{"@context":{"@language":"en","Affiliation":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","AggregatedSourceRepository":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","Campus":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","Creator":"http:\/\/purl.org\/dc\/terms\/creator","DateAvailable":"http:\/\/purl.org\/dc\/terms\/issued","DateIssued":"http:\/\/purl.org\/dc\/terms\/issued","Degree":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","DegreeGrantor":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","Description":"http:\/\/purl.org\/dc\/terms\/description","DigitalResourceOriginalRecord":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","FullText":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","Genre":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","GraduationDate":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","IsShownAt":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","Language":"http:\/\/purl.org\/dc\/terms\/language","Program":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","Provider":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","Publisher":"http:\/\/purl.org\/dc\/terms\/publisher","Rights":"http:\/\/purl.org\/dc\/terms\/rights","RightsURI":"https:\/\/open.library.ubc.ca\/terms#rightsURI","ScholarlyLevel":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","Supervisor":"http:\/\/purl.org\/dc\/terms\/contributor","Title":"http:\/\/purl.org\/dc\/terms\/title","Type":"http:\/\/purl.org\/dc\/terms\/type","URI":"https:\/\/open.library.ubc.ca\/terms#identifierURI","SortDate":"http:\/\/purl.org\/dc\/terms\/date"},"Affiliation":[{"@value":"Science, Faculty of","@language":"en"},{"@value":"Earth, Ocean and Atmospheric Sciences, Department of","@language":"en"}],"AggregatedSourceRepository":[{"@value":"DSpace","@language":"en"}],"Campus":[{"@value":"UBCV","@language":"en"}],"Creator":[{"@value":"Wynands, Eric","@language":"en"}],"DateAvailable":[{"@value":"2022-09-30T07:00:00Z","@language":"en"}],"DateIssued":[{"@value":"2021","@language":"en"}],"Degree":[{"@value":"Master of Science - MSc","@language":"en"}],"DegreeGrantor":[{"@value":"University of British Columbia","@language":"en"}],"Description":[{"@value":"As CO\u2082 emissions continue to rise, the threat posed by anthropogenic climate change grows. To combat this crisis, technologies are needed to capture and store CO\u2082. Carbon mineralization is one approach, which reacts cations with CO\u2082 to form stable carbonate minerals. Ultramafic mine tailings are one cation feedstock that could be used to advance this strategy at scale. CO\u2082 injection into the porous medium is required to accelerate the mineralization rates.\r\nUltramafic mine tailings vary widely in their reactivity to CO\u2082. One important mineral, brucite, reacts rapidly with CO\u2082. Thermogravimetric analysis was used to identify and quantify brucite abundances. The developed method improved brucite detection limits and could quantify brucite when phases, which interfere with identification by X-ray diffraction, are present.\r\nCentimetre-scale injection experiments were conducted on well-graded tailings grain size distributions on serpentinite (brucite-rich) and kimberlite (brucite-poor) samples. Well-graded grain size distributions enabled the permeability to inject into the porous medium. Reactive fine-grained brucite released most of its Mg, while lizardite leached a minority of its Mg in the brucite-poor samples. 0.1 to 1 wt.% CO\u2082 was sequestered into major hydromagnesite and minor nesquehonite. Metre-scale experiments in kimberlite tailings increased the scale and designed methods to achieve mine-scale injection. Compaction testing evaluated the relationship between the dry density and moisture content of the tailings. 6-metre-long pipe experiments replicated the reactivity of the centimetre scale. Scalable pad experiments injected CO\u2082 through a perforated pipe into a layer of mixed coarse and fine tailings. Plume development throughout the pad led to homogeneous reactivity. Heterogeneity and low abundances of sequestered carbon made\r\nconfirming the gas-phase mass balance on the injected CO\u2082 using total inorganic carbon difficult. A similar magnitude of carbon was sequestered in PK, as observed at the centimetre-scale, varying from 0.1 to 0.2 wt.% CO\u2082.\r\nCarbon mineralization in mine tailings via CO\u2082 injection has been shown to be effective in accelerating passive sequestration rates. This observed reactivity equates to a minor reduction in mine emissions at carbon-intensive mines while being enough capacity to make low-carbon mines carbon neutral.","@language":"en"}],"DigitalResourceOriginalRecord":[{"@value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/79796?expand=metadata","@language":"en"}],"FullText":[{"@value":"CARBON MINERALIZATION IN ULTRAMAFIC MINE TAILINGS VIA CO2 INJECTION  by  Eric Wynands  B.A.Sc., Queen\u2019s University, 2018  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Geological Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  September 2021  \u00a9 Eric Wynands, 2021 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis entitled:  Carbon Mineralization in Ultramafic Mine Tailings Via CO2 Injection  submitted by Eric Wynands in partial fulfillment of the requirements for the degree of Master of Science in Geological Sciences  Examining Committee: Dr. Gregory Dipple, Earth, Ocean and Atmospheric Sciences, UBC Supervisor  Dr. Connor Turvey, Earth, Ocean and Atmospheric Sciences, UBC Supervisory Committee Member Dr. Kelly Russell, Earth, Ocean and Atmospheric Sciences, UBC Additional Examiner   Additional Supervisory Committee Members: Dr. Ulrich Mayer, Earth, Ocean and Atmospheric Sciences, UBC Supervisory Committee Member iii  Abstract As CO2 emissions continue to rise, the threat posed by anthropogenic climate change grows. To combat this crisis, technologies are needed to capture and store CO2. Carbon mineralization is one approach, which reacts cations with CO2 to form stable carbonate minerals. Ultramafic mine tailings are one cation feedstock that could be used to advance this strategy at scale. CO2 injection into the porous medium is required to accelerate the mineralization rates. Ultramafic mine tailings vary widely in their reactivity to CO2. One important mineral, brucite, reacts rapidly with CO2. Thermogravimetric analysis was used to identify and quantify brucite abundances. The developed method improved brucite detection limits and could quantify brucite when phases, which interfere with identification by X-ray diffraction, are present. Centimetre-scale injection experiments were conducted on well-graded tailings grain size distributions on serpentinite (brucite-rich) and kimberlite (brucite-poor) samples. Well-graded grain size distributions enabled the permeability to inject into the porous medium. Reactive fine-grained brucite released most of its Mg, while lizardite leached a minority of its Mg in the brucite-poor samples. 0.1 to 1 wt.% CO2 was sequestered into major hydromagnesite and minor nesquehonite. Metre-scale experiments in kimberlite tailings increased the scale and designed methods to achieve mine-scale injection. Compaction testing evaluated the relationship between the dry density and moisture content of the tailings. 6-metre-long pipe experiments replicated the reactivity of the centimetre scale. Scalable pad experiments injected CO2 through a perforated pipe into a layer of mixed coarse and fine tailings. Plume development throughout the pad led to homogeneous reactivity. Heterogeneity and low abundances of sequestered carbon made confirming the gas-phase mass balance on the injected CO2 using total inorganic carbon difficult. iv  A similar magnitude of carbon was sequestered in PK, as observed at the centimetre-scale, varying from 0.1 to 0.2 wt.% CO2. Carbon mineralization in mine tailings via CO2 injection has been shown to be effective in accelerating passive sequestration rates. This observed reactivity equates to a minor reduction in mine emissions at carbon-intensive mines while being enough capacity to make low-carbon mines carbon neutral.   v  Lay Summary Climate change is occurring because of human impacts on the atmosphere. Ways to capture and store CO2 are needed. The largest sink for carbon on the planet is in natural carbonate minerals. It is possible to put CO2 into minerals, and if this can be done in large amounts, it may help with climate change. Mine tailings are waste rock produced from mining. The waste rock can capture the CO2 emissions as new minerals. By injecting CO2 into the waste rock, the chemical reactions which form new minerals are increased as the supply of CO2 is increased. CO2 injection was studied at centimetre and metre scales, and different conditions were studied to improve the process. The results show that many mines could reduce, or even capture all of, their emissions by using their mine waste. vi  Preface This thesis is composed of an introductory chapter, three research chapters, and a concluding chapter. The overarching research program was designed by Prof. Greg Dipple, with contributions to the objectives made by the author and Dr. Connor Turvey. Prof. Ulrich Mayer contributed at committee meetings and through discussions. Chapter 1 was written by the author and benefitted from edits from Irene Fabris, Dr. Connor Turvey and Prof. Greg Dipple. The research conducted in Chapter 2 was done in collaboration with Dr. Connor Turvey. Prof. Greg Dipple and Dr. Connor Turvey designed the study, with Dr. Connor Turvey and the author preparing samples and conducting the thermogravimetric analysis. The author analyzed the results from all samples. Dr. Connor Turvey and the author interpreted the results, and the author wrote the thesis chapter. Prof. Greg Dipple and Dr. Connor Turvey edited the chapter. Chapter 3 was written as a thesis chapter. The research was designed by Prof. Greg Dipple and the author. The author conducted the experiments and analytical work. Ethan Alban assisted in the lab with monitoring experiments and logging data. Durjoy Baidya sampled and provided the diesel flue gas. Xueya Lu conducted the batch dissolution experiments. Suleiman Mohamed assisted with sample preparation. Maureen Soon and Frances Jones analyzed for total inorganic carbon, and Gethin Owen conducted the SEM analysis. The author is not responsible for the analysis of the whole-rock chemistry, BET surface area, PSD, or the QEMSCAN analysis. The author interpreted all results with advice from Prof. Greg Dipple. The author wrote the chapter, and edits were provided by Sterling Vanderzee, Xueya Lu, Peter Scheuermann, Irene Fabris, Dr. Connor Turvey, and Prof. Greg Dipple. vii  Chapter 4 was also written as a thesis research chapter. Prof. Greg Dipple and the author designed the research approach and objectives. The author performed all the experiments and analytical work. Ethan Alban assisted in the field and lab with experimental set-up, sampling, and tear down. Ethan Alban performed the Standard Proctor compaction testing. Suleiman Mohamed assisted with sample preparation. Maureen Soon analyzed for total inorganic carbon. The author is not responsible for the analysis of the whole-rock chemistry, total carbon, BET surface area, PSD, or the QEMSCAN analysis. The author interpreted all the results with input from Prof. Greg Dipple. The author wrote the chapter, and it was edited by Katrin Steinthorsdottir, Dr. Connor Turvey and Prof. Greg Dipple. Chapter 5 was written by the author and was edited by Dr. Connor Turvey and Prof. Greg Dipple. viii  Table of Contents  Abstract ............................................................................................................................. iii Lay Summary .....................................................................................................................v Preface ............................................................................................................................... vi Table of Contents ........................................................................................................... viii List of Tables .................................................................................................................. xxi List of Figures ............................................................................................................... xxiii List of Symbols ............................................................................................................. xxxi List of Abbreviations ................................................................................................. xxxiv Acknowledgements ................................................................................................... xxxvii Dedication ................................................................................................................... xxxix Chapter 1: Introduction ....................................................................................................1 1.1 Carbon Sequestration .......................................................................................... 1 1.2 Carbon Mineralization ........................................................................................ 1 1.3 Ultramafic Mine Tailings .................................................................................... 3 1.4 Accelerating Carbon Mineralization in Mine Tailings via CO2 Injection .......... 5 1.5 Thesis Objectives ................................................................................................ 7 1.6 Thesis Structure .................................................................................................. 8 Chapter 2: Brucite Identification by Thermogravimetric Analysis ............................10 2.1 Synopsis ............................................................................................................ 10 2.2 Introduction ....................................................................................................... 11 2.3 Methods............................................................................................................. 13 ix  2.3.1 Mineral Standard Preparation ....................................................................... 13 2.3.2 Analytical Methods ....................................................................................... 16 2.3.2.1 Thermogravimetric Analysis ................................................................. 16 2.3.2.2 Quantitative Thermogravimetric Analysis ............................................. 16 2.3.2.2.1 Total Mass Loss .............................................................................. 18 2.3.2.2.2 Linear Extrapolation ....................................................................... 18 2.3.2.2.3 Exponential Interpolation ................................................................ 19 2.4 Results ............................................................................................................... 20 2.4.1 Brucite Detection Limit ................................................................................ 20 2.4.2 Brucite Abundance Measurement ................................................................. 23 2.4.3 Correction Factor .......................................................................................... 25 2.5 Discussion ......................................................................................................... 27 2.5.1 Literature Methods and Results .................................................................... 27 2.5.2 Comparison with qXRD ............................................................................... 29 2.5.3 TGA Limitations ........................................................................................... 30 2.6 Implications ....................................................................................................... 31 Chapter 3: Pneumatic Permeability and Centimetre-Scale Reactivity ......................32 3.1 Synopsis ............................................................................................................ 32 3.2 Introduction ....................................................................................................... 33 3.3 Methods............................................................................................................. 36 3.3.1 Unreacted Sample Characterization .............................................................. 36 3.3.2 Experimental Set-Up ..................................................................................... 36 3.3.2.1 Permeability and CO2 Injection ............................................................. 36 x  3.3.2.2 Batch Dissolution ................................................................................... 40 3.3.3 Analysis of Experimental Results ................................................................. 40 3.3.3.1 Permeability ........................................................................................... 40 3.3.3.2 Sequestered CO2 Mass ........................................................................... 41 3.3.3.2.1 Gas Mass Balance ........................................................................... 41 3.3.3.2.2 TIC Increase .................................................................................... 41 3.3.3.2.3 Excluded Methods .......................................................................... 41 3.3.3.3 Water Mass Balance .............................................................................. 42 3.4 Results ............................................................................................................... 44 3.4.1 Characterization Results ............................................................................... 44 3.4.2 Permeability .................................................................................................. 49 3.4.3 Reaction Progression .................................................................................... 52 3.4.3.1 Processed Serpentinite CO2 Profiles ...................................................... 52 3.4.3.2 Processed Kimberlite CO2 Profiles ........................................................ 53 3.4.3.3 Instantaneous Carbonation Rates ........................................................... 55 3.4.4 Sequestered Carbon ...................................................................................... 56 3.4.4.1 Quantification ........................................................................................ 56 3.4.4.2 Precipitate Characterization ................................................................... 59 3.4.4.3 Reaction Efficiency ................................................................................ 64 3.4.5 Water Mass Balance ..................................................................................... 67 3.5 Discussion ......................................................................................................... 69 3.5.1 Permeability .................................................................................................. 69 3.5.2 Reactivity Controls ....................................................................................... 72 xi  3.5.2.1 Mineralogy ............................................................................................. 72 3.5.2.1.1 Source of Mg ................................................................................... 72 3.5.2.1.1.1 Brucite ...................................................................................... 72 3.5.2.1.1.2 Serpentine ................................................................................ 75 3.5.2.1.2 Source of Ca .................................................................................... 77 3.5.2.2 Particle Size Distribution ....................................................................... 78 3.5.2.3 Pore Saturation ....................................................................................... 80 3.5.2.4 CO2 Supply ............................................................................................ 82 3.5.3 Carbonate Stability ........................................................................................ 86 3.6 Implications ....................................................................................................... 90 3.6.1 Sequestration Magnitude .............................................................................. 90 3.6.2 Large-Scale Implementation ......................................................................... 93 3.6.2.1 Characterization ..................................................................................... 93 3.6.2.2 Physical Design ...................................................................................... 95 3.6.2.3 Monitoring and Verification .................................................................. 96 Chapter 4: Metre-Scale Reactivity and Field Pilot Design ..........................................98 4.1 Synopsis ............................................................................................................ 98 4.2 Introduction ....................................................................................................... 99 4.3 Methods........................................................................................................... 