You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The assessment of mineralogical properties and hydrological or physicochemical controls on the drainage… St-Arnault, Mélanie 2020

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2020_may_starnault_melanie.pdf [ 8.36MB ]
Metadata
JSON: 24-1.0389805.json
JSON-LD: 24-1.0389805-ld.json
RDF/XML (Pretty): 24-1.0389805-rdf.xml
RDF/JSON: 24-1.0389805-rdf.json
Turtle: 24-1.0389805-turtle.txt
N-Triples: 24-1.0389805-rdf-ntriples.txt
Original Record: 24-1.0389805-source.json
Full Text
24-1.0389805-fulltext.txt
Citation
24-1.0389805.ris

Full Text

  THE ASSESSMENT OF MINERALOGICAL PROPERTIES AND HYDROLOGICAL OR PHYSICOCHEMICAL CONTROLS ON THE DRAINAGE CHEMISTRY  OF MINE WASTE ROCK   by MÉLANIE ST-ARNAULT B.Sc., Université du Québec à Montréal, 2001 M.Sc., Université du Québec à Montréal, 2005    A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY   in   THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Mining Engineering)        THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)     April 2020  © Mélanie St-Arnault, 2020 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  The assessment of mineralogical properties and hydrological or physicochemical controls on the drainage chemistry of mine waste rock submitted by Mélanie St-Arnault in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Mining Engineering  Examining Committee: Prof. K. Ulrich Mayer, Geological Sciences, UBC Co-supervisor/committee member Prof. Roger Beckie, Geological Sciences, UBC Co-supervisor/committee member Prof. Bern Klein, Mining Engineering, UBC Supervisor Prof. Maria Holuszko, Mining Engineering, UBC University Examiner Prof. Craig Hart, Geological Sciences, UBC University Examiner Mansour Edraki, Sustainable Mineral Institute, UQ External Examiner iii  Abstract Mining exploitation produces a substantial amount of waste rock that, when exposed to water and oxygen, may liberate metals and acidity in mine drainage. Waste rock and mine drainage should therefore be carefully managed to avoid the propagation of harmful elements in the environment. Proper management strategies include geochemical characterization of waste rock, water quality monitoring, long-term predictions, and thorough knowledge of mobility controls of solutes in mine waste. This study investigates the relationship between waste rock weathering, processes controlling mineral reactivity or solute mobility, and drainage chemistry in field barrel kinetic tests or within the full-scale waste-rock pile. The main findings of this work are that: 1) Waste rock reactivity, evaluated from mineral liberation and association indices that are calculated from automated mineralogy data, complements bulk geochemistry and lithological characterization and improves drainage quality predictions; 2) Mechanisms of secondary mineral precipitation, sorption, surface passivation and galvanic reactions are linked to potential mobilization or inhibition of metals in the drainage data; 3) Mineralogical composition and weathering patterns are interrelated with hydrological responses, providing an initial qualitative indication of hydrological and geochemical processes controlling drainage chemistry in field barrels; 4) Reactive transport modeling provides site-specific calibrated mineral weathering rates illustrating the control of hydrological, physicochemical, and mineralogical processes on drainage chemistry; 5) High-resolution sampling and analyses of waste rock weathering and drainage of reactive zones within the waste-rock pile reveal high degrees of physicochemical heterogeneity that can hardly be described by static laboratory testing or sensor measurements alone. Overall, this thesis is facilitating the interpretation of waste-rock reactivity using waste-rock quantitative mineralogical evaluations; providing mineralogical perspectives on geochemical and hydrological iv  processes; expanding the identification of mechanisms affecting metal mobility from waste-rock drainage; and, providing information to achieve more accurate drainage prediction models. As improvements are made to the automation of mining operations at large scales, more information can be gathered cost-effectively. This novel mineralogical approach, adding new perspectives on weathering processes impacting mine drainage, could therefore be applied to other mine sites, ultimately allowing for the optimization of waste-rock management as well as long-term drainage predictions.     v  Lay Summary The extraction of metals from mining can lead to the production of large quantities of waste rock. The weathering of waste rock may liberate elements that could potentially affect the quality of mine drainage. Accurate leaching predictions may help with implementation of effective measures to prevent or contain potential negative outcomes on the environment. Unfortunately, the complexity and lack of information on processes occurring within large-scale waste-rock piles complicate these predictions. This study aims to improve knowledge on mineral reactivity and processes controlling the transfer of elements to drainage water. The main findings provide valuable knowledge on the heterogeneous distribution of processes within the waste-rock pile, as well as new understandings on processes that govern the mobility of elements during the weathering of waste rock by adding perspectives on the reactivity of minerals. These new insights will ultimately contribute to achieve more accurate prediction models and the optimization of waste-rock management.    vi  Preface The research chapters were originally prepared following a manuscript format for publication in peer reviewed journals. I am the main author of all manuscripts/chapters and the contributions of co-authors and colleagues are listed by chapter.  Chapter 2 The manuscript version of this chapter was published in Applied Geochemistry. The full reference is: St-Arnault, M., Vriens, B., Klein, B., Mayer, K.U., Beckie, R.D., 2019. Mineralogical controls on drainage quality during the weathering of waste rock. Appl. Geochemistry 108, 104376.  The technicians and professionals of the Antamina mine provided technical and logistical support during the construction, monitoring, and sampling of the field barrels. The waste rock and water samples were sent to an external laboratory for analyses. I personally sampled and sieved the samples from the field barrels to prepare for mounting on thin sections and round mounts. I analyzed the thin sections with SEM-EDS with the analytical support of Jenny Lai. The elemental content was measured using a portable field FP-XRF with coaching from Libin Tong. The Teck Technical Services from Trail, BC allowed access to the Mineral Liberation Analyzer (MLA). The technicians and professionals provided time and support to perform the data collection and analysis with the MLA. I chose the approach used to analyze the data and the equations were adapted from Aranda (2010) and Lund (2015). Dr Bas Vriens assisted, in part, with the interpretation of data and with the editing of the manuscript. Professors Bern Klein, Roger Beckie, and Uli Mayer provided feedback on the manuscript. As part of the peer reviewed process, two anonymous reviewers also commented on the manuscript prior to publication.   vii  Chapter 3 The manuscript version of this chapter will be submitted for publication in a peer-reviewed journal.   I implemented the tracer tests experiment based on the design of Blackmore (2016). I sampled the waste rock and drainage from field barrels, conducted mineralogical characterization, interpreted data, and wrote the manuscript. The technicians and professionals of the Antamina mine provided technical and logistical support during the application of the tracer tests as well as on-going sampling of drainage from the field barrels. An external laboratory analyzed the waste-rock and water samples from the field barrels. I sieved and prepared the waste-rock samples for mounting on polished thin sections and round mounts. Libin Tong coached me on the analytical instrumentation of FP-XRF. I scanned the thin sections and performed the point counting. The access to the Mineral Liberation Analyzer (MLA) was granted by Teck Technical Services (Trail, BC). The technician and professionals at Teck provided technical support for data collection and analysis with MLA. Professors Bern Klein, Roger Beckie, and Uli Mayer contributed guidance and comments throughout the project. Bas Vriens, Mehrnoush Javadi, Elliott Skierszkan, and Maria Eliana Lorca commented on the manuscript and/or helped with the interpretation of data.  Chapter 4 The manuscript version of this chapter was published in Minerals Engineering. The full reference is: St-Arnault, M., Vriens, Blaskovich, R., Aranda, C., B., Klein, B., Mayer, K.U., Beckie, R.D., 2019. Geochemical and mineralogical assessment of reactivity in a full-scale heterogeneous waste-rock pile. Minerals Engineering 145, 106089. The technicians and professionals of the Antamina mine provided technical and logistical support during the drilling and sampling of the boreholes. Antamina’s geologists performed the viii  lithological description of the drill core and I did the qualitative descriptions of the core. The waste-rock and water samples were sent to an external laboratory for analyses. I sieved and prepared the waste-rock samples for mounting on polished thin sections and round mounts. I performed Raman spectroscopy analyses with coaching from Dr. Matthijs Smit, XRD analyses with the support of Jenny Lai, as well as FP-XRF elemental measurements with the help of an undergraduate student Elaine Baluyut. The Teck Technical Services in Trail, BC allowed access to the Mineral Liberation Analyzer and provided time and technical support for the collection and analysis of data. I selected the data analyses approach and the equations applied were adapted from Aranda (2010) and Lund (2015). I interpreted the data and wrote the manuscript with the support of Bas Vriens as well as contributions from Professors Bern Klein, Roger Beckie, and Uli Mayer. An anonymous peer-reviewer also suggested improvements to the manuscript as part of the publication process.     ix  Table of Contents Abstract ......................................................................................................................................... iii Lay Summary .................................................................................................................................v Preface ........................................................................................................................................... vi Table of Contents ......................................................................................................................... ix List of Tables ................................................................................................................................xv List of Figures ............................................................................................................................ xvii List of Abbreviations ................................................................................................................ xxii Acknowledgements .................................................................................................................. xxiv Dedication ................................................................................................................................. xxvi Chapter 1: Introduction ................................................................................................................1 1.1  Controls on drainage chemistry ...................................................................................... 2 1.1.1  Acid rock drainage, neutral rock drainage and metal leaching ................................... 3 1.1.2  Microbial activity ........................................................................................................ 6 1.1.3  Mineral reactivity ........................................................................................................ 7 1.1.3.1  Mineral coating or passivation ............................................................................ 8 1.1.3.2  Galvanic reactions ............................................................................................... 9 1.2  Research site ................................................................................................................. 11 1.3  Thesis scope and organization ...................................................................................... 14 1.3.1  Objectives ................................................................................................................. 14 1.3.2  Thesis organization ................................................................................................... 20 Chapter 2: Mineralogical controls on drainage quality during the weathering of waste rock ................................................................................................................................................21 x  2.1  Introduction ................................................................................................................... 21 2.2  Materials and methods .................................................................................................. 23 2.2.1  Field barrel kinetic testing experiments .................................................................... 23 2.2.2  Chemical analyses of waste rock and drainage ......................................................... 24 2.2.3  Preferential elemental retention or mobilization ....................................................... 25 2.2.4  Quantitative mineralogical analysis .......................................................................... 26 2.2.5  Mineralogical data processing .................................................................................. 27 2.2.6  Simulation of drainage quality .................................................................................. 28 2.3  Results ........................................................................................................................... 29 2.3.1  Evolution of drainage chemistry ............................................................................... 30 2.3.2  Identification of preferential mobilization or attenuation ......................................... 32 2.3.3  Mineralogical observations ....................................................................................... 34 2.3.3.1  Intrusive waste rock .......................................................................................... 34 2.3.3.2  Skarn waste rock ............................................................................................... 37 2.3.4  Reactive transport modeling ..................................................................................... 42 2.4  Discussion ..................................................................................................................... 45 2.4.1  Mineral modal composition, reactivity, and liberation ............................................. 45 2.4.2  Galvanic reactions ..................................................................................................... 46 2.4.3  Secondary mineral formation: surface passivation and sorption .............................. 47 2.4.4  Simulation of mineralogical controls on drainage chemistry ................................... 48 2.5  Conclusions ................................................................................................................... 49 Chapter 3: Relationship between flow paths, patterns of weathered minerals, and drainage chemistry from field barrel kinetic tests ....................................................................................51 xi  3.1  Introduction ................................................................................................................... 51 3.2  Methods......................................................................................................................... 57 3.2.1  Kinetic testing: field barrel experiment and sampling .............................................. 57 3.2.2  Waste rock and drainage elemental analysis ............................................................ 58 3.2.3  Waste-rock mineralogical analysis ........................................................................... 58 3.2.4  Drainage concentration and loads ............................................................................. 59 3.2.5  Tracer test .................................................................................................................. 60 3.2.6  Mineralogical data .................................................................................................... 61 3.2.7  Reactive transport modeling ..................................................................................... 63 3.3  Results and discussion .................................................................................................. 64 3.3.1  Field barrel material characteristics .......................................................................... 64 3.3.2  Tracer tests ................................................................................................................ 65 3.3.3  Mineralogy ................................................................................................................ 67 3.3.4  Comparison of water regime and mineralogical weathering patterns ...................... 70 3.3.5  Drainage quality ........................................................................................................ 72 3.3.5.1  Comparison of Field barrels 2B, 2C, and 3A.................................................... 73 3.3.6  Drainage simulation .................................................................................................. 77 3.4  Conclusion .................................................................................................................... 81 Chapter 4: Geochemical and mineralogical assessment of reactivity in a full-scale heterogeneous waste-rock pile ....................................................................................................83 4.1  Introduction ................................................................................................................... 83 4.2  Materials and methods .................................................................................................. 86 4.2.1  Site description and borehole drilling ....................................................................... 86 xii  4.2.2  Bulk physicochemical analyses ................................................................................ 87 4.2.3  Mineralogical analyses .............................................................................................. 88 4.2.4  Mineralogical data processing .................................................................................. 89 4.3  Results and discussion .................................................................................................. 91 4.3.1  Bulk physical and geochemical properties of the waste rock ................................... 91 4.3.2  Waste rock mineralogy ............................................................................................. 97 4.3.3  Mineral reactivity and associations ......................................................................... 100 4.3.3.1  Mineral surface passivation ............................................................................ 103 4.3.3.2  Galvanic interactions ...................................................................................... 104 4.3.4  Waste-rock leaching experiments ........................................................................... 105 4.3.5  Mineralogical controls on leachate chemistry ........................................................ 108 4.3.6  Practical implications for waste-rock management ................................................ 110 4.4  Conclusions ................................................................................................................. 112 Chapter 5: Conclusion ...............................................................................................................114 5.1  Main findings and contributions ................................................................................. 115 5.1.1  Identification of physico-chemical and mineralogical controls on drainage chemistry ……………………………………………………………………………………………..116 5.1.1.1  Hydrological paths .......................................................................................... 117 5.1.1.2  Passivation of mineral surface ........................................................................ 119 5.1.1.3  Secondary minerals and sorption .................................................................... 120 5.1.1.4  Galvanic Reactions ......................................................................................... 121 5.1.2  Quantification of mineral reactivity and internal weathering processes ................. 123 xiii  5.1.3  Application of quantitative automated mineralogy for characterization of mine wastes ................................................................................................................................. 124 5.1.4  Description of physicochemical heterogeneity at field barrel and waste-rock pile scales ................................................................................................................................. 125 5.1.5  Application of mineralogical weathering textures as qualitative indicators of hydrological responses, complementing tracer test results and drainage chemistry .......... 127 5.1.6  Simulated mineral weathering rates quantitatively supported hydrological, physicochemical, and mineralogical controls on drainage chemistry ................................ 128 5.2  Recommendations for future research directions ....................................................... 129 5.2.1  Sampling and representability ................................................................................. 129 5.2.2  Reactive transport modeling ................................................................................... 130 5.2.3  Characterization of alteration in the full-scale waste-rock pile .............................. 131 Bibliography ...............................................................................................................................132 Appendices ..................................................................................................................................154 Appendix A SUPPLEMENTARY METHODS TO CHAPTER 2 ......................................... 155 A.1  Uncertainty related to sampling method of the drainage from field barrels ........... 155 A.2  Quantitative mineralogical analyses ....................................................................... 155 A.3  MLA calibration for mineral identification ............................................................ 156 A.4  Reactive transport modelling .................................................................................. 156 A.4.1  Transient conditions ............................................................................................ 156 A.4.2   Mineral content in the waste rock in the field cells ............................................ 157 A.4.3  Rate expression and kinetic parameters .............................................................. 157 Appendix B SUPPLEMENTARY TABLES AND FIGURES TO CHAPTER 2 .................. 159 xiv  Appendix C SUPPLEMENTARY METHODS TO CHAPTER 3 ......................................... 175 C.1  Uncertainty related to sampling method of the drainage from field barrels ........... 175 C.2  Quantitative mineralogical analyzes ....................................................................... 175 C.3  Calibration of MLA for mineral identification ....................................................... 176 C.4  Temporal moments and flow parameters from tracer tests ..................................... 176 C.5  Reactive transport modelling .................................................................................. 177 C.5.1  Transient conditions ............................................................................................ 177 C.5.2   Mineral content in the waste rock in the field cells ............................................ 178 C.5.3  Rate expression and kinetic parameters .............................................................. 178 Appendix D SUPPLEMENTARY TABLES AND FIGURES TO CHAPTER 3 .................. 180 Appendix E SUPPLEMENTARY METHODS TO CHAPTER 4 ......................................... 197 E.1  Powder pH .............................................................................................................. 197 E.2  Raman spectroscopy ............................................................................................... 197 E.3  Mineral liberation analyzer ..................................................................................... 197 E.4  Handheld X-Ray fluorescence ................................................................................ 199 Appendix F SUPPLEMENTARY TABLES AND FIGURES TO CHAPTER 4 .................. 200 Appendix G SUPPLEMENTARY INFORMATION……….…………………………….....228  xv  List of Tables Table 1.1 Authors, title, and main focus of relevant previous studies at Antamina ..................... 16 Table 2.1. Initial modal compositions, simulated depletion (%), and mineral reactivity (MR %) for Fe, Cu, and Zn as pyrite, chalcopyrite and sphalerite in the studied field barrels. ................. 44 Table 3.1. Degree of mobile pore water (%) and slow matrix flow/immobile pore water in each field barrels calculated from temporal moments of tracer test observations. ............................... 67 Table 3.2. Distribution of weathering textures of minerals (boxwork/septo Vs Infilling) as well as residual and absolute accumulation of oxidation products compared to the degree of mobility of water (i.e. preferential/fast-matrix or slow-matrix flow paths). .................................................... 72 Table 3.3. Initial mineral content, weathering rates of the minerals in the simulation, calculated Damköhler number (Da), and calibrated effective rate coefficients for Fe, Cu, and Zn as pyrite, chalcopyrite and sphalerite in the studied field barrels. ................................................................ 80 Table B1. Lithology classes, acid-producing potential (AP), acid-neutralization potential (NP), neutralization potential ratio (NPR), as well as waste-rock solid-phase over leachate elemental concentration ratios for the average first and last year. .............................................................. 159 Table B2. Standard deviation and number of measurements for the average concentration of field barrel leachate from the first (I) and last years (W) of sampling. ............................................... 160 Table B3. Grouping of primary phases and oxidation products identified with MLA. .............. 161 Table B4. Modal compositions of the waste rock in the studied field barrels, as well as calculated (literature) and calibrated effective rate constants. ..................................................................... 164 Table B5. Normalized root mean square error (NRMSE) between the aqueous chemistry of the field barrel drainage and the simulation results data. ................................................................. 165 xvi  Table D1. Lithological class, neutralization potential ratio (NPR) of the bulk waste rock in field barrels. ………………………………………………………………………………………….180 Table D2. Standard deviation and number of measurements for the average concentration of field barrel leachate from the first and last years of sampling……………………………………….181 Table D3. Standard deviation and number of measurements (n) for the average mass loading of field barrel leachate from the first and last years of sampling……………………………….....182   Table D4. Grouping of primary phases and oxidation products identified with MLA...……….183 Table D5. Adjusted space-time discretization representative of the calculated flow paths in the simulation of selected field barrels………………………………………………………..……186 Table D6. Modal compositions of the waste rock in the studied field barrels, as well as calculated (literature) and calibrated effective rate constants……………………….…………………..…188 Table D7. Normalized root mean square error (NRMSE) between the aqueous chemistry of the field barrel drainage and the simulation results data………………………….…..……………189 Table D8. Parameters used to calculate the degree of mobile pore water and Damköhler number in selected field barrels………………………………………………………….….……….….190 Table F1. Grouping of primary phases and oxidation products identified with MLA………....200 Table F2. List of minerals identified with Raman Spectroscopy in selected samples from boreholes BH1D2 and BH3D2……………………………………………………………..…….…….….203 Table F3. List of samples from borehole BH1D2 and BH3D2 selected for further mineralogical characterization with MLA……………………………………………………………….…….204   xvii  List of Figures Figure 1.1 The relationship between the main and secondary factors controlling the quality of the drainage water. ................................................................................................................................ 4 Figure 1.2. Representation of the reactive surface of minerals in relation to each other ................ 8 Figure 1.3. Galvanic reaction model between two minerals ......................................................... 10 Figure 1.4. Site location of the Antamina mine in Peru. ............................................................... 12 Figure 1.5. Location of boreholes BH1D2 and BH3D2 ............................................................... 13 Figure 2.1. Evolution of drainage quality from four waste-rock field barrels at Antamina ......... 31 Figure 2.2. Distribution of the elemental concentration ratios ..................................................... 33 Figure 2.3. Overview of the distribution of Fe-, Cu-, and Zn-bearing primary from the unweathered reference waste rock samples .................................................................................. 39 Figure 2.4 Overview of the distribution of Fe-, Cu-, and Zn-bearing primary in the weathered waste rock ..................................................................................................................................... 40 Figure 2.5. Mineral reactivity and pyrite-association indices (bottom) of the initial non-weathered and weathered waste rock mineralogy ........................................................................ 41 Figure 2.6. Comparisons of the temporal evolution of the measured drainage chemistry from field barrels 1A, 1B, 2A, and 2B with the simulated drainage chemistry .................................... 44 Figure 3.1. Schematic infilling and boxwork textures of secondary minerals precipitated in a primary mineral ............................................................................................................................. 55 Figure 3.2. Relationship between water regime and the weathering patterns of minerals  .......... 56 Figure 3.3. Relationship between flow paths, mineral dissolution, solute concentration, and mineral depletion .......................................................................................................................... 61 Figure 3.4. Normalized chloride breakthrough curves from tracer tests ...................................... 65 xviii  Figure 3.5. Proportion of secondary minerals associated with relative and absolute accumulation....................................................................................................................................................... 69 Figure 3.6. Point counting modal distribution of boxwork and infilling textures  ....................... 69 Figure 3.7. The solid-phase content; the average drainage concentrations, and average mass loading........................................................................................................................................... 76 Figure 3.8. Comparison of measured drainage chemistry with the simulated drainage chemistry....................................................................................................................................................... 78 Figure 3.9. Relationship of Damköhler number with the degree of mobile porewater, septo weathering textures, and secondary minerals ............................................................................... 79 Figure 4.1. Qualitative observations in cores retrieved from boreholes BH1D2 and BH3D2 ..... 92 Figure 4.2. Solid-phase concentrations of total S, Fe, Cu and Zn as well as neutralization potential ratios (NPR) of the waste rock in boreholes BH1D2 and BH3D2 ................................ 94 Figure 4.3. Conceptual schematic of the waste-rock pile and the locations of reactive zones ..... 97 Figure 4.4. Mineralogical parameters of a selection of samples along the depth-profiles of boreholes BH1D2 and BH3D2 ................................................................................................... 102 Figure 4.5. Ratios of solid-phase Cu and Zn waste-rock concentrations over respective aqueous leachate concentrations from samples along boreholes BH1D2 and BH3D2 ............................ 107 Figure 5.1 Relationship between physico-chemical processes and the drainage quality ........... 118 Figure B1. Waste rock sampling locations at the top and the bottom of the field barrels. ......... 166 Figure B2. Modal normalized mineralogical distribution amongst all phases in field barrel samples. ....................................................................................................................................... 167 Figure B3. Linear regression of the elemental content of waste rock as measured with FP-XRF and MLA. .................................................................................................................................... 168 xix  Figure B4. Solid-phase Cu, Fe and Zn content (wt-%) in field barrel samples. ......................... 169 Figure B5. Particle size distribution of the bulk waste rock in field barrels. .............................. 170 Figure B6. Comparisons of the temporal evolution of the measured sulfate drainage chemistry with the simulated drainage chemistry ....................................................................................... 171 Figure B7. Sulfide mineral reactivity indices of field barrel samples. ....................................... 172 Figure B8. Backscattered electron images and elemental maps showing alteration rims .......... 172 Figure B9. Backscattered electron images and elemental maps showing the presence of sulfide couples ........................................................................................................................................ 174 Figure D1. Modal normalized mineralogical distribution and distribution of secondary phases of field barrel samples. .................................................................................................................... 191 Figure D2. Particle size distribution of marble (1A), intrusive (2A, 2B and 2C), and skarn (3A) field barrel samples. .................................................................................................................... 192 Figure D3. Waste rock sampling locations at the top and the bottom of the field barrels .......... 193 Figure D4. Linear regressions of waste rock element content measured with XRF and MLA .. 194 Figure D5. Pictures from microscope or SEM from samples 1A, 2A, 2B, and 3A. ................... 195 Figure D6. Normalized distribution of absolute and relative minerals in weathered samples of all field barrels. ................................................................................................................................ 196 Figure F1. Linear regression of the elemental content of waste rock as measured with XRF or MLA ............................................................................................................................................ 206 Figure F2. Gravimetric moisture content of the waste rock in boreholes BH1D2 and BH3D2.. 207 Figure F3. Total Mo and As Waste rock and leachate concentration of samples along boreholes BH1D2 and BH3D2. ................................................................................................................... 208 xx  Figure F4. Rinse-pH tests and qualitative powder-pH tests of the waste rock samples from boreholes BH1D2 and BH3D2 ................................................................................................... 209 Figure F5. Average oxygen content (%) and temperature (oC) over 4 years in boreholes BH1D2 and BH3D2 ................................................................................................................................. 210 Figure F6. Modal and secondary mineral composition of the waste-rock in boreholes BH1D2 and BH3D2 ........................................................................................................................................ 211  Figure F7. Proportion of primary versus secondary minerals and elemental distributions of Cu, Fe and Zn in borehole BH1D2 .................................................................................................... 212 Figure F8. Proportion of primary versus secondary minerals and elemental distributions of Cu, Fe and Zn in borehole BH3D2 .................................................................................................... 213 Figure F9. Modal distribution of liberated As- and Mo-bearing minerals in selected samples of boreholes BH1D2 and BH3D2. .................................................................................................. 214 Figure F10. Composite photographs of the sonic drill cores from borehole BH1D2 ................. 215  Figure F11. Composite photographs of the sonic drill cores from borehole BH3D2 ................ 216 Figure F12. Pictures of agglomerates from the core samples of boreholes BH1D2 ................... 217  Figure F13.  Composite digital images obtained by MLA from microscopic agglomerates observed in samples from borehole BH1D2 ............................................................................... 218 Figure F14. Relationship between the pyrite reactivity index versus the pyrite association index from selected samples of boreholes BH1D2 and BH3D2 .......................................................... 219 Figure F15. Modal mineralogical composition of the liberated sulfides of selected samples from boreholes BH1D2 and BH3D2. .................................................................................................. 220 Figure F16. Composite MLA images of passivated sulfides ...................................................... 221 xxi  Figure F17. Relationship between pyrite reactivity versus total Fe-, Cu-, or Zn- bearing secondary minerals content ......................................................................................................... 222 Figure F18. Concentration ratios of waste-rock solid-phase As and Mo content over the respective aqueous leachate concentrations along boreholes BH1D2 and BH3D2  ................... 223 Figure F19.  MLA composite images of sulfide galvanic couples ............................................. 224 Figure F20. Modal distribution of Zn-bearing phases in selected samples of boreholes BH1D2 and BH3D2. ................................................................................................................................ 225 Figure F21. Composite MLA images of liberated As-bearing particles ..................................... 226 Figure F22. MLA composite images of liberated Mo-bearing minerals .................................... 227  xxii  List of Abbreviations ABA    Acid Base Accounting  AI    Association Index  AP    Acid potential  BSE   Back-scattered electron  Bt    Billion tons C   Concentration  Co    Initial concentration  Da    Damköhler number  DI water  Deionized water EDS    Energy dispersive X-ray analysis system FP-XRF   Field portable-X-ray fluorescence  GCI    Galvanic coupling index  HCl   Hydrochloride HNO3   Nitric Acid ICP-OES   Inductively coupled plasma-optical emission spectrometry  ICP-MS   Inductively coupled plasma mass spectrometry  keff    Effective rate coefficients LiCl    Lithium/Chloride HDPE    High-density polyethylene M0    Initial Content  2MAD   Double median absolute deviation  MLA    Mineral liberation analyzer  xxiii  MR    Mineral Reactivity  MS    Metal Sulfide NNP    Net neutralization potential  NPR    Neutralizing potential ratio  NRMSE  Normalized root mean square error  PAG    Potentially net acid generating  S    Total sulfur  SEM    Scanning electron microscope  QEMSCAN  Quantitative evaluation of minerals by scanning v-%    Volumetric percent wt%   Weight percent XRD    Dispersive X-Ray spectroscopy    xxiv  Acknowledgements  The explorer Edmund Percival Hillary said: “It is not the mountain you conquer but yourself”. The different stages of production of this project could be compared to an expedition to the top of a mountain. Along my academic journey, I acquired some valuable knowledge but also important life lessons. I am grateful for the support of the people that surrounded me each step of the way towards my PhD. I would like to acknowledge my supervisors who trusted me, gave me the opportunity to choose my own path, and shared their time and knowledge. Many thanks to the team of graduate students including Sharon Blackmore, Holly Peterson, Maria Eliana Lorca, Mehrnoush Javadi, Laura Laurenzi, and Elliott Skierskan as well as the Post Doc Bas Vriens and Daniele Pedretti who were always available to exchange ideas and provide feedback and support when needed. I would also like to give a special mention to: 1) Melu for your moral support, translation services, and your wonderful resourcefulness during the field work at Antamina; 2) Merhnoush for your kindness and intensive brainstorming sessions at the Bean; 3) Bas for your mentorship and encouragement to write journal papers.  This project would not have been possible without the technical or administrative support and hard work of the professionals and technicians from the University of British Columbia (UBC) and industrial partners. Many thanks to Maria Liu, Leslie Nichols, Jenny Lai, Pius Lo, Aaron Hope, and Libin Tong at UBC; Celedonio Aranda, Bevin Harisson, and Edsael Sanchez from the Antamina mine; as well as Stephane Brienne, Randy Blaskovich, Anelda van Staden, and Beth Bromley at Teck. This thesis also relied upon the financial and material contributions from UBC, Natural Sciences and Engineering Research Council of Canada (NSERC), Antamina mine, and Teck.  xxv  Finally, many thanks to my family, dear friends, as well as Gary and Tao for being by my side during the many ups and downs of this journey, for being present, making me laugh, and keeping me grounded. Just like companions on an expedition to the summit of a mountain, the people who engaged with me on this journey helped me “conquer myself” and enriched my experience while helping me safely reach the summit.          xxvi  Dedication          Pour ma mère "Je me souviens"  1  Chapter 1: Introduction The exploitation of precious and base metals requires the extraction of non-economical materials known as waste rocks. The minerals from the newly exposed waste-rock surfaces react with water and oxygen to liberate metals and alkalinity or acidity in water. These phenomena are better known in the literature as acid, neutral, or alkaline mine drainage and metal leaching (Jambor et al., 2003; Morin and Hutt, 1997; Price, 2009). Water quality monitoring and predictions are an important part of the planning and management of mining operations, in order to maintain the water-quality standards for the long-term management of mines as well as established governmental regulations (Lottermoser, 2010; Warhurst, 1999; Younger, et al., 2002).    