the regional capital of Alba Iulia, and 60 km north-northeast of the city of Deva. The village of Rosia Montana and the nearby town of Abrud are the two main centres housing staff and associated infrastructure for the project. The proposed mine and mill site are located at the head of a small drainage basin within steep hilly/mountainous terrain at an elevation of approximately 850 metres above sea level (masl). The proposed tailings management area is located in the immediately adjacent valley to the mine/mill complex. The valley elevation in the area below the site is at approximately 600 masl. 22

As a result of historical mining activities, abandoned waste dumps and tailings ponds exist on the Rosia Montana property. In addition, approximately 140 km of historical underground workings, some dating from Roman times, have been identified within the mineralized zones. Acid mine drainage continues to be produced from the historical openings and dumps. This drainage currently discharges, untreated, into local streams. RMGC proposes to treat these effluents, as part of its normal operating procedures. 4.1 LAND TENURE

RMGC, a joint stock company incorporated in Romania, holds the Rosia Montana project. Eighty percent (80%) of RMGC is owned by Gabriel, a Canadian-based, publicly traded company. The balance is owned by Minvest S.A. (Minvest), a Romanian government-owned organization (19.3% interest), and three other Romanian companies (0.7%, collectively). The 20% interest owned by Minvest and others is a carried interest. Gabriel is required to fund all expenditures, up to the point at which procuction commences. The following description of the land tenure for the properties has been provided by RMGC. On June 16, 1998, Romania enacted a mining law providing for, among other things, the granting of exploration and exploitation concessions to both Romanian and foreign entities. Minvest, as the titleholder to the Rosia Montana and other properties, made application under the new mining law to the Romanian government for an exploitation concession for Rosia Montana, which was approved. The formal exploitation concession for the Rosia Montana project was granted to Minvest in June, 1999. The terms and conditions of the concessions provided for the transfer of the property from Minvest to RMGC. This limits RMGCs involvement in the planned closure of the current mining operations run by the State, and leaves related liabilities to State bodies. The liabilities would be items such as environmental issues and redundancy packages. Romanian mining law provides that all mineral resources are administered by the National Agency for Mineral Resources (NAMR). The following is a description of some of the more important aspects of the mining law. 4.2 MINERAL OWNERSHIP

All mineral resources located in Romania and in the portion of the continental shelf of the Black Sea adjoining Romania, belong to the State of Romania. Mineral rights in Romania are acquired by way of prospecting permit, exploration concession or exploitation concession granted by NAMR. Under the mining law, an exploration or exploitation concession is a property-related right, distinct and independent from the ownership of the land on which it is located, even when both belong to the same person. The rights granted by an exploration or exploitation concession are exclusive to the holder, chargeable, defensible against third parties and are transferable with the consent of NAMR. 4.3 EXPLORATION CONCESSIONS

RMGC holds the Bucium property, adjacent to Rosia Montana, under an exploration concession license. An exploration concession may be obtained for a maximum period of five years, with a renewal right of three years. As originally drafted, the regulations called for a 50% reduction in the concession after two 23

years, and a further 50% reduction after four years. The Bucium property was reduced in size by 50% in May, 2002. This provision was rescinded in the revisions to the mining law that took place as of March, 2003, and further reductions are no longer required. An initial annual fee of 50 RON/km2 was payable on the Bucium concession to the government of Romania. The annual fee doubles two years after the issue of the concession and quintuples after five years and may be adjusted for inflation. The holder of an exploration concession must provide NAMR with annual reports of all exploration activities conducted on an exploration concession. Exploration concessions confer on the holder the exclusive right to explore for all mineral substances lying within the perimeter of the concession. The Bucium property presently comprises an exploration concession (license number 218/1999) measuring approximately 32.123 km2 in area. The 1999 to 2004 expenditure commitment under the terms of the license agreement was US$5,000,000. For 2005 to 2007, it was US$3,300,000. The annual license fee for the exploration conesssion is presently 1,000 RON (US$415) per km2. An environmental bond for the project area is also paid annually. 4.4 EXPLOITATION CONCESSIONS

An exploration concession may be converted into an exploitation concession at any time upon the submission to, and approval by, NAMR of a feasibility study on the exploration concession. An exploitation concession is granted for an initial term of 20 years and is renewable for successive fiveyear periods. An initial annual fee of 2,500 RON/km2 was payable to the government of Romania. The fee may be adjusted for inflation. Holders of exploitation concessions must pay to the government of Romania a net smelter royalty on all production. Exploitation concessions confer on the holder the right to explore, exploit, process, refine and trade all mineral substances (except oil, gas and radioactive substances) lying within the perimeter of the concession, as well as the right to use the surface of the land and available water. The Rosia Montana property is held under exploitation concession license number 47/1999 and covers an area of approximately 23.883 km2. The concession was published in June, 1999, and has a 20 year term, with provision for successive five-year extensions. Initial redevelopment license expenditure commitments, which have been fulfilled, totalled US$4,294,700 and the current annual license fees are 25,000 RON (US$10,415) per km2. An annual environmental bond based on the work program is also paid.

24

5.0 ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY

The following description of accessibility, climate, local resources, infrastructure and physiography is reproduced from the RPA report, with minor modifications. 5.1 ACCESS

The Rosia Montana property lies within an historic mining district, near the village of Rosia Montana in Alba County, approximately 45 km northwest of the regional capital of Alba Iulia, and 50 km northnortheast of the city of Deva, in the Transylvania area of west-central Romania. The property is readily accessible by a fully developed network of roads. 5.2 CLIMATE

The climate of the area is designated as a continental temperate climate, characterized by hot summers, cold winters, significant snowfalls, and annual rainfalls varying from 600 mm to 880 mm. 5.3 INFRASTRUCTURE

The proposed project site is well serviced by existing infrastructure including electrical power, roads and telecommunications. Access to the plant site will be from the existing Gura Rosiei to Rosia Montana road north of the proposed plant site. There are several potential sources for construction materials within the project site namely: Sulei quarry. Limestone quarry. Cetate pit pre-strip. La Piriul Porcului quarry.

An existing twin circuit 110kV power line owned and operated by the local distribution company Electrica S.A. traverses the project site. This power line connects the existing Zlatna and Preparare (Rosia Poeni) substations. Fresh water for the project will be provided from a pumping station on the Aries river and will be delivered to the site by a buried pipeline, 12.7 km long. 5.4 PHYSIOGRAPHY

The project area is in the central part of the Transylvania region of Romania. The Carpathian Mountains to the east and the Transylvanian Alps to the south are the dominant geographical features of central Romania. The project lies in an area of the Transylvanian Plateau that is bordered by the two mountain ranges. The project area is characterized by a partly forested, hilly landscape with elevations of 500 masl to 1,000 masl. Valleys in the area are cut to depths ranging from 100 m to 200 m. Rural and urban areas comprise approximately 70% of the land use.

25

6.0 HISTORY
6.1 EXPLORATION AND MINING The Rosia Montana gold-silver deposit has been mined from pre-Roman times and artefacts, headstones and delivery contracts have been recovered both in and near workings dating from the Roman invasion in 105-106 AD. Four principal periods of mining activity have occurred at Rosia Montana: Pre- and post-Roman era. Austro - Hungarian Empire era (end of 17th century to 1918). Inter-war period (1918 to 1939). Modern era (1959 to present).

During Dacian and Roman times, high-grade quartz veins and breccias (e.g. Cetate and Cirnic) were mined initially from the surface and possibly from underground. Evidence of Roman era mining remains clearly visible and articles of mining equipment have been recovered from abandoned underground workings in recent times. Most of the historic underground development and peak gold production occurred during the period of the Austro-Hungarian Empire. Tens of kilometres of underground openings, mainly adits, were developed during the 1700s through to 1918, with the majority of mining carried out from underground workings over 400 m of vertical elevation. Large open stopes were developed in both the Cetate and Cirnic massifs, many of which were intersected during recent underground exploration. Underground mining under the control of the Romanian government began in the early 1960s. Open pit mining also commenced on the Cetate deposit in 1970, being conducted simultaneously until the cessation of underground activities in 1985. Underground mining during this period was carried out by strike development along individual quartz veins (predominantly at Orlea, Tarina, Cirnicel, and within Cetate and Cirnic) and room and pillar stoping within breccias and dacite at Cetate and Cirnic. In 1970, open pit mining commenced at Cetate, extracting ore from new mining areas of the resource, but also recovering remnant pillars from the previous room and pillar mining areas. The open pit was extended to the southwest to access ore hosted within the dacite at Cetate Affinis. Mining in the open pit has removed the upper 120 m of the Cetate Massif from the original surface at 1,008 masl to the current pit floor at 886 masl. Open pit mining at Cetate ceased in 2006. Open pit mining was conducted on the eastern side of Cirnic from 2000 until early 2005 with minor resources being depleted at an annual rate of approximately 21,000 tonnes per year. In addition some material from historical waste dumps was also mined for processing during this period. Surveyed pit pick-ups between 2000 and 2004 confirmed total annual production to be approximately 151,000 tonnes per year with approximately 130,000 tonnes per year being mined from Cetate and 21,000 tonnes per year mined from Cirnic over this period. From the block model it is estimated that the grade of the material mined was approximately 2.0 g/t Au, equating to roughly 9,700 ounces of contained gold being mined per year.

26

Since the 1970s, exploration at Rosia Montana continued under the control of the Romanian state companies S.C. Minexfor S.A. (Minexfor) and Regia Autonoma a Cuprului Deva or Minvest (Regia Deva), and consisted of the following key methods: Underground development (adits, drives and crosscuts). Underground diamond and short hole rotary drilling. Channel sampling of existing and new underground development. Surface rock chip sampling. Limited surface diamond drilling.

Samples collected during this period were routinely annotated onto plans and sections and gold and silver assays recorded by hand in assay ledgers. In 1984, a feasibility study on the project was compiled by Regia Deva, based on information acquired from exploration carried out up to 1984 that essentially comprised compilation of the available data into a series of maps, plans, sections and tables. The study was updated in 1992, although little further exploration had occurred since 1984, due to budgetary constraints. In 1992, Minexfor completed an 18-hole diamond drilling program at VaidoaiaJig (Lespedar) to confirm and extend previous exploration carried out at Vaidoaia. In January, 1998, RMGC initiated a program of surface reverse circulation (RC) and diamond drilling, along with underground channel sampling of all accessible underground drives and crosscuts. Surface channel sampling was undertaken to extend the known surface geochemical and assay database. During 2000, underground diamond drilling was undertaken from the 714 level, which marks the lowest accessible level in the Cetate and Cirnic underground development. Exploration data gathered from January, 1998 to May, 2005 provides the basis for the resource estimate discussed herein. Subsequent to May, 2005, additional geotechnical drilling has been undertaken and underground channel sampling has also been completed, as summarized below and in Table 6.1. These data are not material to the global mineral resource and therefore no additional resource estimation studies have been completed since 2006. Exploration undertaken by RMGC in 1998, and managed by RSG Global, included the following: Surface mapping and channel sampling. Underground mapping and channel sampling. Surface RC and diamond drilling. Underground diamond drilling. Airborne geophysical surveys. Surface geophysical surveys.

27

During 2002, further drill-based exploration was carried out at Rosia Montana in order to undertake the following activities: Infill drilling to a notional 40 m by 40 m drill hole collar spacing at Cirnic and Cetate. The infill drilling was targeted within a 7-year pit defined by pit optimization studies completed by SNCLavalin, Engineers and Constructors (SNC). Close-spaced drilling, emulating likely grade control drill spacing (15 m by 12 m), within the Cetate pit area (adjacent to Gauri). Further sterilization drilling. Geotechnical drilling.

During 2003, additional drill-based exploration was undertaken including the following: Step out drilling at the eastern and western ends of Orlea. Further drilling along the northwest trending structure at Igre. Further sterilization drilling. Geotechnical drilling.

The additional drilling and assay information was used to provide data for a resource update. During 2004, additional drill-based exploration was undertaken including the following: 15 sets of twin drill holes were completed along the western boundary of the Cirnic domains, and in the central regions of the Cetate domains. A further 7 drill holes were completed twining holes previously completed by RMGC. An additional 23 diamond and RC drill holes were completed testing regions of Cirnic, Cetate and Carpeni. 9 drill holes were completed by RMGC in 2004 along the northeastern boundary of the Cirnic mineralization. Surface and underground channel samples were collected in Cirnic and Cetate.

During 2005, additional channel sampling was undertaken including the following: Underground channel samples were collected in Cirnic and Cetate on newly accessed drives. Updated geological and alteration mapping. Updated underground surveying of voids. Updated geological, alteration and void model was completed.

During 2006, additional channel sampling and geological mapping were undertaken at Cirnicel including the following:

28

Underground channel samples (156 samples) were collected in Cirnicel on re-accessed drives (884 level). Updated geological and alteration mapping. Updated underground surveying of voids. Geotechnical drilling was completed and logged, although not assayed for Au, Ag or S.

During 2007, additional channel sampling and geological mapping were undertaken including the following: Underground channel samples (570 samples) were collected at Cetate and Gauri on three reaccessed drives. Updated geological and alteration mapping was completed underground at Cetate and Gauri. Surface geological and alteration mapping was completed at Cirnic. Updated underground surveying of voids. The 2006 geotechnical drilling programs continued. Detailed topographic surveys (0.5 m contour intervals) were completed for the processing and infrastructure site areas.

During 2008, additional channel sampling and geological mapping were undertaken throughout the project on the 855 and 714 levels. In total 1,233 underground channel samples have been collected although the assay data set is incomplete at the time of compilation of this report. A detailed summary of work completed to the end of 2008 is presented in Table 6.1. 6.2 TECHNICAL AND ECONOMIC STUDIES

In 1999, Pincock, Allen and Holt (PAH) completed a prefeasibility study for the Rosia Montana project. In 2000, RMGC retained GRD Minproc Limited (Minproc) to prepare a feasibility study. A number of other consultants participated in the study, including Knight Piesold Limited (Knight Piesold) which was responsible for the design of the TMF. The study, which was completed in 2001, addressed the feasibility of the project at production rates of both 8 and 20 million tonnes of ore per year. Subsequently, SNC completed a study which indicated that 13 million tonnes of ore per year was the optimum rate of production. In January, 2002, RMGC retained SNC to carry out basic engineering for the project at this rate of production. SNCs scope included a review of the earlier Minproc feasibility study. SNC assumed responsibility for the basic engineering design of the TMF and water management dams and systems. The basic engineering studies were completed in January, 2003.

29

Table 6.1 Rosia Montana Gold Project History Year 1996 Completed Work MOU signed Rosia Montana tailings evaluation JV signed Data compilation Independent geologists review Pit mapping/sampling Petrology studies by Terry Leach and Associates. First drill program3,754 m( Cetate only ) Extensive channel sampling commenced6,121 samples collected (Cetate only) First resource estimate (Cetate only) by RSG. Preliminary metallurgical sampling Scoping study completed August by RSG Underground channel sampling expanded to all of Cetate, Cirnic, and Cirnicel Cetate; 7,708 channel samples, 10,910 m drilling Cirnic/Cirnicel; 23,353 channel samples, 1,130 m drilling Additional petrology by Terry Leach & Associates Additional metallurgical sampling (Dawson Laboratories) Geotechnical study by Golders Resource estimate by RSG Pre-feasibility by PAH (Reserve; 85 Mt @ 1.70 g/t Au, 11 g/t Ag) Channel sampling continued during/after pre-feasibility (Not included.15,624 m) Delineation drilling; 38,750 m of surface RC/DC drilling, 11,821 m of underground diamond drilling Channel sampling (Carpeni, Orlea, Jig); 6,819m Geological mapping and interpretation done by Greg Keeley of RSG Feasibility study commenced by Minproc; inc. EIA, RAP, met work, engineering Interim resource; October,2000 by RSG Aeromagnetics flown (Fugro). Interpretation by Van de Merve of RSG Aerial photography flown by Spectrum Surveying and Mapping. Resource update; February,2001 by RSG. Geotechnical drilling Topographic maps completed by Spectrum Survey and Mapping Feasibility study (20 Mtpa completed in August 2001 by Minproc. Followed by 8 Mtpa study. Optimization study completed; SNC-Lavalin- 13 Mtpa Sterilization drilling; 15,153 m Delineation (7 year pit); 16,947 m Grade control/selectivity drilling and report by RSG; 4,098 m Twin drilling and report by RSG; 1,302 m Resource update December, 2002 (RSG) Resource estimate audit completed (IMC) Reserve update/mine plan commenced (IMC) Geotechnical drilling Basic engineering study commenced (SNC-Lavalin) AMT geophysical test survey Exploration drilling (surface drilling in Tarina, Igre, Orlea): 61 holes completed, 10,949.05 m Geotechnical drilling (Corna, V. Porcului, Balmosesti, Sulei, Tapului): 87 holes completed, 3,694.6 m Surface channel sampling (Tarina, Orlea, Gauri): 437 m (29 channels) Underground channel sampling (Gauri): 167 m, 8 channels) Reserve update February, 2003 (IMC) Resource estimate update November, 2003 (RSG) Resource, reserve and mine plan (feasibility study) completed by Ipromin to Romanian law. Exploration drilling (surface drilling in Cirnic): 10 holes completed, 2,514.35 m Infill and twin drilling program (surface drilling in Cetate, Cirnic): 59 holes completed, 10,908.05 m Surface channel sampling (Cirnic): 341 m (13 channels) Underground channel sampling (Cirnic): 564 m (67 channels) Metallurgical sampling (Cirnic, Cetate, Orlea, Jig): 98 samples Colorado School of Mines Post-Doctorate alteration and mineralization mapping study (Dr. Scott Manske) Re-logging of underground workings (Cirnic) using new alteration and lithology codes Underground surveying of voids and update of void model. Underground channel sampling (Cirnic, Cetate) on new levels: 1,841 m (149 channels) Resource update August, 2005 (RSG Global). Additional confirmatory metallurgical testwork (98 variability samples) ; SGS - Lakefield Re-logging of underground workings continued (Cirnic, Cetate) using new alteration and lithology codes Re-logging of drill holes (Cirnic, Cetate) using new alteration and lithology codes Underground surveying and update of void model. Underground channel sampling (Cirnicel) on new level 884 m (156 channels) Geotechnical program of 33 holes for 1000.5m of drilling and 59 2m pits. Additional confirmatory metallurgical testwork (89 11kg samples that were combined to 7 composites) Underground channel sampling (Cetate and Gauri) on 3 levels: 885 m, 855 m and 827 m for 570 channels samples A topographic survey was completed on a 0. 5 m interval in the area of the proposed processing plant and major infrastructure. Additional confirmatory metallurgical testwork Additional surface mapping completed at Cirnic Underground channel sampling on levels: 855 and 714 m (1,233 channels)

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007 2008

30

RMGC also retained MWH to conduct a comprehensive geotechnical audit of the TMF and water management dam designs, with the objectives of determining whether or not appropriate industrial standards were being met and all options for the siting of these facilities had been identified and evaluated. MWH issued its report in February, 2003 IMC was retained to develop the mine design and mineral reserve estimate, based on a mineral resource estimate prepared by RSG late in 2002. IMC also prepared capital and operating cost estimates for the mine, completing this work early in 2003. A group of Romanian and international experts was retained to complete the Environmental Impact Assessment (EIA) for the project. This process was commenced in October, 2002, with the submission of the Technical Memorandum. The government of Romania issued the terms of reference to RMGC for the EIA for the project in December, 2002. Due to planned changes to align Romanian environmental laws with EU directives, the terms of reference were withdrawn in March, 2003. The EIA process was required to be halted and re-started in line with planned new laws. A discussion of all of the foregoing work was included in the RPA report. At the data cutoff date for the RPA report of January 1, 2004, the following work was in progress: MWH was continuing with geotechnical, civil and infrastructure engineering. Stantec Consulting was compiling environmental information and an EIA. Ausenco was conducting an optimization review of the project flowsheet, and was undertaking plant design and preliminary construction planning. RMGC was implementing public disclosure and awareness activities, had implemented the Relocation and Resettlement Action Plan, and was discussing with local, regional and national regulators the proposed scope of the project.

During 2004 and 2005, additional exploration was conducted, and technical, economic and environmental studies continued. As a result of this work, RMGC, in 2005, retained the following consultants to prepare an updated feasibility study, based on the SNC plant designs, but with updated mining plans and infrastructure requirements, and current capital and operating cost estimates: RSG, to prepare an updated resource estimate. IMC, to prepare an updated mine design and production schedule, together with estimates of mine equipment requirements and mine capital and operating costs. MWH, to review the design of the TMF and all water management systems, and to prepare updated capital and operating cost estimates for these facilities. Washington Group International.Inc. (WGI) to prepare updated capital and operating cost estimates for the processing plant and all infrastructure facilites not costed by MWH. David Christensen, to develop an overall financial model for the project.

31

Micon, to review the work performed by all other consultants, and to compile the 2006 Technical Report which discussed the results of the study.

In 2007 RMGC retained SNC to prepare a definitive capital expenditure estimate for the project, in preparation for the commencement of EPCM activities. In September, 2007, however, the Romanian goverment suspended the review process for the project EIA and the work being performed by SNC was halted prior to its completion. Early in 2008, in recognition of the widespread escalation of costs being experienced throughout the worldwide mining industry, RMGC retained the following consultants to conduct a comprehensive review and re-estimation of all components of capital expenditure and operating cost, with the objective of optimizing the design of the project to minimize these costs. IMC was retained to re-estimate mining costs and to re-evaluate the open pit designs, the scheduling of mine production, and mine equipment requirements. Aurifex was retained to re-estimate process operating costs and to review the process flowsheet. MWH was retained to review the design of all project earthworks and to re-estimate the capital cost of these works. Metifex was retained to review all general arrangement drawings, process flow diagrams, single line electrical drawings, and piping and instrumentation diagrams, and to re-estimate the capital cost of all project facilities, other than mining facilities and those items covered by MWH.

This work was completed early in 2009 and is discussed in Section 18 of this report. 6.3 RESOURCE ESTIMATES

Resource estimates have been completed for the Rosia Montana deposit by RSG from 1998 to 2002, and then by RSG Global from 2003 onwards. A summary of the resource estimates completed by RSG Global from 2001 through to 2003, reported at a 0.6 g/t Au lower cutoff grade, is presented in Table 6.2. All resource estimates were prepared in accordance with the Canadian Institute of Mining, Metallurgy and Petroleum (CIM), CIM Standards on Mineral Resources and Reserves, Definitions and Guidelines, as those standards existed at the date of the estimate. The mineral resource estimate discussed in this report is based on all exploration performed by RMGC from January, 1998 to May, 2005. No update of the mineral resource has been completed since that reported in the 2006 Technical Report.

32

Table 6.2 Rosia Montana Grade Tonnage - Comparison Report November, 2003 OK, November, 2002 MIK , 2002 OK versus February 2001 MIK Lower Cutoff Grade 0.6 g/t Au Tonnes (x1000) Au (g/t) Measured OK- Nov03 MIK - Oct '02 OK - Oct '02 MIK-Feb'01 OK- Nov03 MIK - Oct '02 OK - Oct '02 MIK-Feb'01 OK- Nov03 MIK - Oct '02 OK - Oct '02 MIK-Feb'01 OK- Nov03 MIK - Oct '02 OK - Oct '02 MIK-Feb'01 135,515 138,427 135,254 134,584 216,752 159,531 170,206 161,873 352,267 297,958 305,461 296,457 48,146 40,103 50,527 47,624 1.6 1.6 1.6 1.6 Indicated 1.1 1.1 1.1 1.2 Measured + Indicated 1.3 1.3 1.3 1.4 Inferred 1.0 0.9 1.0 0.9 4 3 4 4 1,504,000 1,178,000 1,636,000 1,305,000 5,959,000 3,881,000 5,807,000 5,668,000 6 6 6 6 14,625,000 12,773,000 13,067,000 12,990,000 67,324,000 58,935,000 60,936,000 61,337,000 4 4 5 5 7,724,000 5,618,000 6,177,000 6,127,000 29,508,000 21,169,000 25,109,000 24,909,000 9 8 8 8 6,901,000 7,155,000 6,890,000 6,863,000 37,816,000 37,766,000 35,827,000 36,428,000 Ag (g/t) Contained Gold (oz) Contained Silver (oz)

Note: The Cos region was removed from the 2002 OK model to ensure consistent reporting regions

33

7.0 GEOLOGICAL SETTING

7.1 REGIONAL GEOLOGY Romania includes three major Alpine and older orogens, namely the Carpathian chain that comprises the Southern Carpathians and the Eastern Carpathians, the Apuseni Mountains and the Northern Dobrogea. Tertiary sediments were deposited in the intervening Pannonian and Transylvanian Basins, as well as on the Scythian and Moesian Platforms. Two principal areas of Tertiary volcanic rocks, of predominantly calc alkaline affinity, intrude and overlie these sequences. The first one spreads in the Eastern Carpathians from the north in the Baia Mare area (Oas-Gutai mountains) to the south (Calimani-Gurghiu-Harghita mountains), containing also a subvolcanic median sector (Tibles-ToroiagaRodna- Bargau mountains). The second area with Tertiary volcanic rocks is the Apuseni Mountains in central-western Romania. The famous mining districts of the Metaliferi Mountains of Transylvania, which represent the southern part of the Apuseni Mountains, comprise a 900 km2 region, immediately to the north of the city of Deva, commonly referred to as the Golden Quadrilateral (Figure 7.1). The Golden Quadrilateral has remained Europes most important centre of gold production for more than 2,000 years since Geto-Dacian (preRoman) times, with the Roman conquest of Dacia in 105 AD-106 AD predicated on gaining control over this important goldfield. The district reached peak production during the period of the Austro-Hungarian Empire at the end of the 17th century to 1918, as well as before World War II. The Golden Quadrilateral lies within the Apuseni Mountains, which consist of Mesozoic, shallow marine and non-marine sedimentary rocks overlying Palaeozoic and Precambrian sedimentary and metamorphic basement. North-directed thrust faulting during the late Cretaceous resulted in a series of nappes that are unconformably overlain, and intruded, by Tertiary volcanics associated with high-level gold-silver mineralization and porphyry copper deposits of the Golden Quadrilateral. According to the classical view regarding the Tertiary volcanism from the Apuseni Mountains, three cycles has been distinguished (Ianovici et al., 1976). The earliest cycle is interpreted as lower Badenian age and comprises rhyolitic ignimbrite overlain by rhyodacitic and andesitic volcanics. Volcanogenic sediments occur throughout this cycle and widespread hydrothermal alteration overprints all rock types. The rocks of the second cycle outcrop extensively and are characterized by andesite and dacite overlain by a very thick sequence of quartz andesite that is, in turn, overlain by pyroxene andesite. The sequence is interpreted to be late Badenia-Sarmatian and Pannonian age. The middle (dacite) and upper (quartz andesite) sequence of this cycle represents the principal host to gold-silver mineralization currently being mined in Romania, as well as significant occurrences of copper, lead, zinc and mercury. The third and final cycle of volcanism continued into the Quaternary era and is characterized by pyroxene andesite, basaltic andesite and potassic basalt.

34

Figure 7.1 Regional Geology of Rosia Montana Area

35

According to available K-Ar dating (Pecskay et al., 1995), the main volcanic activity from the Apuseni Mountains ranges between 14.7 and 7.3 Ma, and ended in the Quaternary (1.6 Ma). Three major northwest-trending belts of volcanism (Brad-Sacaramb, Zlatna-Stanija, and Rosia MontanaBucium) and associated mineralization are identified within the Golden Quadrilateral, with the Rosia Montana Complex representing part of the northernmost belt. 7.2 PROJECT GEOLOGY AND MINERALIZATION CONTROLS

The local Rosia Montana deposit is interpreted as a maar-diatreme complex of Neogene age emplaced into a sequence of Cretaceous sediments, predominantly black shales, sandstones and conglomerates. Marza et al. (1997) documented the Rosia Montana deposit as a low sulphidation epithermal deposit. More recently, Rosia Montana has been interpreted as intermediate sulphidation epithermal (Sillitoe and Hedenquist, 2003). The deposit is now considered to have evolved from a low sulphidation epithermal system to a more intermediate epithermal system. The three-dimensional geometry of the area is well established due to the extensive network of underground development that has been undertaken since the Austro-Hungarian Empire period, and from the extensive surface and underground drilling completed in the last 25 years. Lithologies within the diatreme complex are dominated by breccias, including magmatic-phreatic and sub-aqueous reworked breccia, intruded by porphyritic dacitic sub-volcanic intrusives. These intrusions are interpreted as Neogene age and are informally named the Cirnic and Cetate dacites (Cernic and Cetate massifs). The dacite bodies are interpreted to have intruded vertically through the diatreme breccias and to have spread laterally at shallower levels forming surface domes (Figure 7.2). An alternative interpretation is that only one major dacite intrusion has occurred and that this has been split into the now separate Cirnic and Cetate dacite bodies by a northeast trending strike-slip fault. The majority of the Rosia Montana diatreme is made up of a lithology locally referred to as the vent breccia. This is a diatreme breccia produced by numerous magmatic-phreatic eruptions, as hot rising dacitic magma interacted with groundwater. It is of variable composition with clasts of dacite, Cretaceous sediments and basement schist and gneiss. The clast size, degree of rounding, and the proportion of matrix, vary widely. Texturally it exists as both massive breccia units and sub-aqueous reworked breccia indicating that the breccia has erupted into a shallow lake or maar. This reworked vent breccia is fine to coarsely bedded and varies from clay rich to more common sandy and gravely beds, through to beds containing poorly sorted, cobble sized clasts. Graded bedding is common and cross bedding and ripple marks have also been observed. The vent breccia hosts the dacitic intrusives as well as multiple later crosscutting phreato-magmatic breccia bodies.

36

Figure 7.2 Rosia Montana Project Geology

37

Within the vent breccia a large (1 km by 1 km at the surface) body of breccia has been delineated as a separate phreato-magmatic event. This is composed of generally massive (some sedimentary reworking exists in the upper levels), poorly sorted, matrix supported breccia, with sub-angular to subrounded clasts of dacite, Cretaceous sediments and basement metamorphics. The breccia is distinct because it is the only breccia body at Rosia Montana that still contains significant magnetite. The magnetite is coarse and interpreted to be of magmatic origin. Chlorite also occurs as an accessary mineral in altered clasts. Some sulphide is present as pyrite particles in the matrix or in scattered phyllic-argillic altered clasts of dacite. It is interpreted that this breccia was emplaced after the main period of sulphide mineralization, which is supported by a 40Ar/39Ar date on a hornblende separate from juvenile dacite clasts (11.0 + 0.80 Ma), which is over 1 million years after the youngest date recorded for the mineralization (12.71 + 0.13 Ma). A breccia, locally termed the Black Breccia (Glamm), forms a sub-vertical pipe in the centre of the diatreme, between the Cetate and Cirnic dacites. The breccia is matrix supported, with clasts of dacite, Cretaceous sediment and basement garnet-bearing schist and gneiss in a matrix of pulverized (rock flour) Cretaceous shale which gives the breccia its black, clay rich character. The unit only hosts significant gold-silver mineralization along its northeastern margin. A number of intrusive polymictic (XPO) diatreme (phreato-magmatic) breccia bodies crosscut the reworked vent breccia and also the Cirnic and Cetate dacites. Five of these have been identified between Tarina and Jig (Igre), where they have a sub-vertical, pipe or lens geometry up to 100 m in width. Within the dacites they form multiple broad (up to 100 m wide) to narrow (a few centimetres wide) dykes or breccia pipes. In both locations the breccias are structurally controlled, with the pipes and broad dyke breccias forming at the intersection of large structures. These are composed of matrixsupported breccias with sub-rounded clasts of dacite, Cretaceous sediment and crystalline schists. The breccia bodies between Tarina and Jig have a high proportion of metamorphic clasts, making them easily identifiable from the surrounding vent breccia and also indicating that the phreato-magmatic eruption originated in the metamorphic basement at depth. Texturally they vary from poorly sorted to moderately sorted with a matrix of fine clay (pulverized rock) in many of the dykes at Cirnic and Cetate, to a more sandy matrix in the Tarina and Jig breccias. This is largely controlled by the composition and the energy of emplacement of the breccia body, with breccias dominated by Cretaceous shale having a darker, clay rich matrix. The Sandy (XCF) breccia is a narrow (25 m to 50 m diameter) pipe that cuts through the northwest corner of the Cirnic dacite. It is interpreted to have erupted up the intersection of two large structures and has a plunge of 55 to the southwest. In its upper levels the breccia is dark gray to black, strongly silicified with disseminated pyrite and a fairly massive, homogeneous texture and a medium to coarse sand grain size. At depth the breccia becomes coarser with zones where the clasts are 0.5 to 1 cm in size. In the Igre area, polymictic breccia and dacite dykes have also been identified within the Cretaceous sediment. The breccia dykes are usually 0.5 m to 2 m in width and are lithologically similar to the diatreme breccias described above. These are interpreted to be offshoots of the diatreme breccia material along smaller faults. The dacite dykes are usually 0.5 m to 1 m in width and are composed of fine-grained, silicic dacite, with coarse sand-sized phenocrysts of quartz. Dykes containing both polymictic breccia and dacite have also been observed.

38

Andesitic extrusive rocks are mapped mantling the northern and eastern parts of the project area, forming a thin to moderately thick cover over the maar complex. The lowest units within the sequence are pyroclastic block and ash flows; further north and east andesitic lavas overlie the pyroclastics. 7.2.1 Structure

Structure has played an important role at Rosia Montana, firstly supplying dilation for the emplacement of the maar-diatreme complex, and secondly the structural permeability up which the mineralizing fluids have flowed. Two types of structures have been identified at Rosia Montana, regional scale faults and more localized faulting related to the formation of the diatreme. The Rosia Montana diatreme is interpreted to be emplaced at the intersection of two sub-vertical structures that trend north-northwest and northeast (Figure 7.3). The north-northwest trend is interpreted to be the more pervasive of the two structures, and is thought to be a deep seated, basement structure of regional scale that has produced the broad zone of fracturing and veining seen on the surface within the dacites, vent breccia, Cretaceous sediments and in the andesites at Bucium. It can be traced from Tarina through Orlea, Cetate and Cirnic, and down through the Bucium exploration license to the south; along this 13 km trend the dominant vein orientation strikes parallel to it and is either sub-vertical or dipping steeply to the west. A system of large northeast trending sub-vertical faults bounds and cuts through the Cirnic and Cetate dacites (Figure 7.4). At Cirnic these have provided a structural path for polymictic breccia dykes and also offset the dacite-vent breccia contact. A system of northwest trending faults has been mapped along the northeast side of Cirnic (Figure 7.5). These are interpreted as being active during the diatreme formation as these structures can contain phreato-magmatic breccia dykes, during mineralization and also post mineralization, as they can be unmineralized but cut through well mineralized dacite or breccia. The largest of these is the Saturday Morning Fault, which, with another crosscutting fault, is interpreted to have provided the structural dilation for the emplacement of the Sandy (XCF) breccia pipe. Similar, moderate to steeply southwest dipping northwest trending structures have been identified in the Igre area. These are interpreted to have a right lateral sense of movement, possibly with late dip-slip (normal) reactivation. Mapping at Cirnicel and Cetate has also identified a fault structure (the Cirnicel l fault) that strikes eastwest and dips 40-60 to the south. The last phase of movement on the fault post-dates the mineralization as the fault offsets the late stage carbonate-base metal sulphide Argint vein at Cirnicel. The sense of movement appears to be normal, possibly with some right lateral strike slip. Other eastwest structures have also been mapped and interpreted at Cirnic and Cetate, with east-west to westnorthwest trending sub-vertical polymictic breccias cutting through the Cetate dacite. At Orlea, a large east-west trending clay rich vein/fault (the Cruce vein) dipping at 45 to the south, cuts across the vent breccia. This structure is interpreted to be related to diatreme collapse immediately following eruption of material from the vent. Similar structures have also been interpreted in the Tarina and Igre areas and also tend to dip moderately to steeply towards the centre of the diatreme.

39

Figure 7.3 Structural Interpretation of Rosia Montana

40

Figure 7.4 Structural Summary of Mapped and Interpreted Structures at Cetate and Cirnic

Figure 7.5 Section across Cirnic from the WNW to ESE, view looking north

Blue dashed lines are major NE structures

41

7.2.2

Alteration

An extensive zone of strong hydrothermal alteration hosts the Rosia Montana deposits. The distribution of alteration assemblages is quite complex, however, it can be simplified down to the following groupings (Manske 2004): Chlorite-carbonate-smectite alteration, peripheral to the resource. There are not many locations where this alteration has been preserved except for a few small outcrops on Cirnic hill. Although a green alteration that superficially resembles a propylitic imprint, this zone lacks epidote or appreciable albite (it may preserve relicts of the original igneous plagioclase, andesinelabradorite). It most closely matches the intermediate argillic alteration type in the sense of Meyer and Hemley (1967). Phyllic-argillic alteration. The most widespread alteration at Rosia Montana, creating most of the bleached exposures in dacite porphyry. Rocks affected by this alteration show abundant fine-grained sericite (illite) and murky, birefringent clays (smectites) in thin section. XRD analyses and glycolation tests indicate the presence of mixed-layer, illite-smectite minerals as well (cf. Tamas, 2002). Supergene argillization certainly exists in the weathering profile over the pyritic orebodies, but the presence of phyllic-argillic altered rock in the sulphide zone shows that much of the clay is hypogene. QIP (quartz-illite-pyrite) alteration. QIP appears as a subtype of the general phyllic-argillic imprint, around some quartz veins and also as local zones of more pervasive alteration. It imparts a bluish-grey colour to the rock and may be confused with moderate silicification, but may be easily scratched with a knife blade or steel dental pick (H < 5). Silica may or may not have been added to QIP rock, as the quartz may be generated by release of SiO2 during hydrolytic alteration of feldspar (e.g., Barton et al., 1991). This alteration type may destroy much of the primary igneous texture of the rock, except for quartz phenocrysts. Quartz-adularia replacement. Adularia quartz appears in vein halos and as more pervasive alteration in dacite and vent breccia. Fine-grained adularia in some cases effects a wholesale replacement of the dacite matrix and creates a pale rock with ceramic-like texture. In extreme cases, original quartz phenocrysts are attacked and partly replaced with adularia along grain boundaries and micro-cracks. As rock alteration, adularia commonly is partly overprinted by sericite (illite) or clays. As a vein mineral in dacite or a vug-filling phase in breccia, paragenetic studies show that adularia was introduced into the system at two or three distinct stages. Silicification occurs in some quartz vein halos but is not volumetrically widespread in unbrecciated dacite, at least in the Cetate area. Strong silicification is much more characteristic of the margins of breccia zones within or along the edges of the dacite flow-domes. Brecciated dacite but especially the breccia matrix may be densely silicified, replaced by more than 90 vol. % fine-grained quartz as seen in thin section, and much harder than a knife blade in hand sample (H 7).

XRD and TEM analyses (Marza et al, 1997, Tamas, 2002) confirmed the following alteration assemblages at Rosia Montana: potassium silicate assemblages, phyllic assemblages, intermediate argillic, and advanced argillic assemblages.

42

8.0 DEPOSIT TYPES

The local Rosia Montana deposit is interpreted as a maar-diatreme complex of Neogene age emplaced into a sequence of Cretaceous sediments, predominantly black shales, sandstones and conglomerates. Marza et al. (1997) documented the Rosia Montana deposit as a low sulphidation epithermal deposit. More recent re-classification (Sillitoe and Hedenquist, 2003) would suggest that the deposit is now considered to initially have been a low sulphidation epithermal system over-printed by a later more intermediate epithermal system. Mineralization styles include veins, disseminated sulphides, stockworks, and breccia-fillings. Grades vary widely throughout the district, from disseminated-style mineralization grading 0.5 - 2.0 g/t Au, to localized vein and breccia deposits carrying over 30 g/t Au. The principal sulphide mineral is pyrite with very minor accessory marcasite, galena, sphalerite, chalcopyrite, tetrahedrite, and silver sulphides and sulphosalts. Gold occurs as minute grains of electrum (Figure 8.1) in association with the pyrite.
Figure 8.1 Electrum Rosia Montana Mine

43

9.0 PROJECT MINERALIZATION

Mineralization within the Golden Quadrilateral district includes porphyry-related gold-silver, copper-gold and copper deposits associated with Badenian-Pliocene (Neogene) andesitic to dacitic volcanic rocks, and associated intrusive rocks. The gold-silver mineralization outlined at Rosia Montana is interpreted to represent a mid- to shallow-level, low to intermediate sulphidation epithermal system. The mineralization is dominantly disseminated, with associated stockwork and breccia hosted gold-silver mineralization. 9.1 MINERALIZED LITHOLOGIES

Gold-silver mineralization at Rosia Montana is hosted by the following lithologies: 1. Dacite-hosted mineralization: Characterized by wide zones of finely disseminated sulphide (pyrite) hosted within dacite porphyry. QIP and silica-adularia alteration are distinctive features of the mineralized dacite and the best indicator of gold and silver grade. Narrow, usually widely spaced stockwork veining is always present but is minor in terms of contained gold and silver. The veins are generally steeply-dipping, discontinuous and less than 1 m wide; in places the veins have blown out into narrow hydrothermal breccia pipes. Significant gold mineralization of this style occurs at Cetate, Cirnic, Carpeni, Gauri, Cos and parts of the Vaidoaia zone. 2. Sub-vertical breccia zones crosscutting dacite intrusive bodies: Breccias are commonly of mixed lithology and are considered to represent structurally controlled phreato-magmatic breccias. Mineralization occurs within strongly, to intensely, silicified alteration zones which contain low to moderate amounts of disseminated fine-grained sulphide within both the matrix and breccia clasts. Relevant examples of this type are known in the Cetate and Cirnic massifs. Along the margin of the polymictic breccias often occurs a zone of monomictic crackle breccia, with angular blocks of dacite in a fine matrix of rock flour that originated from the polymictic breccia bodies. These can be fairly localized features and formed as the polymictic breccia events crackle brecciated the surrounding dacite or caved blocks of the adjacent dacite into the breccia pipes or dykes. These crackle breccia zones often have a better permeability and therefore focus the mineralizing fluids and become well mineralized. At Cirnic a significant volume of this type of brecciation occurs at the bend in the main XPO breccia. This is also coincident with a zone of well mineralized hydrothermal brecciation and an area that has been heavily mined by large corandas (stopes) in the past. 3. Disseminated and vein hosted gold-silver mineralization within vent breccia: Significant gold-silver mineralization is hosted by the vent breccia surrounding the dacitic intrusions. The mineralization is characterized by silicification and finely disseminated pyrite with veining infrequent, and generally narrow (less than 1 m). Examples of this style of mineralization are at Cirnicel, Vaidoaia, Jig (also known as Lespedari), Igre, Orlea and Tarina.