102 4.3.1 Sample Characterization ............................................................................. 102 4.3.2 Experimental Set-Up ................................................................................... 103 4.3.2.1 Six Metre Pipe Experiments ................................................................ 103 4.3.2.2 Pad Experiments .................................................................................. 107 xii  4.3.2.2.1 Pad-1 ............................................................................................. 107 4.3.2.2.2 Pad-2 ............................................................................................. 111 4.3.3 Analysis of Experimental Results ............................................................... 113 4.3.3.1 Permeability ......................................................................................... 113 4.3.3.2 Sequestered CO2 Mass ......................................................................... 114 4.3.3.2.1 Gas Mass Balance ......................................................................... 114 4.3.3.2.2 TIC Increase .................................................................................. 115 4.3.3.2.3 TC Increase ................................................................................... 115 4.3.3.3 Water Mass Balance ............................................................................ 115 4.4 Results ............................................................................................................. 117 4.4.1 Characterization Results ............................................................................. 117 4.4.2 Standard Proctor Compaction ..................................................................... 120 4.4.3 Pipe Injection Experiments ......................................................................... 122 4.4.3.1 Permeability ......................................................................................... 122 4.4.3.2 Reaction Progression ........................................................................... 123 4.4.3.2.1 Pipe-1 ............................................................................................ 123 4.4.3.2.2 Pipe-2 ............................................................................................ 126 4.4.3.2.3 Instantaneous Reaction Rates ....................................................... 127 4.4.3.3 Sequestered Carbon ............................................................................. 129 4.4.3.3.1 Pipe-1 ............................................................................................ 129 4.4.3.3.1.1 Bulk Sample Total Inorganic Carbon .................................... 129 4.4.3.3.1.2 Sieved Fines Total Inorganic Carbon .................................... 132 4.4.3.3.1.3 Gas Mass Balance .................................................................. 132 xiii  4.4.3.3.1.4 Reaction Homogeneity ........................................................... 134 4.4.3.3.2 Pipe-2 ............................................................................................ 136 4.4.3.3.2.1 Bulk Sample Total Inorganic Carbon .................................... 136 4.4.3.3.2.2 Sieved Fines Total Inorganic Carbon .................................... 138 4.4.3.3.2.3 Bulk Sample Total Carbon ..................................................... 138 4.4.3.3.2.4 Gas Mass Balance .................................................................. 139 4.4.3.3.2.5 Reaction Homogeneity ........................................................... 139 4.4.3.4 Water Mass Balance ............................................................................ 141 4.4.4 Pad Injection Experiments .......................................................................... 141 4.4.4.1 Reaction Progression ........................................................................... 141 4.4.4.1.1 Pad-1 ............................................................................................. 141 4.4.4.1.2 Pad-2 ............................................................................................. 144 4.4.4.2 Sequestered Carbon ............................................................................. 146 4.4.4.2.1 Pad-1 ............................................................................................. 146 4.4.4.2.1.1 Bulk Sample Total Inorganic Carbon .................................... 146 4.4.4.2.1.2 Gas Mass Balance .................................................................. 148 4.4.4.2.1.3 Reaction Homogeneity ........................................................... 151 4.4.4.2.2 Pad-2 ............................................................................................. 153 4.4.4.2.2.1 Bulk Sample Total Inorganic Carbon .................................... 153 4.4.4.2.2.2 Gas Mass Balance .................................................................. 153 4.4.4.2.2.3 Reaction Homogeneity ........................................................... 154 4.4.4.3 Moisture Content ................................................................................. 156 4.5 Discussion ....................................................................................................... 158 xiv  4.5.1 Physical Properties and CO2 Injection ........................................................ 158 4.5.2 6-Metre Pipe Injection Performance ........................................................... 160 4.5.3 Pad Injection Performance .......................................................................... 165 4.5.3.1 Pad-1 .................................................................................................... 165 4.5.3.2 Pad-2 .................................................................................................... 167 4.5.4 Carbon Accounting ..................................................................................... 171 4.6 Implications ..................................................................................................... 176 4.6.1 Magnitude of Carbon Mineralization .......................................................... 176 4.6.2 Large-Scale Design Recommendations ...................................................... 178 4.6.2.1 Porous Medium Design........................................................................ 178 4.6.2.2 Carbon Mineralization Via Injection ................................................... 182 Chapter 5: Conclusion ...................................................................................................185 5.1 Research Objectives ........................................................................................ 185 5.2 Research Outcomes ......................................................................................... 185 5.2.1 Porous Medium Conditions ........................................................................ 185 5.2.1.1 Quantified Parameters .......................................................................... 185 5.2.1.2 Experimental Parameters ..................................................................... 186 5.2.2 CO2 Injection Demonstration and Verification ........................................... 187 5.2.2.1 Injection Reaction at Scale ................................................................... 187 5.2.2.2 Carbon Verification ............................................................................. 188 5.3 CO2 Injection Field Pilot Design .................................................................... 189 5.4 Future Research Suggestions .......................................................................... 192 References .......................................................................................................................194 xv  Appendices ......................................................................................................................217 Appendix 1 : Appendix to Chapter 2 .......................................................................... 217 A1.1 Detailed Methods ........................................................................................... 217 A1.1.1 Sample Preparation .................................................................................. 217 A1.1.2 X-Ray Diffraction .................................................................................... 217 A1.2 Detailed Results ............................................................................................. 219 A1.2.1 X-Ray Diffraction .................................................................................... 219 A1.2.2 Thermogravimetric Analysis .................................................................... 223 Appendix 2 : Appendix to Chapter 3 .......................................................................... 241 A2.1 Detailed Experimental Set-Up ....................................................................... 241 A2.1.1 Pneumatic Permeability ........................................................................... 241 A2.1.2 CO2 Injection ............................................................................................ 242 A2.2 Detailed Methods ........................................................................................... 243 A2.2.1 Sample Characterization .......................................................................... 243 A2.2.1.1 Sample Preparation ........................................................................... 243 A2.2.1.2 X-Ray Diffraction ............................................................................. 243 A2.2.1.3 Thermogravimetric Analysis ............................................................. 245 A2.2.1.4 Total Inorganic Carbon ..................................................................... 246 A2.2.1.5 Major Oxide Composition ................................................................ 246 A2.2.1.6 Particle Size Distribution .................................................................. 246 A2.2.1.7 Surface Area ...................................................................................... 247 A2.2.1.8 Scanning Electron Microscopy ......................................................... 247 A2.2.1.9 Quantitative Evaluation of Materials by SEM .................................. 247 xvi  A2.2.2 Pneumatic Permeability ........................................................................... 248 A2.2.3 CO2 Mass Balance ................................................................................... 248 A2.2.3.1 Correcting Mass Balance to Mineralized Carbon ............................. 249 A2.2.4 Total Inorganic Carbon Increase .............................................................. 251 A2.2.5 Evaporative Moisture Loss ...................................................................... 252 A2.3 Detailed Results ............................................................................................. 254 A2.3.1 Sample Characterization .......................................................................... 254 A2.3.1.1 X-Ray Diffraction ............................................................................. 254 A2.3.1.2 Particle Size Distribution .................................................................. 261 A2.3.1.3 Quantitative Evaluation of Materials by SEM .................................. 262 A2.3.1.4 Thermogravimetric Analysis ............................................................. 263 A2.3.2 Pneumatic Permeability ........................................................................... 265 A2.3.3 CO2 Injection Detailed Results ................................................................ 280 A2.3.3.1 Concentration Profiles ....................................................................... 280 A2.3.3.2 Carbon Mineralization Rates ............................................................ 284 A2.3.3.3 Mass of Sequestered CO2 .................................................................. 286 A2.3.3.4 CO2 Concentration Data .................................................................... 289 A2.3.3.5 Precipitate Characterization .............................................................. 308 A2.3.3.5.1 X-Ray Diffraction ...................................................................... 308 A2.3.3.5.2 Thermogravimetric Analysis ...................................................... 311 A2.3.3.5.3 Scanning Electron Microscopy .................................................. 314 A2.3.3.6 Reaction Efficiency ........................................................................... 316 A2.3.3.6.1 Batch Dissolution Test Comparison .......................................... 316 xvii  A2.3.3.6.2 MgO Reactivity .......................................................................... 318 A2.3.3.6.3 Mineral Reactivity ...................................................................... 320 A2.3.3.7 Water Mass Balance .......................................................................... 322 A2.4 Discussion ...................................................................................................... 323 A2.4.1 FPS Labile Mg ......................................................................................... 323 A2.4.2 Calcium Source ........................................................................................ 323 A2.4.2.1 Smectite Cation Exchange ................................................................ 323 A2.4.2.2 Ca-Silicate Dissolution ..................................................................... 324 A2.4.3 Particle Size Distribution ......................................................................... 325 A2.4.4 CO2 Supply .............................................................................................. 326 A2.4.5 Carbonate Stability ................................................................................... 327 A2.5 Implications .................................................................................................... 331 A2.5.1 Sequestration Magnitude .......................................................................... 331 Appendix 3 : Appendix to Chapter 4 .......................................................................... 332 A3.1 Detailed Experimental Set-Up ....................................................................... 332 A3.1.1 Pipe-1 ....................................................................................................... 332 A3.1.2 Pipe-2 ....................................................................................................... 333 A3.1.3 Pad-1 ........................................................................................................ 334 A3.1.4 Pad-2 ........................................................................................................ 338 A3.2 Detailed Methods ........................................................................................... 343 A3.2.1 Sample Characterization .......................................................................... 343 A3.2.1.1 Sample Preparation ........................................................................... 343 A3.2.1.2 X-Ray Diffraction ............................................................................. 343 xviii  A3.2.1.3 Thermogravimetric Analysis ............................................................. 344 A3.2.1.4 Total Inorganic Carbon ..................................................................... 346 A3.2.1.5 Total Carbon ..................................................................................... 346 A3.2.1.6 Major Oxide Composition ................................................................ 346 A3.2.1.7 Particle Size Distribution .................................................................. 346 A3.2.1.8 Surface Area ...................................................................................... 347 A3.2.1.9 Quantitative Evaluation of Materials by SEM .................................. 347 A3.2.2 Standard Proctor Compaction .................................................................. 347 A3.2.3 Pneumatic Permeability ........................................................................... 348 A3.2.4 CO2 Mass Balance ................................................................................... 349 A3.2.4.1 Correcting Mass Balance to Mineralized Carbon ............................. 350 A3.2.5 Total Inorganic Carbon Increase .............................................................. 352 A3.2.5.1 Sieved Sample TIC Increase ............................................................. 352 A3.2.6 Total Carbon ............................................................................................ 353 A3.2.7 Evaporative Moisture Loss ...................................................................... 353 A3.3 Detailed Results ............................................................................................. 354 A3.3.1 Sample Characterization .......................................................................... 354 A3.3.1.1 X-Ray Diffraction ............................................................................. 354 A3.3.1.2 Particle Size Distribution .................................................................. 358 A3.3.1.3 Quantitative Evaluation of Materials by SEM .................................. 359 A3.3.1.4 Thermogravimetric Analysis ............................................................. 360 A3.3.2 Standard Proctor Compaction .................................................................. 362 A3.3.3 Pipe Experiment Permeability ................................................................. 363 xix  A3.3.4 Pipe and Pad Experiment CO2 Injection .................................................. 365 A3.3.4.1 Pipe-1 Flow Rate Analysis ................................................................ 365 A3.3.4.2 Gas Phase Data .................................................................................. 366 A3.3.5 Solid Phase Analysis ................................................................................ 390 A3.3.6 Reaction Efficiency .................................................................................. 396 A3.3.6.1 MgO Reactivity ................................................................................. 396 A3.3.6.2 Mineral Reactivity ............................................................................. 399 A3.4 Discussion ...................................................................................................... 401 A3.4.1 Reaction Efficiency .................................................................................. 401 A3.4.2 Pad-2 Gas Flow Simulation ..................................................................... 402 A3.5 Implications .................................................................................................... 403 A3.5.1 Sequestration Magnitude .......................................................................... 403 Appendix 4 : Pneumatic Permeameter Development ................................................. 404 A4.1 Introduction .................................................................................................... 404 A4.2 Literature Review ........................................................................................... 405 A4.2.1 Darcy\u2019s Law ............................................................................................. 405 A4.2.1.1 Reynold\u2019s Number and Laminar Flow .............................................. 405 A4.2.2 Pneumatic and Hydraulic Permeability Relationships ............................. 406 A4.2.3 Effect of Physical Properties on Permeability ......................................... 407 A4.2.4 Permeability Measurement Techniques ................................................... 410 A4.2.4.1 Hydraulic Tests ................................................................................. 410 A4.2.4.1.1 Constant Head Permeameter ...................................................... 410 A4.2.4.1.2 Falling Head Permeameter ......................................................... 410 xx  A4.2.4.2 Pneumatic Tests ................................................................................ 411 A4.2.4.2.1 Steady State ................................................................................ 411 A4.2.4.2.2 Pulse Decay ................................................................................ 411 A4.3 Methodology .................................................................................................. 413 A4.3.1 Pneumatic Permeameter ........................................................................... 413 A4.3.2 Constant Head Permeameter .................................................................... 414 A4.3.3 Procedure ................................................................................................. 414 A4.3.3.1 Pneumatic Test .................................................................................. 415 A4.3.3.1.1 Reynold\u2019s Number Verification ................................................. 416 A4.3.3.2 Hydraulic Test ................................................................................... 