The predictions of drainage chemistry deposits are therefore an important step in mine planning but also a challenging one (Amos et al., 2014; Price, 2009). The site conditions, sampling, sample preparation, analyses, test procedures and the interpretation of data are only some of the elements to consider in order to achieve accurate predictions (Price 2009). Dealing with the material heterogeneity and constant evolution of mine plans are challenges for prediction (Price 2009). For example, mixed neutral and acidic-generating waste rock produces neutral and acidic discharges at distinct locations (Morin and Hutt, 2000, 1997). Moreover, different Mo mines and properties in British Columbia, e.g., Boss Mountain, Brenda, Endako, Highland Valley Copper, Kitsault, and Trout Lake, are demonstrating exceptions to the acid base accounting prediction rules, with unexplained causes (Morin et al., 2001).   The waste-rock characterization plays a crucial role in predicting and controlling the quality of drainage (Amos et al. 2014). The mineralogical and physical characteristics of the waste rock have an impact on the hydrological, geochemical, and physical processes. The extent and relationship between these mechanisms will determine the quality of the drainage based on the 2  mineralogical oxidation rate. Static and kinetic tests, soluble and solid phase elemental analyses, particle size distribution, mineralogical properties, and numerical modelling are examples of tools used to characterize the waste rock for drainage predictions (Price 2009; Morin and Hutt 1997). The numerical models using compilations of thermal, hydrological, and geochemical parameters are a means of quantifying the physical and geochemical mechanisms. Once quantified, they assess the accuracy of the conceptual models and provide predictions for the management of similar waste rock (Amos et al. 2014). The characterization of a full-scale waste-rock pile is challenging because of the heterogeneity of material resulting from end dumping depositions. Many factors can influence the geochemistry of the water from the speed water flows through the waste rock, the grain size distribution, to the composition and association of the minerals composing the pile. All of these can impact the dissolution or precipitation reaction rates of the minerals thereby impacting the geochemistry of the water. The current scope of this project will target processes that potentially have an effect on the solute concentration of waste-rock drainage in small and larger scale systems. The main processes under investigation are the flow residence times, mineral passivation, galvanic reactions, and the adsorption or precipitation of secondary minerals. This study aims to improve understanding of the relationships between mineral reactivity, geochemical and hydrological processes resulting in the attenuation or release of metals, and waste-rock drainage chemistry in order to achieve better predictions.  1.1 Controls on drainage chemistry Amos et al. (2014) reviewed the key elements controlling the quality of drainage water which included environmental factors, physical and mineralogical waste-rock characteristics, physical and geochemical processes, and rates of mineralogical oxidation. The main conclusions from this 3  study highlight the importance of: i) on site characterization of waste rock; ii) a better understanding of mechanisms controlling the thermal, hydrological and geochemical processes; iii) a greater understanding of the scale effect for prediction and prevention tools; iv) the quantitative descriptions of processes at different scales; and, v) the use of numerical models. Figure 1.1 illustrates the relationship between these factors and processes controlling the quality of the drainage water. This study will target the waste-rock mineralogical characteristics (i.e. sulfide and neutralizing mineral content and secondary minerals) that might affect geochemical processes such as the neutralization and acidity potentials, mobility of metals, or attenuation reactions. These geochemical processes in combination with the water regime will ultimately influence the mineral reaction rates and drainage water quality.  1.1.1 Acid rock drainage, neutral rock drainage and metal leaching  A considerable number of reviews explain the geochemical and mineralogical mechanisms triggering acid mine drainage (Amos et al., 2014; Banks et al., 1997; Blowes et al., 2003; Moncur et al., 2009; Nordstrom and Alpers, 1999). Studies on geochemical and mineralogical mechanisms linked with neutral mine drainage and metal release are less abundant in the literature (Amos et al., 2014; Banks et al., 2002; Kirby and Cravotta, 2005; MEND and Stantec, 2004; Pettit, 1999). Three main reactions control the presence of acidity, alkalinity, and metals released in mine waters: acid producing sulfide oxidation (1.1), non-acid producing sulfide oxidation (Eq.1.2) and buffering reactions (Eq.1.3). Following a series of oxidation reactions, Fe and elemental sulfides (i.e. pyrite, chalcopyrite, pyrrhotite, Fe-sphalerite) release protons or acidity, metals, metalloids and sulfates 4  (Banks et al., 1997; Lottermoser, 2010). For example, pyrite can react with oxygen or ferric Fe and produce H+ and SO42- (Eq. 1.1 is the net product of the oxidation reactions). 4FeS2 + 14 H2O +15 O2 → 4Fe(OH)3 + 8SO42- + 16 H+ (1.1) Figure 1.1 The relationship between the main and secondary factors controlling the quality of the drainage water. The key processes targeted in this study are circled. Metal release can also occur from the oxidation of elemental sulfides (i.e. non-Fe sphalerite, galena, molybdenite, etc.) which does not directly produce acidity but releases heavy metals, metalloids, and sulfates (Banks et al., 1997; Lottermoser, 2010; Plumlee et al., 1999). For example, sphalerite reaction with oxygen produces Zn2+ and SO42- (equation Eq.1.2). 5   ZnS + 2O2 → Zn2+ + SO42- (1.2)  Buffering reactions of acidity can occur when alkalinity is present in the system as mineral sources (i.e. carbonates, alumino-silicates, hydroxides, clays). For example, calcite dissolution releases Ca2+, carbonate, and bicarbonate (equation Eq. 1.3). Alumino-silicates such as plagioclase-felsdpar or anorthite release Ca2+ and consumes H+ (equation Eq. 1.4).        CaCO3 + 2H+ → H2CO3 (pH < 6.3) or CaCO3 + H+ → HCO-3 (pH > 6.3) (1.3)  CaAl2Si2O8 + 2H+ + H2O → Ca2+ + Al2Si2O5(OH)4 (1.4) Acid Base Accounting (ABA) is a combination of tests and calculations to establish the potential for acid drainage production. The ABA procedures determine the rinse and paste pH, the acid potential (AP) and the neutralization potential (NP) from which the net neutralization potential (NNP = NP - AP) and NP/AP ratio are determined (Price 2009). The production of acid or neutral mine drainage is estimated using the NP/AP ratio and the reaction rate of the minerals (equations 1.1 to 1.4). A sample is considered potentially net acid generating (PAG) when NP/AP < 1; not potentially net acid generating (non-PAG) if NP/AP > 2, and uncertain if NP/AP is between 1 and 2 (Price, 2009). Theoretically, acid mine drainage occurs when NP < AP or when the reaction rate of consumption of buffering minerals is inferior to the acid-producing minerals’ oxidation rate. On the other hand, neutral mine drainage occurs when NP > AP or when the reaction rate of consumption of buffering minerals is greater than the acid-producing minerals’ oxidation rate. These results serve as guidelines, and can be combined with mineralogy, elemental analysis, and kinetic testing for improved accuracy (Price, 2009).  The term acid rock drainage describes acidic pH seepage loaded with high levels of sulphate and other contaminants (i.e. Fe, As, Cd, Zn, Pb, Cu, Al, Mn). Those dissolved species 6  released by the oxidation reactions of sulfides are mobile at low pH. These products may react with in situ minerals, dissolve as complexes or free anions, precipitate as solids, form secondary minerals and coatings, or adsorb at the surface of other minerals. The term neutral drainage describes neutral pH seepage with high levels of sulphate and other contaminants (i.e. As, Ni, Mn, Mo, Zn) which are mobile at neutral pH. The seepage is neutral because of non-acid producing oxidation reactions and/or buffering reactions of acid producing oxidation reactions. The product of these reactions may dissolve as free anions or complexes, precipitate as secondary minerals (i.e. sulphate, oxide, hydroxide or carbonate), adsorb on Fe-Mn-Al oxyhydroxides and/or co-precipitate with solid phases. 1.1.2 Microbial activity Abiotic oxidation rates of sulfide minerals are orders magnitude lower than biotic oxidation rates, the Fe- and S- oxidizing bacteria play an important role in catalyzing the oxidative dissolution of minerals (Singer and Stumm, 1970; Baker and Banfield, 2003; Johnson and Hallberg, 2003). The acidophilic bacteria catalyses the sulfide oxidation reactions by oxidation of ferrous iron (Fe2+), reduction of inorganic sulfur (S2-), and regeneration of ferric iron (Fe3+) (Johnson and Hallberg, 2003). Fe- and S- oxidizing bacteria are more predominant in acidic environment but can also be observed in neutral and alkaline environments (Nordstrom et al., 2015; Baker and Banfield, 2003; Dockrey et al., 2014). In unsaturated heterogeneous waste-rock environments, microenvironments may protect microbes and sustain biodiversity (Mesa et al., 2017; Blowes et al., 1995; Dockrey et al., 2014). For example, acidophilic bacteria were observed and protected in acidic microenvironments from carbonate acid-neutralizing waste rock rich in neutrophilic bacteria (Dockrey et al., 2014). The circum-neutral microbial abundance is less studied than the acidic mine waste environments (Lottermoser, 2010). The heterogeneity of 7  unsaturated waste rock piles have the potential to foster diverse microbial communities but remain less studied to date (Blackmore, 2018b; Bailey et al., 2016; Dockrey et al., 2014; Smith et al., 2013; Schippers et al., 2010; Sand et al., 2007; Schippers et al., 1995). Acid generating waste rock from the Antamina mine had a higher ratio of acidophililic:neutrophilic S-oxidizing bacteria, whereas non-acid generating waste rock had a higher neutrophilic:acidophililic ratio of bacteria (Blackmore et al., 2018b). Blackmore et al., 2018 studied the drainage chemistry of 2 kinetic columns composed of 2/3 acid-generating and 1/3 non-acid generating material and compared the original untreated waste rock in one column with treated microbial-depleted waste rock in the other. The column with untreated waste rock was more reactive and showed 2 units lower pH (4-5), higher sulfide oxidation rates as well as 16 times higher sulfate and heavy metals release (Blackmore, 2018b). 1.1.3 Mineral reactivity  Mineralogical factors such as mineral composition, abundance, texture, arrangement, and exposed surface area can also influence the overall reactivity of waste rock and the resulting weathering as well as drainage composition (Petruk, 2000; Brough et al., 2013; Dold, 2017; Jamieson et al., 2015; Lapakko, 2015; Parbhakar-fox and Lottermoser, 2015; Pedretti et al., 2017). In this study, the term weathering is defining the processes responsible for the partial or complete transformation of minerals characterized by changes in color, texture, hardness, or shape (Delvigne, 1998). Whereas, the term alteration is interpreted as the result of the weathering process and designate replacement of primary minerals by secondary products. The exposed mineral surface area, influenced by grain size distribution and texture (i.e. availability and association), has an effect on the reactivity (Petruk, 2000; Jamieson et al., 2015; Lapakko, 2002, 2015). A single mineral with a fully exposed surface results in an increased surface area and higher reactivity, 8  whereas, two or more minerals in contact with each other may decrease their surface area and availability to react (Lappako, 2002, 2015). Thus, a mineral is considered totally liberated from the matrix when not in contact with another one, partially liberated when touching parts of another mineral or occluded when totally surrounded by another mineral (Figure 1.2, adapted from Blaskovich, 2013). Consequently, the rate of the oxidation or dissolution depends on the availability or liberation of the reactive surface of those minerals (Lapakko 2002, 2015; Jamieson et al., 2015).         Figure 1.2. Representation of the reactive surface of minerals in relation to each other, i.e. fully and partially liberated or occluded minerals. Adapted from Blaskovich 2013.   1.1.3.1 Mineral coating or passivation  Passivation or coating formation occurs when secondary minerals precipitate at the surface of other minerals resulting in decreased reactivity (Lapakko, 2015). The precipitation of secondary minerals accumulates at the surface of the minerals and forms coatings (i.e. passivation or armoring) that decrease their reactivity (Huminicki and Rimstidt, 2009, 2008; Lapakko, 2015) and impact the rate of oxidation of sulfide minerals and the dissolution of carbonate minerals (Lapakko 2002, 2015; Jamieson et al., 2015). 9  The consequences of the presence of coating at the surface of minerals has been intensively studied to assess the performance of the passive treatment of mine drainage with limestone drains (e.g. Cravotta III and Trahan, 1999; Hammarstrom et al., 2003; Ziemkiewicz et al., 1997). The main conclusions show that the reactivity of limestone drains depends on grain size, carbonate composition, flow rate, and secondary mineral armouring (Genty et al., 2012; Huminicki and Rimstidt, 2008). More specifically, the nature, thickness, and porosity of the coating influences the reactivity of the minerals (Hammarstrom et al., 2003; Huminicki and Rimstidt, 2008; Santomartino and Webb, 2007).  At the Antamina mine, mineralogical characterization of the main lithologies has shown that the neutralization potentials of intrusive and skarn material could be potentially decreased by coatings forming at the surface of carbonate minerals (Peterson, 2014). Skarn are metamorphic rocks rich in calcium-magnesium-iron-manganese-aluminium silicate minerals, whereas intrusive rocks refers to quartz-monzonite porphyry (Escalante et al., 2010; Lipten and Smith, 2004). They are also the most susceptible to produce acidic drainage in the long term. More precisely, the combination of waste-rock lithology in full-scale piles could enhance the precipitation of secondary minerals (Hirsche, 2012), promote the coating of carbonates, and then lower the neutralization potential of intrusive and skarn layers. In contrast, the passivation of sulfides might inhibit their reactivity (Huminicki and Rimstidt, 2009) and decrease the metal release and acidity production of sulfide-rich lithology composed of intrusive and skarn.   1.1.3.2 Galvanic reactions  Galvanic corrosion is an electrochemical process whereby one metal is preferentially corroded in the presence of an electrical contact with another metal and electrolyte. Likewise, a galvanic reaction is a spontaneous oxidation-reduction process occurring when minerals rich in metals (e.g. 10  sulfides or oxides) are in contact with each other in a solution that permits the transfer of charges (Liu et al., 2008; Perkins et al., 1995). As a result, one of the minerals will dissolves preferentially compared to the other (Kwong, 1995; Liu et al., 2009, 2008). The sulfide oxidation reactions are preferentially influenced by galvanic interactions because they require a transfer of electrons. When in contact with each other and in the presence of an electrolyte, two sulfides with different rest potentials will act as a galvanic cell. The rest potential is defined when the difference between the anodic and cathodic current is null at the interface of a mineral and solution. The sulfide mineral with high rest potential will act as the cathode (i.e. reduction reaction) and the sulfide with low rest potential will act as the anode (i.e. oxidation reaction). The Figure 1.3 illustrates the exchange of charges between the two sulfides. Consequently, the rate of oxidative dissolution of the mineral with the lowest rest potential will be increased and the one with the highest rest potential will be inhibited. As a result, the galvanic reactions control the oxidation sequence of sulfide minerals and metal release (Kwong et al., 2003).      Figure 1.3. Galvanic reaction model between two minerals (MS) in contact with each other: MS acts as the anode (i.e. lower rest potential and dissolution reaction) and MS2 as the cathode (i.e. higher rest potential and reduction reaction) when in the presence of an electrolyte (adapted from Liu, 2008). Galvanic reactions can occur at different scales and cause preferential mobilization of metals from kinetic tests, heap leach, or drainage of waste-rock piles (Chopard et al., 2017; Dixon et al., 2008; Kwong, 1995; Kwong et al., 2003; Parbhakar-Fox et al., 2013; Perkins, et al., 1995; 11  Qian et al., 2018). Galvanic reactions between pyrite and chalcopyrite are applied in mining processes to increase recuperation of Cu during the heap leaching of primary copper concentrate. The passivation of chalcopyrite surface is prevented and Cu is preferentially leached when galvanically in contact with added pyrite without additional grinding or introduction of bacteria or other chemicals (Dickson et al., 2008). The impact of galvanic reactions on the drainage chemistry has been previously investigated at different mines such as Red Dog, Alaska (Day, et al., 2003) and Keno Hill, Yukon (Kwong et al., 1997). For instance, the galvanic interactions between sphalerite, pyrite and galena caused the release of Zn under non-acidic conditions. In addition, the cathodic protection of pyrite delayed the production of acid mine drainage. The scarceness of available data on galvanic reactions occurring in waste-rock piles prevents from evaluating the importance of this mechanism on the rate of sulfide oxidation (Perkins et al. 1995). Kwong (2003) demonstrated that galvanic sulfide oxidation and metal release occur under oxygenated water and near neutral pH conditions in tailing ponds. A mineralogical evaluation previously observed the association of minerals suggesting the occurrence of galvanic reactions in waste-rock samples at Antamina (Blaskovich, 2013; Haupt et al., 2011). 1.2 Research site The Antamina mine is a polymetallic Cu, Zn, Mo skarn deposit located in Peru at an elevation between 4000 m and 5000 m above sea level (Figure 1.4). The typical annual precipitation is between 1200 and 1500 mm with average temperatures of 5.4 to 8.5 ºC (Harrison et al., 2012). In 2012, the mine produced approximately 340 000 metric tons per day of waste rock, and is projected to produce 2.2 Bt of waste rock by the end of life of the mine in 2029 (Harrison et al., 2012). The drainage from the waste rock has a neutral to alkaline pH, with some localized acidic pH seeps, and is driven by wet and dry season cycles.  12  There are five major waste-rock types identified at the mine: limestone, marble, hornfels, skarn (endoskarn and exoskarn) and intrusive (Harrison et al., 2012). The different waste types are segregated and dispatched using blast-hole assays and polygon delineation. According to the metal zoning of the Antamina deposit, endoskarn and intrusive are most likely associated with Cu and Mo. Comparatively, Zn, Cu, As, Bi and Pb are associated with exoskarn and Zn, Pb, Ag and Bi are related to marbles and hornfels (Lipten and Smith, 2004). The classification system of waste rock at Antamina is based on the geochemical review of the waste rock including the interpretation of ABA tests (Golder 2004). The classification scheme includes the solid phase content of Zn (weight %), As (weight %), and sulfides (visual estimate %) (Harrison et al., 2012).    Figure 1.4. Site location of the Antamina mine in Peru. In 2005, five experimental instrumented waste-rock piles of 36 m x 36 m x 10 m were built, at the mine site, to study the long-term geochemical behaviour of rock types, providing information for prediction purposes (Harrison et al., 2012). Between 2002 and 2006, experimental field kinetic tests (i.e. field barrels) were built using waste-rock material of specific lithologies corresponding to material representatives of the experimental and full-scale waste-rock piles (Peterson, 2014, 13  Blackmore, 2015). In 2012 and 2014, boreholes were drilled in the full-scale waste-rock dump, samples were collected and temperature and gas monitoring instruments were installed (Vriens, et al., 2018). The locations of 2 boreholes BH1D2 and BH3D2 of 140 m and 126 m drilled in the East dump at Antamina in 2014 are shown in Figure 1.5. The field barrels consist of approximately 300 kg of waste rock placed in a 55 gallon plastic drum. More details on their installation and dimensions are available in Aranda, 2010. Selected waste rock placed in the field barrels was characterized to determine: particle size distribution, surface area, as well as chemical and mineralogical compositions (Aranda, 2010; Golder, 2010, 2004). The field barrels are subject to the same weather and atmospheric conditions as the full-scale waste-rock piles unlike other small scale laboratory testing such as columns and humidity tests. They are therefore useful for validating leaching release rates obtained from laboratory tests (Aranda et al., 2009). The drainage from the field barrels is collected monthly in a 20 L container, analyzed for total and dissolved metals, and the average flow rate is calculated.  Figure 1.5. Location of boreholes BH1D2 and BH3D2 in the east waste-rock pile at Antamina. Peterson (2014) studied the metal release from three experimental waste-rock piles at Antamina. The differences in outflow aqueous geochemistry from the three experimental piles BH3D2BH1D214  were directly linked to the elemental composition and mineralogy of the three main lithologies: marbles/hornfels, intrusives, and skarns. Also depending on the composition and flow there were different dominant seasonal controls in the piles. The hydrological controls observed were either matrix flow with long residence times allowing for the dissolution of ions, or preferential flow paths with shorter residence times providing diluted water. The geochemical controls causing seasonal fluctuations in pH as well as attenuation and release of dissolved solutes were: CO2 degassing, mineral dissolution, secondary mineral precipitation, and ionic sorption/desorption. In addition, Blackmore (2015) studied the relationship between hydrology and geochemistry in experimental piles using in situ chloride and applied bromide tracer tests. The results suggest that preferential flow is causing a faster depletion rate with lower solute concentrations due to lower weathering and higher flushing rates. In contrast, matrix flow is triggering slower depletion rates but with higher solute concentrations caused by higher weathering and slower flushing rates. In summary, the geochemistry at Antamina is the result of complex interactions driven by heterogeneous waste-rock composition, particle size distribution, and hydrological regimes creating conditions suitable for the weathering of waste rocks as well as the attenuation and release of dissolved species such as H+, sulfate, carbonate, As, Cu, Fe, Zn, and Mo.  1.3 Thesis scope and organization 1.3.1 Objectives The classification system of waste rock at Antamina relies on the bulk geochemistry of waste rocks, regardless of the lithology or mineralogy (Harisson, 2012). Thus, the prediction of metal release is currently based on the initial solid phase content with a partial understanding of a few observed processes (e.g. hydrology, sorption, secondary mineral precipitation) occurring in the waste-rock pile. Price (2009) mentioned that discrepancies could occur between the solid phase 15  elemental concentrations and the aqueous concentrations. For instance, in some cases the weathering conditions and the mineral composition (e.g. insoluble or unalterable) or texture (e.g. mineral occluded in another phase) can be unfavorable to elemental mobility resulting in lower aqueous concentrations compared to the solid elemental concentrations. In contrast, higher solute concentrations in drainage compared to the solid phase may occur because of changes of solubility during the weathering of phases or preferential dissolution of mineral due to galvanic reactions. The attenuation and release mechanisms commonly studied are the solubility controls of minerals (Al et al., 1997, 2000; Sloot and Zomeren, 2012; Sracek et al., 2004), adsorption and desorption (e.g. Blowes et al., 2003; Langmuir, 1997; Smith, 2007), sulfide oxidation processes (Evangelou, 1998; Evangelou and Zhang, 1995), and passivation of sulfides (Huminicki and Rimstidt, 2009). Galvanic reactions are not commonly studied in the context of waste-rock piles, but they are hypothesized to have an influence on the solute concentrations (Kwong et al 2003; Peterson 2014; Liu et al 2008).  The current and previous focus of the Antamina research project are hydrology, heat and gas transfer, solubility control processes, complexation, and microbial activity (Blackmore, 2015; Conlan, 2009; Hirsche, 2012; Laurenzi, 2016; Lorca et al., 2015; Peterson, 2014). The current study will complete previous studies performed at Antamina (Table 1.1) and identify less studied mechanisms (i.e. passivation of mineral surfaces and galvanic reactions) that might also have an effect on solute concentration from mine drainage. The purpose is to explore the occurrence and impact on the drainage chemistry of: 1) alteration mechanisms such as sorption or precipitation of secondary minerals (Peterson, 2014; Laurenzi, 2016; Hirsche, 2012); and 2) less studied processes like armoring of minerals and galvanic reactions. The occurrence of these processes and impact 16  on mineral reactivity and drainage compositions were investigated through mineralogical observations at the field barrels (Chapter 2) and through waste-rock pile scales (Chapter 4).  The mineralogical parameters of heterogeneous waste-rock lithology from field studies are rarely quantified (Parbhakar-Fox et al., 2013, 2014; Pedretti et al., 2017). However, they are starting to be incorporated into weathering and drainage chemistry prediction tools (Parbhakar-Fox et al., 2018b) along with other properties such as acid-base accounting (Dold, 2017; Parbhakar-fox and Lottermoser, 2015), particle size distribution (Malmstrom et al., 2000) or surface areas (Jurjovec et al., 2004). The chemical composition, texture, arrangement, and association of minerals characterize their reactivity (Brough et al., 2017; Linklater et al., 2005; Parbhakar-Fox et al., 2018a; Pedretti et al., 2017). Those mineralogical controls subsequently influence the waste-rock weathering rates and mine water quality predictions. The objective is therefore to complement the current knowledge of processes with mineral reactivity, in order to improve the prediction of metal release. To this end, the bulk mineralogical parameters of liberation and association were investigated, as indicators of reactivity over 2 scales: field barrels (Chapter 2) and waste-rock piles (Chapter 4). Table 1.1 Authors, title, and main focus of relevant previous studies at Antamina Author Year            Title of the study Main Focus Corazao Gallegos, Juan C. In Master Thesis 2007 The design, construction, instrumentation and initial response of a filed-scale waste rock test pile. Initial hydrological and geochemical response after construction of waste rock piles. Conlan, M.J.W.  In Master Thesis 2009 Attenuation mechanisms for molybdenum in neutral rock drainage.  Investigation of the fate of Mo in neutral rock drainage with batch and column experiments. Dockrey, John W. In Master Thesis 2010 Microbiology and geochemistry of neutral ph waste rock from the Antamina mine, Peru. Evaluation of microbial populations in waste rock samples and field cells at neutral drainage pH. 17  Author Year            Title of the study Main Focus Aranda, Celedonio In Master Thesis 2010 Assessment of waste rock weathering characteristics at the Antamina Mine based on field cells experiment.  Evaluation of weathering behavior of waste rock from field barrel using the mineral availability of specific size fractions with MLA. Hirsche, Trevor In Master Thesis 2012 A field cell and Humidity cell study of metal attenuation in neutral rock drainage from the Antamina Mine, Peru.  Study of Zn and Mo attenuation by waste-rock mixing using mixed-material in field barrels and humidity cells. Blaskovich, Randy In Master Thesis 2013 Characterizing waste rock using automated quantitative electron microscopy. Testing optimal parameters for the analysis of secondary minerals in waste rock with MLA. Peterson, Holly .E In Doctoral dissertation 2014 Unsaturated hydrology, evaporation, and geochemistry of neutral and Acid rock drainage In highly heterogeneous mine waste rock at the Antamina mine, Peru.  Hydrological and geochemical processes controlling the water quality in waste rock experimental piles. Blackmore, Sharon  In Doctoral dissertation 2015 The role of hydrology, geochemistry and microbiology in flow and solute transport through highly heterogeneous, unsaturated waste rock at various test scales.  Understanding of dominant drainage controls from preferential and matrix flow, geochemical composition, and microbial activity. Laurenzi, Laura In Master Thesis 2016 Investigating metal attenuation processes in mixed sulfide carbonate bearing waste rock. Identification of metal attenuation processes in a heterogeneous waste rock dump. Lorca, Maria Eliana et al. In Vadose zone journal 2016 Spatial and temporal fluctuations of pore-gas composition in sulfidic mine waste rock.  Pore gas composition and temperature in experimental waste rock piles. Pedretti, Daniele et al. In Journal of Contaminant Hydrology 2017 Stochastic multicomponent reactive transport analysis of low quality drainage release from waste rock piles: Controls of the spatial distribution of acid generating and neutralizing minerals.  Monte Carlo simulation of the contribution of mineralogical heterogeneity and preferential flow to the drainage composition of a heterogeneous pile. Skierszkan, Elliott K. in Doctoral dissertation 2018 Application of molybdenum (and zinc) stable isotopes to trace geochemical attenuation in mine waste. Application of Mo and Zn stable-isotope analyses to trace geochemical attenuation processes and metal transport in mine waste. Vriens, Bas et al. In Vadose zone journal 2018 Localized sulfide oxidation limited by oxygen supply in a full-scale waste-rock pile.  The reactive and transport controls limiting the sulfide oxidation within two instrumented boreholes in an operational waste rock pile. 18  Author Year            Title of the study Main Focus Vriens, Bas et al. In Applied geochemistry 2019 Poregas distributions in waste-rock piles affected by climate seasonality and physicochemical heterogeneity. Investigation of pore gas spatiotemporal variations in experimental piles of different waste rock composition. Vriens, Bas et al. In ACS Omega 2019 Mobilization of metal(oid) oxyanions through circumneutral mine waste-rock drainage. Geochemical and mineralogical data from a long term kinetic tests program and mobilization of metals under neutral drainage. Javadi, Mehrnoush In Doctoral dissertation 2019 Reactive transport modeling of unsaturated hydrology and geochemistry of neutral and acid rock drainage in highly heterogeneous mine waste rock at the Antamina mine, Peru. Representation of coupled hydrological and geochemical processes in mine waste rock with reactive transport modeling. The use of automated mineralogy characterization such as a mineral liberation analyzer (MLA) or QEMSCAN is increasingly popular and has, so far, been mostly applied to improve recuperation of ore in mineral-ore processing (Brough et al., 2017; Lastra, 2007). However, recent studies applied or recommended the application of automated mineralogy to characterize the composition or degree of liberation of minerals in waste rocks (Blaskovich, 2013; Bouzahzah et al., 2014; Dold, 2017; Elghali et al., 2018; Jamieson et al., 2015; Parbhakar-Fox et al., 2014). The aim of this study is to demonstrate that automated mineralogy can complement conventional bulk characterization, evaluate reactivity of minerals, and explain discrepancies observed between solid phase content and drainage composition in field barrels (Chapter 2) and waste-rock pile samples (Chapter 4). Previous studies have demonstrated the importance of the hydrological controls on the drainage composition at Antamina (Blackmore, 2015; Peterson 2014) and other mine sites (Amos et al., 2014; Eriksson et al., 1997; Neuner et al., 2013). Tracer tests evaluate the hydrological response but they can be costly, onerous, and the results are difficult to upscale (Blackmore et al., 19  2018a, 2014; Eriksson and Destouni, 1997; Neuner et al., 2013; Peterson, 2014; Shook et al., 2004). In addition, the water regimes are controlling the weathering patterns of minerals by affecting the mineralogical textures and composition (Delvigne, 1998; Taylor, G; Eggleton, 2001; Velbel, 1985). The relationship between flow paths and the weathered textures of minerals has not previously been investigated at Antamina or applied in the context of waste-rock weathering. This study aims to investigate the hydrological controls on the drainage chemistry and assess the potential relationships with mineralogical weathering textures in field barrels. Moreover, it investigates the use of weathering textures as hydrological response indicators to complement the use of tracer tests and qualify the presence of geochemical or hydrological processes. In this study, the proportion of fast- and slow-flow paths of different hydrological regimes are quantified by tracer test results from field barrels of different compositions. The dominant flow-path component is then compared with mineralogical composition, main weathering textures, and overall drainage chemistry (Chapter 3).  Previous studies covering portions of in-situ waste-rock piles have mainly focused on gas and heat transfer mechanisms and pyrite oxidation processes (Lefebvre et al., 2001; Anterrieu et al., 2010; Lahmira and Lefebvre, 2015; Vriens et al., 2018, 2019b). The studies reporting physical, geochemical, and mineralogical characterization of full-scale waste-rock piles are more scarce and only focused on the surface of piles up to 50 m deep (Sracek et al., 2004, Linklater et al., 2005). In fact, despite the importance of mineral reactivity, waste-rock studies have rarely incorporated quantitative consideration of mineralogical and petrographical aspects within large-scale systems (Dold, 2017; Parbhakar-fox and Lottermoser, 2015). Because of scarce field data on mineral reactivity, secondary mineral formation, and scaling issues, such mineralogical aspects, are challenging to quantitatively incorporate into prediction models of heterogeneous waste-rock piles 20  (Lefebvre et al., 2001; Linklater et al., 2005; Molson et al., 2005; Pedretti et al., 2017). In this study, two boreholes of up to 140 m were drilled at two locations in a full-size waste-rock pile at Antamina, thereby providing a unique profile of continuous core samples with preserved alteration and lithological contacts. The aim was to explore the heterogeneity of the waste rock and evolution of weathering within a ten year old waste-rock pile. In addition, leachate tests from targeted intervals were performed to relate mineralogical composition or reactivity with elemental mobility within the boreholes. The results of this unique waste-rock profile and characterization in a full-scale waste-rock pile are presented in Chapter 4.   Finally, the weathering of waste rock and drainage chemistry were simulated in chapters 2 and 3 with numerical modeling using quantified bulk mineralogical parameters from MLA and drainage results from field barrels. Reactive transport simulations provide quantitative site-specific and calibrated mineral dissolution rates and insights into the potential role of aforementioned mineralogical or hydrological controls in the context of waste-rock weathering. The objective was to corroborate the impact of the observed mineralogical (Chapter 2) or hydrological (Chapter 3) parameters on mineral weathering and long-term drainage composition from field barrels. 1.3.2 Thesis organization This thesis is divided into 3 research chapters (Chapters 2 to 4) describing the application of automated mineralogical characterizations of waste rock linked to the hydrological regime and/or the leaching behaviour of waste rock after weathering at two scales: field barrels (Chapters 2 and 3) and full-scale waste-rock pile (Chapter 4). The introduction (Chapter 1) and conclusion (Chapter 5) are tying the research chapters together and integrating knowledge acquired throughout the thesis. Some repetition may occur between chapters, since the research chapters were written in manuscript form in order to submit to peer-reviewed scientific journals.  21  Chapter 2: Mineralogical controls on drainage quality during the weathering of waste rock 2.1  Introduction In mining operations, large quantities of non-profitable waste rock are produced and placed in heterogeneous waste-rock piles. This exposed waste rock is subject to weathering which may lead to acid rock drainage and enhanced metal mobilization (Jambor et al., 2003; Morin and Hutt, 1997; Price, 2009). To reduce the potential environmental impacts of waste rock drainage, monitoring and prediction of long-term drainage quality must be performed throughout mine life cycle (Lottermoser, 2010; Warhurst, 1999; Younger, Banwart et al., 2002). Drainage quality predictions rely on a quantitative understanding of the major geochemical and physical transport processes that control waste rock weathering kinetics, hydrological transport, and ultimately drainage chemistry (Amos et al., 2014). These processes can, to a large extent, be inferred from bulk geochemical and physical properties of the waste rock (e.g. sulfide and metal content, acid-production and neutralization potentials, porosity, and particle size distribution) that can subsequently be used to estimate reaction rates or physical transport of heat, pore gas and solutes (Amos et al., 2014; Blackmore et al., 2014; Lefebvre et al., 2001; Lorca et al., 2016; Smith et al., 2013; Vriens et al., 2018; Vriens et al., 2019b). In addition to the processes above, the overall reactivity of waste rock is also affected by mineralogical parameters such as mineral composition, abundance, texture, arrangement, and exposed surface area (Brough et al., 2013; Dold, 2017; Jamieson et al., 2015; Lapakko, 2015; Parbhakar-Fox and Lottermoser, 2015; Pedretti et al., 2017). The mineral surface exposure is strongly affected by particle size, but also by surface occlusion from neighboring minerals and 22  passivation by secondary precipitates (Lapakko, 2015; Parbhakar-Fox et al., 2013). Furthermore, galvanic reactions between sulfide couples may affect the dissolution or precipitation rates through preferential oxidation and reduction reactions at the associated mineral interface (Kwong et al., 2003; Liu et al., 2008; Perkins, et al., 1995). Previous studies have illustrated the importance of such mineralogical and petrographical controls, through the impacts of secondary mineral precipitation (Vriens et al., 2019a), surface passivation (Fan et al., 2018; Ghorbani et al., 2013), and galvanic coupling between sulfides (Chopard et al., 2017; Fan et al., 2018; Kwong, 1995; Kwong et al., 2003; Parbhakar-Fox et al., 2013) on anomalous metal mobilization or retention trends from sulfide oxidation and acid-production rates. Aforementioned mineral characteristics can be expressed in quantifiable parameters such as association and liberation (Brough et al., 2017; Ghorbani et al., 2013; Lund et al., 2015), which together with particle size and mineral content, enable an assessment of mineral reactivity. Unfortunately, mineralogical parameters are seldom quantified for heterogeneous waste rock with variable lithology or in field studies (Parbhakar-Fox et al., 2013; Pedretti et al., 2017), and their effect on drainage quality at larger (practice-relevant) scales therefore remains poorly understood. Mineralogical aspects are being integrated to weathering and drainage quality prediction tools (Parbhakar-Fox et al., 2018b), as is commonly done for other waste rock properties like chemical reactivity (through acid-base accounting) (Dold, 2017; Parbhakar-Fox and Lottermoser, 2015), particle size (Malmström et al., 2000) or surface area (Jurjovec et al., 2004).  Modern instrumentation provides a means to characterize the mineralogical properties that affect bulk waste rock reactivity: quantitative automated mineralogy uses a combination of back-scattered electron (BSE) microscopy with energy dispersive X-Ray spectroscopy (XRD) to associate specific image patterns with mineral phases (Fandrich et al., 2007; Gu, 2003). 23  Quantitative automated mineralogy tools such as the Mineral Liberation Analyzer (MLA) are broadly applied in ore mineral processing for the improvement of metal recovery, (Baum, 2014; Gu et al., 2014; Hunt et al., 2011; Lastra, 2007; Lastra and Paktunc, 2016) e.g. by quantifying mineral liberation and association for metallurgical process predictions (Lund et al., 2015). Quantitative automated mineralogy could similarly be useful for the characterization of waste rock (Dold, 2017; Lapakko, 2015), but it has only been scarcely applied in the context of mine wastes to date (e.g. to investigate bulk mineral reactivity (Aranda, 2010; Elghali et al., 2019a, 2018), the weathering rates of specific sulfides (Barazzuol et al., 2012), Pb-rich tailings (Buckwalter-davis et al., 2012), or Zn heap leaching (Ghorbani et al., 2013)). Unfortunately, such studies generally focus on single elements and mineralogical/lithological rock classes with elevated metal content, whereas mines generate low-grade waste rock with heterogeneous mineralogical/geochemical compositions (Vriens et al., 2019a).  To this end, quantitative automated mineralogy method were used to i) quantify changes in the bulk mineralogical parameters (i.e. degree of liberation and mineral associations) during weathering of waste rock of various lithological classes in on-site field experiments at the Antamina mine, Peru, ii) identify discrete elemental mobilization and attenuation mechanisms, and iii) relate these mineralogical controls to the evolution of long-term drainage quality. Finally, a numerical model constrained by quantified bulk mineralogical parameters was used to corroborate the controls of mineralogical parameters on the long-term drainage compositions. 2.2  Materials and methods 2.2.1 Field barrel kinetic testing experiments A total of 17 field barrels were used for on-site kinetic testing at the Antamina mine in Peru. Field barrels were constructed from 55 gallon open-top polyethylene drums and filled with ~300 24  kg of waste rock of a single lithological class (i.e. skarn, intrusive, or marble). Waste rock drainage was funneled into 20 L HDPE containers and sampled periodically. The field barrels were subjected to local weathering conditions by allowing for atmospheric gas exchange and infiltration of precipitation through the open top of the barrels. The climate on site is characterized by average daily temperatures between 5.4°C and 8.5°C, but also by strong fluctuations in precipitation that falls in distinct wet (i.e. between September to April) and dry seasons (i.e. between May to August, Bay et al., 2009). In parallel to the field barrel experiments, aliquots of the emplaced waste rock were stored in thick plastic bags under controlled conditions (referred to hereafter as the reference waste rock). The initial composition of the bulk waste rock and the average drainage concentrations were characterized for the 17 field barrels at the beginning of the experiment, and four field barrels (i.e. 1A, 1B, 2A, and 2B) were selected for additional mineralogical characterization after 7 years.   2.2.2 Chemical analyses of waste rock and drainage  Unweathered composite waste-rock samples were systematically sampled immediately after their excavation and homogenized using the coning-and-quartering method to assure representativeness of the chemical analysis (Aranda, 2010). Solid-phase elemental compositions were analyzed by an external laboratory using a whole-rock 4-acid digestion method, inductively coupled plasma-mass spectrometry (ICP-MS), and a Leco C/S analyzer. In addition, solid-phase elemental compositions were analyzed with an Olympus handheld Field-Portable X-Ray fluorescence spectrometer (FP-XRF), using a reference standard (Olympus, n°316) for calibration. An overview of lithological classes and initial neutralizing potential ratio (NPR) of the bulk waste rock in the four selected field barrels is given in Table B1.  