44

4.

Diatreme breccia pipe hosted mineralization: This mineralization is hosted by the sub-vertical diatreme breccia pipes at Igre/Jig. It is characterized by intense, pervasive silicification of both the breccia matrix and the diatreme breccia clasts. Disseminated pyrite is also pervasive within the matrix and clasts and will sometimes completely replace the black shale clasts. Zones of rhodochrosite have also been identified, occurring within the matrix of the diatreme breccia.

5.

Cretaceous sediment hosted mineralization: This mineralization has been identified at Igre, Gauri, East Cirnic and Cos. The mineralization occurs directly below the vent breccia-Cretaceous sediment contact and is usually hosted by shale, sandstone and less frequently by conglomerate beds. The mineralization is characterized by both silicification and pervasive fine-grained disseminated pyrite and in some areas (Igre, Gauri and East Cirnic) by hydrothermal crackle brecciation that varies from mm-width widely spaced spidery crackle breccia through to more intense mosaic (jigsaw) brecciation. Clasts are always very angular and made up of locally derived sediment. The brecciation can be over 50 m thick and tends to be most intense close to the vent breccia-Cretaceous contact. The breccia matrix is typically vuggy and crystalline, some coliform banding has been observed and up to five phases of mineralization can be present. Mineralization is dominated by carbonate (both calcite and rhodochrosite), quartz and pyrite with galena and sphalerite not uncommon and rarer chalcopyrite. Gold has been identified by petrography in numerous samples as electrum. Occurrences were noted as minute (4 m) inclusions in pyrite, as minute grains (up to 25 m) intergrown with, and overgrowing Ag-sulphosalts and tellurides. It has also been observed as coarser grains (up to 100 m) intergrown with carbonate and barite with drilling locally intersecting some very coarse (1 cm) occurrences. The electrum is also associated with quartz, galena, and sphalerite. The electrum had a fineness ranging from .537 to .763 (Leach & Hawke, 1997). Fluid inclusion studies indicate that mineralization was precipitated from a dilute NaCl (0.35 to 7.85 wt. %) hydrothermal solution at temperatures between 200 to 340C. Both boiling and the rising mineralizing fluid mixing with oxidizing CO2-rich meteoric waters are proposed as mechanisms for triggering gold precipitation (Leach & Hawke 1997, Tamas, 2002 and Manske, 2004).

9.2

VEINING

Although veining is secondary to alteration in terms of contained gold and silver, veins are important as fluid conduits and therefore control the location and distribution of the gold-silver mineralization. The veins often contain significantly higher grade gold and were therefore the focus of historical mining. Multiple phases of mineralization have been identified at Rosia Montana that can be subdivided into 3 main vein types:

45

1.

Chinga veins: The Chinga is a black, very fine-grained argillic material, intensely silicified. These veins are the earliest phase of veining at Rosia Montana and their textures indicate that the argillic material has been injected into fractures rather than precipitating from a hydrothermal solution. Chinga material can also occur as breccia matrix and the texture is typically massive. Veins are often overprinted by later phase mineralization. This material was probably derived from pulverized Cretaceous black shales and has been emplaced during dacite or phreato-magmatic breccia eruptions. The chinga veins are most common in the upper levels of the Cirnic and Cetate dacites. The grade of chinga material is typically 0.5 to 3 g/t Au.

2.

Quartz-carbonate-sulphide veins: These veins are associated with the main mineralizing phase at Rosia Montana. Veining occurs as multiphase, quartz-carbonate-sulphide (pyrite) + adularia and rare base metal sulphides; texture varies widely from massive to banded, often vuggy and from very fine grained to coarsely crystalline. Carbonate dominated veins are interpreted to have been deposited late in this mineralized phase and are often associated with base metal sulphides. Carbonates are predominately calcite and rhodochrosite, with rarer dolomite and siderite. Base metal sulphides are dominated by galena, low Fe sphalerite and chalcopyrite. The veins are typically narrow (up to 30 cm) and more common in the deeper levels at Cirnic and Cetate. This also appears to be the main mineralizing phase in the Igre Cretaceous sediments and also the commonest vein type around the margins of the Black Breccia. Textures are typically banded and vuggy, sometimes massive. The very last phase is dominated by calcite with marcasite and very little gold or silver mineralization. Two telluride occurrences in rhodochrosite-rhodonite high grade silver veins were recently located (Tamas et al., 2004).

3.

Clay veins: These veins are generally considered late stage, possibly retrograde. Locally referred to as veins as they contain some gold-silver mineralization, many of these could be classified as shear or fault structures. Clays vary from illite-smectite to illite and often occur with pyrite and marcasite. They usually display a massive or sheared texture. Veins are usually discontinuous, widely spaced and only millimetres to centimetres in width, but have been observed up to 1 m wide. They generally become less frequent, but thicker and more continuous at depth. Numerous vein measurements have been taken during the geological logging of underground and surface channel samples, geological mapping and from orientated core. Vein orientations can vary widely but general trends are evident. Table 9.1 and Figure 9.1 summarize and illustrate these trends.

46

Table 9.1 Rosia Montana Gold Project Vein Orientations Grouped by Region Area Cirnic Cirnic - Central/west Cirnic - North/NE Cirnic - South Cirnicel Cetate Gauri Orlea Igre-Tarina Jig Typical Vein Orientations Numerous vein orientations occur but the veins are typically steeply dipping (less than 40 dip is rare) and the commonest vein orientation has a northerly strike and dips steeply to the west. Variations are also seen within the Cirnic area (see below). North striking, steeply dipping to the west. Northwest to northeast strike, moderate to steeply dipping to the west. Dominant vein orientation has a northwest strike and a sub-vertical through to steep southwest dip. Very distinctive north-northwest strike and vertical dip. Dominant vein set has a north to north-northwest strike, a steep westerly dip. A less common carbonate dominated vein set strikes west northwest with a moderate dip to the north. Dominated by north striking veins with a sub-vertical or steep easterly dip. Two vein orientations; sub-horizontal carbonate veins (rhodochrosite dominant), and a second cross cutting set with a strong north to north-northwest strike and subvertical dip. Northwest strike and a moderate to steep southwesterly dip. A second subhorizontal orientation exists in the diatreme breccia bodies. The dominant vein set has a north-northwest strike and a steep westerly dip. A second vein set has a east northeast strike and a shallow dip to the northwest. Typical Dip / Dip Direction 70/265 70/275 70/315 to 75/240 80/225 85/245 85/260 and 50/345 80/080 05/175 and 80/260 65/220 and 15/345 75/240 and 35/295

Figure 9.1 Vein Stereonets for Cetate and Cirnic

10

12

Stereonet summary including only those domains with sufficient data. Data is the poles of veins measured during channel sample logging and from orientated core (over 3500 veins at Cirnic, 1500 at Cetate). Stereonets are equal angle, lower hemisphere.

47

10.0 EXPLORATION
A description of the exploration activities at the Rosia Montana property has been included previously in Section 6.1.

48

11.0 DRILLING
Mineral resource delineation at Rosia Montana is based on a combination of drilling and channel sampling. The drill hole database is comprised of surface RC and diamond drilling, and underground diamond drilling. Drilling has been undertaken by Genfor SRL, the Romanian subsidiary of RB Drilling Ltd. The company used a variety of RC, diamond and multi-purpose drill rigs, including Shramm T985, T64 and T66 RC rigs, Warman UDR1000 multi-purpose rig, G&K 850 multi-purpose rig, Edson 6000 RC rig, KL400 multipurpose rig, Drill Tech DK40 RC rig, Longyear 44 DC rig, and RB57 DC rig. The available drilling and underground channel sampling are summarized in Table 11.1, sub-divided on the basis of data being included into the mineral resource estimate reported in this document (data up to and including 2005) or being collected from 2006. The geotechnical drilling is listed separately as this drilling is also excluded from the mineral resource estimate.
Table 11.1 Rosia Montana Gold Project Exploration Drilling and Channel Sampling Summary Exploration Drilling RC Holes 629 Samples 0 Metres 75,436 73,343 0 Surface Channels Channels 387 Samples Metres 8,884 7,760 RC / DDH Holes 131 Samples 0 Metres Surface DDH Holes Metres Underground DDH Holes Metres 11,820 11,715 0 Holes 1,108 Samples 33 Total Metres 136,578 128,566 999.5

29,237 276 20,085 72 28,605 Samples 14,903 Samples Post 2005 Geotechnical Drilling (Not Sampled) 0 33 Channel Sampling Underground Channels 999.5 0

Total Channels Channels 1,688 Samples Metres 71,952 62,754 2,078 1,838

Total Channels + Holes Channels 2,796 Samples 2,897 Samples Metres 208,530 191,320 210,608 130,404

Channels 1,301 Samples Post 2005

Metres 63,068 54,994

Up to and including 2005

0 0 101 2,078 101 Samples 0 Samples 1,838 Samples Notes: RC: Reverse circulation drilling. RC/DDH: Reverse circulation drilling with a diamond core tail. DDH: Diamond drill hole. Geotech drilling excluded from totals.

11.1

RC DRILLING

RC samples were routinely collected at 1 m intervals and the cuttings split with a Jones riffle splitter. Field duplicates were taken via the splitter every 20 samples. The bags of cuttings were routinely weighed prior to taking the sub-sample via the Jones riffle splitter. 11.2 DIAMOND DRILLING

A combination of surface and underground diamond drilling was completed as part of the 1998 to 2005 resource delineation program. The drilling comprised:

49

PQ (1%). HQ (42%). NQ (57%).

To ensure a high sample quality stringent data collection quality control procedures have been applied. Details of the procedures and equipment used to maximize core recovery and the quality of RC samples are provided in Section 12. The location of the drill hole collars, colour coded by drilling method, is displayed in Figure 11.1. RSG Global is satisfied that all drilling at Rosia Montana has been completed by reputable, Romanian based, drilling contractors. 11.3 CHANNEL SAMPLING

Extensive channel samples have been collected with the collection method described in Section 12. Significant intercepts of the data not included in the resource estimate reported in this document are provided in the following section. 11.4 SIGNIFICANT INTERSECTIONS

The drilling and channel data collected up to and including 2005 have been used in the mineral resource estimate reported in this document. These data have been considered in 3 dimensions and therefore consider the attitude and geometry of the mineralization. Underground channel data collected between 2006 and 2008 have not been included in the resource estimation study as the 1,838 intervals are not considered to materially impact the resource quantum. Significant intercepts for these data are provided as Table 11.2, 11.3 and 11.4 for the 2006, 2007 and 2008 sampling programs, respectively. Sample location plans for the 2006 to 2008 channel sampling are presented as Figures 11.2 through 11.7.
Table 11.2 Summary of Significant Intercepts 2006 Channel Sampling >0.6 g/t Au Lower Cutoff Grade; Minimum 3 m Intercept

Drive G88431 G88433 G88433

Note:

Prospect Carnicel Carnicel Carnicel

From (m) 11.0 14.0 55.0

To (m) 17.0 16.0 58.0

Width (m) 6 2 3

Au (g/t) 1.13 0.71 1.24

Ag (g/t) 2.5 7.5 2.0

Significant Intercepts calculated using weighted averages of the first reported Au values. Lower Cut: 0.6 g/t Au. Upper Cut: 25 g/t Au. Maximum 3 m of consecutive waste

50

Figure 11.1 Drill Hole Collar Locations Yellow: DDH, Blue: RC/DDH, Red: RC

51

Table 11.3 Summary of Significant Intercepts 2007 Channel Sampling >0.6 g/t Au Lower Cutoff Grade; Minimum 3 m Intercept

Drive G8851 G8851 G8851 G8851 G8852 G8853 G8854 G8854 G8854 G8854 G8855 G8856 G8856 G8856 G8858 G8858 G88511 G88512 G88513 G82401 G82401 G82402 G82402 G82402 G82403 G82404 G82405 G82406 G82407 G82408 G82409 G82412 G82413 G82414 G8271 G8272 G85501 G85501 G85501 G85501 G85501 G85501 G85501 G85501 G85501 G85501 G85501
Note:

Prospect Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri

From (m) 5.0 21.0 44.0 57.0 0.0 0.0 11.0 37.0 61.0 72.0 1.0 6.0 19.0 33.0 0.0 32.0 1.0 1.0 1.0 0.0 11.0 0.0 6.0 12.0 6.0 1.0 0.0 0.0 0.0 1.0 16.0 0.0 0.0 4.0 2.0 0.0 5.0 15.0 23.0 32.0 43.0 46.0 53.0 63.0 74.0 84.0 106.0

To (m) 15.0 34.0 53.0 73.0 2.0 2.0 34.0 52.0 62.0 73.0 3.0 9.0 29.0 37.0 26.0 33.0 3.0 9.0 2.0 6.0 22.0 1.0 9.0 16.0 9.0 3.0 4.0 4.0 8.0 2.0 17.0 9.0 33.0 6.0 36.0 7.0 6.0 19.0 25.0 33.0 45.0 52.0 59.0 64.0 76.0 93.0 107.0

Width (m) 10 13 9 16 2 2 23 15 1 1 2 3 10 4 26 1 2 8 1 6 11 1 3 4 3 2 4 4 8 1 1 9 33 2 34 7 1 4 2 1 2 6 6 1 2 9 1

Au (g/t) 1.71 5.37 0.88 1.25 1.02 1.08 13.10 1.42 0.66 0.71 0.80 0.67 1.09 0.95 1.29 0.60 0.84 0.64 0.76 1.08 0.79 0.67 0.60 1.33 0.96 1.61 0.95 0.79 0.92 0.63 12.50 0.72 0.84 1.58 1.04 0.98 1.09 1.42 1.20 0.66 0.95 3.18 0.78 0.68 0.64 2.33 0.60

Ag (g/t) 5.3 7.5 6.3 2.4 5.5 3.0 33.1 3.9 3.0 22.0 9.5 3.0 4.1 2.0 14.5 13.0 2.5 4.3 4.0 3.2 5.5 8.0 2.3 4.8 4.0 3.5 2.8 2.3 2.4 3.0 14.0 4.9 3.0 2.0 3.9 3.3 6.0 3.0 2.5 1.0 1.0 6.0 7.0 <1 3.0 2.2 3.0

Significant Intercepts calculated using weighted averages of the first reported Au values. Lower Cut: 0.6 g/t Au. Upper Cut: 25 g/t Au. Maximum 3 m of consecutive waste.

52

Table 11.4 Summary of Significant Intercepts 2008 Channel Sampling >0.6 g/t Au Lower Cutoff Grade; Minimum 3 m Intercept Drive G85501 G85503 G85506 G85507 G85509 G85510 G85511 G85512 G85513 G85514 G85515 G85516 G85519 G85522 G85522 G85523 G85523 G724110 G724110 G724111 G724112 G724113 G724114 G724114 G724118 G724119 G724121 G724122 G724123 G724124 G724124 G724125 G724125 G724126 G724126 G724126 G724127 G724128 G724128 G724129 G724130 G724131 G724131 G724132 G724132 G724132 G724132 G724132 G724132 G724132 G724132 G724133 G724136 G724137 G724138 G724138 G85501 G85503 G85506 G85507 G85509 G85510 G85511 G85512 G85513 G85514 Note: Significant Intercepts calculated using weighted averages of the first reported Au values. Lower Cut: 0.6 g/t Au. Upper Cut: 25 g/t Au. Maximum 3 m of consecutive waste. Prospect Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri Cetate-Gauri From (m) 121.0 0.0 12.0 3.0 10.0 2.0 1.0 0.0 2.0 1.0 1.0 0.0 7.0 71.0 77.0 2.0 31.0 0.0 10.0 0.0 23.0 4.0 15.0 27.0 3.0 9.0 0.0 2.0 5.0 0.0 11.0 0.0 11.0 8.0 31.0 43.0 3.0 2.0 11.0 10.0 2.0 0.0 23.0 9.0 17.0 27.0 55.0 86.0 98.0 107.0 113.0 1.0 0.0 2.0 6.0 19.0 121.0 0.0 12.0 3.0 10.0 2.0 1.0 0.0 2.0 1.0 To (m) 137.0 6.0 13.0 4.0 12.0 6.0 7.0 5.0 18.0 4.0 5.0 4.0 8.0 73.0 78.0 6.0 37.0 4.0 11.0 6.0 30.0 7.0 16.0 28.0 9.0 11.0 13.0 7.0 11.0 6.0 24.0 6.0 20.0 22.0 33.0 49.0 10.0 3.0 24.0 12.0 7.0 10.0 24.0 14.0 23.0 30.0 58.0 88.0 101.0 109.0 119.0 3.0 10.0 3.0 8.0 26.0 137.0 6.0 13.0 4.0 12.0 6.0 7.0 5.0 18.0 4.0 Width (m) 16 6 1 1 2 4 6 5 16 3 4 4 1 2 1 4 6 4 1 6 7 3 1 1 6 2 13 5 6 6 13 6 9 14 2 6 7 1 13 2 5 10 1 5 6 3 3 2 3 2 6 2 10 1 2 7 16 6 1 1 2 4 6 5 16 3 Au (g/t) 1.94 0.72 1.12 0.63 1.98 1.94 0.93 3.44 1.67 0.66 1.03 2.50 1.01 0.74 0.77 1.07 2.82 1.05 0.85 0.98 1.99 0.74 2.40 0.65 1.62 0.98 3.19 3.27 0.56 2.29 0.75 0.80 0.64 1.15 0.82 0.61 0.70 0.61 0.90 0.63 1.54 0.92 0.80 2.23 0.67 0.53 1.37 0.95 0.62 0.88 1.23 1.55 0.56 1.45 0.73 0.85 1.94 0.72 1.12 0.63 1.98 1.94 0.93 3.44 1.67 0.66 Ag (g/t) 1.3 1.0 <1 <1 <1 1.0 2.0 1.2 1.9 1.0 1.0 4.8 <1 3.0 5.0 25.3 18.3 1.0 <1 1.0 1.0 <1 <1 <1 2.3 3.5 2.4 1.3 1.5 1.7 2.4 5.7 2.6 1.5 1.0 1.3 1.5 1.0 1.7 1.0 4.0 1.0 <1 1.0 1.0 <1 1.0 1.0 1.0 <1 1.3 <1 1.0 1.0 1.0 <1 1.3 1.0 <1 <1 <1 1.0 2.0 1.2 1.9 1.0

53

Figure 11.2 Underground Sample Locations Cetate-Gauri Area (Level +885 m RL)

54

Figure 11.3 Underground Sample Locations Cetate-Gauri Area (Level +855 m RL)

55

Figure 11.4 Underground Sample Locations Cetate-Gauri Area (Level +827 m RL)

56

Figure 11.5 Underground Sample Locations Cetate-Gauri Area (Level +824 m RL)

57

Figure 11.6 Underground Sample Locations Cetate-Gauri Area (Level +724 m RL)

58

Figure 11.7 Underground Sample Locations Carnicel Area (Level +884 m RL)

59

12.0 SAMPLING METHOD AND APPROACH

12.1 INTRODUCTION Prior to 1998, all exploration data collection at Rosia Montana was undertaken by Romanian state companies. From 1998 to 2004, exploration was undertaken under the management of RSG Global in close consultation with RMGC field staff and management. Exploration in and post 2004 was undertaken by RMGC (due diligence sampling procedures used in 2004 are based on RMGC sampling procedures and are included in Appendix 2). 12.2 CHANNEL SAMPLING

Channel sampling has been completed on 1 m intervals under rigorously controlled and supervised conditions from all the accessible and safe drives within the Rosia Montana deposit, and from surface channels and pits. A total of 74,030 metres of channel sampling has been completed. Figure 12.1 shows the collars of channel samples used in the resource estimation studies documented in this report, with Figures 11.2 to 11.7 showing the channel sampling completed in the period of 2006 to 2008 which have not been used in resource estimation studies. The widths and depths of each channel were measured and samples were routinely weighed prior to final bagging in order to maintain an even sample size and to avoid sampling bias in harder rock types, and an average channel sample weight was maintained at 3.7 kg. As shown in Figure 12.2, the channel sampling procedures are summarized as: All surfaces to be channel sampled were cleaned of all scale prior to sampling. The channel sampling line and channel interval (1 m) for each sample were marked up, using spray paint, by geologists prior to the initiation of sampling. The drive names, start point for the channel run, metreage along the drive and sample numbers were also spray painted on each drive. The establishment of a standard channel width and depth (5 cm high, by 3 cm deep). A sample was not accepted unless the channel was checked and approved by the supervising (RSG and/or RMGC) geologists. A channel width and depth template was produced and was run along all channels prior to the sample being accepted. The routine collection of a duplicate field sampling channel, located approximately 20 to 25 cm above the standard channel run. One in 20 of all channel samples was reproduced in this manner. Channel samples were routinely weighed to ensure a constant sample weight (average weight 3.7 kg equivalent to half HQ core). Sampling teams (pairs of samplers) were kept more than 4 metres apart at all times to ensure that there was no chance of cross-contamination of samples. Extensive statistical analysis has been carried out on the channel sampling field duplicates, laboratory duplicates etc. (as is also the standard procedure for RC and diamond drilling).

60

Figure 12.1 Surface (red) and Underground (blue) Channel Sampling Locations

61

Figure 12.2
UNDERGROUND CHANNEL SAMPLING PROCESS

GEOLOGIST

* Identify the sampling level and accessibility * Identify dangerous zones * establish optimum density of sampling

Verify the quality of sampling process Verify and supervise miner's work

MINER

* Establish access from surface to sampling level * Open the galleries in collapse zones * Reinforce the collapse and dangerous zones

Geological logging of samples * Lithology * Mineralisation * Alteration

Record geological information

DATABASE

* Structural logging * Describe the veins, veinlets, faults and geological contacts Assay Results

TECHNICIAN

Identify the sampling line in drive or room and pillar Distribute safety and sampling equipment

* Mark the number of sampling line * Trace with paint the sampling line at the middle of the wall * Mark every meter within the sampling line

* Collect the samples * Make submission sheet for laboratory

* Mark duplicate sample every twenty meters * Organise samplers and give them the numbered calico bags for collecting samples

* Supervise the sampling to ensure quality * Verify dimension and weight of samples * Verify the cleaning of tarpaulins for sampling

Send the samples to laboratory

LABORATORY

SAMPLER

Clean the wall within the sampling interval

Collect sample with the chisel and hammer

The standard dimensions of sample are 3 x 5 x 100cm The sampling is continuous The material is collected on the sampler's tarpaulins The sampler's tarpaulins are cleaned after every sample The sample is put in the numbered calico bags The minimum weight of sample is 2.5Kg

Channel samples have been routinely weighed as part of the channel sampling process. All material chiselled out of the channel is collected for assay. As such there is no percentage recovery, however, the sample weights have been statistically assessed. Table 12.1 summarizes the results the channel sample weight data assessment for those data used in the resource estimate (collected before 2006), subdivided into the following groupings: Cirnicel underground channel sampling. Cirnic underground channel sampling. Cetate underground channel sampling. Cirnic surface channel sampling. Northern area surface channel sampling.

62

Table 12.1 Rosia Montana Gold Project Channel Sample Weight Assessment Group Cirnicel UG Cirnic UG Cetate UG Cirnic Surf North Surf No. 4,132 24,061 13,349 1,138 2,923 Minimum 3.0 2.2 2.1 2.2 2.2 Maximum 5.4 6.4 4.9 4.7 4.5 Mean 3.9 3.7 3.6 3.4 3.4 Median 3.8 3.8 3.6 3.4 3.4 CV 0.073 0.086 0.094 0.082 0.077

Frequency histograms of the channel sample weight data are provided in Appendix 3. It is evident from the very low coefficients of variation (standard deviation divided by the mean) and the frequency histograms that the sample weights are grouped very closely around the mean. As such, a consistent sample weight has been collected. A median weight of 3.7 kg equates very closely to half HQ core. The sampling weights for the 2006 to 2008 underground channel sampling programs are consistent with those presented above. The correlation between channel sample weights and gold and silver grades has been investigated. As shown in Table 12.2 and the correlation plots contained in Appendix 3, there is no correlation between gold and silver grades and the weight of the channel samples.
Table 12.2 Rosia Montana Gold Project Summary of Linear Correlation between Recovery and Gold/Silver Grades Group Cirnicel UG Cirnic UG Cetate UG Cirnic Surf North Surf Au -0.02 0.00 0.01 0.06 0.00 Ag 0.03 -0.01 -0.02 0.07 0.02

While there is variation in the sample weights for the range of sample interval lengths, the sample weights generally increase with increasing interval lengths consistent with 3 kg samples being collected over 1 m intervals. In addition, there is no evidence of any bias in the sample gold grades relating to the variations in the sample weights (Figure 12.3). 12.3 RC DRILLING

The drilling at Rosia Montana consists of a combination of surface RC and diamond drilling, and underground diamond drilling. RC samples were routinely collected at 1 m intervals and the cuttings split with a Jones riffle splitter. Field duplicates were taken via the splitter every 20 samples. The bags of cuttings were routinely weighed prior to taking the sub-sample. The RC sampling process as applied to the Rosia Montana drilling is summarized in the flowsheet presented as Figure 12.4.

63

Figure 12.3 Channel Sample Gold Assay Grades versus Sample Weights

1000 900 800 700 Gold Grade (g/t) 600 500 400 300 200 100 0 0 1 2 3 4 5 6 7 Sample Weight (kg)

Figure 12.4
RC SAMPLING PROCESS

Survey and setup drillsite GEOLOGIST Ensure drillhole is at correct azimuth and dip Verify and supervise drillers work

Verify the quality of sampling process Record geological information in Field Marshall Logger

Geological logging of samples * Lithology * Mineralisation * Alteration

DATABASE

DRILLER

Try to keep sample dry and of consistent size

Collect samples from cyclone every metre, clean cyclone regularly Make submission sheet and send samples to laboratory daily

Assay Results

FIELD ASSISTANT

Number plastic bags with hole ID and meterage

Number calico bags, including duplicate samples every twenty metres

Weigh every sample then put through the riffle splitter and collect in calico bag

LABORATORY

64

Stringent quality control procedures have been used for all RC drilling completed at Rosia Montana to produce high quality samples. The quality control procedures summarized include: The use of face sampling hammers. Routine blowbacks (i.e. each metre is cleared into the sample bag prior to moving on to the next metre of advance). Routine weighing of collected RC cuttings (see below for analysis). The cyclone is cleaned at each rod change. On a rod change, any material in the hole is cleared before the first new metre sample is collected. The riffle splitter is cleaned with compressed air between each sample (using compressed air guns attached to the drilling rig). Every 20 metres a duplicate field sample is collected by riffle splitting the bag of cuttings once more. With the exception of Igre (discussed later in the text), on the extremely rare occasions that moisture was encountered in the hole (please note the emphasis on moisture rather than wet drilling conditions), the hole was conditioned or dried prior to advancing further. Additional compressed air boosters were routinely used.

Sample weights have been routinely measured on a metre by metre basis as part of the standard RC drilling procedures. The resultant data have been statistically analyzed. Frequency histograms of the RC sample weight data are contained in Appendix 3. Theoretical percentage recoveries have been based on calculating the volume of a hole produced using a 5.25 inch hammer (2000 - 2004 drilling programs) or a 4.5 inch hammer (1998 and 1999 drilling programs), and using a density of 2.4 tonnes per cubic metre (deposit average) to determine the theoretical sample weight. For a 5.25 inch hammer, a metre sample should weigh 33 kg, whilst for a 4.5 inch hammer the theoretical sample weight is 25 kg. Table 12.3 summarizes the result of the statistical analysis and the calculated theoretical percentage sample recovery for the RC drilling program. High sample recoveries have been determined, due to the routine use of blow-backs during the course of drilling. The data have been subdivided as follows: Jig-Igre. Orlea. Cirnic. Cetate (2000 program hole numbers designated with CTSD prefix). Cetate (1998 and 1999 programs hole numbers designated with an RMRD prefix). 2002 Cirnic RC drilling. 2002 Cetate RC drilling. 65

2002 Geotechnical RC pre-collar drilling. 2002 RC sterilization drilling. 2003 Geotechnical RC pre-collar drilling. 2003 RC sterilization drilling. 2004 RC drilling.
Table 12.3 Rosia Montana Gold Project Theoretical Sample Recoveries RC Drilling Group Jig-Igre * Orlea * Cirnic * Cetate (CTSD) * Cetate (RMRD) * 2002 Cirnic 2002 Cetate 2002 Grade Control 2002 Geotech 2002 Sterilization 2003 RC 2004 RC Note: * pre 2002 drilling No. 6,299 3,103 3,941 12,644 10,034 6,512 10,597 3,983 108 3,051 7,722 5,276 Minimum 6 6 6 6 6 1 0 1 1 0 0 0 Maximum 56 48 56 53 42 46 50 42 40 42 57 50 Mean 30 27 27 28 18 26 24 25 24 25 27 30 Median 31 29 27 29 18 28 26 27 25 27 30 31 CV 0.242 0.255 0.268 0.233 0.239 0.258 0.281 0.286 0.348 0.306 0.319 0.227 % Recov. 94 88 82 88 73 84 78 82 76 81 83 96

The correlation between the calculated theoretical RC sample recovery and gold and silver grades has been investigated. As shown in Table 12.4 and the correlation plots contained in Appendix 3, there is no correlation between gold and silver grades and the measured theoretical RC sample recovery. A review of the wet RC drilling was completed using the resource definition drill holes up to and including the 2003 program. A number of drill hole intervals were logged to have been wet and also associated with significant anomalous gold mineralization. It is important to note that the wet drilling is limited only to a small portion of the Igre region.
Table 12.4 Rosia Montana Gold Project Theoretical RC Sample Recovery versus Gold and Silver Grades Group Jig-Igre Orlea Cirnic Cetate (CTSD) Cetate (RMRD) 2002 Cirnic 2002 Cetate 2002 Grade Control 2002 Geotech 2002 Sterilisation 2003 RC 2004 RC Au -0.03 -0.04 0.00 -0.01 -0.03 0.01 -0.03 -0.04 0.04 -0.06 -0.03 -0.02 Ag -0.06 -0.10 -0.02 -0.06 -0.05 0.07 -0.04 -0.01 -0.08 -0.08 -0.05 0.02

Statistical review of 2 m composites was completed for the Igre mineralization (2004 estimate) to compare the wet drilling intervals against the dry and moist intervals. The intervals logged as having a 66

water flow are noted to be significantly higher grade than the moist and dry data. Note these data include the diamond drilling as dry. Based on the preliminary review RSG Global requested that the identified problematic drill holes be assessed on a hole-by-hole basis prior to a decision being made to remove all suspected contaminated drill intervals from the exploration database prior to grade estimation. The intervals removed are summarized below in Table 12.5.
Table 12.5 Rosia Montana Gold Project Igre Region RC Intervals Removed Due to Wet Drilling Hole Id JVSD076 JVSD083 From (m) 64 57 To (m) 88 79 Length 24 22 Uncut Gold Grade (Au g/t) 4.41 3.93 Uncut Silver Grade (Ag g/t) 4 6

Note: Reported intervals in interpreted mineralization zones only while the entire interval post the sample from have been removed from the database.

Notwithstanding the excluded intervals, RSG Global believes that RC drill samples have been collected in a robust manner and as such are appropriate for resource estimation studies. Figure 12.5 displays a scatter plot of the RC sample weights versus gold grade. There is no evidence that anomalously low or high sample weights are associated with high gold grades. 12.4 DIAMOND DRILLING

The diamond drilling sampling process, as applied at Rosia Montana, is summarized in the flowsheet presented as Figure 12.6.
Figure 12.5 RC Sample Weight versus Gold Grade 5 m Composites
50 45 40 35 Gold Grade (g/t) 30 25 20 15 10 5 0 0 10 20 30 40 50 RC Sample Weight (kg)

67

Figure 12.6
DIAMOND CORE SAMPLING PROCESS

Survey and setup drillsite GEOLOGIST Ensure drillhole is at correct azimuth and dip Verify and supervise drillers work

Verify core recovery, orientate core and mark top of core for field assistants to cut along, log core onto paper

DATABASE

DRILLER

Maximise sample recovery, orientate core survey hole at regular intervals

Ensure correct handling and laying out of core with depth markers at end of each run, mark hole ID, sample intervals and tray number on core trays

Assay Results

FIELD ASSISTANT

Provide drillers with core trays, marker blocks and marker pens

Collect sample trays daily and take to core yard

Photograph every core tray then cut core on metre intervals

Make submission sheet and send samples to laboratory daily

LABORATORY

To ensure a high sample quality, stringent data collection quality control procedures have been applied. The standard quality control procedures for the diamond drilling programs include: Use of triple tube sample collection in areas of non-optimal ground conditions. Use of sub-three metre core runs in areas of non-optimal ground conditions. Use of specialized drilling mud in areas of non-optimal ground conditions. Recording of core recovery (see below for analysis). Photography of all core prior to sampling.

The diamond core was marked off at 1 m intervals and sampled to produce half-core (lengthways) using a diamond core saw. Field duplicates were produced from the same half-core following jaw crushing. Drill core recoveries were calculated by comparing the measured length of recovered core with the distance recorded on the core blocks between each drill run. Core recoveries for the resource delineation surface and underground drilling programs are noted to be consistently in excess of 95%. The geotechnical drilling carried out during 2000-2001, and which was subsequently assayed (1,409 samples), returned an average core recovery of 86% (median 100%). Table 12.6 and the frequency histograms in Appendix 3 demonstrate that very high core recoveries have been achieved. The data have been grouped into the following categories: Cetate underground diamond drilling. Cirnic underground diamond drilling. Cetate surface diamond drilling. 68

Cirnic surface diamond drilling. Jig-Igre surface diamond drilling. Orlea surface diamond drilling. 2002 geotechnical drilling (which was subsequently assayed).

Table 12.6 Rosia Montana Gold Project Summary Calculated Diamond Drilling Recoveries (%) Group Cetate UG Cirnic UG Cetate Surf Cirnic Surf Jig Surf Orlea Surf 2002 Geotech No. 5475 5256 8321 3992 1842 1095 1409 Minimum 7 4 4 4 6 10 0 Maximum 100 100 125 110 100 100 103 Mean 98.5 95.3 97.2 96.6 93.3 98.5 86.3 Median 100.0 100.0 100.0 100.0 100.0 100.0 100.0 CV 0.080 0.143 0.108 0.115 0.194 0.086 0.269

Figure 12.7 displays a scatter plot of percentage core recovery versus gold grade for the diamond drilling at Rosia Montana. There is no evidence that low core recoveries are associated with high gold grades.
Figure 12.7 DDH Core Recovery versus Gold Grade

100 90 80 70 Gold Grade (g/t) 60 50 40 30 20 10 0 0 20 40 60 80 100 DDH Core Recovery (%)

69

13.0 SAMPLE PREPARATION, ANALYSIS AND SECURITY

Samples were prepared and assayed at an on-site, custom-built assay facility managed by SGS Ltd. (Formerly Analabs Pty. Ltd.), an internationally accredited assay laboratory group. RC and channel samples were dried and the whole sample pulverized to 85% 75 m in a LM5 disc pulverizer. Core samples were crushed in a jaw crusher prior to the pulverizing stage. A 50 g split was submitted for assay for gold by fire assay followed by an AAS finish, and silver was analyzed by multi-acid digest followed by an AAS finish. The fire assay procedure flowsheet is provided as Figure 13.1. Note that during 2002 the sample drying was completed using a temperature of 80C as the samples were then analyzed as composites for Hg. The precision and accuracy of the assay data were assessed by the use of field duplicates, duplicates and internal laboratory standards. In addition, internationally accredited gold-silver standards were independently submitted into the assay stream at the rate of 5% of all samples, representing some 1,100 data for the 2000 exploration program alone. Both the independently submitted quality control samples and the internal laboratory quality control data are assessed in detail in this report (Section 14.3). The data applied to estimate the mineral resource is sourced from exploration programs executed between 1998 and 2005. These exploration programs were undertaken under the management of RSG Global in close consultation with RMGC field staff and management. Exploration in 2004 was undertaken by RMGC. Independent sampling programs are described in Section 14 of this report. The sulphur sampling has been completed in four phases. The first dataset was collected as part of the 2000 feasibility study and analyzed for sulphur by SGS, Perth using a Leco furnace. The 2001 feasibility study data comprise three types of material: Sub-samples of the 30 kg variability samples, collected as part of the metallurgical sampling program. These data have been collected both from underground and from the then Cetate pit surface. Quarter core from selected core intervals. Five metre composites, produced by combining pulps from adjacent one metre sample intervals, obtained from the sample pulp storage system. The 5 m composites were spaced at 50 m intervals down surface and underground drill holes, excluding the first 5 m interval (0 m to 5 m).

The second suite of sulphide and sulphate samples was collected post 2000, initially focused within the seven-year pit, as defined by pit optimization studies completed by SNC, but subsequently expanded to regions outside the 7 year pit. The data collection methodology, as provided by RMGC, is summarized as: Collection of a downhole composite made from five 1 m interval samples collected from the 2002 infill drilling program pulps. The final sample was approximately 25 g of pulp comprised of 5 g from each 1 m sample interval. The collected composites were homogenized by shaking.

70

The composites samples were dispatched to ALS-Chemex Laboratories in Vancouver and analyzed for sulphide sulphur, sulphate sulphur and total sulphur by Leco furnace. One in ten samples was analyzed for 47 elements using ICP-MS/ICP-AES and Hg via cold vapour AAS.
Figure 13.1

CONVENTIONAL FIRE ASSAY PROCEDURE

Underground and Surface samples

1 metre channel samples, 1/20th re-sampled parallel to original sample

RC Samples
1 meter sample interval Riffle split on site. 1/20th sample re-sampled.

Drill Core Samples

1 meter sample interval Core sawn in half. One half returned to tray, other half assayed

LABORATORY
Samples are racked on trolleys in order by following the submission sheet. Sample numbers entered in computer. CCLAS program generates paperwork. It randomly creates replicates (10%) and second splits (~10%) of the samples for quality control. A replicate is a sample taken from the LM5 bowl and assayed twice. A second split is a same sample taken again from the LM5 bowl, placed in a separate bag and assayed.

OVEN
All samples received are dried at 110-120C

JAW CRUSHER
Underground, Surface and Core samples are crushed to -6mm. 1/20th sample is split to produce B Sample.

LM5 DISC PULVERISER

All samples are pulverised for minimum of 5 minutes. 95% passing through a 75 micron screen. A 300-gram scoop is taken from the bowl and placed in a labeled paper bag (pulp). The rest of the sample is placed back in its original bag (residue). Sample pulps are kept for storage, while residue bags are discarded after the final report, or used for screen fire assay.

FIRE ASSAY
50gm charge from the pulp bag. A normal fire has a batch of 50 assays. It consists of 40 samples, 4 replicates, 3 second splits, 2 standards, and a blank. Samples mixed with flux (Lead oxide). Fusion at 1 100C. Cupellation at 1000C. Prill (Ag + Au) placed in test tube, aqua regia digested, made to volume, and analysed by AAS.

BASE METAL ANALYSIS (Ag, Cu, Pb, Zn)

0.4g from pulp bag weighed, placed in test tube, aqua regia digested, made to volume, and analysed by AAS.