417 A4.4 Results ............................................................................................................ 418 A4.5 Discussion ...................................................................................................... 421 A4.6 Conclusion ..................................................................................................... 423 A4.7 Supplementary Material ................................................................................. 424 A4.7.1 Pneumatic Permeability Calculations ...................................................... 424 A4.7.2 Hydraulic Permeability Calculations ....................................................... 429  xxi  List of Tables  Table 2.1. Results from all standards for each quantification method with their respective errors. ................................................................................................................................ 14 Table 3.1. List of tailings samples. ................................................................................... 36 Table 3.2. CO2 injection experimental conditions. ........................................................... 39 Table 3.3. Mean whole-rock chemistry and standard deviations as determined by XRF or ICP-AES. .......................................................................................................................... 45 Table 3.4. TIC, surface area, grain size, and density of raw materials. ............................ 46 Table 3.5. Mean mineralogical abundance and standard deviations as determined by qXRD and TGA. ........................................................................................................................... 47 Table 3.6. Permeability of asbestos-bearing processed serpentinite. ................................ 51 Table 3.7. Permeability measurements before and after CO2 injection reactivity. ........... 52 Table 3.8. Amount of sequestered CO2 and dominant mineral reactivity. ....................... 58 Table 3.9. Comparison of injection and batch dissolution results. ................................... 66 Table 3.10. Pore water loss after CO2 injection experimental completion. ...................... 68 Table 4.1. List of tailings samples from Gahcho Ku\u00e9 Diamond Mine. .......................... 102 Table 4.2. 6-metre pipe injection experimental conditions. ............................................ 106 Table 4.3. Pad injection experimental set-up conditions. ............................................... 110 Table 4.4. Applied methods to assess the magnitude of sequestered carbon in each experiment. ...................................................................................................................... 114 Table 4.5. Mean whole-rock chemistry and standard deviations as determined by ICP-AES.......................................................................................................................................... 117 xxii  Table 4.6. TIC, surface area, grain size, and density of PK samples. ............................. 117 Table 4.7. Mean mineralogical abundance and standard deviations as determined by qXRD and TGA. ......................................................................................................................... 118 Table 4.8. Sequestered carbon in the 6-metre pipe injection experiments as determined by TIC and a gas phase mass balance. ................................................................................. 131 Table 4.9. Sequestered carbon in the pad injection experiments as determined by TIC and a gas-phase mass balance. ............................................................................................... 150 Table 4.10. Moisture contents of the pad injection experiments before and after injection.......................................................................................................................................... 157  xxiii  List of Figures  Figure 1.1. Potential avenues to couple CDR strategies and fossil fuel combustion with carbon mineralization as the optimal form of geological storage. ...................................... 2 Figure 2.1. Schematic of how the various methods identify the mass loss interval to calculate the dehydration of water from brucite. The mass loss interval is indicated by the double-ended arrows for the total mass loss, the linear extrapolation method and the exponential interpolation method in black, blue and red, respectively. These mass loss intervals are equivalent to the area under the curve for the derivative mass loss plot, bordered by the dashed lines at the beginning and end of the brucite decomposition temperature interval. The total mass loss is equivalent to the area between the derivative mass loss curve and the x-axis. The linear method is equivalent to the area bordered by the derivative mass loss curve and the blue extrapolation line. Lastly, the exponential method is equivalent to the area between the derivative mass loss and the red interpolation curves............................................................................................................................................ 17 Figure 2.2. A. TG mass loss from PS-STD-1, PS-STD-2, and PS-STD-3. B. DTG curves of PS-STD-1, PS-STD-2, and PS-STD-3. PS-STD-1 had below detection brucite, while PS-STD-2 was spiked with 0.3 wt.% brucite, and PS-STD-3 had 1.0 wt.% brucite added............................................................................................................................................ 21 Figure 2.3. A. TG mass loss from PK-STD-1, PK-STD-2, and PK-STD-3. B. DTG curves of PK-STD-1, PK-STD-2, and PK-STD-3. PK-STD-1 had below detection brucite, while PK-STD-2 was spiked with 0.3 wt.% brucite, and PK-STD-3 had 0.4 wt.% brucite added............................................................................................................................................ 22 xxiv  Figure 2.4. Calculated brucite abundances by the total mass loss, linear extrapolation, and exponential interpolation methods for the processed serpentinite and kimberlite standards............................................................................................................................................ 24 Figure 2.5. Calculated brucite abundance and relative error results for the exponential method with the correction factor applied to all mineral standards. ................................. 26 Figure 3.1. Pneumatic permeability and CO2 injection experimental set-up. ................... 38 Figure 3.2. Permeability and porosity of tailings grain size distributions ranging from coarse to fine, at 30% (blue) and 60% (red) pore saturation. Error bars indicate the standard deviation from triplicate tests and are smaller than the symbol where not indicated. Dashed lines indicate the general trends. ....................................................................................... 50 Figure 3.3. Gas-phase CO2 concentrations over time for the processed serpentinite injection experiments. ...................................................................................................................... 53 Figure 3.4. Gas-phase CO2 concentrations over time for the processed kimberlite injection experiments PK-E2 (A), PK-E3 (B) and PK-E4 (C). ....................................................... 54 Figure 3.5. Injection and carbonation rates over time in PS and PK injection experiments............................................................................................................................................ 56 Figure 3.6. Mass of sequestered CO2 over time as quantified from a gas mass balance (with a thin-line error envelope) and TIC increases in carbonated versus initial samples, with a 95% confidence interval indicated. The confidence interval for PS-E9 is smaller than the symbol. .............................................................................................................................. 57 Figure 3.7. XRD pattern of a carbonated sample from injection experiment PS-E1, showing the detection of nesquehonite. ........................................................................................... 60 xxv  Figure 3.8. TG and DTG curves of the initial and carbonated samples from injection experiment PS-E9. In the initial sample, the peaks refer to the loss of adsorbed water (40 \u2013 150\u00b0C), dehydroxylation of brucite (300 \u2013 430\u00b0C) and dehydroxylation of serpentine (460 \u2013 850\u00b0C). For the carbonated sample, the mass loss ranges are due to the loss of adsorbed water (40 \u2013 150\u00b0C), dehydration of hydromagnesite (150 \u2013 250\u00b0C), dehydroxylation of brucite and hydromagnesite and decarbonation of hydromagnesite (300 \u2013 460\u00b0C), and the dehydroxylation of serpentine (460 \u2013 850\u00b0C). .................................................................. 61 Figure 3.9. Representative scanning electron micrographs of precipitated Mg-carbonates from PS-E1. A-D. Secondary electron micrographs. A. Pervasive lenticular rosettes of hydromagnesite. B. Lenticular rosettes of hydromagnesite precipitated along the planar edge of serpentine. C. Massive prismatic needles of nesquehonite. D. Prismatic needles of nesquehonite showing desiccation cracks. E-F. Backscattered electron micrographs. E. Mg-carbonates having precipitated at mineral-pore interfaces where the water film would dissolve CO2 and Mg. F. Hydromagnesite rosettes precipitated out of brucite-rich FPS. 63 Figure 3.10. A. The captured mass of CO2 from the PS experiment series is plotted versus the mass of injected CO2, with their respective confidence intervals. A line of best fit through the data points and a trend to estimate the early time reactivity are shown. The flow rate magnitude is shown by the diamond, square, triangle and circle symbols, from highest to lowest. B. The efficiency of the captured CO2 mass against the mass of injected CO2 is plotted with an approximate trendline through the data. ....................................... 84 Figure 3.11. MgO-H2O-CO2 ternary diagram for the possible compositions of the precipitated Mg-carbonates among brucite, magnesite and hydrous Mg-carbonate minerals known to form in CO2 injection experimental conditions. Dotted lines represent a MgO: xxvi  CO2 ratio of 2, 1.25, and 1. The range in the initial precipitated Mg-carbonate composition from the PS experiment series is presented in red for each ratio, with the extended red lines being the standard deviation on the H2O content. Protohydromagnesite is included due to its proximity to one of the possible precipitated compositions. Dark red indicates the most likely precipitated composition. ........................................................................................ 88 Figure 3.12. CO2 emission rates and injection (this study) and passive sequestration rates for the Diavik Diamond Mine (NT, Canada), Gahcho Ku\u00e9 Diamond Mine (NT, Canada), Mount Keith Nickel Mine (WA, Australia), Dumont Nickel Project (QC, Canada), and the Baptiste Nickel Project (BC, Canada). Percentages above the symbols are the amount of the total emissions that could be sequestered at the respective mine. .............................. 91 Figure 4.1. Schematic of the 6-metre pipe injection experimental set-up showing the gas flow control, hydration and injection into the compacted tailings, with the installed sensors shown. ............................................................................................................................. 104 Figure 4.2. Schematic of the Pad-1 injection experimental set-up showing the gas flow control, hydration and injection into the mixed PK layer, with embedded CO2 sensors and effluent ports shown. ....................................................................................................... 108 Figure 4.3. Schematic of the Pad-2 injection experimental set-up showing the gas flow control, hydration and injection into the base CPK layer through the perforated pipe. Locations of the CO2 sensors and the effluent ports are shown. .................................... 112 Figure 4.4. Standard Proctor compaction testing results on PK blends of 25 and 33 wt.% FPK-2 with CPK-2. The maximum dry density (MDD) and the optimum moisture content (OMC) have been determined. The 90% value of the MDD for both mixes and the 60% and 100% pore saturation lines are indicated. ................................................................ 121 xxvii  Figure 4.5. The initial permeability of the 6-metre pipe injection experiments in the context of the permeability of tailings grain size distributions ranging from coarse to fine, at 30% pore saturation. Error bars indicate the standard deviation from triplicate tests and are smaller than the symbol where not indicated. ................................................................. 123 Figure 4.6. Gas-phase results of the reaction and sequestration of simulated flue gas CO2 over time within the Pipe-1 6-metre injection experiment. A. Gas-phase CO2 concentration profiles overtime at the pipe inlet, midpoint, and outlet. A generator power outage is indicated at 52 hours, coincident with the leak that reduced the injected flowrate. B. Relative humidity (RH) at the inlet and outlet during the experiment. C. Ambient, inlet, and outlet gas temperatures over time. ........................................................................... 125 Figure 4.7. Gas-phase results of the reaction and sequestration of simulated flue gas CO2 over time within the Pipe-2 6-metre injection experiment. A. Gas-phase CO2 concentration profiles over time. A building power outage is indicated at 97 hours, and the planned reduction in flowrate at 400 hours. B. Relative humidity (RH) at the inlet and outlet throughout the experiment. C. Inlet, and outlet gas temperatures over time. ................. 127 Figure 4.8. Instantaneous injection (dashed and dash-dotted lines) and carbonation (solid lines) rates for both Pipe-1 and Pipe-2 injection experiments due to carbon mineralization and solubility trapping. The two dash-dotted lines and the two lines with squares indicate the injection and reaction uncertainty during phase 2 injection for the Pipe-1 experiment.......................................................................................................................................... 128 Figure 4.9. Histograms showing the distribution of TIC measurements from unreacted and reacted samples for bulk (A) and sieved (B) samples from Pipe-1. The solid lines show the xxviii  average carbonated and unreacted values, while the dashed lines show the mean value in the unreacted FPK and CPK samples. ............................................................................ 130 Figure 4.10. Gas and solid-phase results of the reaction and sequestration of simulated flue gas CO2 over time within the 6-metre pipe injection experiments. Sequestered CO2 mass over time was quantified from gas concentrations discounting the expected DIC (solid and dashed lines) and TIC increases in carbonated versus initial bulk (diamonds) and sieved fine fraction (squares) samples, with a 95% confidence interval indicated. Pipe-1 has uncertainty around the gas phase estimate due to a leak that reduced the flow rate. ...... 133 Figure 4.11. TIC for each reacted sample along the length of Pipe-1 at upper (triangle), middle (square) and lower (circle) horizons. The solid and dashed lines show the mean unreacted and carbonated TIC values for the bulk and sieved samples, respectively. ... 135 Figure 4.12. Histograms showing the distribution of TIC measurements from unreacted and reacted samples for bulk (A) and sieved (B) samples from Pipe-2. The solid lines show the average carbonated and unreacted values, while the dashed lines show the mean value in the unreacted FPK and CPK samples. ........................................................................ 137 Figure 4.13. TIC for each sample along the length of Pipe-2 at upper (triangle), middle (square) and lower (circle) horizons. The solid and dashed lines show the mean unreacted and carbonated TIC values for the bulk and sieved samples, respectively. .................... 140 Figure 4.14. Aerial view of the interpolated CO2 concentrations throughout the mixed PK layer in the Pad-1 injection experiment measured by six embedded sensors (X\u2019s), with the injection (triangle) and effluent (squares or circle) ports shown. ................................... 143 Figure 4.15. Cross-section view of the interpolated CO2 concentration profiles in Pad-2 measured by 15 embedded CO2 sensors (X\u2019s) in the base CPK (indicated by a dashed black xxix  line) and the mixed PK layer throughout the experiment. The perforated injection pipe (triangle) and the effluent port (circle) are also indicated. .............................................. 145 Figure 4.16. Histograms showing the distribution of TIC measurements from unreacted and reacted samples for Pad-1 (A) and Pad-2 (B). The solid lines show the average carbonated and unreacted values, while the dashed lines show the mean value in the unreacted FPK, CPK and mixed PK samples. ................................................................ 147 Figure 4.17. Gas and solid-phase results of the reaction and sequestration of simulated flue gas CO2 over time within the pad injection experiments. Sequestered CO2 mass over time was quantified from gas concentrations discounting the expected DIC (solid lines) and from TIC increases in carbonated versus initial samples (circles, squares, diamonds and triangles), with a 95% confidence interval indicated. ..................................................... 149 Figure 4.18. TIC increase for each sample (X\u2019s) or sample group as distributed within Pad-1. A. Cross-section view through the middle of the pad, showing the upper FPK, mixed PK and lower FPK layers. B. Aerial view of the mixed PK layer. ....................................... 152 Figure 4.19. TIC increases for each sample (X\u2019s) or sample group distributed within Pad-2 presented as a cross-section view through the middle of the pad, showing the cap CPK and FPK layers, the mixed PK layer and the base CPK layer. ....................................... 155 Figure 4.20. Estimated CO2 concentration fronts through Pipe-2 at 30, 300 and 450 hours of injection. ..................................................................................................................... 163 Figure 4.21. Injection (dashed lines) and carbonation rates (solid lines) for both the 6-metre-long injection experiments Pipe-1 (first half during injection phase 1) and Pipe-2 (injection phase 1) and the 12-centimetre long injection experiment PK-E3 from Chapter 3....................................................................................................................................... 164 xxx  Figure 4.22. Water drainage amongst the layers of Pad-1. ............................................. 167 Figure 4.23. Modeled flow direction (grey) and pressure equipotential lines (blue) of injected gas from the perforated pipe (lower left corner) into Pad-2. ............................. 169 Figure 4.24. Water drainage and capillary action amongst the layers of Pad-2. ............ 170 Figure 4.25. CO2 emissions, injection (this study), and passive sequestration rates for the Diavik Diamond Mine (NT, Canada), Gahcho Ku\u00e9 Diamond Mine (NT, Canada), Mount Keith Nickel Mine (WA, Australia), Dumont Nickel Project (QC, Canada), and the Baptiste Nickel Project (BC, Canada). Percentages are expressed as the amount of sequestration in relation to the emissions from the respective mine. Results from Chapter 3 are presented for Baptiste. Results from Chapter 3 for Gahcho Ku\u00e9 have been replaced with Chapter 4 results, specifically for the TIC increase (filled diamonds) and gas mass balance (unfilled diamonds) for the lab (Pipe-2 and Pad-2) injection experiments. ...... 177 Figure 4.26. An example of density and moisture content conditions for different locations within the tailings storage facility that enable the promotion of carbon sequestration via CO2 injection. .................................................................................................................. 181 Figure 5.1. Conceptual design of the large-scale implementation of CO2 injection into mine tailings management. ...................................................................................................... 