25  Waste rock drainage samples were collected over a period of 7 years with at least a monthly frequency. The accumulated composite drainage samples were filtered (0.45 μm) and acidified with HNO3 upon collection, sent to an external laboratory, and analyzed for (trace) metal concentrations using supercritical fluid extraction, ICP-MS, and a Leco C/S analyzer. An overview of the average drainage concentrations in the four selected field barrels during the study period is presented in Table B2. Triplicate samples and reference standards were analyzed as part of quality control, for all abovementioned analyses. 2.2.3 Preferential elemental retention or mobilization Preferential mobilization or retention of the metals Fe, Cu, and Zn was identified by comparing the elemental ratios between concentrations in the waste rock and drainage for each of the 17 field barrels. The average drainage composition of the first year was divided by the composition of the reference waste rock to assess the initial reactivity. Similarly, the yearly-average drainage composition after 7 years of weathering was divided by the waste rock composition after 7 years of weathering to assess changes in mineral reactivity. For both, the median elemental ratios among all field barrels and the corresponding double median absolute deviation (2MAD) (Reimann et al., 2005) were calculated (Table B1). Subsequently, those elements and field barrels for which elemental ratios were superior or inferior to the median ratio ±2MAD were considered outliers (Escalante et al., 2010). Elements in field barrels for which the elemental ratio between the waste rock and drainage exceeded the median ratio +2MAD, exhibit preferential retention compared to the other field barrels, whereas ratios smaller than the median -2MAD indicate preferential mobilization. Using element ratios normalizes the concentrations for between-barrel differences in material composition and using the MAD statistic allows us to distinguish outliers from within-group variability. Since care was taken to collect representative, pre-homogenized waste rock 26  samples and (composite) drainage samples, it was therefore assumed that the outliers were not an artefact of sample heterogeneity. 2.2.4 Quantitative mineralogical analysis In addition to the composite samples previously described, grab samples (~1 kg) of weathered (pre-homogenized) waste rock from the top (1B) and bottom (1A, 2A, and 2B) of four selected field barrels (Figure B1) were collected after seven years, as were the respective aliquots of stored reference (unweathered) waste rock from the initial excavation. The abovementioned field barrels were selected because they presented either acidic or neutral drainage signatures or strong visual evidence for secondary precipitation. The silt and clay size fractions (<2 mm) of the dried waste rock samples were isolated by sieving to facilitate comparison of the reactive particle size fraction between the field barrels (Strömberg and Banwart, 1999). The elemental composition of this size fraction was determined with FP-XRF as described above. The dried and sieved waste rock samples were mounted with epoxy resin and polished to thin sections and round mounts using a non-aqueous cooling lubricant to minimize potential phase alteration caused by dissolution and/or heat. The quantitative elemental composition, liberation and association of minerals were analyzed using a Thermo-Fisher Mineral Liberation Analyzer (MLA), equipped with a FEI Quanta 600 scanning electron microscope (SEM) and a dual Bruker-AXS silicon-drift energy dispersive X-ray analysis system (EDS). The MLA extended backscatter electron method was applied with a frame resolution of 1.06 μm/pixel. The mineral standards library compiled by Blaskovich, 2013 was used for phase assignment and mineral association and liberation textures were evaluated using the MLA DataView® software. The elemental composition determined by MLA was reconciled with solid-phase FP-XRF measurements (Figure B3) and revealed average agreements ≈85% for Fe, Cu, and Zn. 27  2.2.5 Mineralogical data processing Evaluation of the mineralogical data was focused on the major primary sulfides encountered at Antamina such as pyrite, chalcopyrite and sphalerite. A select number of other (non-sulfide) primary minerals, their oxidation products prior weathering, as well as secondary and tertiary minerals due to weathering of waste rocks were also evaluated (Blowes et al., 2003). Following the classification scheme outlined by Jambor, (1994), this paper focuses on primary minerals and oxidation products, combining secondary and tertiary minerals altogether, because oxidation products and secondary minerals precipitated prior or after weathering are difficult to differentiate by the MLA classification algorithm. Identified primary minerals and all oxidation products (including secondary and tertiary minerals) were divided into the following nine groups to facilitate interpretation with the MLA DataView® software: sulfides, Fe-oxides, Fe-(oxyhydr)oxides, other oxides, carbonates, sulfates, silicates, aggregates, and other minerals (Table B3). Metal(loid) impurities occurred in primary and secondary minerals in solid solution, substituted in their lattices or adsorbed to mineral surfaces. The MLA software was allowed to reconcile elemental composition measured by SEM-EDS with its mineralogical database. The resulting hypothetical stoichiometric formulas of these phases, aggregated or with minor and trace elements, were calculated using the elemental content quantified with SEM-EDS (Blaskovich, 2013).  The collected MLA data were further used to calculate: i) mineral liberation when >70% of the mineral surface was not encapsulated by another soluble or insoluble phase, ii) Mineral Reactivity (MR, equation 1, Aranda, 2010), which quantifies the fraction of Fe, Cu, and Zn of the modal mineral composition available to react, based on minerals with a surface liberation >70%, and iii) the Association Index (AI, equation 2), which quantifies the association between two 28  minerals while correcting for modal mineralogy and liberation. Association indices of > 1 or < 1 indicate that the association of two minerals is more common or less common, respectively, than what is suggested by the modal mineralogy. If AI=0, there is no association between the two minerals, and if AI = 1, the association is as common as the modal mineralogy. Notably, pairs of sulfide minerals with AI > 1 could form a galvanic couple, whereas an AI > 1 involving a primary and a secondary mineral is indicative of potential surface passivation. Equation 1: Mineral Reactivity (MR) (Adapted from Aranda, C., 2010) ܯܴݔ	ሺݓݐ%ሻ ൌ෍ܮܾ݅݁ݎܽݐ݅݋݊ሺݓݐ%ሻ݅	 ∙ ܧ݈݁݉݁݊ݐ݈ܽ	݀݅ݏݐݎܾ݅ݑݐ݅݋݊ሺݓݐ%ሻ݅	௡௜ୀଵ	 where x = element (e.g. Fe, Cu, or Zn) and n = number of mineral phases i Equation 2: Association Index (AI) (From Lund et al., 2015) ܣܫ ൌ ܯ݅݊݁ݎ݈ܽ	ܣ	ܽݏݏ݋ܿ݅ܽݐ݁݀	ݓ݅ݐ݄	ܯ݅݊݁ݎ݈ܽ	ܤ	ሺ݁ݔ݈ܿ. ݈ܾ݅݁ݎܽݐ݁݀	݃ݎܽ݅݊ݏ	݋݂	ܯ݅݊݁ݎ݈ܽ	ܣሻ	ݓݐ%ܯ݋݈݀ܽ	݉݅݊݁ݎ݈ܽ݋݃ݕ	݋݂	ܯ݅݊݁ݎ݈ܽ	ܤ	ሺ݁ݔ݈ܿ.݉݋݈݀ܽ	݉݅݊݁ݎ݈ܽ݋݃ݕ	݋݂	ܯ݅݊݁ݎ݈ܽ	ܣሻ	ݓݐ%  2.2.6 Simulation of drainage quality The effect of mineral reactivity on dissolution rates and drainage chemistry was further investigated using reactive-transport modeling: a quantitative comparison between calibrated model rate constants and experimentally recorded mineralogical liberations and associations provide additional assessment of the role of mineralogical parameters on mineral reactivity. In brief, the recorded temporal evolution of drainage chemistry from field barrels was simulated using a reactive-transport model developed in PHREEQC (version 3.33) (Parkhurst and Appelo, 2013) and the modified WATEQ4F thermodynamic database (Ball and Nordstrom, 1991). Transient flow through a 1-D column was represented by space-time discretization, incorporating the field-monitored outflow data from the field barrels to define hydrological transport during the wet and 29  dry seasons. Mineral dissolution and precipitation were simulated using equilibrium phases for rapid reactions (e.g. calcite, brochantite or gypsum dissolution and precipitation), whereas slower reactions were described with kinetic rate expressions (e.g. sulfide oxidation and silicate dissolution). For the latter reactions, effective rate constants were calibrated to fit the drainage pH, while keeping all other parameters (e.g. modal mineralogy and surface area, based on mineralogical data; Table B4) constant during calibration. This procedure yields site-specific and non-transferable rate constants that were representative of different mineral reactivities and were used for comparison between field barrels only. The agreement between simulated and measured aqueous drainage chemistry was quantified using the normalized root mean square error (NRMSE). Full details of the reactive-transport model are given in Appendix A.  2.3 Results This section focuses on four field barrels with intrusive- (field barrels 1A and 1B) or skarn-type (field barrels 2A and 2B) waste rock, because compared to each other, these field barrels had noticeably different NPR (Table B1), drainage chemistries (i.e. acidic drainage from field barrels 1A and 1B, Figure 2.1) or solid-phase elemental compositions (i.e. elevated Zn or Cu metal contents in field barrels 2A and 2B, Figure B3). The waste rock in the four selected field barrels had a fine-grained (<2 mm) fraction of over 20 wt% (Figure B5), implying a soil-like hydrological regime with predominant matrix flow (Smith and Beckie, 2003). Field barrels 1A and 1B had relatively reactive, net-acid-producing waste rock with elevated (7-20%) sulfide content and little pH-buffering capacity from silicates (<5%) or carbonates (<2%, Figure B2B). In contrast, the field barrels 2A and 2B had less reactive skarn waste rock with a lower (2-6%) sulfide content and more carbonates (5-50%; Figure B2B).   30  2.3.1 Evolution of drainage chemistry The temporal evolution of the drainage chemistry from all field barrels shows distinct annual oscillations as well as long-term trends (Figure 2.1). As a result of the distinct wet and dry seasonality, elemental drainage concentrations were typically elevated at the onset of the wet season (re-dissolution of previously precipitated oxidation products and subsequent flushing), and lower by the end of the wet season. No measurable flow occurred at the end of the dry season and no drainage chemistry could therefore be analyzed (Figure 2.1).  The long-term evolution of drainage chemistry was different for the intrusive and skarn waste rock field barrels: drainage from field barrels 1A and 1B became acidic after a few years of weathering, whereas the pH from field barrels 2A and 2B remained circumneutral for 7 years (Figure 2.1). For both field barrels 1A and 1B, a significant increase in the drainage concentrations of Fe, Zn, and Cu was observed with progressing weathering (Figure 2.1). In field barrel 1A, initial drainage concentrations of Cu and Zn were between 0.005 and 5 mg/L at neutral pH, but 7 years later, orders-of-magnitude higher Cu and Zn concentrations over 100 mg/L were observed in acidic drainage of pH 4.5. For field barrel 1B, a similar acidification trend was observed, although initial Zn and Cu drainage concentrations were already slightly elevated in the first year so that the relative increase in these drainage concentrations was not as pronounced. Nevertheless, the waste rock from field barrel 1B contained less acid-production potential, lower metal content and slightly more neutralization potential than that of field barrel 1A (Table B1, Figure B3), yet more rapidly developed acidic drainage with high metal concentrations (Figure 2.1). Field barrel 2A generated drainage with a relatively stable circumneutral pH and comparably constant Cu and Zn concentrations, whereas drainage from field barrel 2B showed a slight decrease in Cu and Zn concentrations (Figure 2.1). In contrast, Fe concentrations in the drainage from both field barrels 31  2A and 2B increased between the first and last year. Higher Zn concentrations in the drainage of field barrel 2B compared to 2A do not reflect the higher concentrations of Zn in the waste rock composition of field barrel 2A (Figure B3). Different waste rock types (intrusive versus skarn) thus produced clearly different drainage chemistries, and different long-term drainage quality developments arose for geochemically similar waste rock.  Figure 2.1. Evolution of drainage quality from four waste-rock field barrels at Antamina. The measured pH and Cu, Fe, and Zn concentrations in drainage from the intrusive- (1A, 1B) and skarn-type (2A, 2B) field barrels are plotted for the study period between 2008 and 2014 (logarithmic y-axes except for pH). The average pH values and Cu, Fe, and Zn concentrations calculated for the first and last years of the study period are indicated by the correspondingly colored bars left and right in each frame. The analytical detection limits for the Fe and Cu measurements are illustrated by the triangles on the respective y-axes. The legend applies to all frames. 32  2.3.2 Identification of preferential mobilization or attenuation  The ratios of elemental concentrations in the waste rock over those in the drainage spanned several orders of magnitude for Cu, Fe, and Zn (Figure 2.2). This variation is reflective of the wide range of waste-rock lithologies used in the field barrels and their different reactivity (Table B1). Generally, in all field barrels, Fe appeared to be preferentially retained in the waste rock, whereas Zn exhibited relatively high mobility. These overall trends agree with low mobility of Fe(III) down to a pH of 3-3.5 (Smith, 2007) and with observations from larger-scale experimental piles, and reflect the poor mobility of Fe in the oxic circumneutral drainage from the majority of Antamina waste rock (Peterson, 2014; Vriens et al., 2019a).  From all calculated metal concentration ratios, 12 anomalies fall outside the MAD thresholds (Figure 2.2). Six of those outliers occur in field barrels 1A and 1B and are associated with preferential mobilization; another six are the result of preferential retention and are distributed over field barrels 1A, 2A and 2B. Field barrel 1A initially shows high concentration ratios for Cu and Fe (Figure 2.2), indicating that these metals are effectively retained at circumneutral drainage pH in the waste rock of field barrel 1A. However, after 7 years of weathering, the concentration ratios for Cu, Zn, and Fe in field barrels 1A and 1B are comparatively low: this suggests a higher mobility of these metals in acidic drainage (pH≈4) when dissolution of primary minerals is sustained and natural attenuation through adsorption or precipitation of secondary of Cu- and Zn-(oxyhydr)oxides, (hydroxy)sulfates or carbonates is limited. In contrast, the concentration ratios for Fe, Cu, and Zn were consistently high in field barrels 2A and 2B, resulting from the stable circumneutral drainage imposed by abundant carbonate buffering in the skarn-type waste rock. As a result, the Cu and Zn drainage concentrations from skarn field barrel 2A were lower than that 33  from the intrusive field barrels, despite higher solid-phase concentrations (Figure 2.1 and Figure B3).  Thus, overall metal attenuation and mobilization appear to be largely related to waste rock composition and neutralization capacity, as pH strongly affects mineral solubility and metal sorption. However, a comparison of the elemental ratios in waste-rock with similar lithologies also suggests that drainage chemistries are not only controlled by the waste-rock metal content and abundance of acid/base releasing minerals, but also by mineral surface liberation or associations, which affect mineral reactivity.  Figure 2.2. Distribution of the elemental concentration ratios (i.e. waste rock solid phase content over aqueous drainage concentrations) for the initial unweathered and weathered waste rocks in the intrusive (1A, 1B) and skarn (2A, 2B) field barrels. The boxplots indicate the distribution of elemental concentration ratios using median ±2MAD for outlier threshold. Elemental ratios below -2MAD (lower whisker) indicate relative mobilization; elemental ratios above +2MAD (upper whisker) suggest relative retention.  34  2.3.3 Mineralogical observations To investigate the temporal evolution of waste-rock liberation and reactivity, the distribution of Fe, Cu, and Zn minerals in the reference unweathered waste rocks (Figure 2.3) was compared to that of the weathered waste rocks (Figure 2.4). Moreover, the reactivity of sulfides (i.e. sum of pyrite, chalcopyrite, and sphalerite) and total mineral reactivity (i.e. sum of all minerals bearing Fe, Cu or Zn) were calculated, as well as pyrite-association indices (Figure 2.5).  2.3.3.1 Intrusive waste rock Intrusive waste rock in both field barrels 1A and 1B was relatively reactive with high sulfide and low carbonate content (Figure B2B), and therefore developed acidic drainage with elevated metal concentrations (Figures 2.1 and 2.2). This reactivity was reflected in the mineralogy of intrusive waste rock composed of abundant liberated metal-bearing sulfides with an elemental distribution reaching up to 64% Fe, 99% Cu and 89% Zn (Figures 2.3 and 2.4). In contrast, acid-neutralizing minerals were scarce with <1% calcite and ~2% silicates such as plagioclase or wollastonite (Figures 2.3 and 2.4). The total bulk solid-phase concentrations of all metals investigated generally decreased after weathering, whereas the proportion of Fe-oxidation products doubled in 1B, decreased in 1A, and was stable for Cu and Zn in 1A and 1B (Figures 2.3 and 2.4). Iron was the most abundant of the investigated metals and 85% of total liberated Fe bearing primary minerals in the unweathered waste rock were composed of Fe-sulfides (e.g. pyrite and chalcopyrite), Fe-oxides and Fe-silicates (Figures 2.3 and 2.4). Waste rock in field barrel 1A not only had a higher Fe content than that in field barrel 1B, but also slightly more liberated primary Fe-sulfides. The oxidation products associated with Fe constituted <30% of the total Fe in weathered waste rock samples, and included up to 66% Fe-oxides, 45% Fe-silicates (e.g. altered mica/amphiboles), as well as <15% Fe-(hydroxy)sulfates and Fe-carbonates (e.g. siderite, ankerite, 35  jarosite; Figures 2.3 and 2.4). In general, the abundance of Fe-bearing oxidation products did not vary significantly upon weathering and most likely precipitated prior to disposal, with the exception of an increasing fraction of Fe-(hydroxy)sulfates and sulfates observed in the weathered waste rock of field barrel 1B, and slightly decreasing fractions of Fe-carbonate in both field barrels.  The Cu content of intrusive waste rock was roughly an order of magnitude lower than the Fe content. In secondary oxidation products, the amount of Cu present was <6% of the total Cu content, which is less than the analogue fraction of Fe present in oxidation products (i.e. 30% of the total Fe content). Yet, Cu-bearing liberated primary minerals were very abundant (>95% of total sulfides) in both the weathered- and unweathered waste rock of each field barrel (Figures 2.3 and 2.4). In the intrusive waste rock, the distribution of oxidation products associated with Cu included up to 80% silicates such as altered Cu-bearing mica/amphibole, up to 38% (hydroxy)sulfates such as jarosite with traces of Cu, and <25% Cu-Fe-oxides and <19% carbonates with traces of Cu. Similar to Fe, the abundance of Cu-bearing oxidation products was relatively constant during weathering. Such secondary Cu-bearing alteration phases were already present at the beginning of the experiments, although the fraction of secondary (hydroxy)sulfates with Cu increased over weathering and the fraction of secondary Cu-bearing carbonates decreased. Although Zn concentrations were relatively low in the intrusive material, liberated sulfide minerals were similarly dominant in the mineralogy associated with Zn and contributed up to 90% of the total Zn content (Figures 2.3 and 2.4). Interestingly, the contribution of Zn-sulfides to the primary mineral composition increased over weathering, although this may also be an artifact of sample heterogeneity and/or significantly decreasing (>50%) Zn concentrations that were already low in the unweathered waste rock. Field barrel 1A contained relatively more sulfides in association with Zn than field barrel 1B. In weathered waste rock, the deportment of oxidation 36  products associated with Zn was dominated by up to 92% silicates such as altered Zn-bearing mica/amphibole, up to 25% Zn-bearing Fe-(hydroxy)sulfates and up to 8% carbonates such as siderite with traces of Zn.  In line with the mineralogical observations discussed above, the intrusive waste rock exhibited high reactivity indices for Fe-, Cu-, and Zn-sulfides, in line with the presence of up to 50% pyrite, 90% chalcopyrite and 50% sphalerite (Figure 2.5). Whereas the reactivity indices for Fe-sulfides in the intrusive waste rock decreased during weathering, those of Cu- and Zn-sulfides appeared to increase. Calculated association indices indicate that sphalerite and chalcopyrite are oftentimes more commonly associated with pyrite than suggested by the modal mineralogy of intrusive waste rock (Figure 2.5). This is substantiated by elemental mapping that showed coupling of pyrite and chalcopyrite, with altered etch pits at the surface of chalcopyrite and surrounded by alteration products (Figure B9). Association indices in the (un)weathered intrusive waste rock further indicated strong associations between pyrite and various secondary minerals such as (oxyhydr)oxides, carbonates, sulfates, and silicates (Figure 2.5), supporting the potential presence of surface passivation before and after waste rock disposal. This could similarly be corroborated by observations of disseminated Fe-oxide alteration with traces of Si and Cu at the surface of pyrite (Figure B8).  In summary, the drainage chemistries and concentration ratios from the intrusive field barrels (Figures 2.1 and 2.2) revealed that relatively large proportion of Fe, Cu and Zn were mobilized from mineral dissolution in field barrel 1B compared to 1A. Although microbial conditions (Blackmore et al., 2018b) and particle sizes may also affect weathering rates and should be further investigated, the mineral distributions and calculated reactivities suggest that preferential mobilization from field barrel 1B may partially be related to a higher abundance and liberation of 37  acid-producing Cu- and Fe-sulfide minerals, the formation of secondary phases that are unstable under acidic drainage conditions (e.g. oxides and carbonates versus more refractory silicates; Figures 2.3-2.5) and/or their potential galvanic interaction.  2.3.3.2 Skarn waste rock Compared to the more reactive intrusive waste rock, the skarn waste rock of field barrels 2A and 2B had lower sulfide- and higher carbonate content (Figure B2B). Consequently, the mineral distributions of the skarn waste rocks reveal smaller contributions of liberated sulfides (distributed over up to 10% Fe-, 38% Cu-, and 9% Zn-sulfides in field barrel 2A (Figures 2.3 and 2.4). Liberated primary carbonates bearing (traces of) Fe, Cu or Zn were not detected in significant amounts in the skarn waste rock. The skarn waste rock further had high fractions of liberated silicates that contained up to 80% Fe, 63% Cu, or 90% Zn. Similar to the intrusive waste rock, weathering of the skarn waste rock led to generally lower bulk metal contents, but only marginally (<5%) increased the fraction of oxidation products relative to primary minerals (i.e. maximum 15% for Fe in field barrel 2A).  The distribution of Cu- and Zn-bearing primary minerals was quite dissimilar between the two skarn waste rocks: liberated primary minerals in field barrel 2A mainly consisted of silicates such as dioptase, rhodonite, and willemite with impurities of Cu and Zn (62% of the total Cu and 85% of Zn were associated with aforementioned silicates) and Cu- or Zn-bearing sulfides such as chalcopyrite, covellite, or sphalerite (with 37% of Cu and 10% of Zn present in these sulfides). In contrast, field barrel 2B contained significantly higher fractions of liberated >90% sulfides relative to <10% silicates, both also bearing Cu and Zn. The distribution of Fe-bearing primary minerals was relatively comparable between the two skarn waste rocks before and after weathering, and was composed of, an average, >68% Fe-bearing silicates such as epidote, dioptase, and rhodonite.  38  The elemental distribution of oxidation products in the weathered skarn waste rock was similar to that of the reference waste rock, except for decreasing fractions of Fe-oxides and slightly larger fractions of Cu- and Zn-(hydroxy)sulfates in field barrel 2B (Figures 2.3 and 2.4). The oxidation products associated with Fe in the skarn waste rock were similar to that observed in the intrusive waste rock, with predominantly Fe-oxides and Fe-bearing silicates such as amphibole, rhodonite, or talc with Fe impurities. Copper and Zn-bearing silicates were composed mainly of altered rhodonite or fayalite with impurities of Cu or Zn (Table B3), although significant fractions of Cu-bearing Fe-oxides and carbonates such as malachite were identified in the waste rock from field barrel 2A.  In accordance with the lower sulfide content liberated in field barrel 2A (Figure 2.2), this skarn material exhibited low sulfide mineral reactivity indices of 15% (i.e. sum of pyrite, chalcopyrite and sphalerite Fe-Cu-Zn- reactivities) compared to other field barrels (1A, 1B, and 2B; Figure 2.5). This agrees with the fact that the main liberated minerals in the skarn waste rock of field barrel 2A were primary silicates and secondary Fe-oxides that are poorly soluble at neutral pH and provide potential sorption sites. The higher reactivity of Cu- and Zn-sulfides in field barrel 2B compared to 2A reflects their presence in the primary mineral distribution, but could also be an artifact of generally reduced Cu- and Zn-contents in that waste rock (Figures 2.3 and 2.4). Calculated pyrite-association indices suggest poor associations between pyrite, sphalerite, and/or chalcopyrite in field barrel 2A and the presence of contact between these phases in field barrel 2B (Figure 2.5). Elemental mapping images revealed sulfide coupling between pyrite, sphalerite and chalcopyrite in sample 2B (Figure B9). In addition, pyrite was associated with a variety of secondary minerals in the skarn waste rock of field barrel 2A and 2B, which was confirmed by observations of sulfide 39  grains that were coated with amorphous Fe-Si-oxide alteration products enriched in Cu and Zn (Figure B8).   Figure 2.3. Overview of the distribution of Fe-, Cu-, and Zn-bearing primary (Liberated fraction > 70%) and oxidation products/secondary minerals from the unweathered reference waste rock samples from field barrels (1A and 1B as well as 2A and 2B) that are respectively composed of intrusive and skarn waste rock.  Wt% Wt% Wt% Wt% 40   Figure 2.4 Overview of the distribution of Fe-, Cu-, and Zn-bearing primary (liberated fraction >70%), oxidation products/secondary minerals in the weathered waste rock from field barrels (1A and 1B as well as 2A and 2B) that are respectively composed of intrusive and skarn waste rock.   Wt% Wt% Wt% Wt% 41   Figure 2.5. Mineral reactivity (top and middle) and pyrite-association indices (bottom) of the initial non-weathered and weathered waste rock from the intrusive and skarn-type field barrels. The mineral reactivity quantifies the fraction of Fe-, Cu-, and Zn-bearing total minerals (top) or total sulfides (middle) with surface liberation >70% considered available to react in the investigated waste rock samples. The figure on top is the sum of sulfides, Fe-oxides, Fe-(oxyhydr)oxides, other oxides, carbonates, sulfates, silicates, and other minerals, and the figure in the middle is the sum of pyrite, chalcopyrite, and sphalerite. The association indices of a selection of minerals with pyrite (bottom, logarithmic y-axis) illustrate whether that association is more or less common than what is suggested by the modal mineralogy and liberation; an AI of 1 indicates the modal mineralogy (thick dashed line); and an AI of 0 (not visible on logarithmic axis) indicates the absence of association. Legends are indicated next to the frames.   42  Summarized, the lower geochemical reactivities observed for certain skarn- and intrusive waste rocks (Figures 2.1 and 2.2) may not be solely attributable to lower sulfide content (Figure B2B), but also to a reduced sulfide mineral liberation and the precipitation of poorly soluble, passivating secondary oxide minerals (Figures 2.3 and 2.4): mineralogical parameters such as liberation and association may explain differences in reactivity within and between the waste rock classes. For instance, the higher relative attenuation of Fe, Cu and Zn in field barrel 2A compared to 2B may be linked to lower reactivity indices of sulfides, less important galvanic sulfide dissolution (Figure 2.5) or by smaller precipitation of secondary minerals in field barrel 2B (Figure 2.4). Overall, the presented mineralogical data illustrate how information on metal distributions amongst primary and secondary mineral classes or on mineral surface exposure, passivation, or association, can add important subtleties to the interpretation of waste rock reactivity beyond its bulk geochemical reactivity (i.e. metal content and net-neutralization potential).  2.3.4 Reactive transport modeling  The mineralogical parameters of liberation and associations providing information on mineral reactivity support calibrated mineral rate constant obtained from reactive transport model simulations. The measured and simulated drainage chemistries for field barrels 1A, 1B, 2A and 2B are compared in Figure 2.6. Simulations with calibrated rate constants display relatively good agreement with the measured temporal evolution of the drainage pH, Fe, Cu, Zn, and sulfate concentrations for all field barrels. The general concentration ranges, peak concentrations, as well as the long-term trends in drainage concentrations were captured by the simulations (Figures 2.6 and B6). The agreement is corroborated by lower normalized errors (NRMSE<0) between simulated and measured values indicating a better fit except for SO4,  Cu, or Fe in field barrels 1A or 1B and Zn in 2B (Table B5). Zinc concentrations were slightly underestimated for field barrels 43  1A and 1B, and for Cu and Fe in field barrels 2A and 2B. Yet, sulfate levels were slightly overestimated in all simulations as well as Zn in field barrel 2B (Figures 2.6 and B6). Reasons for slight discrepancies between simulated and measured drainage quality include equilibrium modeling versus disequilibrated aqueous solutions, abrupt variations in hydrological regimes (infiltration versus evaporation), or the averaging out from the analysis of composite drainage samples. Further optimization and parameterization of the reactive transport model may therefore be the topic of future work.  The calibrated effective rate coefficients of the sulfide oxidation/dissolution reactions are compared with the reactivity indices of pyrite, chalcopyrite and sphalerite (referred to as: Fe-Pyrite, Cu-chalcopyrite, and Zn-sphalerite reactivity, respectively) in Table 2.1. Overall, the calibrated effective rate coefficients reflected the differences in mineralogical reactivities between phases in the studied field barrels. For instance, the calibrated effective rate coefficients were higher for chalcopyrite and lower for pyrite in field barrel 1B compared to 1A, despite similar modal compositions, and led to a higher simulated depletion of chalcopyrite for barrel 1A (i.e. 74%) and of pyrite for field barrel 1B (i.e. 81%). Furthermore, a lower pyrite reactivity (i.e. lower surface liberation) in intrusive field barrels 1A compared to 1B (9% difference), despite a higher modal content, resulted in a 7% difference in the calibrated effective rate coefficient. Also, the higher simulated depletion of chalcopyrite in barrel 1A agreed with a higher modal composition of 2.6 mol/L, compared to a lower simulated depletion and modal composition (1.3 mol/L) in field barrel 1B, while both field barrels exhibited a similar chalcopyrite reactivity of 80-83 %. Finally, the higher modal content of sphalerite (from 2 to 3 orders of magnitude higher) in the skarn materials (i.e. field barrels 2A and 2B) compared to the intrusives (field barrels 1A and 1B) and the equivalently enhanced sphalerite reactivity (1.4 to 9 times higher in skarn than in the intrusive 44  materials), was reflected in the up to 3.5 times higher effective rate coefficients and up to 3 orders magnitude higher simulated depletion.   Figure 2.6. Comparisons of the temporal evolution of the measured drainage chemistry from field barrels 1A, 1B, 2A, and 2B (adapted from Figure 2.1) with the simulated drainage chemistry (legend applies to all frames). Table 2.1. Initial modal compositions, simulated depletion (%), and mineral reactivity (MR %) for Fe, Cu, and Zn as pyrite, chalcopyrite and sphalerite in the studied field barrels. Calibrated effective rate coefficients (Calibrated keff; different units) as obtained by calibrating drainage chemistry simulations to pH.  Mineral and  Field barrel Modal  composition[Mol L-1] Simulated depletion [%] MR [%]   Calibrated  keff Pyrite Fe-Py Log [s-1] 1A  2.9 11 21 -7.24 1B  1.8 81 30 -6.75 2A  0.2 79 4 -8.5 2B  0.8 4 31 -8.5 Chalcopyrite   Cu-CPY Log [s-1] 1A  2.6 74 80 -3.2 1B  1.3 52 83 -4.5 2A  0.3 0 4 -5 2B  0.3 0 54 -4 45   Mineral and  Field barrel Modal  composition[Mol L-1] Simulated depletion [%] MR [%]   Calibrated  keff Sphalerite Zn-Sphal L Mol -1s-1 1A  0.02 35 9 2.5 1B  0.004 19 5 1 2A  1.2 2 13 6 2B  1.3 12 46 7 Cpy = Chalcopyrite, Pyr = Pyrite, Sphal = Sphalerite 2.4 Discussion 2.4.1 Mineral modal composition, reactivity, and liberation  Mineral reactivity accounts for liberation (Equation 2) and therefore incorporates effects of primary mineral exposure as well as surface passivation due to secondary mineral precipitation. The intrusive waste rock in field barrels 1A and 1B displayed a higher total mineral reactivity (>75% of all mineral) and generated acidic drainage, whereas the skarn waste rock in field barrels 2A and 2B exhibited lower total mineral reactivity (between 45% and 75% of all mineral) and generated neutral drainage for the entire study period (Figure 2.1 and Figure 2.5). These observations agree with larger-scale experimental piles at Antamina that also show higher reactivity of the intrusive compared to the skarn waste rock (Vriens et al., 2019a). The waste rock in field barrels 1A and 1B had similar particle size distributions (Figure B5), comparable elemental solid phase composition (Figure B3), as well as acid-producing and neutralizing potentials (Table B1). However, their elemental concentration ratios are orders of magnitude different (Figure 2.2). This discrepancy may be partially explained by the fact that field barrel 1B exhibited a higher sulfide mineral reactivity compared to field barrel 1A (Figure 2.5), so that the highly reactive (i.e. exposed) <2 mm size fraction of field barrel 1B may have caused rapid acidification. In another example, the orders of magnitude higher concentration ratios for Cu and Zn in field barrel 2B compared to field barrel 2A (Figure 2.2) indicate significant relative retention of Cu and Zn, 46  despite higher grades of these elements in the waste rock of field barrel 2A (Figure B3). Similarly, these lower release rates of Cu and Zn in 2A will hardly be related solely to a smaller reactive particle size fraction (Table B1, Figure B5) or marginally higher pH-neutralizing mineral content (Figure B2A). Instead, this relative retention compared to field barrel 2B may also be explained by lower total mineral and sulfide mineral reactivity (Figure 2.5), which may reflect extensive passivation of Zn-sphalerite and Cu-chalcopyrite. These results therefore illustrate that waste rock classification schemes and drainage quality prediction models should be optimized by accounting for mineral modal composition, liberation, and reactivity patterns for individual lithologies, in addition to solid-phase metal content and acid/base accounting. 2.4.2 Galvanic reactions In addition to controls of drainage pH on metal mobility, associations between sulfidic minerals, including partial or full encapsulation, may also be linked to preferential mobilization and elevated drainage concentrations. Such galvanic interactions between sulfides in contact with each other have previously been shown to cause preferential mobilizations in kinetic tests (Chopard et al., 2017; Parbhakar-Fox et al., 2013; Qian et al., 2018, Liu et al., 2008; Qing You et al., 2007) and residual mineral leaches (Ghorbani et al., 2013; Kwong, 1995; Kwong et al., 2003). In this study, the pyrite-association indices in field barrels 1A, 1B and 2A suggest quantitatively significant sulfide mineral associations (i.e. between pyrite and sphalerite or chalcopyrite) before and after weathering (Figure 2.5). These associations may have contributed to the preferential dissolution of sphalerite and chalcopyrite while preserving pyrite, thereby explaining preferential release of Cu and Zn and/or retention of Fe (Figures 2.1 and 2.2). SEM images show evidence of preferential surface dissolution and alteration at coupled sulfide surfaces (Figure B9). Although Fe is simultaneously attenuated in Fe-(oxyhydr)oxides or other secondary precipitates at drainage of 47  pH>3 in field barrels 1A and 2A, such selective retention is improbable in the acidic drainage (pH<3) of field barrel 1B. Therefore, galvanic dissolution triggered by the association of sulfide couples may have contributed to the observed preferential mineral dissolution of chalcopyrite and sphalerite and metal leaching of Cu and Zn or mineral conservation of pyrite and metal retention of Fe in reactive waste rock.  2.4.3 Secondary mineral formation: surface passivation and sorption Secondary-mineral precipitation can cause selective retention of metals and lead to passivation of reactive mineral surfaces, and also provide effective sorption sites for metals (i.e. indirect retention) (Huminicki and Rimstidt, 2009; Moncur et al., 2009). In this study, low Fe concentrations in the circumneutral drainage (pH>6) from all field barrels during the first year (Figure 2.1), and the anomalously low Fe concentration ratios in field barrels 1A and 2B (Figure 2.2) were most likely caused by pyrite passivation and strong preferential Fe attenuation through precipitation of secondary Fe. This is illustrated by associations between pyrite and secondary Fe-(oxyhydr)oxides and decreased Fe-sulfide reactivities in the weathered waste rock of field barrels 1A, 1B and 2B (Figure 2.5), as well as SEM imaging of pyrite alteration rims (Figures B8 and B9) and reduced pyrite liberations (Figure B7). Fe precipitates remained stable even in the acidic drainage (pH 4) of field barrel 1A, keeping long-term Fe drainage concentrations low despite elevated Fe content in the waste rock (Figure 2.1 and Figure B3). This is consistent with the typical mineral dissolution and pH-buffering sequence in which Fe-(oxyhydr)oxides dissolve at pH 3-3.5 (Amos et al., 2014; Blowes and Ptacek, 2003). In addition, previous analyses of secondary Fe precipitates at Antamina (Vriens et al., 2019a) and other mines (Davies et al., 2011; Gieré et al., 2003; Smuda et al., 2007; Sracek et al., 2004) observed Fe retention through secondary mineral formation, or a reduction of pyrite dissolution rate coated with Fe-(oxyhydr)oxide layer (Fan et 48  al., 2018) in circumneutral to slightly acidic pH. Amorphous secondary Fe-(oxyhydr)oxides are known to provide effective sorption sites for a variety of trace metals (Smith, 2007), and in this study, a potential role of metal sorption onto secondary Fe minerals is implicated by observed enrichments of Zn and Cu in the alteration rim of chalcopyrite using SEM (Figure B8). Finally, the collected mineralogical data further suggest precipitation of various other secondary minerals (e.g. Cu- and Zn-(hydroxy)carbonates, -(hydroxy)sulfates, and -silicates, Figure 2.4), corroborated by SEM observations of amorphous Fe-Cu-Si-oxide and Fe-Cu-Zn-Si-oxide secondary minerals on pyrite and chalcopyrite surfaces (Figures B8 and B9). This agrees with previously observed formation of discrete secondary Cu- and Zn-minerals in weathered Antamina waste rock (Vriens et al., 2019a) and other mines sites (Davies et al., 2011; Gieré et al., 2003; Smuda et al., 2007), and illustrates that discrete secondary minerals controls on metal mobilization under circumneutral and acidic drainage may also exist for Cu or Zn.  2.4.4 Simulation of mineralogical controls on drainage chemistry In this study, effective rates were calibrated to best fit the drainage pH data with a focus on the comparison of differences between four field barrels (i.e. 2 intrusive and 2 skarn) with specific lithology and reactivity. The differences between those field barrels were then quantitatively explained using the mineral reactivity and association indices. The results presented in Table 2.1 reflect an implicit mineral-specific correction for all mechanisms that alter mineral dissolution rates (i.e. obstruction through surface passivation or galvanic interactions, inhibiting or accelerating dissolution). The lower calibrated rate constants for sulfide dissolution in field barrel 1A over 1B agree with mineralogical data showing higher partial surface occlusion for minerals in 1A compared to 1B: i.e. the liberated pyrite surface in the unweathered waste rock of field barrel 1B was ~30% of the modal mineralogy, compared to 21 % in field barrel 1A (Table 2.1 and Figure 49  2.5). Further, the higher calibrated rate constant for sphalerite oxidation in field barrel 2B over 2A agrees with the higher liberated modal Zn-sphalerite surface of 46% compared to 13%, respectively. In addition, the observed association between pyrite and sphalerite in field barrel 2B (Figure 2.5), in combination with a lesser depletion of pyrite at the end of the simulation (Table 2.1) suggest the occurrence of galvanic interactions that inhibit pyrite dissolution while accelerating sphalerite dissolution in this field barrel. These results thus support the relative retention of Fe or Zn as deduced from the elemental concentration ratio outliers observed in field barrels 1A, 2A, and 2B (Figure 2.2). In summary, the mineralogical parameters may explain some of the discrepancy observed between kinetic rate expressions/effective rate constants and field-scale experimental data. To illustrate this, we used mineral-specific corrections in effective rate-constants to account for variations in mineral reactivity. For the studied waste-rock of comparable particle size, these corrections thus reflect differences in reactivity caused by galvanic interactions, occlusion, or texture. Finally, important distinctions between and within material classes may be isolated by site-specific and calibrated dissolution rates that explicitly represent mineralogical properties. Future predictions of long-term waste rock weathering and drainage quality may therefore be further developed by quantitatively incorporating these parameters into effective rate constants, rate expressions and entire reactive-transport simulations. 2.5 Conclusions Multi-year kinetic field barrel experiments under field conditions provided insights into the mineralogical weathering mechanisms and long-term drainage development of different types of waste rock. The presented data show that micro-scale mineralogical processes such as mineral surface passivation and galvanic interactions, in addition to drainage pH and bulk geochemical properties of the waste rock (e.g. acid-neutralization potential or particle size), affect waste rock 50  weathering rates and the resulting drainage quality. In this study, these mineralogical processes were examined through quantification of the modal mineralogical compositions, as well as of the liberations and associations of primary and secondary Fe, Cu, and Zn minerals in variably weathered waste rock using quantitative automated mineralogy. Reactive transport simulations provided insights into the potential role of aforementioned mineralogical controls in the context of waste rock weathering, as mineral oxidation/dissolution rates quantitatively supported mineralogical texture and arrangement observations. Mineralogical assessments further explained the distinct reactivities amongst field barrels of comparable composition. In summary, this study shows that quantitative automated mineralogy may complement conventional bulk geochemical assessment of waste rock reactivity to account for mineralogical processes in improved drainage quality prediction models.    