QUALITY CONTROL
Every month a batch of 52 samples is sent to Analabs Australia and Bondar Clegg Canada for quality control checks (Au, Ag). A final statistics report on these checks is produced by Analabs headquarters, together with a monthly report on all database (replicates, second splits, standards, blanks) generated by the laboratory.

(Every month 108 samples are sent to SGS and ALS-Chemex for third party checks)

71

The third phase of sampling was performed as part of routine exploration completed post the 2002 study. Total sulphur data only were collected with the distribution of sulphur data presented as Figure 13.2. The fourth phase consisted of resampling and compositing of historical channel sampling pulps from the RMGC sample pulp storage facility to cover those parts of the deposit which had insufficient sampling during late 2004 and early 2005. Selected samples were composited as per the above from consecutive five 1 m pulps. The final sample was approximately 25 g of pulp comprised of 5 g from each 1 m sample interval. The collected composites were homogenized by shaking. The composites samples were dispatched to ALS-Chemex Laboratories in Vancouver and analyzed for total sulphur by Leco furnace. Approximately 3,300 samples composites were analyzed this way and added to the database.
Figure 13.2 Distribution of Total Sulphur Data

72

14.0 DATA VERIFICATION

14.1 LOGGING All surface and underground channel samples, RC drilling and diamond core have been geologically logged using the RMGC logging scheme shown in Appendix 1. Geological logging has been carried out mostly on a one metre basis, with particular attention to oxidation type, rock type, tectonic/structural fabrics, veining/intensity, alteration/intensity, sulphides/intensity, and moisture content. In addition, the occurrences of voids and/or insufficient samples have been recorded. Detailed geological drawings of all channel sampled trenches and road cut exposures have also been generated. Geological logging of core has mostly been conducted over intervals equal to the sampling interval (generally one metre), with the exception of the first drilling program, when logging intervals were chosen on a geological basis, and the 2004 due diligence drilling program. All diamond core has been geotechnically logged, including recording RQD, recovery per drill run, and number of fractures per metre. Core recoveries have been calculated by comparing the measured length of recovered core with the distance recorded on the core blocks between each drill run. Detailed structure orientation logging was completed for all orientated intervals of core including recording of structure types and associated alpha/beta measurements. Dip and dip direction measurements were also collected for structures exposed by trenching. The great majority of logging information has been collected digitally on palm top Hewlett Packard IPAQ computers using Field Marshall software. All core since inception of exploration by RMGC has been photographed, both wet and dry, using a digital camera. RSG Global is satisfied that the geological, geotechnical and structure orientation logging of the drilling and channels completed at Rosia Montana has been conducted to high industry standards. 14.2 QUALITY CONTROL PROCEDURES

Over 1,500 sample pulps were sent to both SGS (Analabs) in Perth, Australia and ALS Chemex (Formerly Bondar Clegg) Laboratories in Vancouver, Canada for independent verification of laboratory bias, during the 1998 to 2000 resource delineation phase. Post 2000, regular batches of samples (104 samples) have been sent each month to SGS and ALS Chemex as part of the routine QAQC process. A summary of pertinent quality control statistics for this phase is as follows: Repeat assays Laboratory splits and duplicate samples SGS internal standards and blanks 22% of all primary assays. 10% of all primary assays. 7% of all primary assays.

73

A summary of pertinent quality control statistics for the 2002-2003 drilling programs is as follows: 14.3 14.3.1 Repeat assays Laboratory splits and duplicate samples SGS internal standards and blanks 20% of all primary assays. 13% of all primary assays. 5% of all primary assays.

And blanks (commencing after the main Rosia drill out resulting in few samples). STATISTICAL ANALYSIS OF ASSAY QUALITY CONTROL DATA Introduction

RSG Global, as part of the 2001 feasibility study and the 2003 resource estimate, completed a detailed review of all quality control data relevant to the gold and silver assays for data included in those studies. All results from the investigations are reported in the feasibility documentation and previous report and will not be repeated in this report. The statistical analysis of the QAQC data revealed that no bias existed. Levels of precision consistent with gold-silver deposits of this style were noted. The assay quality control data pertaining to samples collected subsequent to the 2003 estimate were assessed as part of the resource estimate discussed herein with the results summarized below. Comparative data analysis was undertaken to determine the precision of the gold and silver assay data in the Rosia Montana exploration database, to evaluate the use of fire assays as the primary analytical method, and to determine the inter-laboratory analytical precision between the SGS, Gura Rosiei laboratory and the two internationally accredited laboratory facilities used. Quality control data for the 2006 to 2008 exploration periods have been assessed but are not reported in this document as these data have not been included in the resource estimation study and the assay data are incomplete with a number of assays still outstanding. Based on the available information similar levels of precision and accuracy are noted for these data as for previous exploration programs. 14.3.2 Assay Accuracy

The accuracy of the gold and silver assay data at the Rosia Montana project and the potential for crosscontamination of samples during sample preparation were assessed based on the assay results for the laboratory internal blanks and standards analyzed by the SGS, Gura Rosiei facility, and the RSG Global submitted blanks and standards. The statistical analysis was undertaken based on the assay data for the following sample suites: Blanks and gold standards used by the SGS, Gura Rosiei facility for internal laboratory quality control (generally 2 standards in use at any one time). Eight differing gold standards inserted into the exploration sample stream by RSG Global for independent monitoring of gold assay accuracy (four standards in use at any one time). Blanks and two differing silver standards used by the SGS, Gura Rosiei facility for internal laboratory quality control.

74

Four differing silver standards inserted into the exploration sample stream by RSG Global for independent monitoring of silver assay accuracy.

Descriptive and graphical statistics were compiled for both gold and silver standards (and blanks), with a maximum of 500 assays incorporated into each data group. The following graphical methods were used: Standard Control Plot: Standard control plots allow the investigation of analytical drift, accuracy and bias on the basis of assays undertaken on laboratory standard media. CUSUM Plot: CUSUM plots display the aggregated difference between the assays of a standard and the expected value or the mean of the assays. The plots are particularly useful for enhancing trends or drift in assay accuracy over time.

Descriptive statistics and graphical plots of the gold fire assay data for the Gura Rosiei internal laboratory blanks and standards are provided in Appendix 4. Summary comments on the results of the statistical analysis are provided below: Industry acceptable assay results have been reported for the blanks. Most of the anomalous statistical data appear to result from the mislabelling of standards, thus resulting in the inclusion of assay data for incorrect standards with the data for other standards. The 16% positive bias in the assay data for the standard QFA5 in the period from 6 February, 1999 to 5 May, 1999 results from two very high-grade assays. Otherwise, the remaining assay results for QFA5 reflect acceptable analytical accuracy. With the exception of the anomalous assay data noted above, industry acceptable levels of analytical accuracy are evident for the great majority of gold assay data for the SGS, Gura Rosiei gold standards.

Descriptive statistics and graphical plots of the aqua regia / AAS silver assays of the SGS, Gura Rosiei internal laboratory blanks and standards are provided in Appendix 4. Summary comments on the analysis are provided below: Industry acceptable assay results have been reported for the blanks. Significant percentages of the silver assay data report values outside +/-10% of the expected value for the standard GBM997_1. It is noted, however, that if consecutive assays are averaged in nominal size groups, there is little evidence for significant analytical drift outside the +/- 10% control lines. The positive 47.2% bias reported for the standard GBM997_1 in the period from 5 August, 1999 to 25 November, 1999 results from the inclusion of a single 600 g/t Ag assay in the dataset. If this assay is ignored, the remaining data demonstrate an acceptable level of analytical accuracy. Overall, the SGS, Gura Rosiei silver assays for the internal laboratory standards demonstrate an acceptable analytical accuracy in relation to industry standards when also considering that the umpire assaying indicates that no significant bias exists between umpire assay datasets. 75

Detailed discussion of the standards data, including subdivision by time periods, is provided in previous resource estimation studies completed in 2000, 2002 and 2003 and will not be repeated in this document. Summary statistics for the independent gold and silver standards submitted by RSG Global are presented in Tables 14.1 and 14.2. 14.3.3 Assay Precision

The precision of the gold and silver assay data for the Rosia Montana project was statistically assessed based on the following comparative sample/data types: Duplicate RC and channel samples collected in the field - allows assessment of overall precision, reflecting total combined sampling and analytical errors (field and laboratory). Duplicate splits of core and channel samples after jaw crushing - allows assessment of overall laboratory precision, reflecting sample preparation and analytical errors at the laboratory. Duplicate pulp splits of samples after pulverization - allows assessment of laboratory precision inclusive of sampling and analytical errors after sample pulverization. Repeat analyses of laboratory pulp samples - allows assessment of laboratory analytical precision (exclusive of dominant sampling errors). Primary versus umpire laboratory analyses of duplicate pulp splits - allows assessment of interlaboratory precision inclusive of sampling and analytical errors after sample pulverization.

The order of the comparative data types listed above reflects the successive removal of sampling error thus allowing the precision associated with each stage in the sampling process (field and laboratory), and finally the sample analyses, to be assessed. Details of the available datasets and results of the statistical analyses are summarized below for the Gura Rosiei facility, while a full compilation of statistical plots of the comparative datasets is provided in Appendix 4. Statistical analysis of the gold datasets has considered only the assay data greater than or equal to 10 times the SGS analytical detection limit (i.e. data at or above 0.1 g/t Au). Similarly, a lower selection threshold of 3 times the SGS analytical detection for silver (i.e. data at or above 3 g/t Ag) has been used for statistical analysis of the silver assay datasets.

76

Table 14.1 Rosia Montana Gold Project Statistical Summary SGS Gura Rosiei Laboratory Gold Assays of Standards Standard Name Expected Value (EV) 0.01 0.76 0.87 1.06 3.46 0.53 0.819 0.939 1.315 1.535 1.887 2.22 2.643 4.823 4.997 13.89 +/-10% (EV) (g/t) 0.009 to 0.011 0.68 to 0.84 0.78 to 0.96 0.95 to 1.17 3.11 to 3.81 0.477 to 0.583 0.737 to 0.901 0.845 to 1.033 1.184 to 1.447 1.381 to 1.688 1.698 to 2.076 1.998 to 2.442 2.379 to 2.907 4.341 to 5.305 4.497 to 5.497 12.501 to 15.279 Period of Analyses No of Analyses 139 246 266 237 265 111 334 1022 293 191 1445 683 323 631 255 766 Minimum (g/t) 0.01 0.65 0.72 0.93 3.17 0.51 0.72 0.8 1.14 1.37 1.68 1.52 2.38 4.24 4.46 11.84 Maximum (g/t) 0.03 0.88 1.2 1.17 3.6 0.6 0.92 1.18 1.43 1.65 2.19 2.53 2.84 5.22 5.21 15.84 Mean (g/t) 0.011 0.74 0.86 1.04 3.41 0.549 0.822 0.957 1.319 1.504 1.9 2.156 2.593 4.786 4.869 13.93 % Within +/- 10 of EV 89.928 97.37 87.97 99.16 100 99.09 99.401 91.096 99.647 98.953 97.024 92.972 100 98.891 99.608 95.822 3.59 6.62 2.84 1.96 2.848 3.362 5.451 2.765 2.531 4.073 5.296 2.52 3.158 2.694 4.472 -2.38 -1.7 -2.29 -1.41 3.519 -2.146 1.864 0.292 -1.972 0.684 -2.87 -1.878 -0.759 -2.557 0.306 % RSD (from EV) % Bias (from EV)

RSG Global Submitted Blanks Blank st17/8171 st39/6167 st06/8222 st10/5127 SE35 SF12 S3 SH13 S2 S8 S4 SJ10 SK11 S5 S6 Sept 2003 to June 2005 April 2000 - Sept 2000 April 2000 - Sept 2000 April 2000 - Sept 2000 April 2000 - Sept 2000 April 2004 - June 2005 June 2003 - June 2005 Sept 2000 to July 2004 June 2003 - June 2005 Sept 2000 to Oct 2000 Nov 2000 to June 2005 Sept 2000 to Nov 2002 June 2003 - June 2005 March 2002 to June 2005 Sept 2000 to Sept 2001 Sept 2001 to June 2005 RSG Global Submitted Gannet Standards

RSG Global Submitted Rocklabs Standards

77

Note: RSD = Relative Standard Deviation Table 14.2 Rosia Montana Gold Project Statistical Summary SGS Gura Rosiei Laboratory Silver Assays of Standards Standard Name Expected Value (EV) 1 4 14.77 15.23 15.78 16.57 16.952 17.08 +/-10% (EV) (g/t) 0.9 to 1.1 3.6 to 4.4 13.293 to 16.247 13.71 to 16.75 14.2 to 17.36 14.91 to 18.23 15.257 to 18.647 15.37 to 18.79 Period of Analyses No of Analyses 11 108 1474 687 259 1021 189 779 Minimum (g/t) 1 3 11 12 13 12 14 12 Maximum (g/t) 2 5 17 17 17 18 17 19 Mean (g/t) 1.18 3.9 14.5 14.33 14.61 15.24 15.58 16.3 95 85 85 52 77 54 78 4.5 6.7 6.1 5.3 6.4 5.5 6.6 -0.7 -2.1 -5.9 -7.4 -8.1 -8.1 -4.9 % Within +/- 10 of EV % RSD (from EV) % Bias (from EV)

RSG Global Submitted Blanks Blank SE35 S8 S4 S5 S3 S2 S6 Sept 2003 to June 2005 April 2004 - June 2005 Nov 2000 to June 2005 Sept 2000 to Nov 2002 Sept 2000 to Sept 2001 Sept 2000 to July 2004 Sept 2000 to Oct 2000 August 2001 to June 2005 RSG Global Submitted Rocklabs Standards

Note: RSD = Relative Standard Deviation

Descriptive and graphical statistics were compiled on the basis of each comparative dataset. Graphical methods employed were: Thompson and Howarth Plot showing the mean relative percentage error of grouped assay pairs across the entire grade range, used to visualize precision levels by comparing against given control lines. Rank % HARD Plot, which ranks all assay pairs in terms of precision levels measured as half of the absolute relative difference from the mean of the assay pairs (% HARD), used to visualize relative precision levels and to determine the percentage of the assay pairs population occurring at a certain precision level. Mean vs % HARD Plot, used as another way of illustrating relative precision levels by showing the range of % HARD over the grade range. Mean vs % HRD Plot is similar to the above, but the sign is retained, thus allowing negative or positive differences to be computed. This plot gives an overall impression of precision and also shows whether or not there is significant bias between the assay pairs by illustrating the mean percent half relative difference between the assay pairs (mean % HRD). Correlation Plot is a simple plot of the value of assay 1 against assay 2. This plot allows an overall visualization of precision and bias over selected grade ranges. Correlation coefficients are also used. Quantile-Quantile (Q-Q) Plot is a means where the marginal distributions of two datasets can be compared. Similar distributions should be noted if the data are unbiased. Standard Control Plot shows the assay results of a particular reference standard over time. The results can be compared to the expected value, and the 10% precision lines are also plotted, providing a good indication of both precision and accuracy over time.

Precision of Gold Fire Assay Data - Gura Rosiei Facility The precision of the gold fire assay data for the period up to and including 2002 was investigated, grouped according to project sub-region for the 2001 feasibility study. Descriptive statistics and graphical plots of the data comparisons are provided in Appendix 4 for post feasibility data and summary details on the data precision are provided in Table 14.3. The quality control data collected for the 2002 and (pre-June) 2003 period were investigated by sample type with the result presented as Tables 14.4 and 14.5 respectively.

78

Table 14.3 Rosia Montana Gold Project Summary of Data Precision 2001 Feasibility Study SGS Gura Rosiei Gold Fire Assay Data >0.15 g/t Au Number of Data Pairs Data Comparison Channels RC Chips 121 195 603 16 72 307 502 1796 121 221 683 100 170 475 57 131 441 10 15 266 409 1440 38 87 272 51 85 260 29 47 127 DDH Core 224 373 952 Mean %Hard Channels Cirnic Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats 1251 1277 2190 7362 199 201 337 866 539 774 1297 4197 44 50 91 242 141 140 226 559 78 80 136 376 14 22.7 9.4 9.5 6.9 Cirnicel 16.7 8.4 7.5 6.7 Cetate 19.8 8.4 5.7 5.9 Carpeni 25.8 10.9 9.9 10 Orlea 24.6 12.1 9.9 11.1 Jig 21.4 8.1 4.5 4.7 Cos 3.8 3.9 6.6 2.7 3.3 3 12.1 14.3 9.1 14 8.6 9.4 17 4.7 2.9 2.9 7.5 7.1 4.3 9.1 5.3 5.4 16 13.8 11.8 13.2 11.7 10.2 21.5 9.2 5.8 5.7 10.2 8.6 6.3 12.7 6.6 5.4 14.9 9.6 8.9 10.5 8.3 10 18.6 6.2 6.3 4.5 9.8 5.5 4.3 8.3 4.8 5.7 11.2 8.2 6.2 9.7 6.4 5.6 15.1 5.7 3.4 3.2 5.9 4.3 3.4 5.8 4.3 3.4 4.8 5 11.7 5.3 3.6 3.3 4.3 3.2 10 7.9 7.8 9.6 7 6.3 22.7 5.7 3.8 3.5 6.9 3.7 3.8 6.5 4.4 3.5 RC Chips DDH Core Median %Hard Channels RC Chips DDH Core

Table 14.4 Rosia Montana Gold Project Summary of Data Precision - 2002 Drilling program Gura Rosiei Gold Fire Assay Data >0.1 g/t Au Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels RC Chips 905 1561 2737 DDH Core 20 27 63 Mean %Hard Channels RC Chips 13 9 9 DDH Core 10 10 6 Mean HRD% Channels RC Chips 1.4 3.8 1.6 DDH Core -0.3 0.11 0.43

79

Table 14.5 Rosia Montana Gold Project Summary of Data Precision Pre-June 2003 Period Gura Rosiei Gold Fire Assay Data >0.1 g/t Au Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels 19 13 32 79 RC Chips 203 299 605 DDH Core 88 124 303 Mean %Hard Channels 18.7 4.5 7.2 5.7 RC Chips 15.5 11.0 10.2 DDH Core 8.9 11.6 8.6 Median %Hard Channels 17.5 3.6 3.3 2.8 RC Chips 11.1 7.1 5.8 DDH Core 4.5 7.1 5.8

The quality control data collected for the post-June 2003, 2004 and 2005 periods were investigated by sample type with the result presented as Tables 14.6, 14.7 and 14.8 respectively.
Table 14.6 Rosia Montana Gold Project Summary of Data Precision Post-June 2003 Period Gura Rosiei Gold Fire Assay Data >0.1 g/t Au Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels RC Chips 3 9 19 DDH Core 3 4 9 Mean %Hard Channels RC Chips 34.21 5.62 6.14 DDH Core 4.24 6.12 6.1 Median %Hard Channels RC Chips 30.23 4.04 3.98 DDH Core 4.35 4.01 1.41

Table 14.7 Rosia Montana Gold Project Summary of Data Precision 2004 Period Gura Rosiei Gold Fire Assay Data >0.1 g/t Au Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels 46 43 39 153 RC Chips 35 49 116 DDH Core 40 38 114 Mean %Hard Channels 27.06 7.81 7.24 7.31 RC Chips 8.96 9.82 6.89 DDH Core 3.39 5.56 3.91 Median %Hard Channels 23.35 4.08 6.45 5.56 RC Chips 5.56 5.75 5.5 DDH Core 2.36 3.21 2.4

Table 14.8 Rosia Montana Gold Project Summary of Data Precision 2005 Period Gura Rosiei Gold Fire Assay Data >0.1 g/t Au Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels 79 80 109 280 RC Chips DDH Core Mean %Hard Channels 19.93 5.12 4.52 4.9 RC Chips DDH Core Median %Hard Channels 14.81 3.46 3.23 3.36 RC Chips DDH Core -

80

The results can be summarized as follows: The precision predominantly increases with increasing assay pair means, up to approximately 1 g/t Au for pulp repeats, duplicate pulp splits, and duplicate crusher splits. Conversely, the precision decreases or remains constant with increasing assay pair means above 1 g/t Au. The precision tends to decrease or remain constant with increasing grade for duplicate field samples. The precision increases with each sample preparation step and increasing sample homogeneity, with the lowest precision (greatest error) evident for duplicate field samples and the highest precision noted for pulp repeat analyses. The highest level of analytical precision (from the 2001 results) is reported from Cetate, while the lowest precision level is reported for Orlea. The level of precision determined for the duplicate field samples, duplicate crusher and pulp splits, and pulp repeats is predominantly within industry accepted limits for gold fire assays and the styles of gold mineralization observed at the Rosia Montana project.

Statistical comparisons of the available fire assay and screen fire assay pairs, subdivided by project sub-region and primary sample type, are presented in Appendix 4. Most of the comparative screen fire assay data were collected for channel samples from the Cetate and Cirnic sub-regions. The results can be summarized as follows: The precision increases with increasing grade, with the mean grades of most of the assay pairs lying above 1 g/t Au. The overall precision for datasets containing more than 40 assay pairs ranges from approximately +/-13% to +/-19% Mean % HARD. After removing obvious outlier assay pairs from the datasets, the initial fire assays are overall 8% to 9% higher grade than the screen fire assays for both Cetate and Cirnic. Overall, the fire assays and screen fire assays are in close agreement for Cirnicel (107 assay pairs) and Jig (43 assay pairs).

Insufficient assay pairs have been collected for the Carpeni and Orlea regions to allow valid statistical comparisons for these sub-regions. Comparison of Round Robin Gold Fire Assay Analyses The inter-laboratory analytical precision of the gold fire assays from the SGS, Gura Rosiei facility, SGS (Analabs-Balcatta) and ALS-Canada (previously Bondar Clegg) was assessed with the assay data grouped according to those samples submitted over approximately 3 month intervals. Descriptive statistics and graphical plots of the data are provided in Appendix 4. Summary details on the data precision and relative accuracy are presented in Table 14.9 for the 2001 feasibility study.

81

Table 14.9 Rosia Montana Gold Project Summary of Precision and Relative Bias - 2001 Round Robin Gold Fire Assays >0.15 g/t Au Data Comparison Number of Data Pairs %HARD Mean Median 5.6 5.7 6.7 4.7 4.6 5.2 4.2 4.1 3.2 7.3 4.1 7.5 6.6 5.6 5.9 6.8 5.7 6.8 7.3 8.7 9.5 Mean 0.2 -0.4 -0.5 0.6 1.7 1.1 1.4 0.8 -0.4 5 1.6 -3.8 3.6 1.1 -2.3 2.2 -0.4 -2.6 0.4 0.9 0.8 %HRD 95% CL 2.4 2.2 2.5 1.9 2.1 2.2 1.6 1.5 1.7 2.8 3 3.5 2.2 2 1.9 2.1 2.1 2 2.1 2.3 2.6

Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex

December 1998 - March 1999 134 9.5 133 9.1 133 10.4 April 1999 - June 1999 136 7.7 138 8.3 135 8.5 July 1999 - September 1999 134 6.6 133 5.8 134 5.9 October 1999 - March 2000 112 10.8 114 9.4 111 12.3 April 2000 - June 2000 201 11.1 201 9.7 199 9.6 July 2000 - September 2000 259 11.7 254 10.4 255 11.2 October 2000 - December 2000 224 11.8 225 12.1 224 14

The results of the comparative statistical analysis for the 2001 feasibility study data can be summarized as follows: The inter-laboratory precision amongst the three labs is very similar for each period assessed. The inter-laboratory precision is generally best between the Gura Rosiei and ALS Chemex facilities. There is a high degree of relative accuracy in the assays from the three labs. The inter-laboratory precision between the three labs is predominantly within accepted industry limits for round robin gold fire assay analyses.

The inter-laboratory analytical precision of the gold fire assays from the Gura Rosiei facility, SGS (Analabs-Balcatta) and ALS-Chemex (Bondar Clegg), Canada was assessed. Descriptive statistics and graphical plots of the data are provided in Appendix 4. Summary details on the data precision and relative accuracy are presented in Table 14.10 for the period from the close of the 2001 feasibility study data in December, 2000, through December, 2002.

82

Table 14.10 Rosia Montana Gold Project Summary of Precision and Relative Bias 2001 to 2002 Round Robin Gold Fire Assays >0.15 g/t Au Data Comparison Number of Data Pairs %HARD Mean Median 4.23 5.42 3.99 4.78 3.68 4.15 6.41 4.79 3.99 5.83 5.66 6.22 3.99 3.26 4.82 7.39 7.49 6.67 7.23 6.70 5.95 5.26 5.56 4.88 Mean %HRD

Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex

January 2001 - March 2001 232 8.50 232 9.35 229 6.89 April 2001 - June 2001 232 232 224 8.49 6.99 5.31

0.783 0.289 -0.350 6.050 2.380 -3.129 4.577 1.639 -0.394 6.852 4.654 -1.675 1.048 1.812 1.379 5.662 3.814 -1.576 4.921 1.374 -1.679 0.20 -1.58 -1.74

July 2001- September 2001 202 11.77 240 7.76 156 7.63 October 2001 December 2001 188 11.58 154 11.54 151 9.14 February 2002 March 2002 152 6.66 152 7.47 152 7.84 April 2002 June 2002 234 234 231 12.66 12.79 12.79

July 2002 September 2002 232 11.22 230 11.25 225 11.98 Combined 2003 691 685 693 8.72 9.20 9.31

The results of the comparative statistical analysis for the round robin data presented above can be summarized as follows: The inter-laboratory precision amongst the three labs is very similar for each period assessed. There is a high degree of relative accuracy in the assays from the three labs. The inter-laboratory precision between the three labs is within accepted industry limits for round robin gold fire assay analyses.

Summary details on the data precision and relative accuracy are presented in Table 14.11 for the period 2003 to 2004.

83

Table 14.11 Rosia Montana Gold Project Summary of Precision and Relative Bias 2003 to 2004 Round Robin Gold Fire Assays >0.15 g/t Au Data Comparison Number of Data Pairs 5 5 5 %HARD Mean Median 5.75 9.8 5.03 5.41 7.69 6.25 6.81 4.17 5.26 6.25 10.59 7.18 2.09 4.81 4.09 Mean %HRD

Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS

November 2003 9.68 10.6 11.39

-6.89 -3.57 -3.2 2.14 0.62 2.09 1.64 1.83 -0.17 4.53 4.52 -0.01 0.59 7.62 -6.66

February 2004 - March 2004 180 10.85 183 14.12 177 13.42 April 2004 - June 2004 47 8.69 49 7.97 47 11.26 July 2004- August 2004 37 11.74 37 13.76 38 10.41 7 7 7 November 2004 7.56 9.57 9.37

The results of the comparative statistical analysis for the round robin data presented above can be summarized as follows: The inter-laboratory precision amongst the three labs is very similar for each period assessed. There is a high degree of relative accuracy in the assays from the three labs. The inter-laboratory precision between the three labs is within accepted industry limits for round robin gold fire assay analyses.

Precision of Aqua Regia / AAS Silver Assay Data - Gura Rosiei Facility The precision of the aqua regia / AAS silver assay data from the Gura Rosiei facility was assessed in a similar manner to gold assay data for the data used in the 2001 feasibility study, allowing the determination of the average levels of precision associated with sample collection, preparation, and analysis for the Rosia Montana project (Appendix 4 and Table 14.12). The results can be summarized as follows: The precision generally increases with increasing assay pair means up to approximately 25 g/t Ag for pulp repeats and duplicate crusher splits. Above 25 g/t Ag, the precision tends to remain constant. The precision generally decreases with increasing grade for duplicate field samples.

84

The levels of precision reported for the duplicate field samples, sample duplicate crusher splits, and pulp repeats are well within the industry accepted limits for the aqua regia / AAS analytical method and the styles of silver mineralization observed at the Rosia Montana project.

Table 14.12 Rosia Montana Gold Project Summary of Data Precision Gura Rosiei Aqua Regia / AAS Silver Assay Data > 3.0 g/t Au Number of Data Pairs Data Comparison Channels RC Chips 53 538 114 109 981 13 128 15 144 21 210 31 DDH Core 37 536 14.7 5.1 2.3 47 675 11 24 70 4.7 Mean %Hard Channels Cirnic 17.3 5.7 2.4 Cirnicel 3.1 Cetate 16.2 6 2.7 Carpeni 3.3 Orlea 4 Jig 14.6 5.7 2.1 Cos Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats 15 2.7 3.4 0 RC Chips 6.7 2.6 10 4.3 0 6.5 2.7 7.5 4.3 5.3 4.2 3.9 3.3 DDH Core 5.9 3.3 0 8.4 3.1 3.4 3.4 1.8 11.1 4.8 0.2 0 2.6 9.1 5.1 1.1 5.9 0 6.7 0 0 0 0 0 5.3 0 0 0 0 Median %Hard Channels RC Chips 7.7 0 DDH Core 3.7 0.9

Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats

951 959 7128 169 161 1069 250 376 3135 66 109 69 74 471

12.5 4 0.6

Tables 14.13 and 14.14 summarize the pertinent statistics for silver assays for the 2002 and pre-June 2003 drilling program.

85

The following comments can be made, relating to the 2002 and 2003 drilling program: Very similar levels of sampling precision were returned for the 2003 program when compared to the previous program. There is no bias between the sets of sample pairs (% HRD).

Tables 14.15, 14.16 and 14.17 summarize the pertinent statistics for silver assays for the post-June 2003, 2004 and 2005 drilling programs respectively.
Table 14.13 Rosia Montana Gold Project Summary of Data Precision - 2002 Drilling program Gura Rosiei Silver Assay Data 3 g/t Ag Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels RC Chips 394 2327 DDH Core 3 38 Mean %Hard Channels Table 14.14 Rosia Montana Gold Project Summary of Data Precision pre-June 2003 Drilling program Gura Rosiei Silver Assay Data >3 g/t Ag Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Pulp Repeats Channels 20 18 123 RC Chips 304 1361 DDH Core 107 521 Mean %Hard Channels 9.80 3.76 4.3 RC Chips 7.4 6.1 DDH Core 6.4 5.2 Mean HRD% Channels 0.0 0.0 0.0 RC Chips 0.0 0.0 DDH Core 0.0 0.0 RC Chips 8.3 3.3 DDH Core 4.4 Mean HRD% Channels RC Chips 3.1 0.6 DDH Core 0.1

Table 14.15 Rosia Montana Gold Project Summary of Data Precision Post-June 2003 Period Gura Rosiei Silver Fire Assay Data >3 g/t Ag Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels RC Chips 1 21 DDH Core 2 18 Mean %Hard Channels RC Chips 2.03 DDH Core 3.85 2.54 Median %Hard Channels RC Chips 0 DDH Core 3.85 0

86

Table 14.16 Rosia Montana Gold Project Summary of Data Precision 2004 Period Gura Rosiei Silver Fire Assay Data >3 g/t Ag Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels 46 42 252 RC Chips 21 157 DDH Core 25 170 Mean %Hard Channels 20.16 2.98 2.34 Table 14.17 Rosia Montana Gold Project Summary of Data Precision 2005 Period Gura Rosiei Silver Fire Assay Data >3 g/t Ag Number of Data Pairs Data Comparison Duplicate Field Samples Duplicate Crusher Splits Duplicate Pulp Splits Pulp Repeats Channels 81 58 442 RC Chips DDH Core Mean %Hard Channels 15.39 3.95 2.9 RC Chips DDH Core Median %Hard Channels 11.11 2.87 0.73 RC Chips DDH Core RC Chips 5.48 3.66 DDH Core 5.26 3.15 Median %Hard Channels 14.29 0.65 0 RC Chips 2.44 1.37 DDH Core 0 0

The following comments can be made, relating to the 2003 to 2005 drilling programs: Very similar levels of sampling precision were returned for these programs when compared to the previous programs. There is no bias between the sets of sample pairs (% HRD).

Comparison of Round Robin Aqua Regia / AAS Silver Analyses The inter-laboratory analytical precision of the aqua regia / AAS assays from SGS Gura Rosiei, SGSPerth (Analabs-Balcatta) and ALS-Canada (previously Bondar Clegg), Vancouver was assessed with the assay data grouped according to those samples submitted over approximately 3 month intervals. Descriptive statistics and graphical plots of the data are provided in Appendix 4. Summary details on the data precision and relative accuracy are presented in Table 14.18 for the 2001 feasibility study.

87

Table 14.18 Rosia Montana Gold Project Summary of Precision and Relative Bias Round Robin Aqua Regia / AAS Silver Assays >3.0 g/t Ag Data Comparison Number of Data Pairs %HARD Mean Median 6.6 5 5.5 8 3.7 6.4 5.4 6.6 3.2 9.1 5.6 9.7 7.7 6.7 5.3 13 6.5 12.8 8.9 6.2 9.1 Mean 6.3 3.1 -3.8 5.6 0.5 -5.7 3.9 5.4 0.8 -7.2 4.4 9.3 -3.2 3.4 7.2 -9.1 2.2 12.4 -1.4 -2.2 -1.1 %HRD 95% CL 2.5 1.6 2.4 2.8 2.2 2.4 2.5 1.7 1.9 5.7 3 5.8 5.1 3.5 5.6 6 3.3 5.7 5.7 3.7 6.3

Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex

December 1998 - March 1999 74 8.9 86 6.3 76 7.5 April 1999 - June 1999 90 10.5 101 6.5 90 9 July 1999 - September 1999 76 7.5 89 7.9 76 5.4 October 1999 - March 2000 45 13.4 52 8.1 41 14.4 April 2000 - June 2000 32 10.3 46 8.6 31 11.4 July 2000 - September 2000 41 16.7 64 9.5 42 17.2 October 2000 - December 2000 42 12.8 54 9.7 41 14.3

The resultant comparative statistics can be summarized as follows: The inter-laboratory precision is generally best between the Gura Rosiei and ALS-Chemex (previously Bondar Clegg) facilities. Adequate overall levels of relative accuracy are evident amongst the three laboratories, with the exception of the July, 2000 to September, 2000 period, within which the SGS-Perth (AnalabsBalcatta) assays are significantly higher-grade than the assays from both the Gura Rosiei and ALS-Chemex (previously Bondar Clegg) facilities. All resource estimations have been based on the Gura Rosiei analyses and, as such, the accuracy of the SGS-Perth (Analabs-Balcatta) data for this July, 2000 to September, 2000 period is not considered to reduce the overall quality of the assay data. Aside from this period, the inter-laboratory precision between the three laboratory facilities is within accepted industry limits for round robin aqua regia / AAS silver analyses.

88

A similar approach was adopted to assess the round robin silver data for the 2002 drilling program and other exploration from the close of the 2001 feasibility study data in December, 2000 to 2003. Table 14.19 summarizes the results.
Table 14.19 Rosia Montana Gold Project Summary of Precision and Relative Bias 2001 to 2002 Round Robin Silver Assays >3 g/t Ag %HARD Mean Median 5.88 4.38 7.47 5.88 6.10 4.00 4.68 4.00 3.45 3.29 4.17 11.40 3.35 4.93 7.69 6.92 6.82 5.88 3.734 3.961 4.76 4.55 5.26

Data Comparison

Number of Data Pairs

Mean %HRD

Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex Gura Rosiei vs SGS Perth Gura Rosiei vs ALS Chemex SGS Perth vs ALS Chemex

January 2001 - March 2001 160 11.07 88 10.95 102 15.49 April 2001 - June 2001 144 8.09 144 8.48 150 5.73 July 2001- September 2001 75 6.81 123 6.43 80 5.53 October 2001 December 2001 62 5.93 63 7.87 February 2002 March 2002 57 13.44 28 6.12 25 9.06 April 2002 June 2002 96 12.40 96 11.344 104 10.89 July 2002 September 2002 146 8.00 145 7.709 148 7.372 Combined 2003 304 7.92 301 8.15 306 8.70

1.117 7.426 12.35 2.955 1.829 0.416 0.498 1.624 2.272 0.939 2.080 7.844 3.978 -0.350 4.855 7.146 6.854 3.740 2.302 0.094 -1.50 -1.02 1.42

Summary details on the data precision and relative accuracy are presented in Table 14.20, for the period 2003 to 2004.

89

Table 14.20 Rosia Montana Gold Project Summary of Precision and Relative Bias 2003 to 2004 Round Robin Silver Fire Assays >3 g/t Ag %HARD Mean 5.33 6.32 7.29 8.66 10.38 10.95 4.56 6.75 5.36 5.28 5.78 5.52 3.49 4.95 2.93 Median 5.73 6.19 5.37 3.89 5.26 5.26 0 3.2 1.96 4.35 4.19 3.85 0 4.11 2.28

Data Comparison

Number of Data Pairs November 2003 6 5 5 February 2004 - March 2004 120 109 106 April 2004 - June 2004 40 30 31 July 2004- August 2004 38 37 37 November 2004 7 7 7

Mean %HRD

Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS Gura Rosiei vs Analabs Gura Rosiei vs ALS Analabs vs ALS

-1.51 4.69 6.5 1.56 3.66 1.96 1.72 -0.41 -1.53 4.01 -0.06 -3.78 -1.59 -3.16 -1.57

14.4

GRID CONTROL AND SURVEY

All surveying, topography, underground workings and drill hole collars for the Rosia Montana project are reported based on the Stereo70 grid system. Appendix 5 provides detailed discussion of the Stereo70 grid system and various transforms. 14.4.1 Topography

The current topography used in the resource estimation study is based on aerial photography and ground survey. The aerial photography was flown by RMGC as part of the 2001 feasibility study with the topography generated by licensed surveyors Spectrum Survey and Mapping of Perth, Australia. Additional close spaced surveying of the plant site has been completed but this does not impact the mine area and has not been incorporated into the updated digital terrain model. Figure 14.1 displays a perspective view of the topographic model of the Rosia Montana project area. 14.4.2 Drill Hole Collar Locations

Spectrum Survey and Mapping of Perth has routinely surveyed all drill hole collars. The collars are considered accurately located in 3D space.

90

Figure 14.1 Perspective View of the Topographic Model

14.4.3

Downhole Surveying

All surface and underground drill holes have been downhole surveyed using Eastman or Sperry Sun single shot cameras, based on a downhole interval of approximately 50 metres. Due to ground conditions many RC holes had to be surveyed inside the drill rods, resulting in dip measurement rather than a dip and azimuth measurement. Some RC holes have been surveyed after the removal of the drill rig, using PVC piping to protect the Eastman single shot cameras. In these cases the PVC piping was lowered down the hole as far as possible and camera shots were taken every 50 metres down the hole. 14.4.4 Underground Workings

The underground workings have been constructed from digitized historical plans. Check surveying of portals and traverses within the underground development by Spectrum Surveys has confirmed the accuracy of the historical plans, and effectively the small amount of deviation (generally much less than 2 to 3 metres can be ascribed to changes in stability of the original historic plans). 91

The data and methodology applied to the construction of the underground void model are summarized as: Digitize plans of historical workings in the original mine grid system and produce closed solid triangulations using Vulcan and Surpac mining software. Convert all underground wireframes to the hybrid stereo 70. Check survey of all the gallery portals and selected traverses of significant accessible levels. All surveying was completed using EDM in the HYST70 grid system. The check surveying of the selected levels included pickup of the location of backs, location of crosscuts and start position of the underground channel sampling supervised by RSG. Validate the check survey against the solid triangulations in HYST70. The more recently discovered underground voids and workings located in the course of the archaeological program were surveyed by a team of underground surveyors working under the supervision of the University of Toulouse of France who were contracted to the Natural History Museum of Bucharest under the auspices of the Ministry of Culture. They delivered to RMGC the completed surveys, which were all tied to previous underground control points set up by Spectrum for RMGC. During 2004 and 2005 RMGC survey teams used distance measuring equipment to measure the shapes and volumes of all accessible voids and underground workings. These measurements which were tied into the existing grid were used to update the void model.

RSG Global considers that further detailed surveying of accessible underground workings is warranted prior to the onset of mining. It is noted that the significant underground voids (or coranda) have not been internally edm/laser surveyed to date. There is thus uncertainty as to their exact volumes and shapes and it is strongly recommended that surveying is undertaken prior to beginning open pit mining. 14.4.5 Channel Locations

The start points of underground channel traces were located on the underground development plans, which form the basis for the underground development wireframes. The same start points are marked on the walls of the drives using spray paint. During the data validation process, all visible start points for the underground channel samples were surveyed by Spectrum Surveys and were subsequently compared with the underground channel sampling three-dimensional database. The maximum differences in position exactly matched those described above for the drive walls. Thus the channel samples are considered to be accurately located in space, particularly given the block size and hence selective mining unit for the planed operation at Rosia Montana. 14.5 BULK DENSITY

A total of 6,213 determinations have routinely been collected since January, 1998, from both diamond drill core and hand specimens obtained from underground development. Measurements were undertaken at the Cepromin laboratory in Deva, Romania, which is a commercial laboratory previously run by the Romanian government prior to privatization. Senior RSG Global staff audited the bulk density measurement process.