191 xxxi  List of Symbols \u2206 \u2013 delta \u00baC \u2013 degrees Celsius \u00b5m \u2013 micrometre 13C \u2013 carbon 13 14C \u2013 carbon 14 a \u2013 constant Al2O3 \u2013 aluminum oxide Al \u2013 aluminum b \u2013 constant BruciteCorrected \u2013 Brucite abundance with the applied correction factor BruciteExp \u2013 Brucite abundance determine from the exponential interpolation method BruciteLin \u2013 Brucite abundance determine from the linear extrapolation method BruciteTotalMassLoss \u2013 Brucite abundance determine from the total mass loss method c \u2013 constant C \u2013 carbon Ca2+ \u2013 calcium cation Ca \u2013 calcium CaO \u2013 calcium oxide cm \u2013 centimetre cmol \u2013 centimole [CO2] \u2013 carbon dioxide concentration CO2 \u2013 carbon dioxide xxxii  CO \u2013 carbon monoxide d \u2013 constant Fe2+ \u2013 Divalent iron cation Fe2O3 \u2013 iron(III) oxide Fe3+ \u2013 trivalent iron cation Fe \u2013 iron g \u2013 gram Gt \u2013 gigaton H2O(g) \u2013 water vapour HCO3- \u2013 bicarbonate anion H2O \u2013 water hr \u2013 hour hrs \u2013 hours K2O \u2013 potassium oxide K \u2013 potassium kg \u2013 kilogram kt \u2013 kiloton L \u2013 litre m300 \u2013 mass loss at ~300\u00baC m415 \u2013 mass loss at ~415 \u00baC M \u2013 moles per litre m \u2013 metre mExp \u2013 mass loss determined from the exponential interpolation method xxxiii  Mg2+ \u2013 magnesium cation Mg \u2013 magnesium mg \u2013 milligram MgO \u2013 magnesium oxide Mg(OH)2 \u2013 brucite min \u2013 minute mL \u2013 millilitre mLin \u2013 mass loss determined from the linear extrapolation method mm \u2013 millimetre mmol \u2013 millimole Mt \u2013 megaton N2 \u2013 nitrogen Na2O \u2013 sodium oxide Na \u2013 sodium NOx \u2013 nitrogen oxides O2 \u2013 oxygen O \u2013 oxygen pCO2 \u2013 CO2 partial pressure Si \u2013 silicon SiO2 \u2013 silicon dioxide SO4 \u2013 sulphate anion SOx \u2013 sulphuric oxides Ti \u2013 titanium xxxiv  List of Abbreviations BC \u2013 British Columbia BECCS \u2013 bioenergy with carbon capture and storage BET \u2013 Brunauer-Emmett-Teller Brc \u2013 brucite CDR \u2013 carbon dioxide removal CI \u2013 confidence interval CO2 \u2013 carbon dioxide CPK \u2013 coarse processed kimberlite CPS \u2013 coarse processed serpentinite DAC \u2013 direct air capture DACCS \u2013 direct air carbon capture and storage DFG \u2013 diesel flue gas DIC \u2013 dissolved inorganic carbon DTG \u2013 derivative thermogravimetry E# \u2013 experiment number EW \u2013 enhanced weathering FPK \u2013 fine processed kimberlite FPS \u2013 fine processed serpentinite FS \u2013 Free State FTIR \u2013 Fourier-transform infrared spectroscopy HK \u2013 hypabyssal kimberlite Hmg \u2013 hydromagnesite xxxv  IBC \u2013 intermediate bulk container ICP-AES \u2013 inductively coupled plasma-atomic emission spectroscopy ICP-OES \u2013 inductively coupled plasma-optical emission spectroscopy LOI \u2013 loss on ignition MBD \u2013 maximum bulk density MDD \u2013 maximum dry density MS \u2013 mass spectrometry Nsq \u2013 nesquehonite NT \u2013 Northwest Territories OMC \u2013 optimum moisture content PK \u2013 processed kimberlite PS \u2013 processed serpentinite  PSA \u2013 asbestiform processed serpentinite PSD \u2013 particle size distribution QC \u2013 Quebec QEMSCAN \u2013 quantitative evaluation of materials by scanning electron microscopy qXRD \u2013 quantitative X-ray diffraction RH \u2013 relative humidity SEM \u2013 scanning electron microscopy SFG \u2013 simulated flue gas Srp \u2013 serpentine Std \u2013 standard TC \u2013 total carbon xxxvi  TG \u2013 Thermogravimetry TGA \u2013 thermogravimetric analysis TIC \u2013 total inorganic carbon TKB \u2013 tuffisitic kimberlite breccia TOC \u2013 total organic carbon tr \u2013 trace abundance TSF \u2013 tailings storage facility VT \u2013 Vermont WA \u2013 Western Australia wt.% - weight percent XRD \u2013 X-ray diffraction XRF \u2013 X-ray fluorescence xxxvii  Acknowledgements \u201cThe LORD God took the man and put him in the Garden of Eden to work it and take care of it.\u201d Genesis 2:15 NIV I would never have been interested in this research project had it not been for the Biblical principle of stewardship. It has indeed been the basis for my environmental moral compass, even though I once did not acknowledge it as I do now. It is to the LORD, who made every bit of grit I studied, who gave me purposeful work suitable to my interests and abilities, and who encouraged me every day with his love and hope, that I owe everything. I am indebted to my supervisor, Dr. Greg Dipple, for bringing me on as a student and redefining what I see as possible to accomplish both in research and in life. Under your tutelage, I have grown in understanding and in confidence. Thank you for your encouragement, wisdom, and trust. But above all, thank you for always being willing to joke and laugh with me regardless of how things were going. To Dr. Connor Turvey, I am grateful for your availability and approachability to ask the simple and difficult questions. Thank you for teaching me the magical ways of XRD and for all the commiseration along the way. Thank you to Dr. Ulrich Mayer for serving on my supervisory committee, for teaching me so much about geochemistry, and for being prone to scare me (in a good way!) with your thoughts related to modelling. Among all my colleagues, the first to thank is Ethan Alban for being the best undergraduate assistant I could have asked for and for suffering with me in the field. Thank you for your patience with me and for bringing your cheerful golden retriever attitude every day. Thank you to Durjoy Baidya for collaborating with me and proving that the economics of CO2 injection is not hopeless. I am grateful to Suleiman Mohamed for his tireless work in helping me with my samples and to Maureen Soon for analyzing my hundreds of TIC samples. Thank you to Gethin Owen for his xxxviii  perseverance with me in locating Mg-carbonates under the SEM. A big thank you to the other Carbmin team members: Katrin Steinthorsdottir, Sterling Vanderzee, Xueya Lu, Peter Scheuermann, Frances Jones, Anne-Martine Doucet, Xue Wang, Dave Zeko, Jaime Cutts, and Bethany Ladd, for everything along the way. I am grateful to my fellow students, Danielle Cocchetto and Eve Wicksteed, and to the rest of my church family, for their friendship and fellowship these past years. Thank you to Irene Fabris for kindly editing my work. Lastly, I owe a special thank you to my parents, Janet and Ed, who instilled in me a passion for stewardship and the environment and raised me to work hard and overcome obstacles. I am grateful for their love and moral support. Thank you to my siblings, Karen, Jenn and John, for their love and support. This research was funded by Natural Resources Canada\u2019s Clean Growth Program, De Beers Group, and FPX Nickel Corp. I am grateful for the scholarship support from the Department of Earth, Ocean and Atmospheric Sciences. xxxix  Dedication       To my parents, Janet and Ed.  1  Chapter 1: Introduction 1.1 Carbon Sequestration Under the threat of anthropogenic climate change, the world must steward the climate for the well-being of future generations. At the latest climate accord (Paris, 2015), the Intergovernmental Panel on Climate Change outlined emissions pathways to reduce greenhouse gas emissions and limit warming below 1.5\u00baC above pre-industrial levels (United Nations Framework Convention on Climate Change, 2016). All paths rely on carbon dioxide removal (CDR) strategies to capture and store 100 to 1000 Gt CO2 during the 21st century (IPCC, 2018). Biogenic methods such as afforestation and bioenergy with carbon capture and storage (BECCS) are expected to contribute the most (IPCC, 2018). Other proposed CDR strategies, such as direct air carbon dioxide capture and storage (DACCS) and enhanced weathering (EW), are less technologically ready and still under active research (de Coninck et al., 2018). All CDR strategies rely upon a mechanism to capture CO2 from a source and a sink to store the sequestered carbon. Geological storage has greater capacity and stability than either biogenic or ocean storage and consists of CO2 being either delivered into geological formations in-situ or exposed to extracted rocks ex-situ (de Coninck et al., 2018; Lackner et al., 1995, 1997, 2003; Seifritz, 1990).  1.2 Carbon Mineralization One form of carbon sequestration is carbon mineralization, which is the most desirable result of geological storage. Carbon mineralization occurs naturally as part of the carbon cycle due to rock weathering and sees cations released via weathering react with atmospheric CO2 to produce carbonate minerals (Berner et al., 1983). As a storage method, carbon mineralization must be coupled with a CO2 source and a strategy to capture and supply the CO2 (Figure 1.1). From the 2  atmosphere, CO2 can be captured and stored directly via EW. CO2 can also be captured from point sources, such as fossil fuel or biofuel combustion emissions (BECCS) and concentrated streams from direct air capture systems (DACCS), before being directed to geological storage.  Figure 1.1. Potential avenues to couple CDR strategies and fossil fuel combustion with carbon mineralization as the optimal form of geological storage. For carbon mineralization to be significant in the context of global CO2 emissions, large-scale cation feedstocks are required. Ultramafic rocks are typically targeted for carbon mineralization as they are rich in divalent cations such as Mg2+, Ca2+ and Fe2+, and form out of thermodynamic equilibrium with the atmosphere, meaning they weather more rapidly than other rock types (Kelemen et al., 2011). However, many industrial wastes are also suitable as reactive feedstocks for carbon mineralization, including ultramafic mine tailings, cement waste, steel slag, fly ash, and alumina waste (Bobicki et al., 2012; Power et al., 2013). Ultramafic mine tailings, produced from diamond, nickel, asbestos, chromite and platinum group element mines (Beinlich & Austrheim, 2012; Pronost et al., 2012; Vogelia et al., 2011; Wilson et al., 2006, 2009b, 2014), CO2CO2CO2Fossil Fuel ExtractionBiomass GrowthFossil Fuel & Biofuel CombustionDirect Air CaptureEnhanced WeatheringEx-situ Carbon MineralizationIn-situ Carbon Mineralization3  exemplify a waste stream representative of ultramafic rocks, the largest cation feedstock available, making mine tailings a model feedstock.  1.3 Ultramafic Mine Tailings The study of carbon mineralization in ultramafic mine tailings advances the technological readiness of carbon mineralization as a geological storage method coupled with CO2 capture strategies. Ultramafic mine tailings are rich in Mg-bearing minerals such as serpentine [Mg3Si2O5(OH)4], olivine [Mg2SiO4] and brucite [Mg(OH)2], are highly reactive due to the reduced grain size from comminution and have been historically and are currently produced in abundance significant to industrial and global scales (Gras et al., 2020; Oskierski et al., 2013; Power et al., 2013, 2020; Wilson et al., 2006, 2009b, 2014; Zarandi et al., 2016). The use of tailings reduces costs associated with acquiring and preparing ultramafic rocks, which is energy-intensive. Ultramafic rocks (<45 wt.% SiO2) can be divided into two broad categories, peridotite and pyroxenite, and are representative of the earth\u2019s mantle. Ultramafic rocks can be present at the earth\u2019s surface as part of obducted ophiolite complexes, intrusive complexes, and extrusive komatiites and pyroclastic deposits (Boudier & Nicolas, 1985; Milidragovic et al., 2018; Nixon, 1998; O\u2019Hanley & Dyar, 1993). Kimberlites are rare, potassic, intrusive ultramafic rocks, rich in volatiles, with distinct texture characteristics that form as volcanic pipes, dykes and sills (Clement et al., 1984). However, they typically have less Mg than other ultramafic rocks. Metamorphism of olivine to serpentine through serpentinization, a hydration process, commonly alters peridotite, pyroxenites, and kimberlites. This process is notable as it can lead to the formation of brucite. Ultramafic mine tailings are produced from several commodity types, with two examples being nickel and diamonds. Mine tailings reflect the mineralogical variability within and between 4  mineral deposits. Two exemplary deposits which posit the potential to mineralize carbon within their tailings are the Baptiste Nickel Project (BC, Canada) and the Gahcho Ku\u00e9 Diamond Mine (NT, Canada). The Batiste Nickel Project is situated 90 kilometres northwest of Fort St. James, BC, and is located in the Trembleur ultramafite, which is part of the Cache Creek terrane (Steinthorsdottir, 2021). The geology consists of largely peridotite (mostly harzburgite and lesser dunite), either partially or heavily serpentinized, with rare, significant carbonate alteration (Steinthorsdottir, 2021). Nickel mineralization occurs as awaruite, a nickel-iron alloy. Awaruite and brucite abundances increase with the degree of serpentinization and are related to the protolith type (Britten, 2017; Steinthorsdottir, 2021). Brucite is present at abundances up to 13 wt.% (Steinthorsdottir, 2021). As a mine, Baptiste would produce 43.8 Mt of tailings and 108 kt of CO2 per year for 24 years (Power et al., 2020; Turenne, 2021). Gahcho Ku\u00e9 is located 280 kilometres northeast of Yellowknife, NT, and consists of several kimberlite pipes of Cambrian age situated within the Slave Craton (Johnson & Pilotto, 2020). Surrounding the four main kimberlite pipes are felsic granites and gneisses. The kimberlite textures vary from tuffisitic kimberlite breccia (TKB) to hypabyssal kimberlite (HK)(Hetman et al., 2004). Within the TKB, country rock xenoliths are unaltered, olivine is serpentinized, clay minerals are abundant, and the phlogopite and serpentine matrix lacks carbonate minerals. The HK contains primary olivine, clay minerals are absent, and the matrix consists of monticellite, phlogopite, serpentine, and carbonate minerals (Hetman et al., 2004). Xenolith digestion in the HK is significant and leads to the formation of clinopyroxene, garnet, phlogopite and pectolite (Hetman et al., 2004; Niyazova, 2020). Gahcho Ku\u00e9 began mining in 2016, producing on average 3.2 Mt of 5  processed kimberlite and 122 kt of CO2 per year (De Beers Group, 2019; Johnson & Pilotto, 2020). Mine lifespan is expected to be until 2030 (Johnson & Pilotto, 2020).  1.4 Accelerating Carbon Mineralization in Mine Tailings via CO2 Injection Exposed to the atmosphere in the tailings storage facility (TSF), tailings passively mineralize CO2 from the air, as observed at diamond, nickel, asbestos and chromite mines (EW route from Figure 1.1)(Beaudoin et al., 2008; Beinlich & Austrheim, 2012; Nowamooz et al., 2018; Oskierski et al., 2013; Pronost et al., 2012; Turvey et al., 2018b; Wilson et al., 2006, 2011, 2014, 2009b). This passive carbon mineralization encompasses three main processes, 1) CO2 transportation, dissolution, hydration and acidity generation, 2) mineral dissolution and alkalinity generation, and 3) solid carbonate mineral precipitation. Passive reaction rates in ultramafic mine tailings are limited by the first process. Low diffusive transportation rates of CO2 from the atmosphere to the mineral surface at depth occur due to low tailings permeabilities, high water saturation, high tailings deposition rates, and climatic conditions, which can lead to excess precipitation or freezing (Bea et al., 2012; Gras et al., 2017; Kandji et al., 2015; Li et al., 2018; Power et al., 2020; Wilson et al., 2014; Zarandi et al., 2017a). Approaches that target the acceleration of ambient processes by focusing on the most reactive fraction of the material are required to lower the cost of the technology (Hitch & Dipple, 2012; Lu, 2020; Vanderzee et al., 2019). Increasing the rate of CO2 supply could be achieved by various methods, including enhancing passive reaction rates or by injecting CO2. Suggested solutions to promote passive carbon mineralization have included storing the tailings subaerially, increasing the surface area of the TSF, reducing the deposition rate, churning the surface, and co-disposal of fine-grained tailings with coarse-grained tailings or waste rock (Assima et al., 2014b; Hamilton et al., 2021; Li et al., 6  2018; Power et al., 2020; Wilson et al., 2014; Zarandi et al., 2017a). Each strategy has specific benefits and intervention costs. Intervening into tailings management practices creates new possibilities to enhance CO2 supply, such as delivering it by advection via injection (Hamilton et al., 2021; Harrison et al., 2013b; Li et al., 2018; Mervine et al., 2018; Nowamooz et al., 2018; Power et al., 2020; Pronost et al., 2012; Wilson et al., 2014). Carbon mineralization via injection would take place in addition to passive surface mineralization processes, as injection would take place in the subsurface. CO2 injection need not be limited by the pCO2 of air but by the availability of higher pCO2 sources, opening the opportunity to mineralize captured CO2 from point sources (BECCS, DACCS, and fossil fuels routes from Figure 1.1). Since mines commonly combust fossil fuels for electric power generation, concentrated CO2 point sources may be available. Advective delivery enables supplying enough CO2 to shift the rate-limiting process towards mineral dissolution (Harrison et al., 2016). Mineral dissolution is controlled by the mineral chemical structure and the strength of the bonds that need to be broken (Schott et al., 2009). For some key Mg-bearing minerals, such as brucite, dissolution rates are pCO2 dependent and increase linearly with pCO2 (Harrison et al., 2013a; Lu, 2020). Injection of enriched CO2 gas has been investigated experimentally on mineral phases and mine tailings in saturated and unsaturated conditions (Assima et al., 2012, 2013a, 2014d, 2014c; Hamilton et al., 2020; Harrison et al., 2013a, 2015, 2016, 2017; Lu, 2020; Paulo et al., 2021). This research has established guidelines on the rate-limiting conditions. Mineral dissolution rates are mineral-specific and grain size and surface area dependent. On a per mass basis, minerals carbonate in the following order of decreasing reactivity, brucite > wollastonite > chrysotile > lizardite\/antigorite > olivine (Assima et al., 2012, 2013a; Lu, 2020; Paulo et al., 2021). Finer grain 7  sizes react faster and to completion (Assima et al., 2013a; Harrison et al., 2015; Wilson et al., 2009a). Passivation of the mineral surface area occurs due to the formation of Si-rich layers on Mg-silicates, which resist dissolution, and the precipitation of Fe-hydroxides and non-porous Mg-carbonates (Andreani et al., 2009; Assima et al., 2012, 2013a, 2014d, 2014a, 2014c; B\u00e9arat et al., 2006; Daval et al., 2011; Harrison et al., 2015, 2016; H\u00f6velmann et al., 2012; Huijgen et al., 2006; Johnson et al., 2014; Park & Fan, 2004; Sissmann et al., 2014; Zarandi et al., 2016, 2017b). Water is required to facilitate CO2 and mineral dissolution and Mg-carbonate precipitation and can be a reactant in forming hydrous Mg-carbonates (Assima et al., 2012, 2013a, 2014b, 2014d; Harrison et al., 2015, 2016; Pronost et al., 2011). Water also plays an important physical control in the gas distribution during the injection. With excess saturation, gas will displace the pore water from preferentially larger pores, leading to heterogeneous flow paths, depriving zones of CO2 (Harrison et al., 2016). The energy requirements to compress and inject CO2 are limited by the permeability of the porous medium (Andreani et al., 2009; Power et al., 2014; Wilson et al., 2014). Therefore, idealized CO2 injection prefers high brucite abundances, pore saturations between 30% and 60%, fine grains for their reactivity and coarse grains for their permeability.  1.5 Thesis Objectives Before the injection of CO2 into mine tailings can be deployed at the mine scale, field conditions must be simulated in the lab and at the field scale. This thesis seeks to advance this technology by: 1. Analyzing porous medium conditions and their impact on injection reactivity and feasibility (Chapters 2, 3 and 4). 8  a. Parameters such as brucite abundance, gas permeability and tailings compaction were quantified to evaluate their relationship to successful injection practices (Chapters 2, 3 and 4). b. Experimental parameters such as mineralogy, grain size, moisture content, compaction, and injection rate were qualitatively investigated to examine their impact on carbon mineralization processes (Chapter 3). 2. Demonstrating and verifying the capability for CO2 injection to accelerate the mineralization of carbon across scales (Chapters 3 and 4). a. Increasing injection experiments from centimetre to metre scales (Chapter 4). b. Verifying the magnitude of sequestered carbon by a mass balance on the injected gas volume and an analysis of the initial and final total inorganic carbon (TIC) (Chapters 3 and 4).  1.6 Thesis Structure Five chapters make up this thesis. Chapter 1 describes the project\u2019s context and objectives. Chapter 2 establishes a new methodology for brucite detection and quantification using thermogravimetric analysis due to the importance of brucite for tailings reactivity. Chapter 3 determines the permeability of grain size blends varying from coarse to fine tailings. Centimetre-scale injection experiments were conducted on a selected grain size distribution chosen to maximize reactivity and permeability to evaluate the impact of various factors such as mineralogy, grain size, water content, compaction, and injection rate. Two methods, 9  a mass balance on the injected CO2 volume, and an assessment of the increase in TIC are used to quantify the abundance of carbon captured. Chapter 4 increases the scale of the pipe injection experiments to 6 meters. Pad injection experiments were tested and designed to simulate mine scale conditions. New methods to quantify the captured carbon are applied. Physical design considerations are evaluated through Standard Proctor compaction testing, providing recommendations for the grain size distribution, permeability, moisture content and degree of compaction. Chapter 5 provides the project\u2019s outcomes, conclusions, and future work recommendations. 