51  Chapter 3: Relationship between flow paths, patterns of weathered minerals, and drainage chemistry from field barrel kinetic tests 3.1  Introduction When waste rock is exposed to weathering, physical and geochemical processes are occurring simultaneously and controlling the mineralogical reaction rates and water discharge quality (Amos et al., 2014; Jambor, 2003; Morin and Hutt, 1997; Price, 2010). Consequently, physical processes such as hydrology, heat and gas transport, combined with geochemical processes such as microbial activity, mineral solubility, and metal mobility, will impact mineralogical reactivity (Amos et al., 2014; Blackmore et al., 2014; Lefebvre et al., 2001; Lorca et al., 2016; Smith et al., 2013; Vriens et al., 2018, 2019a). This study explores the relationship between hydrological and geochemical processes, more specifically the relationship between hydrological regime, weathering patterns of residual and secondary minerals, and drainage chemistry from waste-rock field barrels of different composition. Mineralogical parameters such as composition, abundance, and textures may influence mineral weathering and reactivity which were investigated in Chapter 2 (Dold, 2017; Parbhakar-fox and Lottermoser, 2015; St-Arnault et al., 2019a), while hydrological regimes impact the residence time and mineral dissolution or precipitation rates (Blackmore et al., 2014; Colombani, 2008; Evans and Banwart, 2006; Lasaga et al., 1994; Neuner et al., 2013; Velbel, 1993). Tracer tests were applied in previous studies to measure the residence time of water in a system, characterize the hydrological response (e.g. preferential flow paths), and help predict the long-term impact of fluid flow on solute transport (Eriksson et al., 1997; Neuner et al., 2013; Peterson, 2014; Blackmore et al., 2014 and 2018). However, the implementation of tracer tests at large scales can be expensive and often present technical challenges (Shook et al., 2004). In 52  addition, the hydrological regimes from smaller to larger scales experiments are difficult to extrapolate due to the increase complexity of systems at larger scales (Eriksson and Destouni, 1997; Stockwell et al., 2006; Blackmore et al., 2014; Vriens et al., 2019a). This study investigates the use of indicators of hydrological response, such as grain size or mineralogical textures from small-scale experiments. It is hypothesized that the morphology of secondary or residual minerals is controlled in part by hydrology, such that it could be used by proxy to infer hydrology. These markers could therefore be first qualitative indicators of hydrological and geochemical processes in larger systems. The solute concentrations and loads are determined by the flow paths and volume of water, reactive surface area of exposed minerals, geochemical controls (e.g. pH, secondary mineral precipitation or dissolution, and sorption), and hydrodynamic mixing (Evans and Banwart, 2006; Peterson, 2014; Vriens et al., 2019a). The water flowing through waste rock can be characterized by preferential flow channels through the coarse-sized fraction (i.e. cobbles and boulders) of the waste rocks or voids, whereas matrix flow is associated with the flux of water in the finer grain material (i.e. sand or clay) (Blackmore et al., 2014). The water will flow rapidly in preferential flow paths and more slowly in matrix flow paths (Neuner et al., 2013; Blackmore et al., 2018a). The presence of fast-flow channels and fast- or slow-flow matrix material influences the contact time between rock and solute which impacts the precipitation or dissolution of minerals and solute concentrations and loadings to the environment (Neuner et al., 2013; Peterson, 2014; Blackmore et al., 2014). The flow volumes impact the residence time of water and mineral dissolution and precipitation along with other solubility controls such as pH (Neuner et al., 2013; Nordstrom et al., 2015; Stockwell et al., 2006). For instance, seasonal variation of solute concentrations and loads were previously attributed to mixing of different proportions of higher and lower velocity 53  flow paths, demonstrated from tracer tests results in experimental piles at the Antamina mine, Peru (Peterson, 2014). Higher metal concentrations were associated with matrix flow because of longer water-rock contact time, while lower metal concentrations were associated with preferential flow due to shorter water-rock contact time (Peterson, 2014). Higher solute loadings occurred in the wet season due to increased water volume, while lower solute loadings were observed in the dry season due to decreased water volume (Peterson, 2014). In addition, the characterization of the degree of preferential flow in unsaturated experimental waste-rock piles from bromide tracer tests at Antamina revealed both dominant preferential flow and faster matrix flow components (Blackmore, 2015). This observation suggested that lower water-rock contact and depletion rates were sustained over long periods of time. In contrast, dominant matrix flow infiltration suggested faster depletion of the metal source assuming similar mineral composition, due to longer waste-rock interactions sustained over a long period of time compared with the previous scenario (Blackmore, 2015).   The heterogeneity of flow will limit the participation of potential available mineral surfaces to pore fluids reaction (Velbel, 1993). For instance, under water-unsaturated conditions and  the vadose zone, where the mineral surfaces are less likely to be in contact with water, the chemically reactive fraction of the surface of the mineral (e.g. gypsum) is limited and the release of its dissolution products is determined by flow velocity (Kuechler et al., 2004). Transport-controlled minerals (i.e. readily soluble salts) have a dissolution rate proportional to the flow rates, whereas the dissolution rates of surface controlled minerals, such as silicates, are too slow to be affected by flow rates (Evans and Banwart, 2006; Lasaga et al., 1994). The dissolution rates of carbonate and sulfate minerals have mixed transport and surface controls (Colombani, 2008). Stoops et al., (1979) and Delvigne, (1998) reviewed the mineral patterns caused by weathering and reported that 54  textures of residual and secondary minerals were based on the weathering conditions, as well as the internal structural discontinuities of primary minerals. Depending on the flow regime and quantity of water available, primary minerals are weathered into residual and secondary products with typical morphological textures (Delvigne, 1998; Taylor and Eggleton, 2001; Velbel, 1985). The concentration of less mobile and moderately mobile elements or minerals (i.e. relative versus absolute accumulation), resulting from the weathering of minerals produce boxwork and infilling textures indicative of flow effects (Figure 3.1, adapted from Delvigne, 1998). The boxwork textures (i.e., linear banded patterns, Figure 3.1) are associated with faster more abundant water circulation; whereas, the infilling textures (i.e., rims or agglomerate patterns, Figure 3.1) are more representative of slower circulation and less abundant water (Figure 3.2, adapted from Delvigne, 1998). Faster and abundant water circulation promotes advection and relative accumulation of oxidation products and results in the leaching of easily weathered minerals and dissolved mobile elements. As a result, the less mobile elements (e.g. Al, Fe, Ti) are conserved as uniform and more stable secondary products within boxwork textures  which correspond to the relative accumulation of secondary products (e.g. gibbsite, goethite, anatase). In contrast, slower and less abundant water promotes the diffusion of solutes and dissolution of  less weatherable minerals, and the absolute accumulation of minerals (e.g. silica, carbonate, sulfate, or oxyhydroxide of Mn) by precipitation of moderately mobile elements (e.g. Si, Fe, Ca, Mg, Mn) secondary minerals with infilling textures (Figure 3.1). The relationship between flow and the mineral weathering patterns is summarized in Figure 3.2. For instance, the mineralogical study of hardpans and cemented layers in tailings reported that layers with low saturation of water and capillary effects were associated with the accumulation of secondary mineral precipitation resulting in decreased porosity (Elghali et al., 2019b; Graupner et al., 2007; Redwan et al., 2012). 55   Figure 3.1. Schematic infilling (up) and boxwork (down) textures of secondary minerals precipitated in a primary mineral (adapted from patterns of glomero (up) or septo (down) alteromorphs presented in Delvigne, 1998). Infilling textures are agglomeration of small compact areas or homogeneous layers of secondary minerals. Boxwork or septo textures are regular or irregular banded partitioned secondary minerals along mineral cleavage planes or fissures. The main characteristics of absolute and relative accumulations are summarized in relation with infilling and septo textures (Delvigne, 1998).  56   Figure 3.2. Relationship between water regime and the weathering patterns of minerals (adapted from Velbel, 1985; Lasaga et al., 1994; Delvigne, 1998; Taylor and Eggleton, 2001; Colombani, 2008). Previous studies at Antamina have mainly focused on the hydrological and geochemical aspects (Peterson, 2014, Blackmore et al., 2014 and 2018a). They proposed that flow from large experimental waste-rock piles as well as column kinetic tests were unsaturated and not uniform and that solutes were controlled by a multi-porosity system (Blackmore et al., 2014, 2018a). The mineralogical composition of the waste rock at Antamina has previously been investigated, although little information was gathered on the mineralogical textures related to weathering. This study aims to: i) better qualify the weathering patterns and composition of minerals in association with hydrological systems characterized by tracer tests; ii) investigate the relationship between flow paths, mineralogical composition, weathering patterns, and waste-rock drainage chemistry in long-term kinetic experiments. The qualitative description of mineralogical weathering patterns 57  will be combined with tracer tests results, indicative of water flow regime, as an initial evaluation of the impact of hydrological processes on drainage chemistry. Finally, numerical modelling will compare the impact of these processes occurring in field barrels of different compositions on the mineral reactivity and long-term drainage compositions.  3.2 Methods 3.2.1 Kinetic testing: field barrel experiment and sampling Seventeen field barrels of different waste rock composition were installed at the Antamina mine (Table D1). Each field barrel consisted of a 55 gallon opened plastic drum containing about 300 kg of waste rock. The field barrels, located adjacent to the waste dumps, were thus subjected to the same weather and atmospheric conditions. Each field barrel contained distinct waste rock type of skarn, intrusive and marbles  representative of the main lithological classes excavated onsite. The field barrels were exposed to the seasonal variations in temperature and precipitation that characterize the Antamina mine, which has distinct wet and dry seasons with 80% of precipitation usually occurring between October and April, a mean annual precipitation of 1200 mm, and mean annual temperature of 5.6 oC. The flow regime was first estimated using the proportion of fine grain size portion < 2 mm of the grain size distribution (Figure D2), corresponding to be soil like behaviors and supporting matrix flow when greater than 20% (Smith and Beckie, 2003). For the interpretation of the results, it was considered that all field barrels were exposed to the same climatic conditions (i.e., temperature, precipitation) but with different grain size distribution resulting in different components of flow regimes (i.e. portions of fast preferential versus slow matrix flow). 58  3.2.2 Waste rock and drainage elemental analysis  The initial solid phase elemental composition of the waste rock was analyzed from a quartered composite sample using whole rock 4 acid digestion (i.e., hydrochloric, nitric, hydrofluoric, and perchloric acids) and inductively coupled plasma mass spectrometry (ICP-MS) for metals, as well as solid phase S and C by Leco Furnace at ALS commercial laboratory in Peru. Drainage from the field barrels was collected in 20L containers in which the volume of water was regularly measured in order to provide estimates of volumetric flow rates. Monthly composite water samples were collected from the 20L containers over a period of 7 years. The containers were emptied after each sampling except for 5L that was kept as part of the composite sample. The samples were acidified and filtered upon collection, sent to ALS commercial laboratory in Peru, and analyzed for sulfate and trace metal concentrations with ICP-MS. For all above mentioned analyses, triplicate samples were taken and reference standards analyzed for quality control. 3.2.3 Waste-rock mineralogical analysis   Grab samples of waste rock (~1 kg) were collected from bags of waste rock used in the construction of the field barrels and kept in controlled conditions to characterize the initial elemental abundances before weathering (i.e. the “initial” waste-rock composition). In addition, grab samples (~1 kg) of waste rocks were collected from the top and bottom of the field barrels (Figure D3) after 7 years of weathering (i.e. weathered waste rock). The most reactive portion of waste rock, represented by the silt and clay (< 2mm) size fractions (Strömberg and Banwart, 1999), was isolated by sieving. Subsequently, the dried and sieved waste-rock samples were mounted with epoxy resin and polished to thin sections using a non-aqueous cooling lubricant to minimize potential phase alteration caused by dissolution and/or heat and carbon coated. The weathered and 59  initial waste-rock elemental abundances were analyzed with an Olympus handheld Field portable X-Ray fluorescence spectrometer (FP-XRF). A reference standard no 316 was used for calibration.    The quantitative composition, liberation and association of minerals were analyzed using a Thermo-Fisher Mineral Liberation Analyser (MLA), equipped with a FEI Quanta 600 scanning electron microscope (SEM), dual Bruker-AXS silicon-drift energy dispersive X-ray analysis system (EDS), and the Thermo-Fisher MLA software. The MLA extended backscatter electron method was applied with a frame resolution set at 1.06 μm/pixel. The mineral standards library compiled by Blaskovich (2013) was used for X-ray pattern phase assignment, and mineral association and liberation textures were interpreted using the MLA DataView® software. The elemental content data from MLA were reconciled with FP-XRF measurements and the linear regressions were > 83% for Fe, Cu and Zn (Figure D4). 3.2.4 Drainage concentration and loads   The mass loading per pore volume (mg/kg per pore volume) correspond to the solute concentration (mg/L) multiplied by the outflow volume (L) representing the pore volume in unsaturated conditions divided by the waste rock mass (kg) (Neuner et al., 2013).  The drainage composition averaged over the first and last years were considered in order to represent the behavior of initial and weathered waste rocks, respectively. They were then compared to the mineralogical textures observed from the initial and weathered grab samples collected in the field barrels. The average concentration and loads of the drainage over the first year of the experiment was considered in order to assess the reactivity of initial waste rock before weathering (Tables D2 and D3). In turn, the average from the last year of the experiment was selected to account for the reactivity of weathered waste rocks when samples were collected from the field barrels and tracer tests performed, after 7 years of weathering (Tables D2 and D3).  60  Solute concentrations were considered an instantaneous measure of the impact of flow on the contact time with the mineral as well as the solubility controls. Comparatively, the mass loading corresponds to solute concentrations over volumetric flux which demonstrate the combined effects of concentration and volume of pore water over time. The conceptual modal processes summarized from the literature are illustrated in Figure 3.3, where the flow regimes influence the dissolution or precipitation of soluble and less soluble minerals in turn impacting solute concentrations, mineral weathering, and elemental depletion (Velbel, 1985; Lasaga et al., 1994; Delvigne, 1998; Taylor and Eggleton, 2001; Colombani, 2008; Neuner et al., 2013; Peterson, 2014; Blackmore et al., 2014). For instance, high solute concentrations or loads are the result of slower velocities, longer residence time, and slower reactions reaching equilibrium. In contrast, low solute concentrations or loads result in partial equilibrium, faster velocities, shorter residence time, and faster reactions.    3.2.5 Tracer test The application of a tracer composed of LiCl was selected to characterize fast-flow regions associated with higher-rates recharge events characteristic of the wet season at Antamina (Blackmore et al., 2018a). The test was performed to collect data on the hydrological behavior of five selected field barrels (i.e. 1A, 2A, 2B, 2C, and 3A). This test provided information on mass recovery and residence time in the field barrels. A volume of 2.5 L of 1 g/L LiCl solution was sprayed at the surface of the field barrels to simulate a 10 mm rainfall at an intensity that approximates a 2 to 5-year return period at Antamina (Golder Associates, 1999). Instantaneous samples were collected at least every hour for the first 48 hours during the day and composite samples (i.e. bi-weekly to monthly collections) were analyzed for chloride with ion 61  chromatography at the ALS commercial laboratory in Peru. The samples were taken in triplicate as part of the quality controls of sampling and analyses. The transport behavior of the tracer tests was investigated based on breakthrough curves and temporal moment analysis using the flow-corrected time approch to allow comparison between field barrels as described in supplementary information (Eriksson et al., 1997). This method calculates the proportion of preferential and fast matrix flow (%) indicative of fast-flowing water and the remainder corresponds to the proportion of slow moving matrix flow or immobile domain (%) of the total water content (Eriksson et al., 1997; Blackmore et al., 2014).          Figure 3.3. Relationship between flow paths, mineral dissolution, solute concentration, and mineral depletion (based on Velbel, 1985; Lasaga et al., 1994; Delvigne, 1998; Taylor and Eggleton, 2001; Colombani, 2008; Neuner et al., 2013; Peterson, 2014; Blackmore et al., 2014).  3.2.6 Mineralogical data This study focuses on a select number of primary minerals as well as the oxidation products of major sulfide minerals within the waste rock. These include pyrite, chalcopyrite, and sphalerite 62  as well as their oxidation products of secondary Fe-, Cu-, Zn-, and S-bearing minerals, respectively. The primary and secondary minerals or aggregates were identified with MLA and categorized into different groups: sulfides, oxides, Fe-hydroxy-sulfates, other oxides, carbonates, sulfates, silicates, and phosphates (Table D4). From these groups of minerals the secondary phases included both secondary and tertiary minerals, which precipitated before or after sampling as defined by Blowes, et al., 2003.  The mineralogical data and weathering patterns from polished thin sections of the initial and weathered waste rocks were investigated using an optical microscope. This study considered the concentration of moderately mobile precipitates (e.g. silica, carbonate, sulfate, or oxyhydroxide of Mn) through absolute accumulation resulting in weathered minerals with infilling textures (adapted from Delvigne, 1998, Figure 3.1).  In contrast, the concentration of less mobile elements/minerals (e.g. goethite, gibbsite, anatase) from relative accumulation are associated with boxwork textures of weathered minerals (adapted from Delvigne 1998, Figure 3.1). Infilling textures therefore refers to the agglomeration of secondary minerals into small compact areas or homogeneous layers (Figures 3.1 and D5A). Whereas, boxwork or septo textures show the distribution of regular or irregular banded partitioned secondary minerals along mineral cleavage planes or fissures (Figures 3.1 and D5B). The proportion of micromorphological textures resulting from weathering of minerals was estimated by following the point counting procedure. The constituents of each polished thin-section were estimated by random observations made over its total area (based on Jones, 1987).  The point counting of mineral textures was done using pictures from the scanned surface of polished thin sections and with the help of the Jmicrovision software (Roduit, 2007). To assure the appropriate statistical representation of the textures in each sample, a minimum of 300 points were randomly distributed at the surface of the thin-sections. 63  Consequently, the point counting was terminated when the evolution plots of the different classes were showing steady lines (Roduit, 2007).  3.2.7 Reactive transport modeling The evolution of drainage chemistry from selected field barrel was simulated using the reactive transport code PHREEQC (version 3.33) (Parkhurst and Appelo, 2013) with the modified WATEQ4F thermodynamic database (Ball and Nordstrom, 1991). The objective was to calibrate the simulation with the drainage results from the field barrels and evaluate the differences observed between the calibrated effective kinetic rate constants of the main reactive minerals. A one-dimensional model was used to describe the temporal evolution of the chemistry of drainage infiltrating through field barrels of: 1) intrusive (i.e. 2B and 2C) and skarn (i.e. 3A) composition; 2) field barrels of different composition and dominant matrix flow regimes (i.e. 2C and 3A); or 3) intrusive field barrels with either dominant preferential or matrix flow (i.e. 2B and 2C). Transient conditions were induced by space-time discretization incorporating the recorded length of the wet and dry season of those field barrels from 6 to 7 months. The space-time discretization was also adjusted based on the proportion of the preferential and matrix flow paths, calculated from the flow-corrected results of the tracer test, as presented in Appendices C and D (Table D5). Faster mineral dissolution/precipitation reactions, such as calcite or gypsum, were simulated as equilibrium phases. Conversely, kinetically limited minerals were described with first order kinetic reactions (i.e. pyrite) and second order kinetic reactions (i.e. chalcopyrite and sphalerite) based on equations from the literature Williamson and Rimstidt, 1994; Acero et al., 2009; Pan et al, 2012. The effective kinetic rate constants were calibrated in order to best fit the results of drainage pH from the field barrels. Further boundary conditions and model details are described in the Appendix C. The pH and dissolved Fe, Cu, Zn, and S from the oxidation of pyrite, 64  chalcopyrite, and sphalerite were targeted for the simulations. The aqueous chemistry of the field barrel drainage was compared to the modelling results graphically, and by calculating the normalized root mean square error (NRMSE). In addition, the dimensionless Damköhler number, which divides fluid mixing and chemical reaction timescales, was calculated to evaluate the competition between transport versus reaction rates within this system (Maher and Mayer, 2019). Subsequently, the residence time, measured in days, was calculated from the tracer tests and divided by the time, in days, to reach a depletion of 0.1 mole of pyrite in the simulation.  3.3 Results and discussion 3.3.1 Field barrel material characteristics Samples from five field barrels composed of marble (1A), intrusive (2A, 2B and 2C) and skarn (3A) were selected to represent various waste-rock composition and flow regimes. The lithological classes and initial neutralizing potential ratio (NPR) overview of the bulk waste rock in the five selected field barrels (1A, 2A, 2B, 2C, and 3A) are given in Table D1. The mineralogy of field barrel 1A was composed of 53-60% carbonate with 1-2% sulfides; 2A of < 3% carbonate with < 1% sulfides; 2B and 2C of 8-20% sulfides and < 2% carbonates; and 3A of 4-6% sulfides and 20-25% carbonates (Figure D1). The water flow regime was first estimated in the field barrels accordingly to the proportion of < 2mm grain size portion of the particle size distribution greater or inferior to 20% indicative of soil versus rock like behaviors, respectively (Smith and Beckie, 2003). For the field barrels 1A and 2A, the < 2 mm content was lower than 20%, whereas for 2B, 2C and 3A it was greater than 20% (Figure D2). Therefore, field barrels 1A and 2A were associated with ''rock-like'' preferential flow behavior and 2B, 2C, 3A with “soil-like” matrix flow behavior. This classification was then compared to the water regime results deduced from the tracer tests. 65  3.3.2 Tracer tests The solute transport from the field barrels were analyzed based on the breakthrough curves from the normalized Li-chloride tracer test results (i.e. chloride concentration (C)/Initial chloride concentration (Co)). All the field barrels show the rapid arrival of the first temporal moment of the chloride tracer within a few hours after application (Figure 3.4). The intensity and/or concentration of the initial tracer peak was more pronounced for field barrels 1A, 2A, and 2B suggesting dominant preferential or fast matrix flow component. This initial peak was then followed by a long steady increase and tailing suggesting a slow matrix flow or immobile component. In contrast, the initial intensity and/or concentration of the tracer was less pronounced for field barrels 2C and 3A and was followed by a stronger subsequent increase of tracer concentration suggesting a dominant component of slow matrix flow. These results are comparable to bromide tracer tests showing small preferential flow component with dominant high velocity matrix flow-paths from column tests filled with waste rock from Antamina (Blackmore et al., 2014).       Figure 3.4. Normalized chloride breakthrough curves for field barrels 1A, 2A, 2B, 2C, and 3A showing evidence of preferential and matrix flow paths. The dashed lines correspond to period of time with missing samples. The arrow indicates the location of the initial breakthrough after tracer application. 66   The degree of mobility of pore water was calculated from temporal moment results of the Li-chloride tracer test observations using the flow-corrected time approch (Eriksson et al. 1997) as described in supplementary information and presented in Table 3.1 and D8. The results allow comparison of the proportion of mobile preferential/fast matrix flow versus slow matrix/immobile flow between different field barrels. The degree corresponding to fast mobile water showed similarities with the estimated values from the proportion of grain-size < 2 mm (Table 3.1) as well as the breakthrough curves (Figure 3.4). For instance, the field barrels 1A and 2A with a rock-like behavior, had higher proportions of mobile water of 65% and 32% which were most likely associated with preferential or fast-matrix flow. However, field barrels 2B, 2C, and 3A with soil-like behavior, presented lower proportion of fast mobile water between 3% and 19%, therefore corresponding to higher proportion of slower matrix/Immobile flow (Table 3.1). In a previous study, columns filled with a composite of Antamina waste-rock similar to field barrels 1A, 2A, and 2C had comparable calculated mobile pore-water ratio between 40 to 58% (Blackmore et al., 2014). In addition, despite similarities in composition, the intrusive field barrels 2A, 2B, and 2C showed 3 distinct degrees of mobility of pore-water. The higher proportion of faster flow in 2A is consistent with its lower proportion of < 2 mm particle size fraction and rock-like classification. The higher proportion of mobile water in field barrel 2B compared to 2C, despite similar proportion of < 2 mm particle size fraction between them, might be explained by higher proportion of coarser grain-size distribution > 20 mm (Figure D2). Based on these observations, the impact of water regime on mineralogical patterns of weathering and drainage composition is investigated in the following sections.   67  Table 3.1. Degree of mobile pore water (%) corresponding to mobile preferential and fast matrix flow (v) and slow matrix flow/immobile (1-v) pore water in each field barrels calculated from temporal moments of tracer test observations. Field barrel Lithology < 2mm  Fraction % Preferential/Fast matrix flow % (v) Slow matrix flow % (1-v) 1A* Marble 16 65 35 2A Intrusive 15 32 68 2B Intrusive 30 12 88 2C Intrusive 28 3 97 3A Skarn 28 19 81       * Averaged results from 2 field barrels (1B and 1F) of similar composition 3.3.3 Mineralogy The relationship between flow paths and the weathering patterns of minerals observed in the field barrels were adapted from the classification of morphological textures of alteromorphs, resulting from the weathering and alteration of minerals identified by Delvigne (1998) which was previously summarized in introduction. According to this classification, the residual and secondary products, observed in this study, were divided into relative or absolute accumulation of minerals, resulting respectively from the concentration of generally less mobile elements (e.g. Al, Fe, Ti) or moderately mobile elements (e.g. Si, Ca, Mg, Mn) (Figure 3.1) at pH>5 with Fe substrate (Smith, 2007). The minerals associated with relative accumulation were groups of more stable oxidation products such as oxides, Fe-Oxi/hydroxyl-Sulfates, sulfates (e.g. alunite or jarosite), silicates, and phosphates (e.g. apatite, goyazite, or monazite). However, the minerals associated with absolute accumulation were part of less stable groups such as carbonates, sulfates (e.g. celestine, goslarite, gypsum), and phosphates (e.g. fluorite; the selected individual minerals or phases are shown in 68  Table D4). The mineralogical composition from field barrels 1A and 2A only had <2 wt% secondary phases mainly composed of silicates. Field barrels 2B and 2C were composed of up to 10 wt% oxidation products such as Fe-oxy/hydroxyl-sulfates, silicate, and sulfates. Field barrel 3A had <4 wt% oxidation products composed of silicate, sulfate, and Fe-oxides (Figure D1B). In the weathered samples in contrast to the initial samples, the proportion of more soluble secondary products (i.e absolute accumulation) generally increased as the secondary minerals associated with less soluble minerals (i.e. relative accumulation) generally decreased after weathering (Figure 3.5).   Microscopic mineral textures indicative of weathering and flow effects were observed in thin sections from samples of field barrels. The inventory of those textures, by point counting in thin sections, were represented by infilling or boxwork/septo textures (Figure 3.1) which are the product of low or high drainage environment, respectively (Delvigne, 1998; Figure 3.2). Examples of infilling textures were observed in thin section of field barrels 1A, 2A, and 3A; boxwork/septo textures were observed in 1A, 2B, and 3A; whilst a combination of both was observed in 2A, 2B, and 3A (Figure D5). The boxwork/septo textures counts were up to 52% in field barrels 1A and 2A, while up to 83% infilling textures were dominating in field barrels 2B, 2C, and 3A (Figure 3.6). Independently of the composition of primary minerals, high drainage boxwork/septo textures were dominant in field barrels 1A and 2A which were mainly composed of stable secondary minerals (Figures D1A, D1B, and Table D4). Instead, predominantly low drainage infilling textures were observed in field barrels 2B, 2C, and 3A with more-abundant, less-stable secondary minerals and independent of the primary mineral compositions (Figures D1A, D1B, and Table D4). In addition, increased sulfate content in the weathered samples compared to the initial samples (Figure D1B and Table D4) was indicative of absolute accumulation of moderately mobile minerals (Figure 3.2). These features suggest a relationship between the degree of mobility of 69  water with mineralogical content and weathering patterns observed in the field barrels as supported by previous studies (Colombani, 2008; Delvigne, 1998; Kuechler et al., 2004; Lasaga et al., 1994; Stoops et al., 1979; Taylor and Eggleton, 2001; Velbel, 1985). This relationship will be investigated by comparing these observations on mineral composition and textures with results from tracer tests presented in the following section. Figure 3.5. Proportion of secondary minerals associated with relative and absolute accumulation (i.e. less or more soluble minerals, respectively) from initial (I) and weathered (W) waste-rock samples of field barrels.      Figure 3.6. Point counting modal distribution of boxwork and infilling textures associated with residual and secondary minerals from initial (I) or weathered (W) waste-rock samples of field barrels. 70  3.3.4 Comparison of water regime and mineralogical weathering patterns Preferential and fast-matrix flow most likely promoted faster and more abundant water circulation, preferentially leaching more mobile elements and conserving immobile elements, which resulted in the relative accumulation of weathered minerals with boxwork textures (Figures 3.1 and 3.2; Delvigne, 1998). In contrast, slow-matrix flow was associated with slower circulation and less abundant water, which promoted absolute accumulation of moderately mobile elements and more soluble minerals with infilling textures of weathered minerals (Figures 3.1 and 3.2; Delvigne, 1998). As a result, preferential and fast-matrix flow lead to less leaching, because of faster water flow and less contact with waste rocks, whereas slow-matrix flow promotes even and more complete leaching because of longer contact time. To illustrate this relationship between flow, mineral composition, and patterns of weathering the proportion of preferential/fast-matrix flow and slow-matrix flow were compared to the composition of secondary products and distribution of weathering textures. The dominant flow paths were determined with the temporal moment of tracer tests (Table 3.1). The selected secondary phases were calculated from the proportion of residual (i.e. less soluble) and absolute (i.e. more soluble) secondary products over the sum of soluble and less soluble phases (Figure 3.5). The distribution of weathering textures (i.e. boxwork vs infilling) were semi-quantified by point counting in thin sections (Figure 3.6). The results compiled in Table 3.2 show that the mineralogical weathering patterns (Figure 3.6) and residual or absolute accumulation distribution of secondary products (Figure 3.5) were present in similar proportions compared to the proportion of preferential/fast-matrix vs slow-matrix flow paths (Table 3.1) with standard deviations between 4 to 15%.   71  The weathering patterns of minerals and secondary mineral content/composition were consistent with the degree of preferential/fast-matrix or slow-matrix flow paths calculated from the tracer tests (Table 3.2) as well as the water regime previously inferred from the particle size distributions (Figure D2 and Table 3.1). Field barrels 1A and 2A had higher degrees of preferential or fast matrix flow, and were associated with higher boxwork/septo textures and absolute accumulation of more soluble secondary minerals. Field barrels 2B, 2C, and 3A had higher degree of slower matrix flow with elevated infilling textures and relative accumulation of less soluble secondary minerals. In addition, the degree of mobility, patterns of weathering observations, and secondary mineral content were representative of the water regime projected from the particle size distributions. The observations from the field barrels were consistent with observations from the literature associating the effect of water and flow with the secondary mineral products content and/or weathering patterns (Velbel, 1985; Lasaga et al., 1994; Delvigne, 1998; Taylor and Eggleton, 2001; Kuechler et al., 2004; Colombani, 2008). In summary, the morphology and textures in primary and secondary minerals that were developed through weathering were governed by preferential/fast-matrix and slow-matrix flow paths, and the link with water chemistry will be investigated next.        72  Table 3.2. Distribution of weathering textures of minerals (boxwork/septo Vs Infilling) as well as residual and absolute accumulation of oxidation products compared to the degree of mobility of water (i.e. preferential/fast-matrix or slow-matrix flow paths). The boxwork/septo textures and relative secondary mineral content, were compared to the degree of faster preferential/fast-matrix flow. The infilling textures and absolute secondary mineral content, were compared to the degree of slow-matrix flow.  Field Barrel Preferential/Fast-Matrix flow Slow-Matrix Flow Degree of mobile pore water Boxwork/ Septo texture  Relative accumulation of secondary minerals Standard deviationDegree of  slow or immobile pore water Infilling texture Absolute accumulation of secondary minerals Standard deviation1A 65% 40% 38% 12% 35% 60% 62% 12% 2A 32% 52% 45% 10% 68% 48% 55% 10% 2B 12% 26% 42% 15% 88% 74% 58% 15% 2C 3% 17% 32% 15% 97% 83% 68% 15% 3A 19% 23% 15% 4% 81% 77% 85% 4% 3.3.5 Drainage quality The elemental concentrations and mass loading in the drainage extended over several orders of magnitude for Cu, Fe, and Zn (Figure 3.7) and were reflective of the wide range of waste-rock lithologies and neutralization potential ratios (NPR) in the field barrels (Table D1). The field barrels 1A and 2A had the lowest concentration of Fe (up to 2 wt%), Cu (<1 wt%), and Zn (<1 wt%) in waste rock. In contrast, field barrels 2B and 2C had the highest concentration of Fe (up to 13 wt%) and field barrel 3A had the highest concentration of Cu (up to 4 wt%) and Zn (up to 10 wt%) in the waste rock (Figure 3.7A). The concentration of Zn, Cu, or Fe in the drainage were also higher in the same field barrels (up to 2300 mg/L) after 7 years of weathering and at low pH between 2.5 and 4.5 (Figure 3.7B). The average mass loading for sulfate, Cu, Zn, and Fe were in 73  general orders of magnitude higher in field barrels 2B and 2C (up to 151 mg/kg) compared with field barrels 1A, 2A, and 3A (up to 0.02 mg/kg) (Figure 3.7C). The high solid-phase concentrations of Cu and Fe were proportional to the average drainage concentration of 2B, 2C, and 3A and mass loading for 2B and 2C. In addition to the high waste-rock concentration and acidic pH, 58 to 85% of the sum of soluble and less soluble selected secondary phases were associated with the absolute accumulation of minerals composed of more mobile elements such as: gypsum, ankerite, siderite, and amorphous Fe-sulphate (Table 3.2 and Figure D6). These conditions might have promoted the potential for Fe-, Cu-, or Zn-mobility in the drainage of field barrels 2B and 2C. In contrast, neutral pH and 38-45 wt% of the  selected oxidation products were associated with the relative accumulation of minerals composed of less mobile elements such as amorphous Fe-oxi-hydroxysulfates or Fe-silicates (Table 3.2 and Figure D6), which might have decreased the potential mobility of elements in field barrels 1A and 2A.   3.3.5.1 Comparison of Field barrels 2B, 2C, and 3A Despite the higher metal content, slightly lower NPR, and similar fractions of particle size distribution < 2mm (Table D1, Figures D2 and 3.7A), slower development of acidic conditions and lower Fe drainage concentration and mass loading were observed in field barrel 2B compared with 2C (Figure 3.7). This could be attributable to a coarser grain size > 2mm (Figure D2) resulting in higher proportion of fast preferential/fast-matrix flow path (i.e. 12% in 2B versus 3% in 2C, Table 3.1) and decreasing contact time with lesser leaching of dissolved species. In addition, field barrel 2B had 10% more secondary products associated with relative accumulation of less soluble minerals (Figure 3.5 and Table 3.2) such as amorphous Fe-oxi/hydroxysulfates and Fe-silicate than 2C (Figure D6). In contrast, field barrel 2C had a higher proportion of slow-matrix flow paths (i.e. 97%, Table 3.1) promoting higher contact time and even leaching of minerals. This is supported 74  by 68% absolute accumulation of less stable minerals (i.e. content of more soluble minerals over the sum of soluble and less soluble minerals; Figure 3.5 and Table 3.2) such as gypsum, ankerite, and siderite (Figure D6). Lastly, despite higher zinc solid phase concentrations, the drainage concentration and loads in 3A were lower than those in field barrels 2B and 2C (Figure 3.7). This could be attributable to circumneutral pH between 7 and 8 (Figure 3.7B) and that 25% of secondary minerals are non-soluble Zn-bearing minerals such as amorphous Fe-silicates and Ca-sulfates aggregates (Figure D6). Secondary Zn-minerals precipitated in weathered Antamina waste rock in similar conditions were observed in larger scale experimental piles or full-scale waste rock pile in Chapter 4 (Vriens et al., 2019a; St-Arnault, 2019b) and other mine sites (Davies et al., 2011; Gieré et al., 2003; Smuda et al., 2007). In addition, the possible desorption of Zn from Zn-bearing phases under circumneutral pH, resulting in effective mobilization of Zn (Roberts et al., 2003; Smith, 2007) were observed from leachate test of samples from the full-scale waste rock pile at Antamina mine as observed in Chapter 4 (St-Arnault, 2019b). In summary, the intrusive field barrel 2A, with lower waste-rock metal content and NPR as well as 32% dominant preferential/fast-matrix flow paths, resulted in relative accumulation of 45% minerals composed of less mobile elements and 52% boxwork/septo weathering texture. After 7 years, the drainage in field barrel 2A had neutral pH, lower concentrations and loadings of Fe, Cu, and Zn than field barrels 2B and 2C. In contrast, higher waste-rock metal content and low NPR of the intrusive field barrels 2B and 2C, with up to 97% dominant slow-matrix flow, resulted in dominant absolute accumulation of up to 68% minerals composed of more mobile elements and up to 83% infilling weathering texture. The corresponding final drainage from those field barrels showed acidic pH with higher solute concentrations as well as mass loading of Fe, Cu, and Zn. Finally, field barrel 3A composed of skarn with high waste-rock metal content, high NPR, and 75  81% dominant slow-matrix flow, lead to proportionate absolute accumulation of 85% secondary minerals composed of more soluble elements, and 77% infilling weathering texture. The drainage from field barrel 3A, after 7 years of weathering, remained near neutral pH with low concentration or mass loading of Fe, Cu, and Zn, which could be in part due to less soluble primary minerals. The proportions of preferential/fast-matrix and slow-matrix flow paths in the waste-rock field barrels influenced the metal concentrations and loads in drainage. The degree of mobility of water inhibited or accelerated water-rock interaction, therefore releasing or retaining solutes based on the dissolution or precipitation of absolute or relative oxidation products. The impact of flow on the water-rock interaction was also reflected by proportional counts of weathering textures and mineralogical content representative of the dominant flow regime observed in the field barrels. When preferential/fast-matrix flow is dominant, lower depletion may result from the decrease in contact time and when slow-matrix flow is dominant, higher depletion may result from increased contact time between water and waste rocks (Blackmore, 2015). The weathering rate of minerals in selected field barrels will be estimated next using calibrated drainage simulations in order to further assess the relationship between these processes.  76                          A                                B                              C         Figure 3.7. The solid-phase content of Cu, Fe and Zn (wt%) as measured with FP-XRF (A); the average drainage concentrations (mg/L) (B) and average mass loading (mg/kg) (C) of the last year from weathered (W) field barrels of marble (1A), intrusive (2A, 2B and 2C), and skarn (3A) composition.   