92

Samples were collected and data recorded according to detailed mineralized zone location, lithology and style and intensity of alteration. Diamond core samples were prepared by squaring off the ends of approximately 15 cm billets of half core. Bulk density determination was by the standard water immersion method with each sample coated in wax prior to immersion. Standard laboratory samples were used to calibrate the scales between each measurement. All samples were returned to site and the samples placed back into the core trays, without removing the wax coating, as a record. Results are supplied in hardcopy format with the bulk density measurement reported to two decimal places. A detailed description of the procedures used is provided in Appendix 6. 14.6 VERIFICATION SAMPLING

The channel sampling and drilling data represent different sampling techniques and, as such, required comparison before both datasets were accepted into the final exploration database applied in resource evaluation. A detailed investigation comparing the different data types was completed as part of the 2001 feasibility study and will not be repeated in this document. The reviews completed as part of the feasibility include: A review of the datasets in terms of their individual data quality (e.g. precision and accuracy) and on the impact of combining different data types, characterized by different sampling error, for resource estimation purposes. Comparative estimates of a selected region of Cetate using the total database, only the channel database and only the drilling database. Reporting of check underground channel samples taken by RSG. Comparative downhole variography generated for the Cetate SIK region using only channel data and only drilling data.

The 2001 feasibility study resulted in the conclusion that no bias was evident in the datasets and the combination of the different data was possible. After the feasibility study a number of investigations have been completed to test the veracity of the drilling and channel data and also to investigate the appropriateness of combining the datasets. The investigations undertaken post feasibility includes: Twin drilling to compare DDH-DDH, DDH-RC, RC-DDH and RC-RC. Independent underground channel sampling completed by ACA Howe.

A review of co-located data pairs of underground channels versus drilling was completed. Drill Hole Twinning A total of 34 pairs of drill holes, representing twinned holes, have been completed during the exploration programs undertaken since 1998 at Rosia Montana (Table 14.21). Detailed discussion of the 2002 and 2004 twin drilling programs is the subject of separate reports (see Appendix 7). Most of the 2002 twinned drill holes were completed as specific tests of pre-existing holes, however, a number 93

were completed as re-drills of holes which failed to reach the target depth. Both the original and twin holes were drilled for the majority of the twin holes in the 2004 program. It should be noted that the following types of drill twins have been carried out during the course of exploration at Rosia Montana: Diamond Core: Diamond Core. Diamond Core: Reverse Circulation. Reverse Circulation: Reverse Circulation. Reverse Circulation: Diamond Core.

The study reinforces the following conclusions which were also noted during the 2001 feasibility study: At a local level there is a high degree of variation. There is no bias between the primary drilling and the twin drilling. As the scale of comparison is increased the degree of variation is reduced significantly. The twin drilling reinforces the observation that at a global level the drilling is an excellent estimator of the metal in the neighbourhood, however, on a local level a high degree of variation can be expected. Detailed grade control drilling will be required in order to produce a robust local estimate for short term scheduling purposes.

Review of Underground Channels and Drilling RSG Global completed a review of the underground channels versus the drilling (combined DDH and RC) to investigate potential bias between the datasets. The investigation was designed to supplement the investigations reported in the 2001 feasibility study and to include the infill drilling. In general, the drilling and channel data represent different regions of sampling and therefore conclusive investigation is difficult due to the lack of co-location. In addition, the infill drilling completed post the 2001 feasibility study has generally been targeted at the periphery of the main siliceous alteration, which is sampled predominately by channels and not drilling. The methodology applied to this study was to migrate the nearest channel data to the drilling data (10 m composites) such that spatially co-located pairs could be reviewed. An isotropic search of 10 m was applied to the migration. A review of the co-located pair data was then completed via means of summary sample statistics, scatter plots and quantile-quantile plots.

94

Table 14.21 Rosia Montana Gold Project Rosia Montana Drill Hole Database : Drill Twins Twin Type RC-DDH RC-DDH RC-DDH RC-DDH RC-RC RC-RC RC-RC RC-RC RC-RC RC-RC (39m), RC-DDH (17m) RC-DDH (40m), DDH-DDH (94m) DDH-DDH DDH-DDH Twins 2002 Twin drilling CTSD054 CTSD016 RMRD054 RMRD091 CTSD001 RMRD070 CNSD051 CNSD102 CTSD109 CTSD217 CNSD034 CNSD036 CNSD037 CNSD010 CNSD015 CNSD058 CNSD091 CTSD115 CTSD218 CNSD018 CNSD003 CNSD024 CNSD092 RMRD027 CTSD219 CTSD070 CTSD220 2004 Twin Drilling CNSD126 CNSD127 CNSD130 CNSD131 CNSD112 CNSD113 CNSD114 CNSD115 CNSD120 CNSD121 CNSD134 CNSD134D1 CTSD221 CTSD222 CTSD233 CTSD234 CTSD227 CTSD228 CTSD229 CTSD229D1 CTSD223 CTSD224 CTSD235 CTSD236 CTSD225 CTSD225D1 CTSD225 CTSD226 RMRD012 CTSD260 RMRD081 CTSD266 CTSD024 CTSD241 CNSD078 CNSD132 CNSD081 CNSD138 CTSD021 CTSD225D1 CNSD071 CNSD118 Company RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC RMGC Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence Due Diligence RMGC Due Diligence RMGC Due Diligence RMGC Due Diligence RMGC Due Diligence RMGC Due Diligence RMGC Due Diligence RMGC Due Diligence Metres Twinned 150.3 150.3 140.1 140.1 109 109 160 160 120 120 37 37 37 85 85 110 110 140 140 56.3 56.3 134 134 129 129 130 130 145 145 172 170 140 140 216 217 90 89 38 38 194 193 261 260 172 169 60 60 273 275 243 244 117 123 110 109 173 161 170 166 173 156 134 133 200 200 135 132 110 111

DDH-RC DDH-RC DDH-RC DDH-RC DDH-RC DDH-DDH DDH-RC DDH-RC DDH-RC DDH-DDH DDH-RC DDH-RC DDH-RC DDH-RC RC-RC RC-RC RC-DDH RC-DDH RC-DDH RC-RC RC-DDH

95

A total of 199 sample pairs were collected. The mean gold and silver grades of the channel data are higher than the combined drilling dataset, although relatively few data exist for the comparison. While the mean grades are noted to be higher, similar levels of variability are noted for both data types as reflected by relatively similar standard deviations. Table 14.22 presents the summary statistics comparing the drilling and channel data pairs.
Table 14.22 Rosia Montana Gold Project Summary Statistics 10 m Composite Data Pairs (+/-10 m) Drilling versus Underground Channel Sampling Gold (g/t) Drilling No of Pairs Minimum Maximum Mean Median Std Dev CV 199 0.05 24.50 1.17 0.67 2.16 1.85 Underground Channel 199 0.03 24.30 1.31 0.72 2.23 1.70 Drilling 199 0.50 102.5 5.93 2.90 10.78 1.82 Silver (g/t) Underground Channel 199 0.50 72.17 6.56 3.43 9.90 1.51

A review of the correlation plots reveals a broad scatter and little correlation between the datasets. The QQ plots show similar distributions for gold, although slightly higher grades are noted for the channel data between 2 and 5 gt Au. The silver data show similar poor correlation with the QQ plots indicating that the channel data are consistently of higher grade. RSG Global believes that the results of the comparison are inclusive although little strong evidence exists that the channel data are biased for gold. Silver is more problematic, although the studies completed as part of the 2001 feasibility study and included as Appendix 8 support a finding that no bias is present between the two datasets. In addition, RSG Global understands that various parties have completed check channel sampling and have indicated high variability but no bias.

96

15.0 ADJACENT PROPERTIES

Gabriel is continuing to explore the adjacent Bucium exploration license, which is situated to the immediate southeast of the Rosia Montana license and covers an area of 32.123 km2. Exploration work consists of geological mapping, sampling and drilling. An independent resource has been defined by RMGC in the Bucium license area. As the holder of the exploration concession license, RMGC has the right to apply to the NAMR for conversion of that license into an exploitation concession license. RMGC made such application in May, 2007 and is currently awaiting a response from the NAMR as to when negotiations can begin on the terms and conditions of the new exploitation concession license.

97

16.0 MINERAL PROCESSING AND METALLURGICAL TESTING

16.1 GENERAL Metallurgical testing of the Rosia Montana ores has been conducted by several parties during the recent development of the project. Testwork programs have been managed by: Pincock Allen & Holt, in January, 2000, as part of a prefeasibility study. Minproc with testwork performed by Ammtec Limited (Ammtec) and Amdel Limited as part of a feasibility study. SNC with testwork performed by Lakefield Research as part of design. Ausenco International Limited (Ausenco) with testwork performed by Ammtec as part of detailed design. Newmont Limited (Newmont) with testwork performed by the (in-house) Plato Malozemoff Technical Facility in Englewood Colorado. J.R. Goode and Associates (J.R. Goode) metallurgical consultants, in liaison with RMGC, with the testwork performed by SGS-Lakefield Research.

The program conducted by Pincock Allen & Holt was on a limited sample base and could be considered preliminary at best. These limited results were used as the basis of a prefeasibility study completed in 2000. The results obtained can be considered to be aligned with those of the later testwork programs. Consequently, this work is not considered formative with regard to the development of the process design and is not discussed any further in this report. The most substantial testwork program was that conducted by Minproc. This program forms the basis of the flowsheet, provides variability behaviour, was the basis of subsequent testwork programs and provides a database that can be used for flowsheet optimization and recovery determinations as used in the mine modelling. A large number of samples were collected including a significant number of variability samples, originating from both drill core and channel samples collected from underground workings. The SNC program investigated some flowsheet refinements as required by SNC in its flowsheet development and as had been recommended by Minproc as part of its metallurgical testwork reporting. The samples used were from the same sample set used by Minproc. Ausenco conducted some further work biased towards cyanide detoxification with some additional gravity and leach extraction determinations. The samples used for this program were new composites representing the first seven years of operation and were derived from drill core. Newmont undertook a testwork program to both confirm the metallurgical performance as had been reported by earlier programs, as well as to explore the presence of gravity gold. Newmont managed a sampling program on site, collecting samples from a number of locations adjacent to the same sample points that had been used for the collection of the original Minproc variability samples, in order to 98

confirm sample source in addition to metallurgical performance. The samples were sent to the Plato Malozemoff Technical Facility in Englewood, Colorado for processing. The program managed by J. R. Goode and Associates was conducted on splits of the same samples collected under the management of Newmont. This program developed some concepts identified by J.R. Goode and Associates after review of all the previous testwork, including the Newmont program. 16.2 THE RPA TECHNICAL REPORT

The RPA report of March, 2004, summarized the testwork conducted by Minproc and SNC. Apart from some subtle differences in the cyanide detoxification area of the plant, the current process flowsheet is based on the Minproc and SNC testwork. The metallurgical discussion of these two testwork programs, as presented in the RPA report, thus remains relevant, and is summarized below. Comments in square brackets are not as written by RPA, but are provided by Aurifex. 16.2.1 Samples and Flowsheet Development RPA Report Fifty 20 kg samples were collected on behalf of Minproc from the Cetate and Cirnic deposits and composites prepared to represent each deposit. The composites were subjected to a comprehensive program conducted primarily at Ammtec, with some specialist work conducted by other parties. Testwork included: Assay and mineralogical study. Diagnostic leaching to estimate association of residual gold. Cyanidation under various conditions including various grind sizes, cyanide concentrations, pulp densities, in the presence of activated carbon, pre-oxidation and leach rate enhancement. Gravity concentration. Flotation. Gravity and flotation concentrate treatment (grinding, leaching). Oxygen uptake. Viscosity as a function of pulp density. Settling characteristics and flocculation demands.

16.2.2 Ore Characterization RPA Report The analyses of the ten composites are summarized in Table 16.1. The analysis suggested that none of the elements often associated with cyanide extraction difficulties is present at elevated levels.

99

Table 16-1 Summary of Head Assays Ore Type Sample Au g/t (assay) Au g/t (calc) Ag g/t As ppm Bi g/t C organic % Cu g/t S % (total) Sb g/t Cetate 3 1.68 1.77 9.8 110 < 10 0.05 30 2.49 7 Cirnic 8 1.91 1.62 8.6 130 < 10 0.05 30 1.48 5

1 1.81 1.55 12.6 182 < 10 0.05 50 2.72 5

2 1.72 2.21 17.4 165 < 10 0.05 80 3.08 4

4 1.70 1.70 11.6 234 < 10 0.06 60 3.51 7

5 1.87 1.87 7.1 130 < 10 0.07 120 2.11 10

6 1.6 1.57 15.4 200 < 10 0.06 30 1.17 8

7 1.69 1.95 21.4 125 < 10 0.05 80 1.41 8

9 1.87 1.86 10.8 170 < 10 0.07 45 1.57 11

10 1.82 1.63 33.6 135 < 10 0.08 300 2.18 124

Mineralogical studies concluded that: The samples contained approximately 5% sulphides with pyrite dominant. Gold and silver are associated with the sulphides and with quartz. All ten samples had similar characteristics, being dominant in quartz, feldspar and muscovite. Jarosite found in most samples indicated that oxidation of sulphides is present [suggesting that similar processing behaviour could be expected]. Gold grains were found in seven of the ten samples, ranging from 1 to 180 m, but generally less than 25 m, suggesting low gravity recoveries. This gold was generally associated with pyrite; however, examples associated with chalcopyrite and acanthite were observed, as well as quartz. The gold grains are high in silver which may retard the leaching process, compared to low silver gold grains [and could be expected to reduce the gold extraction unless conditions are conducive to silver leaching]. No native silver was identified. Silver associations with acanthite, galena, chalcopyrite and tetrahedrite could be expected. These silver associations would be expected to give low cyanidation extractions.

Diagnostic leaching was conducted on leach residues that had been subjected to gravity concentration to recover coarse gold prior to leaching. The results of the diagnostics are summarized in Table 16.2. [The diagnostic leaching highlights that the dominant residual phase is that of gold locked in sulphides.]
Table 16.2 Diagnostic Leach Data % Gold Distribution Source Sample Gravity recovered Cyanide soluble Sulphide locked Silicate locked Cetate 1 15 51 31 4.0 2 18 58 23 1.5 6 29 50 16 4.3 Cirnic 7 18 64 17 1.5 Low Grade 11 60 35 4.7

100

16.2.3 Cyanidation Leaching and Grind Sensitivity RPA Report Cyanide leaching included bottle roll tests of coarse samples, a column heap leach of a low grade composite, and samples ground to the nominal liberation size of 150 m. The results are summarized in Table 16.3 and reveal some sensitivity to size distribution. Further grind sensitivity analysis was conducted at a range of P80 sizes from 75 to 220 m. Leaching was conducted under testwork standard conditions of 0.1% NaCN, pH of 10.5 and for a duration of 24 hours. The results summarized in Table 16.4 present a low grind sensitivity for most samples once a nominal P80 of 150 m is reached. Minproc compared the additional capital and operating costs of finer grinding with the revenue from the additional recovery gained. It was concluded that 80% passing 150 m was the optimum grind [thereby defining a key design criterion].
Table 16.3 Comparison of Heap Leach and Conventional Leaching Bottle Roll Gold Extraction at 150 m (%) 61.2 78.6 79.2 68.8 62.3 75.4 83.3 80.1 82.0 75.7 60.6 Bottle Roll Gold Extraction at 10 mm (%) 31.8 66.9 53.4 36.5 53.7 58.5 57.6 37.7 63.8 57.3 45.9 Column Gold Extraction at 10 mm (%)

Sample

Head Grade (g/t Au) 1.55 2.21 1.77 1.70 1.87 1.57 1.95 1.62 1.86 1.63 0.78

1 Cetate 2 Cetate 3 Cetate 4 Cetate 5 Cetate 6 - Cirnic 7 Cirnic 8 Cirnic 9 Cirnic 10 - Cirnic 11 (Low/Grade)

46.1

Table 16.4 Gold Recovery as a Function of Grind Size Ore Type Grind P80 75 106 150 180 220 Cetate Extraction to Solution (%) 1 64.3 63.0 61.2 57.4 55.1 2 81.7 80.1 78.6 74.7 73.7 3 82.1 79.0 79.2 6 75.9 75.3 75.4 74.6 69.0 Cirnic Extraction to Solution (%) 7 84.3 82.5 83.3 79.8 78.6 8 81.6 80.7 80.1 79.0 75.7 9 82.7 81.8 82.0 10 79.9 76.9 75.7

Testwork conducted to determine cyanide concentration sensitivity suggested that increased extractions were possible at elevated levels; however, Minproc determined that the additional cyanide cost was not economically justified [given the economic conditions at that time]. An extended leach time test suggested that the optimum leach time was 15 hours [given the economic conditions at that time].

101

16.2.4 Gravity Concentration and Sulphide Recovery RPA Report Gravity concentration tests were conducted that, in general, revealed little free gold, and total gold recoveries to concentrate of less than 20% [it should be noted that these are recoveries to a high mass pull concentrate and are not extractions]. Sulphide recovery by gravity and flotation was trialled, both giving similar results. Gravity sulphide concentrates were leached as recovered and also following fine grinding. The silver extraction for the ground gravity concentrate was improved and an improvement in the Cetate gold extraction was found. The Cirnic sample did not show a significant increase in extraction. The overall recovery for these process routes was not acceptable, compared with leaching of whole ore. Pressure oxidation tests were conducted and gave very high recoveries. These results combined with the flotation results were used to evaluate such a flowsheet, along with several other options. 16.2.5 Grinding Testwork RPA Report Grinding testwork was been conducted at both bench scale and pilot scale. The grinding energy has found to be a function of lithology, as would be expected. The dominant dacite ores were found to have the highest work indices (approximately 20 kWh/t) with the peripheral and breccia ore types being somewhat softer (13 15 kWh/t). Pilot scale semi-autogenous grinding testing was conducted on three samples totalling some 170 tonnes. This work confirmed the bench scale testwork [and provided design data for the grinding circuit]. 16.2.6 Variability Testing and Algorithm Development RPA Report A large number of samples for Cirnic and Cetate [approximately 280 samples from Cirnic and 220 from Cetate] and samples representing other smaller mineralized bodies were subjected to a standard leach flowsheet. [Early in the project development Minproc determined that a simple crush, grind, leach circuit was all that was required to process the Rosia Montana ores. This standard leach flowsheet was applied to the variability samples]. The variability testing results were used to establish recovery algorithms. Weak relationships were found, but these correlations improved when a moving average of 20 variability tests was assessed. [This methodology and the results of this process are discussed in more detail below as further algorithm development was undertaken post the RPA report]. To adjust the correlations from the averaged variability test results with the proposed flowsheet that included gravity and a different leach time, eleven composite samples were prepared from the variability samples, and subjected to the proposed flowsheet. The actual gold and silver extractions were compared with the averages from the variability testing and the recovery algorithms were adjusted accordingly. This was simply achieved by subtracting 1.3% for Cetate and 2.8% for Cirnic for gold, and subtracting 10.5% for Cetate and 13.9% for Cirnic for silver. Adjustments were similarly made for the other deposits represented by the variability samples [the final Minproc flowsheet gave lower extractions than the standard variability tests].

102

16.2.7 Cyanide Destruction [Detoxification] RPA Report Cyanide detoxification testwork was conducted by both Minproc and SNC. Minproc focused on treatment of tailings decant liquor whilst SNCs work was based on the actual tailings slurry prior to discharge to the tailings storage, due to a subtle flowsheet revision in this area. SNC tested a number of detoxification alternatives and selected the SO2/air process as it was the lowest cost and most effective evaluated. 16.2.8 Testwork to Develop Design Criteria A number of testwork programs were carried out by both vendors and metallurgical laboratories to define key design criteria not defined by the developmental testwork programs. This work included: Settling and flocculant screening by Supaflo (Outokumpu). Slurry rheological studies performed by Ammtec and Curtin University (process plant aspects), Knight Piesold for tailings behaviour and Lightnin for agitator design. Carbon kinetics tests.

16.2.9 Flowsheet Options Several flowsheet options were evaluated by Minproc including flotation, gravity sulphide options, pressure oxidation and variations on these alternatives. This evaluation concluded that carbon-in-leach (CIL) for the whole ore, ground to 80% passing 150 m was the most economic option for most ore types. The traditional multi-stage crushing and grinding flowsheet was rejected on the grounds of capital and maintenance cost. Various autogenous and semi-autogenous (SAG) with ball mill grinding configurations were considered. Due to the likely build up of pebbles, the final circuit included SAG, ball mill and pebble crushing [SABC circuit], in which pebbles are extracted from the SAG mill and re-cycled through a cone crusher to the SAG mill feed conveyor. A later study by Minproc concluded that the extraction of a gravity concentrate improved the kinetics of the subsequent leaching, so the gravity concentration circuit was included in Minprocs design together with an intensive leach of the gravity concentrate and a 15 hour residence time in the CIL. [Minproc recommended that further work be conducted in this area which was duly undertaken by SNC. SNC retained the gravity circuit in its designs.] 16.3 AUSENCO MANAGED TESTWORK AND FLOWSHEET MODIFICATIONS

16.3.1 Introduction Ausenco conducted two stages of engineering design. The first was as a joint venture partner with Bechtel Australia Pty. Ltd. (Bechtel) during which time a flowsheet evaluation was carried out using primarily the Minproc and SNC testwork results. This resulted in some flowsheet changes based upon the economic criteria of the time. Of significance was deletion of the leach feed thickener, increase in CIL residence time to a nominal 24 hours (from 15 hours) and the deletion of the gravity circuit.

103

The second stage of engineering design was by Ausenco in its own right. Some minor equipment changes were made at this time but basically the flowsheet remained as it had been during the earlier Bechtel-Ausenco joint venture period. Ausencos assessment of the cyanide detoxification work conducted by Lakefield (managed by SNC) was that the work was not adequate for process design. Consequently, a testwork program was devised to generate cyanide detoxification criteria as well as to provide some confirmatory testwork on other aspects of the flowsheet. In addition, water chemistry, additional slurry rheology, geotechnical and geochemistry work was conducted. 16.3.2 Samples - Ausenco The selection of samples for the Ausenco program was based on obtaining representative samples to characterize the mine plan and ore production schedule from the Cirnic and Cetate pits in the first seven years of operation only. Four composite samples were selected to represent the principal ore feed scheduled to the plant in terms of ore blend from pits, rock type, alteration and gold/silver/sulphur grades. Samples were selected with the assistance of the mining and geological consultants retained by RMGC, and were derived from diamond drill core, typically from historical drilling, although some new core was drilled to represent areas that had not been adequately assessed in the past. The samples were inspected by representatives of Ausenco and, in some instances, significant oxidation had occurred in the old core. Such samples were rejected as non-representative. The four composites were referred to as RM1, RM2, RM3A and RM3B. The composite components are summarized in Table 16.5.
Table 16.5 Composite Details Composite RM1 RM2 RM3A RM3B Pits Mined Cirnic Cetate Cirnic Cetate Cirnic Cirnic Cirnicel Relative Ore tonnage 80 20 33 67 100 70 30 Years Treated 1 to 3 1 to 3 4 4 5 to 7 6 to 7 6 to 7 Rock Types Dacite dominant Dacite dominant Dacite/mixed breccia Dacite/vent breccia Dacite dominant Dacite breccia g/t Au 2.16 1.83 1.77 1.34 g/t Ag 15.3 4.3 3.1 10.5 Weight (kg) 179 170 143 111

16.3.3 Program - Ausenco The Ausenco testwork program included: Head analyses and mineralogical studies. Influence of cyanide concentration and grind sensitivity.

104

Effect of leach pulp density on leaching behaviour as earlier work had suggested some dependence. Effect of leaching in the presence of carbon. The Minproc testwork suggested a benefit when leaching in the presence of carbon, albeit minor. Review of leach kinetics. Effect of temperature on leaching as it had been observed that temperature effects could influence leach rate. Oxygen demand and pre-aeration effects. Gravity concentration. Diagnostic leaching. Viscosity behaviour at various pulp densities. Acid mine drainage neutralization methodology and design criteria development. Thickener sizing and agitation testwork. Cyanide detoxification testwork conducted by CyPlus gmbH (CyPlus).

16.3.4 Mineralogy and Diagnostic Leaching - Ausenco Mineralogically the four composites were quite similar. Differences between the samples tended to be in the accessory mineral abundances. The mineralogical results are summarized in Table 16.6.
Table 16.6 Relative Mineralogical Abundances Mineral Orthoclase Quartz Muscovite Pyrite Fe-Mn Carbonate Rutile Apatite Sphalerite Chalcopyrite Barite RM1 60.5 22.1 10.9 4.7 1.30 0.22 0.20 0.08 0.05 0 RM2 59.1 22.4 12.0 5.1 0.80 0.38 0.18 0 0 0 RM3A 67.2 17.5 13.0 2.3 0 0.13 0 0 0 0 RM3B 60.0 24.8 10.0 3.5 0.86 0.25 0.03 0.16 0.04 0.04

Diagnostic leaching of leach residues showed that most of the residual gold was associated with sulphides, except for sample RM1 which had a significant component of silicate locked gold reported. This suggests that a relationship between sulphide level and residue grade would be expected. 16.3.5 Cyanide Conditions - Ausenco Cyanide concentration gave improved extraction rates at the higher concentrations; however, the final extractions were similar if oxygen sparging was used. These results suggested that the use of oxygen sparging would be a design feature. Cyanide consumption was found to be significantly reduced with pre-oxidation and a corresponding increase in extraction was achieved. The plant design uses high 105

levels of oxygenation in the first two tanks to exploit this observation. Oxygen uptake testwork provided parameters for system design. Leaching was conducted at three grind sizes for samples RM1 and RM2. Results are summarized in Table 16.7. This analysis was undertaken to assess the losses that may be attributable to a coarser grind and a corresponding increase in plant throughput. Whilst losses at 48 hours were similar, some sensitivity was shown at 24 hours. It was considered that at grinds coarser than 150 m there may be issues with agitation and screening, and so the 150 m grind size was retained as the design value.
Table 16.7 Grind Sensitivity Ore Composite RM1 P80 Grind m 212 180 150 212 180 150 Residue Grade, g/t Au Calculated 24 h Assay 48 h 0.61 0.47 0.59 0.48 0.51 0.46 0.59 0.50 0.76 0.54 0.53 0.49

RM2

Leaching at various pulp densities did not show any consistent trends. Similarly CIL versus CIP processing (leaching in the presence of carbon or not) did not show any consistent trends. Consequently the design pulp density was not dictated by extraction issues and the use of CIL over CIP was not considered mandatory; however, it remains as the design basis. Leach kinetic tests showed that leaching was continuing to 24 hours and therefore the design leach time of 24 hours was retained. The Rosia Montana slurry temperature in the CIL circuit is expected to get to a nominal 15oC in winter. A series of tests were conducted at 15, 22 and 30C to establish if the leach behaviour was adversely influenced. The final results showed no influence with low temperatures and therefore no heat retention measures were to be provided as part of the CIL design. 16.3.6 Carbon Kinetics - Ausenco Carbon loading kinetic testwork for both gold and silver was conducted using aged and attritioned carbon. Values achieved are summarized in Table 16.8. The k value is typical, although the n factor is at the lower end of typical values.
Table 16.8 Carbon Kinetic Parameters Bulk Leach Solution Tenor Au (ppm) 1.68 1.84 0.92 Gold Fleming and Nicol Constants k (h-1) n 158 0.54 157 0.54 157 0.55 Bulk Leach Solution Tenor Ag (ppm) 6.14 1.95 1.34 Silver Fleming and Nicol Constants k (h-1) n 97 0.53 116 0.53 108 0.55

Ore Composite RM1 RM2 RM3A

106

16.3.7 Gravity - Ausenco Samples were ground to a P80 of 250 m and sulphide concentrates were collected using a laboratory scale Knelson concentrator, and then the concentrates were ground to a P80 of 38 m. These concentrates were intensively cyanided whilst the gravity tailing was leached under standard conditions of 700 ppm NaCN, oxygen sparged for 24 hours (45% solids). The results are compared with the leach of whole ore samples in Table 16.9. The results suggest that some opportunity exists to increase extractions with such a flowsheet; however, earlier economic assessment of this option by Ausenco had suggested that the additional operating and capital costs did not provide an acceptable pay back. It should be noted that the current economic conditions are somewhat different from those of the time of this testwork and the assessment as undertaken by Ausenco as to the cost effectiveness of including a gravity circuit.
Table 16.9 Gravity + CIL Leach versus CIL Leach Ore Composite Gravity + CIL Recovery (%) 80.9 83.1 64.4 Leach/CIL Recovery (%) 77.8 79.9 67.6 Difference in Recovery (%) + 3.1 +3.2 - 3.2 Range Leach/CIL Recovery (%) 73.9 83.4 74.3 82.9 64.1 72.2

RM1 RM2 RM3

16.3.8 Cyanide Detoxification - Ausenco The SO2/air cyanide detoxification process was evaluated by CyPlus on three Rosia Montana leach effluent slurries emanating from samples RM1, RM2 and RM3A. In addition, scoping tests using the CyPlus CombinOx process were performed on effluents from RM 1 and RM 2. The laboratory work showed that the SO2/air process was successful in reducing the CNWAD (weak acid dissociable cyanide) from a feed concentration range of 159-184 mg/L down to a residual concentration range of 0.2-0.8 mg/L. The CNTOT (total cyanide) was reduced to a residual concentration range of 1.25.1 mg/L. The RM1 ore leach effluent required a retention time of 60 minutes and reagent dosages of 4 g SO2 per g CNWAD. For the RM2 and RM3A leach effluents, the same conditions as for RM1 leach effluent were found adequate for the fresh leach feed, but for a series of confirmatory tests the retention time had to be increased to 90 minutes for leach effluent feeds left standing for several days. The SO2 dosage had to be increased from 4 to approximately 6 g per g CNWAD. Treated effluent slurries were left standing for 48 hours at 5C and re-analyzed. A slight decrease in residual cyanide levels and metals resulted. Scoping tests using the CyPlus CombinOx process did not remove the CNWAD to acceptably low levels in an economical justifiable way and this process is therefore not considered to be a viable alternative.

107

16.3.9 Thickening and Agitation - Ausenco GL&V Eimco and Outokumpu were commissioned to carry out thickening testwork to allow thickener sizing to be conducted. The results of this work are summarized in Table 16.10. The results achieved by the two vendors were comparable with a settling rate of 1.2 t/m2/h for a 15% solids feed density being selected for design. A nominal flocculant dose of 30 g/t was similarly used for design.
Table 16.10 Summary Thickening Results Vendor Sample Feed density % w/w Overflow Clarity, ppm Flocculant Dosage, g/t Underflow Density, % w/w Settling Rate, t/m2/h RM1 15 110 / 30 23 / 30 62 / 63 1.2 / 1.5 Outokumpu RM2 15 60 / 70 22 / 23 62 / 60 1.2 / 1.5 RM3A 15 60 / 80 29 / 30 59 / 58 1.2 / 1.5 RM3B 15 110 / 90 30 / 30 62 / 61 1.2 / 1.5 RM1 10 < 300 20 / 30 60 / 60 1.2 / 1.3 GLV Eimco RM2 10 < 300 20 / 30 64 / 64 1.2 / 1.3 RM3A 10 < 300 20 / 30 60 / 60 1.2 / 1.3

Samples were also sent to Lightnin and Ekato for agitator selection. 16.3.10 Viscosity

Pulp viscosity measurements were conducted at four different pulp densities on each of the samples RM 1, RM 2 and RM 3A using a Bohlin Visco 88 viscometer. At the design pulp density of 45% solids, a pH of 10.5 and at a nominal 25C, the slurries were found to have moderate viscosity and no design issues are anticipated. Samples of cyanide detoxified slurries gave similar results to the CIL samples. 16.3.11 Comparison with Previous Testwork

The initial leaching testwork was found to give lower extractions than anticipated when compared with earlier characterization and variability testwork. The Minproc results are summarized as follows: Cirnic results - Minproc: A leach recovery of 80 to 83% with residue typically 0.30 0.35 g/t Au was achieved for four characterization composites from a head grade of 1.8 2.0 g/t Au. The fifth composite, mainly microconglomerate breccia, achieved 77% recovery with residue approximately 0.35 g/t Au from a 1.6 g/t Au head.

Cetate results Minproc: A leach recovery of 72 to 77% with residue typically 0.45 0.50 g/t Au was achieved for three characterization composites from a head grade of 1.7 1.9 g/t Au. The initial pit composite achieved 66% recovery with residue typically 0.50 0.60 g/t Au from a 1.6 g/t Au head.

108

One sample (argillic/dacite) achieved 81% recovery and 0.30 0.35 g/t Au residue from a 1.7 g/t Au head.

Compared with the Ausenco results: Ausenco program results A leach recovery of 76 to 79% with residues typically in the range of 0.50 0.55 g/t Au was achieved from a head grade of nominally 2.4 g/t Au. This sample was prepared from core representing ore for the first three years of operation (RM1) and was made up of a 70:30 by weight blend of Cirnic:Cetate. Variable leach recoveries were reported from a sample prepared from core representing ore in the fourth year of operation (RM2) with a blend of Cirnic:Cetate of 30:70 by weight. Typically, a recovery of 78 to 81% with residue typically 0.50 0.60 g/t Au was achieved from a head grade of nominally 2.6 g/t Au. The recoveries varied from a low of 69% to a high of 83% Au. Variable leach recoveries were reported from samples prepared from core representing ore in the fifth to seventh years of operation (RM3A, RM3B). Samples came exclusively from Cirnic and Cirnicel, (an extension of the main Cirnic pit). Recovery varied from 61 to 72% with a range of residues from 0.45 to 0.60 g/t Au from a head grade of 1.5 to 1.7 g/t Au. The majority of recoveries achieved were in a narrower range from 64% to 67% Au.

Study of the results and the methodology revealed a number of factors: Minprocs samples were defined by characteristic rock types and alteration profiles, and composited to provide a grade that approximated the mine schedule (1.7 to 1.8 g/t). The Ausenco samples were blends with weights proportional to the tonnages to be mined from each pit over the period represented by the sample. The Ausenco sample grades reflected the then mine schedule which had higher grades in the 2.2 to 2.4 g/t range in the first four years. Higher residue grades for Cirnic ore were reported in the Ausenco work compared to the Minproc work, whilst for the Minproc Cetate characterization samples the residue grades were similar to grades found in the Ausenco program, at 0.45 to 0.60 g/t Au. The natural pH, for pulped samples before addition of lime, in the Ausenco program was typically 2 pH units higher compared to the Minproc work and the Minproc lime consumptions were higher. This suggested that the Minproc samples may have been partially oxidized; however, an analysis by Peter J. Lewis and Associates Pty. Ltd. concluded that this was not a factor in the apparent extraction differences. The grind size for the Minproc work was 90% passing 150 m compared to the design and Ausenco test grind size of 80% passing 150 m. This would be expected to give higher extractions for the Minproc work.

The differing extractions achieved between the Ausenco program and the Minproc program led to a detailed review of the Minproc variability results and the recovery algorithms developed by Minproc. Investigations conducted by Ausenco, Investor Resources Limited, Peter J. Lewis and Associates and Aurifex established that: 109

Oxidation of the Minproc samples did not appear to have influenced the extractions to any level of significance. Minprocs extraction algorithms relied on grouped data sets for various sulphur levels. When the Ausenco composites were plotted with consideration of metal head grade and sulphur grade, they could be seen to fall into the same broad band of results achieved for the variability samples. Minprocs method of averaging the unsorted and un-cut variability data to smooth the results could not be endorsed. The Ausenco testwork results appeared to correlate with the Minproc testwork when the various factors were considered. The extraction algorithms required re-calculation (this is discussed further below). ARD Testwork - Ausenco

16.3.12

Ausenco conducted an acid rock drainage (ARD) treatment testwork program to develop design criteria for an ARD treatment plant. Water samples of ARD generated by the existing Rosia Montana workings were sent to Ammtec in Australia for the program. The program established the dose rates of hydrated lime slurry to various ratios of the ARD and evaluated settling rates of the treatment sludge. The anticipated flowsheet as originally proposed by SNC was mimicked in the laboratory and comprises: Mixing of ARD with hydrated lime slurry in a three stage reactor train. Settling of the sludge in a 28 metre diameter clarifier and re-cycling of some sludge to the head of the circuit to assist seeding of precipitates. The remaining sludge is discharged (to the tailings storage). Overflow is pH adjusted from +10 down to nominally 8.5 in a two stage reactor train and then discharged to local watercourses.

Analysis of the final effluent produced showed that the solution would not comply with the Romanian regulations for the discharge of calcium sulphate. Removal of calcium sulphate to the levels required by the Romanian authorities was not considered practical. It is common to find that the discharges from such facilities exceed such limits and normally dispensation from the regulators is sought. Ausenco considered this to be the preferred course of action and no further work was performed. 16.4 16.4.1 NEWMONT TESTWORK PROGRAM Background

Following its purchase of a significant shareholding in the Rosia Montana project, Newmont undertook a metallurgical testwork program to further explore some of the aspects of the proposed flowsheet that Newmont considered could be improved and to confirm the metallurgical performance of the Rosia Montana ore types. This testwork was conducted after the Ausenco design reviews and could therefore be influential in flowsheet modifications. This is commented on subsequently. 110

Samples were collected from site in a number of locations adjacent to the same sample points that had been used for the collection of the original Minproc variability samples. The samples were sent to the Plato Malozemoff Technical Facility in Englewood, Colorado for processing. 16.4.2 Program - Newmont

Newmont prepared seven Cirnic and five Cetate master composites to be used for various metallurgical studies. The program included: Grind sensitivity. Gravity concentration. Sensitivity to cyanide concentration. Addition of lead nitrate. High pH (high lime) additions. Preg-robbing potential. Variability testing.

The gold grade of the Cirnic master composites ranged from 0.86 ppm to 8.45 ppm. The composites were not preg-robbing, in alignment with previous work by others. The sulphide sulphur content ranged from 1.1% to 1.9%. Silver ranged from 7.5 ppm to 26.8 ppm. The composites did not contain high mercury levels and tellurium contents (a specific interest of Newmonts) ranged from 22 to 34 ppm. The gold grade of the Cetate master composites ranged from 0.99 ppm to 4.12 ppm. The sulphide sulphur content was 1.9% to 3.9%. The silver contents ranged from 5.4 ppm to 15.7 ppm. The composites did not contain high mercury levels and the tellurium contents were 27 to 41 ppm. 16.4.3 Gravity Evaluations Newmont

Newmont ground the various samples to P80 sizes of 170 m (80 mesh), 140 m (100 mesh), 106 m (150 mesh) and 75 m (200 mesh) and used hand panning to recover coarse gold. The gravity tails were then used for various tests to evaluate sensitivity to grind and other parameters. The gravity results showed that the gold recovery increased significantly with finer grind, implying liberation of these particles is very grind dependent. This would suggest diminishing gravity gold returns in that as the grind size is reduced, the liberation of free gold will increase; however, the efficiency of full-scale process equipment will decrease. Similarly, the use of hand panning shows that significant quantities of free gold can be recovered, but the direct comparison of hand panned extractions to those achieved by full-scale equipment cannot be made with any confidence. The gravity gold recoveries are summarized in Table 16.11.