10  Chapter 2: Brucite Identification by Thermogravimetric Analysis 2.1 Synopsis One of the most important minerals for carbon mineralization in ultramafic mine tailings is brucite. However, brucite is commonly present in only trace abundances. Detecting and accurately quantifying brucite abundances is critical for estimating carbon sequestration potential and selecting suitable materials for experimental work to characterize reactivity and accelerate carbon mineralization. Here, the quantification of brucite is investigated through thermogravimetric analysis of the thermal decomposition of brucite, which releases one mole of water per mole of brucite between 300 to 450\u00baC. Using an exponential fit for the background mass loss, brucite can be consistently detected at abundances of at least 0.3 wt.%. Relative errors for quantities below 2 wt.% are on average 50% and decrease as the abundance increases, being 4% for abundances higher than 5 wt.%. TGA has advantages over XRD with superior detection limits and the ability to quantify brucite abundances when chlorite minerals are present, which can interfere with brucite peaks in XRD. TGA and XRD can be used to support one another\u2019s strengths and overcome weaknesses to collaboratively improve geochemical investigations into the carbon sequestration potential of mine tailings.  11  2.2 Introduction Carbon mineralization is a natural process that occurs as part of the greater carbon cycle (Berner et al., 1983). Weathering of minerals releases divalent cations, which combine with aqueous carbon dioxide (CO2) to precipitate as carbonate minerals. The rate of weathering is primarily controlled by the rates of mineral dissolution, which are mineral-specific and differ significantly (Lu, 2020; Pokrovsky et al., 2005). These natural reactions could be accelerated and harnessed to sequester CO2 to mitigate climate change. For engineered carbon mineralization, particular minerals are of critical importance. Ultramafic rocks, such as serpentinites and kimberlites, host olivine and serpentine in high abundances, which react relatively rapidly with CO2 (Andreani et al., 2009; B\u00e9arat et al., 2006; Daval et al., 2011, 2013; Wilson et al., 2006). However, brucite [Mg(OH)2], a trace mineral, outpaces even these minerals by orders of magnitude (Lu, 2020; Pokrovsky et al., 2005). Therefore, the detection of brucite is critical when evaluating the potential of ultramafic rocks to rapidly sequester carbon and the rates at which they do so. Brucite is present in serpentinized rocks in abundances from 0 wt.% to as high as 15 wt.% (Pronost et al., 2011; Steinthorsdottir, 2021). X-ray diffraction (XRD) is commonly used to determine and quantify mineralogy. XRD can quantify the abundance of brucite to estimate carbon mineralization potential (Mervine et al., 2018; Turvey et al., 2018a). However, peak interference in XRD patterns can complicate the identification of brucite and its inclusion in the quantitative assessment. In kimberlites, for example, the 0 0 3 peak of chlorite minerals can overlap with brucite\u2019s 0 0 1 peak. Even in a clear pattern, detection of brucite at low abundances (<1 wt.%) is difficult, can lead to false positives, and results in high relative errors for the calculated abundance (Mervine et al., 2018; Turvey et al., 2018a). 12  An alternative to XRD for brucite identification and quantification is thermogravimetric analysis (TGA), which measures the sample mass loss as it is heated. Minerals release volatiles at characteristic temperature intervals, and this enables their detection (F\u00f6ldv\u00e1ri, 2011). Brucite releases water on a mole per mole basis between ~285 to ~450\u00b0C, via the following process in Equation 2.1. Since this mass loss occurs at a stoichiometric ratio (mass ratio of 3.24), the mass loss can be used to quantify the abundance of brucite (Assima et al., 2013b; F\u00f6ldv\u00e1ri, 2011). Equation 2.1. Mg(OH)2 \u2192 MgO + H2O(g) Several studies have measured the mass loss from brucite, and a few have even used the stoichiometric ratio to estimate brucite abundance (Assima et al., 2013b; F\u00f6ldv\u00e1ri, 2011; Liu et al., 2018; Nahdi et al., 2009; Trittschack et al., 2014; van der Merwe & Strydom, 2004; Wang et al., 1998). Assima et al. (2013b) developed a method that spiked a sample with an increasing amount of pure brucite and involved three rounds (13-hour long analyses) of TGA. An ideal analytical method would reduce the required analysis time. Further, how this method performed when other phases that lost mass at the same temperatures were present was not investigated. Chlorite and serpentine minerals are both common in ultramafic rocks and lose some mass at the same time as brucite (F\u00f6ldv\u00e1ri, 2011). Mineral standards with known brucite abundances were prepared to assess TGA\u2019s ability to aid in estimating carbon mineralization potential. The objectives of this study were to 1) determine the detection limit of brucite in TGA, 2) develop a method to quantify the abundance of brucite, and 3) provide recommendations on the use of TGA for the estimation of carbon mineralization potential and the assessment of carbon mineralization rates. 13  2.3 Methods 2.3.1 Mineral Standard Preparation Three sets of mineral standards were prepared. The first consisted of a processed serpentinite (PS) sample sourced from the Baptiste Nickel Project (BC, Canada) spiked with known quantities of a natural brucite sample from Brucite Mine (NV, USA). The second standard set consisted of a processed kimberlite (PK) sample sourced from Gahcho Ku\u00e9 Diamond Mine (NT, Canada) spiked with known quantities of the natural brucite. The third set of standards consisted of quartz sand from Lane Mountain Materials spiked with known amounts of natural brucite or synthetic brucite (95% purity Mg(OH)2) purchased from Ward\u2019s Science and Alfa Aesar. Synthetic Mg(OH)2 samples were run due to the natural brucite having impurities, as determined by quantitative X-Ray Diffraction (qXRD). Details and brucite abundances for each sample can be found in Table 2.1. The serpentinite and kimberlite samples had no detectable brucite according to TGA and XRD (detection limit of ~0.5 wt.%). Samples were weighed using a Mettler Toledo TLE104E scale to add known quantities of natural brucite or synthetic Mg(OH)2. All samples were milled using a McCrone micronizing mill under anhydrous ethanol for seven minutes to achieve a consistent grain size between samples. Samples were then dried and homogenized using an agate mortar and pestle.14  Table 2.1. Results from all standards for each quantification method with their respective errors. Sample Added Brucite (wt.%) Total Mass Loss Linear Extrapolation Exponential Interpolation Exponential + Correction Brucite (wt.%) Relative Error (%) Brucite (wt.%) Relative Error (%) Brucite (wt.%) Relative Error (%) Brucite (wt.%) Relative Error (%) PS-STD-1 0.0 1.8 - 0.4 - 0.0 - 0.0 - PS-STD-2 0.3 3.6 1100.0 1.6 433.3 0.5 66.7 0.6 100.0 PS-STD-3 1.0 4.4 340.0 2.1 110.0 1.1 10.0 1.4 40.0 PS-STD-4 3.0 5.9 96.7 3.3 10.0 2.3 -23.3 2.7 -10.0 PS-STD-5 5.0 7.7 54.0 5.4 8.0 4.4 -12.0 5.2 4.0 PK-STD-1 0.0 1.5 - 0.2 - 0.1 - 0.1 - PK-STD-2 0.3 1.7 466.7 0.3 0.0 0.2 -33.3 0.2 -33.3 PK-STD-3 0.4 2.2 450.0 0.7 75.0 0.8 100.0 0.9 125.0 PK-STD-4 0.6 2.4 300.0 1.0 66.7 0.8 33.3 0.9 50.0 PK-STD-5 0.9 2.5 177.8 1.0 11.1 0.7 -22.2 0.8 -11.1 PK-STD-6 1.3 3.2 146.2 1.3 0.0 1.5 15.4 1.8 38.5 PK-STD-7 1.7 3.2 88.2 1.6 -5.9 1.5 -11.8 1.8 5.9 PK-STD-8 2.1 3.7 76.2 1.9 -9.5 1.4 -33.3 1.6 -23.8 PK-STD-9 2.6 4.0 53.8 2.1 -19.2 2.3 -11.5 2.7 3.8 PK-STD-10 3.4 4.6 35.3 2.7 -20.6 2.6 -23.5 3.0 -11.8 PK-STD-11 4.3 5.5 27.9 3.7 -14.0 3.8 -11.6 4.4 2.3  15  Table 2.1 continued. Sample Added Brucite (wt.%) Total Mass Loss Linear Extrapolation Exponential Interpolation Exponential + Correction Brucite (wt.%) Relative Error (%) Brucite (wt.%) Relative Error (%) Brucite (wt.%) Relative Error (%) Brucite (wt.%) Relative Error (%) PK-STD-12 5.1 5.9 15.7 4.1 -19.6 3.8 -25.5 4.4 -13.7 PK-STD-13 5.1 6.2 21.6 4.2 -17.6 4.7 -7.8 5.4 5.9 PK-STD-14 6.0 7.1 18.3 5.0 -16.7 5.1 -15.0 5.8 -3.3 PK-STD-15 6.8 7.4 8.8 5.5 -19.1 6.0 -11.8 7.0 2.9 PK-STD-16 7.6 8.2 7.9 6.1 -19.7 6.4 -15.8 7.4 -2.6 PK-STD-17 8.5 8.8 3.5 6.5 -23.5 7.3 -14.1 8.4 -1.2 BRC-STD-1a 83.1 d - - - - 72.3 -13.0 85.6 3.0 BRC-STD-2a 41.6 d - - - - 36.4 -12.5 43.1 3.6 BRC-STD-3a 8.3 d - - - - 7.3 -12.0 8.7 4.8 BRC-STD-4b 100.0 - - - - 82.5 -17.5 97.7 -2.3 BRC-STD-5b 50.0 - - - - 41.2 -17.6 48.7 -2.6 BRC-STD-6b 10.0 - - - - 8.4 -16.0 10.0 0.0 BRC-STD-7c 100.0 - - - - 83.9 -16.1 99.3 -0.7 BRC-STD-8c 50.0 - - - - 43.7 -12.6 51.7 3.4 BRC-STD-9c 10.0 - - - - 8.1 -19.0 9.5 -5.0 a Natural brucite from Brucite Mine, NV, United States. b Synthetic brucite from Alfa Aesar. c Synthetic brucite from Ward\u2019s Science. d Brucite values for the natural brucite samples include the abundance of minor hydromagnesite as it contributes to the mass loss within the brucite interval. 16  2.3.2 Analytical Methods 2.3.2.1 Thermogravimetric Analysis A Perkin Elmer TGA 4000 with a Polyscience chiller and an AS 6000 autosampler performed the thermogravimetric analysis at the University of British Columbia. N2 was used as an inert carrier gas and was pumped at a rate of 19.8 mL min-1. Samples (~50 mg) were heated from 100 to 900\u00b0C at a rate of 10\u00b0C per minute. Recordings of the temperature and sample mass were taken every 0.5 seconds. Thermogravimetric (TG) curves of mass versus temperature were plotted as wt.%. Derivative thermogravimetric (DTG) curves were plotted as every 50th data point to reduce noise and smoothen out the curves. 2.3.2.2 Quantitative Thermogravimetric Analysis All methods rely on identifying the specific window during which brucite rapidly loses mass. For each sample, this can vary, with the brucite interval being selected based on the peak from the DTG curve. Methods of quantifying brucite abundance are described in order of increasing complexity. Figure 2.1 demonstrates the differences between each of the three methods. 17   Figure 2.1. Schematic of how the various methods identify the mass loss interval to calculate the dehydration of water from brucite. The mass loss interval is indicated by the double-ended arrows for the total mass loss, the linear extrapolation method and the exponential interpolation method in black, blue and red, respectively. These mass loss intervals are equivalent to the area under the curve for the derivative mass loss plot, bordered by the dashed lines at the beginning and end of the brucite decomposition temperature interval. The total mass loss is equivalent to the area between the derivative mass loss curve and the x-axis. The linear method is equivalent to the area bordered by the derivative mass loss curve and the blue extrapolation line. Lastly, the exponential method is equivalent to the area between the derivative mass loss and the red interpolation curves. 18  2.3.2.2.1 Total Mass Loss The first method involves determining the mass loss across the brucite interval and converting this to wt.% based on the stoichiometric ratio of water to brucite, which is 3.24. This abundance is calculated using the formula in Equation 2.2. \u2018m300\u2019 is the mass in wt.% at ~300\u00b0C (280 \u2013 300\u00b0C), which is roughly the start of the brucite mass loss interval, and \u2018m415\u2019 is the mass in wt.% at ~415\u00b0C (400 \u2013 475\u00b0C), which is roughly the end of the brucite mass loss interval. Equation 2.2. BruciteTotalMassLoss (wt.%) = 3.24\u00b7(m300 \u2013 m415) 2.3.2.2.2 Linear Extrapolation The linear extrapolation method builds on the total mass loss method. For the total mass loss method, the assumption is that the mass loss from minerals apart from brucite is zero during the brucite decomposition interval. If brucite were absent, no mass would be lost over the temperature interval from 300 to 415\u00b0C. However, this is unlikely as many other minerals found in serpentinites and kimberlites lose some of their mass over this same interval, including chlorite and serpentine (Assima et al., 2013b; F\u00f6ldv\u00e1ri, 2011). While the mass loss from other minerals is not large enough to prevent brucite from being quantified, it is significant and is especially so in cases with low brucite abundances. Improving the quantification of brucite requires a method to remove the mass loss contribution from these other minerals. One way to do this is to extrapolate the data trend immediately before the brucite decomposition window, typically using from ~220 to ~300\u00b0C. The average slope of this data enables the mass loss to be extrapolated. This extrapolation estimates a new mass that would have been lost if brucite was not present. Instead of attributing the total mass loss in the interval to brucite, a portion is discounted and is attributed to other mineral phases. This method leads to brucite abundance estimates being lower than found 19  for the total mass loss method. This calculation is performed in Equation 2.3. \u2018mLin\u2019 is the linearly projected mass in wt.% if there was no brucite in the sample. Equation 2.3. BruciteLin (wt.%) = 3.24\u00b7(mLin \u2013 m415) 2.3.2.2.3 Exponential Interpolation Finally, the exponential interpolation builds upon the linear method. Since the rate of mass loss from minerals depends on the initial abundance of the mineral, this makes an exponential function an ideal candidate to model the background mass loss from other minerals. Similar to the linear method, the initial derivative mass loss data before the brucite interval was used (220 \u2013 300\u00b0C), along with the derivative mass loss data immediately after the brucite interval (415 \u2013 500\u00b0C). These two data intervals were then used to fit an exponential function of the form shown in Equation 2.4. Equation 2.4. y = abx + cdx Where y is the derivative mass loss (wt.%\u00b7\u00b0C-1), x is the temperature (\u00b0C), and a, b, c, and d are constants determined by the \u2018fit\u2019 function in MATLAB R2020a (MathWorks). The exponential fit for the DTG curve provides the slope of the TG curve, allowing a corresponding TG curve to be back-calculated. This fitted function matches the slope of the provided data and enables a new quantification of mass loss attributable to brucite. This calculation is accomplished through Equation 2.5. \u2018mExp\u2019 is the exponentially interpolated mass in wt.% if there had been no brucite in the sample. Equation 2.5. BruciteExp (wt.%) = 3.24\u00b7(mExp \u2013 m415) 20  2.4 Results 2.4.1 Brucite Detection Limit Brucite can be qualitatively identified by observing the DTG curve for a peak indicating a rapid mass loss in the TG curve, between 285 and 450\u00b0C. The mass loss and derivative mass loss curves for PS standards 1, 2, and 3 are displayed in Figure 2.2. PS standards 2 and 3 had 0.3 and 1.0 wt.% brucite added, respectively. PS standard 1 has no identifiable brucite. The derivative peak for PS standard 3 is quite clear. The peak for PS standard 2 is less distinct but still discernable. 21   Figure 2.2. A. TG mass loss from PS-STD-1, PS-STD-2, and PS-STD-3. B. DTG curves of PS-STD-1, PS-STD-2, and PS-STD-3. PS-STD-1 had below detection brucite, while PS-STD-2 was spiked with 0.3 wt.% brucite, and PS-STD-3 had 1.0 wt.% brucite added. PK standards 1, 2, and 3 are presented in Figure 2.3. PK standards 2 and 3 had 0.3 and 0.4 wt.% brucite added, respectively. There is no peak for PK standard 1. While PK standard 3 is quite clear, PK standard 2 has an observable peak. Regardless of the sample type, the detection limit of brucite by TGA is, therefore, less than 0.3 wt.%. 22   Figure 2.3. A. TG mass loss from PK-STD-1, PK-STD-2, and PK-STD-3. B. DTG curves of PK-STD-1, PK-STD-2, and PK-STD-3. PK-STD-1 had below detection brucite, while PK-STD-2 was spiked with 0.3 wt.% brucite, and PK-STD-3 had 0.4 wt.% brucite added.   23  2.4.2 Brucite Abundance Measurement The results from all three methods are presented in Figure 2.4 and Table 2.1. The total mass loss method overestimates the brucite abundance for all samples, with the total errors between the known and calculated brucite abundance decreasing as the brucite abundance increases. Notably, the brucite estimates for the PS and PK mineral standards are not consistent for the same brucite abundance. While the trend of both standard sets is the same, the calculated estimates for the PS standards overestimates the abundances by twice as much as the PK standards. The relative errors are substantial for low quantities of brucite (average relative error of 380% for <2 wt.% brucite) and improve at higher brucite abundances (average relative error of 35% for >2 wt.% brucite). The linear and exponential methods produce similar results to each other. Both are close to the known brucite abundance at amounts below 2.5 wt.% and begin to underestimate at higher brucite abundances. Again, there is an inconsistency between the PS and PK samples for the linear extrapolation method, as the estimates for the PS samples are significantly higher than those for the PK standards, as occurred for the total mass loss method. However, the exponential method is consistent at the same brucite abundance regardless of the sample type. The relative errors for the exponential method are, on average less than for the linear method. At abundances below 2 wt.%, the exponential method has a relative error of 37% versus 88% for the linear method. At abundances greater than 2 wt.%, the methods are comparable with relative errors of 17 and 16% for the exponential and linear methods, respectively. 24   Figure 2.4. Calculated brucite abundances by the total mass loss, linear extrapolation, and exponential interpolation methods for the processed serpentinite and kimberlite standards.25  2.4.3 Correction Factor Of the three methods, the exponential interpolation produced the best results for the processed serpentinite and kimberlite standards. To examine the entire brucite concentration range, brucite-quartz standards were also prepared and run at abundances of 8 to 100 wt.%. The exponential method was used to calculate the brucite abundance (Figure 2.5). Again, the exponential method underestimated the brucite abundance, even for pure synthetic brucite samples. What is more, the brucite-quartz standards also follow the trend found from the serpentinite and kimberlite standards. This trend is linear but underestimates the known brucite abundances more as concentrations increase. Since there exists a strong, consistent linear trend in the data, a correction factor was determined and applied to improve the accuracy of the results. This calculation was accomplished as shown in Equation 2.6. Equation 2.6. BruciteCorrected (wt.%) = BruciteExp\u00b71.19 The correction factor of 1.19 was calibrated using the slope of a regression line that passed through the origin and fitted the results for all serpentinite, kimberlite and quartz standards. This factor was applied for all 40 mineral standards with brucite values ranging from 0 to 100 wt.%. Figure 2.5 shows the results of using the calibration factor. The effect upon the relative errors is remarkable. The relative errors increase slightly for some low values, but the errors decrease by a factor of ~2 for the rest of the samples. At abundances below 2 wt.%, the absolute average relative error is 50%. For 2 to 5 wt.%, the average relative error is 9%, and at higher abundances (>5 wt.%), it is on average 4%. 26   Figure 2.5. Calculated brucite abundance and relative error results for the exponential method with the correction factor applied to all mineral standards. 27  2.5 Discussion 2.5.1 Literature Methods and Results Other studies have attempted to quantify brucite abundances using TGA, including via the total mass loss method (Assima et al., 2013b; F\u00f6ldv\u00e1ri, 2011). Measurements of the mass loss from pure brucite have varied from 81 to 115% of the expected stoichiometric mass loss (Assima et al., 2013b; Liu et al., 2018; Nahdi et al., 2009; Trittschack et al., 2014; van der Merwe & Strydom, 2004; Wang et al., 1998). Overestimates have been due to identified impurities contributing more volatiles than expected (Assima et al., 2013b; Nahdi et al., 2009). Underestimates were attributed to brucite not completely decomposing over the characteristic peak temperature interval, but rather continuing to lose mass up until 500 or even 600\u00baC (Liu et al., 2018; Nahdi et al., 2009; Wang et al., 1998). This mass loss at higher temperatures is also observed in the pure synthetic brucite samples run in this study (Appendix 1). The signature peak ends around 450\u00baC, but the mass loss continues, and the derivative does not decline to zero. One estimate which found 96% of the expected mass loss used a temperature interval of 200 to 600\u00baC (Trittschack et al., 2014). However, this poses a problem in mineralogically diverse samples, as this becomes a large temperature interval to require no mass loss contributions from other minerals. For example, in the serpentinite and kimberlite standards, chlorite and serpentine lose significant mass above 500\u00baC. Using a narrower temperature interval avoids the mass loss from chlorite and serpentine minerals but prevents observing the full decomposition of brucite. Underestimates of brucite abundances have been 81% (Wang et al., 1998) and 87% (Liu et al., 2018; van der Merwe & Strydom, 2004), of which the 84% found here is in good company (Figure 2.5). Therefore, the applied correction factor is not arbitrary but in line with the literature decomposition behaviour of brucite within its characteristic temperature interval. Even in using the total mass loss method, Assima et al. (2013b) 28  applied a correction factor of 1.086, as it was recognized that not all of the brucite mass loss occurred inside the characteristic window. The percentage of brucite mass loss beyond 450\u00baC is likely within error of the exponential method at low brucite abundances. As brucite abundances increase, this proportion of mass loss becomes more significant, leading to underestimates. Assima et al. (2013b) extensively investigated using TGA to quantify brucite abundances using a method of spiking an individual sample of inert forsterite several times with increasing quantities of brucite. This enabled a linear regression of the relationship between the mass loss within the brucite decomposition interval and the brucite abundance to determine a factor to relate the mass loss to the brucite abundance. This method achieved excellent results with relative errors of 10% for abundances less than 1 wt.