77  3.3.6 Drainage simulation The mineralogical content and the degree of mobility of water calculated from tracer tests were used to simulate chemistries with reactive transport modal simulations in PHREEQC. The reaction rate constants of minerals were calibrated to achieve the best fit between measured and simulated chemistry for 3 selected field barrels 2B, 2C, and 3A which are presented in Figure 3.8. Relatively good agreement was achieved between simulations and measured drainage pH in all field barrels. Despite orders of magnitude differences between measured and simulated chemistry, the simulations captured the general and long-term trends and timing of concentration ranges and peaks for sulfate, Fe, Cu, and Zn (Figure 3.8). Zinc concentrations were underestimated for all field barrels, and for Cu and Fe in field barrel 3A. Dissolved sulfate concentrations were slightly overestimated in field barrels 2C and 3A as well as Cu and Fe in field barrel 2C (Figure 3.8). Low normalized errors support good agreement between simulated and measured values except for Fe or Cu in 2B or 2C (Table D7). Not accounting for dis-equilibrated chemistry of aqueous solutions, missing a Fe- or Cu-bearing minerals such as bornite or covellite, the presence of an immobile domain, or averaging composite drainage samples might be the cause of these discrepancies observed between simulated and measured drainage chemistry. Also, overestimation of calibrated sulfate, Cu and Fe curves in field barrel 2C might be indicative of or an overestimation of the calibrated effective rate coefficients and weathering rates of pyrite and chalcopyrite. In future study, the application of a more complex reactive transport model could better capture the subtleties of solute concentrations evolution impacted by flow (Maher and Mayer, 2019).   78  Figure 3.8. Comparison of measured drainage chemistry (points) from field barrels 2B, 2C, and 3A with the simulated drainage chemistry (lines). The x axis labels for all barrels and y axis labels for SO4, Cu, Fe, and Zn are the same for all frames. Reactive transport simulations incorporated the effect of geochemical and hydrological components on calibrated effective dissolution rates of minerals. The calibrated effective dissolution rates of pyrite, chalcopyrite and sphalerite presented in Table 3.3 and differences with dissolution rates from the literature (Williamson and Rimstidt, 1994; Acero et al., 2009; Pan et al, 2012 referenced in Table D6) implicitly reflected a mineral-specific correction for mechanisms that might alter mineral dissolution rates (i.e. water-rock interaction, reactive surfaces, and/or 79  mineral association). Field barrel 3A had the highest calibrated effective rate coefficients for pyrite and sphalerite followed by 2B and 2C (Table 3.3). Whereas, field barrel 2C had the highest calibrated effective rate coefficients for chalcopyrite followed by field barrels 3A and 2B (Table 3.3). The calculated Damköhler number (Da) reflected the relationship between transport and reaction (Tables 3.3 and D8); in which a low Da corresponds to a fast-flowing system with minimal time to react and vice versa (Maher and Mayer, 2019). The lower Da of 0.003 and 0.007 from field barrels 2B and 3A, respectively compared with higher Da of 0.012 for field barrel 2C (Table 3.3), also corresponded with a higher proportion of matrix flow (Table 3.1) and faster pyrite weathering rate (Table 3.3) in field barrel 2C. Field barrels 2B and 3A, had lower Da, higher degree of preferential/fast-matrix flow and boxwork/septo textures as well as relative accumulation of secondary mineral composed of less mobile elements for 2B, in contrast to field barrel 2C (Figure 3.9).   Figure 3.9. Relationship of Damköhler number with degree of mobile porewater (i.e. preferential/fast matrix flow), septo weathering textures, and secondary minerals composed of less mobile elements.     80  Table 3.3. Initial mineral content, weathering rates of the minerals in the simulation, calculated Damköhler number (Da), and calibrated effective rate coefficients for Fe, Cu, and Zn as pyrite, chalcopyrite and sphalerite in the studied field barrels. Calibrated effective rate coefficients (Calibrated keff; different units) were obtained by calibrating drainage chemistry simulations to pH. Mineral Initial Content Mo  Rate Coefficient (Keff) Weathering Rate Damköhler Number  (Da) Pyrite Mol L-1 Log [s-1] Mol L-1Day-1 (-) 2B  2.88 -7.75 3.21E-05 0.004 2C  1.76 -6.75 4.33E-04 0.070 3A  0.15 -8.5 3.88E-05 0.003 Chalcopyrite Mol -1 Log [s-1] Mol L-1 Day-1  2B  2.6 -3.5 3.01E-04  2C  1.3 -4.5 2.12E-04  3A  0.3 -4 1.06E-10  Sphalerite Mol L-1 L Mol-1 s-1 Mol L-1 Day-1  2B  0.015 2.5 6.30E-07  2C  0.004 1 2.31E-07  3A  1.22 6 5.68E-06  Despite lower mineral content and similar NPR (Table D1), the field barrel 2C had faster simulated rate coefficients and sulfide weathering rates (Table 3.3) than 2B, which may be the result of longer residence times (Table 3.1). A similar discrepancy was also observed for pyrite between field barrels 3A and 2B, where 3A, with 81% slow-matrix flow (Table 3.1) and neutral pH (Figure 3.7), displayed similar weathering rates than field barrel 2B with 88% slow-matrix flow and acidic pH (Table 3.3, Figure 3.8). Accelerated depletion rates linked to dominant slow-matrix flow were also observed at small- and large-scale experiments at Antamina mine (Blackmore, 2015; Peterson, 2014). The faster weathering rates of chalcopyrite compared with slower rates of pyrite in field barrel 2B (Table 3.3), despite higher modal compositions and faster flow regime within field barrel 2B compared with 2C, indicate that other factors may also alter 81  mineral dissolution rates. For instance, several experimental studies and previous observations from Chapter 2 demonstrated that mechanisms such as surface liberation of minerals or galvanic interactions may inhibit or accelerate mineral depletion (Chopard et al., 2017; Lapakko, 2015; Parbhakar-fox et al., 2013; Pedretti et al., 2017; St-Arnault et al., 2019a). In summary, those site-specific calibrated dissolution rates, captured important distinction between and within material classes as well as partially based on the flow regime, and might also indicate that other mechanisms affect the mineral reactivity. In addition, the Da, reflecting the relationship between transport and reactivity between simulation and tracer tests results, was also consistent with the proportionality that was previously observed between tracer tests results and other indicators. Effectively, the dominant flow paths were proportionate to the content of secondary mineral products and patterns of mineral weathering which were reflected in the drainage composition of the field barrels.   3.4 Conclusion The evolution of drainage chemistry, mineralogical content, and patterns of weathering in field barrels, exposed to the natural climatic conditions, provided insights on the development of different types of water regime and waste-rock weathering over a period of seven years. The results show that, in addition to the waste-rock bulk chemistry, the degree of preferential and matrix flow paths affected the simulated mineralogical weathering rates, the observed weathered mineral textures, and the resulting drainage chemistry. Higher proportion of boxwork/septo textures and accumulation of less soluble minerals were consistently observed in marble/intrusive/skarn waste rock with higher proportion of preferential/fast-matrix flow paths and circumneutral drainage with lower concentrations and loadings of Fe, Cu, and Zn. In contrast, elevated proportion of infilling textures and accumulation of moderately soluble minerals were linked to intrusive waste rock with higher proportion of slow-matrix flow path which corresponded to acidic pH drainage and higher 82  solute concentrations as well as mass loading of Fe, Cu, and Zn. The effective mineral oxidation/dissolution rates calibrated from reactive transport model simulations and measured data provide a site-specific correction for the flow regime and mineral reactivity. The simulations also captured the controls of water mobility on calibrated mineral weathering rates from field barrels of similar composition. In summary, the mineralogical and textural assessments of weathering can be integrated with bulk geochemistry and water flow regime data, which is interpreted as a first qualitative indication of processes ultimately controlling the drainage quality. This assessment could be easily applied at any sites to estimate the proportion of flow paths and their controls on drainage chemistry in environment where tracer tests are difficult to implement such as full-size waste-rock piles. The next steps would be to i) quantify the contribution of these individual processes by comparing transport and reaction rates using the Damköhler numbers from drainage results of kinetic tests performed in controlled environment such as humidity cells and ii) investigate controls of flow path on weathering textures of minerals and drainage chemistry in samples taken at the full-scale waste-rock pile.   83  Chapter 4: Geochemical and mineralogical assessment of reactivity in a full-scale heterogeneous waste-rock pile 4.1  Introduction Mining operations produce non-economical waste rock that, once exposed to atmospheric conditions, is prone to weathering. This weathering (i.e. oxidation and migration with infiltrating meteoric water) can release acidity and metals via the drainage effluent. A combination of physical, geochemical and microbial processes control drainage quality, including oxidation kinetics, hydrological pore-water transport, internal drainage mixing, oxygen ingress, microbial catalysis and secondary mineral precipitation (Nordstrom, 2011; Amos et al., 2014; Moncur et al., 2014; Blackmore et al., 2018b). The effects of the physical transport of water and gas on drainage quality are largely related to the effective permeability of waste rock, controlled by particle size distributions and water saturation (Vriens et al., 2019). In turn, geochemical controls on drainage quality include primary sulfide oxidation reactions, acid-buffering reactions from the dissolution of carbonates and silicates, and attenuation mechanisms such as the precipitation of secondary minerals and ion sorption to mineral surfaces (Moncur et al., 2009; Amos et al., 2014; Vriens et al., 2019b). The abovementioned physico- and geochemical processes are often intricately (and non-linearly) coupled. Therefore, the prediction of the long-term drainage quality and timely planning of mitigation measures requires accurate prediction models that include all relevant mechanisms (Lapakko, 2015; Parbhakar-Fox and Lottermoser, 2015; Dold, 2017).  The complexity of drainage quality prediction models is often proportional to the scale of the studied problem, i.e. the waste-rock storage facility. While the properties of the local waste rock can to a large extent be characterized from field samples, drainage quality prediction models 84  to date are typically derived from conventional static testing of bulk waste-rock properties (e.g. its chemical composition, mineralogy, acid-base-accounting (ABA) parameters, or particle size). However, these static tests have limitations, including a poor representativeness of small sample sizes and the difficulty of scaling-up prediction model results to practice-relevant dimensions (Malmström et al., 2000; Dold, 2017). Therefore, complementary (long-term) kinetic testing procedures are increasingly used to evaluate the weathering behavior of waste rock under (variable) field conditions such as observations made in Chapters 2 and 3 (Vriens et al., 2019c; St-Arnault et al., 2019a). These advanced tests better capture long-term waste-rock weathering and drainage quality evolution, but they do not resolve all relevant small-scale heterogeneities or physicochemical interactions (Vriens et al., 2019b). In addition, many waste-rock piles have a highly heterogeneous composition and internal structure (Anterrieu et al., 2010), resulting, for instance, in grain-size sorting from dumping (e.g. upward fining) or compaction due to traffic circulation. Thus, dedicated studies into the heterogeneity and in-situ conditions in large-scale waste-rock piles are necessary to assess the long-term behavior of these complex storage facilities (Ritchie, 1994; Molson et al., 2005; Pedretti et al., 2017; Vriens et al., 2018).  Only few studies have investigated the physical and geochemical characteristics of large-scale waste-rock piles (Lefebvre et al., 2001; Molson et al., 2005). For instance, reactive waste-rock piles of up to 55 m height have been sampled at the surface to investigate the role of secondary mineral precipitation on the overall drainage quality (Gieré et al., 2003; Sracek et al., 2004; Smuda et al., 2007). Further work on large experimental piles has shown that reactive waste-rock fractions as small as <10% of the total pile mass may dominate the overall drainage quality, and that oxygen transport limitations may effectively change drainage quality (Pedretti et al., 2017; Vriens et al, 2019). In the majority of these studies on large-scale systems, the focus has been on physical 85  controls affecting pore-gas distributions (Lefebvre et al., 2001b; Anterrieu et al., 2010; Lahmira and Lefebvre, 2015; Vriens et al., 2019b) or on geochemical and mineralogical processes governing metal(loid) mobility (Gieré et al., 2003; Smuda et al., 2007), but less so on the combination of both (Sracek et al., 2004, Linklater et al., 2005; Vriens et al., 2018). In fact, few waste-rock studies to date have given quantitative consideration to mineralogical and petrographical aspects and their role in large-scale systems (Parbhakar-Fox et al., 2014; Parbhakar-Fox and Lottermoser, 2015; Dold, 2017), even though mineral texture, liberation and arrangement are known to critically determine mineral reactivity as was observed in Chapter 2 (Linklater et al., 2005; Brough et al., 2017; Pedretti et al., 2017; Parbhakar-Fox et al., 2018b; St-Arnault et al., 2019a). For instance, in laboratory settings and small-scale field experiments, secondary mineral precipitation (Vriens et al., 2018), surface passivation (Ghorbani et al., 2013; Fan et al., 2018), and galvanic coupling (Kwong, 1995; Kwong et al., 2003; Parbhakar-Fox et al., 2013; Chopard et al., 2017; Fan et al., 2018) have been directly linked to variable weathering rates. Unfortunately, such mineralogical aspects are challenging to incorporate quantitatively into prediction models to date, because of scarce field data on mineral reactivity and secondary mineral formation, as well as scaling issues related to large, heterogeneous waste-rock piles (Lefebvre et al., 2001; Linklater et al., 2005; Molson et al., 2005; Pedretti et al., 2017).  To that end, in-situ waste-rock samples were collected from two continuous >125 m deep boreholes, while preserving alteration and lithological contacts. Conventional geochemical testing and qualitative analysis of bulk physicochemical properties were combined with advanced quantitative mineralogical analyses and leaching tests to establish a detailed mineral weathering and reactivity profile. The main objectives of this study were to: i) evaluate the effect of physical and chemical heterogeneity on waste rock weathering; ii) characterize reactivity and associations 86  of minerals and their interplay with alteration in reactive zones of the boreholes; iii) assess the impact of heterogeneity and mineralogical factors on the leachate chemistry from the boreholes.    4.2 Materials and methods 4.2.1 Site description and borehole drilling The Antamina mine is a polymetallic Cu, Zn, and Mo mine located in the Peruvian Andes at ~4,500 m elevation. The annual precipitation and temperature at Antamina vary between 1,200 to 1,500 mm and 5.4 to 8.5 ºC, respectively (Harrison et al., 2012), driven by distinct wet (i.e. from October to May) and dry season cycles (Bay et al., 2009). The main waste-rock types produced by the mine are limestone, marble, hornfels, skarn, and intrusives (Harrison et al., 2012). The different waste-rock types are segregated and stored based on a classification scheme that accounts for solid-phase sulfide and metal(loid) content (i.e. Zn, Cu, As), amongst others (Harrison et al., 2012). The waste rock is stored on-site in ~300 m tall end-dumped piles that are composed of distinct benches and separated by traffic surfaces.  In 2014, two boreholes, BH1D2 and BH3D2, were drilled at two different locations in the waste rock pile using a rotary vibratory sonic drill without introducing water. The boreholes served a dual purpose of 1) collecting a depth profile of waste rock heterogeneity and 2) installing instrumentation to measure temperature and pore gases as reviewed in another study (Vriens et al., 2018). Borehole BH1D2 (146 m deep) was drilled at an elevation of 4,473 m while BH3D2 (126 m deep) was drilled approximately 1 km east on a lower bench at an elevation of 4,273 m. The drill cores, composed of in situ continuous samples, were collected in bags of 1 m, preserving the spatial arrangement of materials.  Qualitative descriptions of the major and minor waste-rock lithologies, particle size distributions, and sulfide dissemination grades were logged by geologists. The occurrence of 87  disseminated liberated sulfides and oxidized alterations in the drill core matrix was qualified as trace, medium, or high, based on abundance and coloration intensity, respectively. The drill core was composed of loose waste-rock particles varying from cobbles (>12 cm) to clay size (<9 μm) particles. Because the rotary vibratory drilling may produce excess fine fractions in zones with dry cobbles and gravels with a diameter >2 mm (Horton et al., 2015), the abundance of boulders and cobbles was estimated visually from the occurrence of intense drill breakage and rock flour.  4.2.2 Bulk physicochemical analyses  Based on the lithological description and visual observations of prevalent secondary minerals, composite as well as discrete waste-rock samples were collected from the borehole drill core. Representative composite samples were taken from homogeneous intervals of up to 2 m length and excluded particles >10 cm diameter. In addition, discrete grab samples were taken from smaller, 1 to 20 cm intervals and excluded particles >10 cm diameter. Selected grab samples were sieved to <6 mm for determination of gravimetric water moisture content by mass according to ASTM procedure D2216-10 (ASTM, 2010) and sieved to <2 mm to perform rinse-pH tests (Price, 2009) and qualitative powder-pH tests with a Hellige-TruogTM test kit. Additional aliquots of collected waste-rock samples were homogenized, air dried, and their elemental content analyzed with i) inductively coupled plasma – mass spectrometry (ICP-MS) after a 4-acid digestion method for metals and ii) a LECO elemental analyzer for total sulfide and carbon content. The modified Sobek method was used to determine sulfate lixiviation in HCl, paste pH, and acid-base accounting parameters, as described in Lawrence and Wang (1996). Finally, a selection of samples was subjected to shake-flask leachate tests, i.e. 24 hours DI water extractions, according to the methods outlined in Price (2009), and the resultant leachates were analyzed by ICP-MS. All aforementioned analyses were conducted by a commercial laboratory (ALS in Peru, methods reference codes: 88  9207, 8585, 8580, 9306, 8490, 8079, 7497, and 8494), with quality control consisting of the analysis of triplicate samples and standard reference materials.  To evaluate metal(loid) mobilization behavior from the various waste-rock samples, solid-phase concentrations were compared with aqueous concentrations obtained from the leachate tests: elemental concentration ratios of solid-phase over leachate fractions were calculated and outliers identified by calculating the double median absolute deviation (2MAD) for asymmetric datasets (Reimann et al., 2005). Elements with waste rock/leachate ratios exceeding the median ratio +2MAD indicate relative retention in a borehole interval compared to the other intervals, whereas waste rock/leachate ratios less than the median -2 MAD suggest relative mobilization. Representative, pre-homogenized composite waste rock samples were collected, it was therefore assumed that the outliers were not an artefact of sample heterogeneity.  4.2.3 Mineralogical analyses Twenty-four waste-rock samples were selected amongst altered and non-altered layers for detailed mineralogical analysis with automated mineralogy and Raman spectroscopy. The reactive particle size-fraction (< 2 mm; Strömberg and Banwart, 1999) of dried waste-rock samples was isolated by sieving, homogenized, and subsequently mounted and polished on rounded epoxy resin thin sections. Thin sections were polished using a non-aqueous cooling lubricant to minimize potential phase dissolution or alteration, and carbon coated. The mineralogy of the waste rock was investigated with a Thermo-Fisher Mineral Liberation Analyzer (MLA), consisting of a FEI Quanta 600 scanning electron microscope (SEM), a dual Bruker-AXS silicon-drift energy dispersive X-ray analysis system (EDS), and the MLA DataView® software. Raman spectroscopy was performed on samples mounted on polished thin sections using a Horiba XploRA ONETM Raman and results were interpreted with the CrystalSleuth software (Laetsch, 2006). Full 89  instrumental details for MLA and RAMAN are provided in Appendix E. For quality control purposes, the solid-phase elemental composition of the investigated samples measured by MLA was additionally compared with compositions measured by field portable-X-ray fluorescence (FP-XRF) spot analyses. The latter were collected with an InnovX Olympus FP-XRF, calibrated with a reference standard (Olympus, n°316). The agreement between elemental compositions of Fe, Cu, and Zn recorded by MLA and FP-XRF was >82% for BH1D2 and between ~60-70% for BH3D2 (see linear regressions in Figure F1). 4.2.4 Mineralogical data processing The waste-rock mineralogy was evaluated by distinguishing between primary minerals and ‘secondary’ alteration/oxidation products (including both secondary and tertiary minerals per definition of Blowes et al., 2003) in the following mineral groups: sulfides, Fe-oxides, Fe-(hydroxy)sulfates, other (oxyhydr)oxides, carbonates, other sulfates, silicates, and other minerals (Table E1). The mineralogical evaluation was focused on minerals bearing Fe, Cu, Zn, As, and Mo. From the obtained mineralogical data, the following quantitative evaluation indices, also applied in Chapter 2 (St-Arnault et al., 2019a), were calculated: i) Mineral Reactivity (MR, in units of wt-% as defined in Aranda, 2010), as a measure of the availability of metal-bearing minerals to react. Mineral reactivity is defined as the product of the elemental grade in a mineral and the fraction of that mineral with a >70% liberated surface (i.e. 70% of the phase perimeter not being covered by another phase, because 100% liberation may underestimate results from unsieved fractions, Petruk 2000; equation 1): ܯܴ	ሺݓݐ%ሻ ൌ ∑ ܮܾ݅݁ݎܽݐ݅݋݊ሺ݅ሻ 	 ∙ ܧ݈݁݉݁݊ݐ݈ܽ	݀݅ݏݐݎܾ݅ݑݐ݅݋݊ሺ݅ሻ	௡௜ୀଵ 	    eq. 1 where n = number of mineral phases, i, that contain the metal of concern.  90  ii) Association Index (AI, unitless as defined in Lund et al., 2015) to indicate the association between pyrite and other minerals and thus the potential for pyritic surface passivation or galvanic interactions (equation 2):  ܣܫ ൌ ெ௜௡௘௥௔௟	஺	௔௦௦௢௖௜௔௧௘ௗ	௪௜௧௛	௠௜௡௘௥௔௟	஻	ሺ௘௫௖௟௨ௗ௜௡௚	௟௜௕௘௥௔௧௘ௗ	௚௥௔௜௡௦	௢௙	௠௜௡௘௥௔௟	஺ሻீ௥௔ௗ௘	௢௙	௠௜௡௘௥௔௟	஻	ሺ௘௫௖௟௨ௗ௜௡௚	௚௥௔ௗ௘	௢௙	௠௜௡௘௥௔௟	஺ሻ    eq. 2 An association index >1 or <1 indicates that the association of two minerals is more or less common, respectively, than what is suggested by the modal mineralogy. If AI = 0, there is no association between both minerals and if AI = 1, the association is as common as the modal mineralogy.  iii) Galvanic coupling index (GCI; unitless), to indicate the degree of contact between liberated sulfide grains in the matrix of borehole samples. It is defined as the liberated amount of a cathode mineral (i.e. pyrite, in wt-%) divided by the liberated amount of an anode mineral (i.e. chalcopyrite or sphalerite [wt-%]); equation 3):  ܩܥܫ ൌ ∑௅௜௕௘௥௔௧௘ௗ	஼௔௧௛௢ௗ௘	ெ௜௡௘௥௔௟	஼௢௡௧௘௡௧	∑௅௜௕௘௥௔௧௘ௗ	஺௡௢ௗ௘	ெ௜௡௘௥௔௟	஼௢௡௧௘௡௧          eq. 3 The GCI only accounts for liberated minerals as this allows contact within the core sample matrix between anodic and cathodic minerals that can undergo galvanic interaction. A GCI < 1 suggests a galvanic protection of pyrite (i.e. less liberated pyrite is in contact with more abundant chalcopyrite and sphalerite), whereas a GCI > 1 indicates limited galvanic protection of abundant pyrite and preferential dissolution of less abundant chalcopyrite and sphalerite.  91  4.3 Results and discussion 4.3.1 Bulk physical and geochemical properties of the waste rock  The bulk waste-rock properties measured along the profiles of boreholes BH1D2 and BH3D2, including its lithology, grain size distribution (Figure 4.1), mineralogy, chemical composition (Figure 4.2), and moisture content (Figure F2), all illustrate the significant heterogeneity in the studied waste-rock pile.  Cores collected from both boreholes were mainly composed of carbonate- and silicate-rich waste rock such as marbles and hornfels (Figure 4.1). This carbonate-rich waste rock has a high neutralization potential (NPR>2; Figure 4.2) which was similarly observed from samples of field barrels in Chapter 2 and boreholes at Antamina (Vriens et al., 2018; St-Arnault et al., 2019a). However, both boreholes also contained appreciable amounts of skarn and intrusive waste rock, distributed along 33% of the total length of BH3D2 and 25% of BH1D2 (Figure 4.1). These more reactive waste-rock types have previously been identified as potentially acid-producing (NPR≤1; Figure 4.2) in waste rock from other boreholes at Antamina and field barrels analyzed in Chapter 2 (Vriens et al., 2018; St-Arnault et al., 2019a). Strong to medium intensity waste-rock oxidation and higher disseminated sulfide contents were observed at 6 m, 22 m, 35 to 43 m, and 85 to 103 m depth in BH1D2 and at 4-15 m, 28 m, 35-41 m, 91 m, and 95 m depth in BH3D2 (Figure 4.1).   The qualitative particle size distribution in both boreholes was dominated by fine to medium-sized grains (i.e. gravels, sands, and clays), but coarser particles (medium to large grains such as boulders, pebbles and gravels) existed at depths between 40 to 70 m and 115 m to 125 m in BH1D2, and between 50 to 78 m and 97 to 119 m depth in BH3D2 (Figure 4.1). Fine-grained clayey alluvial material was also observed between 20-50 m depth in BH1D2 and between 20-22 m, as well as 79 m and 121 m depth in BH3D2 (Figure 4.1). The gravimetric moisture content in  92  Figure 4.1. Qualitative observations in cores retrieved from boreholes BH1D2 and BH3D2: sulfide dissemination, oxidation intensity (both categorized as trace, medium or high; upper x-axes), particle size (grouped as alluvial, sand/silt, or gravel/pebbles; lower x-axes), and qualitative bulk lithology of average section intervals along the profile (pie-chart insets; legend in the bottom right). Absent rectangles for oxidation intensity and disseminated sulfides 93  indicate that they were not visible in the rock matrix. The average lithology of the entire borehole profile is indicated above each borehole. The grey shading marks the reactive zones as indicated by pore-gas analyses and discussed in the text. the waste rock ranged between 1 wt% and 14 wt% in the two boreholes, with maximum moisture contents recorded at 20 m, between 35 to 50 m, and 95 to 100 m depth in BH1D2 as well as at 20 m and 80 m depth in BH3D2 (Figure F2). These higher moisture intervals correlate with observations of clayey alluvial deposits as described in the geological logs, and are likely the result of compaction of traffic surfaces (Figure 4.1). Similar to previous work in the same waste-rock pile (Vriens et al., 2018), five reactive zones can be demarcated in the two boreholes by records of elevated temperature and CO2 content as well as lower oxygen content (Figure F5). These reactive zones or hotspots were located between 10-15 m, 35-55 m, and 80-100 m depth in BH3D2 and between 25-48 m, and 75-100 m depth in BH1D2 (see grey shading in Figures 4.1, 4.2 and F5). Although the vertical resolution of these pore-gas sensor records was several meters, the deduced location of these reactive zones from pore-gas analyses coincides with an abundance of fine- to medium-sized particles interlayered with low permeable clayey alluvial layers with higher moisture content of up to 11 wt% (Figures 4.1 and F2). In contrast, the non-reactive zones are mainly composed of porous and permeable layers with coarse particles and low moisture content (as low as 1 wt%; Figures 1 and S2). The qualitative grain size distribution thus indicated potential moisture traps along the profiles, coinciding with zones of more evident alteration and higher disseminated sulfide contents (Figures 4.1 and F2). Overall, the qualitative characterization of the internal structure of the pile provides a useful first assessment of the distribution of waste-rock reactivity, as conceptualized in Figure 4.3.  In future study, as applied in Chapter 3, the first indication of impact of dominant flow paths on the reactivity of minerals within those intervals  94  Figure 4.2. Solid-phase concentrations of total S, Fe, Cu and Zn as well as neutralization potential ratios (NPR) of the waste rock in boreholes BH1D2 and BH3D2, combined with the concentrations of SO4, Cu and Zn in the leachates from the rinse tests of the same samples. The upper detection limit for solid-phase Zn was 1 wt-%; dissolved Fe in the 95  leachates was not detected below the detection limit of 0.1 mg/L. The grey shading marks the reactive zones as indicated by pore-gas analyses and discussed in the text. may be further investigated by comparing the weathering textures of secondary minerals between reactive and non-reactive zones.   In addition to the qualitative lithology and particle size distributions, quantitative geochemical composition measurements also reflected the heterogeneity of the waste-rock pile. While the median total sulfur (S) content of the waste rock was moderate at ~1%, peak concentrations of up to 19% total S occurred between 10-13 m, 35-40 m, 80-105 m and at 118 m depth in BH3D2 and peaks of up to 4% total S existed between 77-107.5 m and at 130 m depth in BH1D2 (Figure 4.2). Similarly irregular distributions were observed for other elements, with peak concentrations reaching up to 10 times their respective median concentrations: e.g.,  for Fe (>10% at 87 m, 100 m, and 109.7 m in BH1D2 and between 20-25 m, 59-65 m, and at 91.3 m in BH3D2), Zn (>1 % at 79 m in BH1D2 and between 13.5-18 m in BH3D2), Cu (>1 % at ~87 m depth in BH1D2 and between 87-94 m in BH3D2; Figure 4.2), as well as for As (>350 ppm between 77-94 m in BH1D2 and between 19-21.45 m in BH3D2) and Mo (>750 ppm between 7-10 m in BH3D2; Figure F3). In general, total S concentrations were higher in BH3D2 than in BH1D2, whereas other measured elemental concentrations were more comparable between the two boreholes. Although relatively elevated concentrations of S, Zn, Cu, Fe, As, and Mo were prevailing in the reactive zones of both boreholes and particularly in BH1D2, substantial concentrations of Fe, As, Cu, and Mo could also be observed in the non-reactive zones of BH3D2 (Figure 4.2). The agreement between the depth-profiles of S versus Fe, Cu, and Zn observed in BH1D2 but not BH3D2 suggest different (non-sulfidic mineral) sources for these metals in BH3D2. The median elemental concentrations recorded in the boreholes are in line with elemental 96  compositions measured in previous experiments conducted at the Antamina mine site (Hirsche et al., 2017; Vriens et al., 2018; Vriens et al., 2019).  The neutralization potential ratios (NPR) from composite waste-rock samples along the boreholes were highly variable and ranged from 0.3 to 124 (Figure 4.2). Sizeable intervals (e.g. between 12-13 m, 35-40.2 m, and 83-106 m in BH3D2 or between 90-95.25 m in BH1D2) exhibited relatively low NPR values (i.e. NPR<3), indicating pockets of potentially acid-generating waste rock (Figure 4.2). Comparatively low rinse- and powder-pH measurements in selected grab samples from discrete locations at 20 m and 22.9 m depth in BH1D2 and 11.5 m depth in BH3D2 (pH<4.5; Figure F4) corroborate the acid-producing potential of this waste rock and suggest that localized acidic environments may exist throughout the pile. This supports previous observations of acidic (micro)environments in smaller-scale experiments at Antamina (Dockrey et al., 2014; Vriens et al., 2019a). In summary, the qualitative and quantitative physicochemical properties of the waste rock indicate distinctive heterogeneities along the continuous profile of the studied waste-rock pile: qualitative and quantitative physicochemical parameters differed on spatial scales ranging from the meter to the centimeter scale. Reactive zones previously defined from in-situ pore gas and temperature data are largely confirmed by the presented data, but the high-resolution sampling and analysis reveals a degree of physicochemical heterogeneity that could not have been resolved from static laboratory testing or sensor measurements alone.    97  Figure 4.3. Conceptual schematic of the waste-rock pile and the locations of reactive zones (in pink) with approximate particle size variations and grading in the pile benches separated by traffic surfaces along boreholes BH1D2 and BH3D2 (not to scale).  4.3.2 Waste rock mineralogy  The mineral composition of the waste rock in the two boreholes was, similar to the bulk geochemistry discussed above, highly heterogeneous and varied from micrometer to meter scales. The bulk mineralogy in the waste rock was dominated by silicates (29-92 wt%) carbonates (2-83 wt%) and sulfides (<1-40 wt%), with smaller amounts of sulfates (<1-10 wt%), Fe-oxides (0-8 wt%) and other minerals (up to 2 wt%; Figure F6). On a per-element basis, the major identified and liberated (i.e. >70% surface perimeter exposed) Fe, Cu, and Zn-bearing primary minerals were silicates such as epidote, biotite, chlorite, grunerite, zincsilite, titanite and dioptase; sulfides such as pyrite, pyrrhotite, chalcopyrite, and sphalerite; as well as various Fe-(oxyhydr)oxides (Figures F7 and F8). In addition to primary minerals, up to 16 wt% of the total modal mineralogy was identified as secondary alteration/oxidation product that occurred at various depths throughout 98  both boreholes. Of this 16 wt% secondary fraction, up to 63 wt% of minerals were Fe-bearing, up to 74 wt% of minerals were Cu-bearing and up to 66 wt% Zn-bearing (Figures F6 to F8). Major identified Fe-, Cu-, and Zn-bearing secondary phases (both discrete mineral phases as well as minerals containing traces of lattice-substituted or adsorbed metals) were silicates (e.g. mica, talc, zincsilite, Ti-silicate), Fe-oxides, (hydroxy)sulfates (e.g. goslarite, jarosite) and other (oxyhydr)oxides (e.g. cuprite, fornacite, wulfingite) and carbonates (e.g. siderite, ankerite; Figures F7 and F8). Due to lower overall abundance, the As and Mo mineralogy was primarily investigated in samples containing high elemental concentrations. In those samples, major As- or Mo-bearing phases were primary sulfides and secondary Fe-(hydroxy)sulfates, oxides and carbonates, and/or sulfates (Figure F9).  Pervasive precipitation of oxidation products in the matrix of the waste-rock was mainly observed in samples from the reactive (high-S) zones, whereas lesser or non-oxidized waste rock dominated in samples from the non-reactive zones (compare Figures 4.1, F6, F10, and F11). Surprisingly, minimally altered waste rock also occurred in high-S intervals, e.g. in BH1D2 between 77-78 m and 96-97 m or in BH3D2 between 84-86 m and 100.75-101.65 m; (Figures 4.1, F6, F10, and F11). Pore-gas data suggest that these poorly altered layers were located in anoxic or low O2 (<10 v-%) and lower temperature (10-17°C) zones (Figure F5), where O2 transport limitations may have inhibited sulfide oxidation (Lefebvre et al., 2001; Vriens et al., 2018). Although the reactive zones generally contained mostly altered waste rock, there thus appears to be a mechanism causing preferential oxidation of certain sulfidic materials while maintaining ‘inertness’ of neighboring sulfidic waste rock.  Sharp and progressive contacts, over several centimeters to meters, were observed between oxidized and non-oxidized layers in the matrix of waste rock from BH3D2 (between 11-13 m, 34-99  35 m, 95-97 m) and that from BH1D2 (between 20-24 m and 104-107 m; Figures F10 and F11). The existence of contacts between altered and unaltered layers likely resulted from transient water table conditions and localized capillary barriers, as has been previously described in homogeneous tailings materials (Moncur et al., 2005; Graupner et al., 2007; Redwan et al., 2012; Lindsay et al., 2015; Elghali et al., 2019b) and to a lesser extent in smaller-size waste-rock piles (Gieré et al., 2003; Sracek et al., 2004). Layered mineralogy patterns in mine tailings have previously been attributed to capillary fringes and cementation of less permeable layers with localized precipitation of secondary sulfates (Graupner et al., 2007; Redwan et al., 2012; Liu et al., 2018;), which has also been corroborated by reactive-transport simulations (Meima et al., 2012). In the boreholes studied here, the observed agglomeration and transitional contact from white to orange alterations (e.g. in BH3D2 between 34-35 m, at 47.2 m, and between 95-96 m and in BH1D2 between 20-23 m; 95-97 m and 104-108 m; see Figures F10 and F11) may similarly be evidence of transient conditions and localized capillary barriers in the waste-rock pile: secondary minerals were generally enriched in the altered waste rock layers compared to non-altered layers (average of 7 wt-% compared to 2 wt-%, respectively; Figure F6). In addition, the major secondary minerals identified in the altered layers (i.e. Fe-(oxyhydr)oxides such as ferrihydrite and lepidocrocite and Fe-(hydroxy)sulfates such as alunite, brochantite, or posnjakite; Table E2), have also been previously implied in the cementation of unsaturated oxidized tailings (Blowes and Jambor, 1990; Moncur et al., 2005; Graupner et al., 2007; Lindsay et al., 2015; Liu et al., 2018). Although quantitative waste rock porosity along the boreholes was not determined in this study, qualitative observations show particle agglomeration on the centimeter to micrometer scale (e.g. in BH1D2 between 20-21 m, 22.3-22.5 m, or 96-97 m, and in BH3D2 between 11.25-11.8 m, 34.4-34.6 m, 91-91.25 m, 93.5-93.7 m, or 95-95.8m; Figures F12 and F13). This observed particle agglomeration has potentially 100  contributed to reduce the porosity of the oxidized layers, as previous work demonstrated that secondary mineral precipitation may lead to porosity reductions of at least 33% (Redwan et al., 2012). Further characterization of porosity gradients in the altered zones of the studied waste-rock pile is therefore warranted.  In summary, quantitative mineralogical analyses revealed a highly heterogeneous mineralogical assemblage in the waste-rock pile. Secondary minerals, as markers of oxidation intensity, revealed discrepancies between mineralogical alteration and geochemical reactivity and confirmed the existence of inert zones. Sharp and progressive mineralogical contacts indicate that discrete internal alterations are preserved within sequential benches in this large-scale waste-rock pile. The results show that this genetic layering (Figure 4.3), resembling a hardpan formation in tailings, can thus occur below compacted traffic surfaces at significant depths in waste-rock piles, and is, in general, not restricted to less permeable layers in tailings (Moncur et al., 2005; Graupner et al., 2007; Redwan et al., 2012; Lindsay et al., 2015; Elghali et al., 2019b) or the near–surface layers of waste-rock piles (Gieré et al., 2003; Sracek et al., 2004; Smuda et al., 2007). 4.3.3 Mineral reactivity and associations  Mineral reactivity indices for Fe-sulfides were generally lower than those for Cu- and Zn sulfides (Figure 4.4). Reactivity indices for Fe-, Cu- and Zn-sulfides were highly variable along the boreholes, with relatively high and low reactivity variations between adjacent samples occurring throughout various intervals (Figure 4.4). The reactive zones generally exhibited higher total-S levels and Fe- and Cu- sulfide reactivity indices (i.e. through higher modal content and liberation), which corroborates the oxygen depletion and temperature rise recorded in those intervals (Figure F5). Pyrite in the waste-rock samples was mainly associated with other sulfides and secondary minerals (i.e. AI>1), although such associations were less abundant in some 101  samples, e.g. from BH1D2 at 96 m and 107.55 m, as well as BH3D2 at 11.25 m, 11.85 m, 34 m, 62 m, 84.4 m, 95 m, 95.3 m, and 100.75 m with AI<1 (Figure 4.4). Independent of reactivity, the selected samples of waste rock from both boreholes generally exhibited strong pyritic associations with chalcopyrite, sphalerite, Fe-(hydroxy)sulfates and carbonates (i.e. AI>10;) as well as Fe oxides, silicates, and sulfates (i.e. AI between 1 to 10, Figure F14). Finally, contact between cathodic pyrite and anodic chalcopyrite and sphalerite that would enable galvanic interaction was generated by more abundant quantity of pyrite than chalcopyrite and sphalerite (i.e. GCI between 1 and 10) along BH1D2, whereas in BH3D2, liberated pyrite was less abundant than chalcopyrite and sphalerite (i.e. GCI<1) in samples up to 63 m depth, but more abundant than chalcopyrite and sphalerite (i.e. GCI>1) in samples from 84 m downward (Figure 4.4). In both BH1D2 and BH3D2, the higher GCIs (i.e. lower pyrite inhibition by galvanic interaction) appeared related to higher pyrite reactivity indices (Figure 4.4), which suggests a higher abundance of liberated pyrite compared to liberated chalcopyrite and sphalerite in these intervals (Figure F15). In contrast, there was little correlation between pyritic association indices with other sulfides and the GCI, which is explained by the fact that the association index only reflects contact between phases within the same particle (Lund et al., 2015), where the GCI accounts for contact of liberated grains within the matrix. In general, the previously recorded reactive zones from gas and temperature observations are mirrored by elevated reactivity indices for pyrite and other sulfides and corroborated by higher GCIs corresponding to limited pyrite inhibition by galvanic interactions. However, the primary and secondary mineral assemblage associated with pyrite was similar, independent of the degree of liberation or the abundance of pyrite. 102  Figure 4.4. Mineralogical parameters of a selection of samples along the depth-profiles of boreholes BH1D2 (top) and BH3D2 (bottom). The location of the investigated waste-rock samples is indicated in the left panel, Fe-, Cu-, and Zn-sulfide reactivity indices and the fraction of mineral phases of the total assemblage classified as secondary are 103  given in the middle panel. Galvanic coupling indices and pyritic mineral association indices are illustrated on the right panel. Legends are indicated on top of each panel. The grey shading in the borehole and panels on the right mark the reactive zones as indicated by pore-gas analyses and as discussed in the text. 4.3.3.1  Mineral surface passivation  Relatively low sulfide reactivity indices (RI<30%; Figure 4.4) and strong associations between pyrite and secondary oxidation products such as Fe-oxides or Fe-(hydroxy)sulfates (AI>10; Figure 4.4) suggest significant occlusion of primary minerals by secondary precipitates in specific zones in both boreholes (e.g. BH1D2 at 86.65m and 95.25m or BH3D2 at 11.25m). In fact, a lower reactivity index of pyrite compared to that of chalcopyrite and sphalerite suggests that pyritic surfaces may be more efficiently passivated by secondary minerals than chalcopyrite and sphalerite (Figure 4.4). This may be related to the higher abundance of Fe or to a higher stability of secondary Fe-oxides compared to their Cu- or Zn-analogues (WATEQ4F thermodynamic database from Ball and Nordstrom, 1991). Further analysis of the secondary mineral assemblage in altered waste rock from the reactive zones of both boreholes confirmed the presence of abundant amorphous phases surrounding primary silicate, carbonate or sulfides minerals (Figure F16, Table E2) or as part of agglomerate particles (Figure F13). Several samples had particularly high contents of secondary Fe-(hydroxy)sulfates and/or Fe-oxides (e.g. BH1D2: 5% between 86.65-87 m and 6% between 95.25-96 m, and BH3D2: 9% between 11.25-11.8 m; Figure F6). Colored SEM images of these samples show sulfides completely occluded by coatings (Figure F16) or as part of agglomerates (Figure F13). The amorphous nature of these coatings complicated the identification of their mineral structure with MLA, but investigation with Raman spectroscopy confirmed the presence of Fe-(oxyhydr)oxides (i.e. goethite, ferrihydrite, lepidocrocite, and hematite), sulfates such as gypsum, brochantite, antlerite, and alunite, as well carbonates such as siderite (Table E2). 