111

Table 16.11 Gravity Recovery as a Function of Grind Size Grind P80 m 170 140 106 75 Cirnic % Gravity Gold Recovery 19.9 21.4 29.8 33.1 Cetate % Gravity Gold Recovery 27.3 30.1 40 (spotty gold?) 34.5

16.4.4

Cyanide Concentration - Newmont

Newmont evaluated the effect of increased cyanide strength and used a concentration range as had been employed by Ausenco. The conclusion was that a cyanide concentration in the 700 and 1,000 ppm NaCN did not influence the gold or silver recoveries by more than approximately 1% and that the increased cyanide levels would most likely not be justified. 16.4.5 Lead Nitrate - Newmont

Lead nitrate trials did not show significant benefit, increasing silver extraction by about 10%. The cost effectiveness of this addition requires evaluation as well as optimization. Plant trials would be expected to be the most effective way of establishing the influence of this reagent. In general, a reduction in cyanide consumption was also noted from the presence of lead nitrate. 16.4.6 Additional Lime to the Leach to Increase the pH - Newmont

The addition of excess lime did not increase gold recoveries for either composite at either grind size. Silver recoveries were increased by approximately 10% for both composites at a 170 m grind. The increase in silver recovery was reduced to around 2% at the 75 m grind. It has been mooted that the lime/high pH may have aided silver extraction from tellurides present in the ore. 16.4.7 Grind Sensitivity - Newmont

Table 16.12 provides a summary of the gravity and cyanidation extractions as a function of grind size. The results present a nominal 4% increase in gold extraction for Cirnic and 6% for Cetate as the grind is reduced from 170 m to 75 m. Silver extraction is some 12% greater for Cirnic and 17% for Cetate for the same range of grinds. There is some indication that between 106 m and 75 m the grind versus recovery curve flattens out. These results are similar in characteristics to those reported by Minproc (Table 16.4) and similar findings as to the most economic grind size would be anticipated if evaluated under the same economic conditions.
Table 16.12 Comparisons of Baseline Gravity + CIL Gold and Silver Extractions vs. Grind 170 m Test A 73.3 65.8 Gold Extraction (%) 140 m 106 m Test A Test A 74.9 75.3 68.1 72.4 75 m Test A 77.6 71.7 170 m Test A 31.6 26.2 Silver Extraction (%) 140 m 106 m Test A Test A 41.8 49.1 31.9 39.2 75 m Test A 43.4 37.7

Grind P80 Cirnic average Cetate average

112

16.4.8

Newmont Testwork Influence on Design

As Newmont did not conduct testwork without gravity, it is difficult to quantify the benefit of gravity as there is no direct comparison with the previous program, even though the overall results could be considered similar. As the samples are not identical (even though similar in location source to Minprocs samples) and the bulk of the Newmont work was conducted at a coarser grind than the Minproc data set, there are some disconnects. The comparison of a hand panned concentrate and the extraction achievable from a full-scale production gravity concentrator are not directly comparable due to the differences in scale and mass split to concentrate. Similarly, it appears (although not confirmed) that the hand panned concentrates have been assigned an extraction of the contained values of 100% in the metallurgical accounting. Such an extraction will not be achieved in a full-scale plant due to the processing steps required to convert concentrates to bullion. Consequently the gravity recovery reported by Newmont would not be expected to be as high in a full-scale plant. This is not to say the gold would be lost, it would report to the downstream CIL for leaching. Similarly, a full-scale circuit is closed by hydrocyclones which will recirculate coarse gold preferentially until it is ground or flattened, minimizing the amount of coarse gold jettisoned by the grinding circuit or providing high surface area particles and certainly having much different physical shape and size distribution than the laboratory milled samples. This work, therefore, does not quantify the benefit that may be achieved by including a gravity circuit. Just the same, in the event a gravity circuit is not included in the final design, the ability to retro-fit a gravity circuit would appear mandatory based on these results. Given the Ausenco results that presented an additional extraction of gold with the inclusion of gravity (around 2 to 3% at typical head grades), this again reinforces the need to provide for a gravity system as either an inclusion in the design or the ability to retro-fit. The Newmont results on the individual samples, in effect a partial repeat of the Minproc variability testwork albeit with the inclusion of the gravity step, were included in recovery algorithm derivation in late 2005 for extraction predictions. Whilst it is arguable whether the data should be included given the differences in flowsheet, the similarity of results and the benefit of additional data points suggested that the Newmont data would be beneficial in providing an improved understanding of the ore behaviour. 16.5 16.5.1 J. R. GOODE AND ASSOCIATES / RMGC TESTWORK PROGRAM - PHASE 1 Background

This program was managed by J. R. Goode with input from both RMGC and SGS-Lakefield Research. Two programs were undertaken. The first, referred to here as Phase 1, made a number of valuable additions to the metallurgical knowledge of the project and provided a scope of work for a second phase of work (reported on below as Phase 2) conducted some time later when detailed design of the process plant recommenced in 2007. The aim of the Phase 1 program, conducted at the SGS Lakefield laboratories at Lakefield, was primarily to investigate and confirm the effects of gravity and coarser grinding on the Rosia Montana 113

ores. Secondary aims were to provide additional design data with regard to gravity and cyanidation circuits, depending on outcomes. Grinding at 150 m (design grind) and comparison at 190 m were the grinds selected for assessment. The program investigated the grind sensitivity combined with gravity recovery, leaching of gravity tails, whole ore leaching, some washing testwork and exploration of variation in leaching time. Seven individual ore composites were prepared with the assistance of RMGC. Four from Cirnic (Cirnic 1, Cirnic 2, Cirnic 3 and Cirnic 4) and three from Cetate (Cetate 1, Cetate 2 and Cetate 3). The results of the program resulted in a second phase of work being scoped and executed which is discussed further in Section 16.6. 16.5.2 Grind Sensitivity and Gravity Recovery J. R. Goode Phase 1

The testwork used a Knelson concentrator to provide a primary gravity concentrate. This was then cleaned up using a Mozley table, thereby giving a mass split to concentrate closer to the type of ratio that could be experienced in the full-scale plant A mass split of nominally 0.05% was achieved, which would be equivalent to a 20 tonne per day production at full scale. The results showed 8% to 39% gold recovery to a Mozley concentrate for the various samples. A low silver gravity recovery of between 1% and 6% resulted. The gold recovery is some 6 to 8 times that of the silver and this ratio was displayed in most tests, suggesting that the gold recovered has a common association with silver. The results showed a weak relationship of coarser grind giving an increased recovery to gravity. The relationship being weak, and the fact that mechanical devices were used compared with some earlier hand-panning tests, could suggest that this observation is a function of machine and operational characteristics, and not metallurgy of the samples. 16.5.3 Extended Leach Times J. R. Goode Phase 1

Extending the leach times from 24 to 48 hours resulted in an additional gold extraction of 0.06 g/t average for the combined Cirnic and Cetate results. This result was determined by a line of best fit applied to the data set of results for gravity recovery followed by CIL leach of the gravity tails. The silver residue was decreased by 0.22 g/t for the same data set when increasing the leach time to 48 hours. 16.5.4 Cyanide Consumption J. R. Goode Phase 1

Comparing the CIL only results with the gravity and CIL leach of the gravity tail showed that the CIL only leach gave significantly higher cyanide and lime consumption. The cause of this was identified as removal of cyanicides (particularly iron) from the sample during the gravity step, due to the quantities of fresh water used. The CIL leach of the gravity tail was then conducted on a pulp low in soluble species giving lower cyanide and lime consumption. The whole ore CIL was conducted with elevated levels including iron resulting from the size reduction steps. 114

This observation raises some points of note: The Ausenco pre-aeration tests showed similar decreases in reagent consumptions which would be due primarily to the removal of soluble iron. Whilst Ausenco allowed for oxygenation there may be benefit in dedicated pre-aeration tankage. This would be expected to assist in the cyanide detoxification as well. Leach extractions conducted with solutions containing high soluble levels of iron may well report lower extractions. Benefit from gravity processing may be due to this washing effect and not due to the removal of coarse gold.

The differences in reagent consumptions are shown in Tables 16.13 and 16.14. 16.5.5 Gravity + CIL Leach of Gravity Tail versus Whole of Ore CIL Leaching Phase 1

The four Cirnic composites showed a trend of increased total extractions for gold and silver when processed via the gravity + CIL leach option compared to a whole ore CIL leach. Subtle differences in grind size and head grade need to be considered when evaluating the results; however, the leach residues provide a good indication of the apparent gravity (or washing) benefit. Table 16.13 summarizes the results for Cirnic. Cetate composites did not show the same benefit, and differences in head grade and grind further complicate the matter. The Cetate results can be considered inconclusive as far as the benefit of the flowsheet is concerned. Table 16.14 summarizes the results for Cetate. Reagent consumptions are observed to increase as the fineness of grind increases and as the sulphur level increases. 16.5.6 Bond Work Indices - Phase 1

Work indices were determined for three of the composites. The results showed little variability and are of the same order as indices determined by previous work. 16.5.7 J. R. Goode Testwork Influence on Design Phase 1

The Phase 1 program managed by J. R. Goode has provided some valuable information with regard to influence on design: The observations with regard to the washing effect and the removal of soluble species prior to leaching (primarily soluble iron) provides opportunity to reduce operating costs by way of reduced reagent consumption in the CIL and the detoxification facility. Whilst washing of the pulp is not a practical option, this mechanism may well be achieved by the pre-aeration step. There was inadequate work completed by the Phase 1 program to be able to justify flowsheet modifications and this provided direction for the scope of the Phase 2 program.

115

Table 16.13 Comparison of Cirnic Gravity + CIL and Whole of Ore CIL Size K80 m Cons NaCN kg/t Final Residue Mean g/t g/t Au Ag g/t Au Au Au Note Cirnic #1 samples have been leached for differing leach times Cons CaO kg/t 1.93 1.94 80.1 82.6 47.2 51.6 0.39 0.38 0.39 0.39 0.39 0.39 % Extraction CN Head g/t Ag calc. g/t Au calc. g/t Ag direct ST

Test No.

Sample

Grav + CIL G2-CN1 24 h GS-CN2 48 h

Cirnic #1 Cirnic #1

174 174

0.29 0.47

6.3 6.7

1.20 1.35

11.5 13.3

1.6 1.6

Whole Ore CN CN29 21 h CN43 24h CN30 24 h Grav + CIL G4-CN8 G6-CN12 G8-CN16 Average Whole Ore CN CN32 CN34 CN36 Average

Cirnic #1 Cirnic #1 Cirnic #1

166 166 166

0.89 2.01 63.4 42.8 0.45 0.48 0.52 1.77 62.4 41.6 0.85 0.49 1.09 2.01 58.9 43.3 0.74 0.91 Cirnic #2, #3 and #4 samples averages only relate to these samples 0.70 0.50 0.74 0.65 1.31 1.31 1.41 1.34 80.7 82.1 74.1 59.9 58.2 59.3 0.37 0.33 0.32 0.38 0.35 0.33

0.47 0.67 0.83

7.0 7.7 7.4

1.27 1.78 2.01

12.2 13.2 13.1

1.6 1.6 1.6

Cirnic #2 Cirnic #3 Cirnic #4

132 134 128

0.38 0.34 0.33 0.35

3.2 3.0 3.7 3.3

1.73 1.74 1.04

7.8 7.1 9.0

1.9 1.7 1.8

Cirnic #2 Cirnic #3 Cirnic #4

149 153 140

1.06 0.99 0.99 1.01

1.79 1.52 1.88 1.73

76.8 79.0 69.0

54.7 57.0 51.0

0.38 0.41 0.29

0.40 0.53 0.31

0.39 0.47 0.30 0.39

3.6 3.1 3.8 3.5

1.68 2.23 0.97

8.0 7.2 7.8

1.9 1.7 1.8

Table 16.14 Comparison of Cetate Gravity + CIL and Whole of Ore CIL Size K80 m Cons NaCN kg/t Cons CaO kg/t % Extraction Au Ag g/t Au Final Residue Mean g/t g/t Au Au 0.40 0.25 0.51 0.37 0.24 0.46 0.36 CN Head g/t Ag calc. g/t Au calc. g/tAg direct ST

Test No. Grav + CIL GI0-CN20 G12-CN24 G14-CN28 Average Whole Ore CN CN38 CK40 CK42 Average

Sample

Cetate #1 Cetate #2 Cetate #3

131 146 135

0.64 0.73 0.74 0.70

1.40 1.29 1.69 1.46

85.5 79.9 78.4

52.5 46.9 51.5

0.34 0.23 0.41

4.8 2.1 5.2 4.0

1.85 1.01 1.90

9.5 3.9 10.6

2.4 2.2 3.0

Cetate #1 Cetate #2 Cetate #3

145 149 129

1.30 1.08 1.42 1.27

2.90 2.31 2.82 2.68

79.2 80.3 82.1

52.4 42.8 52.1

0.36 0.24 0.39

0.51 0.24 0.47

0.44 0.24 0.43 0.37

4.6 2.2 5.2 4.0

2.10 1.22 2.40

9.7 3.8 10.9

2.4 2.2 3.0

The benefit of the gravity step was not definitive. It was decided at the time that the flowsheet should stay as planned (no gravity circuit) yet with room for a retro-fit gravity circuit if plant trials suggest economic benefit. As per the Newmont work, the gravity concentrate is attributed a 100% recovery of the contained gold. For the same reasons as described for the Newmont work, this will not be achieved in practice and therefore the gravity extraction values quoted are inflated. A 20 t/d concentrate treatment circuit will have significant capital and operating cost associated with it. An extended leach time over 24 hours is most likely not justified.

116

16.6 16.6.1

J. R. GOODE AND ASSOCIATES / RMGC TESTWORK PROGRAM PHASE 2 Background - Phase 2

This program was managed by J. R. Goode with input from both RMGC and SGS-Lakefield Research. The aim of this program, conducted at the SGS Lakefield laboratories at Lakefield, was to investigate and quantify the effect of pre-aeration and high pH on solution chemistry and cyanide credits in what would be tailngs thickener return water (mill solution) in the full-scale process. The opportunity was undertaken to repeat gravity recovery assessments as part of the program in addition to providing design data should gravity concentration and intensive cyanidation be employed in the flowsheet. Samples of Cirnic and Cetate ores as were used in the Phase 1 work were again used for the Phase 2 work. 16.6.2 Gravity Separation Phase 2

A number of 10 kg charges of the Cirnic composite were subjected to Knelson concentration and the concentrates then upgraded using a Mozley table. The upgrade ratio was similar to that which would be achieved in a full-scale gravity circuit thereby addressing some of the issues associated with earlier gravity work. A 35% gravity gold recovery was achieved to 0.044% of the feed mass. This equates to a nominal 20 t/d concentrate production rate from a full-scale operation which is a practical tonnage to subject to intensive cyanidation processes. A similar program evaluated Cetate composite samples resulting in a gravity recovery of 36% to 0.055% of the feed mass. 16.6.3 Cyanidation Testwork Phase 2

Cyanidation leaching was undertaken on whole ore and on the gravity concentrates, middlings and tailings samples. The whole ore leaching for the Cirnic and Cetate composites presented similar results of 79.1% and 80% gold recovery, respectively. The intensive cyanidation of the various high grade gravity concentrates presented extractions in excess of 99% in all cases suggesting that intensive cyanidation would be a valid process route for such products. The leaching of the gravity tail presented extractions as anticipated; however, the combined leach extraction was found to be slightly higher than the whole ore leach that excluded gravity. The results are summarized in Table 16.15.
Table 16.15 Comparison of Whole of Ore Leaching and Gravity/Tail Leaching Composite Gravity Concentrate % Mass Overall (P2) Cirnic MC Cetate MC 0.044 0.055 0.051 % Au dist 35.1 36.4 50.4 Whole ore -79.1 80.0 Overall % Au Extraction by Cyanidation of: Garvity Conc 34.9 36.1 49.9 Gravity Tail 46.0 44.6 32.0 Grav Conc CN + Grav Tail CN 81.0 80.7 82.0

117

16.6.4

Iron Chemistry Testwork

Several tests were conducted to study the impact of quicklime addition and pH on the oxidation and release of iron during grinding and the influence this has on cyanide consumption. The tests evaluated the influence of time in addition to the oxidation state of the iron and the ratio of the ferrous and ferric species. The first round of work showed that significant amounts of ferrous iron are released upon contact with water as would occur in a grinding mill. The Cetate sample was the highest in iron release. The work also showed that any cyanide present was rapidly consumed. The second and third rounds of work included quicklime additions to simulate the proposed addition of quicklime to the mill feed as was present in the process plant design. This work showed that the cyanide present was preserved if the ore was mixed with the water at a high pH, nominally greater than 9.6, with some 70% of the cyanide being retained. 16.6.5 J. R. Goode Testwork Influence on Design Phase 2

The Phase 2 program managed by J. R. Goode has provided some valuable information with regard to influence on design: The program reinforced that the likely benefit of gravity concentration would be in excess of 1% extraction based on these programs and that of Ausenco. Given the economic conditions of the time (elevated gold price), the amenability of the concentrates to intensive cyanidation and the fact that the work represented practical concentrate production from a full-scale plant, it would be cost effective to implement gravity concentration as part of the flowsheet and plant design. The program suggested that the recirculation of tailings thickener water (mill solution) would result in a considerable saving in cyanide consumption. Indications are that 70% or more of the cyanide recycled would be utilized if grinding was conducted at high pH in the SAG mill. This would further suggest that SAG milling should also preferentially be undertaken in process water with mill solution (thickener overflow) typically added to the mill discharge hopper after iron precipitation has all but completed. A dedicated pre-aeration tank would not be required. DESIGN CRITERIA

16.7

The key design criteria as used by Ausenco should be retained based on the testwork to date, with the addition of a gravity circuit and intensive cyanidation plant to process the concentrate. The criteria are summarized in Table 16.16.

118

Table 16.16 Process Design Criteria Criterion Design Basis Nominal plant throughput Gold head grade (mass balance) Silver head grade (mass balance) Gold head grade (design) Silver head grade (design) Ore Characteristics Ore specific gravity Unconfined compressive strength (soft to hard ore) Bond Crushing Work Index (soft to hard ore) Bond Rod Mill Work Index (soft to hard ore) Bond Ball Mill Work Index (soft to hard ore) Abrasion Indices (soft to hard ore) Comminution Primary crusher type Primary crushed ore stockpile capacity Grinding circuit type SAG mill power installed Ball mill power installed Classification Pebble crusher type Gravity circuit Leaching and Adsorption Circuit type Nominal throughput Residence time Pulp density Design gold extraction Design silver extraction Number of contact stages Configuration Carbon loading, gold Carbon loading, silver Desorption and Regeneration Circuit type Capacity Percentage of carbon to be regenerated Tailings Thickening and Detoxification Thickener type Unit thickener rate (minimum) Rise rate (maximum) Flocculant dose Thickener underflow density Detoxification feed density post dilution Detoxification residence time (minimum) ARD Treatment Design inflow of ARD Configuration Neutralization tankage Polishing tankage Thickener type Rise rate (maximum) Flocculant dose (new solids) Unit tonnes/annum g/t g/t g/t g/t Value 13,000,000 1.62 7.3 2.20 15.0 2.61 54 - 222 6.4 9.4 14.0 20.3 13.8 19.5 0.11 0.62 Gyratory 39,000 live SABC 15,000 2 x 10,000 Hydrocyclones Cone Centrifugal primary and Intensive cyanidation CIL 1,625 24 40 42% 85 65 7 2 parallel 7 tank trains 1,400 6,300 Continuous acid wash and Zadra 4 75 High rate 1.4 5.4 20 60 50 1.5 (not provided by Ausenco) 400 2 stage with recycle 3 x 1 h residence 2 x 1 h residence High rate 1.0 10

MPa kWh/t kWh/t kWh/t kWh/t

t kW kW

t/h h wt% % %

g/t g/t

t/h %

t/h/m2 m/h g/t wt% wt% h m3/h

m/h g/t

119

16.8

FLOWSHEET

A summary flowsheet is provided in Figure 16.1, displaying the major unit processes and configuration planned for the processing plant at Rosia Montana. With exception of the use of continuous elution, the process is conventional and consists of crushing, grinding, gravity concentration, cyanide leaching in a CIL circuit and recovery of gold and silver by absorption on to carbon, electrowinning and smelting. 16.9 EXTRACTION ALGORITHMS

The Minproc variability program provided variability results from some 560 samples that allowed the generation of recovery algorithms. These algorithms were then used as inputs to aid the development of a mine plan and financial model. Minprocs original flowsheet comprised comminution to a P80 of 150 m followed by whole ore CIL leach for 24 hours. The variability samples on which the Minproc recovery algorithms were based were processed to this flowsheet. Further flowsheet development and economic considerations at the then gold price of US$280 per ounce and processing rate of 20 Mt/y led Minproc to revise the flowsheet. This revised flowsheet was modified to include a basic gravity circuit for coarse gold removal and a reduction in CIL residence time to a nominal 15 hours. SNC undertook the basic engineering phase of the project. During 2001-2002, SNC conducted testwork to finalize a number of design parameters as well as to further investigate the gravity circuit options. The final SNC flowsheet employed jigs to recover a sulphide concentrate. This concentrate was subjected to fine grinding followed by recombining with the gravity tail/cyclone overflow for leaching in a CIL circuit, retaining the 15 hour residence time. The variability testwork data based on a 24 hour CIL circuit with no gravity were used by Minproc to develop recovery algorithms. The flowsheet evolution through to the final SNC version resulted in these algorithms being manipulated a number of times to describe the recovery differences from one phase of development to the next. These manipulated algorithms originating from the Minproc work were the basis of the project financial model up until mid-2004. Ausenco conducted a review of the flowsheet in 2004 with consideration given to the economic factors of the time. The assessment suggested that the original 24 hour residence CIL circuit with no gravity as originally proposed by Minproc was the most suitable. A confirmatory testwork program on composite samples representing the first seven years of production was completed early in 2004. The extraction results obtained from this program did not match those anticipated based on the previous Minproc results as discussed above. A review of the Minproc data and Ausenco data commenced with the aim of identifying the reasons for the apparent discrepancy. These investigations found that the Ausenco results were within the bounds of the variability displayed by the previous Minproc results when due consideration was given to gold head grade and sulphur relationships.

120

Figure 16.1 Simplified Process Flowsheet (Page 1 of 2)

121

Figure 16.1 Simplified Process Flowsheet (Page 2 of 2)

122

The investigations also identified a number of issues with the algorithm development methodology applied by Minproc and later extended by SNC. The investigations could not define the impact to the project economics without further work, but it appeared that the Minproc/SNC algorithms could be overstating the recoveries, particularly early in the mine life. Revised equations for the prediction of process recovery were developed by Aurifex and Geostats Pty. Ltd. (Geostats). These were compared with the recovery predictions used for the financial model and with kriged metallurgical test results as applied to the block model of the time. A number of iterations in the development of the algorithms resulted until a final suite of equations was selected. These final algorithms were again compared with kriged metallurgical data to establish the degree of error/variation and to provide some level of confidence in the predictions. The most suitable algorithms developed by Aurifex and Geostats were then selected for final recovery estimations of the total reserve and input to a revised financial model in July, 2004. Further refinement since 2004 has led to a revised set of algorithms. These revised algorithms have been applied to the current mine plan and economic model. The recovery algorithms are: Gold Equations Cetate Cirnic Jig Orlea Residue = (((0.0685*S+0.1796)*Au^(0.7423-0.092*S)) + 0.004) Residue = (((0.1315*S+0.0569)*Au^(0.64-0.0725*S)) + .000) Residue = (0.02 + 0.08*S + 0.17*Au - 0.053) Residue = (((0.0928*S-0.0313)*Au^(0.4013+0.1772*S)) + 0.018)

Silver Equations Cetate Cirnic Jig Orlea Residue = (((0.0539*S+0.372)*Ag^(0.871+0.0276*S) - 0.026) Residue = (((0.1146*S+0.1954)*Ag^(0.0266 0.00287*S) + 0.028) Residue = (-0.29 + 0.46*S + 5.4/Ag + 0.268) Residue = 0.44*Ag

123

17.0
17.1 17.1.1

MINERAL RESOURCE AND MINERAL RESERVE ESTIMATES

DATABASE DEVELOPMENT Introduction

The mineral resource estimate is based on the data collected up to and including 2005. Additional underground channel sampling has been collected post this date, but the additional sampling (Section 11; approximately 1,838 additional samples) is not considered material to the global mineral resource. These data should be included in the estimation database for future study once additional data are collected. The database used for resource estimation studies was compiled by RMGC technical staff on site and independently validated by RSG Global geological staff in Perth. The data from the initial exploration program were compiled on site using Microsoft Excel, Vulcan, Microsoft Access and Micromine. In 2003, the then current database was transferred into a full relational database using the acQuire software product. Most of the data from the most recent exploration program have been compiled and validated on site using acQuire. The database channel and drill hole information for the current resource estimate were exported by the RMGC site database manager to a series of comma delimited files and forwarded to the RSG Global Perth office, along with a series of Excel files containing the quality control data. Detailed surface topography models (based on contour, breakline and survey points) over each deposit were also supplied. RSG Global undertook a detailed review and validation of the supplied channel and drill hole database files using MS Excel, Micromine and GeoAccess software products, prior to loading separate channel and drill hole databases into the Vulcan software package, wherein further visual validation of the data was carried out prior to the commencement of resource modelling. 17.1.2 Validation of Supplied Database

The following database validation activities were carried out by RSG Global: Checking of underground and surface channel sampling traces against the locations of the surveyed underground workings. Surface channels that were noted to deviate substantially from the surveyed topography have been adjusted to the topography where appropriate. Ensure compatibility of total hole depth data in the collar, survey, assay and geology database files. Checking of drill hole survey data for unusual or suspect downhole deviations. Ensure sequential downhole depth and interval data in the survey, assay and geology files. Checking of high-grade assays in the primary gold and silver assay fields against the laboratory assay reports. Replacement of less than detection limit assays with 0.005 g/t Au and 0.5 g/t Ag, character entries, and blanks for unsampled intervals with nominal low-grade values (Table 17.1). 124

Checking of lithology and alteration codes. Removal of non-essential information from validated database files.
Table 17.1 Rosia Montana Gold Project Assay Database : Character Replacements Numeric Replacement Gold (g/t) 0.005 0.001 0.002 Silver (g/t) 0.5 0.001 0.002

Character <0.01 ns is/dip/nr

No anomalies were identified in the manual cross-checks of the digital assay data and hard copy assay certificates. No material errors were identified during the RSG Global database validation activities. 17.1.3 Database Development

The validated drill hole and channel database files were imported into the Vulcan software package as two databases: Drill hole database, containing collar, survey, assay, geology, vein class, density, and recovery/RQD tables. Channel database, containing collar, survey, assay, geology, and density tables.

Summary statistics of the contents of the master resource databases, including metres of channel sampling and drilling, metres sampled, and the number of assayed samples, subdivided into collection methodology, are given in Table 17.2. Note the following abbreviations: RC: RC/DDH: DDH: Reverse circulation drilling. Reverse circulation drilling with a diamond core tail. Diamond drill hole.
Table 17.2 Rosia Montana Gold Project Mineral Resource Estimate - Drilling and Channel Sampling Summary RC Holes 629 Samples Metres 75,436 73,343 RC / DDH Holes 131 Samples Metres 29,237 28,605 Surface DDH Holes 276 Samples Metres 20,085 14,903 Underground DDH Holes 72 Samples Total Channels Channels 1,688 Samples Metres 71,952 62,754 Metres 11,820 11,715 Holes 1,108 Samples Total Metres 136,578 128,566

Surface Channels Channels 387 Metres 8,884

Underground Channels Channels 1,301 Metres 63,068

Total Channels / Holes Channels 2,796 Samples Metres 208,530 191,320

Samples 7,760 Samples 54,994 Notes: RC: Reverse circulation drilling. RC/DDH: Reverse circulation drilling with a diamond core tail. DDH: Diamond drill hole. Excludes 2006 to 2008 data.

In addition to the gold and silver data a total of 8,672 sulphur composites were available for investigation. 125

A full listing of drilling and channels used in this resource estimation study is included as Appendix 9. 17.2 17.2.1 GEOLOGY AND MINERALIZED DOMAIN MODELLING Introduction

Lithology, oxidation and mineralized domain boundaries have been interpreted and wireframe modelled to constrain resource estimation for the Rosia Montana project. Modelling was initially completed by RMGC technical staff and then validated in detail by RSG Global in Perth. The interpretation and wireframe models were validated by RSG Global using the Vulcan software package. The project area has been subdivided into nine estimation regions, based primarily on the spatial distribution of the historic mining centres. The estimation regions are consistent with previous studies and allow the effective reporting of the resource model but do not represent unique or major changes in geology and/or grade characteristics. The estimation regions (Figure 17.1) constructed are as follows: 17.2.2 Orlea (code 10). Carpeni (code 20). Cetate (code 30). Cirnicel (code 40). Cirnic (code 50). Cos (code 60). Jig (code 70). Igre (code 80). Tarina (code 90). Lithological Model

A total of 31 unique lithology codes were recorded in the detailed geological logging undertaken. The highly detailed geological subdivisions represented in the drill hole logs were considered unsuitable at the scale of the resource study, and were consequently grouped into 8 major subdivisions with the following major lithologies and wireframe solids constructed by RMGC and reviewed by RSG Global (Figure 17.2): Dacite. Vent breccia. Corna breccia. Internal breccia (breccia zones occurring in Cirnic).

126

Figure 17.1 Estimation Regions with Drilling Data

127

Figure 17.2 Perspective View of the Geological Model

17.2.3

Cretaceous sediments. Black breccia. Andesite. Polymict breccias. Alteration Model

RMGC updated the alteration models for the 2005 resource update. For the majority of estimation regions, an alteration model is used, grouping quartz-adularia alteration or silicic zones (SIK), against data with little or no quartz-adularia alteration (NSIK). Silicic and potassic alteration recorded with an alteration intensity of moderate, strong or intense was considered SIK and the remaining was coded as NSIK. During 2005, RMGC added a third alteration phase which can be mapped and modelled. The third phase incorporates QIP (quartz-illite-pyrite) alteration which is essentially intermediate between the quartz-adularia (SIK) and the argillic zones. This new intermediate alteration code is only applied to the Cetate and Cirnic regions. Regions of strong quartz-adularia alteration are also modelled in the Cirnic area. 128

Major zones of SIK/QIP alteration are interpreted to occur at Cetate, Cirnic/Cirnicel, Igre, Jig and Orlea. At Cetate, the QIP alteration zone is interpreted as subvertical, with a northeast-southwest strike. At Cirnic, the previous separated SIK zones were combined into a single QIP zone. These zones are interpreted to coalesce to the north of Cirnic, and the SIK/QIP zone that traverses Gauri and Cetate is considered likely to extend through to Jig, directly northeast along strike. East-west trending alteration zones are interpreted at Orlea and Carpeni. At Igre the SIK alteration has been interpreted to be associated with structure and the polymict breccia A perspective view of the alteration zone wireframes is provided in Figure 17.3.
Figure 17.3 Perspective View of the Silicic Alteration Zone Wireframes

17.2.4

Mineralization Modelling

The mineralization zone model developed for the 2001 feasibility study and the 2002 resource estimation study has not been updated for the current study. The majority of the mineralization models were constructed for the feasibility study, with only the Orlea and Igre zones updated in 2002. The interpretations were reviewed using the new drill and sample data and still found to be a suitable representation of the mineralization. The general approach taken to constructing the mineralization zone interpretation is based on capturing regions of anomalous gold mineralization incorporating a notional 0.3 g/t Au lower cutoff grade and geology. Mineralized zone interpretation based solely on a lower cutoff grade was rejected as the nature of the mineralization is such that significant grade variation is noted and elevating the lower 129

cutoff grade results in poor continuity of mineralization. The apparent lack of continuity is especially evident at lower cutoff grades above 0.6 g/t Au as displayed in Figures 17.4 through 17.7. The final modelled mineralization zones previously developed and used for the current study are presented in Figure 17.8. More detail on the criteria used to construct the mineralization zones can be found in previous reports.
Figure 17.4 Gold 5 m Composites Indicator Coding

(Red >=0.1 g/t Blue < 0.1 g/t Au)

130

Figure 17.5 Gold 5 m Composites Indicator Coding

(Red >=0.3 g/t Blue < 0.3 g/t Au)

Figure 17.6 Gold 5 m Composites Indicator Coding

(Red >=0.6 g/t Blue < 0.6 g/t Au)

131

Figure 17.7 Gold 5 m Composites Indicator Coding

(Red >=0.8 g/t Blue < 0.8 g/t Au)

17.2.5

Hardness Modelling

A hardness coding appropriate for processing was updated in December, 2005 by RMGC based on the original hardness samples and composites from the Minproc feasibility study conducted in 2001 (Table 17.3). Updated alteration and rock type codes were used to assign appropriate hardness categories (hard, medium and soft) to the updated geological and alteration model based on the original testwork. The new model consisted of three main alteration phases, quartz-adularia (silica), QIP (quartz-illite-pyrite) and argillic (including phyllic) based on the post-doctorate mapping studies conducted by Dr. Scott Manske of the Colorado School of Mines. The project defined a third and additional (transition) phase of alteration (QIP) over the previous model which consisted of silicified (including potassic) and argillic alteration. The processing throughput rates were confirmed and refined to incorporate the findings of the SNC metallurgical testwork and design, as well as the later Ausenco testwork and design updates based on the 13 million tonne per year plant. Wireframes were constructed for the model based on the original variability and composite samples used for the Bond Work Index testwork and confirmed with underground and surface re-mapping as well as the re-logging of drill core to constrain the hard and medium hard rock types. All material outside of these is considered to be soft in hardness.

132

Figure 17.8 3D view of the Interpreted Mineralization Zones

133

Table 17.3 Rosia Montana Gold Project Hardness Characteristics Subdivided by Lithology and Rock Type Ore Type Lithology and Alteration Plant Throughput ( t/h) Hard si+po Dacite si+po Monomictic breccia 1,300 Medium qp Dacite qp Monomictic breccia si+po+qp Polymictic breccia 1,625 Soft All ph-ar+ar alteration and all vent breccia ore 1,950

In summary, hardness was modelled at Rosia Montana such that: Parts of the Carpeni region (dacite) were modelled as medium. Parts of the Cetate dacite and XPO were modelled medium and hard. Parts of the Cirnic dacite were modelled as medium and hard. Parts of Igre were modelled as medium. All SIK material at Jig was considered as hard. All or the remainder of Orlea, Carpeni, Cetate, Cirnic, Cirnicel, Jig NSIK and Igre were considered soft. Validation of Geological Interpretation and Wireframe Models

17.2.6

The geological constraints interpreted and modelled in the current studies were reviewed and validated in detail, with the following validation activities undertaken: 17.3 17.3.1 Review all modelled boundaries in cross-section, long section, oblique section and plan views. Ensure the integrity of all wireframe surface and solid models and that wireframe solid models are closed. STATISTICAL ANALYSIS Introduction

Extensive statistical analysis was undertaken based on 5 m gold and silver composites, the 5 m sulphur samples, and the bulk density data for the drilling and channel sampling completed at Rosia Montana. The activities completed in this phase of the study were as follows: Coding of the drill hole and channel sampling databases based on the geological interpretation. Compositing of the drill hole and channel sampling data to 5 m unit lengths. Compilation of descriptive statistics and histogram plots of the composite gold, silver, sulphur and bulk density datasets, subdivided by the geological interpretation. Outlier grade analysis and determination of upper cuts for gold and silver (completed pre and post compositing).

134

Assessment of data clustering and calculation of declustered grade statistics. Correlation analysis between composite gold and silver datasets subdivided by estimation domain/group.

17.3.2

Data Coding

The wireframe models of the regions, lithology, alteration and mineralization domains were used to assign a series of codes into the drill hole and channel sampling databases to allow assessment of the variations in grade and bulk density amongst the various domains. The coding applied to both databases is summarized in Table 17.4. The lithology, oxidation, mineralized domain and mineralized region coding assigned to the drill hole and channel samples was visually compared with the corresponding wireframe boundaries in cross-section and plan views to ensure that all coding was robust. No errors were detected. Domains are commonly referred to using region names, followed by alteration style, for example, Cetate QIPSI refers to the region where the Cetate mineralization is altered to quartz-illite-pyrite and is silicified, whereas Cetate NQIPSI refers to the region where the Cetate mineralization is unaltered. A series of underground channel samples were highlighted by RMGC to have been sampled along veins and considered not representative of the local grade. These intersections have been coded into a variable called VEIN using the binary code of 1 = sampled along a vein and 0 = all other composites. The composites sampled along veins are only used to estimate cells that they are touching. The veins are only located in the Orlea, Cetate and Cirnic regions (Figure 17.9). 17.3.3 Compositing

The lengths of the samples were statistically assessed prior to selecting an appropriate length to composite the data in preparation for undertaking detailed statistical analyses, variography and grade estimation. Figure 17.10 indicates that at Rosia Montana, approximately 97% of the drill hole samples and >99.99% of the channel samples have been collected over 1 m. Most of the remaining samples have been collected over various intervals ranging 0.5 m to 1 m long. The drill hole and channel databases coded by the geological interpretation were composited as a means of achieving a uniform sample support. It should be noted, however, that equalizing sample length is not the only criterion for standardizing sample support. Factors such as angle of intersection of the sampling to mineralization, sample type, sample diameters, drilling conditions, recovery, sampling/sub-sampling practices and laboratory practices all affect the support of a sample. Exploration/mining databases which contain multiple types and/or sources of data provide challenges in generating composite data with equalized sample support and uniform support may be difficult to achieve. A 5 m unit length was used for data compositing, allowing subsequent selective mining resource estimation to be conducted. Any composites less than 2 m long were excluded for the remainder of the study.

135

Table 17.4 Rosia Montana Gold Project Domain Coding Domain Description default Orlea Region Carpeni Region Cetate Region Cirnicel Region Region Cirnic Region Cos Region Jig Region Igre Region Tarina Region No Rock Dacite Dacite Dacite Dacite Dacite Dacite Dacite Dacite Dacite Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Lithology Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Phreatic Breccia Phreatic Breccia Phreatic Breccia Sandy Breccia Cretaceous Sediments Andesite Vent Breccia (Default) Corna Vally Breccia Black Breccia Waste Dumps Unalteration (default) Waste Dumps quartz-adularia quartz-adularia quartz-illite-pyrite quartz-adularia Alteration quartz-adularia quartz-adularia quartz-adularia quartz-illite-pyrite silicification silicification Unmineralised (Default) Waste Dumps Orlea Mineralization Carpenic Mineralization Cetate Mineralization Cetate East Mineralization Cetate East Mineralization Cetate East Mineralization Gauri Mineralization Mineralization Cirnic Mineralization Jig Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Rock (Default) Air Waste Dumps Air Type Name (*.00t) region_orlea_ros017 region_carpeni_ros017 region_cetate_ros017 region_cirnicel_ros017 region_cirnic_ros017 region_cos_ros017 region_jig_ros017 region_igre_ros017 region_tarina_ros017 topo_jul04_ros017 VDA_cos_ros017 VDA_jig_ros017 VDA_carp_ros017 VDA_CLAST1_ros017 VDA_CLAST2_ros017 VDA_cirnicel_ros017 VDA_cir_ros017 VDA_cet_ros017 nda_gaur_02a_ros017 igre_XPO_ros017 igre_XPO2_ros017 igre_XPO3_ros017 igre_XPO4_ros017 XPO1_cir_ros017 XPO2_cir_ros017 XPO3_cir_ros017 XPO1_cet_ros017 XPO2_cet_ros017 XPR1_cir_ros017 XPR2_cir_ros017 XPR3_cir_ros017 XCF_cir_ros017 SSE_uppersurf_ros017 VAN_lowersurf_ros017 corna_bx_ros017 XBB_ros017 wastebase_ros017 wastebase_ros017 gauri_sik_west_ros017 gauri_sik_ros017 cetate_qipsi_ros017 orlea_sik_ros017 jig_sik_ros017 igre_sik_ros017 carpeni_sik_ros017 cirnic_qipsi_ros017 cirnic_si1_ros017 cirnic_si2_ros017 wastebase_ros017 orlea_min_ros017 carpeni_min_ros017 cetate_min_ros017 cetate_east_min1_ros017 cetate_east_min2_ros017 cetate_east_min3_ros017 gauriw_min_ros017 cirnic_min_ros017 jig_min_ros017 igrmin4_ros017 igrmin5_ros017 igrmin6_ros017 igrmin7_ros017 igrmin8_ros017 igrmin9_ros017 igrmin10_ros017 wastebase_ros017 topo_jul04_ros017 Wireframe Description Orlea Solid Carpeni Solid Cetate Solid Cirnicel Solid Cirnic Solid Cos Solid Jig Solid Igre Solid Tarina Solid Topography Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Top of Cretaceous Sediments Base of Andesite Breccia Solid Breccia Solid Base of Waste Dumps Base of Waste Dumps Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Base of Waste Dumps Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Base of Waste Dumps Topography Variable Name region region region region region region region region region region lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith alt alt alt alt alt alt alt alt alt alt alt alt zone zone zone zone zone zone zone zone zone zone zone zone zone zone zone zone zone zone rock rock rock Code 0 10 20 30 40 50 60 70 80 90 0 50 50 50 50 50 50 50 50 50 150 150 150 150 150 150 150 150 150 160 160 160 165 250 350 450 500 550 850 0 5 98 99 100 100 100 100 101 102 200 200 0 5 10 20 30 31 32 33 34 50 70 80 81 82 83 84 85 86 10 100 1000

136

Figure 17.9 Channels Sample Along Veins

Figure 17.10 Log Probability Plot of Sample Intervals Drilling

137

17.3.4

Statistical Analysis of Composite Data - Gold and Silver

Detailed statistical analysis of the Rosia Montana composite gold, silver and sulphur data has been conducted, sub-divided by region and alteration. Descriptive statistics of the composite gold data subdivided by the region and alteration domains are presented in Table 17.5. The QIP/SI and SIK domains have the highest mean gold grades while the Jig SIK has the overall highest gold grade. The Igre 9 domain has the lowest mean gold grade. The relative variability of the grades indicated by the coefficient of variation values (CV; calculated by dividing the standard deviation by the arithmetic mean) is highest for Carpeni and the Cetate NQIPSI compared to the other datasets.
Table 17.5 Rosia Montana Gold Project Summary Statistics - Gold Orlea SIK Count Minimum Maximum Mean Median Standard Deviation Variance Coefficient of Variation 962 0.01 48.18 1.39 0.80 2.56 6.55 1.84 Jig SIK Count Minimum Maximum Mean Median Standard Deviation Variance Coefficient of Variation 266 0.063 25.29 2.80 2.17 2.63 6.93 0.94 NSIK 114 0.005 3.12 0.60 0.37 0.66 0.43 1.10 4 194 0.013 7.29 1.05 0.67 1.14 1.29 1.09 5 132 0.005 9.88 0.89 0.64 1.25 1.57 1.41 6 1267 0.005 5.83 0.62 0.31 0.88 0.78 1.43 NSIK 1,139 0.019 46.85 0.69 0.36 2.01 4.05 2.91 Carpeni ALL 1,819 0.005 272.34 0.91 0.39 6.64 44.12 7.33 QIPSI 5,200 0.016 179.80 1.44 0.63 4.87 23.72 3.39 Cetate NQIPSI 3,982 0.009 168.59 0.79 0.35 4.41 19.42 5.56 Igre 7 215 0.042 25.06 1.38 0.59 2.33 5.44 1.69 8 418 0.006 27.44 0.85 0.35 2.22 4.93 2.60 9 68 0.014 2.04 0.31 0.20 0.34 0.11 1.08 10 37 0.03 8.33 1.02 0.28 1.78 3.16 1.75 East 829 0.02 30.53 0.81 0.40 1.60 2.54 1.96 Gauri ALL 143 0.013 5.80 0.51 0.24 0.81 0.65 1.59 Cirnicel ALL 967 0.006 63.22 0.93 0.48 2.64 6.99 2.83 Cirnic QIPSI 6,941 0.01 112.12 1.49 0.84 3.27 10.70 2.19 All 167 0.02 4.23 0.62 0.50 0.52 0.28 0.85 NQIPSI 3,861 0.005 129.47 0.77 0.38 3.10 9.62 4.02

Waste Dumps

Log probability plots (significant domains in Figure 17.11) of these datasets indicate that they all form positively skewed distributions typical of gold deposits. Individual log histogram plots and probability plots for each domain are also presented in Appendix 10. Descriptive statistics of the composite silver data subdivided by the region and alteration domains are presented in Table 17.6. The QIP/SI and SIK domains have the highest mean silver grades while the Jig SIK has the overall highest mean silver grade followed by Cetate QIPSI and Cirnic QIPSI/SI. The Igre 9 domain has the lowest mean silver grade. The relative variability of the grades indicated by the coefficient of variation values is highest for Cetate QIPSI compared to the other datasets.