% and relative errors below 2% for abundances from 1 to 5 wt.%, significantly more accurate than the results presented in this study. However, there are several limitations to this method and its application. First, the quantitative results and the relative errors were attained on samples with no other minerals present that would lose mass over the same temperature interval. Second, while the method was tested on brucite-bearing ultramafic mine tailings samples, it was not applied to samples where the brucite content was known. Third, the method of spiking a sample several times, and conducting multiple 13-hour long TGA analyses (compared to the single 2-hour analyses used here), means that the method here is considerably less time-intensive. This rapid turnaround time is an important consideration considering the scale and heterogeneity intrinsic in assessing mineral deposits for their carbon mineralization potential, as dozens or hundreds of analyses may be required (Vanderzee et al., 2019).  29  2.5.2 Comparison with qXRD qXRD and TGA can both quantify the abundance of brucite. From the results of this study, the detection limits of TGA for brucite (~0.3 wt.%) are lower and appear to be more consistent than those for qXRD (~0.5 wt.%). This lower detection limit is significant as it allows brucite at low abundances to be detected and quantified. It also enables an explicit assessment of when brucite abundance in a sample is below detection, which is more difficult in qXRD even at higher abundances and may lead to false positives. TGA allows the challenges related to peak interference in qXRD to be overcome. While minerals such as chlorite and serpentine begin to lose mass over the same interval, the magnitude is small and was discounted by the methods used here. A comparison of the relative errors with those of qXRD is possible from the results of Turvey et al. (2018), who investigated the accuracy of different structure-less fitting methods on mineral standards with brucite below 2.2 wt.%. Notably, the study was conducted on serpentinite standards that were devoid of chlorite minerals. The best-fitting method produced relative errors of on average ~20%, while the other fitting methods yielded relative errors from 25 to 100%. All in all, this is similar to the results attained here with the applied correction factor. This comparison suggests that under ideal conditions, both methods can achieve the same level of accuracy. However, ideal conditions are less likely for qXRD, as chlorite minerals are common alteration products of serpentine. Further, quantifying brucite by qXRD is challenging, and the errors tend to be inconsistent, with brucite being either over or underestimated depending on the sample, creating more random errors (Turvey et al., 2018a). Compared to qXRD, the consistent errors in TGA can be partially overcome by using a correction factor.  30  2.5.3 TGA Limitations TGA is limited in its ability to identify and distinguish mass loss peaks from mineral phases with similar thermal decomposition behaviour. This limitation also affects the ability to quantify brucite as multiple minerals can lose mass in the same temperature interval. While contributions to the mass loss from chlorite and serpentine minerals have been managed here, this was only possible because those minerals lose a very small percentage of their mass in the brucite temperature interval and mostly lose their mass outside this interval. However, minerals that directly lose mass in the same interval will make it very difficult to quantify brucite abundances. It may still be possible to identify brucite, as the peaks will not perfectly overlap, but a method would be needed to differentiate between the mass loss contribution when interference occurs. Prevalent minerals in ultramafic settings that interfere with brucite thermal decomposition include Mg-carbonates (i.e. nesquehonite, dypingite, hydromagnesite) and hydrotalcites (i.e. pyroaurite) (F\u00f6ldv\u00e1ri, 2011; Frost & Erickson, 2004; Jauffret et al., 2015; Lin et al., 2018; Spratt et al., 2008). These minerals are also of importance for carbon mineralization as they host sequestered carbon. However, quantifying their abundance would be inhibited by remnant brucite. Dehydration peaks from these minerals, which occur at lower temperatures (<300\u00b0C) (F\u00f6ldv\u00e1ri, 2011; Frost & Erickson, 2004; Jauffret et al., 2015; Lin et al., 2018; Spratt et al., 2008) may facilitate their qualitative identification in the DTG curve. TGA systems with either Fourier transform infrared spectroscopy (FTIR) or a coupled evolved gas mass spectrometer (MS) would enable contributions from each mineral to be differentiated as they can determine the proportion and type of gas being released. Alternatively, an approach is needed to determine the ratio of brucite to the complicating mineral, allowing the stoichiometric factor for the mass loss to be altered and the total mineral abundance determined. qXRD could be used if both mineral abundances were high enough. 31  2.6 Implications Thermogravimetric analysis has been shown to accurately and repeatedly identify and quantify the presence of brucite in ultramafic mine wastes at typical abundances of occurrence. TGA has improved the ability to detect brucite at low abundances and reduces the potential for false-positive identification. Further, TGA enables brucite identification and quantification when interfering minerals are present in XRD. The relative errors from the methods used here are similar to qXRD under ideal conditions and produce results favourable compared to literature. Finally, methods have been suggested if brucite and an interfering mineral are both present in TGA. Brucite is a critical mineral for carbon sequestration. Therefore, this new development enhances the ability to target suitable materials for carbon sequestration potential estimations and experimental work to characterize ultramafic samples and enhance their reactivity. To detect brucite presence, TGA is more suitable than XRD. However, this identification is aided by XRD\u2019s characterization of the bulk mineralogy to ensure minerals with similar thermal decomposition behaviour are not misidentified as brucite. If brucite is absent, this means carbon sequestration can be positively attributed to other minerals, such as serpentine or olivine. In chlorite-bearing samples, brucite should be quantified by TGA and not by qXRD. If Mg-carbonates or hydrotalcites are identified in XRD quantitative TGA is not reliable. Either XRD alone or XRD combined with TGA must be used under these conditions unless more powerful analytical tools are available (i.e. FTIR, MS). The methods used here enable TGA to rapidly and accurately identify and quantify brucite. While TGA is currently specialized to brucite, it may be possible for the methods used here to be applied to other mineral phases. Ultimately, TGA is a valuable tool in the geochemist\u2019s toolbox, ensuring that suitable mines can take advantage of their mineralogy and reduce their carbon footprint. 32  Chapter 3: Pneumatic Permeability and Centimetre-Scale Reactivity 3.1 Synopsis Processed ultramafic mine wastes passively sequester carbon dioxide (CO2) from air into magnesium carbonate minerals. These passive reactions are rate-limited by the CO2 supply and have a limited impact on the carbon footprint of the respective mines. The rate of carbon mineralization can be enhanced by increasing the supply rate of CO2 via the direct injection of elevated pCO2 gases, such as diesel flue gas from mine power generation. This chapter assesses the feasibility and controls on the injection of CO2 into processed nickel and diamond mine wastes. Simulated or diesel flue gas was injected into unsaturated, compacted, well-graded grain size distributions of tailings. The coarse-grained fraction of the distribution enabled sufficient permeability to inject, with permeability maintained after carbon mineralization. The fine-grained fraction promoted significant Mg dissolution from brucite and lesser dissolution from lizardite, leading to the precipitation of nesquehonite and hydromagnesite. Gas-phase CO2 concentrations were used to assess reaction rates, while the total amount of reaction was determined by a mass balance on the injected CO2 and the increase in total inorganic carbon. Overall, short-term injection demonstrated enhanced sequestration rates that are an order of magnitude faster than observed passive rates compared between similar deposit types. Carbon mineralization rates from injection are sufficient to achieve the magnitude of sequestration needed to make carbon-efficient mines carbon neutral.  33  3.2 Introduction Carbon mineralization has been suggested as a possible carbon dioxide (CO2) removal strategy to mitigate anthropogenic climate change as it possesses significant storage capacity (Gt-scale) and long-term stability (>1000 years) (Lackner et al., 1995, 1997, 2003; Seifritz, 1990). Ultramafic mine wastes can contain the appropriate hydroxide and silicate minerals to readily and rapidly provide Mg needed to mineralize CO2 into carbonate minerals under ambient conditions (Assima et al., 2012; Assima et al., 2013a; Beinlich & Austrheim, 2012; Oskierski et al., 2013b; Pronost et al., 2011, 2012; Wilson et al., 2006, 2009a, 2009b, 2014). In the case of passive carbon mineralization at active mine sites, the carbon sequestration rate is limited by the supply rate of CO2 due to both the slow diffusion of CO2 into the tailings and the slow kinetics of CO2 hydration (Assima et al., 2013a; Harrison et al., 2013a, 2015; Wilson et al., 2010). Increasing the supply rate of CO2 causes mineral dissolution to become rate-limiting (Harrison et al., 2016), with Mg dissolution rates being mineral and grain-size dependent (Assima et al., 2013a; Harrison et al., 2015, 2016; Lu, 2020; Oskierski et al., 2013b; Wilson et al., 2006, 2009a; Zarandi et al., 2016). Mg that can be readily leached under ambient conditions is deemed \u2018labile\u2019 and represents a fraction of the total Mg in the sample (Lu, 2020). Mineral dissolution is typically limited in ambient conditions by the passivation of mineral surfaces, either by Si or Fe layers or precipitating phases (Andreani et al., 2009; Assima et al., 2012, 2013a, 2014d, 2014a, 2014c; B\u00e9arat et al., 2006; Daval et al., 2011; Harrison et al., 2015, 2016; H\u00f6velmann et al., 2012; Huijgen et al., 2006; Johnson et al., 2014; Park & Fan, 2004; Sissmann et al., 2014; Zarandi et al., 2016, 2017b) and by the moisture content of the mine waste (Assima et al., 2012, 2013a, 2014b, 2014d; Chakravarthy et al., 2020; Harrison et al., 2015, 2016; Pronost et al., 2011). 34  Injection of CO2 rich gases into processed mine waste is one proposed strategy to capture and store CO2 by enhancing carbon mineralization, as it increases both the mass transport and the hydration kinetics of CO2 (Hamilton et al., 2021; Harrison et al., 2013b; Li et al., 2018; Mervine et al., 2018; Nowamooz et al., 2018; Power et al., 2020; Pronost et al., 2012; Wilson et al., 2014). This strategy necessitates reasonable injection rates and permeability conditions to minimize energy requirements. Injection rates could be minimized by capturing and using concentrated sources of CO2, such as flue gases from local mine point sources or low purity streams from direct air capture (DAC) systems (Kelemen et al., 2020). Local point sources eliminate the need for CO2 transportation. Under ambient conditions, passive carbon mineralization is limited by the permeability of the mine wastes (Assima et al., 2014b, 2014c; Gras et al., 2017, 2020; Hamilton et al., 2021; Kandji et al., 2015), with some form of co-disposal of coarse and fine tailings being suggested to increase diffusion rates (Assima et al., 2014b; Hamilton et al., 2021; Zarandi et al., 2017a). The tailings grain size distribution is an important control on the reactivity (Mg2+ supply) and permeability (CO2 supply)(Assima et al., 2014b; Hamilton et al., 2021; Zarandi et al., 2017a). While finer grain sizes promote rapid dissolution and reaction due to their higher surface area to mass ratio, coarser grain sizes react slower (Assima et al., 2013a; Harrison et al., 2015) but promote higher permeabilities. The objectives of this study were to, 1) map the permeability of nickel (Baptiste Nickel Project, BC, Canada), diamond (Gahcho Ku\u00e9 Diamond Mine, NT, Canada) and asbestos (Lowell Asbestos Mine, VT, USA) tailings grain size distributions, and 2) assess the rate, magnitude, and controls on the accelerated reactivity of nickel and diamond processed mine wastes to CO2. Permeability mapping was accomplished by mixing various coarse and fine tailings fractions at a minimum and maximum moisture content and compacting the mix inside a horizontal cylinder. 35  Air was injected, and the measured differential pressure was used to determine the pneumatic permeability. Using the same experimental set-up, a grain size distribution was selected, compacted and injected with a CO2-rich gas stream to examine the reactivity rate and magnitude. Variables including the mineralogy, grain size, moisture content, degree of compaction and injection flow rate were investigated to examine the controls on reactivity.  36  3.3 Methods 3.3.1 Unreacted Sample Characterization Nine tailings samples were used in this study (Table 3.1). Unreacted tailings samples were characterized for their initial total inorganic carbon (TIC) by coulometry, mineralogy by quantitative X-Ray Diffraction (qXRD), thermogravimetric analysis (TGA) and QEMSCAN automated mineralogy (samples CPK-1 and CPK-2), whole-rock chemistry by X-Ray Fluorescence or ICP-AES, particle size distribution by sieving or laser diffraction, surface area by multipoint BET with N2 adsorption and density by pycnometry or water displacement. The CPK-Perm and the PSA samples were only used in the permeability experiments and were not chemically characterized. Detailed characterization methods and results may be found in Appendix 2. Table 3.1. List of tailings samples. Sample Type Locality Sample Date CPK-Perm Tailings Gahcho Ku\u00e9 Diamond Mine July 2017 CPK-1 and FPK-1 Tailings Gahcho Ku\u00e9 Diamond Mine August 2019 CPK-2 and FPK-2 Tailings Gahcho Ku\u00e9 Diamond Mine March 2020 FPS Drill core pulp Baptiste Nickel Project February 2017 CPS-1 and CPS-2 Crushed drill core Baptiste Nickel Project February 2019 PSA Tailings Lowell Asbestos Mine July 2020   3.3.2 Experimental Set-Up 3.3.2.1 Permeability and CO2 Injection Grain size distributions of nickel and diamond mine wastes were produced by mixing coarse (<6000 \u00b5m) and fine (<425 \u00b5m) processed kimberlite (CPK, FPK) and processed 37  serpentinite (CPS, FPS). Asbestos-bearing processed serpentinite (PSA) samples were already well-graded and did not require mixing. Deionized water was mixed into the grain size distribution to reach a moisture content specific to the tailings type and grain size distribution. Samples were compacted into ~5 cm diameter pipes, with samples varying in length from 12 to 20 cm. Pneumatic permeability was measured on an array of grain size distributions from pure coarse (CPK-Perm) to pure fines (FPS) at both 30% and 60% pore saturation. Permeability and reactivity were assessed using grain size distributions of 100 wt.% coarse and at a ratio of 25 wt.% fine to 75 wt.% coarse. Samples were injected with air to determine their pneumatic permeability and simulated flue gas (SFG; 90% N2, 10 vol.% CO2) or diesel flue gas (DFG; 7.6 vol.% CO2) to determine their reactivity (Figure 3.1). SFG was bubbled through a hydration flask to simulate humidified DFG and reduce evaporative moisture loss from the tailings. DFG was captured from the exhaust of a DuroStar DS7000Q diesel generator and stored in Tedlar gas sampling bags. SFG flow was controlled by either a Matheson FM-1050 series rotameter flowmeter, a Masterflex variable area flowmeter (model RK-03227-06), or a Masterflex L\/S precision standard pump with an Easy-Load II pump head. DFG flow was controlled by the Masterflex pump. Bosch Sensortec BMP280 or BMP388 pressure sensors and Vaisala GMP251 CO2 sensors were installed on either end of the compacted tailings sample. Experiments were terminated based on the degree of reaction progression as assessed by effluent CO2 concentrations and, therefore, varied among experiments. 38   Figure 3.1. Pneumatic permeability and CO2 injection experimental set-up. Permeability test conditions may be found in Appendix 2. Reactivity test conditions for all experiments are presented in Table 3.2. Detailed and representative results will be presented for the experiments with the shaded rows. Detailed results for the remaining experiments may be found in Appendix 2.  VAISALA| GMP251VAISALA| GMP251BMP388T+PBMP388T+PEASY LOAD IIHydration FlaskPeristaltic PumpCO2 SensorCompacted TailingsPressure SensorSFG: 10% CO2DFG: 7.6% CO21 \u2013 27 mL min-112 \u2013 19 cm5.1 cm39  Table 3.2. CO2 injection experimental conditions. Experiment Grain Size Blend Fines, <425 \u00b5m (wt.%) Gas Flux (cm\u00b7min-1) Length (cm) Dry Mass (kg) Bulk Density (g\u00b7cm-3) Moisture Content a (wt.%) Duration (hours) PS-E1 CPS-1:FPSb 25 0.97 14.7 0.674 1.7 4.6 191 PS-E2 CPS-1:FPSb 25 0.43 12.2 0.673 2.1 4.7 285 PS-E3 CPS-1:FPSb 25 0.25 11.9 1.925 2.1 4.6 484 PS-E4 CPS-1:FPSb 25 0.97 34.3 1.811 2.0 4.6 266 PS-E5 CPS-2:FPSb 25 0.97 12.2 0.674 2.1 4.6 189 PS-E6 CPS-2:FPSb 25 0.43 12.5 0.674 2.0 4.6 177 PS-E7 CPS-2:FPSb 25 0.25 12 1.926 2.1 4.6 428 PS-E8 CPS-2:FPSb 25 0.26 19.1 0.533 1.4 4.9 241 PS-E9 CPS-2:FPSb 25 0.27 16.5 0.509 1.6 8.1 241 PS-E10 CPS-2 0 0.44 12.4 0.598 1.8 2.9 1299 PK-E1 CPK-1:FPK-1b 42 0.34 12.4 0.674 2.1 6.5 1085 PK-E2 c CPK-1:FPK-1b 42 0.046 13.8 0.539 1.9 6.2 188 PK-E3 CPK-2:FPK-2b 39 0.047 12.1 0.517 2.2 6.2 685 PK-E4 CPK-2 18 0.042 14.5 0.589 2.0 3.4 677 a Moisture content measured as the mass of water over the total sample mass. b PS and PK coarse: fine blends were prepared at a 3:1 mass ratio. c Experiment PK-E2 was injected with diesel flue gas with 7.6 vol.% CO2. All other experiments used SFG with 10 vol.% CO2. 40  3.3.2.2 Batch Dissolution The dissolution properties of the FPS, FPK-1 and FPK-2 samples were examined in detail using the method of Lu (2020). Coarse grain size fractions were not evaluated by this methodology as they are less reactive due to their lower surface area to mass ratio. Samples were prepared as three slurries with continuous equilibration of the fluid phase with 10% CO2. Flasks were placed on a shaker table, and pH and solution chemistry were monitored periodically throughout the 55 (FPS) to 100 (FPK-1 and FPK-2) hours experimental duration. Solution samples were analyzed for the total dissolved Al, Ca, Fe, K, Mg, Na, and Si by a Varian 725-ES inductively coupled plasma-optical emission spectrometer (ICP-OES) at the University of British Columbia.  3.3.3 Analysis of Experimental Results During and after the permeability and injection experiments, a variety of data sets were collected and analyzed. A summary of the analytical methods is presented here, and detailed methods may be found in Appendix 2. 3.3.3.1 Permeability During air injection, pressure sensors were used to measure the differential pressure required to calculate intrinsic permeability following a methodology modified from ASTM D4525 (ASTM International, 2013). Pressure readings were recorded every 2 seconds, for 25 to 50 seconds, to ensure pressure stabilization. Permeabilities were measured at three flow rates, with the average taken as the final value.41  3.3.3.2 Sequestered CO2 Mass 3.3.3.2.1 Gas Mass Balance Vaisala GMP251 CO2 sensors were used to estimate the rate and magnitude of CO2 sequestration due to mineral trapping alone by performing a mass balance on the amount of injected CO2. CO2 concentrations were measured in intervals varying from 5 to 15 minutes, depending on the experiment. CO2 sequestration due to solubility trapping was discounted as its stability in mine tailings is uncertain, as alkalinity can fluctuate with environmental conditions. Conservative dissolved inorganic carbon (DIC) concentrations of 0.2 M for the PS and 0.1 M for the PK were assumed to quantify the solubility trapped abundance of CO2. These DIC concentrations are equivalent to a pH of 9.1 and 8.8, respectively. The amount of carbon in the aqueous phase was calculated from the final measured moisture content and was subtracted from the total sequestered mass of CO2. 