104  The inverse relationship between pyrite reactivity and the abundance of Fe-, Cu-, or Zn- bearing secondary phases (Figure F17), as well as the occasionally strong associations between pyrite and secondary phases (e.g. AI>1 in BH1D2 between 86.65-87 m and 95.25-97m as well as in BH3D2 between 11.25-12.45 m and 47.2-48 m; Figure 4.4), both confirm that surface passivation plays an important role in reducing pyritic reactivity in selected waste-rock samples.  4.3.3.2 Galvanic interactions Selected intervals in boreholes BH1D2 (20-40 m and 90-100 m) and borehole BH3D2 (10-15 m, 30-40 m, 80-90 m, and 95-102 m) contained substantial amounts of disseminated sulfide grains (Figure 4.1) and up to 20% total sulfur (Figure 4.2) of which up to 25 wt% was liberated (Figure F15). In these intervals, as well as in selected other samples throughout the boreholes, elevated sulfide contents did not always correspond with higher secondary mineral precipitation. This discrepancy suggests that processes might be inhibiting the oxidation of sulfides such as oxygen depletion or the preferential dissolution of one sulfide over the other. Amongst those samples, strong associations between pyrite and chalcopyrite or sphalerite (Figures 4.1 and 4.4) suggest potential galvanic interactions. Indeed, galvanic coupling indices >1 occurred in altered intervals (e.g. BH1D2 between 27.25-28 m, 86.65-87 m, and BH3D2 between 95-95.3 m; Figure 4.4) but illustrate that galvanic protection of pyrite was partially limited and pyrite still remained relatively reactive. Instead, poorly altered layers generally exhibited GCI<1, suggesting preferential dissolution of chalcopyrite and sphalerite and galvanic protection of pyrite in these intervals (e.g. BH1D2 between 96-97 m and 107.5-108 m, Figure 4.4). Consequently, the combined effect of galvanic interactions between sulfides liberated in the matrix or coupled in the same particle may have protected pyrite-rich layers from oxidation and alteration in the reactive zones despite a higher sulfide content, explaining at least in part some of the abovementioned 105  discrepancies between bulk geochemical reactivity and mineral alterations. In contrast, sulfidic associations suggest that physical constraints may be responsible for the occurrence of inert zones observed earlier (e.g. in BH3D2 at 84 to 86 m and 100.75 to 101.65 m), as minimal alteration was observed in disseminated sulfide-rich layers with elevated pyrite reactivity indices (Figures 4.2 and 4.4). In addition, a limited pyrite protection (GCI>1) and minimal preferential association with traces of chalcopyrite (Figure 4.4) suggest that physical constraints such as low oxygen or water supply may have been responsible for protecting those layers from alteration. The quantitative contribution of such physical constraints should be further investigated as part of future research. In summary, the presented quantitative mineralogy data show that mineral liberation and associations may explain observations of anomalous alteration heterogeneities and small-scale contacts. Reduced sulfide reactivity in reactive zones can be at least partially attributed to sulfide passivation resulting from secondary mineral precipitation whereas a lack of alteration in reactive zones may to some degree be explained by galvanic inhibition of pyrite.  4.3.4 Waste-rock leaching experiments While the drainage composition of a heterogeneous waste-rock pile will carry the aggregate signature of the entire pile, the leachate tests on individual borehole samples provide insights into metal mobility within the pile (Figure 4.2). In the 24-h leaching tests, pH values were on average 7; reflective of the net-non-acid-generating character of the waste rock at Antamina. Median leachate concentrations of sulfate, Cu, and Zn were 0.3 g/L, 4-6 μg/L, and 0.01 mg/L, respectively, with localized peaks of up to 1.5 g/L for SO4; 100 μg/L for Cu, and 0.5 mg/L for Zn (Figure 4.2). Cu and Zn concentrations were generally low and Fe could not be detected in the circumneutral DI water tests, reflective of the fact that a significant fraction of these metals precipitated out of solution as insoluble secondary phases (Smith, 2007; Vriens et al., 2019). Overall, total S and Zn 106  leachate concentrations were disproportional to solid-phase concentrations along the borehole profiles, with the exception of selected intervals (e.g. in BH1D2 between 90-105 m and in BH3D2 at 12 m, between 62-63 m, and 80-95 m). The agreement between Cu concentrations observed in leachates and the respective waste rock solid-phases was comparatively better than that for S and Zn. Calculation of solid-phase/leachate concentration ratios and outliers therein (Figure 4.5) indicate both relative mobilization of Cu and Zn, e.g. localized in BH1D2 between 84-105 m and in BH3D2 between 11-12 m and 55-80 m, but also selective retention of Cu and Zn in other intervals, e.g. in BH1D2 between 75-85 m and in BH3D2 between 35-55 m and 90-105 m. In the reactive zones of both boreholes, Zn and Cu are preferentially leached from altered and non-altered layers with marbles and skarn waste rock. Interestingly, relative retention of Zn and Cu was also observed in altered and non-altered layers of the reactive zones in BH3D2 and in the less reactive zones and non-altered layers in BH1D2 (compare Figures 4.2 and 4.5, Table E3). Similar preferential release of Zn from waste rock composed of marbles and skarn and attenuation of Zn in intrusive and hornfels waste rock was observed with kinetic tests previously performed at Antamina (Hirsche et al., 2017). Median As and Mo concentrations in the leachate tests were 4-7 μg/L and 0.2 mg/L, respectively, with peaks of >200 μg/L As (e.g. in BH1D2 between 70-90 m and in BH3D2 at 49 m and 78.55 m) and >4,000 mg/L Mo (e.g. in BH1D2 at 71 m and in BH3D2 at 28 m, between 45-60 m, and 73-83 m), respectively (Figure F3). In comparison with Cu and Zn, leachate concentrations of As and Mo were more proportional to their solid-phase concentrations, with leachate over solid-phase concentration ratios outliers indicating preferential leaching of As, e.g. between 77-97 m in BH1D2, and preferential mobilization of both As and Mo in BH3D2 between 45-60 m (Figures F3 and F18). Similarly effective leaching of As and Mo, proportional to 107  elemental grades, was observed in previous kinetic tests performed with Antamina waste rock (Vriens et al., 2019c). Release and retention of As and Mo were predominantly recorded in non-altered layers of the non-reactive zones with marble and intrusive layers, whereas As appeared retained only in altered layers composed of marble and skarn from the reactive zones of BH1D2 Figure 4.5. Ratios of solid-phase Cu and Zn waste-rock concentrations over respective aqueous leachate concentrations from samples along boreholes BH1D2 and BH3D2 (logarithmic x-axis). Reactive zones are indicated by the grey horizontal shading. Vertical dotted lines represent the median concentration ratio in that profile; the ±2MAD outlier threshold is indicated by the red shading. Samples right of the +2MAD threshold suggest relative retention, whereas samples left of the -2MAD threshold indicate relative mobilization. 108  between 93-102 m (compare Figures 4.2 and F3, Table E3). Again, the preferential release of As associated with intrusive waste rock and retention of As with marble waste rock reflects results from smaller scale kinetic tests with combined lithologies at Antamina (Hirsche et al., 2017). Lithological composition might therefore be a dominant control on As and Mo mobilization, mostly occurring within non-altered layers in both boreholes.    The circumneutral leachate pH values in the tests generally do not promote the dissolution of more stable phases and desorption of metal(loid)s from Fe bearing phases (Vriens et al., 2019c). Nevertheless Cu, Zn, As and Mo may be mobile in the absence of sufficient Fe-mineral substrate (Smith, 2007), which may explain the previously discussed preferential retention of these elements within altered layers enriched in secondary Fe-phases and release from non-altered layers with little secondary Fe-substrate. For example, As and Mo may be expected to be poorly attenuated by sorption to amorphous secondary Fe-phases under circumneutral conditions in non-altered layers of the non-reactive zones. In contrast to As and Mo, Cu and Zn were more effectively released from the reactive zones (Figure 4.2). This suggests that mechanisms other than sorption on Fe-substrates and the primary mineralogical composition such as surface liberation or galvanic reactions might control their leaching behavior. 4.3.5 Mineralogical controls on leachate chemistry In several non-altered layers (e.g. BH1D2 between 96-97 m, BH3D2 between 11.85-12.45 m and 62-63 m), pyritic associations with other sulfides (i.e. AI > 1) and low galvanic coupling indices (GCI ≤ 1) suggested galvanic protection of pyrite and preferential dissolution of chalcopyrite and sphalerite (Figure 4.4). In the leaching tests of these samples, relative mobilization of Zn and Cu was revealed (Figures 4.2, 4.5 and Table E3). The Fe concentrations in leachate were below detection limits, despite Fe-sulfide content up to 6 wt% and abundant 109  secondary Fe-oxide precipitation, suggesting retention of Fe (Figures 4.2, 4.5 and Table E3). Sulfide coupling was observed from MLA images and identified with Raman spectroscopy in multiple locations within these intervals (Figure F19). The absence of matrix alteration combined with elevated pyritic-sulfide associations and galvanic indices in these locations all suggest galvanic protection of pyrite. The mineralogical associations between reactive sulfides within the same particle or within the matrix contributed to reduced alteration of pyrite in the reactive zones, while preferentially leaching Cu and Zn with undetected Fe from poorly altered layers, regardless of the solid-phase abundance of these minerals. The effective mobilization of Zn from altered and non-altered layers composed of predominantly stable Zn-bearing minerals such as sphalerite, silicates (i.e. willemite and amorphous Fe-silicate phases with co-precipitated or adsorbed Zn) and apatite (see phase definitions in Table E1; Figure F20), may have originated from the possible desorption of Zn from these phases under circumneutral pH (Roberts et al., 2003; Smith, 2007). The observed preferential mobilization of As and Mo at select intervals, including in BH1D2 between 104.15-105 m and in BH3D2 at 47.2 m, 95.3-95.5 m, and 95.8-96 m (Figure F3, Table E3), may have been facilitated by a lack of sorption sites in these Fe-poor zones (Figure 4.2). In addition, the fact that up to 75% of the As and Mo was present in more soluble and liberated minerals such as carbonates (i.e. siderite) or amorphous Fe-sulfates (Figure F9) might have facilitated their mobilization. This was supported by liberated As/Mo-bearing minerals observed in these layers with MLA (Figures F21 and F22).  The non-detect Fe concentration in most leachates, despite up to 8 wt% Fe-sulfide content in the host waste rock, is consistent with the effective precipitation of amorphous secondary Fe-minerals (i.e. Fe-oxides or hydroxysulfates, carbonates, and/or silicates) in the matrix of altered 110  layers (e.g. in BH1D2 between 86.65-87 m and 95.25-96 m and in BH3D2 between 11.25-12 m; Figures 4.2, F7, and F8). Low pyrite reactivity indices in these locations are indicative of low liberation and modal abundance and poor associations with secondary minerals: they suggest potential surface passivation of pyrite in those layers (Figures 4.4 and F17), which is corroborated by observations of pyrite coated with secondary minerals (Figure F16).  Thus, both sorption and secondary mineral precipitation appear to play a major role in controlling the drainage quality, particularly under circumneutral to slightly acidic conditions. This has been abundantly observed at Antamina (Blackmore et al., 2018b; Vriens et al., 2019) and other mine sites (Gieré et al., 2003; Sracek et al., 2004; J. Smuda et al., 2007; Davies et al., 2011). Nevertheless, galvanic interactions may explain the anomalous abundance of secondary Cu and Zn adsorbed on silicates and Fe substrates as well as the limited surface liberation through precipitation of stable secondary mineral agglomerate around sulfides. Further investigations on altered layers could distinguish the roles of passivation and galvanic interactions from adsorption and therefore quantify the extent to which mineralogical parameters like liberation (or reactivity) reflect the passivating effect of secondary mineral precipitation on weathering rates and therefore drainage quality. 4.3.6 Practical implications for waste-rock management The variable leachate chemistries of the studied waste-rock samples reflect the highly heterogeneous waste-rock properties identified along the boreholes. This corroborates previous findings that heterogeneous distributions of material in waste-rock piles can result in variable hydrological regimes and development of multiple ‘seeps’ with different drainage qualities (Amos et al., 2014; Vriens et al., 2019a). Similar preferential attenuation and release of metals based on the lithological composition, as well as the presence of acidic micro-environments observed in 111  smaller scale experiments from Chapter 2 or previous studies (St-Arnault et al., 2019a, Vriens et al., 2019a, 2019c; Hirsche et al., 2017; Dockrey et al., 2014) were also observed at the larger scale suggesting that comparable processes occur at different scale. In full-scale waste-rock storage facilities, weathering and drainage of reactive zones within heterogeneous composite piles may yield i) local alteration patterns based on the bulk geochemistry, ii) oxygen depletion patterns extending across the scale with which mineralogical variations occur (i.e. meter-versus-micrometer scales) as well as iii) drainage mixing resulting in aggregate chemical leachate signatures. Bulk geochemical analyses may suffice to predict local drainage chemistries originating from homogeneous and well-characterized waste-rock types, yet evidence is mounting that the drainage quality is not proportional to the average bulk solid-phase composition of heterogeneous composite piles (Pedretti et al., 2017; Parbhakar-Fox et al., 2018b; Vriens et al., 2019a). In that regard, additional information on the elemental distribution, liberation, and mineral associations used in tandem with bulk mineralogy was useful in this study to identify mineralogical controls (e.g. sorption, surface passivation, and galvanic reactions) on drainage chemistry. This demonstrates that further identification of mineralogical phases including their crystallinity, impurities and their degree of liberation, may complement bulk geochemical analyses to improve drainage quality prediction models at other sites. Thus, information gathered on mineral reactivity, liberation and the abundance of amorphous phases may support choices of the appropriate mineral assemblage and its stability which may help reduce uncertainties normally related to the degree of mineral reactivity and the selection of appropriate secondary minerals represented in prediction models (Linklater et al., 2005). Mineralogical textures and other automated mineralogical data may be collected cost-effectively in collaboration with existing automated analytical investigations for other disciplines at mine sites (e.g. metallurgy and ore processing; Brough et al., 2013, 112  Parbhakar-Fox et al., 2018b). The additional interpretation of mineralogical textures demonstrated by this study could therefore bring valuable information to optimize storage conditions and drainage predictions, even during mine operation. 4.4 Conclusions Two continuous boreholes drilled in a heterogeneous waste-rock pile revealed depth-profiles of highly variable geochemical composition, grain size, and oxygen content. Geochemical and mineralogical data revealed distinct reactive and non-reactive zones located within the pile: reactive zones exhibited elevated temperatures, lower oxygen content, higher sulfide content, and were generally characterized by ubiquitous orange/brown matrix coloring from secondary mineral precipitation. These reactive zones were located within finer-grained layers, within and below traffic surfaces located in waste-rock benches featuring upward finer grading. Sharp and progressive contact between altered and non-altered layers indicated small-scale heterogeneities likely caused by variable moisture content and capillary barriers created by cementation. Poorly-altered sulfide-rich layers existed within reactive zones, showing that random dumping may create conditions not amenable for weathering. Localized inert zones of low alteration within high-sulfide hotspots was linked to potential galvanic interactions between sulfide couples and/or low oxygen conditions. Finally, disproportionalities between solid-phase metal concentrations and metal mobilization in leaching tests appeared partially linked to mineralogical observations: effective Cu and Zn leaching seemed to be facilitated by galvanically promoted dissolution, whereas relative As and Fe retention could be ascribed to secondary mineral formation and sorption as well as pyritic surface passivation. In summary, this study provided unique insights into the material distribution and weathering processes in hundred-meter tall heterogeneous waste-rock pile. Data on mineralogical composition, association and liberation were used to interpret mineral reactivity, 113  complementing more readily available physical and geochemical data. Quantitative mineralogical evaluations may therefore improve the interpretation of waste-rock reactivity, internal weathering processes and ultimately contribute to more accurate drainage prediction models.   114  Chapter 5: Conclusion The characterization of waste rock and drainage concentration plays a key role in achieving realistic conceptual models and reliable predictions of acid mine drainage or metal leaching. However, the influence of hydrological or physico-geochemical processes might affect the weathering rate of waste rock resulting in mass-balance discrepancies between solid and aqueous phase concentrations. The overarching focus of this dissertation was to refine the existing knowledge of geochemical and hydrological processes responsible for element mobilization and attenuation in relation to mineral reactivity and waste-rock drainage chemistry in order to improve drainage predictions. The sampling of waste rock and drainage water from small-scale field barrels (i.e. 1 m) occurred over a period of seven years allowing for geochemical assessment over time. The recording of outflow and application of Cl-Li tracer tests evaluated the hydrological controls in these field barrels. Similarly, the sampling of waste-rock samples and leachate tests from selected intervals of two boreholes allowed the geochemical characterization of a heterogeneous full-scale waste-rock pile at Antamina. The qualitative assessment of alteration distribution along the two boreholes also completed the geochemical characterization of the in-situ weathering located within the full-scale waste-rock pile.  The specific research objectives were sometimes overlapping between multiple chapters and aimed to:   Identify the presence of mechanisms resulting from alteration of minerals such as the precipitation of secondary minerals, elemental sorption, armoring of minerals, or galvanic interaction and relate them to potential impact on drainage quality (Chapters 2 and 4).   Complement the current geochemical processes with mineral reactivity indices based on the bulk mineralogical parameters of the liberation and association of minerals to enhance 115  the conceptual models of element inhibition or mobilization related to weathering (Chapters 2 and 4).  Demonstrate that mineralogical data, collected with automated mineralogy, provides useful information to evaluate the reactivity of minerals and explains discrepancies between waste rock and drainage compositions (Chapters 2 and 4).  Explore the connections between hydrological processes, mineralogical weathering textures, and drainage chemistry in weathered field barrels, as well as the application of these textures as qualitative indicators of geochemical or hydrological processes (Chapter 3).   Investigate the heterogeneity and weathering evolution of waste-rock samples from field barrels or full-scale waste-rock pile and relate their characterized mineralogical composition, weathering textures, or mineral reactivity with the corresponding leachate or drainage quality (Chapters 2, 3 and 4).   Corroborate the observed mineralogical, geochemical, or hydrological processes with the mineral weathering rates simulated from the long-term drainage composition of field barrels (Chapters 2 and 3). 5.1 Main findings and contributions The Chapter 2 investigated the relationship between mineral reactivity and drainage quality, whereas Chapter 3 additionally focused on the impact of hydrology, both at small scales. The Chapter 4 inferred, at larger scales, the presence of the same processes observed at smaller scales in Chapter 2 and 3. Weathering rates are difficult to upscale, however geochemical and mineralogical features, observed at small scales, were also observed at larger scales. Similar processes affecting or controlling reactivity can therefore be inferred from smaller to larger scales. 116  5.1.1 Identification of physico-chemical and mineralogical controls on drainage chemistry The drainage composition of field barrels measured the evolution of solutes from weathered waste rock over a period of seven years. In addition, the leachate of waste rock collected in the full-scale waste-rock pile measured the solutes readily dissolved from selected weathered and unweathered intervals. The micro-scale mineralogical analyses identified that processes such as sorption, secondary mineral formation, mineral surface passivation, and galvanic interactions contributed, to some extent, to the resulting drainage quality in combination with hydrological regime, drainage pH, and bulk geochemical properties. For instance, discrete secondary mineral controls on Cu and Zn mobilization were observed with circumneutral and acidic drainage in the field barrels (Chapter 2). Sulfate, Cu, and Zn were leached, whereas As and Fe were retained from altered reactive zones of the full-scale waste-rock pile (Chapter 4). The Cu-Zn effective leaching appeared to be facilitated by galvanic dissolutions observed in the field barrels and waste-rock pile samples (Chapter 2 and 4). In contrast, As and Fe retention was attributed to sorption, surface passivation, or secondary mineral formation (Chapter 4). In addition, As and Mo were released from altered layers of the reactive zone, suggesting that surface liberation of mineral source might control the leaching behavior. The information collected on elemental distribution, liberation, as well as mineral associations identifies the mineralogical controls on drainage chemistry, such as sorption, surface passivation, and galvanic reactions. In addition to hydrological regime and bulk geochemical properties, these processes show evidence of non-proportional mass balance of element concentrations between solid phase and drainage. Figure 5.1 summarizes the relationship between these processes and the drainage chemistry observed from samples of field barrels and full-scale waste-rock pile in Chapters 2 to 4. These information on the processes and mobility of 117  metals within the waste-rock pile provide important insights for the long-term management of the waste-rock drainage quality. For instance, the relationship between the mobility of As or Mo with primary mineral liberation and secondary mineral precipitation and sorption may contribute to find better strategy to prevent leaching of those metalloids that are difficultly removed with treatment plant. 5.1.1.1 Hydrological paths The hydrological flow-paths influence the residence time of water and interaction between water and mineral surfaces, which ultimately control the reactivity of minerals and other processes controlling the drainage chemistry. The higher proportion of fast-matrix flow paths, bulk geochemical composition of the waste rock, circumneutral pH, and the accumulation of more stable secondary minerals composed of less mobile elements might have promoted lower mobility of Fe, Cu, and Zn from marble, intrusive, and skarn field barrels 1A, 2A, and 3A (Chapter 3). In contrast, the higher proportion of slow-matrix flow paths, bulk geochemical composition of the waste rock, acidic pH from 2.5 to 4.5, and the accumulation of acid-soluble secondary minerals composed of more mobile elements might have promoted higher mobility of Fe, Cu, and Zn from intrusive field barrels 2B and 2C (1A and 1B in Chapter 2). In addition, the intrusive field barrels 2B and 2C (named 1A and 1B in Chapter 2) have similar fractions of particle size distribution < 2 mm and develop long-term acidic drainage. However, despite higher metal content and slightly lower NPR, field barrel 2B had a slower development of acidic conditions and lower long term Fe-drainage concentration and mass loading. Such discrepancy could be attributable to higher proportion of fast-matrix flow paths in field barrel 2B resulting in decreased contact time and lower concentration of dissolved species in the drainage. The higher proportion of fast matrix-flow paths from field barrel 2B (i.e. 1A in chapter 2) were associated with higher accumulation of more stable 118   Figure 5.1 Relationship between physico-chemical processes and the drainage chemistry observed from samples of the field barrels and full-scale waste-rock pile in Chapters 2 to 4 (FB: Field barrel, C2: Chapter 2, C3: Chapter 3). 119  secondary minerals, which might have resulted in passivation of pyrite surfaces and lower surface liberation and reactivity indices. This agrees with lower calibrated rate constants of pyrite in field barrel 2B over 2C (named 1A and 1B in Chapter 2) observed in simulations from Chapters 2 and 3 performed with PHREEQC. 5.1.1.2 Passivation of mineral surface The passivation of mineral surface was observed in waste rock samples from the field barrels and the full-scale waste-rock pile in Chapters 2 and 4. For example, in the skarn field barrels 2A and 2B (Chapter 2) pyrite was associated with various secondary minerals and confirmed by observations of amorphous Fe-Si-oxide alteration products enriched in Cu and Zn coating at the surface of sulfide grains. Concentration of Zn and Cu were significantly higher in drainage of field barrel 2B compared to 2A despite higher grades in waste rock of 2A. This relative retention of Zn and Cu can’t rely solely on smaller reactive size fractions or slightly higher pH-neutralizing mineral content. It may also be explained by lower sulfide mineral liberation reflecting passivation of sphalerite and chalcopyrite in field barrel 2A against 2B (Chapter 2). In addition, Fe-precipitates and pyrite passivation contributed to maintain low Fe-drainage concentration in circumneutral and slightly acidic drainage in field barrels 1A, 2A, and 2B, despite elevated Fe content in the waste rock. Pyrite passivation was illustrated by its association with Fe-(oxyhydr)oxides as well as lower Fe-sulfide liberation or reactivity indices, and observed by SEM imaging of pyrite alteration rims. In chapters 2 and 3, higher simulated calibrated rate constants for pyrite and sphalerite agree with higher proportion of fast-matrix flow and/or higher liberation and mineral reactivity indices in field barrels 1B over 1A and 2B over 2A. In Chapter 4, the occlusion of sulfide surfaces by secondary precipitates were observed in specific zones of both boreholes drilled in the waste rock pile. This was supported by strong 120  association between pyrite and secondary oxidation products combined with lower sulfide reactivity index, and SEM images of sulfides occluded by coatings. In addition, inverse relationship between pyrite reactivity and the abundance of Fe-, Cu-, or Zn- bearing secondary phases were observed in selected intervals and confirmed that surface passivation plays an important role in reducing pyritic reactivity. Finally, the circumneutral pH leachate, the effective precipitation of amorphous secondary Fe-minerals, and lower sulfide reactivity index are reflected in non-detect Fe concentrations in most leachates despite up to 8 wt% Fe-sulfide content in the waste rock. 5.1.1.3 Secondary minerals and sorption Under circumneutral and slightly acidic conditions, sorption and secondary mineral precipitation played an important role in controlling the drainage quality and were observed in waste rock samples from the field barrels and the full-scale waste-rock pile in Chapters 2 and 4. For instance, in the skarn waste-rock field barrel 2A with low concentration of Cu- and Zn in the drainage, the main liberated minerals were silicates and Fe-oxides, which are poorly soluble at neutral pH and provide potential Cu- and Zn-sorption sites. The contribution from Zn- or Cu- sorption or co-precipitation with amorphous secondary minerals were supported by SEM observations of Zn- and Cu- enrichments in alteration rims of sulfides of field barrels 1A, 1B, and 2A (Chapter 2).  In Chapter 4, the circumneutral pH of the leachate from the boreholes of the full-scale waste-rock pile, generally do not promote desorption of metal(loid)s from Fe-bearing minerals or dissolution of more stable phases. However, the absence of Fe-mineral substrate may be responsible for the mobility of Cu, Zn, As, or Mo. This may explain preferential retention of those elements within altered layers enriched in Fe-phases and release from non-altered layers with 121  limited Fe-substrate. Yet, As, Mo, Cu, and Zn were also released from altered layers of the reactive zone, suggesting that other mechanisms might control the leaching behavior such as galvanic reactions or surface liberation. For instance, more than 75% of As and Mo bearing-minerals were indeed present in more soluble and/or liberated minerals.  5.1.1.4 Galvanic Reactions The GalvanoxTM process uses galvanic reactions between pyrite and chalcopyrite to increase recuperation of Cu from chalcopyrite-rich concentrates (Dickson et al., 2008). This process demonstrates that galvanic reactions can simply occur through contact with added pyrite, without additional grinding or introduction of bacteria or other chemicals (Dickson et al., 2008). The development of automated mineralogy provided a novel way to identify the association of sulfide minerals that could potentially lead to the presence of galvanic reactions within waste-rock samples from the field barrels and in the full-scale waste rock pile. In chapter 2, calculated association index indicated associations between sulfide minerals in the waste rock. Elemental mapping and SEM images validated signs of preferential galvanic alteration between non-altered pyrite surfaces coupled with altered chalcopyrite with etch pits. The presence of galvanic reaction might also preferentially release Cu and Zn and retained Fe in the drainage of field barrel 2B compared with 2A or field barrel 1A against 1B (chapter 2). The effect of galvanic reactions is added to higher sulfide mineral liberation and lower secondary minerals precipitation identified in field barrel 2B or higher Cu- waste rock content and precipitation of less soluble phases in 1A. Despite faster matrix flow component and slightly higher pH as well as similar reactivity indices and particle size fraction < 2mm, the field barrel 1A had similar long-term Cu- and Zn- concentrations than 1B. Yet, the association index of pyrite with chalcopyrite and sphalerite was slightly higher in field barrel 1A, which support some controls from galvanic reactions. This is 122  corroborated by faster calibrated weathering rates of chalcopyrite and lower rates of pyrite from intrusive field barrels 1A compared with 1B (respectively 2B and 2C in chapter 3) from simulations performed in PHREEQC in chapters 2 and 3. The field barrels 1A have slightly higher pH and flow regimes than 1B, which indicate that liberation and/or galvanic reactions might alter dissolution rates of sulfides. In addition, a significant lower depletion of pyrite compared with high depletion of sphalerite and observed association between pyrite and sphalerite in skarn field barrel 2B suggested the occurrence of galvanic inhibition of pyrite and dissolution of sphalerite.     Similar observations of potential galvanic reactions and their impact on drainage chemistry were also made in the full-scale waste rock pile samples (Chapter 4). Selected intervals with elevated sulfide content did not necessarily correspond to higher secondary mineral precipitation suggesting processes such as oxygen depletion or galvanic inhibition of pyrite. In general, poorly altered layers exhibited less abundant liberated pyrite than chalcopyrite and sphalerite which might have enhanced the galvanic protection of pyrite. In contrast, some altered layers exhibited more abundant liberated pyrite than the other sulfides illustrating partial galvanic protection of pyrite, while remaining relatively reactive. The leachate concentrations from layers with sulfide couples show the relative mobilization of Zn and Cu and retention of Fe at circumneutral pH that suggest galvanic protection of pyrite and preferential dissolution of chalcopyrite and sphalerite. Sulfide coupling were observed with MLA and Raman spectroscopy in multiple locations within the boreholes. In addition, the anomalous abundance of secondary Cu and Zn adsorbed on silicates and Fe substrates as well as the limited surface liberation through precipitation of stable secondary mineral agglomerated around sulfides observed with MLA might be the result of galvanic dissolutions.    123  5.1.2 Quantification of mineral reactivity and internal weathering processes   Field data on mineral reactivity and secondary mineral formation are not commonly collected, which makes it challenging to quantitatively incorporate them into prediction models (Lefebvre et al., 2001; Linklater et al., 2005; Molson et al., 2005; Pedretti et al., 2017). In this study, mineral reactivity was deduced from mineralogical data such as composition, association, and liberation, as a complement to physical and geochemical data collected from samples in field barrels and full-scale waste-rock piles. The reactivity, association and galvanic coupling indices quantified the reactivity of minerals based on association and liberation data, which in turn inferred processes of mineral passivation and galvanic interactions and implied drainage quality controls. For example, field barrels of intrusive waste-rock composition with low neutralization potential, Fe-, Cu-, or Zn-bearing sulfide minerals with >75% liberated surfaces, and, galvanic coupling of pyrite with chalcopyrite and sphalerite led to more effective mobilization of Cu, Fe, and Zn (Chapter 2). By comparison, less effective mobilization of Cu, Fe, and Zn was observed in waste rock of skarn composition with high neutralization potential, in Cu-, Fe-, and Zn-bearing sulfide minerals with <70% liberated surfaces, and where there were occurrences of passivation on pyrite surfaces (Chapter 2). In addition, the differences in drainage chemistry between two intrusive field barrels with similar waste-rock metal content and neutralization potentials suggested additional controls from the mineral reactivity influenced by water-mineral contact time, surface liberation and associations (Chapters 2 and 3).  Similarly, the observations of anomalous alteration heterogeneities and small-scale contacts along the boreholes were linked to mineral liberation and associations (Chapter 4). For instance, sulfide passivation from secondary mineral precipitation may result in reduced sulfide reactivity, whereas galvanic inhibition of pyrite may partially explain the lack of alteration in 124  reactive zones (Chapter 4). Overall, elevated reactivity indices for pyrite and other sulfides as well as galvanic couple indices corresponding to the limited galvanic inhibition of pyrite were consistent with the reactive zones identified in the full-scale waste-rock pile from gas and temperature observations (Chapter 4). Altogether, in field barrel and borehole samples, these indices illustrate that besides the bulk geochemical reactivity based on metal content and neutralization potential, the mineral surface exposure, passivation, or association can contribute additional information on waste-rock reactivity at two different scales (Chapter 2 and 4). These quantitative mineralogical evaluations may easily be applied to other mine sites and therefore facilitate the interpretation of waste-rock reactivity, internal weathering processes, and contribute to improved drainage prediction models. 5.1.3 Application of quantitative automated mineralogy for characterization of mine wastes To date quantitative automated mineralogy has not been widely applied in the context of mine waste-rock characterization (Dold, 2017; Lapakko, 2015). Also, even though mineral texture, liberation and arrangement have shown to critically define mineral reactivity (Linklater et al., 2005; Brough et al., 2017; Pedretti et al., 2017; Parbhakar-Fox et al., 2018), only a few waste-rock studies quantified mineralogical and petrographical characteristics and evaluated their role within large-scale systems (Parbhakar-Fox and Lottermoser, 2015; Dold, 2017). Selected samples from the field barrels and boreholes from the full-scale waste-rock pile were analyzed with MLA. The weathering and reactivity of waste rocks were characterized using mineral composition, liberation, and association. This information was useful to complement bulk geochemistry with information on processes such as sorption, surface passivation and galvanic reactions in relation to drainage chemistry. Bulk geochemical analyses complemented by mineralogical composition, crystallinity, 125  impurities, and degree of liberation would significantly improve prediction models of drainage quality. With the application of widespread automated analytical investigations across all disciplines at mine sites, such as metallurgy and ore processing (Brough et al., 2013, Parbhakar-Fox et al., 2018), the automated mineralogical data may be collected cost-effectively at larger scales. As a result, the interpretation of mineralogical textures, as applied in this study, may further optimize storage conditions and drainage predictions at mine sites.  5.1.4 Description of physicochemical heterogeneity at field barrel and waste-rock pile scales Mines generally produce low-grade waste rock of heterogeneous composition that are stored in large-scale waste-rock piles. In contrast, the mineralogical parameters are generally quantified at small scales in homogeneous settings (Parbhakar-Fox et al., 2013; Pedretti et al., 2017) and tend to focus on single elements and mineralogical/lithological rock classes with elevated metal content (Price, 2009). As a result, the effect of physicochemical heterogeneity on the mineralogical and drainage composition measured in the field and at larger scales remains poorly understood. The kinetic tests in field barrels captured long-term waste-rock weathering and drainage chemistry evolution under field conditions. For instance, different drainage chemistries resulted from different waste rock types (i.e. intrusive versus skarn) as well as geochemically similar intrusive waste rock (Chapter 2). The characterization of distinct mineralogical reactivities amongst field barrels of comparable composition explained the cause of those drainage discrepancies (Chapter 2).  Unfortunately, small-scale kinetic tests do not capture all significant interactions related to heterogeneities or physicochemical composition (Vriens et al., 2019b) found in full-scale waste-rock piles. The heterogeneity, in-situ conditions, and long-term behavior of weathering in these 126  complex storage facilities need to be assessed through dedicated studies in large-scale waste-rock piles (Ritchie, 1994; Molson et al., 2005; Pedretti et al., 2017; Vriens et al., 2018). The reactive zones, previously defined by in-situ pore-gas and temperature data, were largely confirmed by geochemical and mineralogical data collected as part of this study (Chapter 4). Sharp and progressive contact between altered and non-altered layers were observed along two boreholes drilled in the waste-rock pile (Chapter 4). These meter to centimeter-wide layers indicated small-scale heterogeneities likely caused by variations in moisture content and capillary barriers created by cementation or traffic surfaces, as well as strong variability of mineral reactivity. In addition, the random end-dumping of waste rock may create conditions unfavorable to weathering as was observed when parts of sulfide-rich layers of the reactive zones were poorly altered. Low oxygen levels, surface passivation, and galvanic interactions between sulfide couples limited alteration within these high-sulfide zones. This study provided unique insights into the material distribution and weathering processes throughout a hundred-meter tall heterogeneous waste-rock pile (Chapter 4). For example, the genetic layering resembled those observed in hardpan formations of tailings and could possibly occur below compacted traffic surfaces, even at significant depths in waste-rock piles, and are therefore not restricted to less permeable layers of tailings. Static laboratory testing or sensor measurements alone could not have revealed this high degree of physicochemical heterogeneity resolved by the high-resolution sampling and analyses from this study (Chapter 4). Comparable processes are occurring at different scales; equivalent preferential attenuation and release of metals originating from similar lithological compositions and the presence of acidic micro-environments were observed in smaller and larger scale experiments. In summary, the weathering and drainage of reactive zones observed within heterogeneous waste-rock piles may develop patterns of: i) local alterations based on bulk geochemistry; ii) oxygen depletion extending 127  beyond mineralogical variations from micrometer to meter scales; and, iii) aggregate chemical leachate signatures resulting from mixed drainage.  5.1.5 Application of mineralogical weathering textures as qualitative indicators of hydrological responses, complementing tracer test results and drainage chemistry Information gathered in the context of weathered waste-rock characterization rarely combine together interpretations from flow paths, mineralogical content, and weathering textures. At Antamina, the mineralogical composition of the waste rock has been substantially described; yet, up until now, very little information on mineralogical textures of weathering has been gathered. Tracer tests are widely applied to measure residence times of water to gain valuable knowledge of the hydrological responses of a system (Eriksson et al., 1997; Neuner et al., 2013; Peterson, 2014; Blackmore et al., 2014 and 2018). However, their application can be expensive and onerous (Shook et al., 2004) and the results difficult to up-scale (Eriksson and Destouni, 1997; Stockwell et al., 2006; Blackmore et al., 2014; Vriens et al., 2019). The characterization of mineralogical content, weathering textures, and drainage chemistry, from field barrels exposed to weathering over a period of seven years, provided insights on the development of water regimes and waste-rock weathering (Chapter 3). For example, field barrels with dominant preferential or fast-matrix flow paths had: i) shorter tracer residence times; ii) higher content of secondary minerals composed of less mobile elements with boxwork textures; and, iii) lower drainage concentrations/loads. In contrast, slower-matrix flow in field barrels had longer residence times, more abundant secondary minerals composed of moderately mobile elements with infilling textures, and higher drainage concentrations/loads. This study demonstrates the interrelationship between hydrological regimes, mineralogical composition, and weathering textures through the characterization of flow paths, secondary mineral products, and mineral weathering textures in 128  field barrels samples. This characterization provides an initial qualitative assessment of the hydrological and geochemical processes ultimately controlling the long-term drainage quality of these field barrels. The application of hydrological indicators, in more complex systems such as full-scale waste-rock piles, could give an initial overview of these processes and/or complement tracer test results and be a starting point for more detailed investigations. 