138

Figure 17.11 Log Probability Plots of Composite Gold Data

Table 17.6 Rosia Montana Gold Project Summary Statistics - Silver Orlea SIK Count Minimum Maximum Mean Median Standard Deviation Variance Coefficient of Variation 962 0.5 14.6 2.2 2.0 1.3 1.6 0.6 Jig SIK Count Minimum Maximum Mean Median Standard Deviation Variance Coefficient of Variation 266 0.9 395.8 26.9 14.2 40.5 1640.2 1.5 NSIK 114 0.5 68.4 5.2 3.3 7.6 58.4 1.5 4 194 0.5 22.4 3.2 2.2 2.9 8.4 0.9 5 132 0.5 45.2 6.5 4.6 6.0 35.8 0.9 6 127 0.5 52.0 3.2 2.0 5.4 29.5 1.7 NSIK 1,139 0.5 63.2 1.9 1.8 2.3 5.4 1.2 Carpeni ALL 1,819 0.5 85.9 2.0 1.4 3.3 10.6 1.6 Cetate QIPSI 5,198 0.5 665.6 9.2 3.6 22.5 507.4 2.5 NQIPSI 3,982 0.5 246.6 3.4 2.0 7.9 61.7 2.3 Igre 7 215 0.5 138.4 4.4 2.8 9.9 98.0 2.2 8 418 0.5 50.8 2.5 2.0 3.5 12.2 1.4 9 68 0.5 28.2 1.5 0.9 3.4 11.8 2.3 10 37 0.5 15.6 2.9 2.4 2.9 8.3 1.0 East 829 0.5 128.6 2.7 2.0 5.0 24.6 1.9 Gauri ALL 143 0.5 38.2 3.1 2.2 4.4 19.0 1.4 Cirnicel ALL 967 0.5 98.0 9.5 6.4 10.3 106.6 1.1 Cirnic QIPSI and SI 6,941 0.5 269.6 9.8 6.8 11.9 140.9 1.2 All 167 0.5 16.7 4.1 3.5 2.4 5.6 0.6 NQIPSI 3,861 0.5 156.0 5.6 3.2 8.4 70.3 1.5

Waste Dumps

139

Log probability plots (significant domains in Figure 17.12) of these datasets indicate that they all form positively skewed distributions typical of silver deposits. Individual log histogram plots and probability plots for each domain are also presented in Appendix 10.
Figure 17.12 Log Probability Plots of Composite Silver Data

17.3.5

Outlier Analysis and Determination of Upper Cuts

Assessment of the high grade gold composites was completed to determine the requirement for high grade cutting for each of the input datasets to be used for resource estimation. The approach taken to the assessment of the high grade composites and outliers is summarized as follows: Assess holes with visible gold and histograms and probability plots of uncomposited samples. Detailed review of histograms and probability plots of 5 m composites, with significant breaks in populations used to interpret possible outliers. Investigation of clustering of the higher-grade data. High grade data that are clustered were considered to be real while high grade composites not clustered with other high grade data were considered to be possible outliers, requiring further consideration via cutting. The ranking of the composite data and the investigation of the influence of individual composites on the mean and standard deviation (mean versus std-dev plots). Plots of all datasets are provided in Appendix 10.

Based on this assessment, a strategy for managing outliers was adopted for Rosia Montana, including: 140

Uncomposited gold samples were cut to 200 g/t based on an inflection in the combined probability plot affecting a total of 23 one-metre samples. Uncomposited gold samples with visible gold were cut to 100 g/t based on the disruption of probability plots for the Cetate and Cirnic QIPSI above 100, affecting 11 samples of between 0.5 and 1 m in length. Uncomposited silver samples were cut to 500 g/t based on the disruption of the combined probability silver plot at 500 g/t, affecting 29 one-metre samples.

The pre-compositing cutting is equivalent to cutting less than seven 5 m gold composites and less than six 5 m silver composites. Following the compositing of the sample data to 5 m lengths a series of high grade cuts or caps were determined and applied to resource estimation as presented in Table 17.7. It should be noted that no gold upper cut was considered necessary for Domain 4, due to the absence of outliers. The gold upper cuts applied to the SIK or QIPSI domains result in cut mean grades ranging from 2% to 11% less than the uncut mean grades. The silver upper cuts applied to the SIK or QIPSI domains result in cut mean grades ranging from 0% to 9% less than the uncut mean grades. 17.3.6 Declustered Grade Statistics

Clustering of the exploration sample data is evident at Rosia Montana. Cell declustering has been undertaken to assess the effects of the data clustering on the global mean grade of each of the defined estimation domains/groups. Table 17.8 presents a comparison of the nave and declustered mean grades for the SIK/QIPSI estimation domains based on the cut composite datasets. It is evident that declustering results in significantly reduced mean gold grades. 17.3.7 Statistical Analysis of Composite Data - Sulphur

The statistics of total sulphur data composited to 5 m were reviewed and are presented in Table 17.9 grouped by region. Histograms and probablility plots for each region are included in Appendix 10. Low variability is noted for the Orlea and Igre sulphur datasets with the Orlea region displaying only slight skew, while the Igre dataset shows significant positive skew and bi-modal behaviour, due to grouping of the mineralization zones.

141

Table 17.7 Rosia Montana Gold Project Outlier Analysis Number 5m Composites 962 1139 1819 5200 3982 829 143 967 6941 3861 266 114 194 132 127 215 418 68 37 962 1139 1819 5198 3982 829 143 967 6941 3861 266 114 194 132 127 215 418 68 37 Number 0.5-1.0m Samples Cut Gold (g/t) Orlea Carpeni Cetate Gauri Cirnicel Cirnic Jig SIK NSIK All QIP NQIP East All All QIP NQIP SIK NSIK 4 5 6 7 8 9 10 SIK NSIK All QIP NQIP East All All QIP NQIP SIK NSIK 4 5 6 7 8 9 10 48.18 46.85 272.34 179.80 168.59 30.53 5.80 63.22 112.12 129.47 25.29 3.12 7.29 9.88 5.83 25.06 57.44 2.04 8.33 14.60 63.20 85.90 665.60 246.60 128.60 38.20 98.00 269.60 156.00 395.80 68.40 22.40 45.20 52.00 138.40 50.80 28.20 15.60 1.39 0.69 0.91 1.44 0.79 0.81 0.51 0.93 1.49 0.77 2.80 0.60 1.41 0.89 0.62 1.38 0.85 0.31 1.02 2.23 1.95 2.01 9.15 3.44 2.65 3.10 9.46 9.84 5.57 26.94 5.19 3.22 6.46 3.16 4.43 2.55 1.48 2.90 2.56 2.01 6.64 4.87 4.41 1.60 0.81 2.64 3.27 3.10 2.63 0.66 1.14 1.25 0.88 2.33 2.22 0.34 1.78 1.27 2.33 3.26 22.53 7.86 4.96 4.36 10.33 11.87 8.39 40.50 7.65 2.90 5.98 5.43 9.90 3.49 3.44 2.88 1.84 2.91 7.33 3.39 5.56 1.96 1.59 2.83 2.19 4.02 0.84 1.10 1.09 1.41 1.43 1.69 2.60 1.08 1.75 Silver (g/t) Orlea Carpeni Cetate Gauri Cirnicel Cirnic Jig 0.57 1.20 1.62 2.46 2.29 1.87 1.41 1.09 1.21 1.51 1.50 1.47 0.90 0.92 1.72 2.23 1.37 2.33 0.99 0 0 0 16 0 0 0 0 5 2 1 0 0 0 0 1 0 0 0 10 11 14 170 65 17 11 78 135 65 120 14 10 14 12 15 14 2 7 2.22 1.88 1.92 8.75 3.31 2.50 2.73 9.44 9.79 5.45 24.43 4.39 3.11 5.71 2.71 3.72 2.39 0.94 2.60 1.19 1.19 1.71 16.29 5.62 2.22 2.37 10.12 11.22 7.08 27.54 3.70 2.44 3.26 2.73 3.17 2.03 0.42 1.90 0.54 0.63 0.89 1.86 1.70 0.89 0.87 1.07 1.15 1.30 1.13 0.84 0.78 0.57 1.01 0.85 0.85 0.44 0.73 4 2 5 9 8 3 4 2 3 11 7 6 6 8 4 5 3 3 2 100% 97% 96% 96% 96% 94% 88% 100% 100% 98% 91% 85% 97% 88% 86% 84% 94% 64% 90% 1 1 1 12 3 0 0 1 10 4 0 0 0 0 0 1 0 0 0 21 7 10 43 20 8 1.6 12 25 15 6 0 0 4 3 8 10 0 3 1.36 0.61 0.70 1.34 0.68 0.77 0.42 0.84 1.43 0.69 2.49 N/A N/A 0.80 0.58 1.29 0.76 N/A 0.77 2.13 0.83 1.10 2.85 1.45 1.06 0.40 1.32 2.29 1.28 1.35 1.57 1.36 1.56 2.13 2.14 1.38 0.95 1.57 1.60 1.86 0.54 2 5 8 2 8 3 5 4 18 10 13 98% 88% 77% 93% 85% 95% 83% 90% 96% 89% 89% Upper Cut Cut Mean Cut Std Dev Cut CV Number 5m Compsites Cut (Cut Mean / Mean) X 100

Region

Domain

Maximum

Mean

Std Dev

CV

Igre

0.71 0.69 1.68 1.36 0.97

0.89 1.20 1.31 1.79 1.26

2 2 3 3 2

90% 94% 93% 89% 75%

142

Igre

Table 17.8 Rosia Montana Gold Project Gold (g/t) 5 m Composites (Cut data) Orlea Mineralization Zone SIK Cell Size 40 x 40 x 5 40 x 40 x 10 60 x 60 x 10 80 x 80 x 10 Kriging Weights Nave No Data 962 962 962 962 962 Mean 1.361 1.361 1.361 1.361 1.361 Nave No Data 5,200 5200 5200 5200 5200 Mean 1.335 1.335 1.335 1.335 1.335 Nave No Data 6,941 6941 6941 6941 6941 Mean 1.413 1.413 1.413 1.413 1.413 Nave No Data 266 266 266 266 266 Mean 2.49 2.49 2.49 2.49 2.49 Std Dev 1.35 1.35 1.35 1.35 1.35 Std Dev 2.294 2.294 2.294 2.294 2.294 Std Dev 2.848 2.848 2.848 2.848 2.848 Std Dev 2.133 2.133 2.133 2.133 2.133 Declustered Mean 1.038 1.086 1.001 0.987 1.154 Std Dev 1.718 1.83 2.133 1.635 1.684 Mean 24% 20% 26% 27% 15% %Diff Mean 12% 11% 15% 16% 18% %Diff Mean 12% 11% 16% 17% 8% %Diff Mean 4% 7% 9% 9% -4% Std Dev -8% -5% -6% -7% -10% Std Dev 8% 8% 13% 12% 2% Std Dev 9% 8% 10% 12% 14% %Diff Std Dev 19% 14% 0% 23% 21%

Cetate Mineralization Zones QIPSI Cell Size 40 x 40 x 5 40 x 40 x 10 60 x 60 x 10 80 x 80 x 10 Kriging Weights Declustered Mean 1.174 1.191 1.14 1.115 1.092 Std Dev 2.6 2.621 2.561 2.501 2.449

Cirnic Mineralization Zone - QIPSI Cell Size 40 x 40 x 5 40 x 40 x 10 60 x 60 x 10 80 x 80 x 10 Kriging Weights Declustered Mean 1.248 1.26 1.19 1.168 1.304 Std Dev 2.1 2.105 1.997 2.019 2.252

Jig Mineralization Zone - SIK Cell Size 40 x 40 x 5 40 x 40 x 10 60 x 60 x 10 80 x 80 x 10 Kriging Weights Declustered Mean 2.39 2.31 2.26 2.26 2.6 Std Dev 1.46 1.42 1.43 1.44 1.48

143

Table 17.9 Rosia Montana Gold Project Sulphur Data (%) - Descriptive Statistics Total Count Minimum Maximum Mean Median Standard Deviation Variance Coefficient of Variation 8,672 0.01 15.45 1.71 1.72 0.86 0.74 0.50 Orlea 712 0.06 4.03 1.60 1.58 0.57 0.33 0.36 Carpeni 530 0.09 3.53 1.36 1.34 0.67 0.45 0.49 Cetate 3,048 0.02 15.45 1.89 1.95 0.95 0.90 0.50 Cirnicel 176 0.15 3.68 1.70 1.81 0.82 0.68 0.49 Cirnic 3,149 0.01 8.48 1.74 1.72 0.74 0.55 0.43 Jig 39 0.01 1.37 0.64 0.67 0.34 0.12 0.53 Igre 969 0.01 6.90 1.46 1.40 1.03 1.05 0.70 Tarina 4 0.3 0.38 0.35 0.35 0.03 0.00 0.09 Not Region 45 0.03 2.92 0.76 0.17 0.96 0.92 1.27

17.3.8

Correlation Analysis of Composite Gold and Silver Data

Bivariate analysis was completed between gold and silver. Little linear correlation is noted between gold and silver. The linear correlation co-efficient r, determined for the entire dataset, is 0.25. Figure 17.13 displays a scatter plot between gold and silver for the entire dataset with Figure 17.14 displaying the same data on log scales. There is no correlation between the sulphur and either the gold or silver data with linear correlation coefficients of -0.01 and 0.04 respectively (for samples located within the estimated domains). 17.3.9 Statistical Analysis Bulk Density Data

A detailed statistical analysis was undertaken on the bulk density database. The distribution of all bulk density data displays little skew and has a very low coefficient of variation, indicating that the great majority of density measurements cluster strongly around the mean. Summary statistics for the density database grouped by the key lithological subdivisions (Table 17.10) demonstrate that, with the exception of only the black breccia, the lithologies display very similar distributions and all units are characterized by very low coefficients of variation.
Table 17.10 Rosia Montana Gold Project Bulk Density Data - Statistical Summary Subdivided by Grouped Logged Lithology Item Number Minimum Maximum Mean Median Std Dev Coeff Var Total 6,209 1.83 3.18 2.43 2.42 0.17 0.07 Dacite 2,964 1.83 3.18 2.45 2.44 0.16 0.07 Vent Breccia 1,761 1.95 3.14 2.39 2.37 0.18 0.07 Corna Breccia 200 2.07 2.64 2.25 2.25 0.09 0.04 Cret Seds 709 2.01 2.98 2.49 2.51 0.15 0.06 Black Breccia 81 2.02 2.83 2.32 2.33 0.15 0.06 Andesite 2 2.18 2.44 2.31 2.31 0.13 0.06 XPO/XPR 492 1.84 2.81 2.41 2.42 0.12 0.05

Since the previous resource estimate, there have been some changes to the lithological model, (for example incorporating the Corna Breccia as a separate unit) and, therefore, the density values assigned have been updated using the latest data. 144

Figure 17.13 Scatter Plot - Gold (g/t) versus Silver (g/t)

Figure 17.14 Scatter Plot - Gold (g/t) versus Silver (g/t)

145

The density values used are as follows: Dacite Vent breccia Corna breccia Black breccia Phreatic breccia Cretaceous sediments 2.45 t/m3 2.39 t/m3 2.25 t/m3 2.32 t/m3 2.41 t/m3 2.49 t/m3

The adopted values, and their low levels of associated variance, reflect a robust estimate of the bulk density of the rocks hosting the Rosia Montana gold-silver deposit. 17.4 17.4.1 VARIOGRAPHY Introduction

Variography is used to describe the spatial variability or correlation of an attribute (gold, silver, sulphur, etc.). The spatial variability is traditionally measured by means of a variogram, which is generated by determining the averaged squared difference of data points at a nominated distance (h), or lag. The averaged squared difference (variogram or (h)) for each lag distance is plotted on a bivariate plot where the X-axis is the lag distance and the Y-axis represents the average squared differences ((h)) for the nominated lag distance. The spatial measures applied by RSG Global for the Rosia Montana study are a combination of traditional variography, back transformed gaussian variography and the correlogram, which is normalized for the variance of the data for the given lag (see Srivastava and Parker, 1989). In this document, the term variogram is used as a generic word to designate the function characterizing the variability of variables versus the distance between two samples. Fitted to the determined experimental variography is a series of mathematical models which, when used in the kriging algorithm, will recreate the spatial continuity observed in the variography. For the Rosia Montana study, all the variography generated for Ordinary Kriging estimation is based on a 5 m composite. A combination of software packages have been employed to generate and model the variography, including the geostatistical software Isatis and the mining planning software package Vulcan. The rotations are reported as input for grade estimation, with X (rotation around Z axis), Y (rotation around Y`) and Z (rotation around X``) also being referred to as the major, semi-major and minor axes.

146

17.4.2

Variography Gold and Silver

Detailed variography was completed for the 5 m run length composites coded within the interpreted mineralization zones. The modelled variography typically displays a high level of short scale variability that is comprised of moderate (22% to 39%) relative nuggets and a short scale structure that has been modelled with a range of approximately 25 metres. It is typical for the higher grade domains, for example the Cetate and Cirnic QIPSI zones, to exhibit the highest relative nugget effects, which are closer to 40%. This has been interpreted to represent the high variability of this mineralization, as observed underground. The interpreted Igre mineralization zones, which have similar geometry and statistical character, have been grouped together to allow robust variography to be generated. The combined variography has then been applied to estimate the separate zones with the variogram rotations adjusted based on the individual mineralization zone geometry. Interpreted zones with too few data to allow robust generation of variograms apply variogram models applied from like domains. The fitted variogram models for gold and silver are presented as Table 17.11 and 17.12 respectively. Variogram plots are included as Appendix 11. 17.4.3 Variography Sulphur

Sulphur variography was generated to allow sulphur estimates to be generated. To allow robust generation of variography, the sulphur variography was generated for the following grouped mineralization zones: Orlea. Igre. Cetate including Cetate east domains and Gauri. Cirnic including Cirnicel. Carpeni and remaining regions (includes Igre).

The modelled sulphur variography was applied to the remaining mineralization zones with adjustments made to the rotation such as to be consistent to the interpreted zone geometry. All variogram plots for sulphur are presented as Appendix 11. The sulphur variogram models are characterized by low relative nuggets and extended ranges of continuity. The extended variogram ranges indicate that the data spacing is sufficient to produce a globally robust estimate, with the low nugget effect reflecting relatively low levels of close spaced (down the hole) variability. The variogram models fitted for sulphur are provided as Table 17.13.

147

Table 17.11 Rosia Montana Gold Project Summary Variogram Models Gold Domain Orlea (SIK) Orlea (NSIK) Carpeni Cetate (qipsi) Cetate (east) Cetate (nqipsi) Gauri Cirnicel Cirnic (qipsi and si) Cirnic (nqipsi and si) Jig (SIK) Jig (NSIK) Igre4 Igre5&6 Igre7&8 Rotation X 90 90 0 40 0 40 0 20 30 30 0 0 300 300 300 Y' 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Z'' 0 0 0 0 0 0 0 0 0 0 0 0 0 25 35 % Rel Nugget 38% 29% 34% 39% 30% 39% 34% 34% 39% 32% 23% 23% 26% 22% 28% C0 1.25 0.20 0.41 2.60 0.34 0.81 0.06 0.49 1.80 0.32 0.42 0.10 0.34 0.11 0.63 C1 1.35 0.30 0.47 2.35 0.63 0.97 0.04 0.57 1.85 0.30 0.64 0.15 0.58 0.22 1.19 Range (m) X 25 85 37 31 19 33 65 15 29 29 20 20 70 125 200 Y 20 85 44 30 19 33 65 32 35 21 40 40 50 30 30 Z 17 18 12 16 14 19 65 13 14 15 25 25 35 20 33 C2 0.68 0.18 0.34 1.72 0.17 0.28 0.07 0.38 1.00 0.38 0.77 0.18 0.37 0.18 0.43 Range (m) X 147 150 136 195 76 193 73 58 205 120 125 125 70 125 200 Y 78 150 136 184 31 129 73 79 142 128 60 60 50 180 205 Z 48 60 44 113 35 64 73 23 120 85 60 60 35 25 33

Table 17.12 Rosia Montana Gold Project Summary Variogram Models Silver Domain Orlea (SIK) Orlea (NSIK) Carpeni Cetate (qipsi) Cetate (east) Cetate (nqipsi) Gauri Cirnicel Cirnic (qipsi and si) Cirnic (nqipsi and si) Jig (SIK) Jig (NSIK) Igre4 Igre5&6 Igre7&8 Rotation X 90 90 0 40 0 40 0 20 30 30 0 0 300 300 300 Y' 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Z'' 0 0 0 0 0 0 0 0 0 0 0 0 0 25 35 % Rel Nugget 36% 24% 29% 27% 28% 28% 41% 31% 38% 15% 18% 18% 19% 22% 25% C0 0.51 0.34 0.76 52 1.2 9 2.3 31 49 6 136.6 2.46 1.1 2.49 1.63 C1 0.53 0.80 0.69 56 1.3 14.5 1.3 25 42 12 258.0 4.64 3.0 5.21 1.95 Range (m) X 12 153 60 29 23 39 35 35 22 55 28 28 145 135 155 Y 22 105 30 29 50 45 35 95 25 68 45 45 145 90 25 Z 24 73 65 33 31 24 35 95 9 26 17 17 65 30 33 C2 0.37 0.27 1.15 85 1.8 8.2 2 44 38 21.5 364.2 6.56 1.8 3.63 2.93 Range (m) X 52 175 178 222 43 204 99 168 321 168 65 65 165 135 240 Y 85 120 118 252 91 130 99 95 244 138 45 45 165 95 180 Z 40 73 102 135 43 76 99 95 157 136 35 35 70 35 46

Table 17.13 Rosia Montana Gold Project Summary Variogram Models - Sulphur Domain Orlea Carpeni Cetate Cirnic Igre Other Rotation X 90 0 40 30 320 90 Y' 0 0 0 0 0 0 Z'' 0 0 0 0 30 0 C0 0.10 0.08 0.07 0.08 0.15 0.08 C1 0.40 0.30 0.52 0.62 0.38 0.22 Range (m) X 53 99 20 69 190 35 Y 62 106 16 83 230 40 Z 18 28 16 63 65 45 C2 0.50 0.62 0.41 0.30 0.47 0.70 Range (m) X 118 245 215 225 203 295 Y 181 146 163 202 260 220 Z 75 74 191 115 70 90

148

17.4.4

Summary

The variogram models are considered consistent with the interpreted controls on mineralization wherein a high degree of short scale variability exists between intercepts from adjacent drill holes and section lines. The implication of the moderate to high nugget variance is that a moderate to high level of smoothing will be produced by kriging. In addition to the estimation implications discussed above, the variography indicates that a close spacing of grade control drilling is required to effectively implement selective mining practises. It is considered that angle RC drilling is the most suitable drilling methodology although blasthole sampling maybe effective for larger bulk mining scenarios. Variogram plots for all domains are provided as Appendix 11. 17.5 17.5.1 BLOCK MODEL DEVELOPMENT Introduction

A block model was constructed using the Vulcan mining software package. The block model contains sufficient variables to record the results of Ordinary Kriging (OK) grade estimates and other required parameters. 17.5.2 Block Construction Parameters

Table 17.14 summarizes the extents of the Rosia Montana block model. The resource block model was developed using block dimensions of 10 m East by 10 m North by 10 m vertical with sub-blocking to 5 m Easting by 5 m Northing by 2.5 m vertical for the purpose of providing appropriate definition of the topographic surface, geological and mineralization zone boundaries.
Table 17.14 Rosia Montana Gold Project Block Model Dimensions Origin East North Elevation 354100.0 534080.0 450.0 Extent 2940.0 2760.0 750.0 Number 294 276 75 Block Size Parent 10 10 10 Sub-block 5 5 2.5

The interpreted lithology, alteration (siliceous and non siliceous zones), interpreted mineralization zones, waste dump material and topography were coded to the block model. A separate block model was constructed with subcelling down to 0.5 m to accurately represent the underground development and mining. This sub-block model was then regularized up to the parent size allowing each parent block to carry a percentage mined field that is incorporated into the volume prior to reporting.

149

A variable called VEIN was set to 1 where parent blocks are intersected by underground channel samples collected along veins. The remaining cells were coded as 0. Table 17.15 displays a full listing of the variables in the Rosia Montana block model, while domain coding incorporated into the Rosia Montana block model is summarized in Table 17.16.
Table 17.15 Rosia Montana Gold Project Block Model Variables Variables region alt rock lith zone domain vein bound au auestvar auestflag auestflag_c aunsamps aunholes nnau nnaudist id2au au2 ag nnag id2ag ag2 s_ok sestflag insitu_density prop_density resclass totalvoid Type integer integer integer integer integer integer integer integer double double integer integer integer integer double double double double float float float double double integer float float integer float Default 0 0 10 450 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Description Estimation Region Alteration (0 - NSAIK, 50 to 100 SIK, 20 rock 10, waste 100 and air 1000 DEFAULT IS VENT BRECCIA interp Mineralization Zone simplified domain code Blocks containing samples collected alon Blocks tounching soft boundaries OK estimate - Gold OK estimation variance - Au Au OK estimation flag (1 to 5) Au OK estimation flag combined (1 to 3) Au OK nsamples Au OK number of holes Au Nearest Neighbour Au Nearest Neighbour distance Au Inverse distance x2 OK estimate - Uncut Gold OK estimate - Silver Ag Nearest Neighbour Ag Inverse distance x2 OK estimate - Uncut Silver S OK estimate S OK estimate flag Insitu density based on rock type Proportion density considering pugw, air Resource Classification - TSE - JORC Total Void Proportion

17.5.3

Hardness Coding

Hardness was coded into the block model using solids supplied by RMGC and a series of scripts. The hardness variable was coded using the following scheme: 1 2 3 = = = hard medium soft

150

Table 17.16 Rosia Montana Gold Project Block Model Domain Coding Domain Type Description default Orlea Region Carpeni Region Cetate Region Cirnicel Region Cirnic Region Cos Region Jig Region Igre Region Tarina Region No Rock Dacite Dacite Dacite Dacite Dacite Dacite Dacite Dacite Dacite Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Intrusive Polymictic Breccia Phreatic Breccia Phreatic Breccia Phreatic Breccia Sandy Breccia Cretaceous Sediments Andesite Vent Breccia (Default) Corna Vally Breccia Black Breccia Waste Dumps Unalteration (default) Waste Dumps quartz-adularia quartz-adularia quartz-illite-pyrite quartz-adularia quartz-adularia quartz-adularia quartz-adularia quartz-illite-pyrite silicification silicification Unmineralised (Default) Waste Dumps Orlea Mineralization Carpenic Mineralization Cetate Mineralization Cetate East Mineralization Cetate East Mineralization Cetate East Mineralization Gauri Mineralization Cirnic Mineralization Jig Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Igre Mineralization Rock (Default) Waste Dumps Air Wireframe Name (*.00t) region_orlea_ros017 region_carpeni_ros017 region_cetate_ros017 region_cirnicel_ros017 region_cirnic_ros017 region_cos_ros017 region_jig_ros017 region_igre_ros017 region_tarina_ros017 topo_jul04_ros017 VDA_cos_ros017 VDA_jig_ros017 VDA_carp_ros017 VDA_CLAST1_ros017 VDA_CLAST2_ros017 VDA_cirnicel_ros017 VDA_cir_ros017 VDA_cet_ros017 nda_gaur_02a_ros017 igre_XPO_ros017 igre_XPO2_ros017 igre_XPO3_ros017 igre_XPO4_ros017 XPO1_cir_ros017 XPO2_cir_ros017 XPO3_cir_ros017 XPO1_cet_ros017 XPO2_cet_ros017 XPR1_cir_ros017 XPR2_cir_ros017 XPR3_cir_ros017 XCF_cir_ros017 SSE_uppersurf_ros017 VAN_lowersurf_ros017 corna_bx_ros017 XBB_ros017 wastebase_ros017 wastebase_ros017 gauri_sik_west_ros017 gauri_sik_ros017 cetate_qipsi_ros017 orlea_sik_ros017 jig_sik_ros017 igre_sik_ros017 carpeni_sik_ros017 cirnic_qipsi_ros017 cirnic_si1_ros017 cirnic_si2_ros017 wastebase_ros017 orlea_min_ros017 carpeni_min_ros017 cetate_min_ros017 cetate_east_min1_ros017 cetate_east_min2_ros017 cetate_east_min3_ros017 gauriw_min_ros017 cirnic_min_ros017 jig_min_ros017 igrmin4_ros017 igrmin5_ros017 igrmin6_ros017 igrmin7_ros017 igrmin8_ros017 igrmin9_ros017 igrmin10_ros017 wastebase_ros017 topo_jul04_ros017 Description Orlea Solid Carpeni Solid Cetate Solid Cirnicel Solid Cirnic Solid Cos Solid Jig Solid Igre Solid Tarina Solid Topography Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Dacite Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Breccia Solid Top of Cretaceous Sediments Base of Andesite Breccia Solid Breccia Solid Base of Waste Dumps Base of Waste Dumps Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Alteration Solid Base of Waste Dumps Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Mineralization Solid Base of Waste Dumps Topography Variable Name region region region region region region region region region region lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith lith alt alt alt alt alt alt alt alt alt alt alt alt zone zone zone zone zone zone zone zone zone zone zone zone zone zone zone zone zone zone rock rock rock Code 0 10 20 30 40 50 60 70 80 90 0 50 50 50 50 50 50 50 50 50 150 150 150 150 150 150 150 150 150 160 160 160 165 250 350 450 500 550 850 0 5 98 99 100 100 100 100 101 102 200 200 0 5 10 20 30 31 32 33 34 50 70 80 81 82 83 84 85 86 10 100 1000 Fill Direction Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (above) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (below) Z (above) Z (inside) Z (inside) Z (above) Z (above) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (above) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (inside) Z (above) Z (above) Priority (lowest=1) 90 50 80 40 85 10 20 30 5 91 1 4 9 11 12 13 14 15 17 5 6 7 8 18 19 20 10 21 22 23 24 25 2 3 89 16 90 89 40 30 20 60 70 80 50 10 11 12 95 80 10 50 47 48 49 40 90 70 30 21 22 23 24 25 31 50 99

Region

Lithology

Alteration

Mineralization

Rock

151

17.5.4

Bulk Density Assignment

A block model script was used to assign the mean density based on lithology. In Table 17.17, the internal vent breccia (Cirnic) refers to a major body of vent breccia which cuts through the Cirnic dacite. A theoretical bulk density was used for andesite, as no bulk density data for this rock type were collected during exploration.
Table 17.17 Rosia Montana Gold Project Block Model Bulk Density Assignments Block Model Code Dacite Phreatic Breccia Black Breccia Vent Breccia Corna Breccia Cretaceous Sediments Andesite 50 150 550 450 500 550 350 Bulk Density (t/m3) 2.45 2.41 2.32 2.39 2.25 2.49 2.70

17.5.5

Coding of Underground Workings

A 3-dimensional model of the underground workings and development was produced by RMGC technical staff (Figure 17.15) and has been applied to remove underground voids from the block model. The relative volume of each individual block affected by the underground development was calculated and applied to reduce the block volume and subsequently tonnage. As presented in Table 17.18, the model estimates that approximately 2.03 million cubic metres have been removed by previous underground mining.
Table 17.18 Rosia Montana Gold Project Summary of Coded Stope Volumes versus the Block Model Volumes Triangulation Name ct_dev_ros017.00t ct_dev2005a.00t ct_dev2005b.00t ct_gall_ros017.00t ct_stope_ros017.00t cn_dev_ros017.00t cn_gall_ros017.00t cn_mstope1_ros017.00t cn_mstope2_ros017.00t cn_stope_ros017.00t cn_roman_ros017.00t or_dev_ros017.00t jig_dev_ros017.00t adits_ros017.00t shaft_rso017.00t Triangulation Volume (m3) 403,942 13,120 663 220,343 311,354 258,074 186,729 352,486 65,769 57,244 31,772 41,967 9,689 63,439 11,230 2,027,820 Block Model Volume (m3) 400,308 13,129 661 220,434 311,305 257,963 186,789 352,927 65,870 57,218 31,782 41,951 9,751 63,452 11,251 2,024,791

Cetate

Cirnic

Orlea Jig-Vaidoaia General Total

152

Figure 17.15
Plan View of the Underground Workings

17.5.6

Block Model Validation

The block model was extensively validated against the geological model wireframes and the surface topography. The model was validated by viewing in multiple orientations using the 3-D viewing tools in Vulcan. Based on the visual review the block model was considered robust. 17.6 17.6.1 GRADE ESTIMATION Introduction

Resource estimation for the Rosia Montana project was undertaken using Ordinary Kriging as the principal estimation methodology for the gold, silver, and sulphur. Other check estimates, including an inverse distance squared estimate, were also generated. The OK estimate was completed using the Vulcan implementation of the GSLib software library with a check of a block estimate made between Vulcan and Isatis software to ensure a robust estimation algorithm. 17.6.2 Ordinary Kriging - Gold and Silver

Ordinary Kriging grade estimates were generated for gold and silver based on the 5 m composite data and applying a restricted number of composite data. A staged sample search was applied to the generation of the OK grade estimate with the generalized approach as summarized below: Pass 1: minimum of 6 and maximum of 12 composites collected within a 50 m E x 50 m N x 50 m vertical sample search. A maximum of 3 composites per drill hole was applied. An exception to this was the Igre domains for which an extended 80 m x 80 m x 30 m sample search was applied due to the more sparse data. 153

Pass 2: minimum of 6 and maximum of 12 composites collected within an approximate 120 m E x 120 m N x 90 m vertical sample search. A maximum of 3 composites per drill hole was applied. Pass 3: minimum of 6 and maximum of 12 composites collected within an approximate 150 m E x 150 m N x 90 m vertical sample search. A maximum of 3 composites per drill hole was applied.

Adjustments were applied to sample searches to reflect the model orientation of mineralization zones, variography and data spacing. Table 17.19 provides the sample search parameters applied for gold and silver. The sample search parameters were based on iterative runs and were designed to restrict smoothing to appropriate levels for the levels of mining selectivity applied. Note that on a block-by-block basis the approach of restricting the sample search results in conditionally biased estimates but the results are globally reflective of the predicted grade tonnage distribution. Another robust approach would be to estimate grade into large panels with an increased sample search and produce a selective mining model using a change of support methodology such as Uniform Conditioning (completed as a comparative estimate). Upper cuts, as presented in Table 17.7, were applied to estimation.
Table 17.19 Rosia Montana Gold Project Sample Search Parameters - Ordinary Kriging Gold and Silver Bearing (Z) 90 90 90 90 90 90 90 90 90 90 90 90 0 0 0 Plunge (Y) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Dip (X) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Major Axis (m) Orlea 50 50 120 50 120 150 70 50 120 50 120 150 Carpeni ALL ALL ALL 1 2 3 50 120 160 50 120 160 50 90 120 6 6 6 12 12 12 3 3 3 Semi-Major Axis (m) 50 50 120 50 120 150 70 50 120 50 120 150 Minor Axis (m) 40 40 60 40 60 100 60 50 90 50 90 100 Min Samp 6 6 6 6 6 6 6 6 6 6 6 6 Max Samp 12 12 12 12 12 12 12 12 12 12 12 12 Max Per Holes 3 3 3 3 3 3 3 3 3 3 3 3

Domain

Pass

SIK Vein SIK ALT SKIN SIK ALT SKIN SIK SIK SIK NSIK Vein NSIK ALT SKIN NSIK ALT SKIN NSIK NSIK NSIK

1 1 2 1 2 3 1 1 2 1 2 3

154

Table 17.19 (continued) Rosia Montana Gold Project Sample Search Parameters - Ordinary Kriging Gold and Silver Bearing (Z) 40 40 40 40 40 40 0 0 0 40 40 40 40 40 40 0 0 20 20 30 30 30 30 30 30 30 30 30 30 30 30 0 0 0 0 0 0 0 0 0 300 300 300 300 300 300 Plunge (Y) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Dip (X) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 25 25 25 Major Axis (m) Cetate 70 50 120 50 120 150 50 120 150 70 50 120 50 120 150 Gauri ALL ALL ALL ALL QIP Vein QIP ALT SKIN QIP ALT SKIN QIP QIP QIP NQIP Vein NQIP ALT SKIN NQIP ALT SKIN NQIP NQIP NQIP SIK ALT SKIN SIK ALT SKIN SIK SIK NSIK ALT SKIN NSIK ALT SKIN NSIK NSIK NSIK ALL ALL ALL ALL ALL ALL 1 2 1 2 1 1 2 1 2 3 1 1 2 1 2 3 1 2 1 2 1 2 1 2 3 1 2 3 1 2 3 60 130 Cirnicel 50 120 Cirnic 50 50 120 50 120 150 50 50 120 50 120 150 Jig 50 100 50 100 50 120 50 120 150 Igre4 80 120 160 Igre5 80 120 160 80 120 160 30 45 90 6 6 6 12 12 12 3 3 3 30 45 90 60 90 120 6 6 6 12 12 12 3 3 3 50 100 50 100 50 120 50 120 150 50 75 50 75 50 90 50 90 100 6 6 6 6 6 6 6 6 6 12 12 12 12 12 12 12 12 12 3 3 3 3 3 3 3 3 3 50 50 120 50 120 150 50 50 120 50 120 150 50 50 90 50 90 100 50 50 90 50 90 100 6 6 6 6 6 6 6 6 6 6 6 6 12 12 12 12 12 12 12 12 12 12 12 12 3 3 3 3 3 3 3 3 3 3 3 3 50 100 50 80 6 6 12 12 3 3 60 130 50 100 6 6 12 12 3 3 Semi-Major Axis (m) 70 50 120 50 120 150 50 120 150 70 50 120 50 120 150 Minor Axis (m) 60 50 90 50 90 90 50 90 100 60 50 90 50 90 90 Min Samp 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 Max Samp 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 Max Per Holes 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

Domain

Pass

QIP Vein QIP ALT SKIN QIP ALT SKIN QIP QIP QIP CETATE EAST CETATE EAST CETATE EAST NQIP Vein NQIP ALT SKIN NQIP ALT SKIN NQIP NQIP NQIP

1 1 2 1 2 3 1 2 3 1 1 2 1 2 3

155

Table 17.19 (continued) Rosia Montana Gold Project Sample Search Parameters - Ordinary Kriging Gold and Silver Bearing (Z) 300 300 300 300 300 300 310 310 310 30 30 30 305 305 305 Plunge (Y) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Dip (X) 25 25 25 35 35 35 25 25 25 25 25 25 0 0 0 Major Axis (m) Igre6 80 120 160 Igre7 ALL ALL ALL ALL ALL ALL ALL ALL ALL ALL ALL ALL 1 2 3 1 2 3 1 2 3 1 2 3 80 120 160 Igre8 80 120 160 Igre9 80 120 160 Igre10 80 120 160 30 45 90 60 90 120 6 6 6 12 12 12 3 3 3 80 120 160 30 45 90 6 6 6 12 12 12 3 3 3 80 120 160 30 45 90 6 6 6 12 12 12 3 3 3 80 120 160 30 45 90 6 6 6 12 12 12 3 3 3 Semi-Major Axis (m) 80 120 160 Minor Axis (m) 30 45 90 Min Samp 6 6 6 Max Samp 12 12 12 Max Per Holes 3 3 3

Domain

Pass

ALL ALL ALL

1 2 3

Note: Jig SIK ALT SKIN and SIK - Silver Search distances for pass 2 are 120m x 120m x 90m

The SIK/QIP Vein and NSIK/NQIP Vein estimates in Table 17.19 refer to the estimate of cells that are touching underground channel samples that have been sampled along veins. All subsequent estimates within each domain exclude the samples collected along veins. The estimation domains, as described in Table 17.20, were used for data collection. The mineralization zones were considered as hard boundaries (ie: no sample collection across boundaries). The exceptions to this rule are: Contact between SIK/QIP and NSIK/NQIP material within a domain. Blocks at the boundary (ie: one SIK and one NSIK) were allowed to search both SIK and NSIK data. Cirnic and Cirnicel were considered a soft boundary, as both regions are included in the Cirnic mineralization zone. Different cuts and restrictions were applied to the Cirnic/Cirnicel domains. Overlapping region between Igre 7 and 8 zones which has been estimated with the combined Igre 7 and Igre 8 composite dataset.