3.3.3.2.2 TIC Increase TIC was measured on the carbonated samples after experimental completion. The initial TIC content was then subtracted to determine the TIC increase from mineralized carbon. The mineralogy of carbonated samples was also qualitatively investigated by XRD, TGA, and secondary electron microscopy (SEM) to determine the form of the Mg-carbonate minerals. 3.3.3.2.3 Excluded Methods While XRD and TGA were used to assess the mineralogy of carbonate minerals, they were not used to quantify the amount of carbon captured. qXRD can only account for crystalline minerals, meaning that amorphous carbonates cannot be detected, with amorphous carbonates being a common carbon mineralization product (Hamilton et al., 2021; Harrison et al., 2015, 2016; Kandji et al., 2017; Rausis et al., 2020; Turvey et al., 2018a; Zarandi et al., 2017b). Further, qXRD 42  is known to underestimate the abundances of carbonate and hydrotalcite minerals such as hydromagnesite and pyroaurite (Turvey et al., 2017, 2018a). qXRD estimates of the mass of sequestered carbon have been found to underestimate by a factor of two to four in comparison to elemental carbon approaches, which account for inorganic carbon regardless of the mineral structure (Hamilton et al., 2021; Turvey et al., 2018b). While TGA can be used to quantify the abundance of a single mineral phase, as has been done for brucite (Chapter 2), many Mg-carbonates, including hydromagnesite, nesquehonite, dypingite, artinite, and pyroaurite, lose mass over the same interval as brucite (Canterford et al., 1984; Davies & Bubela, 1973; F\u00f6ldv\u00e1ri, 2011; Frost et al., 2008; Frost & Erickson, 2004; Hales et al., 2008). Therefore, while TGA can identify the mineral form from the mass loss pattern, it is difficult to determine the proportion of mass loss attributable to remnant brucite versus the mass loss attributable to the Mg-carbonate or hydrotalcite. 3.3.3.3 Water Mass Balance A mass balance was performed upon the amount of pore water by comparing the initial and final moisture contents. The initial moisture content was known as a specific mass of water was added to the dry sample mass. The final moisture content was determined by weighing the final sample before and after drying under ambient conditions. Water lost to evaporation was accounted for by assuming a differential relative humidity of 10% between the injected and effluent gas. This assumption was thought to be a reasonable estimate as the injected gas was humidified before entry into the experimental apparatus, and the validity of this assumption has been confirmed by experimental results presented in Chapter 4. From the known injected gas volume, the carrying capacity for water vapour in the injected gas and the effluent was determined, with the difference being the amount of water lost to evaporation. This mass was then subtracted from the difference 43  between the initial and final moisture contents to determine the mass of pore water lost to hydrated Mg-carbonate precipitation. This calculation is presented clearly in Equation 3.1. Equation 3.1. !\"##!!\"#$%&'(')*)'+, = %!\"##!!\"-,')'*. \u2212!\"##!!\"\/',*.' \u2212!\"##!!\"01*(+$*)'+,  44  3.4 Results 3.4.1 Characterization Results For the PS and PK samples, the characterization results for the whole rock chemistry (>1.0 wt.%), TIC, surface area, grain size, density and mineralogy are presented in Tables 3.3, 3.4, and 3.5.45  Table 3.3. Mean whole-rock chemistry and standard deviations as determined by XRF or ICP-AES. Abundance (wt.%) FPS a CPS-1 b CPS-2 b FPK-1 b CPK-1 b FPK-2 b CPK-2 b SiO2  34.3 \u00b1 0.1 39.6 \u00b1 0.2 40.2 \u00b1 0.8 47.2 \u00b1 0.5 48.4 \u00b1 0.4 50.1 \u00b1 0.2 45.3 \u00b1 0.4 Al2O3 0.2 \u00b1 0.0 1.2 \u00b1 0.0 1.3 \u00b1 0.0 6.8 \u00b1 0.1 7.0 \u00b1 0.1 7.3 \u00b1 0.0 5.8 \u00b1 0.1 Fe2O3  9.6 \u00b1 0.0 8.3 \u00b1 0.2 8.7 \u00b1 0.2 6.2 \u00b1 0.1 5.9 \u00b1 0.1 6.1 \u00b1 0.1 6.7 \u00b1 0.1 CaO 0.0 \u00b1 0.0 1.0 \u00b1 0.1 0.7 \u00b1 0.0 2.5 \u00b1 0.0 2.4 \u00b1 0.1 3.3 \u00b1 0.0 5.0 \u00b1 0.2 MgO 41.9 \u00b1 0.0 38.1 \u00b1 0.3 38.2 \u00b1 1.1 22.9 \u00b1 0.4 22.6 \u00b1 0.4 20.3 \u00b1 0.1 24.0 \u00b1 0.4 Na2O 0.1 \u00b1 0.0 <0.01 <0.01 0.7 \u00b1 0.0 0.8 \u00b1 0.1 1.1 \u00b1 0.0 1.1 \u00b1 0.0 K2O <0.01 <0.01 <0.01 2.0 \u00b1 0.1 2.6 \u00b1 0.1 2.9 \u00b1 0.0 2.3 \u00b1 0.0 LOI 12.9 \u00b1 0.1 12.7 \u00b1 0.0 12.1 \u00b1 0.3 10.6 \u00b1 0.1 9.8 \u00b1 0.3 8.6 \u00b1 0.1 9.1 \u00b1 0.1 a Determined by XRF. b Determined by ICP-AES. 46  Table 3.4. TIC, surface area, grain size, and density of raw materials. Sample TIC (wt.% CO2) BET N2 Adsorption (m2\u00b7g-1) Mean Particle Size (\u00b5m) Density (g\u00b7cm-3) FPS 0.18 \u00b1 0.03 4.1 109 2.70 CPS-1 0.16 \u00b1 0.04 - 1990 2.73 CPS-2 0.32 \u00b1 0.01 - 1980 2.73 FPK-1 0.75 \u00b1 0.08 39.1 42 2.67 CPK-1 0.81 \u00b1 0.13 30.8 1590 2.63 FPK-2 0.71 \u00b1 0.05 - 182 2.73 CPK-2 0.53 \u00b1 0.07 - 1740 2.75 CPK-Perm a - - 2780 2.68 a Used in the permeability experiments only.47  Table 3.5. Mean mineralogical abundance and standard deviations as determined by qXRD and TGA. Abundance (wt.%) FPS CPS-1 CPS-2 FPK-1 CPK-1 FPK-2 CPK-2 Brucite a 12.8 \u00b1 0.2 - 1.4 \u00b1 0.1 - - - - Serpentine b 80.9 \u00b1 1.1 83.8 \u00b1 0.8 84.5 \u00b1 1.2 37.7 \u00b1 9.4 30.0 \u00b1 2.0 d 27.2 \u00b1 1.5 28.1 \u00b1 1.3 d Forsterite 3.8 \u00b1 0.7 7.1 \u00b1 0.5 5.9 \u00b1 0.5 1.0 \u00b1 0.1 1.5 \u00b1 1.1 d 6.2 \u00b1 0.5 7.8 \u00b1 1.0 d Smectite - - - 14.2 \u00b1 5.5 17.9 \u00b1 4.5 d 10.9 \u00b1 3.1 11.0 \u00b1 1.5 d Magnetite 1.8 \u00b1 1.1 7.0 \u00b1 0.3 6.3 \u00b1 0.7 1.2 \u00b1 0.1 1.1 \u00b1 0.2 d 1.2 \u00b1 0.1 1.4 \u00b1 0.3 d Diopside 0.7 \u00b1 0.6 2.2 \u00b1 0.4 1.9 \u00b1 0.1 7.4 \u00b1 7.0 2.9 \u00b1 0.1 d 3.7 \u00b1 0.9 6.2 \u00b1 1.1 d Phlogopite - - - 10.0 \u00b1 3.5 11.8 \u00b1 0.9 d 11.8 \u00b1 1.2 17.7 \u00b1 1.0 d Clinochlore - - - 6.9 \u00b1 2.7 9.7 \u00b1 2.3 d 7.9 \u00b1 1.3 5.8 \u00b1 0.5 d Calcite - - - 0.5 \u00b1 0.5 0.9 \u00b1 0.1 d 0.9 \u00b1 0.2 0.8 \u00b1 0.2 d Quartz - - - 2.9 \u00b1 1.1 3.8 \u00b1 1.0 d 5.1 \u00b1 0.7 2.1 \u00b1 0.5 d Albite - - - 4.2 \u00b1 1.6 6.0 \u00b1 0.5 d 8.5 \u00b1 1.7 5.7 \u00b1 1.9 d K-Feldspar - - - 6.6 \u00b1 2.0 9.2 \u00b1 1.5 d 12.8 \u00b1 1.3 7.1 \u00b1 0.8 d Talc - - - 6.6 \u00b1 1.4 5.4 \u00b1 0.3 d 3.0 \u00b1 0.1 3.9 \u00b1 0.2 d Dolomite - - - 0.1 \u00b1 0.1 tr. cd 0.8 \u00b1 0.1 0.7 \u00b1 0.2 d Cuspidine - - - 0.6 \u00b1 0.1 tr. cd - 0.1 \u00b1 0.4 d Andradite - - - tr. c tr. cd tr. c 1.1 \u00b1 1.3 d 48  Table 3.5 continued. Abundance (wt.%) FPS CPS-1 CPS-2 FPK-1 CPK-1 FPK-2 CPK-2 Wollastonite - - - - tr. cd tr. c 0.4 \u00b1 0.8 d Grossular - - - - - - tr. Tremolite - - - - - - tr. a Brucite detected and quantified by TGA with a detection limit of 0.3 wt.%. All other minerals were detected and quantified by qXRD. b Dominant serpentine type identified by TGA. PS: Lizardite and antigorite. PK: Lizardite. c Trace abundance. Peak identified in the XRD pattern but not included in the quantitative analysis or identified by QEMSCAN at an abundance < 1wt.%. d Confirmed by QEMSCAN. Smectite mineral identified as saponite. Other trace minerals include muscovite, biotite, illite, chromite, epidote, Fe-sulphides, melilite, Ti-oxides, apatite, zircon, and barite.  49  3.4.2 Permeability In Figure 3.2, the average permeabilities (Figure 3.2A) and porosities (Figure 3.2B) of 19 and 11 different coarse and fine tailings blends at pore saturations of 30% and 60% are plotted against their percentage of fines. From 0 to 12.5 wt.% fines, there is little change in permeability for the distribution at 30% pore saturation. After ~15 wt.% fines, the permeability of the blend starts to decrease significantly, from 3 \u00d7 10-10 m2 at 20 wt.% fines to 3 \u00d7 10-13 m2 at 100 wt.% fines. The gas porosities first decrease and then increase as the abundance of fines increases. Under the same compaction energy and pore saturation, the required moisture content varies as water influences the achieved degree of compaction. Gas porosity reaches a minimum value of 21% at 30 wt.% fines. The permeability trend for the grain size distribution at 60% pore saturation is more complex than the trend at 30% pore saturation. Below 7.5 wt.% fines, the pore water was free-draining, and the permeability could not be representatively measured. The permeability increases from 7.5 to 15 wt.% before sharply decreasing and tailing off from 20 to 70 wt.% fines. The permeability remains in a narrow range between 10-10 to 10-12 m2. Gas porosity follows the same trend as before but is generally 10% lower. The minimum porosity was 9% at 30 wt.% fines. 50   Figure 3.2. Permeability and porosity of tailings grain size distributions ranging from coarse to fine, at 30% (blue) and 60% (red) pore saturation. Error bars indicate the standard deviation from triplicate tests and are smaller than the symbol where not indicated. Dashed lines indicate the general trends.   51  Asbestos tailings permeabilities at ~30% pore saturation ranged from 5 \u00d7 10-12 m2 to 7 \u00d7 10-11 m2 (Table 3.6). Two samples were sieved below 500 \u00b5m and found the fines content varied from 24 (PSA-1) to 32 wt.% (PSA-4), with the other samples likely being similar. The sample with more fines content had a lower permeability by an order of magnitude. Table 3.6. Permeability of asbestos-bearing processed serpentinite. Sample PSA-1 PSA-2 PSA-3 PSA-4 Permeability (m2) 6.7 \u00d7 10-11 4.8 \u00d7 10-12 2.1 \u00d7 10-11 5.8 \u00d7 10-12  Permeabilities were measured on the PS and PK reactivity experiments before and after CO2 injection (Table 3.7). Experiment PS-E10, which had no fines but was well compacted, had a measured permeability of 8 \u00d7 10-10 m2. Samples with 25 wt.% FPS that were loosely compacted resulted in permeabilities of ~5 \u00d7 10-10 m2 (PS-E8 and PS-E9). High degrees of compaction yielded permeabilities at the low end of the 10-11 m2 order of magnitude (PS-E2 \u2013 PS-E7). For the PK, which had 18 to 22 wt.% fines in the coarse phase, experiment PK-E4 was done on compacted CPK and had a permeability of 4 \u00d7 10-11 m2. Adding 25 wt.% FPK but compacting less resulted in a permeability of 1 \u00d7 10-11 m2 for experiment PK-E2. Compacting to a further degree yielded permeabilities ranging from 6 \u00d7 10-13 m2 to 3 \u00d7 10-12 m2 (PK-E1 and PK-E3). The addition of fines and increasing the degree of compaction consistently lowers the permeability for both PS and PK samples. The average relative change in permeability after the reactivity injection on seven experiments was 14 \u00b1 15%.  52  Table 3.7. Permeability measurements before and after CO2 injection reactivity. Experiment Initial Permeability (m2) Final Permeability (m2) % Change PS-E1 5.0 \u00d7 10-11 3.0 \u00d7 10-11 41 PS-E2 2.3 \u00d7 10-11 - - PS-E3 7.9 \u00d7 10-12 7.3 \u00d7 10-12 8 PS-E4 2.3 \u00d7 10-11 2.1 \u00d7 10-11 9 PS-E6 1.8 \u00d7 10-11 - - PS-E8 5.0 \u00d7 10-10 - - PS-E9 4.9 \u00d7 10-10 a - - PS-E10 8.2 \u00d7 10-10 5.7 \u00d7 10-10 30 PK-E1  6.5 \u00d7 10-13 - - PK-E2 c 1.5 \u00d7 10-11 1.5 \u00d7 10-11 0 PK-E3 3.1 \u00d7 10-12 3.4 \u00d7 10-12 -9 PK-E4 4.5 \u00d7 10-11 4.4 \u00d7 10-11 2 a Single rather than triplicate measurements.   3.4.3 Reaction Progression 3.4.3.1 Processed Serpentinite CO2 Profiles Gas-phase inlet and outlet concentrations were measured over time for the processed serpentinite injection experiments (Figure 3.3). Inlet CO2 concentrations rise rapidly before reaching steady-state concentrations, with PS-E1 reaching a steady-state ahead of PS-E2. Outlet concentrations take longer to increase. In PS-E1, the outlet concentration increases almost instantaneously, while breakthrough in PS-E2 takes nearly 100 hours. Outlet concentrations steadily increase and gradually approach the concentrations of the inlet sensor. Diurnal, periodic fluctuations may be observed in inlet and outlet concentrations for both experiments. 53   Figure 3.3. Gas-phase CO2 concentrations over time for the processed serpentinite injection experiments. 3.4.3.2 Processed Kimberlite CO2 Profiles The recorded CO2 concentrations have been similarly plotted against time for the processed kimberlite injection experiments (Figure 3.4). Inlet concentrations initially increase before starting to stabilize at 6 vol.%, 8 vol.% and 7 vol.% CO2 in experiments PK-E2, E3 and E4, respectively. Outlet concentrations mimic the injected concentrations before stabilizing and showing no indication of equilibrating with the inlet concentration. Diurnal fluctuations in the gas concentrations are visible in experiments PK-E3 and E4. 54   Figure 3.4. Gas-phase CO2 concentrations over time for the processed kimberlite injection experiments PK-E2 (A), PK-E3 (B) and PK-E4 (C). 55  3.4.3.3 Instantaneous Carbonation Rates Gas concentrations were used to calculate the carbonation rate over time by relating the mass balance on the sequestered CO2 to the time interval between concentration measurements on a sample mass basis. The carbonation rates are plotted in Figure 3.5. PS-E1 begins at a rate of nearly 400 mg CO2\/kg tailings\/hr and decreases, apart from the diurnal fluctuations, quasi-linearly over time to 10 mg CO2\/kg tailings\/hr on the semi-log plot. Diurnal variations lead the rate to fluctuate by as much as a factor of 6. PS-E9 sees the rate increase initially before reaching a maximum at around 100 hours of 85 mg CO2\/kg tailings\/hr. The rate then begins to decrease quasi-linearly but not as quickly as occurs for PS-E1. The three PK experiments all show similar trends. Rates increase and reach a maximum within the first 50 hours at 14, 7, and 4 mg CO2\/kg tailings\/hr for experiments PK-E3, E2, and E4. The rates then fall gradually and become asymptotic with little further change over time after 300 hours. Diurnal variations lead to rates fluctuating by a factor of 0.5. 56   Figure 3.5. Injection and carbonation rates over time in PS and PK injection experiments.  3.4.4 Sequestered Carbon 3.4.4.1 Quantification Characterization of the initial and the carbonated tailings using TIC assessed the CO2 captured in the solid phase for all experiments. Results are presented in Figure 3.6 and Table 3.8. The TIC increase for the PS experiments with fines ranged from 1.3 to 1.5 wt.% CO2, while the PK experiments with fines captured 0.1 to 0.2 wt.% CO2. The CPS also captured 0.2 wt.% CO2, while the CPK captured 0.05 wt.% CO2. The DFG injection experiment captured the least carbon at only 0.02 wt.%. However, this is consistent with the shorter duration and lower pCO2. 57  A mass balance on the injected and effluent CO2 can be used to estimate the mass of sequestered CO2 over time. This mass balance was accomplished by subtracting the effluent mass of CO2 from the injected CO2 mass. CO2 captured as solubility trapping was not counted, as the stability of alkalinity as a storage mechanism in mine tailings is uncertain. The gas-phase mass balance shows that the amount of captured carbon in all experiments increases quickly before tapering off with extended injection. The mass balance results generally agree with the TIC measurements, being within the confidence interval or slightly overestimating (Figure 3.6). The experiments with the lowest amount of sequestered CO2 have the highest relative errors and the greatest uncertainty, which is notable as they show less agreement with the mass balance.  Figure 3.6. Mass of sequestered CO2 over time as quantified from a gas mass balance (with a thin-line error envelope) and TIC increases in carbonated versus initial samples, with a 95% confidence interval indicated. The confidence interval for PS-E9 is smaller than the symbol.58  Table 3.8. Amount of sequestered CO2 and dominant mineral reactivity. Experiment Mass Balance (wt.% CO2) Initial TIC (wt.% CO2) Final TIC  (wt.% CO2) TIC Increase (wt.% CO2) Leached MgO (wt.%)a Most Reactive Mineral Leached Reactive Mineral Mg (wt.%)a PS-E1 1.61 \u00b1 0.42 b 0.16 \u00b1 0.03 c 1.59 \u00b1 0.17 c 1.43 \u00b1 0.22 d 4.2 \u00b1 0.6 d FPS Brucite 74.0 \u00b1 11.2 d PS-E2 1.65 \u00b1 0.31 0.16 \u00b1 0.03 1.49 \u00b1 0.23 1.33 \u00b1 0.24 3.9 \u00b1 0.7 FPS Brucite 68.7 \u00b1 12.5 PS-E3 1.84 \u00b1 0.33 0.16 \u00b1 0.03 1.60 \u00b1 0.18 1.44 \u00b1 0.23 4.2 \u00b1 0.7 FPS Brucite 74.5 \u00b1 11.9 PS-E4 1.68 \u00b1 0.35 0.16 \u00b1 0.03 1.69 \u00b1 0.19 1.53 \u00b1 0.14 4.5 \u00b1 0.4 FPS Brucite 79.4 \u00b1 7.2 PS-E5 1.67 \u00b1 0.41 0.29 \u00b1 0.01 1.69 \u00b1 0.19 1.41 \u00b1 0.20 4.1 \u00b1 0.6 FPS Brucite 72.8 \u00b1 10.4 PS-E6 1.46 \u00b1 0.20 0.29 \u00b1 0.01 1.65 \u00b1 0.10 1.36 \u00b1 0.13 4.0 \u00b1 0.4 FPS Brucite 70.6 \u00b1 6.5 PS-E7 1.54 \u00b1 0.28 0.29 \u00b1 0.01 1.71 \u00b1 0.06 1.42 \u00b1 0.08 4.2 \u00b1 0.2 FPS Brucite 73.6 \u00b1 4.0 PS-E8 1.13 \u00b1 0.15 0.26 \u00b1 0.01 1.61 \u00b1 0.15 1.35 \u00b1 0.12 3.9 \u00b1 0.4 FPS Brucite 69.8 \u00b1 6.4 PS-E9 1.45 \u00b1 0.16 0.26 \u00b1 0.01 1.65 \u00b1 0.05 1.39 \u00b1 0.04 4.1 \u00b1 0.1 FPS Brucite 72.1 \u00b1 2.3 PS-E10 -e 0.32 \u00b1 0.01 0.53 \u00b1 0.02 0.21 \u00b1 0.03 0.63 \u00b1 0.09 CPS Brucite 24.8 \u00b1 3.7 PK-E1 -e 0.81 \u00b1 0.10 1.00 \u00b1 0.10 0.19 \u00b1 0.09 0.95 \u00b1 0.46 Fine Lizardite 3.1 \u00b1 1.5 PK-E2 0.053 \u00b1 0.016 0.94 \u00b1 0.07 f 0.98 \u00b1 0.02 f 0.02 \u00b1 0.02 f 0.09 \u00b1 0.10 Fine Lizardite 0.3 \u00b1 0.3 PK-E3 0.23 \u00b1 0.08 0.59 \u00b1 0.04 0.80 \u00b1 0.06 0.21 \u00b1 0.05 1.0 \u00b1 0.3 Fine Lizardite 4.6 \u00b1 1.2 PK-E4 0.10 \u00b1 0.06 0.53 \u00b1 0.07 0.58 \u00b1 0.03 0.05 \u00b1 0.06 0.23 \u00b1 0.28 Fine Lizardite 2.5 \u00b1 3.1 a Assumes the precipitation of hydromagnesite.   b Calculated uncertainty. c One standard deviation.      d 95% confidence interval. e Not calculated as the disparity between the reactivity rate and the injected gas flux yielded significant uncertainty on the estimate. f Measured on the fraction below 425 \u00b5m (41.9 wt.% of the sample). TIC increase was attributed to the whole sample mass. 59  3.4.4.2 Precipitate Characterization In the PK experiments, the TIC increase was equal to or less than 0.2 wt.% CO2. Neither qXRD nor TGA detected any Mg-carbonate minerals, and efforts to locate them in SEM proved unsuccessful. However, characterization of the precipitates in the PS experiments was possible. qXRD identified nesquehonite in PS-E1, E2, E6 and E10, while hydromagnesite was identified from the DTG curves for every sample from each experiment (Figure 3.7 and Figure 3.8). 60   Figure 3.7. XRD pattern of a carbonated sample from injection experiment PS-E1, showing the detection of nesquehonite.   2-Theta Degrees8075706560555045403530252015105stnuoC16,00015,00014,00013,00012,00011,00010,0009,0008,0007,0006,0005,0004,0003,0002,0001,0000PS-E1 CarbonatedLN BLLFNFLMFFFDMLMDL FL L F MLLF FMBL - LizarditeN - NesquehoniteF - ForsteriteB - BruciteD - DiopsideM - Magnetite61   Figure 3.8. TG and DTG curves of the initial and carbonated samples from injection experiment PS-E9. In the initial sample, the peaks refer to the loss of adsorbed water (40 \u2013 150\u00b0C), dehydroxylation of brucite (300 \u2013 430\u00b0C) and dehydroxylation of serpentine (460 \u2013 850\u00b0C). For the carbonated sample, the mass loss ranges are due to the loss of adsorbed water (40 \u2013 150\u00b0C), dehydration of hydromagnesite (150 \u2013 250\u00b0C), dehydroxylation of brucite and hydromagnesite and decarbonation of hydromagnesite (300 \u2013 460\u00b0C), and the dehydroxylation of serpentine (460 \u2013 850\u00b0C). 62  SEM micrographs of material from experiment PS-E1 are presented in Figure 3.9. Two phases were identified: prismatic, massive, 100 \u00b5m-scale needles of nesquehonite that exhibited desiccation cracks (Pronost et al., 2011), and a hollow internal structure (Figure 3.9C and D) and pervasive lenticular rosettes of hydromagnesite at the 5 \u00b5m-scale, which were plentiful along grain boundaries (Figure 3.9A, B, E and F). Energy dispersive spectroscopy confirmed the presence of Mg, O and C, and a low Si signature. Both textures are characteristic of nesquehonite and hydromagnesite, respectively (Boschi et al., 2017; Hamilton et al., 2021; Harrison et al., 2015, 2016; Hopkinson et al., 2012; Power et al., 2013; Pronost et al., 2011; Zarandi et al., 2016). The blending method of the coarse and fines smears the fines along the coarse-grain surfaces, with carbonates having precipitated out of these masses of fines (Figure 3.9F). Both nesquehonite and hydromagnesite were observed to penetrate and cement grains together. 63   Figure 3.9. Representative scanning electron micrographs of precipitated Mg-carbonates from PS-E1. A-D. Secondary electron micrographs. A. Pervasive lenticular rosettes of hydromagnesite. B. Lenticular rosettes of hydromagnesite precipitated along the planar edge of serpentine. C. Massive prismatic needles of nesquehonite. D. Prismatic needles of nesquehonite showing desiccation cracks. E-F. Backscattered electron micrographs. E. Mg-carbonates having precipitated at mineral-pore interfaces where the water film would dissolve CO2 and Mg. F. Hydromagnesite rosettes precipitated out of brucite-rich FPS.64  3.4.4.3 Reaction Efficiency Having determined the mass of sequestered CO2 by measuring the TIC increase, this amount of CO2 can be converted to the Mg abundance needed to form the Mg carbonates (Table 3.8). This conversion is possible due to the precipitate characterization, which identified hydromagnesite (5 MgO: 4 CO2) and nesquehonite (1 MgO: 1 CO2). Hydromagnesite and its MgO: CO2 stoichiometry were dominant in abundance compared to nesquehonite, as evidenced by their prevalence in TGA and SEM. This conversion enabled three comparisons, first to the total available MgO content, second to the available cation abundance from specific reactive minerals, and third to the cation abundance leached in the batch dissolution experiments. Reactivity was compared against the total MgO content as determined by whole-rock chemistry. PS sample blends had between 38 to 39 wt.