5.1.6 Simulated mineral weathering rates quantitatively supported hydrological, physicochemical, and mineralogical controls on drainage chemistry  The long-term drainage composition of field barrels was simulated using reactive transport modelling and calibrated with drainage composition from field data (Chapters 2 and 3). The effects of water residence times and mineral reactivity were investigated by performing a quantitative comparison between calibrated model rate constants with quantitative liberation and association indices or with a proportion of fast and slow-flow paths in the field barrels. The mineral reactivity quantified by mineralogical parameters of liberation and associations supported the calibrated mineral rate constant obtained from reactive transport model simulations. For example, lower calibrated rate constants for sulfides agreed with mineralogical data showing lower surface liberation suggesting the occurrence of mineral passivation in skarn field barrel 2A (Chapter 2). While, the association of sulfide couples in combination with lower simulated pyrite weathering rates suggested the potential galvanic inhibition of pyrite in skarn field barrel 2B (Chapter 2). The Damköhler number, representing the relationship between transport (i.e. residence time from tracer tests) and reactivity (i.e. calibrated rate constant from simulation), was also consistent with the dominant flow paths and other indicators of hydrological responses such as secondary mineral composition and/or weathering textures (Chapter 3). For instance, the calibrated sulfide dissolution rates linked higher calibrated mineral weathering rates with higher Damköhler and dominant slow-129  matrix flow paths. In summary, the effective rate-constants represented an implicit site-specific correction for mechanisms that alter dissolution rates and illustrated variations in mineral reactivity caused by water rock interactions, the availability of reactive surfaces, or galvanic interactions. These calibrated dissolution rates provide a site-specific correction and capture important distinctions between and within material compositions, as well as mineralogical and hydrological properties. Reactive transport simulations therefore provide valuable understandings of the potential role of these hydrological and mineralogical controls in the context of waste-rock weathering. 5.2 Recommendations for future research directions This study improves knowledge of geochemical and hydrological processes by adding perspectives on mineral reactivity in association with waste-rock drainage chemistry, and is ultimately improving predictions of drainage quality. However, this study has some limitations and further refinements are recommended concerning: i) the representability of samples and interpretation of processes at large scales; ii) the level of complexity of the chosen reactive transport model as well as the estimation of parameters used in the model; and, iii) the assessment of alteration in the waste-rock pile and relative importance of processes. 5.2.1 Sampling and representability Even though good care was taken to achieve representative samples from quartering, the interpretation of processes was deduced from mineralogical observations of samples representing only an infinitesimal portion of the material from the field barrels and full-scale waste-rock pile. Consequently, additional samples should be characterized to achieve more accurate statistical representability of the processes identified in this study. In addition, we recommend further quantitative characterization of the particle size distribution of the waste rock collected from the 130  boreholes to support the qualitative observations made in this study. The identification of phases with the mineral liberation analyzer is limited by phases previously identified in the library. It would be recommended to further expand the number of phases recorded in this library by identifying additional unknown phases and crosscheck them with other mineral analysis tools. Finally, in order to avoid errors related to the averaging of concentrations, the composite samples used to assess drainage quality should be taken on a regular basis and should not extend over the capacity of the sampling containers.  5.2.2 Reactive transport modeling The effective dissolution rate-constants obtained from the calibration of field data are site-specific and would not be representative of other mine sites. However, the calibration methods and the implied processes could be applied at other sites. The drainage chemistry predictions of long-term waste-rock weathering could be further optimized by incorporating the indices of galvanic interactions, occlusion, or mineral textures into effective rate constants, rate expressions and entire reactive-transport simulations (Chapter 2). However, the occurrence of other important parameters such as microbial conditions and particle sizes that could affect weathering rates, as well as the interpretation of weathering mechanisms, were not assessed in this study and should be further investigated (Chapter 2 and 3). Non-equilibrated aqueous solutions, irregularity or incomplete representations of the hydrological regime, and composite samples are limiting factors explaining discrepancies observed between simulated and measured drainage chemistry (Chapter 2 and 3). Better representation of the parameters in the reactive transport model as well as exploration of other multi-component reactive transport models is therefore recommended. In addition, in order to better quantify the contribution of transport versus reaction rates using the Damkölher number, 131  it is recommended to better target each process by performing kinetic tests (e.g. humidity cells) in laboratory settings (Chapter 3). 5.2.3 Characterization of alteration in the full-scale waste-rock pile  Recommendations are proposed for future research to further characterize the alteration in a full-scale waste-rock pile and the relative importance of processes from Chapter 4:  Agglomerated particles were observed in the altered zones of the waste-rock pile. As a result, they were hypothesized to reduce the porosity of the oxidized layers based on observations from previous studies. Thus, additional characterization of porosity in these altered zones using automated mineralogy or other characterization tools is recommended.    The contribution of physical constraints (i.e. low oxygen or water supply) explaining the occurrence of inert zones, where minimal alteration was observed in sulfide rich layers, should be further quantified as part of future research.  Further investigate controls of flow path on alteration, weathering textures, and drainage chemistry within the waste-rock pile by applying the qualitative indicators of hydrological responses and complement with tracer test.  It is recommended to further investigate the distinction between the roles and the relative importance of passivation, galvanic interactions, and adsorption in altered layers. This would quantify to what extent the mineralogical parameters, such as reactivity indices based on mineral liberation, reflect the effect of passivation from secondary mineral precipitation on weathering rates and the resulting drainage quality.  132  Bibliography Acero, P., Cama, J., Ayora, C., Asta, M.P., 2009. Chalcopyrite dissolution rate law from pH 1 to 3. Geol. Acta 7, 389–397. doi.org/10.1344/105.000001444 Al, T.A., Blowes, D.W., Martin, C.J., Cabri, L.J., Jambor, J.L., 1997. Aqueous geochemistry and analysis of pyrite surfaces in sulfide-rich mine tailings. Geochim. Cosmochim. Acta 61, 2353–2366. doi.org/10.1016/S0016-7037(97)00113-0 Al, T.A., C.J. Martin, D.W., Blowes, 2000. Carbonate-mineral / water interactions in sulfide-rich mine tailings. Geochim. Cosmochem. Acta 64, 3933–3948. doi.org/10.1016/S0016-7037(00)00483-X Amos, R.T., Blowes, D.W., Bailey, B.L., Sego, D.C., Smith, L., Ritchie,  A.I.M., 2014. Waste-rock hydrogeology and geochemistry. Appl. Geochemistry 57, 140–156. doi.org/10.1016/j.apgeochem.2014.06.020 Anterrieu, O., Chouteau, M., Aubertin, M., 2010. Geophysical characterization of the large-scale internal structure of a waste rock pile from a hard rock mine. Bull. Eng. Geol. Environ. 69, 533–548. doi: 10.1007/s10064-010-0264-4 Aranda, C.A., Klein, B., Beckie, R.D., Mayer, K.U., 2009. Assessment of waste rock weathering characteristics at the Antamina mine, in: ICARD. Skelleftea, Sweden. Aranda, C., 2010. Assessment of waste rock weathering characteristics at the Antamina Mine based on field cells experiment. (Master’s thesis) The University of British Columbia. Vancouver, Canada. ASTM, 2010. Standard Test Methods for Laboratory Determination of Water (Moisture) Content of Soil and Rock by Mass. D2216-10, 1–7. doi.org/10.1520/D2216-10.N Bailey, B.L., Blowes, D.W., Smith, L., Sego, D.C., 2016. The Diavik waste rock project: 133  geochemical and microbiological characterization of low sulfide content large-scale waste rock test piles. Appl. Geochem. 65, 54–72. doi.org/10.3390/min4020257 Baker, B.J., Banfield, J.F., 2003. Microbial communities in acid mine drainage. FEMS Microbiol. Ecol. 44 (2), 139–152. doi: 10.1016/S0168-6496(03)00028-X Ball, J.W., Nordstrom, D.K., 1991. User’s Manual for WATEQ4F, with revised thermodynamic data base and test cases for calculating speciation of major, trace, and redox elements in natural waters. U.S. Geol. Surv. Water-Resources Investig. Rep. 91–183, 1–188. Banks, D., Parnachev, V.P., Frengstad, B., Holden, W., Vedernikov, A.A., Karnachuk, O.V., 2002. Alkaline mine drainage from metal sulphide and coal mines; examples from Svalbard and Siberia. Mine water Hydrogeol. geochemistry 198, 287–296. doi:10.1144/GSL.SP.2002.198.01.19 Banks, D., Younger, P., Arnesen, R.T., Iversen, E.R., Banks, S.B., 1997. Mine-water chemistry : the good, the bad and the ugly. Environmental Geology 32, 157-174. doi.org/10.1007/s002540050204 Barazzuol, L., Sexsmith, K., Bucknam, C., Lopez, D., 2012. Application of an Advanced Mineralogical Technique : Sulphide Mineral Availability and Humidity Cell Interpretations based on MLA Analysis, in: ICARD, Ottawa, Canada. Baum, W., 2014. Ore characterization, process mineralogy and lab automation a roadmap for future mining. Miner. Eng. 60, 69–73. doi: 10.1016/j.mineng.2013.11.008 Bay, D.S., Peterson, H.E., Singurindy, O., Aranda, C., Dockrey, J.W., Vargas, F.S., Gallegos, J.C.C., Mayer, K.U., Smith, L., Beckie, R.D., 2009. Assessment of Neutral pH Drainage from Three Experimental Waste-Rock Piles at the Antamina Mine, Peru, in: Securing the Future and 8th International Conference on Acid Rock Drainage, Skellefteå, Sweden, p. 11. 134  Blackmore, S., Smith, L., Ulrich Mayer, K., Beckie, R.D., 2014. Comparison of unsaturated flow and solute transport through waste rock at two experimental scales using temporal moments and numerical modeling. J. Contam. Hydrol. 171, 49–65. doi: 10.1016/j.jconhyd.2014.10.009 Blackmore, S., 2015. The role of hydrology, geochemistry and microbiology in flow and solute transport through highly heterogeneous, unsaturated waste rock at various test scales. (Doctoral dissertation) University of British Columbia. Vancouver, Canada. Blackmore, S., Pedretti, D., Mayer, K.U., Smith, L., Beckie, R.D., 2018a. Evaluation of single- and dual-porosity models for reproducing the release of external and internal tracers from heterogeneous waste-rock piles. J. Contam. Hydrol. 214, 65–74. doi.org/10.1016/j.jconhyd.2018.05.007 Blackmore, S., Vriens, B., Sorensen, M., Power, I.M., Smith, L., Hallam, S.J., Mayer, K.U., Beckie, R.D., 2018b. Microbial and geochemical controls on waste rock weathering and drainage quality. Sci. Total. Environ. 640-641, 1004-1014. doi: 10.1016/j.scitotenv.2018.05.374 Blaskovich, R., 2013. Characterizing waste rock using automated quantitative electron microscopy. (Master’s thesis) The University of British Columbia. Vancouver, Canada. Blowes, D.W., Jambor, J.L., 1990. The Pore-Water Geochemistry And The Mineralogy Of The Vadose Zone Of Sulfide Tailings, Waite-Amulet, Quebec, Canada. Appl. Geochemistry 5, 327–346. doi.org/10.1016/0883-2927(90)90008-S Blowes, D.W., Al, T., Lortie, L., Gould, W.D., Jambor, J.L., 1995. Microbiological, chemical, and mineralogical characterization of the Kidd Creek mine tailings impoundment, Timmins area, Ontario. Geomicrobiol J. 13, 13–31. doi.org/10.1080/01490459509378001 Blowes, D.W.; Ptacek, C.J.; Jambor, J.L.; Weisener, C.G., 2003. The Geochemistry of Acid Mine 135  Drainage. Treatise on Geochemistry 9, p.149-204. doi.org/10.1016/B0-08-043751-6/09137-4 Bouzahzah, H., Benzaazoua, M., Bussiere, B., Plante, B., 2014. Prediction of Acid Mine Drainage: Importance of Mineralogy and the Test Protocols for Static and Kinetic Tests. Mine Water Environ. 33, 54–65. doi.org/10.1007/s10230-013-0249-1 Brough, C.P., Warrender, R., Bowell, R.J., Barnes, A., Parbhakar-fox, A., 2013. The process mineralogy of mine wastes 52, 125–135. doi.org/10.1016/j.mineng.2013.05.003 Brough, C., Strongman, J., Bowell, R., Warrender, R., Prestia, A., Barnes, A., Fletcher, J., 2017. Automated environmental mineralogy; the use of liberation analysis in humidity cell testwork. Miner. Eng. 107, 112–122. doi.org/10.1016/j.mineng.2016.10.006 Buckwalter-davis, M., Jaggard, H., Jamieson, H., 2012. Automated Mineralogy of Mine Tailings Using Mineral Liberation Analysis Automated mineralogy Case Study : Weathered tailings from the New Calumet Pb-Zn mine, in: ICARD, Ottawa, Canada. Cebeci, A.T., Tong, L., Klein, B., 2016. Regression analysis of XRF and electromagnetic sensor responses for improved sorting of a gold ore. In XXVIII International Mineral Processing Congress (IMPC 2016), Quebec, Canada. Chopard, A., Plante, B., Benzaazoua, M., Bouzahzah, H., Marion, P., 2017. Geochemical investigation of the galvanic effects during oxidation of pyrite and base-metals sulfides. Chemosphere 166, 281–291. doi.org/10.1016/j.chemosphere.2016.09.129 Colombani, J., 2008. Measurement of the pure dissolution rate constant of a mineral in water. Geochim. Cosmochim. Acta 72, 5634–5640. https://doi.org/10.1016/j.gca.2008.09.007 Conlan, M.J.W., 2009. Attenuation Mechanisms For Molybdenum In Neutral Rock Drainage. (Master’s thesis) The University of British Columbia. Vancouver, Canada. 136  Cravotta III, C.A, Trahan, M.K., 1999. Limestone drains to increase pH and remove dissolved metals from acidic mine drainage. Appl. Geochemistry 14, 581–606. doi.org/10.1016/S0883-2927(98)00066-3 Davies, H., Weber, P., Lindsay, P., Craw, D., Peake, B., Pope, J., 2011. Geochemical changes during neutralisation of acid mine drainage in a dynamic mountain stream, New Zealand. Appl. Geochemistry 26, 2121–2133. doi: 10.1016/j.apgeochem.2011.07.010 Day, S.; Kuit, W.; Filion, M., 2003. Evolution of Mine Drainage Chemistry, Red Dog Mine, Alaska, in: Proceeding of the 10th Annual British Columbia MEND ML/ARD Workshop Vancouver. Vancouver, Canada. De Baere, B., François, R., Mayer, K.U., 2015. Measuring mineral dissolution kinetics using on-line flow-through time resolved analysis (FT-TRA): an exploratory study with forsterite. Chem. Geol. 413, 107–118. doi.org/10.1016/j.chemgeo.2015.08.024 Delvigne, J.E., 1998. Atlas of Micromorphology of Mineral Alteration and Weathering, Special pu. ed. Mineralogical Association of Canada (in collaboration with ORSTOM, Paris), Ottawa, Ontario. Dixon, D., Mayne, D., Baxter, K., 2008. Galvanox™ – a novel galvanically-assisted atmospheric leaching technology for copper concentrates. Canadian Metallurgical Quarterly. 47(3), 327–336. doi:10.1179/000844308794408317 Dockrey, J.W., Lindsay, M.B.J., Mayer, K.U., Beckie, R.D., Norlund, K.L.I., Warren, L.A., Southam, G., 2014. Acidic Microenvironments in Waste Rock Characterized by Neutral Drainage: Bacteria–Mineral Interactions at Sulfide Surfaces. Minerals 4, 170–190. doi.org/10.3390/min4010170 Dold, B., 2017. Acid rock drainage prediction : A critical review. J. Geochemical Explor. 172, 137  120–132. doi.org/10.1016/j.gexplo.2016.09.014 Elghali, A., Benzaazoua, M., Bouzahzah, H., Bussière, B., Villarraga-Gómez, H., 2018. Determination of the available acid-generating potential of waste rock, part I: Mineralogical approach. Appl. Geochemistry 99, 31–41. doi: 10.1016/j.apgeochem.2018.10.021 Elghali, A., Benzaazoua, M., Bussière, B., Bouzahzah, H., 2019a. Determination of the available acid-generating potential of waste rock, part II: Waste management involvement. Appl. Geochemistry 100, 316–325. doi.org/10.1016/j.apgeochem.2018.12.010 Elghali, A., Benzaazoua, M., Bussière, B., Kennedy, C., Parwani, R., Graham, S., 2019b. The role of hardpan formation on the reactivity of sulfidic mine tailings : A case study at Joutel mine (Québec). Sci. Total Environ. 654, 118–128. doi: 10.1016/j.scitotenv.2018.11.066 Eriksson, N., Gupta, A., Destouni, G., 1997. Comparative analysis of laboratory and field tracer tests for investigating preferential flow and transport in mining waste rock. J. Hydrol. 194, 143–163. doi.org/10.1016/S0022-1694(96)03209-X Eriksson, N. and Destouni, G., 1997. Combined effects of dissolution kinetics, secondary mineral precipitation, and preferential flow on copper leaching from mining waste rock. Water Resour. Res., 33: 471–483. doi/10.1029/96WR03466 Escalante, A.; Dipple, G.M.; Barker, S.L.L.; Tosdal, R., 2010. Defining trace-element alteration halos to skarn deposits hosted in heterogeneous carbonate rocks: Case study from the Cu–Zn Antamina skarn deposit, Peru. J. Geochemical Explor. 105, 117–136. doi: 10.1016/j.gexplo.2010.04.011 Evangelou, V.P., 1998. Pyrite Chemistry: The Key for Abatement of Acid Mine Drainage, in: Geller, P.D.W., Klapper, P.D.H., Salomons, P.D.W. (Eds.), Environmental Science. Springer Berlin Heidelberg, pp. 197–222. 138  Evangelou, V.P., Zhang, Y.L., 1995. A review: Pyrite oxidation mechanisms and acid mine drainage prevention. Crit. Rev. Environ. Sci. Technol. 25, 141–199. doi.org/10.1080/10643389509388477 Evans, K.A., Banwart, S.A., 2006. Rate controls on the chemical weathering of natural polymineralic material. I. Dissolution behaviour of polymineralic assemblages determined using batch and unsaturated column experiments. Appl. Geochem.21, 352–376. doi: 10.1016/ j.apgeochem.2005.10.001 Fan, R., Short, M.D., Zeng, S.J., Qian, G., Li, J., Schumann, R.C., Kawashima, N., Smart, R.S.C., Gerson, A.R., 2018. The Formation of Silicate-Stabilized Passivating Layers on Pyrite for Reduced Acid Rock Drainage. Environ. Sci. Technol. 52, 5349–5357. doi : 10.1021/ acs.est.7b03232 Fandrich, R., Gu, Y., Burrows, D., Moeller, K., 2007. Modern SEM-based mineral liberation analysis. Int. J. Miner. Process., Special Issue To Honor The Late Professor R. Peter King 84, 310–320. doi: 10.1016/j.minpro.2006.07.018 Genty, T., Bussière, B., Potvin, R., Benzaazoua, M., Zagury, G.J., 2012. Dissolution of calcitic marble and dolomitic rock in high iron concentrated acid mine drainage: application to anoxic limestone drains. Environ. Earth Sci. 66, 2387–2401. doi.org/10.1007/s12665-011-1464-3 Ghorbani, Y., Becker, M., Petersen, J., Mainza, A.N., Franzidis, J., 2013. Investigation of the effect of mineralogy as rate-limiting factors in large particle leaching. Miner. Eng. 52, 38–51. doi: 10.1016/j.mineng.2013.03.006 Gieré, R., Sidenko, N. V., Lazareva, E. V., 2003. The role of secondary minerals in controlling the migration of arsenic and metals from high-sulfide wastes (Berikul gold mine, Siberia). Appl. Geochemistry 18, 1347–1359. doi: 10.1016/S0883-2927(03)00055-6 139  Golder Associates, 1999. Hydrology design parameters for Antamina Mine Tailings Facility, Antamina Project. Golder Associates, 2004. Final Version 4.0 Waste Rock Geochemistry - Phase 2 Antamina Peru. Golder Associates, 2010. Waste Rock and Tailings Geochemestry - Field Cell Monitoring December 2002 to May 2009. Graupner, T., Kassahun, A., Rammlmair, D., Meima, J.A., Kock, D., Furche, M., Fiege, A., Schippers, A., Melcher, F., 2007. Formation of sequences of cemented layers and hardpans within sulfide-bearing mine tailings (mine district Freiberg, Germany). Appl. Geochemistry 22, 2486–2508. https://doi.org/10.1016/j.apgeochem.2007.07.002 Gu, Y., 2003. Automated Scanning Electron Microscope Based Mineral Liberation Analysis An Introduction to JKMRC/FEI Mineral Liberation Analyser. J. Miner. Mater. Charact. Eng. 02, 33-41. doi: 10.4236/jmmce.2003.21003 Gu, Y., Schouwstra, R.P., Rule, C., 2014. The value of automated mineralogy. Miner. Eng. 58, 100–103. doi: 10.1016/j.mineng.2014.01.020 Hammarstrom, J.M., Sibrell, P.L., Belkin, H.E., 2003. Characterization of limestone reacted with acid-mine drainage in a pulsed limestone bed treatment system at the Friendship Hill National Historical Site, Pennsylvania, USA. Appl. Geochemistry 18, 1705–1721. doi.org/10.1016/S0883-2927(03)00105-7 Harrison, B., Aranda, C., Sanchez, M., Vizconde, J., 2012. Waste Rock Management at the Antamina Mine : Overall Strategy and Data Application in the Face of Continued Expansion. In Proceedings of the 9th International Conference on Acid rock drainage, Ottawa, Canada. Haupt, C., Kabwe, L., Klein, B., Jorge, A., Altun, E., 2011. Assessing weathering in field cell experiments using quantitative mineralogy and sequential leaching. 140  Hirsche, D.T., 2012. A Field Cell and Humidity Cell Study of Metal Attenuation in Neutral Rock Drainage from the Antamina Mine, Peru. (Master’s thesis) The University of British Columbia. Vancouver, Canada. Hirsche, D.T., Blaskovich, R., Mayer, K.U., Beckie, R.D., 2017. A study of Zn and Mo attenuation by waste-rock mixing in neutral mine drainage using mixed-material field barrels and humidity cells. Appl. Geochemistry 84, 114–125. doi.org/10.1016/j.apgeochem.2017.06.005 Horton, J., Butler, L., Overton, D.D., Garcia, D., 2015. Variability of samples from various drilling and sampling methods. In Proceedings Tailings and mine wastes, Vancouver, Canada. Jones, M.P., 1987. Applied mineralogy : a quantitative approach, Graham & T. ed. London. 229p. Huminicki, D.M.C., Rimstidt, J.D., 2009. Iron oxyhydroxide coating of pyrite for acid mine drainage control. Appl. Geochemistry 24, 1626–1634. doi.org/10.1016/j.apgeochem.2009.04.032 Huminicki, D.M.C., Rimstidt, J.D., 2008. Neutralization of sulfuric acid solutions by calcite dissolution and the application to anoxic limestone drain design. Appl. Geochemistry 23, 148–165. doi.org/10.1016/j.apgeochem.2007.10.004 Hunt, J., Berry, R., Bradshaw, D., 2011. Characterising chalcopyrite liberation and flotation potential : Examples from an IOCG deposit. Miner. Eng. 24, 1271–1276. doi: 10.1016/j.mineng.2011.04.016 Jambor, J.L., 1994. Mineralogy of sulfide-rich tailings and their alteration products., in: Jambor, J.L.; Blowes, D.W. (Ed.), The Environmental Geochemistry of Sulfide Mine-Wastes. Mineralogical Association of Canada, Nepean, ON, pp. 59–102. Jambor, J.L., 2003. Mine-waste mineralogy and mineralogical perspectives of acid-base accounting. In Environmental aspects of mine wastes, Jambor, J.L., Blowes, D.W., Ritchie, 141  A.I.M. (Eds.) Short Course Series. Mineralogical Association of Canada, pp. 117–146. Jambor, J.L., Blowes, D.W., Ritchie, A.I.M., 2003. Environmental aspects of mine wastes, Short Cour. ed. Mineralogical Association of Canada. Jamieson, H.E., Walker, S.R., Parsons, M.B., 2015. Mineralogical Characterization of Mine Waste. Appl. Geochemistry. doi.org/10.1016/j.apgeochem.2014.12.014 Johnson, D.B., Hallberg, K.B., 2003. Themicrobiology of acidic mine waters. Res. Microbiol. 154, 466–473. doi.org/10.1016/S0923-2508(03)00114-1 Jones, M.P., 1987. Applied mineralogy : a quantitative approach, Graham & T. ed. London Jurjovec, J., Blowes, D.W., Ptacek, C.J., Mayer, K.U., 2004. Multicomponent reactive transport modeling of acid neutralization reactions in mine tailings. Water Resour. Res. 40, 1–17. doi.org/10.1029/2003WR002233 Kirby, C.S., Cravotta, C. a., 2005. Net alkalinity and net acidity 1: Theoretical considerations. Appl. Geochemistry 20, 1920–1940. doi.org/10.1016/j.apgeochem.2005.07.002 Kuechler, R., Noack, K., Zorn, T., 2004. Investigation of gypsum dissolution under saturated and unsaturated water conditions. Ecol. Modell. 176, 1–14. doi.org/10.1016/ j.ecolmodel.2003.10.025 Kwong, Y.T.J., 1995. Influence of galvanic sulfide oxidation on mine water chemistry, in: Sudbury ’95 Mining and the Environment, Sudbury, Canada, p.477 Kwong, T.J., Roots, C.F., Roach, P., Kettley, W., 1997. Post-mine metal transport and attenuation in the Keno Hill mining district, central Yukon, Canada. Environ. Geol. 30, 98-107. doi.org/10.1007/s002540050137 Kwong, Y.T. J., Swerhone, G.W., Lawrence, J.R., 2003. Galvanic sulphide oxidation as a metal-leaching mechanism and its environmental implications. Geochemistry Explor. Environ. 142  Anal. 3, 337–343. doi.org/10.1144/1467-7873/03/013 Laetsch T, D.R.T., 2006. Software For Identification and Refinement of Cell Parameters From Powder Diffraction Data of Minerals Using the RRUFF Project and American Mineralogist Crystal Structure Databases., in: Abstracts from the 19th General Meeting of the International Mineralogical Association, Kobe, Japan, 23-28 July 2006. Lafuente, B., Downs, R.T., Yang, H., Stone, N., 2015. The power of databases: the RRUFF project.  In Highlights in Mineralogical Crystallography, Walter de Gruyter GmbH. 29p. doi.org/10.1515/9783110417104-003 Langmuir, D., 1997. Aqueous Environmental Geochemistry. Prentice Hall, Upper Saddle River, N.J. Lahmira, B., Lefebvre, R., 2015. Numerical modelling of transfer processes in a waste rock pile undergoing the temporal evolution of its heterogeneous material properties. Int. J. Mining, Reclam. Environ. 29, 1–22. doi.org/10.1080/17480930.2014.889362 Lapakko, K., 2002. Metal Mine Rock and Waste Characterization Tools : An Overview. 29p. http://pubs.iied.org/pdfs/G00559.pdf Lapakko, K. a., 2015. Preoperational assessment of solute release from waste rock at proposed mining operations. Appl. Geochemistry 57, 106–124. doi.org/10.1016/j.apgeochem.2015.01.010 Lasaga, A.C., Soler, J.M., Ganor, J., Burch, T.E., Nagy, K.L., 1994. Chemical weathering rate laws and global geochemical cycles. Geochim. Cosmochim. Acta 58, 2361–2386. doi.org/10.1016/0016-7037(94)90016-7 Lastra, R., 2007. Seven practical application cases of liberation analysis. Int. J. Miner. Process. 84, 337–347. doi.org/10.1016/j.minpro.2006.07.017 143  Lastra, R., Paktunc, D., 2016. An estimation of the variability in automated quantitative mineralogy measurements through inter-laboratory testing. Miner. Eng. 95, 138–145. doi: 10.1016/j.mineng.2016.06.025 Lawrence, R.W., Wang, Y., 1996. Determination of neutral potential for acid rock drainage prediction, in: MEND Project report 1.16.3. p. 89. Lefebvre, R., Hockley, D., Smolensky, J., Gélinas, P., 2001. Multiphase transfer processes in waste rock piles producing acid mine drainage. 1: Conceptual model and system characterization. J. Contam. Hydrol. 52, 137–164. doi.org/10.1016/S0169-7722(01)00156-5 Lipten, E.J., Smith, S.W., 2004. The geology of the Antamina copper-zinc deposit, Peru, South America, in: Porter, T.M. (Ed.), Super Porphyry Copper & Gold Deposits: A Global Perspective. PGC Publishing, Adelaide, Australia. Lindsay, M.B.J., Moncur, M.C., Bain, J.G., Jambor, J.L., Ptacek, C.J., Blowes, D.W., 2015. Geochemical and mineralogical aspects of sulfide mine tailings. Appl. Geochemistry 57, 157–177. https://doi.org/10.1016/j.apgeochem.2015.01.009 Linklater, C.M., Sinclair, D.J., Brown, P.L., 2005. Coupled chemistry and transport modelling of sulphidic waste rock dumps at the Aitik mine site, Sweden. Appl. geochemistry 20, 275–293. doi.org/10.1016/j.apgeochem.2004.08.003 Liu, Q., Li, H., Zhou, L., 2009. Experimental study of pyrite–galena mixed potential in a flowing system and its applied implications. Hydrometallurgy 96, 132–139. doi.org/10.1016/j.hydromet.2008.09.002 Liu, Q., Li, H., Zhou, L., 2008. Galvanic interactions between metal sulfide minerals in a flowing system: Implications for mines environmental restoration. Appl. Geochemistry 23, 2316–2323. doi.org/10.1016/j.apgeochem.2008.02.024 144  Liu, Y., Wu, S., Nguyen, T.A.H., Southam, G., Chan, T., Lu, Y., Huang, L., 2018. Microstructural characteristics of naturally formed hardpan capping sulfidic copper-lead-zinc tailings. Environ. Pollut. 242, 1500–1509. doi.org/10.1016/j.envpol.2018.08.027 Lorca, M.E., Mayer, K.U., Beckie, R., 2015. Investigation of pore gas oxygen and internal temperature in a waste rock dump, Antamina mine, Peru. In ICARD, Santiago, Chile. Lorca, M. E., Mayer, K. U., Pedretti, D., Smith, L. & Beckie, R. D., 2016. Spatial and Temporal Fluctuations of Pore-Gas Composition in Sulfidic Mine Waste Rock. Vadose Zo. J. 15. Doi: 10.2136/vzj2016.05.0039 Lottermoser, B.G., 2010. Mine Wastes : Characterization, Treatment and Environmental Impacts, Third edition, Springer, 2009. ProQuest Ebook Central, https://ebookcentral.proquest.com/lib/ubc/detail.action?docID=338317. Lund, C., Lamberg, P., Lindberg, T., 2015. Development of a geometallurgical framework to quantify mineral textures for process prediction. Miner. Eng. 82, 61–77. doi: 10.1016/j.mineng.2015.04.004 Maher, K., Mayer, K.U., 2019. Tracking Diverse Minerals, Hungry Organisms, and Dangerous Contaminants Using Reactive Transport Models. Elements 15: 81-86. doi.org/10.2138/gselements.15.2.81 Malmström, M.E., Destouni, G., Banwart, S., Stromberg, B.H.E., 2000. Resolving the scale dependency of mineral weathering rates. Int. J. Miner. Process. 34, 1375–1377. doi: 10.1021/es990682u Meima, J.A., Graupner, T., Rammlmair, D., 2012. Applied Geochemistry Modeling the effect of stratification on cemented layer formation in sulfide-bearing mine tailings. Appl. Geochemistry 27, 124–137. doi.org/10.1016/j.apgeochem.2011.09.024 145  MEND, Stantec, 2004. Review of Water Quality Issues in Neutral pH Drainage : Examples and Emerging Priorities for the Mining Industry in Canada. In MEND Report 10.1. Mesa, V., Gallego, J.L.R., González-Gil, R., Lauga, B., Sánchez, J., Méndez-García, C., 2017. Bacterial, archaeal, and eukaryotic diversity across distinct microhabitats in an acid mine drainage. Front. Microbiol. 8, 1756. doi:10.3389/fmicb.2017.01756 Molson, J.W., Fala, O., Aubertin, M., Bussière, B., 2005. Numerical simulations of pyrite oxidation and acid mine drainage in unsaturated waste rock piles. J. Contam. Hydrol. 78, 343–371. doi.org/10.1016/j.jconhyd.2005.06.005 Moncur, M.C., Ptacek, C.J., Blowes, D.W., Jambor, J.L., 2005. Release, transport and attenuation of metals from an old tailings impoundment. Appl. Geochemistry 20, 639–659. doi.org/10.1016/j.apgeochem.2004.09.019 Moncur, M.C., Jambor, J.L., Ptacek, C.J., Blowes, D.W., 2009. Mine drainage from the weathering of sulfide minerals and magnetite. Appl. Geochemistry 24, 2362–2373. doi.org/10.1016/j.apgeochem.2009.09.013 Moncur, M.C., Ptacek, C.J., Hayashi, M., Blowes, D.W., Birks, S.J., 2014. Applied Geochemistry Seasonal cycling and mass-loading of dissolved metals and sulfate discharging from an abandoned mine site in northern Canada. Appl. Geochemistry 41, 176–188. doi.org/10.1016/j.apgeochem.2013.12.007 Morin, K.A., Hutt, N.M., 1997. Environmental geochemistry of minesite drainage: practical theory and case studies. MDAG Publishing, Vancouver. Morin, K.A., Hutt, N.M., 2000. Discrete-zone mixing of net-acid-neutralizing and net-acid-generating rock: Avoiding the argument over appropriate ratios. Society for Mining, Metallurgy, and Exploration, Littleton, CO, USA, Denver, USA, pp. 797–803. 146  Morin, K.A., Hutt, N.M., Price, W.A., Coffin, V., 2001. Violation of Common ABA Prediction Rules by Molybdenum-Related Minesites in British Columbia, Canada, in: ICARD, Skelleftea, Sweden, pp. 566–575. Neuner, M., Smith, L., Blowes, D.W., Sego, D.C., Smith, L.J.D., Fretz, N., Gupton, M., 2013. The Diavik waste rock project : Water flow through mine waste rock in a permafrost terrain. Appl. Geochemistry 36, 222–233. doi.org/10.1016/j.apgeochem.2012.03.011 Nordstrom, D.K., Alpers, C.N., 1999. Geochemistry of acid mine waters. In Environ. geochemistry Miner. Depos. 6, Plumlee, G.S. and Logsdon, M.J., (Eds.) Society of Economic Geologists, pp. 133–160. Nordstrom, D.K., 2011. Hydrogeochemical processes governing the origin, transport and fate of major and trace elements from mine wastes and mineralized rock to surface waters. Appl. Geochemistry 26, 1777–1791. doi.org/10.1016/j.apgeochem.2011.06.002 Nordstrom, D.K., Blowes, D.W., Ptacek, C.J., 2015. Applied Geochemistry Hydrogeochemistry and microbiology of mine drainage : An update. Appl. Geochemistry 57, 3–16. doi.org/10.1016/j.apgeochem.2015.02.008 Pan, Ziping; Ye, Lin; Liu, Tiegeng; Cheng, Zengtao; Gao, Wei; Yang, Y., 2012. A simulation experimental study on oxidative kinetics of sphalerite under hypergene condition. Chinese J. geochemistry 31, 457:464. doi.org/10.1007/s11631-012-0597-8 Parbhakar-Fox, A., Lottermoser, B., Bradshaw, D., 2013. Evaluating waste rock mineralogy and microtexture during kinetic testing for improved acid rock drainage prediction. Miner. Eng. 52, 111–124. doi.org/10.1016/j.mineng.2013.04.022 Parbhakar-Fox, A.K., Edraki, M., Hardie, K., Kadletz, O., Hall, T., 2014. Identification of acid rock drainage sources through mesotextural classification at abandoned mines of Croydon, 147  Australia: Implications for the rehabilitation of waste rock repositories. J. Geochemical Explor. 137, 11–28. doi.org/10.1016/j.gexplo.2013.10.017 Parbhakar-Fox, A. and Lottermoser, B.G., 2015. A critical review of acid rock drainage prediction methods and practices. Miner. Eng. 82, 107–124. doi.org/10.1016/j.mineng.2015.03.015 Parbhakar-Fox, A., Fox, N., Hill, R., Ferguson, T., Maynard, B., 2018a. Improved mine waste characterisation through static blended test work. Miner. Eng. 116, 132–142. doi.org/10.1016/j.mineng.2017.09.011 Parbhakar-Fox, A., Fox, N., Jackson, L., Cornelius, R., 2018b. Forecasting Geoenvironmental Risks: Integrated Applications of Mineralogical and Chemical Data. Minerals 8, 541. doi.org/10.3390/min8120541 Parkhurst, D.L., Appelo, C.A.J., 2013. Description of input and examples for PHREEQC version 3 - A computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations: U.S. Geological Survey Techniques and Methods, book 6. Pedretti, D., Mayer, K.U., Beckie, R.D., 2017. Stochastic multicomponent reactive transport analysis of low quality drainage release from waste rock piles: Controls of the spatial distribution of acid generating and neutralizing minerals. J. Contam. Hydrol. 201, 30–38. doi.org/10.1016/j.jconhyd.2017.04.004 Perkins, E.H.; Nesbitt, H.W.; W.D., Gunter; L.C., St-Arnaud; Mycroft, J.R., 1995. Critical review of geochemical processes and geochemical models adaptable for prediction of acid drainage from waste rock, in: MEND Project 1.42.1. p. 283. Peterson, H.E., 2014. Unsaturated Hydrology, Evaporation, And Geochemistry Of Neutral And Acid Rock Drainage In Highly Heterogeneous Mine Waste Rock At The Antamina Mine, Peru. (Doctoral dissertation) University of British Columbia. Vancouver, Canada. 148  Petruk, W., 2000. Applied Mineralogy in the Mining Industry, 1st Edition. Elsevier Science. doi.org/10.1016/B978-044450077-9/50012-2 Pettit, C.M., 1999. Neutral Mine Drainage, in: Proceedings of Sudbury 1999 Mining and the Environment International Conference, Vol. 2, Sudbury, Canada. Plumlee, G.S., Smith, K.S., Montour, M.R., Ficklin, W.H., Mosier, E.L., 1999. Geologic controls on the composition of natural waters and mine waters draining diverse mineral-deposit types. Reviews in Economic Geology, vol 6B, Society of Economic Geologists, 373-432pp. Pooler, R., Dold, B., 2017. Optimization and Quality Control of Automated Quantitative Mineralogy Analysis for Acid Rock Drainage Prediction. Minerals 7, 12. doi.org/10.3390/min7010012 Price, W.A., 2009. Prediction manual for drainage chemistry from sulphidic geologic materials, Natural Resources Canada. MEND Report 1.20.1., 579 p Qian, G., Fan, R., Short, M.D., Schumann, R.C., Li, J., St.C. Smart, R., Gerson, A.R., 2018. The Effects of Galvanic Interactions with Pyrite on the Generation of Acid and Metalliferous Drainage. Environ. Sci. Technol. 52, 5349-5357. doi : 10.1021/acs.est.7b05558 Qing You, L., Heping, L., Li, Z., 2007. Study of galvanic interactions between pyrite and chalcopyrite in a flowing system: Implications for the environment. Environ. Geol. 52, 11–18. doi: 10.1007/s00254-006-0444-5 Redwan, M., Rammlmair, D., Meima, J.A., 2012. Application of mineral liberation analysis in studying micro-sedimentological structures within sulfide mine tailings and their effect on hardpan formation. Sci. Total Environ. 414, 480–493. doi.org/10.1016/j.scitotenv.2011.10.038 Reimann, C., Filzmoser, P., Garrett, R.G., 2005. Background and threshold : critical comparison 149  of methods of determination. Sci. Total Environ. 346, 1–16. doi: 10.1016/j.scitotenv.2004.11.023 Ritchie, A.I.M., 1994. Rates Of Mechanisms That Govern Pollutant Generation From Pyritic Wastes, in: ACS Symposium Series. pp. 108–122. Roberts, D.R., Ford, R.G., Sparks, D.L., 2003. Kinetics and mechanisms of Zn complexation on metal oxides using EXAFS spectroscopy. J. Colloid Interface Sci. 263, 364–376. doi.org/10.1016/S0021-9797(03)00281-9 Roduit, N., 2007. JMicroVision : un logiciel d’analyse d’images pétrographiques polyvalent. (Doctoral Dissertation) Université de Genève, Genève, Suisse. Sand, W., Jozsa, P.-G., Kovacs, Zs.-M., Săsăran, N., Schippers, A., 2007. Long-term evaluation of acid rock drainage mitigation measures in large lysimeters. J. Geochem. Explor. 92, 205–211. doi.10.1016/j.gexplo.2006.08.006 Santomartino, S., Webb, J. a., 2007. Estimating the longevity of limestone drains in treating acid mine drainage containing high concentrations of iron. Appl. Geochemistry 22, 2344–2361. doi.org/10.1016/j.apgeochem.2007.04.020 Schippers, A., Hallmann, R., Wentzien, S., Sand, W., 1995. Microbial diversity in uranium mine waste heaps. Appl. Environ. Microbiol. 61, 2930–2935. Schippers, A., Breuker, A., Blazejak, A., Bosecker, K., Kock, D., Wright, T.L., 2010. The biogeochemistry and microbiology of sulfidic mine waste and bioleaching dumps and heaps, and novel Fe(II)-oxidizing bacteria. Hydrometallurgy 104, 342–350. doi:10.1016/j.hydromet.2010.01.012 Shook, M.G., Ansley, S.L., Wylie, A., 2004. Tracers and Tracer Testing: Design, Implementation, and Interpretation Methods. U.S. Department of Energy. INEEL Report EXT-03-01466. 150  Singer, P.C., Stumm, W., 1970. Acidic mine drainage: the rate-determining step. Science 167, 1121–1123. doi.org/10.1126/science.167.3921.1121 Sloot, H.A. van der, Zomeren, A. van, 2012. Characterisation Leaching Tests and Associated Geochemical Speciation Modelling to Assess Long Term Release Behaviour from Extractive Wastes. Mine Water Environ. 31, 92–103. doi.org/10.1007/s10230-012-0182-8 Smith, J.L., Beckie, R.D., 2003. Hydrologic and geochemical transport processes in mine waste rock, in: Jambor, J.L., Blowes, D.W., Ritchie, A.I.M. (Eds.), Short Course Series. Mineralogical Association of Canada, Ottawa, pp. 51–72 Smith, K.S., 2007. Strategies to predict metal mobility in surficial mining environments. Eng. Geol. XVII, 25–45. doi: 10.1130/2007.4017(03). Smith, L.J., Blowes, D.W., Jambor, J.L., Sego, D.C., Neuner, M., 2013. The Diavik Waste Rock Project: Particle size distribution and sulfur characteristics of low-sulfide waste rock. Appl. Geochemistry. doi: 10.1016/j.apgeochem.2013.05.006 Smuda, J., Dold, B., Friese, K., Morgenstern, P., Glaesser, W., 2007. Mineralogical and geochemical study of element mobility at the sulfide-rich Excelsior waste rock dump from the polymetallic Zn-Pb-(Ag-Bi-Cu) deposit, Cerro de Pasco, Peru. J. Geochemical Explor. 92, 97–110. doi : 10.1016/j.gexplo.2006.08.001 Sracek, O., Choquette, M., Gélinas, P., Lefebvre, R., Nicholson, R. V, 2004. Geochemical characterization of acid mine drainage from a waste rock pile, Mine Doyon, Québec, Canada. J. Contam. Hydrol. 69, 45–71. doi.org/10.1016/S0169-7722(03)00150-5 St-Arnault, M., Vriens, B., Klein, B., Mayer, K.U., Beckie, R.D., 2019a. Mineralogical controls on drainage quality during the weathering of waste rock. Appl. Geochemistry 108, 104376. doi.org/10.1016/j.apgeochem.2019.104376 151  St-Arnault, M., Vriens, B., Blaskovich, R., Aranda, C., Klein, B., Mayer, K.U., Beckie, R.D., 2019b. Geochemical and mineralogical assessment of reactivity in a full-scale heterogeneous waste-rock pile Minerals Engineering 145, 106089. doi.org/10.1016/j.mineng.2019.106089  St-Arnault, M., 2020, Replication Data for: The assessment of mineralogical properties and hydrological or physicochemical controls on the drainage chemistry of mine waste rock, Scholars Portal Dataverse. doi.org/10.5683/SP2/DI9OSN Stockwell, J., Smith, L., Jambor, J.L., Beckie, R., 2006. The relationship between fluid flow and mineral weathering in heterogeneous unsaturated porous media: A physical and geochemical characterization of a waste-rock pile. Appl. Geochemistry 21, 1347–1361. doi.org/10.1016/j.apgeochem.2006.03.015 Stoops, G., Dobrovolsky, V.V, Fitzpatrick, E.A., Paneque, G., Sleeman, J., 1979. Guidelines for the description of mineral alterations in soil micromorphology. Pedologie 29(1):121-135. Strömberg, B., Banwart, S.A., 1999. Experimental study of acidity-consuming processes in mining waste rock: some influences of mineralogy and particle size. Appl. Geochemistry 14, 1–16. doi: 10.1016/S0883-2927(98)00028-6 Sylvester, P.J., 2012. Use of the mineral liberation analyzer (MLA) for mineralogical studies of sediments and sedimentary rocks, in Mineralogical Association of Canada Short Course 42, St. John’s NL, May 2012, p. 1-16 Taylor, G; Eggleton, R.A., 2001. Regolith Geology and Geomorphology. Ed. John Wiley and Son, Chichester. Velbel, M.A., 1985. Hydrogeochemical constraints on mass balances in forested watersheds of the southern Appalachians. In: J.I. Drever (Editor), The Chemistry of Weathering. D. Reidel, Dordrecht, pp. 231-247. 152  Velbel, M.A., 1993. Constancy of silicate-mineral weathering-rate ratios between natural and experimental weathering: implications for hydrologic control of differences in absolute rates. Chem. Geol. 105, 89–99. doi.org/10.1016/0009-2541(93)90120-8 Vriens B., St.Arnault M., Laurenzi L., Smith L., Mayer K.U., Beckie R.D., 2018. Localized sulfide oxidation limited by oxygen supply in a full-scale waste-rock pile. Vadose Zo. J. 17. doi: 10.2136/vzj2018.06.0119 Vriens B., Peterson H.E., Laurenzi L., Aranda C., Smith L., Mayer K.U., Beckie, R.D., 2019a. Long-term monitoring of waste rock weathering at the Antamina mine, Peru. Chemosphere 215, 858-869. doi: 10.1016/j.chemosphere.2018.10.105 Vriens B., Smith, L., Mayer, K.U., Beckie, R.D. 2019b. Poregas distributions in waste-rock piles affected by climate seasonality and physicochemical heterogeneity. Applied Geochemistry 100, 305-315. doi: 10.1016/j.apgeochem.2018.12.009 Vriens, B., Skierszkan, E.K., St-Arnault, M., Salzsauler, K., Aranda, C., Mayer, K.U., Beckie, R.D., 2019c. Mobilization of Metal(oid) Oxyanions through Circumneutral Mine Waste-Rock Drainage. ACS Omega 2019, 4, 10205-10215. doi.org/10.1021/acsomega.9b01270 Warhurst, A.C., 1999. Best available practices in mining on land of minerals., volume 2. ed. Routledge, L, London. Williamson, M.A., Rimstidt, J.D., 1994. The kinetics and electrochemical rate-determining step of aqueous pyrite oxidation. Geochim. Cosmochim. Acta 58, 5443–5454. doi.org/10.1016/0016-7037(94)90241-0 Younger, P.L.; Banwart, A.A.; Hedin, R.S., 2002. Mine Water Hydrology, Pollution, Remediation, Springer S. ed. Springer Science+Business. Doi:10.1007/978-94-010-0610-1 Ziemkiewicz, P.F., Skousen, J.G., Brant, D.L., Sterner, P.L., Lovett, R.J., 1997. Acid mine 153  drainage treatment with armored limestone in open limestone channels. J. Environ. Qual. 26, 1017–1024. doi.org/10.2134/jeq1997.00472425002600040013x   154  Appendices   155  Appendix A  SUPPLEMENTARY METHODS TO CHAPTER 2  A.1 Uncertainty related to sampling method of the drainage from field barrels Source of water drainage sampling bias from the field barrels: 1- If the volume of water surpassed the 20 L capacity of the sampling container during the sampling period, the container was emptied and 5L was kept to produce a composite sample. This could therefore result in dilution or concentration of elements depending on the length and timing of the last discharge.  