The soft boundary at the contact between the SIK/QIP and NSIK/NQIP was defined by coding all blocks within the SIK touching the NSIK blocks and reverse. A typical cross-section highlighting the soft boundary blocks, coding into the block model variable as dilution, is presented as Figure 17.16.

156

Figure 17.16 Typical Cross-Section Highlighting Blocks (pink) that Share Data

Table 17.20 Rosia Montana Gold Project Ordinary Kriging Estimation Domains Wireframe carpeni_ore_ros006tc.00t cetate_east_ore1_ros006tc. 00t cetate_east_ore2_ros006tc. 00t cetate_east_ore3_ros006tc. 00t cetate_ore_ros006tc. 00t cetate_ore_ros006tc. 00t gauriw_ros006tc.00t cirnic_ore_ros006tc.00t cirnic_ore_ros006tc.00t cirnic_ore_ros006tc.00t Igre4_ore_ros006tc.00t Igre5_ore_ros006tc.00t Igre6_ore_ros006tc.00t Igre7_ore_ros006tc.00t Igre8_ore_ros006tc.00t Igre9_ore_ros006tc.00t Igre10_ore_ros006tc.00t jig_ore_ros006tc.00t jig_ore_ros006tc.00t orlea_ore_ros006tc.00t orlea_ore_ros006tc.00t Estimation Domain Carpeni cetate_e1 cetate_e2 cetate_e3 Cetate QIPSI Cetate NQIP Gauri Cirnic QIPSI Cirnic NSIP Cirnicel Igre_4 Igre_5 Igre_6 Igre_7 Igre_8 Igre_9 Igre_10 Jig SIK Jig NSIK Orlea SIK Orlea NSIK Description All Carpeni with no separation based on alteration Combined Cetate east domains with no separation based on alteration Cetate zone siliceous alteration Cetate zone not siliceous alteration All Gauri zone with no separation based on alteration Main Cirnic siliceous zone (not Cirnicel region code) Cirnic non siliceous zone (not Cirnicel region code) All Cirnicel not subdivided alteration Igre zone Igre zone Igre zone Igre zone Igre zone Igre zone Igre zone Jig siliceous zone Jig non siliceous zone Orlea siliceous zone Orlea non siliceous zone

17.6.3

Comparative Estimates

Comparative estimates were undertaken for both gold and silver as a means of assessing the sensitivity of the estimate to cutting and estimation methods, including the following: 157

A comparative Uniform Conditioned (UC) check estimate based on a 30 m x 30 m x 30 m panel OK estimate. An OK estimate with more aggressive grade top cuts. An OK estimate with no grade top cuts. An inverse distance squared estimate. A nearest neighbour estimate (NN).

The reader is also directed to previous resource estimation studies completed in 2000 and 2002 wherein comparative Multiple Indicator Kriging estimates were undertaken. No detailed of these comparative estimates is provided in this report. 17.6.4 Sulphur Estimate

Based on additional data a sulphur model was generated with OK. Potential existed to produce an over-smoothed sulphur estimate during grade estimation, which would then result in unrealistic recovered grades being applied to mine design and cash flow models. RSG Global therefore considered a restricted sample search for estimation to limit smoothing to appropriate levels. A minimum of 2 and a maximum of 6 composites were used for the OK grade estimation to ensure mapping of local variability and to limit the smoothing effect of estimation. Comparisons of the input data and the sulphur model were undertaken as part of a detailed validation process to ensure that the sulphur model robustly mapped the input data. The validation process included both visual and statistical comparisons. A high degree of correlation is noted between the input data and the sulphur model. The visual and statistical validation indicates that the sulphur model is globally robust but that additional data are required to produce increased confidence in local estimates. 17.6.5 Mineralized Waste Dump Estimate

Following statistical review, a cut inverse distance squared grade estimate was completed for the Cetate waste dumps, using only those samples collected within the waste dumps. The resource estimate was validated by detailed visual inspection of estimated blocks versus the input grade data and is considered a reasonable global estimate of the metal. 17.6.6 Validation

Extensive visual and statistical validation of the grade estimates was completed. This process included: Review of the block estimate and the composite data in cross-section, long section, plan and oblique views. Comparison of the mean grade of the estimate versus the declustered mean grade, subdivided by estimation domain. Stacked transects, comparing the NN gold grade and the OK gold grade (Appendix 12).

158

The validation indicates that the resource model replicates the source input data well in regions of higher density drilling, such as the SIK/QIP domains. In the regions where the data density is lower, smoothing is evident, however, the estimates are considered appropriate. The restriction of the sample search for high-grade composites has ensured reasonable limits on the extrapolation of high-grade data. When the mean gold grade of the input data, declustered where appropriate, is compared with the overall gold grade (Table 17.21), a high degree of correlation is noted, except in regions of lower data density that are classified as Inferred, or as a combination of Indicated and Inferred. Transects comparing the OK gold grades versus the average 5 m cut composite grades (composite data blue, block model data in red) are provided as Figures 17.17, 17.18 and 17.19 respectively for easting, northing and elevation sections. The model generally honours the input data well, with the peaks and troughs noted in the input composite data adequately reproduced. 17.6.7 Resource Reporting

The RSG Global resource estimate for the Rosia Montana project has been categorized in accordance with the criteria laid out in Canadian National Instrument 43-101 (NI 43-101) and the JORC code. A combination of Measured, Indicated and Inferred Resources have been defined using definitive criteria determined during the validation of the grade estimates, with detailed consideration of the NI 43-101 categorization guidelines.
Figure 17.17 Au OK vs Au 10 m Composites : Easting
3
Composite Grades

800
Model Grades Number of Composites

Average Gold Grade (g/t)

2.5 2 1.5 1 0.5 0 354305 354805 355305 355805

700 600 500 400 300 200 100 0

356305

356805

Easting (mE)

159

Number of composites

Figure 17.18 Au OK vs Au 10 m Composites : Northing

3
Composite Grades

1200
Model Grades Number of Composites

Average Gold Grade (g/t)

2 1.5 1 0.5 0 534245

800 600 400 200 0 534745 535245 535745 536245

Northing (mN)
Figure 17.19 Au OK vs Au 10 m Composites : Elevation
1.4 3500 3000
Composite Grades Model Grades Number of Composites

Average Gold Grade (g/t)

1.2 1 0.8 0.6 0.4 0.2 0 465

2500 2000 1500 1000 500 0

565

665

765

865

965

1065

Elevation (mRL)

160

Number of composites

Number of composites

2.5

1000

Table 17.21 Rosia Montana Gold Project Comparison of the Mean Block Model Grade : Gold (g/t) Zones Orlea SIK Orlea NSIK Carpeni Cetate QIPSI Cetate NQIP Cetate East Combined Gauri Cirnicel Cirnic QIPSI Cirnic NQIP Jig SIK Jig NSIK Igre4 Igre5 Igre6 Igre7 Igre8 Igre9 Igre10 OK Model 1.133 0.683 0.639 1.069 0.681 0.672 0.402 0.858 1.33 0.63 2.568 0.724 1.022 0.599 0.644 1.144 0.71 0.303 0.878 5m Cut Composites Raw 1.36 0.61 0.7 1.33 0.68 0.77 0.42 0.84 1.43 0.69 2.49 0.60 1.05 0.80 0.58 1.18 0.75 0.31 0.77 Declustered 1.09 0.59 0.64 1.14 0.67 0.67 0.42 0.7 1.28 0.61 2.31 0.62 1.00 0.75 0.55 1.13 0.75 0.32 0.84 Number of Composites 962 1139 1819 5200 3982 829 143 967 6941 3861 266 114 194 132 127 215 418 68 37

Table 17.22 Rosia Montana Gold Project Confidence Levels of Key Categorization Criteria

Items Drilling Techniques Logging Drill Sample Recovery Sub-sampling Techniques & Sample Preparation Quality of Assay Data Verification of Sampling and Assaying Location of Sampling Points Data Density and Distribution

Discussion RC/Diamond/Channel - industry standard approach. Geological and Geotechnical logging is completed using standard nomenclature and apparent high quality. Acceptable recoveries determined for the majority of the drilling. Industry standard for both RC and diamond drilling and Channelling Comprehensive internal laboratory and external quality control data available for the majority of the drilling and channel sampling. Data is of industry standard quality. The verification drill hole program confirmed previously collected data. All drill hole collars and channels have been surveyed and most drillholes have been downhole surveyed. The sample data array varies considerably throughout the Rosia Montana project. In regions of historical underground workings substantial data exist in the easting and northing directions, which in general is better than 40mE x 40mN in spacing. The sampled underground workings are spaced between 20 and 50 metres apart. Underground diamond drilling has been completed in regions of low sample data, both drilling and channels, from points of available access. The surface drilling represents a notional 80mE x 80mN pattern in regions not covered by the higher density channel sampling and underground diamond drilling, or the post 2000 infill drilling program. The current data spacing is considered appropriate for resource evaluation The current resource estimation has been audited by independent consulting firm SRK. No material errors identified. The interpreted lithological alteration and mineralization boundaries are considered robust and of high confidence. The gold, silver and sulphur resource estimate has been generated via OK. RSG Global considers OK estimation approach appropriate for gold and silver where bulk mining is being considered based on low cutoff grades. A notional 0.3g/t Au cutoff grade criterion has been used to complete the interpretation mineralization domain envelopes to exclude material that is definitely waste. The estimation methods used in the resource estimation study are not cutoff grade dependent, however the resource estimate is targeted for 0.4g/t Au to 0.6g/t Au lower cutoff range and is not appropriate at higher cutoff grades. The resource estimate has used geostatistical techniques which take into account mining dilution given that the proposed minimum mining unit represents the selected block size. The resource model thus takes into account proposed mining dilution and mining recovery. Detailed statistical evaluation of the dry bulk density data subdivided by lithology has been completed prior to assignment of average bulk density per rock type. The bulk density is considered to be well established.

Confidence High High High High High Moderate High Moderate

Audits or Reviews Database Integrity Geological Interpretation Estimation and Modelling Techniques Cutoff Grades

NA High Moderate Moderate to High NA

Mining Factors or Assumptions Tonnage Factors

High

High

161

Criteria for Resource Categorization The RSG Global resource estimate for Rosia Montana has been classified as a combination of Measured, Indicated and Inferred Mineral Resources based on the confidence levels of the key criteria that were considered during resource classification as presented in Table 17.22. Applying the above confidence levels, resource classification codes were assigned to the resource block model using the following criteria applied to estimation statistics recorded during estimation of the resource gold grades: Measured Resources Only blocks coded as occurring within the SIK Zone for Orlea and Jig. Only regions coded as QIP or NQIP for Cetate and Cirnic. Estimation pass 1. A minimum of 8 composites. Where the distance to the nearest composite is not more than 30 m. Data collected from a minimum of 3 drill holes or underground drives.

Indicated Resources OR Not Measured and where the distance to the nearest composite is not more than 70 m. A minimum of 6 composites. Data collected from a minimum of 2 drill holes or underground drives. Estimation pass 1.

Inferred Resources Blocks lying further than 70 m from the closest composite. A minimum of 6 composites Data collected from a minimum of 2 drill holes or underground drives.

Igre was assessed on a zone by zone basis with adjustment of higher confidence zones to Indicated Resource which did not meet the above criteria. In addition, the southern extension of Orlea was also assessed and a more restrictive classification criterion applied with blocks that would be considered Indicated Resource in the above criteria categorized as Inferred Resource. A solid wireframe was constructed for the measured category based on the criteria presented above, and cells whose centroids were located within the wireframe were classified as a Measured Resource. Assignment of indicated and inferred resource categories into the block model was achieved by the use of a block model manipulation script.

162

Grade Tonnage Reporting Measured, Indicated and Inferred Resources are reported for the Rosia Montana project, based on the 10 m x 10 m x 10 m block model. Resources reported are based on the OK gold and silver estimates generated using a 10 m x 10 m x 10 m block size, using the lower cutoff grades stated, and excluding the volume contained within the models of the underground voids. Summary grade tonnage reports are provided as Table 17.23 and a detailed grade tonnage report subdivided by region is provided in Table 17.24.
Table 17.23 Rosia Montana Gold Project Categorized Grade Tonnage of accepted Ordinary Kriged Estimate 10 mE x 10 mN x 10 mRL, Cut Gold and Silver 5 m composites Gold Cutoff Grade 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 Mt 171.51 139.83 113.11 90.70 341.22 210.52 137.65 94.40 512.73 350.35 250.76 185.10 44.81 30.29 22.20 17.53 Au (g/t) 1.3 1.5 1.7 1.9 0.9 1.2 1.4 1.6 1.0 1.3 1.5 1.8 1.0 1.2 1.4 1.5 Ag (g/t) 8 8 9 10 3 4 4 5 5 6 6 7 3 3 3 3 Contained Gold (Moz) 7.3 6.7 6.1 5.5 9.9 7.8 6.2 5.0 17.1 14.6 12.3 10.5 1.4 1.2 1.0 0.9 Contained Silver (Moz) 43.2 38.1 32.8 27.8 38.0 26.8 19.5 14.6 81.1 64.9 52.3 42.4 4.1 3.0 2.1 1.6

Measured

Indicated

Measured and Indicated

Inferred

The Mineral Resources stated in Table 17.23 include the Mineral Reserves described below. In addition to the Inferred Resource reported (above) for the insitu gold-silver mineralization at Rosia Montana, an Inferred Resource in the Cetate waste dumps, Halda Hop and Halda Valea Verde, is reported based on the inverse distance squared estimates in 10 m x 10 m x 10 m blocks with no lower cutoff grade applied. 17.6.8 Inferred Resource: 4.0 million tonnes at 0.6 g/t Au, 4 g/t Ag. Comparative Estimates

Check estimates of the Rosia Montana project are also presented in Table 17.25 based on a recovered estimate. Uniform Conditioning was implemented to achieve the recovered estimate using the Ordinary Kriged estimate in a 30 m E x 30 m N x 10 m vertical block model. Sensitivity analysis of the gold cutting strategy was undertaken by completing an estimate using aggressive cuts (Table 17.26) for Orlea SIK (cut to 10 g/t Au), Cetate QIP (cut to 31 g/t Au) and Cirnic QIP (cut to 19 g/t Au). Further analysis of the effectiveness of the cutting strategy was undertaken in an estimate using no topcuts (Table 17.27). Comparative inverse distance squared (Table 17.28) and nearest neighbour (Table 17.29) estimates were also completed. All comparative and check models are categorized for comparison purposes only using the same criteria/coding as the accepted model, and the comparative resource estimates do not constitute categorized Mineral Resources.

163

Table 17.24 Rosia Montana Gold Project Grade Tonnage Report: Ordinary Kriging - Gold and Silver Estimates Subdivision on Resource Category and Region 10 mE x 10 mN x 10 mRL Size Blocks Whole Block Grades - Excluding Gallery Material Tonnes
(x 1000)

Cutoff 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0

Au
(g/t)

Measured Ag Au
(g/t) (Ounces x 1000)

Ag
(Ounces x 1000)

Tonnes
(x 1000)

Au
(g/t)

Indicated Ag Au
(g/t) (Ounces x 1000)

Ag
(Ounces x 1000)

Sub-Total Measured and Indicated Tonnes Au Ag Au Ag

(x 1000) (g/t) (g/t) (Ounces x 1000) (Ounces x 1000)

Tonnes
(x 1000)

Au
(g/t)

Inferred Ag Au
(g/t) (Ounces x 1000)

Ag
(Ounces x 1000)

Orlea

Carpeni

Cetate

Cirnicel

Cirnic

Cos

Jig

Igre

Total

9,661 8,524 7,505 6,284 15 5 0 0 49,486 37,020 28,521 22,665 7,314 5,427 3,917 2,759 103,268 87,082 71,408 57,252 0 0 0 0 1,771 1,769 1,761 1,740 0 0 0 0 171,513 139,827 113,112 90,701

1.50 1.63 1.76 1.92 0.57 0.60 0.00 0.00 1.26 1.52 1.77 1.99 1.01 1.18 1.36 1.56 1.32 1.48 1.65 1.83 0.00 0.00 0.00 0.00 2.63 2.63 2.64 2.66 0.00 0.00 0.00 0.00 1.32 1.50 1.69 1.89

2 2 2 2 2 2 0 0 6 7 8 8 10 11 12 12 9 9 10 10 0 0 0 0 25 25 25 26 0 0 0 0 8 8 9 10

466 447 424 388 0 0 0 0 2,007 1,808 1,619 1,450 236 206 172 139 4,397 4,136 3,783 3,375 0 0 0 0 150 150 150 149 0 0 0 0 7,256 6,746 6,147 5,501

665 601 532 445 1 0 0 0 9,948 8,412 7,142 6,127 2,449 1,981 1,509 1,105 28,662 25,658 22,234 18,719 0 0 0 0 1,433 1,432 1,432 1,429 0 0 0 0 43,157 38,084 32,848 27,825

79,349 46,917 29,021 18,306 32,137 19,916 12,417 8,206 73,160 44,214 28,107 19,001 9,894 7,079 4,886 3,368 90,494 52,651 34,048 23,808 4,838 2,747 1,515 573 4,545 3,419 2,530 2,191 46,591 33,468 25,113 18,946 341,215 210,521 137,650 94,399

0.82 1.06 1.28 1.50 0.85 1.08 1.31 1.52 0.87 1.11 1.35 1.57 0.99 1.18 1.40 1.62 0.92 1.23 1.53 1.81 0.71 0.86 1.01 1.20 1.14 1.35 1.59 1.69 1.06 1.28 1.47 1.66 0.90 1.16 1.40 1.63

2 2 2 2 2 2 2 2 3 4 4 4 10 12 13 15 4 5 6 7 7 8 8 9 6 8 9 10 3 3 4 4 3 4 4 5

2,103 1,592 1,192 885 882 689 522 401 2,039 1,582 1,223 961 314 269 219 176 2,684 2,086 1,676 1,382 110 76 49 22 167 148 129 119 1,584 1,374 1,188 1,011 9,886 7,818 6,199 4,957

5,199 3,253 2,058 1,320 1,888 1,287 846 577 7,475 5,185 3,685 2,741 3,285 2,634 2,054 1,617 13,011 9,247 6,800 5,167 1,061 669 385 158 930 829 746 695 5,087 3,727 2,894 2,314 37,960 26,845 19,470 14,587

89,010 55,440 36,526 24,590 32,151 19,921 12,417 8,206 122,646 81,234 56,628 41,666 17,208 12,507 8,803 6,127 193,762 139,732 105,456 81,059 4,838 2,747 1,515 573 6,316 5,189 4,291 3,931 46,591 33,468 25,113 18,946 512,729 350,348 250,762 185,100

0.90 1.14 1.38 1.61 0.85 1.08 1.31 1.52 1.03 1.30 1.56 1.80 0.99 1.18 1.38 1.60 1.14 1.39 1.61 1.83 0.71 0.86 1.01 1.20 1.56 1.79 2.02 2.12 1.06 1.28 1.47 1.66 1.04 1.29 1.53 1.76

2 2 2 2 2 2 2 2 4 5 6 7 10 11 13 14 7 8 9 9 7 8 8 9 12 14 16 17 3 3 4 4 5 6 6 7

2,568 2,038 1,616 1,273 882 689 522 401 4,046 3,390 2,842 2,412 550 474 391 314 7,082 6,222 5,459 4,757 110 76 49 22 316 298 279 268 1,584 1,374 1,188 1,011 17,142 14,565 12,346 10,458

5,865 3,854 2,589 1,764 1,889 1,287 846 577 17,422 13,598 10,828 8,868 5,735 4,614 3,563 2,722 41,672 34,905 29,034 23,885 1,061 669 385 158 2,362 2,262 2,177 2,124 5,087 3,727 2,894 2,314 81,117 64,929 52,318 42,413

25,168 19,589 16,316 13,762 837 831 828 828 2,032 981 287 65 658 523 370 326 8,328 3,332 1,738 1,164 2,941 1,978 1,081 354 1,986 1,544 848 547 2,227 1,132 650 487 44,810 30,285 22,199 17,533

1.15 1.34 1.47 1.57 1.56 1.57 1.57 1.57 0.63 0.77 0.97 1.20 1.17 1.34 1.60 1.70 0.70 1.02 1.33 1.55 0.74 0.85 0.99 1.18 0.85 0.95 1.16 1.29 0.77 1.04 1.30 1.44 0.98 1.22 1.41 1.55

2 2 2 2 2 2 2 2 2 2 2 3 14 16 18 19 3 4 5 5 7 8 8 9 5 5 7 7 3 4 4 5 3 3 3 3

929 841 768 695 42 42 42 42 41 24 9 3 25 23 19 18 188 109 74 58 70 54 34 13 54 47 32 23 55 38 27 23 1,416 1,187 1,008 873

1,548 1,236 1,034 874 57 57 56 56 130 70 21 5 295 262 210 197 813 422 253 178 673 490 278 104 295 251 181 126 179 129 93 76 4,100 2,985 2,142 1,617

164

Note: Appropriate rounding Tonnes 10,000, Ounces to the nearest 1,000 and Au to 1 decimal place and Ag to nearest gm.

Table 17.25 Rosia Montana Gold Project Categorized Grade Tonnage of Uniform Conditioned check estimate 30 mE x 30 mN x 30 mRL, Cut Gold and Silver 5 m composites Gold Cutoff Grade 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 Contained Gold (Moz) 169.27 140.40 113.45 90.15 345.58 221.58 142.39 94.42 514.84 361.99 255.84 184.58 45.06 31.79 23.46 17.73 1.3 1.5 1.7 1.9 0.9 1.1 1.3 1.6 1.0 1.3 1.5 1.7 1.0 1.2 1.4 1.5 8 8 9 9 3 4 4 5 5 6 6 7 3 3 3 3 7.2 6.7 6.1 5.4 9.9 7.9 6.2 4.8 17.0 14.6 12.2 10.2 1.4 1.2 1.0 0.9 Contained Silver (Moz) 42 37 32 27 37 27 19 14 80 64 51 41 4 3 2 2

Mt

Au (g/t)

Ag (g/t)

Measured

Indicated

Measured and Indicated

Inferred

Table 17.26 Rosia Montana Gold Project Grade Tonnage Distribution of RSG Global Comparative Ordinary Kriged Estimate with Aggressive Cuts 10 mE x 10 mN x 10 mRL, Cut Gold and Silver 5 m composites Gold Cuts, Orlea:- 10, Cetate:- 31 and Cirnic:- 19 Gold Cutoff Grade 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 Contained Gold (Moz) 171.51 139.83 113.11 90.68 341.22 210.52 137.65 94.40 512.73 350.35 250.76 185.08 44.81 30.29 22.20 17.53 1.3 1.5 1.7 1.9 0.9 1.2 1.4 1.6 1.0 1.3 1.5 1.7 1.0 1.2 1.4 1.5 8 8 9 10 3 4 4 5 5 6 6 7 3 3 3 3 7.2 6.6 6.0 5.4 9.9 7.8 6.2 4.9 17.0 14.4 12.2 10.3 1.4 1.2 1.0 0.9 Contained Silver (Moz) 43.2 38.1 32.8 27.8 38.0 26.8 19.5 14.6 81.1 64.9 52.3 42.4 4.1 3.0 2.1 1.6

Resource Category

Mt

Au (g/t)

Ag (g/t)

Measured (as assigned to the RSG Global Preferred whole block estimate) Indicated (as assigned to the RSG Global Preferred whole block estimate) Measured and Indicated (as assigned to the RSG Global Preferred whole block estimate) Inferred (as assigned to the RSG Global Preferred whole block estimate)

165

Table 17.27 Rosia Montana Gold Project Grade Tonnage Distribution of RSG Global Comparative Ordinary Kriged Estimate with Uncut Composites 10 mE x 10 mN x 10 mRL, Uncut Gold and Silver 5 m composites Gold Cutoff Grade 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 Contained Gold (Moz) 171.51 139.83 113.16 90.82 341.47 211.63 140.11 97.80 512.98 351.46 253.27 188.62 44.80 30.35 22.41 17.96 Table 17.28 Rosia Montana Gold Project Grade Tonnage Distribution of RSG Global Comparative Inverse Distance Estimate 10 mE x 10 mN x 10 mRL, Cut Gold and Silver 5 m composites Gold Cutoff Grade 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 Contained Gold (Moz) 167.95 136.08 110.04 88.38 329.28 204.02 134.19 92.03 497.24 340.10 244.23 180.41 47.03 32.27 22.27 17.81 1.3 1.5 1.7 1.9 0.9 1.2 1.4 1.7 1.1 1.3 1.6 1.8 1.0 1.2 1.4 1.5 8 9 9 10 4 4 4 5 5 6 7 7 3 3 3 3 7.2 6.7 6.1 5.5 9.7 7.7 6.2 5.0 16.9 14.4 12.3 10.5 1.5 1.2 1.0 0.9 Contained Silver (Moz) 42.9 37.8 32.6 27.7 37.3 26.7 19.4 14.5 80.2 64.4 52.0 42.2 4.4 3.3 2.1 1.6 1.4 1.6 1.8 2.1 1.0 1.3 1.6 2.0 1.1 1.4 1.7 2.0 1.0 1.3 1.5 1.6 8 9 9 10 4 4 5 5 5 6 7 7 3 3 3 3 7.8 7.3 6.7 6.0 11.1 9.0 7.4 6.2 18.9 16.3 14.1 12.2 1.5 1.3 1.1 1.0 Contained Silver (Moz) 44.1 38.9 33.7 28.7 39.6 28.4 21.0 16.1 83.7 67.3 54.7 44.7 4.6 3.4 2.5 1.9

Resource Category

Mt

Au (g/t)

Ag (g/t)

Measured
(as assigned to the RSG Global Preferred whole block estimate)

Indicated
(as assigned to the RSG Global Preferred whole block estimate)

Measured and Indicated

(as assigned to the RSG Global Preferred whole block estimate)

Inferred
(as assigned to the RSG Global Preferred whole block estimate)

Resource Category

Mt

Au (g/t)

Ag (g/t)

Measured
(as assigned to the RSG Global Preferred whole block estimate)

Indicated
(as assigned to the RSG Global Preferred whole block estimate)

Measured and Indicated

(as assigned to the RSG Global Preferred whole block estimate)

Inferred
(as assigned to the RSG Global Preferred whole block estimate)

166

Table 17.29 Rosia Montana Gold Project Grade Tonnage Distribution of RSG Global Comparative Nearest Neighbour Estimate 10 mE x 10 mN x 10 mRL, Cut Gold and Silver 5 m composites Gold Cutoff Grade 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 Contained Gold (Moz) 135.47 102.07 79.63 62.70 235.10 151.99 104.57 74.56 370.56 254.05 184.19 137.26 36.05 23.89 19.38 16.45 1.6 2.0 2.3 2.7 1.2 1.6 2.0 2.4 1.4 1.7 2.1 2.6 1.5 2.0 2.4 2.6 9 10 11 12 4 4 5 6 6 7 8 9 3 3 3 3 7.0 6.5 6.0 5.5 9.1 7.8 6.7 5.8 16.1 14.3 12.7 11.4 1.8 1.6 1.5 1.4 Contained Silver (Moz) 39.2 33.4 28.7 24.4 29.7 21.8 17.2 13.9 69.0 55.2 45.9 38.2 3.7 2.5 2.1 1.8

Resource Category

Mt

Au (g/t)

Ag (g/t)

Measured
(as assigned to the RSG Global Preferred whole block estimate)

Indicated
(as assigned to the RSG Global Preferred whole block estimate)

Measured and Indicated

(as assigned to the RSG Global Preferred whole block estimate)

Inferred
(as assigned to the RSG Global Preferred whole block estimate)

17.7

MINERAL RESERVES

As described subsequently in Section 18.1.1, all material within the design open pits at the Rosia Montana project which, on the basis of economic criteria, has been scheduled for processing for the recovery of gold and silver, has been classified as Proven and Probable Mineral Reserves. As summarized in Table 17.30, the Proven and Probable Mineral Reserves of the Rosia Montana property are estimated at 214.9 million tonnes at average grades of 1.46 g/t gold and 6.88 g/t silver.
Table 17.30 RMGC Mineral Reserves - February 26, 2009
Category Cutoff Net/hour ($ x 1000) Proven Probable Total $2.24 to $22.00 $2.24 to $22.00 Ore (kt) 112,455 102,476 214,931 Direct Plant Feed Ore Gold (g/t) 1.630 1.270 1.46 Silver (g/t) 9.01 4.55 6.88 Recovered Grade (g/t Au) 1.290 1.000 1.15 Recovered Grade (g/t Ag) 5.54 2.69 4.18 Contained Gold (Moz) 5.9 4.2 10.1 Contained Silver (Moz) 32.6 15.0 47.6 Recovered Recovered Gold Silver (Moz) (Moz) 4.6 3.3 7.9 20.0 8.9 28.9

The Mineral Resources and Mineral Reserves reported herein have been estimated in accordance with the Canadian Institute of Mining, Metallurgy and Petroleum standards and definitions. As described elsewhere in this report, in September, 2007, the Romanian government suspended its review of the EIA submitted for the project by RMGC. RMGC, therefore, does not presently have the permits necessary to commence construction or operation. It is the opinion of RMGC, however, that the EIA review process is likely to recommence in 2009. The authors have relied on this opinion.

167

To the best of the knowledge of the authors, except for the suspension of the EIA review process, there are no legal, taxation, socio-economic, marketing or other factors which would materially affect the Mineral Reserves estimated herein.

168

18.0 OTHER RELEVANT DATA AND INFORMATION

The principal purpose of this Technical Report is to discuss updated estimates of capital expenditure and operating costs, and an updated analysis of overall economics, for the Rosia Montana project. In the course of this work, the open pit designs were reviewed in light of updated gold price projections and operating cost estimates. As a result of this review, the ultimate limits of the open pits remained unchanged from those discussed in the 2006 Technical Report. The process flowsheet was also reviewed, and has been modified to incorporate a gravity circuit for the recovery of coarse gold. The batch elution process described in the 2006 Technical Report has now also been replaced by a continuous elution process. The Rosia Montana project is based on producing gold and silver dor bars by conventional open pit mining using shovels and trucks followed by conventional processing for the recovery of precious metals by gravity concentration and cyanide leaching of the whole ore, after grinding to a size of 80% minus 150 microns. The dor bars produced at site will be shipped to commercial facilities for refining. The project is designed for a mining life of 14 years and a processing life of approximately 16 years, at a plant throughput of between 13 million and 15.5 million tonnes per year. Total life-of-mine production is estimated at 7.9 million ounces of gold and 29 million ounces of silver. Figure 18.1 is a general site plan showing the locations of the open pits and other facilities planned to be constructed at Rosia Montana. A comprehensive Environmental Impact Assessment was completed early in 2006 and was submitted to the Romanian authorities for review. The review process was suspended by the Romanian government in September, 2007, and remains suspended at the date of this report. The timing at which construction will commence thus remains uncertain. In the interim, RMGC has placed orders totalling approximately $44 million for major equipment items with long lead times, including the primary crusher, the SAG mill and ball mills, and mill drive systems. Once construction of the project begins, it is expected to take approximately 24 months to complete. The organizations which have prepared the engineering designs and cost estimates discussed in this section of the report are: IMC, which was responsible for the design of the open pits, the scheduling of mine production, the selection of mining equipment, and the estimation of mine capital expenditures and operating costs. Aurifex, which was responsible for the review and modification of the process flowsheet, and the estimation of process operating costs. MWH, which was responsible for the design of all earthworks and certain infrastructure, including the embankment for the TMF, all water impoundments and all roads, and for the estimation of capital expenditures for these components of the project.

169

Figure 18.1 Overall Site Plan

170

Metifex, which was responsible for the estimation of capital expenditures for all project facilities, other than mining facilities and those items covered by MWH. RMGC, which was responsible for the estimation of the Owners component of total capital expenditure, and the general and administration component of operating cost, as well as the analysis of overall project economics.

RMGC also prepared the sections of this report dealing with environmental and socio-economic considerations (Section 18.4) and permitting (Section 18.5). The authors have relied upon RMGCs description of these matters. 18.1 MINING

As shown in Figure 18.1, mining is planned from four open pits on the Rosia Montana property: Cirnic, Cetate, Orlea and Jig. 18.1.1 Pit Design

The ultimate limits of the Rosia Montana open pits used conventional floating cones as a basis for design, based on the RSG block model, and the design criteria summarized in the Table 18.1.
Table 18.1 Pit Design Criteria Criterion Overall pit slope Gold recovery Silver recovery Gold price Silver price Mining cost Haulage increment Process and G&A cost -hard ore -medium ore -soft ore Gold refining cost Silver refining cost US$/oz US$/oz US$/tonne mined US$/bench/tonne US$/tonne processed US$/tonne processed US$/tonne processed US$/ounce US$/ounce Unit degrees Value 40 Aurifex algorithm Aurifex algorithm 735 11.50 1.29 0.021 11.78 10.60 9.19 3.11 0.467

The pit limits were also constrained by certain physical features which restricted the size of the economic pits in several areas. These physical features included: The Protected Area shown on Figure 18.1, and the buffer zone around this area. The Piatra Corbuli, which restricts the east wall of the Cirnic pit. The Mausoleum area to the south of the Cetate pit. Various historic buildings throughout the Rosia Montana valley. 171

All material within the design pits which, on the basis of economic criteria, has been scheduled for processing for the recovery of gold and silver has been classified as Proven and Probable Mineral Reserves, as shown previously in Table 17.30. The tonnages and grades scheduled for mining from each of the four open pits are summarized in Table 18.2.
Table 18.2 Mine Plan Summary, by Pit Pit Cetate Cirnic Jig Orlea Total Tonnes (thousand) 54,633 112,652 5,547 42,100 214,931 Grade (g/t) 1.539 1.529 1.370 1.181 1.459 5.67 9.28 4.54 2.34 6.88

For production scheduling, the Cetate, Cirnic and Orlea pits were divided into a series of sequential mining phases, designed to maximize the present value of the project while honouring practical operational criteria. The small Jig pit was scheduled as a single phase. Provision was made for the stockpiling of low grade material, for processing at the end of the mine life. The design mine production schedule is summarized in Table 18.3.
Table 18.3 Mine Production Schedule Year Cutoff Waste (thousand t) Tonnes (thousand) Plant Feed Tonnes (thousand) To Stockpile

Net $x1000 per hour

Grade (g/t) Gold Silver 11.21 12.77 12.65 11.14 7.82 8.61 8.64 9.84 5.65 2.93 3.38 2.75 3.63 2.50 1.61 6.71

Grade (g/t) Gold Silver 7.73 9.62 8.95 7.11 6.03 7.01 6.07

-1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Total

20.00 22.00 20.00 16.00 10.00 11.00 10.00 5.04 5.04 5.04 5.04 5.04 3.00 2.24 2.24

335 16,398 16,006 18,939 20,013 19,774 20,007 21,889 20,305 19,638 18,896 18,852 19,076 19,201 6,570 256,899

479 10,904 12,952 13,057 13,480 13,432 13,535 14,179 13,984 14,867 15,382 15,358 14,223 14,117 5,591 185,540

1.808 1.871 2.030 2.113 1.805 1.693 1.599 1.474 1.385 1.243 1.082 1.432 1.383 1.417 1.140 1.548

261 7,007 8,022 5,028 2,644 3,630 2,800

0.852 0.937 0.935 0.928 0.829 0.803 0.775

29,392

0.892

8.01

Process cutoff grades vary by year in order to maximize project return on investment. Cutoff grades were based on net benefit per hour milled in order to account for the different throughput rates of hard, medium and soft ores. The low grade stockpile is located north of the process plant and west of the Cetate pit. Most of the waste from the Cetate pit will be delivered to the Cetate waste storage area, located to the west of the low grade stockpile. Most of the waste from the Cirnic pit will be delivered to the Cirnic waste storage area, to the south of the Cirnic pit. The Cetate waste storage is scheduled to receive 23.6 million 172

tonnes over the life of the mine, and the Cirnic waste storage 117.1 million tonnes. An additional 49.0 million tonnes of waste are scheduled to be used in construction of the tailings impoundment. The remainder of the mined waste will be used to completely backfill the Jig pit, and partially backfill the Orlea and Cirnic pits. 18.1.2 Mine Equipment

The open pits at Rosia Montana will be mined in benches 10 m high, using the following equipment and procedures: Drilling will be performed with conventional rotary blast hole drills utilizing 25 cm diameter bits. Three such drills will be required throughout most of the mine life. Blasting will be performed with conventional ANFO explosives and ANFO-slurry blends. Blast hole cuttings will be assayed for grade control purposes. geologic data will also be recorded. Hardness information and

Ore and waste will be loaded by 19 m3 hydraulic shovels, of which three will be required throughout most of the mine life. Ore and waste will be hauled by 146-t capacity rear-dump trucks. Ore will be delivered to the primary crusher or to the low grade stockpile, and waste to the waste storage areas, the tailings impoundment or, later in the mine life, to the Jig, Orlea and Cirnic pits. A maximum fleet of 21 haul trucks will be required in production years of 8 through 13. The major mining equipment will be supported by bulldozers, graders, front-end loaders, maintenance equipment and miscellaneous mobile units.

Mine equipment requirements were estimated in accordance with standard industry procedures, based on the productivity, availability and utilization of each unit. Haul truck requirements were estimated from detailed haulage profiles for each time period, for each material type and for each destination. Truck productivity was calculated using haul time simulations over the measured haul profiles. 18.2 PROCESSING

The selection of the processing flowsheet for the Rosia Montana project has been based on extensive metallurgical testwork, as discussed in Section 16 of this report. The process is entirely conventional and consists of crushing, grinding, gravity concentration, cyanide leaching in a CIL circuit, and recovery of gold and silver by adsorption on to carbon, electrowinning and smelting. The only relatively new technology will be the use of a continuous elution circuit. The process flowsheet has been provided previously in Figure 16.1. Table 18.4 summarizes the design schedule of mill throughput and metal production. Tailings slurry from the processing plant will be treated in the cyanide detoxification circuit and pumped for permanent storage in the TMF which will be located in the Corna valley, as shown in Figure 18.1. The TMF, which will ultimately cover an area of approximately 360 ha, has sufficient capacity to store all 173

tailings produced during the life of the mine, together with run-off resulting from two Probable Maximum Precipitation events. Reclaim water from the tailings will be re-cycled back to the processing plant for reuse. The area of the TMF impoundment is underlain by a layer of colluvium which, when compacted, provides a liner of sufficient impermeability to protect groundwater from contamination. At present, it is not planned to install any additional geotextile components. Tailings will be impounded behind an engineered dam, which will be constructed in stages to an ultimate height of approximately 180 m. The dam has been designed to withstand the Maximum Credible Earthquake for this region of Romania. A secondary containment dam will be constructed, downstream of the TMF, to collect any seepage from the TMF. Water retained by the secondary dam will be pumped back to the TMF. Both the tailings dam and the secondary containment dam have been designed to comply with Romanian standards, and with the guidelines of the International Commission on Large Dams and the Canadian Dam Association. The overall design of the TMF is based on considerable field investigations of the geological, geotechnical and hydrogeological characteristics of the site.
Table 18.4 Process Plant Production Schedule Year 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Total Throughput (thousand t) 11,383 12,952 13,057 13,480 13,432 13,535 14,179 13,984 14,867 15,382 15,358 14,223 14,117 14,268 13,536 7,178 214,931 Feed Grade (g/t) Gold Silver 1.868 12.70 2.030 12.65 2.113 11.14 1.805 7.82 1.693 8.61 1.599 8.64 1.474 9.84 1.385 5.65 1.243 2.93 1.082 3.38 1.432 2.75 1.383 3.63 1.417 2.50 0.989 5.50 0.892 8.01 0.892 8.01 1.459 6.88 Recovery (%) Gold Silver 83.1 65.6 80.4 61.0 79.3 60.2 77.5 61.0 78.6 61.3 79.2 61.2 76.7 59.6 79.2 60.0 82.9 58.0 81.8 58.9 82.5 54.5 79.3 57.9 74.5 57.6 71.7 61.0 70.3 61.4 70.3 61.4 78.7% 60.8% Recovered Metal (thousand oz) Gold 567.8 679.7 703.6 606.3 574.4 551.4 515.6 493.2 492.8 437.7 583.6 501.6 479.3 325.2 272.9 144.7 7,929.8 Silver 3,048.2 3,212.5 2,816.8 2,067.3 2,280.2 2,302.0 2,671.4 1,524.2 812.6 984.2 740.7 960.3 653.6 1,540.0 2,138.6 1,134.1 28,886.7

18.3

INFRASTRUCTURE AND ANCILLARY FACILITIES

The Rosia Montana project is generally well served by existing infrastructure. Electric power will be supplied from an existing 110 kV transmission line which passes through the project site. It has sufficient capacity to supply the peak project demand of 55 MW. Some relocation of the existing line and interconnection facilities, including a substation, will be required. Fresh water requirements will be met from a pumping station located on the Aries river, near the village of Campeni. Water will be delivered to the project site through a buried pipeline 12.7 km long. Required road works include a new access road to the project site, along the left bank of Rosia Montana creek, and a by-pass road east of the Corna creek.