2- Low recuperation of drainage water in the sampling container compared to the rain data might affect the average concentration and loads recorded depending on the timing of the missing information occurring at the beginning or at the end of the wet season.  The calculated percentage of the volume of discharge water in all field barrels was under 20%. In addition, after excluding data from low water recuperation, 38% to 52% of the volume of rain was recuperated in the drainage volume from the field barrels. Differences in recuperation may be explained by: experimental error caused by lost of water because of leaks or human error in recording data; evaporation of water; or presence of slow to immobile water component in the field barrels. A.2 Quantitative mineralogical analyses Each aliquots of waste-rock samples were sieved at <2 mm fraction and mounted on one polished thin section and one epoxy round that were carbon coated. The sample representativeness and degree of phase precision were assessed by collecting at least 87223 particles and 171752 grain counts per thin section or mount, which is above previous recommendations of at least 20000 particles and 120000 grains (Pooler and Dold, 2017; Sylvester, 2012). The relative error calculated from the number of measurements (i.e., total grain counts) and proportion of grouped minerals 156  such as sulfides, carbonates, sulfates, phosphates, silicates, Fe-oxides, and Fe-(hydroxy)sulfates was <10%, according to the method outlined by (Jones 1987). A.3 MLA calibration for mineral identification The calibration of mineral identification with MLA was part of a previous study described in Blaskovich (2013). Energy dispersive X-Ray spectroscopy (EDS) classification standards provided X-ray spectra to discriminate phases as well as the elemental composition of mineral phases. As outlined in Blaskovich (2013), the calculated elemental composition data obtained by MLA were in good agreement with geochemical analyses on sample using ICP-MS and atomic absorption spectroscopy. For this study, additional verification of elemental composition measurements by MLA was obtained by collecting FP-XRF measurements. Furthermore, the mineral phase composition attributed by the software was evaluated by EDS spot checks and visual inspection before assigning the phase name and composition. More than 1000 inspections indicated a correct phase assignment rate of 97% (Blaskovich, 2013).   A.4 Reactive transport modelling The geometric parameters of the field barrels are from specifications provided by Aranda, 2010: • Total length L = 0.65 m • Diameter d = 0.585 m  • Area of field cells: Afc = d2/4 = 0.2688 m2 • Volume of field cells: Vfc = Afc x L = 0.1613 m3 (161 300cm3) A.3.1 Transient conditions For space-time discretization in the 1D PHREEQC model, the field barrel length specified above was divided into 12 cells. A cell length and time step was assigned according to the duration of the wet and dry season, which was variable each year. The length and the number of shifts per 157  cells were calculated using the recorded time of water recuperation from the field barrels. During the dry season, no precipitation could be registered, so that the model was assumed to shift one cell for the entire duration of the dry season. A.3.2  Mineral content in the waste rock in the field cells Based on field barrel drainage analyses, the numerical model was focused on simulating the release of Ca, SO4, Fe, Zn and Cu. Reactive minerals incorporated in the model are given in Table B4, along with the initial concentrations and the literature (Keff) and calibrated effective dissolution rate constants (Cal-Keff). The concentrations of reactive minerals (sulfides), calcite, gibbsite and gypsum were calculated using the bulk mineralogy and were converted to molar mass. A.3.3 Rate expression and kinetic parameters The rapid precipitation/dissolution of select minerals (e.g., calcite and gypsum) was simulated assuming thermodynamic equilibria between the solution and the solid phase based on K values from the WATEQ4 database (PHREEQC equilibrium block). The dissolution of other minerals (e.g., pyrite, chalcopyrite, and sphalerite) was simulated using kinetic rate expressions (PHREEQC kinetic / rate blocks). For these kinetic rate expressions, the effective kinetic rate constants were defined by first order (pyrite) or second order (chalcopyrite and sphalerite) rate expressions from the literature (Acero et al., 2009; Williamson and Rimstidt, 1994; Pan et al, 2012). Effective kinetic rate constants (keff) were calibrated using the field cells pH drainage data only, while assuming that other parameters such as modal mineralogy and surface area remained constant (i.e. yielding site-specific and non-transferable rate constants used for intercomparison of rates between the field barrels only). The A/V factor (A is the surface area (m2) and V is the amount 158  of solution (kg water)) was calibrated to fit the drainage pH of the field barrels. The initial moles of solid (Mo) was given by MLA mineralogical content considered 100% liberated.    159  Appendix B  SUPPLEMENTARY TABLES AND FIGURES TO CHAPTER 2  Table B1. Lithology classes, acid-producing potential (AP), acid-neutralization potential (NP), neutralization potential ratio (NPR), as well as waste-rock solid-phase over leachate elemental concentration ratios for the average first and last year. Median and double-mean average deviation ranges are indicated at the bottom. Field Barrels Lithology AP     NP NPR First year Log Solid (mg)/ Leachate (mg/L) ratio Last year Log Solid (mg) / Leachate (mg/L) ratio   tCaCO3/ 1000t tCaCO3/ 1000t  S/SO4 Cu Fe Zn S/SO4 Cu Fe Zn 1A Intrusive 19.4 29 0.06 1.50 6.60 8.31 2.52 1.15 1.25 5.96 0.69 1B Intrusive 6.3 12 0.14 1.17 2.47 7.81 1.01 0.43 1.08 1.14 0.82 1C Intrusive 133.1 8 1.49 1.14 5.95 7.50 2.88 1.58 3.78 6.22 2.06 1D Intrusive 20.0 8 1.9 1.40 5.43 7.30 2.60 1.68 4.86 5.20 2.57 1E Intrusive 48.8 7 0.4 1.39 5.81 7.47 2.86 1.84 4.16 6.27 1.90 2A Skarn 77.5 126 7.7 1.45 6.34 8.09 4.81 1.84 5.98 6.38 4.98 2B Skarn 235.3 36 3.5 1.22 3.67 8.25 2.60 1.74 4.59 6.59 3.05 2C Skarn 290.0 135 1.6 1.73 3.94 7.77 3.11 2.18 4.88 6.05 3.63 2D Skarn 47.5 366 0.2 1.85 4.32 8.39 3.47 2.21 5.23 6.34 3.68 2E Skarn 34.7 122 0.5 1.87 4.69 7.28 2.94 2.05 5.32 6.50 3.06 3A Marble 4.7 769 163.6 1.05 5.53 7.33 4.60 1.16 5.29 4.61 4.53 3B Marble 3.8 838 220.5 0.90 5.62 7.13 4.43 1.20 5.82 5.52 4.42 3C Marble 8.1 605 74.7 1.60 5.81 7.42 4.27 1.79 5.52 5.96 3.97 3D Marble 34.7 184 5.3 1.88 4.05 7.85 2.50 2.16 3.97 4.76 2.37 3E Marble 30.6 175 5.7 1.68 3.65 7.85 2.03 1.82 4.13 6.11 2.40 3F Marble 20.3 707 34.8 1.68 3.79 7.50 2.78 1.67 3.91 5.63 2.77 3G Marble 3.4 990 291.2 1.19 3.92 6.76 2.65 1.72 4.02 4.89 2.62 Median (M)     1.45 4.69 7.50 2.86 1.74 4.59 5.96 2.77 M+2MAD     1.97 6.55 8.12 3.55 2.07 5.94 6.79 4.49 M-2MAD     0.92 2.83 6.87 2.17 1.42 3.24 5.13 1.05  160  Table B2. Standard deviation and number of measurements for Zn, Fe, Cu (n) and pH (n*) for the average concentration of field barrel leachate from the first (I) and last years (W) of sampling.   Field Barrel 1A (I) mg/L 1A (W) mg/L 1B (I) mg/L 1B (W)mg/L 2A (I) mg/L 2A (W)mg/L 2B (I) mg/L 2B (W)mg/L  n=14 n=7 n=15 n=7 n=17 n=17 n=14 n=14  n*=32 n*=38 n*=34 n*=36 n*=34 n*=38 n*=34 n*=23 ZN 0.38 54 9.1 45 0.29 0.19 4.7 2.2 FE 0 0.22 0 2104 0 0.02 0 0.02 CU 0.01 1770 12.15 546.2 0.01 0.01 0.29 0.01 PH* 0.22 0.38 0.62 0.16 0.37 0.34 0.26 0.39  161  Table B3. Grouping of primary phases and oxidation products identified with MLA. The phases were determined by combining database stoichiometric formulae with elemental content measurements by SEM-EDS (Blaskovich, 2013). The amorphous phases are indicated by asterisks and potentially sorbed, co-precipitated or substituted trace elements are indicated by brackets. Phases Primary  Oxidation d tSulfides (Ag)Sulphosalt X  Arsenopyrite X  Bismuthinite X  Bornite X X Chalcocite X  Chalcopyrite X  Covellite X X Enargite X  Enargite(Zn) X  Galena(Se) X  Galenobismutite X  Molybdenite X  Pyrite X  Pyrite(Cu) X  Pyrrhotite X  Siegenite X  Siegenite(CuZn) X  Sphalerite X  Sphalerite(Cu) X  Stibnite X  Tennantite(ZnFe) X  Watanabeite(Zn) X  Oxides Bismutostibiconite X  Cassiterite X  Cuprite X  Magnetite X  Paramelaconite(Zn) X  Portlandite X  Scheelite X  Spinel X  Wulfingite X  Conichalcite X X Molybdofornacite(ZnCu) X X PbMoOxide* X X Powellite X X Srebrodolskite X  Tyrolite(Pb) X X Wulfenite X X Fe-(oxyhydr)oxides/sulfates FeOxide* X X FeOxide(Cu)* X X FeOxide(Ti)* X X 162  Phases Primary  Oxidation d tFe-oxy(hydr)oxide-SO4(SiCuAsMoZn)* X X Carbonates Calcite X X Ankerite X X Dolomite X X Malachite X Otavite(ZnCu) X X Rhodochrosite X X Siderite X X Siderite(MnAsZnCrCu) X X Smithsonite X X Sulfates Alunite X X Barite X X Celestine X X FeSulphate X X Goslarite X X Gypsum X X Jarosite(Cu) X X MoCaSO4(MnCuFeZn)* X X Silicates Andalusite X  Apophyllite X  Augite X  Biotite X  Chlorite X  Dioptase X  Epidote X  Fayalite X  Fayalite(Cu) X  Grossular X  Grunerite X  Grunerite(Mn) X  Muscovite X  Orthoclase X  Phlogopite X  Plagioclase X  Pyroxene X  Quartz X  Rhodonite(Zn) X  Willemite X  Wollastonite X X Zincsilite X X Zircon X  Kaolinite X X Melilite X  Sericite X X Talc(Fe) X X Ti silicate(CuPb)* X X Phosphates Fluorite X X Apatite X  163  Phases Primary  Oxidation d tGoyazite X  Monazite(Ce) X  Apatite(PbCuZn) X  Aggregates RealgarOrpiment X  Fornacite-Conichalcite X  (PbCa)Oxide-(MoZnW) Srebrodolskite X X MoSO4_AltGrossul* X X MoSO4-PowelliClay* X X Amphibole_Anthophylite X  MicaAltered(CuZn) X X Rhodonite-FeSO4(PbZnCu)* X X Fe-Silicate(CuPbZn)* X X FeOxideSO4(CuPbZnAs)* X X * Amorphous phase 164  Table B4. Modal compositions of the waste rock in the studied field barrels, as well as calculated (literature) and calibrated effective rate constants.    a Calculated with rate expression from Acero et al., 2009 at 20 oC .    b Calculated using rate expression from Williamson and Rimstidt, 1994 at 20 oC .    c Calculated using rate expression from Pan et al, 2012 at 20 oC .      Modal composition Mol L-1 Phase 1A 1B 2A 2B Chalcopyrite 2.6 1.3 0.3 0.3 Pyrite 2.9 1.8 0.1 0.8  Sphalerite 0.02 0.004 1.2 1.3   Plagioclase 1.9 3.6 0.3 0.4   Calcite 0.02 0.7 12.9 3.7   Gypsum 0.2 0.08 0.1 0.1   Fe(oxide) 0.8 0.3 8.3 1.5           Calibrated effective rate coefficients   Calculated  rate coefficient  Units Phase 1A  1B  2A  2B    Chalcopyrite  -3.2 -4.5 -5.0 -4.0 -5.2 a Log [s-1 ] Pyrite  -7.24 -6.75 -8.5 -8.5 -8.19 b Log [s-1 ] Sphalerite 2.5 1.0 6.0 7.0 1.76 c L Mol-1 s-1 165  Table B5. Normalized root mean square error (NRMSE) between the aqueous chemistry of the field barrel drainage and the simulation results data.  Field Barrel n SO4 Zn Fe Cu pH 1A 74 1.89E+01 1.54E+00 8.03E+01 2.40E+01 1.32E-01 1B 77 8.44E+00 1.26E+00 7.12E+00 7.92E+00 1.78E-01 2A 134 3.25E+00 3.15E+00 2.92E+00 2.26E+00 9.89E-02 2B 77 3.00E+00 6.13E+00 3.47E+00 1.53E+00 3.10E-02  166  Figure B1. Waste rock sampling locations at the top and the bottom of the field barrels. Top samples were taken directly from the open surface of the barrel, whereas a small opening was made on the side of the field barrels to access waste rock located at the bottom of the field barrel.      167  Figure B2. Modal normalized mineralogical distribution amongst all phases (A) as well as the distribution of selected acid producing (sulfides) and acid buffering (carbonates, plagioclase, and wollastonite) minerals (B) of non-weathered (I) and weathered (W) intrusive (IA and 1B) and skarn (2A and 2B) field barrel samples.  A           B            168   Figure B3. Linear regression of the elemental content of waste rock as measured with FP-XRF and MLA.    y = 1.1752xR² = 0.75y = 0.8618xR² = 0.99y = 0.9287xR² = 0.83024681012140 2 4 6 8 10 12 14Elemental Content from MLA (Wt%)Elemental Content from FP‐XRF (Wt%)CuZnFe1:01Linear (Cu)Linear (Zn)Linear (Fe)169  Figure B4. Solid-phase Cu, Fe and Zn content (wt-%) as measured with FP-XRF in initial “non-weathered” (I) and weathered (W) intrusive (IA and 1B) and skarn (2A and 2B) field barrel samples.      170  Figure B5. Particle size distribution (passing weight-fraction) of the bulk waste rock from intrusive (IA and 1B) and skarn (2A and 2B) field barrels.    171  Figure B6. Comparisons of the temporal evolution of the measured sulfate drainage chemistry from field barrels 1A, 1B, 2A, and 2B with the simulated drainage chemistry (legend applies to all frames).   172  Figure B7. Sulfide mineral reactivity indices (wt%) of the initial non-weathered (I) and weathered (W) waste rock from the intrusive (IA and 1B) and skarn (2A and 2B) field barrel samples. 173  Figure B8. Backscattered electron images and elemental maps showing the presence of alterations at the surface and in cracks of sulfide grains from intrusive 1A (A-B) and skarn 2A (C) field barrels.     A  BC 174  Figure B9. Backscattered electron images and elemental maps showing the presence of sulfide couples surrounded by alteration of intrusive field barrel 1A (frames A-B) and skarn field barrels 2B (frame C) and 2A (frame D).     A B CD175  Appendix C  SUPPLEMENTARY METHODS TO CHAPTER 3  C.1 Uncertainty related to sampling method of the drainage from field barrels Source of water drainage sampling bias from the field barrels: 1- If the volume of water surpassed the 20 L capacity of the sampling container during the sampling period, the container was emptied and 5L was kept to produce a composite sample. This could therefore result in dilution or concentration of elements depending on the length and timing of the last discharge.  2- Low recuperation of drainage water in the sampling container compared to the rain data might affect the average concentration and loads recorded depending on the timing of the missing information occurring at the beginning or at the end of the wet season.  The calculated percentage of the volume of discharge water in all field barrels was under 20%. In addition, after excluding data from low water recuperation, 38% to 52% of the volume of rain was recuperated in the drainage volume from the field barrels. Differences in recuperation may be explained by: experimental error caused by lost of water because of leaks or human error in recording data; evaporation of water; or presence of slow to immobile water component in the field barrels. C.2 Quantitative mineralogical analyzes  Each aliquots of waste-rock samples were sieved at <2 mm fraction and mounted on one polished thin section and one epoxy round that were carbon coated. The sample representativeness and degree of phase precision were assessed by collecting at least 87223 particles and 171752 grain counts per thin section or mount, which is above previous recommendations of at least 20000 particles and 120000 grains (Pooler and Dold, 2017; Sylvester, 2012). The relative error calculated from the number of measurements (i.e., total grain counts) and proportion of grouped minerals 176  such as sulfides, carbonates, sulfates, phosphates, silicates, Fe-oxides, and Fe-(hydroxy)sulfates was <10%, according to the method outlined by Jones 1987. C.3 Calibration of MLA for mineral identification The calibration of mineral identification with MLA was part of a previous study described in Blaskovich (2013). Energy dispersive X-Ray spectroscopy (EDS) classification standards provided X-ray spectra to discriminate phases as well as the elemental composition of mineral phases. As outlined in Blaskovich (2013), the calculated elemental composition data obtained by MLA were in good agreement with geochemical analyses on sample using ICP-MS and atomic absorption spectroscopy. For this study, additional verification of elemental composition measurements by MLA was obtained by collecting FP-XRF measurements. Furthermore, the mineral phase composition attributed by the software was evaluated by EDS spot checks and visual inspection before assigning the phase name and composition. More than 1000 inspections indicated a correct phase assignment rate of 97% (Blaskovich, 2013).   C.4 Temporal moments and flow parameters from tracer tests The tracer test results were analyzed using temporal moments and the flow parameters were quantified from breakthrough curves according to the flow-corrected time approach of Eriksson et al. 1997 as applied in Blackmore, 2014 and 2018a. The degree of mobile pore water (ν; [-]) was calculated by comparing the ratio of apparent mobile water content (Ɵ*m) to the measured mobile content (Ɵ) by assuming no transfer between immobile and mobile domains or the absence of immobile domain (equation 1). This ratio gives the proportion of mobile pore water corresponding to preferential/fast-matrix flow, whereas the remaining proportion (1-v; [-]) provides the degree of matrix flow associated with slow or immobile pore water. The apparent 177  mobile water content Ɵ*m [L3L-3] was calculated according to equation 2, where ߬ ̅  is  the  first temporal moment (i.e. the arrival of center of tracer solute mass at the base of the field barrels graphically evaluated from tracer tests results), ܳ [L3T-1] is the steady state water flow corresponding to the flow-corrected time (τ) (determined graphically), A is the discharge area [L2] of drainage and Z is the travel distance [L]. The flow-corrected time (τ) is the proportion of cumulative flow V(t) [L3] divided by total flow Vtot [L3] relative to the duration of the tracer experiment over the entire period of the tracer test experiment ϒ [T] (equation 3). Results are presented in Table D8. ߥ ൌ Ɵ೘∗Ɵ     Equation 1 Ɵ௠∗ ൌ த̅୕஺௭   Equation 2 ߬ ൌ ϒ ௏ሺ೟ሻ௏೟೚೟				Equation 3 C.5 Reactive transport modelling The geometric parameters of a typical field barrel are from specification of Aranda, 2010:  Total length L = 0.65 m  Diameter d = 0.585 m   Area of field cells: Afc = d2/4 = 0.2688 m2  Volume of field cells: Vfc = Afc x L = 0.1613 m3 (161 300cm3) C.5.1 Transient conditions For space-time discretization in the 1D PHREEQC model, the field barrel length specified above was divided into 12 cells. A cell length and time step was assigned according to the duration 178  of the wet and dry season, which was variable each year. The length and the number of shifts per cells were calculated using the recorded time cumulative volume of infiltration from the field barrel multiplied by the preferential and matrix flow velocities calculated with the breakthrough curves from the tracer test (Table D5, Eriksson et al., 1997). During the dry season, no precipitation could be registered, so the model was assumed to shift one cell for the entire duration of the dry season. C.5.2  Mineral content in the waste rock in the field cells Based on field barrel drainage analyses, the numerical model was focused on simulating the release of Ca, SO4, Fe, Zn and Cu. Reactive minerals incorporated in the model are given in Table D6, along with the initial concentrations and the literature (Keff) and calibrated effective dissolution rate constants (Cal-Keff). The concentrations of reactive minerals (sulfides), calcite, gibbsite and gypsum were calculated using the bulk mineralogy and were converted to molar mass. C.5.3 Rate expression and kinetic parameters The rapid precipitation/dissolution of select minerals (e.g., calcite and gypsum) was simulated assuming thermodynamic equilibria between the solution and the solid phase based on K values from the WATEQ4 database (PHREEQC equilibrium block). The dissolution of other minerals (e.g., pyrite, chalcopyrite, and sphalerite) was simulated using kinetic rate expressions (PHREEQC kinetic / rate blocks). For these kinetic rate expressions, the effective kinetic rate constants were defined by first order (pyrite) or second order (chalcopyrite and sphalerite) rate expressions from the literature (Acero et al., 2009; Williamson and Rimstidt, 1994; Pan et al, 2012). Effective kinetic rate constants (keff) were calibrated using the field cells pH drainage data only, while assuming that other parameters such as modal mineralogy and surface area remained constant (i.e. yielding site-specific and non-transferable rate constants used for comparison of rates 179  between the field barrels only). The A/V factor (A is the surface area (m2) and V is the amount of solution (kg water)) was calibrated to fit the drainage pH of the field barrels. The initial moles of solid (Mo) was given by MLA mineralogical content considered 100% liberated.    180  Appendix D  SUPPLEMENTARY TABLES AND FIGURES TO CHAPTER 3  Table D1. Lithological class, neutralization potential ratio (NPR) of the bulk waste rock in field barrels.    Field Barrels Lithology NPR1A  Marble  164 1B  Marble  221 1C  Marble  75 1D  Marble  5 1E  Marble  6 1F  Marble  35 1G  Marble  291 2A  Intrusive  0.1 2B  Intrusive  0.1 2C  Intrusive  1.5 2D  Intrusive  1.9 2E  Intrusive  0.4 3A  Skarn  7.7 3B  Skarn  3.5 3C  Skarn  1.6 3D  Skarn  0.2 3E  Skarn  0.5    181  Table D2. Standard deviation and number of measurements for SO4, Zn, Fe, Cu (n) and pH (n*) for the average concentration of field barrel leachate from the first and last years of sampling.   Field Barrel  1A (I)  1A (W)  2A (I)  2A (I)  2B (I)  2B (W)  2C (I)  2C (W)  3A (I) 3A (W)   n = 16  n= 4  n= 16  n= 6  n=14  n=7  n=15  n=7  n=17  n=17   n*=31  n*=31  n*=32  n*=31  n*=32  n*=38  n*=34  n*=36  n*=34  n*=38 SO4‐DISS  52.6  29.5  72.7  24.1  265  937.7  542.9  4870.3  209.7  148.9 ZN‐DISS  0.03  0.02  0.04  0.02  0.4  53.6  9.1  44.6  0.3  0.2 FE‐DISS  0  0.2  0  0.1  0  0.2  0  2104.1  0  0.02 CU‐DISS  0.002  0.01  0.004  0.00  0.01  1770.3  12.2  546.2  0.01  0.01 PH‐FIELD*  0.4  0.2  0.3  1.3  0.2  0.4  0.6  0.2  0.4  0.3                         182  Table D3. Standard deviation and number of measurements (n) for the average mass loading of field barrel leachate from the first and last years of sampling.   Field Barrel  1A (I)  1A (W)  2A (I)  2A (I)  2B (I)  2B (W)  2C (I)  2C (W)  3A (I)  3A (W)   n = 16  n= 4  n= 16  n= 8  n=14  n=9  n=15  n=9  n=17  n=19 Sulfate  2.6E+00  1.3E+01  2.0E+00  7.8E‐01  3.5E+01  2.4E+02  3.6E+01  4.1E+02  1.7E+01  1.0E+01 Zn  1.1E‐03  1.4E‐02  2.2E‐03  2.1E‐03  3.3E‐02  5.7E+00  9.9E‐01  2.8E+00  2.1E‐02  2.0E‐02 Fe  7.5E‐03  1.9E‐02  5.8E‐03  3.9E‐03  5.1E‐03  1.4E‐02  6.8E‐03  1.5E+02  5.8E‐03  2.5E‐03 Cu  6.9E‐03  3.8E‐04  4.8E‐03  6.4E‐04  3.5E‐03  9.4E+01  8.4E‐01  3.6E+01  1.2E‐03  1.4E‐03      183  Table D4. Grouping of primary phases and oxidation products identified with MLA. The phases were determined by combining database stoichiometric formulae with elemental content measurements by SEM-EDS (Blaskovich, 2013). The amorphous phases are indicated by asterisks and potentially sorbed, co-precipitated or substituted trace elements are indicated by brackets. Phase/Aggregate Primary Oxidation products Soluble products Less-soluble Sulfides  (Ag)Sulphosalt X X Arsenopyrite X X Bismuthinite X X Bornite X X X Chalcocite X X Chalcopyrite X X Covellite X X X Enargite X X Enargite(Zn) X X Galena(Se) X X Galenobismutite X X Molybdenite X X Pyrite X X Pyrite(Cu) X X Pyrrhotite X X Realgar/Orpiment* X X Siegenite X X Siegenite(CuZn) X X Sphalerite X X Sphalerite(Cu) X X Stibnite X X Tennantite(ZnFe) X X Watanabeite(Zn) X X Oxide  Bismutostibiconite X X Cassiterite X X Cuprite X X Fornacite/Conichalcite* X X Magnetite X X Paramelaconite(Zn) X X Portlandite X X Scheelite X X Spinel X X Wulfingite X X Conichalcite X X X FeOxide(Ti)* X X X Molybdofornacite(ZnCu) X X X (PbCa)Oxide-(MoZnW)Srebrodolskite* X X X PbMoOxide* X X X Powellite X X X Srebrodolskite X X Tyrolite (Pb) X X X Wulfenite X X X Fe Hydroxy-Sulfates  FeOxide* X X X 184  Phase/Aggregate Primary Oxidation products Soluble products Less-soluble FeOxide(Cu)* X X X FeOxide(SO4)* X X X FeOxideSO4(CuPbZnAs)* X X X FehydroxideSO4(SiCuAsMoZn)* X X X Carbonate  Calcite X X X  Ankerite X X X  Dolomite X X X  Malachite X X X  Otavite (ZnCu) X X X  Rhodochrosite X X X  Siderite X X X  Siderite(MnAsZnCrCu) X X X  Smithsonite (Zn, Cd) X X X  Sulfate  Alunite X X X Barite X X X  Celestine X X X  FeSulphate* X X X  Goslarite X X X  Gypsum X X X  Jarosite(Cu) X X X MoSO4 AltGrossular* X X X MoCaSO4(MnCuFeZn)* X X X MoSO4 PowelliClay* X X X Silicate  Andalusite X X Apophyllite X X Augite X X Biotite X X Chlorite X X Dioptase X X Epidote X X Fayalite X X Fayalite(Cu) X X Grossular X X Grunerite X X Grunerite(Mn) X X Muscovite X X Orthoclase X X Phlogopite X X Plagioclase X X Pyroxene X X Quartz X X Rhodonite(Zn) X X Willemite X X X Wollastonite X X X Zincsilite X X X Zircon X X Amphibole Anthophylite* X X Fe-silicate(CuPbZn)* X X Kaolinite X X X Melilite X X X MicaAltered(CuZn)* X X X 185  Phase/Aggregate Primary Oxidation products Soluble products Less-soluble RhodonitFeSO4(PbZnCu)* X X Sericite X X X Talc(Fe) X X X Titanite(CuPb)* X X X Phosphate  Fluorite X X X  Apatite X X Goyazite X X Monazite(Ce) X X Apatite(PbCuZn) X X  * Aggregated or amorphous phases    186  Table D5. Adjusted space-time discretization representative of the calculated flow paths (i.e. proportion of preferential and matrix flow) used in the simulation of 3 selected field barrels 2B, 2C, and 3A.  Wet season 40% Preferential Flow 60%  Matrix Flow 100% Matrix Flow 100% Preferential Flow Year (Day) Shift  Shift  Shift  Shift  Field barrel 2B 2007-2008 220 14 21 23 0 2008-2009 189 12 18 19 0 2009-2010 161 10 16 17 0 2010-2011 175 11 17 18 0 2011-2012 231 15 22 24 0 2012-2013 181 12 17 19 0 2013-2014 280 18 27 29 0 2014-2015 233 15 22 24 0 2015-2016 196 13 19 20 0 Field barrel 2C2007-2008 184 1 14 14 0 2008-2009 162 1 12 12 0 2009-2010 195 1 15 15 0 2010-2011 189 1 14 14 0 2011-2012 232 1 17.5 17.5 0 2012-2013 188 1 14 14 0 2013-2014 280 2 21 21 0 2014-2015 211 1 16 16 0 2015-2016 223 1 17 17 0 Field barrel 3A2007-2008 198 1 7 7 0 2008-2009 196 1 7 7 0 187   Wet season 40% Preferential Flow 60%  Matrix Flow 100% Matrix Flow 100% Preferential Flow Year (Day) Shift  Shift  Shift  Shift  2009-2010 182 1 6 6 0 2010-2011 187 1 6 6 0 2011-2012 210 1 7 7 0 2012-2013 174 1 6 6 0 2013-2014 280 2 10 10 0 2014-2015 239 1 8 8 0 2015-2016 203 1 7 7 0  188  Table D6. Modal compositions of the waste rock in the studied field barrels, as well as calculated (literature) and calibrated effective rate constants.    a Calculated with rate expression from Acero et al., 2009 at 20 oC .    b Calculated using rate expression from Williamson and Rimstidt, 1994 at 20 oC .    c Calculated using rate expression from Pan et al, 2012 at 20 oC .     Modal composition Mol L-1 Phase 2B 2C 3A Chalcopyrite 2.6 1.3 0.3 Pyrite 2.9 1.8 0.1  Sphalerite 0.02 0.004 1.2   Plagioclase 1.9 3.6 0.3   Calcite 0.02 0.7 12.9   Gypsum 0.2 0.08 0.1   Fe(oxide) 0.8 0.3 8.3          Calibrated effective rate coefficients   Calculated  rate  coefficients  Units Phase 2B  2C  3A    Chalcopyrite  -3.5 -4.5 -4.0 -5.2 a Log [s-1 ] Pyrite  -7.75 -6.75 -8.5 -8.19 b Log [s-1 ] Sphalerite 2.5 1.0 6.0 1.76 c L Mol-1 s-1 189  Table D7. Normalized root mean square error (NRMSE) between the aqueous chemistry of the field barrel drainage and the simulation results data. Field Barrel % Matrix flow n      pH  Fe          S(6)   Zn      Cu  2B  60  77  9.5E‐02  2.1E+01  5.3E+00  1.6E+00  7.2E+00 2C  7  73  1.8E‐01  7.0E+00  9.1E+00  1.2E+00  1.1E+01 3A  10  69  4.9E‐02  2.8E+00  3.3E+00  7.7E‐01  2.4E+00    190  Table D8. Parameters used to calculate the degree of mobile pore water and Damköhler number in selected field barrels using tracer tests results, the flow-corrected time approach, and pyrite depletion rate from simulation. Field Barrel R ߬̅ Ɵ Ɵ*m Q v 1-v ߬̅* Pyrite Depletion Da Day Day % (-) L/D (-) (-) Day Day (-) 1B* 11.99 5.8 3.0 0.021 0.0007 0.70 0.30 -- -- -- 1F* 12.01 5.5 3.8 0.022 0.0008 0.59 0.41 -- -- -- 2A 45.00 1.9 1.0 0.003 0.0003 0.32 0.68 -- -- -- 2B 45.13 2.4 3.8 0.004 0.0004 0.12 0.88 9.0 3130 0.0032C 13.07 0.5 2.6 0.001 0.0003 0.03 0.97 2.3 186 0.0123A 45.01 6.7 1.0 0.019 0.0005 0.19 0.81 15.4 2138 0.007r: Flow-corrected time ߬̅: First temporal moment (flow-corrected time) Ɵ: Measured water content Ɵ*m: Apparent mobile water content Q: Steady state water flow v: Degree of mobile pore water 1-v: Degree of slow or immobile pore water  ߬̅*: First temporal moment (tracer test results non-corrected) Pyrite Depletion: Time to reach pyrite depletion of 0.1 mole in simulation Da: Damköhler number  * The average of field barrels 1B and 1F was used to represent field barrel 1A  191  Figure D1. Modal normalized mineralogical distribution amongst all phases (i.e primary and residual) (A) and distribution of secondary phases of non-weathered (I) and weathered (W) marble (1A), intrusive (2A, 2B and 2C), and skarn (3A) field barrel samples.       192  Figure D2. Particle size distribution (passing weight-fraction) marble (1A), intrusive (2A, 2B and 2C), and skarn (3A) field barrel samples.   193  Figure D3. Waste rock sampling locations at the top and the bottom of the field barrels. Top samples were taken directly from the open surface of the barrel, whereas a small opening was made on the side of the field barrels to access waste rock located at the bottom of the field barrel.     194  Figure D4. Linear regressions of waste rock element content measured with XRF and MLA     y = 1.1752xy = 0.8618xy = 0.9287x024681012140 2 4 6 8 10 12 14Elemental Content from MLA (Wt%)Elemental Content from FP‐XRF (Wt%)CuZnFe1:01Linear (Cu)Linear (Zn)Linear (Fe)195  Figure D5. Infilling agglomerate/rims of Fe-oxides (orange) and sulfate needles (A) Boxwork textures (i.e. cracks and cleavage planes) (B) or both textures (C) observed with the microscope or SEM in thin sections of samples 1A, 2A, 2B, and 3A.                      A BC196  Figure D6. Normalized distribution of absolute and relative minerals (i.e. moderate and less soluble minerals, respectively) in weathered samples of all field barrels.   197   Appendix E  SUPPLEMENTARY METHODS TO CHAPTER 4  E.1 Powder pH Hellige-TruogTM soil powder pH reaction tests from Orbeco analytical systems inc. were performed on <2 mm grab samples of waste rock collected from the drill cuttings. Two drops of the reagent # 697-27 were mixed with <1 g of the dried sample and a film of reaction powder # 697-26 was sprinkled on top. The resulting color was compared to the standard color chart to estimate pH on a scale from 1 to 10. E.2 Raman spectroscopy For Raman spectroscopy, an XploRA ONETM Horiba Raman spectrometer was mounted on an optical microscope equipped with a laser beam with a frequency of 532 μm and a range of 50 to 1600 cm-1. The laser power was set between 0.1 to 25% at a 30 to 120 s counting time to avoid thermal destruction of the sample. The spectrograph settings were 1800 gr/mm, a 200 μm slit width, and a 300 μm confocal hole opening to analyze the scattered light. The obtained spectra were matched using the RRUFF database (Lafuente, 2015) and the CrystalSleuth identification software (Laetsch, 2006). E.3 Mineral liberation analyzer For automated mineralogical analyses, thin sections were prepared from selected representative waste-rock samples by drying and sieving the <2 mm fractions. These samples  were mixed with epoxy resin, mounted on a glass slides, allowed to cure, and then polished using a non-aqueous cooling lubricant to minimize potential phase alteration caused by dissolution and/or heat. The mineralogical assemblage of the waste rock samples was investigated using Thermo-Fisher’s Mineral Liberation Analyzer (MLA®), which includes a FEI Quanta 600 scanning electron 198  microscope (SEM), a dual Bruker-AXS silicon-drift (10 mm2) energy dispersive X-ray analysis system (EDS) and the Thermo-Fisher MLA DataView® software. Electron beam operating conditions were set at 25 kV accelerating voltage and a ~6 nÅ beam current. The MLA extended backscatter electron method (XBSE) was used with the frame resolution set at 800 x 680 μm or 1.06 μm/pixel.  The mineral standards library, collected by Blaskovich 2013, was applied for X-ray pattern phase assignment. As outlined in Blaskovich (2013), the calculated elemental composition data obtained by MLA was in good agreement with geochemical analyses performed using ICP-MS and atomic absorption spectroscopy. For this study, additional verification of elemental composition measurements by MLA was obtained by collecting FP-XRF measurements (see below). Furthermore, the mineral phase composition attributed by the software was evaluated by EDS spot checks and visual inspection before assigning the phase name and composition. More than 1000 inspections performed by Blaskovich (2013) indicated a correct phase assignment rate of 97%.   The MLA DataView® software provided collated mineral association and liberation data for the interpretation of mineral textures and produced colored SEM images of the particles. Each sieved <2 mm sample fraction was mounted on one thin section and one epoxy round mount and coated with carbon. The sample representability and degree of precision for phase assignment were verified by analyzing at least 66669 particles and 177,474 grains counted per thin sections or mounts, which is above previous recommendations for 20,000 particles and 120,000 grains (Sylvester 2012, Pooler, 2017). A relative error <10% was calculated (according to the method outlined in Jones 1987) for the phase-assignment of grouped sulfides, carbonates, sulfates, 199  phosphates and silicate minerals, while a relative error of <14% was attained for Fe-(oxyhydr)oxides and -sulfates. E.4 Handheld X-Ray fluorescence A field portable Olympus InnovX X-Ray fluorescence (FP-XRF) analyzer measured lower grade elements such as Ca, S or As in ‘soil mode’, and more abundant metals such as Fe or Cu in ‘processing mode’ (Cebeci et al, 2016). Waste-rock samples were analyzed through plastic bags containing an aliquot of the sieved <2 mm fraction. An average of 3 measurements of ≥30 sec were performed and the samples were homogenized by shaking the bag in between each measurements. The FP-XRF was calibrated with a reference standard (Olympus, n°316) every 15 readings. The elemental composition from selected samples measured by FP-XRF was compared with assays results from ICP-MS for quality control purposes. The results from FP-XRF are most of the time equivalent (i.e. y=0.6 to 0.9) or overestimated (y= 1.5 to 2) compared to the ICP-MS assays.   200  Appendix F  SUPPLEMENTARY TABLES AND FIGURES TO CHAPTER 4  Table F1. Grouping of primary phases and oxidation products identified with MLA. The name of the phases or aggregates were attributed based on stoichiometric formulae calculated from elemental content measurements by SEM-EDS (Blaskovich, 2013). The amorphous phases are indicated by asterisks and potentially adsorbed, co-precipitated or substituted trace elements are indicated by brackets. Phases Primary  Oxidation d tSulfides (Ag)Sulphosalt X  Arsenopyrite X  Bismuthinite X  Bornite X X Chalcocite X  Chalcopyrite X  Covellite X X Enargite X  Enargite(Zn) X  Galena(Se) X  Galenobismutite X  Molybdenite X  Pyrite X  Pyrite(Cu) X  Pyrrhotite X  Siegenite X  Siegenite(CuZn) X  Sphalerite X  Sphalerite(Cu) X  Stibnite X  Tennantite(ZnFe) X  Watanabeite(Zn) X  Oxides Bismutostibiconite X  Cassiterite X  Cuprite X  Magnetite X  Paramelaconite(Zn) X  Portlandite X  Scheelite X  Spinel X  Wulfingite X  Conichalcite X X Molybdofornacite(ZnCu) X X PbMoOxide* X X Powellite X X Srebrodolskite X  Tyrolite(Pb) X X Wulfenite X X Fe-(oxyhydr)oxides/sulfates FeOxide* X X 201  Phases Primary  Oxidation d tFeOxide(Cu)* X X FeOxide(Ti)* X X FeOxideSO4(CuPbZnAs)* X X Fe-oxy(hydr)oxide-SO4(SiCuAsMoZn)* X X Carbonates Calcite X X Ankerite X X Dolomite X X Malachite X Otavite(ZnCu) X X Rhodochrosite X X Siderite X X Siderite(MnAsZnCrCu) X X Smithsonite X X Sulfates Alunite X X Barite X X Celestine X X FeSulfate X X Goslarite X X Gypsum X X Jarosite(Cu) X X MoCaSO4(MnCuFeZn)* X X Silicates Andalusite X  Apophyllite X  Augite X  Biotite X  Chlorite X  Dioptase X  Epidote X  Fayalite X  Fayalite(Cu) X  Grossular X  Grunerite X  Grunerite(Mn) X  Muscovite X  Orthoclase X  Phlogopite X  Plagioclase X  Pyroxene X  Quartz X  Rhodonite(Zn) X  Willemite X  Wollastonite X X Zincsilite X X Zircon X  Kaolinite X X Melilite X  Sericite X X Talc(Fe) X X Ti silicate(CuPb)* X X 202  Phases Primary  Oxidation d tPhosphates Fluorite X X Apatite X  Goyazite X  Monazite(Ce) X  Apatite(PbCuZn) X  Aggregates RealgarOrpiment X  Fornacite-Conichalcite X  (PbCa)Oxide-(MoZnW) Srebrodolskite X X MoSO4_AltGrossul* X X MoSO4-PowelliClay* X X Amphibole_Anthophylite X  MicaAltered(CuZn) X X Rhodonite-FeSO4(PbZnCu)* X X Fe-Silicate(CuPbZn)* X X * Amorphous phases   203  Table F2. List of minerals identified with Raman Spectroscopy in selected samples from boreholes BH1D2 and BH3D2. Mineral BH3D2 11.25-11.8 BH3D2 39.3-39.5 BH3D2 39.5-39.6 BH3D1 95-95.3 BH3D2 100.8-100.9 BH1D2 22.3-22.5 BH1D2 22.5-22.85 BH1D2 95.25-96 BH1D2 96-97 Adamite    X  Alunite X   X   Anglesite X   X   Antlerite X   X   Argentopyrite   X   Aurichalcite X     Brochantite X     Brookite    X   Bukovskyite X   X X   Chalcopyrite X X  X  X Covellite    X   Diopside X     Ferrihydrite X X    Glauberite    X   Goethite X X  X X X X  Gypsum X   X X X  Hematite X X  X X X X X Lepidocrocite X X  X X  Linarite X     Magnesite X     Magnetite X   X  X Mimetite    X  Nesquehonite X   X   Posnjakite X     Pyrite X X X X X X X X Pyrrhotite   X   Satterlyite X     Siderite X   X  Sphalerite    X X  X 204  Table F3. List of samples from borehole BH1D2 and BH3D2 selected for further mineralogical characterization with MLA. Selection was based on lithology, color of alteration in the matrix, presence of disseminated sulfides, average metal leaching, or solid phase/leachate ratio outliers showing relative retention (+2MAD) or release of elements (-MAD). The reactive zones are indicated in bold, No alt: No alteration, NA: Not applicable, -- : None. Borehole Depth (m) Lithology  Color of Alteration in matrix Disseminated Sulfides (DS) Secondary mineral abundance (%) Metal(loid) in Leachate  Relative retention (+2MAD) Relative Release (-2MAD) BH1D2 20.3-20.5 Intrusive/Skarn Orange NA NA NA NA BH1D2 22.3-22.5* Intrusive Orange 9 NA NA NA BH1D2 23.25-23.75* Intrusive Green 5 NA NA NA BH1D2 27.25-28* Marble Orange 6 NA NA NA BH1D2 28-28.5* Marble DS 4 NA NA NA BH1D2 83-83.3 Intrusive Orange NA NA NA NA BH1D2 83.3-83.6* Marble White 1 NA NA NA BH1D2 86.65-87* Intrusive/Marble Orange 7 Zn Cu Fe -- BH1D2 95.25-96* Skarn/Marble Orange 8 Zn Cu Fe As Zn BH1D2 96-97* Skarn No alt 3 Zn Cu Fe -- BH1D2 104.15-105* Marble  No alt 2 Mo -- Mo Cu BH1D2 105-106.3* Marble  Orange 5 Zn -- -- BH1D2 107.55-108* Marble/Marble  No alt 2 Zn As -- BH3D2 11.25-11.8* Intrusive/Marble Orange 15 Zn Fe As Cu Zn BH3D2 11.85-12* Intrusive/Marble Green NA Zn Fe As Cu Zn BH3D2 12-12.4* Intrusive/Marble No alt 2 Zn Cu Fe Mo As Cu Zn 205  Borehole Depth (m) Lithology  Color of Alteration in matrix Disseminated Sulfides (DS) Secondary mineral abundance (%) Metal(loid) in Leachate  Relative retention (+2MAD) Relative Release (-2MAD) BH3D2 34-34.1* Marble No alt NA NA NA NA BH3D2 34.1-34.3 Marble White 1 NA NA NA BH3D2 34.4-34.6* Marble Orange 1 NA NA NA BH3D2 47.2-48* Intrusive/Marble White 2 Mo As Cu Pb  Fe As BH3D2 62-63* Marble No alt 1 Zn Fe Mo As -- BH3D2 84.4-85.7* Skarn No alt, DS 2 Zn -- -- BH3D2 85.7-86* Marble Orange 2 NA NA NA BH3D2 95-95.3* Intrusive/Limestone Orange DS 11 ND -- ND BH3D2 95.3-95.5* Intrusive/Limestone White 4 As -- As BH3D2 95.8-96* Skarn/Limestone White 2 As -- As BH3D2 100.75-101.65* Skarn/Limestone White, DS 2 -- Cu -- 206  Figure F1. Linear regression of the elemental content of waste rock as measured with XRF or MLA, for selected samples from boreholes BH1D2 and BH3D2.         207  Figure F2. Gravimetric moisture content of the waste rock in boreholes BH1D2 and BH3D2.    208  Figure F3. Total Mo and As Waste rock and leachate concentration of samples along boreholes BH1D2 and BH3D2.   209  Figure F4. Rinse-pH tests and qualitative powder-pH tests were compared with paste-pH results of the waste rock samples from boreholes BH1D2 and BH3D2. Paste pH for BH1D2 only started at 60 m.  210  Figure F5. Average oxygen content (%) and temperature (oC) over 4 years in boreholes BH1D2 and BH3D2 adapted from Vriens et al. (2018). The red shading is indicative of reactive zones, as based on low oxygen content and elevated temperatures.     211  Figure F6. Modal composition (wt-%) of the total mineral assemblage (top frame, normalized to 100 wt-%) and the secondary mineral composition of the waste-rock in boreholes BH1D2 and BH3D2  212   Figure F7. Proportion of primary versus secondary minerals and elemental distributions of Cu, Fe and Zn among liberated primary minerals (>70% mineral perimeter exposed) and total secondary minerals in borehole BH1D2 (PM: Primary minerals, SM: Secondary minerals).       213  Figure F8. Proportion of primary versus secondary minerals and elemental distributions of Cu, Fe and Zn among liberated primary minerals (> 70% mineral perimeter exposed) and total secondary minerals in borehole BH3D2 (PM: Primary minerals, SM: Secondary minerals).    214  Figure F9. Modal distribution of liberated As- and Mo-bearing minerals (> 70% mineral perimeter exposed) including As and Mo potentially adsorbed or co-precipitated with other phases in selected samples of boreholes BH1D2 and BH3D2.    215  Figure F10. Composite photographs of the sonic drill cores from borehole BH1D2. Crosses indicate mineralogical sampling locations and rectangles indicate reactive zones.     216   Figure F11. Composite photographs of the sonic drill cores from borehole BH3D2. Crosses indicate mineralogical sampling locations and rectangles indicate reactive zones.    217  Figure F12. Pictures of agglomerates retrieved from the core samples of boreholes BH1D2 between 20.3-20.4 m and of BH3D2 at 11.4 m, 91-91.25 m, and 93-94 m.  218    Figure F13.  Composite digital images obtained by MLA from microscopic agglomerates observed in samples from borehole BH1D2 at 22.3-22.5 m (1) and 96-97 m (2-3) depth, and BH3D2 at 34.4-34.6 m (7), 95-95.3 m (4-5-8) and 95.3-95.5m (6) depth. The mineral legends are indicated next to the respective images. The scale at the bottom left applies to all images.    219  Figure F14. Relationship between the pyrite reactivity index [expressed in wt-%; see equation 1] versus the pyrite association index [unitless, see equation 2] from selected samples of boreholes BH1D2 and BH3D2 (same as Figure 4.4).      220  Figure F15. Modal mineralogical composition of the liberated sulfides [wt%] obtained by MLA of selected samples from boreholes BH1D2 and BH3D2.    221  Figure F16. Composite MLA images of passivated sulfides from samples BH3D2 11.25-11.8m (1-2), 11.85-12m (3), 95-95.3m (4-5), 95.3-95.5m (6) as well as BH1D2 22.3-22.5m (7), 86.65-87m (8-9), 95.25-96 (10), 96-97m (11). Scale on bottom left applies to all frames.    222  Figure F17. Relationship between pyrite reactivity wt% versus total Fe-, Cu-, or Zn- bearing secondary minerals content (wt%) in selected weathered (triangles) and unweathered (circles) samples of boreholes BH1D2 and BH3D2.   223  Figure F18. Concentration ratios of waste-rock solid-phase As and Mo content over the respective aqueous leachate concentrations from samples along boreholes BH1D2 and BH3D2 (logarithmic x-axis). Reactive zones are indicated by the grey horizontal shading. Vertical dotted lines represent the median concentration ratio in that profile; the ±2MAD outlier threshold is indicated by the red shading. Samples right of the +2MAD threshold suggest relative retention, whereas samples left of the -2MAD threshold indicate relative mobilization. 224  Figure F19.  MLA composite images of sulfide galvanic couples from samples BH3D2 11.25-11.8m (1), 12-12.45m (2), 95-95.3 (4-5) as well as BH1D2 28-28.55m (6), 86.65-87m (3), 96-97m (7-8). Scale at the bottom right applies to all frames.    225  Figure F20. Modal distribution (wt%) of Zn-bearing phases (refer to Table E1 for phase group) including Zn potentially adsorbed or co-precipitated with other phases in selected samples of boreholes BH1D2 and BH3D2.    226  Figure F21. Composite MLA images of liberated As-bearing particles from samples BH1D2 104.15-105 m (1) and BH3D2 47.2-48 m (2), 95.3-95.8 m (3) and 95.8-96 m depth (4).Only the As bearing phases (including As adsorbed or co-precipitated with phases) are presented in the legend (refer to Table E1 for phase group).      1 2 3 4 227  Figure F22. MLA composite images of liberated Mo-bearing minerals from samples BH1D2 104.15-105m (1), BH3D2 47.2-48m (2), 95.3-95.8m (3), 95.8-96m (4). Only the Mo bearing phases (including phases with adsorbed or co-precipitated Mo) are presented in the legend (refer to Table E1 for phase group).   1 2 4 3 228  Appendix G  SUPPLEMENTARY INFORMATION  Raw data are available for consultation online in the public data repository Scholars Portal Dataverse (St-Arnault, 2020). doi.org/10.5683/SP2/DI9OSN   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0389805/manifest

Comment

Related Items