174

Ancillary facilities to be constructed at the site include a mine workshop for maintenance of all mobile equipment, a tank farm for fuel storage, an explosives storage magazine, fire protection systems, communications systems and office, laboratory and warehousing facilities. Extensive water management facilities will also be provided, principal among which are the Cetate water management dam and a state-of-the-art water treatment plant. Historical mining operations in the Rosia Montana area continue to produce ARD which presently flows, untreated, into local watercourses. Additional ARD is expected to be produced as a result of the proposed Rosia Montana project. The management and treatment of ARD, therefore, will be an important component of the project. Interceptor ditches, seepage collection ponds and secondary dams will be constructed at various locations across the site. These facilities will collect contaminated seepage and run-off, both from previous mining activities and from the new project. Water collected in these secondary facilities will be delivered directly to the ARD plant or to the Cetate dam, where it will be stored for subsequent processing through the water treatment plant, prior to discharge. These water management facilities have been designed not only to minimize the environmental impact of the Rosia Montana project, but also to improve significantly the quality of effluent produced from previous mining activities. 18.4 18.4.1 ENVIRONMENTAL AND SOCIO-ECONOMIC CONSIDERATIONS Overview

An EIA for the Rosia Montana project, prepared in accordance with international best practice, was filed with the Romanian government in May, 2006. In May, 2007, in response to the EIA public consultation process which took place during the summer of 2006, RMGC responded to 5,610 questions and 93 statements received by the Romanian government and judged by it to merit a response. As required under Romanian law, the Romanian government then set up the Technical Assessment Committee (TAC), comprised of officials from the various Ministries of the Romanian government involved in the permitting process to review the project, the EIA and RMGCs responses to the questions asked during the public consultation. The TAC held meetings on July 10, July 19 and August 9, 2007 which covered the first four chapters of the EIA representing the bulk of the EIA. Overall, there are 10 chapters to the EIA. These include Chapter 5 Assessment of Alternatives, Chapter 6 Monitoring, Chapter 7 Risk Cases, Chapter 8 Description of Difficulties, Chapter 9 Non Technical Summary and Chapter 10 Transboundary Impacts. The TAC meetings to date have been constructive with a thorough technical analysis of the project. Separately, RMGC participated in intergovernmental meetings between the Romanian and Hungarian governments on July 30 and 31, 2007, as required under the Espoo Convention. On September 12, 2007, RMGC received a letter from the Ministry of Environment and Sustainable Development (MESD) indicating that the review process for the EIA had been suspended. The MESD based its action on a court challenge by an NGO, opposing the project, regarding the validity of an urbanism certificate wholly unrelated to the EIA review process. RMGCs position, supported by local counsel, is that an urbanism certificate is not required for the TAC review process. Administrative complaints filed by RMGC failed to resolve the suspension issue and, as a result, RMGC filed a lawsuit against the MESD, as well as the Minister of the MESD personally and the State Secretary, in November, 2007. The lawsuits are ongoing and RMGC expects the court to rule 175

sometime during 2009. Notwithstanding the ongoing litigation relating to actions taken by a prior Romanian government, RMGC continues to pursue all options to re-start the review of its EIA with representatives of the new government of Romania, which was elected at the end of 2008 and officially took office in January, 2009. RMGC expects that the TAC process will be re-started in 2009 and that the EIA for the project will be given the fair and transparent review it is entitled to under Romanian law. 18.4.2 Environmental Considerations

Rosia Montanas long history of mining has left an environmental legacy that the project will have to address. As a result of uncontrolled, under-funded mining, toxic heavy metals in Rosias rivers and streams are high above the legal limits, which has a damaging influence on the local communitys health. Baseline and impact studies have indicated that the principal environmental considerations for the Rosia Montana project are ARD, the preservation of culturally important and archaeological structures or artifacts, and the effective reclamation and closure of the site at the end of its operating life. As discussed in Section 18.3, extensive measures are planned for the collection, control and treatment of ARD. Estimates suggest that the cost of a clean-up of the area to address ARD, and other environmental impacts from past mining activities would be likely to surpass 35 million Euro. RMGC has committed to remediate the past damage at no cost to the Romanian state. This clean-up commitment has been built into the EIA. While remediating past pollution is one benefit provided by the Rosia Montana project, it does not obviate a careful consideration of the environmental impacts of the new project. As such, the Rosia Montana project has been designed to comply with all applicable EU and Romanian laws, and seeks no exemption or exception to any applicable laws. With respect to areas of cultural and archaeological interest, the following are two of the key issues that RMGC has addressed in its design of the project: The existence of buildings of cultural significance within the Rosia Montana valley. The existence of mine working, artifacts, and traces of habitation and infrastructure which date from Roman times.

The majority of the buildings of cultural significance are contained in relatively close proximity to each other, in a location which has been designated as the Protected Area. RMGC has undertaken not to allow mining activities to encroach upon the Protected Area or the buffer zone which surrounds it. Other buildings of cultural significance, outside of the limits of the Protected Area, will be preserved to the extent possible. RMGC has funded a major program of archaeological research in the Rosia Montana area, with expenditures currently in excess of $10 million. Underground openings dating from Roman times have been made accessible and are being mapped by independent teams of French, German and Romanian archaeologists. Artifacts discovered both underground and on surface have been collected in a

176

temporary facility, open to the public, pending the planned construction of a more formal museum. Funerary monuments discovered on surface are being preserved. The effect of these measures to preserve sites of cultural and archaeological significance has been to limit the design open pits to a materially smaller surface area than would have been dictated solely by economic analysis. Closure plans for mining operations are required under Romanian law. developed by RMGC and its consultants, and is included in the EIA. A closure plan has been

The Mine Rehabilitation and Closure Plan presents the first version of a comprehensive plan for reclaiming mined areas and other areas affected by mining operations. This plan will be implemented concurrently with active mining operations as well as at closure. Closure objectives and the design criteria that will be implemented in actual closure designs are described as discrete elements of the plan. The closure and design objectives presented in the plan meet or exceed international and Romanian standards pertaining to mine closure and rehabilitation. The Mine Rehabilitation and Closure Plan will be periodically reviewed and adjusted to accommodate changes in mining operations, regulatory requirements, and other factors affecting rehabilitation or closure that may change over the life of the mine. 18.4.3 Socio-Economic Considerations

The principal socio-economic impacts of the Rosia Montana project will be: The creation of employment in the local area, with multiplier effects at a local, regional and national level. Significant financial contribution to the local, regional and federal governments of Romania. The need to relocate families from areas which will be disturbed by the project.

Mining has a 2,000 year history in Rosia Montana, starting with the underground works of the Romans in the first century AD, through the extensive mining of the Austro-Hungarian imperial era, into the 20th century period of exploitation under Romanias communist regime. The state-run mine of the communist era continued operations even after Romanias revolution in 1989. The loss-making enterprise was shut down in 2006, in advance of Romanias accession to the European Union. Today, for the first time in the living memory of Rosia Montanas residents, there is no mining taking place in the village. Unemployment in Rosia Montana, 50% even while the state-run mine was operating, now exceeds 70%; a significant percentage of the residents depend on barter and subsistence farming of family plots. The area has been officially designated a Disadvantaged Zone by the Romanian government. The economic impact of the Rosia Montana project will reverse Rosia Montanas downward economic spiral, infusing significant revenues into a rural region that has proved difficult for Romanias national government to revitalize. The project will provide an average of 1,200 jobs during its construction phase, and more than 600 jobs when it commences operation. In addition to direct employment, a project of this size and scope typically generates thousands of indirect jobs. In an area where those able to find paid work often earn a fraction of Romanias national average wage, RMGCs high-wage jobs will pay a 177

multiple of that national average. In this context, it is considered that the Rosia Montana project will have a beneficial impact on economic reality in Rosia Montana. For Romania as a whole, with two-thirds of project procurement likely to occur in the country, the Rosia Montana project will generate $1.5 to $2 billion in economic activity, making RMGC one of the countrys larger consumers of goods and services. In addition, the project will generate between $1.2 billion and $2.0 billion of revenue for the Romanian Treasury, consisting of royalties, taxes and the governments 20% share of profits from the project. With respect to the need to relocate certain residents of Rosia Montana, RMGC has designed and implemented a Resettlement and Relocation Action Plan (RRAP) which governs its actions with respect to the relocation of families or individuals. A number of families have already been relocated, and are being monitored by RMGC, which reports that all are satisfied with their new circumstances. RMGC reports that, in general, the necessary relocations are proceeding well, although, in situations such as this, it is not uncommon for a few individuals or families to resist relocation. Third party consultants acting on behalf of financial institutions have evaluated the project, including the RRAP, and detemined it to be Equator Principle compliant. Construction of the first resettlement site, Alba Iulia, consisting of approximately 128 homes, commenced in July, 2007 and is expected to be completed by June, 2009. As a result of the suspension of the EIA review process, and in order to align the RMGCs activities to the pace of the approval process, RMGC suspended the home purchase program on February 1, 2008 and plans to obtain construction permits for the second resettlement site, Piatra Alba, during the summer of 2009. As of February, 2009, RMGC owns or has options on 77 percent of the homes in the industrial zone, Protected Area and buffer zone. Once the agreements for institutional properties are completed, RMGCs ownership will rise to approximately 85 percent of the three zones of the project, further demonstrating strong local support for the project. RMGC has established a community relations office within the Protected Area in the village of Rosia Montana. This office is open to any members of the public who wish to discuss the project, or any aspect of it. The office also contains a well laid out display of Roman artifacts recovered from the area. Local support for the Rosia Montana project is strong, as evidenced by the fact that, in the local elections of June, 2008, both candidates for village mayor campaigned on a pro-mine platform. An independent report completed at roughly the same time in mid-2008, commissioned by the Rosia Montana Local Council, funded by NADMZ, the Romanian National Agency for Development of Mining Zones, and conducted by the NADMZ/Alba Regional Office, noted that: "those supporting the project represent the majority and hope that the project implementation will generate a better life, jobs and a good standard of living. 18.5 PERMITTING

As is common to most countries, Romania requires that a significant number of permits, licenses, authorizations and endorsements be obtained before the construction and/or operation of a mining

178

project can commence. Listed below are the significant permits, licenses and authorizations that need to be obtained before the Rosia Montana project can proceed to commercial operation. Environmental Impact Assessment. Industrial (Zonal) Urbanism Plan. General Urbanism Plan. Land Use Changes (forest & agricultural). Water Supply Agreement. Power Supply Agreement. Dam Safety Permit. Archaeological Discharge Certificate.

While the EIA is by far the most important project permit, there are a number of other permits and approvals required beyond those listed above, relating principally to roads, resettlement, and acquisition of surface rights, which must be received prior to applying for a construction permit. The processes for each of these permits and approvals are underway and are expected to be completed within approximately six months of EIA approval. As Gabriel, through RMGC, is the first company to permit a new mining project under the new European legislation applicable in Romania, it is pioneering with the government of Romania the permitting process. The issuance of several permits, including those permits necessary to commence construction, is contingent upon approval of the EIA. In the absence of any other extraordinary events, legal or otherwise, RMGC anticipates that it would take at least 6 months from the re-start of the permitting process to complete the EIA approval process; complete the purchase of the outstanding properties; receive all other permits and approvals, including initial construction permits; complete the control estimate for the project, and complete the required financing. Ultimately, RMGCs ability to obtain construction permits for the mine and plant is predicated on securing 100 percent of the surface rights in the industrial zone. It is has been well documented in prior public disclosure that the project has been the target of opposition of several NGOs. These NGOs have attempted to use use the courts as a tactic for delaying the permitting process for the project. While the majority of these lawsuits have been frivolous and abusive, some of them have resulted in the loss of certain permits issued to RMGC, most notably Archaeological Discharge Certificate #4. However, in most cases involving key permits, approvals or processes relating to the project, the efficacy of the actions of the relevant Romanian government agencies have been upheld by the courts. RMGC will continue to defend all permits granted to it by the Romanian government. Similarly, RMGC expects that the Romanian government agencies charged with issuing permits and approvals will do so in a manner consistent with all applicable legal provisions.

179

For much of 2007 and 2008, progress on the Rosia Montana project was blocked by political factors and a largely gridlocked governing environment. With the results of Romanias 2008 parliamentary elections, prospects for the advancement of the project are objectively more favourable now than at any point since April, 2007. Romanias new unity coalition is comprised of the countrys two largest political parties, the centreright PD-L and centre-left PSD. Both parties in the new government have shown themselves inclined towards resource development as an economic driver of Romanias growth, even as the global financial crisis has sharpened the need for Romania to welcome large-scale foreign direct investment of the sort the project would bring. 18.6 18.6.1 CAPITAL EXPENDITURE Summary

As shown in Table 18.5, the total life-of-mine capital expenditures for the Rosia Montana project are currently estimated at $1,242 million, comprising $876 million of initial, pre-production capital, and $366 million of sustaining capital to be incurred during the operating life of the property.

Table 18.5 Estimated Capital Expenditures Area Initial Capital ($ million) 89 115 99 137 158 203 73 Sustaining Capital ($ million) 89 13 152 33 45 34 366 Total Capital ($ million) 179 128 251 170 158 248 107 1,242

Mining Processing TMF Infrastructure, Utilities EPCM(1) and Indirect Costs Owners Cost Contingency

(1)

Total Capital 876 EPCM: Engineering, procurement and construction management

The capital expenditure estimate is expressed in US dollars of fourth quarter 2008 value. Excluding uncertainties arising from the global economic crisis, and the associated volatility in both commodity prices and currency exchange rates, the estimate is judged to be accurate to approximately 15%. The contingency provision included in the estimate of initial capital is approximately 9%, and is judged to be reasonable, given the quality of the underlying data. The estimate excludes the payment of value-added taxes, which are understood to be recoverable by RMGC. With the exception of surface rights to be acquired after December 31, 2008, the base date for the capital expenditure summarized in Table 18.5 is a notional three months prior to the commencement of construction, a date which is presently indeterminate. This approach has led to the exclusion from the estimate of the following items:

180

All costs for surface rights previously acquired and for the construction of the Alba Iulia resettlement site, estimated at $90 million. All payments, including future payments, on the items of long-lead-time equipment already ordered, estimated at $44 million. The cost of engineering activities undertaken prior to March, 2009, estimated at $13 million. Basis for the Estimate

18.6.2

The capital expenditure estimate was initially prepared in terms of the currency in which the expenditures are expected to be incurred. These are principally RON, Euros and US dollars, with lesser amounts in Australian and Canadian dollars. All estimates were then converted to US dollars at the exchange rates shown previously in Table 2.1. Construction labour rates have been based on surveys conducted in Romania during 2007, escalated to fourth quarter 2008 on the basis of supplemental data provided by RMGC. Labour productivity has been adjusted downward from North American and Australian standards, to account for Romanian conditions. Mine capital expenditures were estimated by IMC in accordance with procedures in common use within the mining industry. The initial mine capital expenditures consist principally of pre-production stripping, equipment costs and certain mine infrastructure. Budgetary quotations were obtained for all major equipment. The cost of minor equipment was obtained from budgetary quotations and data on file. The cost of pre-production stripping was estimated in the same manner as mine operating costs, which is described in Section 18.7. The cost of earthworks and certain infrastructure was estimated by MWH. Direct costs were developed for unit rates of production for each cost item. These unit costs were then multiplied by the relevant material quantities. Indirect costs were developed for the entire scope of work and allocated pro-rata among the individual cost items. For the estimation of direct costs, first principles estimates were prepared for every significant task for which sufficient data were available. These estimates were prepared by reducing major activities to the most detailed, or lowest level activities that could be interpreted from the original estimate. Activity sheets were then prepared for each of these detailed activities. The individual activity sheet identifies the total production quantity required and itemizes each piece of equipment needed to perform the task, the total staffing requirements and any materials required per crew. The productivity for each crew was calculated by using the assumed machine capacities, operating speeds and efficiencies, factoring in travel distances, load sizes, loading cycle times, manoeuvering times and waiting periods as appropriate. The total crew hours required were then calculated using the crew production rate and required production quantity. The hours per shift worked by each crew, the number of shifts worked per day and the assumed overall efficiency of the crew were used to determine total production days required for input to the construction schedule. Cost per unit rate of production was obtained by adding the hourly cost of each piece of equipment or staff member

181

and dividing the result by the hourly production rate. This cost per unit of production was then multiplied by the required quantity to obtain the activity total cost. Material quantities were obtained from estimates prepared by SNC in 2007, with some optimization of earthworks for roads and plant site excavations performed by MWH. The capital expenditure estimates for all project facilities, other than those developed by IMC and MWH, were prepared by Metifex. The starting point for the Metifex estimate was the work performed by SNC in 2007 in the preparation of the definitive capital expenditure estimate. SNCs work was halted, prior to completion, following suspension by the Romanian government of the EIA review process. SNC provided a close-out report on the status of its work, and that report was reviewed and expanded upon by Metifex. Based on the information provided by SNC, Metifex performed all of its basic cost estimates in US dollars of mid-2007 value. The estimates were then escalated to US dollars of fourth quarter 2008 value, using appropriate inflation indices. In the course of its work, Metifex made certain modifications to the designs prepared by SNC. The capital expenditures estimates for the process plant and those items of infrastructure included in the Metifex scope of work were developed on the following bases. Budgetary quotations were obtained by SNC for all major equipment units. The cost of minor units of equipment was obtained from data on file, or by factoring. Quantities of concrete, structural steel and like items were based on quantity take-offs from general arrangement drawings, single line electrical diagrams, and piping and instrumentation diagrams. Unit prices of bulk materials were based on budgetary quotations obtained from potential Romanian suppliers. The major components of EPCM and indirect costs were estimated from first principles. Minor components were factored. Contingency allowances were allocated on the basis of the adjudged quality of the underlying estimate.

The Owners component of capital expenditure was estimated by RMGC, and the estimates were reviewed by one of the authors of this report. Owners capital expenditures were estimated on the following bases: Budgetary quotations were obtained for resettlement site construction. The cost of acquiring surface rights was based on the approved purchase prices used by RMGC since mid-2007. These prices have not been escalated, as Romanian property values deflated in the latter half of 2008. Engineering consulting services were estimated from first principles. Provision was made for the costs of RMGCs own project team. Provision was made for the payment of those Romanian taxes which will be levied during the construction period. Principal among these is the forestry land use change tax, which 182

was estimated on the basis of a legislative proposal. This proposal has been recently enacted. 18.7 OPERATING COSTS

Average life-of-mine unit cash operating costs have been estimated at $12.57 per tonne processed, as summarized in Table 18.6. These costs are expressed in US dollars of fourth quarter 2008 value. The general and administration (G & A) cost includes the off-site costs to be incurred for the transport and refining of the dor bars produced at site.
Table 18.6 Estimated Average Cash Operating Cost Component Mining Processing G & A and off-site Total Unit Cost ($/tonne processed) 2.88 8.23 1.46 12.57

The unit mining cost of $2.88 per tonne processed is equivalent to $1.31 per tonne of material mined. Operating costs have been estimated in accordance with standard industry procedures. Labour costs have been based on the project manning charts and a survey of salaries and wages conducted by RMGC in 2007, which has been escalated to the fourth quarter of 2008. Provision has been made for a small contingent of expatriate managers. The cost of process reagents and consumables has been based on consumption levels determined through metallurgical testwork and application of recognized empirical consumption calculations, coupled with budgetary price quotations obtained from Romanian and international suppliers. Mine consumables have been based on manufacturers recommendations, in-house data and representative unit prices. Maintenance costs have either been factored as a percentage of capital cost, or have been based on manufacturers recommendations and in-house data. The operating cost estimates do not include value-added taxes. The cost of diesel fuel, delivered to the site, has been estimated by RMGC at $0.89 per litre. Annual consumption is estimated at 14.6 million litres. The cost of electric power has been estimated at $0.083/kWh, based on an analysis of the Romanian electric market conducted by a consultant. The cost of sodium cyanide has been estimated at 1,600 Euros per tonne. The average cash cost of producing gold, after deducting silver credits, but including royalty and other taxes, is estimated to average $272 per ounce over the first five years of production and $335 per ounce over the life of the mine, as summarized in Table 18.7.
Table 18.7 Unit Cash Cost of Gold Production Component Mining Processing G&A Offsite First 5 Years 70 181 29 3 Unit Cost ($/oz) Life-of-Mine 78 223 37 3

183

Royalty, other taxes Silver credit Total

33 (44) 272

32 (38) 335

18.8 18.8.1

PROJECT ECONOMICS Base Case

The overall economics of the Rosia Montana project have been evaluated by conventional discounted cash flow techniques, based on the production schedules, capital expenditures and operating costs discussed in this report, together with the following additional parameters: The metal prices used for the base case economic analysis are $750 per ounce for gold and $10.50 per ounce for silver. The analysis is based on 100% equity financing, with no debt component. Revenues received from the sale of silver are treated as a by-product credit against operating costs. A tax rate of 16% has been applied, based on information provided by RMGC. Provision has also been made for a net smelter return royalty of 4% payable to the government of Romania. A provision of $128 million has been included for final reclamation and closure.

Under the estimates and assumptions used for the base case analysis, the Rosia Montana project would be expected to generate an undiscounted, life-of-mine cash flow, after tax, of $1.66 billion, a net present value of $997 million at a discount rate of 5% per year, and an after tax internal rate of return of 20.4% per year. The payback period is estimated at 3.5 years. Table 18.8 presents a summary of the life-of-mine results of the base case financial evaluation of the Rosia Montana project, and a comparison of these results with the estimates contained in the 2006 Technical Report. It is evident that, in common with mining projects throughout the world, the Rosia Montana project has experienced significant escalation in both capital expenditures and operating costs over the last three years. On the other hand, the price of gold has also increased, with the result that the overall economics of the project today are more favourable then they were in early 2006.
Table 18.8 Summary of Financial Analysis (life-of-mine)

Item Gold produced Silver produced Average gold production Average silver production Cash cost Pre-production capital Sustaining capital Closure cost

Units Moz Moz koz/y koz/y $/oz $M $M $M

This Report 7.93 28.89 511 1,860 335 876 366 128

2006 Report 7.94 28.89 509 1,852 237 638 208 70

184

Undiscounted cash flow, after tax NPV after tax, 5% discount IRR after tax Payback

$M $M %/y y

1,662 997 20.4 3.5

1,022 498 17.6 3.8

18.8.2

Sensitivity Analysis

The economics of the Rosia Montana project are most sensitive to variations in gold price and mined grade. They are less sensitive to variations in capital expenditure or operating cost. Table 18.9 summarizes the sensitivity to variations in the base case gold price of $750/oz.
Table 18.9 Sensitivity to Gold Price Gold price ($/oz) Undiscounted NPV ($ million) Internal rate of return (%/yr) Payback period (years) 600 688 10.9 5.2 750 1,662 20.4 3.5 900 2,621 28.0 2.7

Additional sensitivity analyses are shown in the following illustrations: Figure 18.2: Sensitivity to operating cost, capital expenditures and gold grade. Figure 18.3: Sensitivity to the cost of major consumables. Figure 18.4: Sensitivity to variations in currency exchange rates. Figure 18.5: Sensitivity of cash cost per ounce to variations in power costs and currency exchange rates. Figure 18.6: Sensitivity of initial capital expenditure to variations in currency exchange rates.
Figure 18.2 Project Sensitivity

$4,000,000 Thousands of US dollars - NAV 0% $3,500,000 $3,000,000 $2,500,000 $2,000,000 $1,500,000 $1,000,000 $500,000
100% 90% 75% 80% 70% 85% 95% Processing Costs Capital Costs Grade Au Mining Costs 115% 120% 105% 110% 125% 130% 135%

$0

185

% Variable

of

Figure 18.3 Operating Cost: Commodity Sensitivity

$1,900,000 Thousands of US dollars - NAV 0% $1,850,000 $1,800,000 $1,750,000 $1,700,000 $1,650,000 $1,600,000 $1,550,000 $1,500,000 -30% -25% -20% -15% -10% -5% 0% 5% 10% 15% 20% 25% 30%
Diesel Sodium Cyanide Sodium Metabisulfite Process Power Grinding Media

% Change Price

in

Commodity

Figure 18.4 Operating Cost: Currency Sensitivity

Thousands of US dollars - NAV 0%

$2,200,000 $2,000,000 $1,800,000 $1,600,000 $1,400,000 $1,200,000 $1,000,000 $800,000 -30% -25% -20% -15% -10% -5% 0% 5% 10% 15% 20% 25% 30%
EURO RON

% Change in Currency Compared to the US dollar

186

Figure 18.5 Cash Cost per Ounce Sensitivity: Life-of-Mine

$60 $40 $20 $-30% -25% -20% -15% -10% -5% $(20) $(40) $(60) 0% 5% 10% 15% 20% 25% 30%

US dollars

Process Power EURO RON

% Change in Variable

Figure 18.6 Capital to Completion Sensitivity Capital Exchange Rate Foreign to Completion Sensitivity Foreign Exchange Rates

$100,000
Thousands of US dollars

$50,000 $0

EURO

0%

-2 5% -2 0% -1 5% -1 0%

5%

0%

-5 %

10 %

20 %

-$50,000

15

25

-$100,000
Change in cur r ency relative to US dollar

On the basis of the updated estimates discussed in this report, it is concluded that the Rosia Montana project remains both technically feasible and economically viable.

-3

187

30

RO N

19.0 INTERPRETATION AND CONCLUSIONS

The exploration activities undertaken by RMGC since 1998 have delineated a significant gold deposit, with by-product silver, on the Rosia Montana property. Updated estimates of capital expenditure and operating costs, recently completed, have confirmed the technical feasibility and economic viability of producing an estimated 7.9 million ounces of gold and 29 million ounces of silver from the property, over an operating life of approximately 16 years, from an estimated Proven and Probable Mineral Reserve of 215 million tonnes at an average grade of 1.46 g/t Au and 6.9 g/t Ag. This reserve is contained in four open pits which will be mined conventionally by shovels and trucks. The process plant feed will be ground to 80% minus 150 m and gold and silver will be recovered as dor bars by conventional gravity concentration, CIL processing, electrowinning and smelting techniques. A comprehensive Environmental Impact Assessment was completed early in 2006 and was submitted to the Romanian authorities for review. The review process was suspended by the Romanian government in September, 2007, and remains suspended at the date of this report. The timing at which construction will commence thus remains uncertain. In the interim, RMGC has placed orders totalling approximately $44 million for major equipment items with long lead times, including the primary crusher, the SAG mill and ball mills, and mill drive systems. A number of families also remain to be relocated before construction can commence. For much of 2007 and 2008, progress on the Rosia Montana project was blocked by political factors and a largely gridlocked governing environment. With the election of a new coalition government in Romania, which appears to favour resource development, it is RMGCs opinion that the prospects for advancement of the project are more favourable now than at any point since April, 2007. The geological model of the mineralization has been developed over a number of years, and is considered to be well established and of high confidence. Variographic studies, however, indicate that the mineralization is characterized by a high nugget effect, suggesting that the mineral resource estimate, while globally robust, is unlikely accurately to have predicted all local variations in grade. Strict grade control procedures will therefore be required during operation. It is unlikely, also, that all of the underground workings resulting from previous mining have yet been accurately located. Further work in this area is recommended prior to the commencement of open pit operations. Since, with the exception of continuous elution, only conventional, well-proven processes will be used at Rosia Montana, the overall level of technical risk is considered to lie within the range of normal mining risk. On the basis of the discussion contained within the body of this report, it is concluded that the Rosia Montana project is both technically feasible and economically viable, and that the main challenge to be overcome before the project can be brought to fruition lies in the area of permitting. While RMGC is considered to have appropriate plans and strategies in place to deal with this challenge, the outcome of the permitting process is not fully within its control.

188

20.0 RECOMMENDATIONS
The principal conclusion arising from this review of the Rosia Montana project is that project implementation remains contingent on obtaining all of the permits necessary to enable construction to commence. It is recommended, therefore, that RMGC maintain its focus on the entire permitting process. It is recommended, also, that the resettlement and relocation process be advanced to the extent consistent with maintaining the support of the local community for the project. The effective date of this report is March 4, 2009.

Brett L. Gossage Brett L. Gossage, MAusIMM Coffey Mining Pty. Ltd.

Patrick G. Corser Patrick G. Corser, P.E. MWH Americas, Inc.

John M. Marek John M. Marek, P.E. Independent Mining Consultants, Inc.

Stuart Smith Stuart Smith, MAusIMM Aurifex, Pty. Ltd.

Christopher R. Lattanzi Christopher R. Lattanzi, P.Eng. Micon International Limited.

189

21.0 REFERENCES
Gossage, B.L. et al., March, 2006. Technical Report on the Rosia Montana Gold Project, Transylvania, Romania. Ianovici, V., Borco, M., Bleahu, M., Patrulius, D., Lupu, M., Dimitrescu, R., Savu, H. (1976): Geologia Munilor Apuseni. Ed. Acad., 631 p., Bucureti. Isaaks, EH and Srivastava, RM 1989. An Introduction to Applied Geostatistics. Oxford University Press, New York. Leach, T., Hawke, M., (1997): Petrographic study of 32 samples from Rosia Montana. Unpublished company report. Manske, S., Ullrich, T., Reynolds, T.J., and OConnor, G., 2004, Vein sets and hydrothermal alteration in the Cetate Cirnic area, Roia Montana district, Romania, in Rosu, E., ed., Gold in the Metaliferi Mountains: Romanian Journal of Mineral Deposits Special Issue v. 81, p. 122-125. Marza, I., Tamas, C. G., and Ghergari, L., 1997, Low sulfidation epithermal ore deposits from Rosia Montana, Metaliferi Mountains, Romania: Stud. Cerc. Geol., v. 42, p. 3-12. Meyer, C., and Hemley, J., 1967. Wall Rock Alteration, in Barnes, H.L.(ed.), Geochemistry of hydrothermal ore deposits; New York, Holt, Reynolds and Winston, p. 166-203. Pcskay, Z., Edelstein, O., Sehedi, I., Szakcs, A., Kovacs, M., Crihan, M., and Bernad, A., 1995, K-Ar datings of Neogene-Quaternary calc-alkaline volcanic rocks in Romania: Acta Vulkanologica, v. 7, p. 53-61. Roscoe Postle Associates, Inc., March, 2004. Technical Report on the Rosia Montana Project, Romania. Rosu, E., Udubasa, G., Pcskay, Z., Panaiotu, C., and Panaiotu, C. E., 2004, Timing of Miocene-Quaternary magmatism and metallogeny in the south Apuseni mountains, Romania, in Rosu, E., (ed.), Gold in the Metaliferi mountains, Romanian Journal of Mineral Deposits Special Issue v. 81, p. 33-38. Sillitoe, R. H., and Hedenquist, J. W., 2003, Linkages between volcanotectonic settings, orefluid compositions, and epithermal precious metals deposits, in Simmons, S. F., and Graham, I., eds., Volcanic, geothermal, and ore-forming fluids: Rulers and witnesses of processes within the earth: SEG Special Publication No. 10, p. 315-343. Srivastava, R.M., and Parker, H.M., 1989. Robust measures of spatial continuity. M. Armstrong, editor, Geostatistics, pages 295-308. Kluwer, Dordrecht. In

Tma, C. G., 2002, Breccia pipe structures related to some hydrothermal ore deposits in Romania: Unpublished Ph.D. dissertation (abridged), Cluj-Napoca, Romania, BabeBolyai University, 33 p.

190

Tma, C. G., Bailley, L., and Cauuet, C., 2004, Breccia structures and Au-Ag mineral assemblages in the Roia Montana ore deposit, Apuseni mountains, Romania, in Cook, N. L., and Ciobanu, C., eds., Au-Ag-telluride deposits of the Golden Quadrilateral, Apuseni mountains, Romania: Guidebook of the international field workshop of the IGCP project 486, IAGOD Guidebook Series 12, p. 254-255.

191

22.0 CERTIFICATES

192

CERTIFICATE OF AUTHOR BRETT L. GOSSAGE, MAusIMM

As a co-author of the report entitled Technical Report on the Rosia Montana Gold Project, Transylvania, Romania, dated March 4, 2009, I, Brett L. Gossage, do hereby certify that: 1. 2. I am employed by, and carried out this assignment for, Coffey Mining Pty. Ltd., 1162 Hay Street, West Perth, Western Australia, tel. 61-8-9324-8800, fax. 61-8-9324-8877. I hold the following academic qualifications: I am a graduate of Curtin University and hold a Bachelor of Applied Science in Geology (1988) and have obtained a Post Graduate Certificate in Geostatistics from Edith Cowan University (1999). 3. 4. 5. 6. 7. 8. 9. I am a member of the Australasian Institute of Mining and Metallurgy (112606). I have practiced my profession continuously since 1989. I am familiar with NI 43-101 and, by reason of education, experience and professional registration, I fulfill the requirements of a Qualified Person as defined in NI 43-101. I have read NI 43-101 and this Technical Report has been compiled in compliance with that instrument. I am responsible for the preparation of Sections 7 through 14, 17.1 through 17.6, and portions of Section 6 of the Technical Report. I visited the Rosia Montana property in 2002. I supervised the work of colleagues who were on site and directly involved in the project up to June, 2005. I was a co-author of a prior Technical Report on the Rosia Montana property, dated March, 2006. I have performed consulting services for Rosia Montana Gold Corporation between 1998 and February, 2006. As of the date of this certificate, 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 this report not misleading. I am independent of the parties for which this report is required, other than providing consulting services. Dated this 4th day of March, 2009 Brett L. Gossage Brett L. Gossage, MAusIMM

10.

11.

193

CERTIFICATE OF AUTHOR PATRICK G. CORSER, P.E.

As a co-author of the report entitled Technical Report on the Rosia Montana Gold Project, Transylvania, Romania, dated March 4, 2009, I, Patrick G. Corser, do hereby confirm that: 1. 2. I am employed by, and carried out this assignment for, MWH Americas, Inc., 1475 Pine Grove Road, Steamboat Springs, Colorado, 80487, tel. (970)-879-6260, fax (970)-879-9048. I hold the following academic qualifications: Bachelor of Science, Civil Engineering from the University of Minnesota, 1977 Master of Science, Geotechnical Engineering from Northwestern University, 1979 I am a Professional Engineer who is licensed in thirteen states. I have practiced my profession for more than 29 years. I am familiar with NI 43-101 and, by reason of education, experience and professional registration; I fulfill the requirements of a Qualified Person as defined in NI 43-101. I have read NI 43-101 and this Technical Report has been compiled in compliance with that instrument. I am responsible for the preparation of portions of Section 18.6 of the Technical Report. I have visited the Rosia Montana property on several occasions, most recently in 2006. I was a co-author of a prior Technical Report on the Rosia Montana property, dated March, 2006. I have performed geotechnical site investigations, design studies, permitting studies and public consultation for Rosia Montana Gold Corporation since 2002. As of the date of this certificate, 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 this report not misleading. I am independent of the parties for which this report is required, other than providing consulting services. Dated this 4th day of March, 2009 Patrick G. Corser Patrick G. Corser, P.E.

3. 4. 5. 6. 7. 8. 9.

10.

11.

194

CERTIFICATE OF AUTHOR JOHN M. MAREK, P.E.

As a co-author of the report entitled Technical Report on the Rosia Montana Gold Project, Transylvania, Romania, dated March 4, 2009, I, John M. Marek, do hereby certify that: 1. 2. I am employed by, and carried out this assignment for, Independent Mining Consultants, Inc., 3560 E. Gas Road, Tucson, Arizona 85714. 520-294-9861. I hold the following academic qualifications: B.Sc. (mineral engineering physics) M.Sc. (mining engineering) 3. Colorado School of Mines Colorado School of Mines 1974 1976

I am a registered Professional Mining Engineer in Arizona (12772) and a registered Professional Engineer in Colorado (16191); as well, I am a registered member in good standing of the American Institute of Mining and Metallurgical Engineers, Society of Mining Engineers. I have worked as a mining engineer in the minerals industry for 33 years. I am familiar with NI 43-101 and, by reason of education, experience and professional registration, I fulfill the requirements of a Qualified Person as defined in NI 43-101. I have read NI 43-101 and to the best of my knowledge, this Technical Report has been compiled in compliance with that instrument. I am responsible for the preparation of Sections 17.7 and 18.1, and portions of Sections 18.6 and 18.7 of the Technical Report. I visited the Rosia Montana property in September, 2002 and August, 2003. I was a co-author of a prior Technical Report on the Rosia Montana property, dated March, 2006. I have performed mine planning and cost estimation services for Rosia Montana Gold Corporation since 2002. As of the date of this certificate, 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 this report not misleading. I am independent of the parties for which this report is required, other than providing consulting services. Dated this 4th day of March, 2009 John M. Marek John M. Marek, P.E.

4. 5. 6. 7. 8. 9.

10.

11.

195

CERTIFICATE OF AUTHOR STUART J. SMITH, MAusIMM

As a co-author of the report entitled Technical Report on the Rosia Montana Gold Project, Transylvania, Romania, dated March 4, 2009, I, Stuart J. Smith, do hereby certify that: 1. 2. I am employed by, and carried out this assignment for, Aurifex Pty. Ltd., 44 Coral Tree Avenue, Subiaco, Western Australia, tel. 61-4-1793-2509, fax. 61-8-9388-0997. I hold the following academic qualifications: B. App. Sc. (Metallurgy) 3. 4. 5. 6. 7. 8. 9. 10. Bendigo College of Advanced Education, Australia 1984

I am a member of the Australasian Institute of Mining and Metallurgy. I have worked as a metallurgist in the minerals industry for more than 20 years. I am familiar with NI 43-101 and, by reason of education, experience and professional registration, I fulfill the requirements of a Qualified Person as defined in NI 43-101. I have read NI 43-101 and this Technical Report has been compiled in compliance with that instrument. I am responsible for the preparation of Sections 16 and 18.2, and portions of Sections 18.7 of the Technical Report. I visited the Rosia Montana property in November of 2003. I was a co-author of a prior Technical Report on the Rosia Montana property, dated March, 2006. As of the date of this certificate, 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 this report not misleading. I am independent of the parties for which this report is required, other than providing consulting services. Dated this 4th day of March, 2009 Stuart J. Smith Stuart J. Smith, MAusIMM

11.

196

CERTIFICATE OF AUTHOR CHRISTOPHER R. LATTANZI, P.Eng.

As a co-author of the report entitled Technical Report on the Rosia Montana Gold Project, Transylvania, Romania, dated March 4, 2009, I, Christopher R. Lattanzi, do hereby certify that: 1. 2. I am employed by, and carried out this assignment for, Micon International Limited, Suite 900, 390 Bay Street, Toronto, Ontario, M5H 2Y2, tel. (416) 362-5135, fax (416) 362-5763. I hold the following academic qualifications: B. Eng. (Mining) 3. University of Melbourne, Australia 1959

I am a registered Professional Engineer with the Association of Professional Engineers of Ontario (membership number 25705013); as well, I am a member in good standing of the Canadian Institute of Mining, Metallurgy and Petroleum. I have worked as a mining engineer in the minerals industry for more than 45 years. I am familiar with NI 43-101 and, by reason of education, experience and professional registration, I fulfill the requirements of a Qualified Person as defined in NI 43-101. I have read NI 43-101 and this Technical Report has been compiled in compliance with that instrument. I am responsible for the preparation of Sections 1, 2, 3, 4, 5, 15, 18.3, 18.8, 19 and 20, and portions of Sections 6, 18.6 and 18.7 of the Technical Report. I visited the Rosia Montana property in March, 2006. I was a co-author of a prior Technical Report on the Rosia Montana property, dated March, 2006. As of the date of this certificate, 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 this report not misleading. I am independent of the parties for which this report is required, other than providing consulting services. Dated this 4th day of March, 2009 Christopher R. Lattanzi Christopher R. Lattanzi, P.Eng

4. 5. 6. 7. 8. 9. 10.

11.

197