% MgO, while PK sample blends had 23 to 24 wt.% MgO. Experiments PS-E1 to PS-E9, on average, required 4.1 wt.% of the MgO content to mineralize the mass of sequestered carbon. PS-E10, on CPS, must have leached 0.6 wt.% of the total MgO content. The experiments PK-E1 and PK-E3 reacted with 1 wt.% of the MgO content, while PK-E2 and PK-E4 leached 0.1 and 0.2 wt.%, respectively. Two simplifying assumptions enabled a determination of the degree of mineral reactivity. First, only the fine fraction contributed cations and second, the most reactive mineral present dominates the cation release (brucite when present, otherwise lizardite). TGA characterized the brucite abundance at 12.8 wt.% in the FPS, and qXRD characterized the serpentine abundance at 27.2 \u2013 37.7 wt.% in the FPK and 28.1 \u2013 30.0 wt.% in the CPK fines (Table 3.5). The average captured carbon from experiments PS-E1 to E9 was 1.41 wt.% CO2, and this requires 73% of the FPS brucite to have reacted. For the average sequestered mass from PK-E1 and PK-E3 at 0.20 wt.% CO2, 3.8% of the Mg from lizardite was leached. PK-E2 captured 0.02 wt.% CO2, requiring 65  0.3% of the lizardite Mg to react. For the experiments on the coarse, 25% of the brucite in experiment PS-E10 reacted to sequester 0.21 wt.% CO2 and 2.5 % of lizardite from the CPK fines in experiment PK-E4 reacted to capture 0.05 wt.% CO2. Batch dissolution experiments on the FPS and the two FPK samples were conducted as slurries at a constant pH of 4.4 and measured the concentration of cations leached into the solution. These cation abundances can be compared with the degree of carbonation from injection experiments (Table 3.9). The first step for this comparison requires accounting for the cation contributions from carbonate minerals (Paulo et al., 2021). Carbonate mineral dissolution in solution adds to the cation total but would release CO2. The initial TIC enables an accurate assessment of the cation abundance released from carbonate minerals assuming a 1:1 cation to CO2 stoichiometry. For the FPS, Mg was the only cation released in abundance, while for the FPK, calcite was identified in XRD, and therefore Ca was assumed to be the cation released from the carbonate minerals. Having corrected the leached cation abundance, the cation abundance was converted to an equivalent amount of captured CO2 by assuming the formation of calcite and hydromagnesite. TIC increases from the injection experiments must be multiplied by a factor of 4 since the FPS and FPK only constituted 25 wt.% of the mix. For the PK experiments, the contribution from the CPK fines also needed to be removed by considering the proportion of CPK fines within the total amount of fines. Noncarbonate Ca still accounted for ~20% of the divalent cation total in the FPK samples. Injection experiments leached 72%, 67%, and 57% of the labile cation abundance determined from the batch experiments on FPS, FPK-1 and FPK-2, respectively.  66  Table 3.9. Comparison of injection and batch dissolution results. Experiment FPS FPK-1 FPK-2 Batch Mg2+ Leached (mmol\u00b7g-1) 2.25 a 0.15 0.20 Batch Ca2+ Leached (mmol\u00b7g-1) - 0.20 0.22 Initial TIC (mmol\u00b7g-1) 0.04 \u00b1 0.01 0.17 \u00b1 0.02 0.16 \u00b1 0.01 Noncarbonate Mg2+ (mmol\u00b7g-1) b 2.21 (100%) 0.15 (84%) 0.20 (79%) Noncarbonate Ca2+ (mmol\u00b7g-1) b - 0.03 (16%) 0.05 (21%) Total Divalent Cations (mmol\u00b7g-1) 2.21 0.18 0.26 Batch TIC Increase (wt.% CO2) c 7.80 0.67 0.95 Injection TIC Increase (wt.% CO2) 5.64 \u00b1 0.64 d 0.45 \u00b1 0.22 e 0.54 \u00b1 0.14 e Injection\/Batch (%) 72 \u00b1 8 67 \u00b1 33 57 \u00b1 14 a Batch dissolution results leached 1.58 mmol\u00b7g-1 Mg from a sample with 9.0 wt.% brucite. This result has been normalized to the brucite abundance used in the injection experiments (12.8 wt.%). b Initial inorganic carbon attributed to Ca-carbonates first, and Mg-carbonates second. c Assumes the precipitation of calcite and hydromagnesite. d From the average TIC increase of experiments PS-E1 \u2013 PS-E9, multiplied by a factor of 4 to attribute the reactivity to the 25 wt.% FPS in the mix. e From experiments PK-E1 and PK-E3, respectively. TIC increases were multiplied by the percentage of fines corresponding to the FPK to remove the contributions of the CPK fines and by a factor of 4 to attribute the reactivity to the 25 wt.% FPK in the mix. 67  3.4.5 Water Mass Balance A mass balance on the amount of water was performed by comparing the initial and final moisture contents and accounting for the water mass lost to evaporation (Table 3.10). The initial moisture content was known as a specific mass of water was added to the dry sample mass. The final moisture content was determined by weighing the final sample mass before and after drying under ambient conditions. Evaporation was quantified by determining the carrying capacity of the injected gas volume to remove water vapour and was found to be small relative to the moisture contents. Comparing the change in moisture contents revealed that water was consistently lost throughout the experiments. PS experiments lost between 1.3 and 2.1 wt.% water from their initial moisture content. For the PK experiments, the water loss was less, between 0.4 to 0.8 wt.% water.68  Table 3.10. Pore water loss after CO2 injection experimental completion. Experiment Initial Moisture Content a  (wt.% H2O) Final Moisture Content a  (wt.% H2O) Evaporative Moisture Loss a (wt.% H2O) Pore Water  Loss a  (wt.% H2O) PS-E1 4.9 3.0 0.07 1.8 PS-E2 5.0 2.9 0.07 2.1 PS-E3 4.8 - 0.05 - PS-E4 4.8 3.0 0.04 1.7 PS-E5 4.8 2.9 0.07 1.8 PS-E6 4.8 3.3 0.03 1.5 PS-E7 4.8 3.1 0.05 1.6 PS-E8 5.2 3.3 0.02 1.8 PS-E9 8.8 6.8 0.03 2.0 PS-E10 3.0 1.4 0.27 1.3 PK-E1 7.0 6.0 0.16 0.8 PK-E2 6.6 6.2 0.004 0.4 PK-E3 6.6 6.0 0.01 0.7 PK-E4 3.6 3.0 0.01 0.6 a Defined as mass of water over dry sample mass, enabling comparisons as the water mass changes, which would affect the total sample mass.  69  3.5 Discussion 3.5.1 Permeability Permeability is a critical constraint on the efficiency of carbon sequestration in mine tailings as it has the most significant impact upon the pressure and energy requirements to inject into porous media (Andreani et al., 2009; Power et al., 2014; Wilson et al., 2014). The impact of the grain size distribution and the degree of pore saturation upon the permeability was investigated by testing a variety of grain size distributions at the minimum (~30%) and maximum (~60) degrees of pore saturation considered suitable for CO2 injection (Assima et al., 2013a; Harrison et al., 2015, 2016), and by keeping the applied compaction energy constant. At 30% pore saturation, the addition of fines consistently lowers the permeability (Figure 3.2). The porosity captures this evolution over the grain size distribution regime. The minimum porosity is attained at around 30 wt.% fines, and while this occurs after the permeability begins to decrease at an increasing rate (Figure 3.2), the decreasing permeability coincides with the decreasing porosity. This trend occurs because fines begin occupying the pore space between the coarse grains, and as more fines are added, they expand out, isolating coarse grains from one another. As minimum porosity is approached, the gas pathways progressively shift from travelling through the large pore spaces of the coarse to going through the smaller pore spaces of the fines, increasing the permeability (Budhu, 2011). As the portion of fines is increased beyond 30 wt.%, porosity increases as permeability decreases. Porosity increases because the addition of fines expands coarse grains apart, meaning that a solid grain is replaced by a group of fines and fine pore spaces. Permeability decreases because greater and greater portions of the pore pathways are through fine pores as the fine fraction increases. 70  The trend at 60% pore saturation is more complex than the trend at 30%. In hydrophilic porous media, at low moisture contents, the water is bound to the surface of grains forming thin films, and as the moisture content increases, the water films invade the pores. At low fines content, the pore water could drain as there was insufficient grain surface area to retain all the pore water. Adding fines increases the permeability as the pore water is drawn out of the large pore throats and onto the surface of the fine grains. With more fines added to the mixture, at ~20 wt.% fines, the fines begin to obstruct the gas, forcing it to travel through their smaller pore spaces. As the portion of fines is increased beyond 20 wt.%, the permeability decreases. However, the trend remains similar to that seen in the 30% pore saturation case. That higher degrees of pore saturation would not impact permeability requires explanation. It may be that while the pore saturation was increased, gas flow may have occurred through more permeable channels resulting in preferential flow paths and higher measured permeabilities than expected (Harrison et al., 2016). The permeabilities of asbestos samples were measured to compare the produced grain size distributions against naturally well-graded mine tailings. The permeabilities varied from 10-11 to 10-12 m2, consistent with the permeabilities measured at a pore saturation of 30% with similar amounts of fines, validating the materials and methods used for the PS and PK blends. As another point of comparison, these permeabilities are slightly higher but similar to those observed in the asbestos wastes at Thetford Mines with a permeability of 10-13 m2 (Lechat et al., 2016). FPK permeabilities have been estimated at 10-12 to 10-15 m2 at Diavik Diamond Mine (Moncur & Smith, 2012), a range in which the measured value of 10-14 m2 for 100% FPK (pore saturation of 30%) falls within. Selecting a grain size distribution for CO2 injection requires compromises between reactivity (fines) and permeability (coarse) (Assima et al., 2014b, 2014c; Gras et al., 2017, 2020; 71  Hamilton et al., 2021; Kandji et al., 2015; Zarandi et al., 2017a). A gradation with 25% added fines was selected for the reactivity experiments because it was in the upper range of what is nominally \u2018semi-pervious\u2019 and is comparable to fine-grained sand (Bear, 1972; Budhu, 2011; Nield & Bejan, 2006). Of note for this gradation was the similarity in permeability regardless of whether the moisture content was 30% or 60% (Figure 3.2). In all the experiments from this study, the material was relatively well compacted. The CO2 injection reactivity experiments evaluated a significant repercussion upon permeability. Carbon mineralization has the potential to alter permeability as it reduces porosity due to the precipitation of Mg-carbonates, which are volumetrically larger than the dissolving minerals (Assima et al., 2013a; Harrison et al., 2016; Vanderzee et al., 2018). At the same time, precipitation of hydrated Mg-carbonates incorporates the liquid pore water into the solid mineral structure (Harrison et al., 2015, 2016). From testing the permeability before and after the injection experiments, it is clear that there is little impact on the permeability, consistent with Assima et al. (2013a) and Andreani et al. (2009). Vanderzee et al. (2018) found that the porosity of carbonated nickel tailings remained above 0.3, indicating significant remaining permeability. This remaining permeability is reasonable because precipitation occurs in saturated pore spaces, which already impede gas flow (Harrison et al., 2016). This result indicates that the decrease in moisture content counterbalanced the volume change of the solids. Results may differ in the field where precipitation could recharge the moisture content, reducing the permeability to a greater extent than observed here (Assima et al., 2013a).  72  3.5.2 Reactivity Controls 3.5.2.1 Mineralogy Mineralogy has been well established as a dominant control on reactivity as dictated by the mineral crystal structure and the strength of the bonds needed to be broken to release Mg (Schott et al., 2009). Mineralogy was contrasted between the PS and PK materials which differed significantly in reactive mineral abundances, such as brucite, serpentine, forsterite and smectite (Table 3.5). In batch dissolution experiments, 100% of the Mg from brucite, 4% from serpentine, and 1% from forsterite are readily leached (Lu, 2020). Mg that can be accessed rapidly has been termed labile, while the inaccessible or slowly released Mg is recalcitrant (Lu, 2020; Vanderzee et al., 2019). Not only is the magnitude of cation release mineral specific, but the dissolution rate is as well. From the reactivity rates in Figure 3.5, three stages can be identified: I) the \u2018initial stage\u2019 during which equilibrium conditions are approached in the gas and aqueous phase, rapidly labile Mg is accessed, and reaction rates remain high, preventing CO2 in the effluent, II) a \u2018bulk reactivity stage\u2019 in which a majority of the labile Mg is accessed where outlet concentrations and reactivity rates progressively increase and decrease, respectively, and III) a \u2018residual stage\u2019 of reactivity where outlet concentrations and reactivity rates stabilize. 3.5.2.1.1 Source of Mg 3.5.2.1.1.1 Brucite Within PS experiments E1-E9, the FPS possessed 12.8 wt.% brucite. This abundance represents an idealized situation in which brucite-rich tailings have been designated explicitly for carbon mineralization use. The contribution of the coarse phase in the grain size distribution was examined by comparing experiments that used CPS-1 (<0.3 wt.% brucite; PS-E1, PS-E2 and PS-73  E3) versus those that used CPS-2 (1.4 wt.% brucite; PS-E5, PS-E6 and PS-E7). The addition of coarse brucite had no significant effect on the carbonation rate or the amount of CO2 captured (Table 3.8). In the coarse and fine mixtures, the fines were observed to coat the coarse grain surfaces, which may impede coarse brucite dissolution. Serpentine in the FPS can also be expected to contribute alongside brucite. Hydromagnesite precipitating off the edge of a planar serpentine grain may be evidence of serpentine Mg dissolution from highly reactive sites (Figure 3.9B). However, serpentine reactivity is dominated by the abundance of Mg available from brucite (six times the labile Mg content) and its faster dissolution rate, which is accelerated by inorganic ligands, such as HCO3- (Lu, 2020; Pokrovsky et al., 2005). The reactivity of brucite can be observed in the gas phase data for PS-E1 and PS-E9. In the CO2 profiles of Figure 3.3, inlet concentrations increase in correspondence with the experimental injection rate and are delayed by mixing with the initial air of the apparatus headspace. Outlet concentrations are delayed by flushing the initial air, CO2 dissolution into the aqueous phase and rapid carbon mineralization of very fine brucite grains (Harrison et al., 2015, 2016). Delayed breakthrough of CO2 in the effluent is due to brucite dissolution outpacing the injection rate for PS-E9. In Figure 3.5, reaction rates from brucite within experiments PS-E1 and PS-E9 were an order of magnitude higher than in the PK experiments and compare favourably with the rates observed by Harrison et al. (2015, 2016), who tested CO2 injection into mixtures of 10 wt.% brucite and quartz sand. The initial reactivity stage was not observed in PS-E1 due to its higher flow rate and is extended in PS-E9 through 120 hours due to the lower injection flux. As brucite was consumed in the bulk reactivity stage, this led to reaction rates falling by a factor of 20 over the short course of the experiment (Figure 3.5). An end to the bulk reactivity in PS-E1 and PS-E9 was not observed. 74  With an average TIC increase of 1.41 wt.% CO2 observed across the nine PS experiments, this corresponds to an average reaction of 4.1% of the total MgO and 73% of the FPS brucite assuming the formation of hydromagnesite, as evidenced by TGA and SEM. In comparing the injection results with those from the batch dissolution experiments, a very similar result was found in that 72% of the Mg leached in the batch dissolution test was accessed in the injection experiments. This similarity can be explained by the fact that brucite is known to release all its Mg under batch dissolution experimental conditions (Lu, 2020). Other studies have ascribed the incomplete reactivity of brucite under injection conditions to surface passivation and reduced local water content, and these factors likely played a role here (Harrison et al., 2015, 2016; Zarandi et al., 2016, 2017b). Assuming similar brucite reactivity grain-size dependent limitations observed by Harrison et al. (2015), the expected reactivity of brucite would be 83%, which is only slightly higher than observed. The formation of crystalline phases such as nesquehonite passivates the surface of brucite (Harrison et al., 2015, 2016; Zarandi et al., 2016), with nesquehonite being identified in XRD and SEM. Consumption of water due to hydrated Mg-carbonate precipitation inhibits brucite carbonation due to water\u2019s role as both a reactant and a reaction medium (Harrison et al., 2015). As a reaction medium, Mg-carbonate precipitation occurs in the aqueous phase and is constrained by the boundaries of the pore water (Harrison et al., 2017). The reactive surface area in mine tailings may also be limited due to the intergrowth of brucite with other minerals compared to pure mineral sample reactivity (Hamilton et al., 2020; Zarandi et al., 2016). Further, as shown in Figure 3.9F, large agglomerates of fines may have protected interior grains from CO2 exposure and mineral dissolution.75  3.5.2.1.1.2 Serpentine While brucite is known to be present at Gahcho Ku\u00e9, it was below detection (<0.3 wt.%) in all PK samples. As such, the Mg seen in the batch dissolution results is coming from another source. Lizardite is likely the dominant Mg source as it is in much higher abundance than forsterite and is known to release Mg through transient surface reactions in batch dissolution experiments (Daval et al., 2013; Lu, 2020; Paulo et al., 2021). Daval et al. (2013) observed an increase in lizardite dissolution with higher pCO2 potentially promoted by HCO3- ligands. In diffusive carbon mineralization experiments at elevated pCO2, there is some evidence for serpentine reactivity based on the release of Si (Zarandi et al., 2016). Mg-silicate reactivity is limited from complete dissolution due to surface passivation by the formation of Si-rich layers (Andreani et al., 2009; B\u00e9arat et al., 2006; Daval et al., 2011; Johnson et al., 2014; Park & Fan, 2004; Sissmann et al., 2014; Zarandi et al., 2016). From the gas phase data, CO2 concentrations at the inlet rise quickly and are delayed due to mixing with the initial air in the headspace, while effluent CO2 concentrations follow the inlet concentrations relatively quickly due to limited reactivity (Figure 3.4). However, the effluent concentrations stabilize, indicating a consistent reactivity rate as the difference between the inlet and outlet concentrations is maintained. The initial stage of reactivity for the PK experiments was short-lived, at less than 20 hours (Figure 3.5). During the stage of bulk reactivity, the PK samples had some rapid reactivity before experiencing a rate decline by a factor of three. Bulk reactivity ended at 200 and 400 hours in PK-E4 and PK-E3, respectively. An end to the bulk reactivity in PK-E2 was not observed. Residual reactivity was observed in the longer duration of the PK-E3 and E4 experiments and was not observed to end. PK reactivity differed from the PS principally in the magnitude between the bulk reactivity rates. 76  The difference in these dissolution behaviours reflects the difference in the dissolution rates of brucite and serpentine. Brucite dissolves two orders of magnitude faster and reacts to completion. In contrast, Mg-silicates react slower and continue to react over time (Lu, 2020). When rapidly reactive phases are absent or have been consumed, reactivity from lizardite is significant as it governs the reactivity both in rate and magnitude. 1% of the total MgO content and 3.1 \u2013 4.6% of serpentine in the fines from experiments PK-E1 and PK-E3 reacted. Contributions of Mg from forsterite and possibly some Ca mean this is an overestimate, but it agrees well with the expectation of 4% labile Mg from serpentine (Lu, 2020). The higher reactivity in experiment PK-E3 agrees with the batch dissolution results, which found the FPK-2 sample to release 1.4 times as much cations as FPK-1. However, there is no clear correlation for this higher reactivity with either mineralogy or grain size. The amount of CO2 captured from injection experiments was 67% and 57% of the equivalent results from batch dissolution tests for FPK-1 and FPK-2. These results, along with that from the FPS, are notable given the differences between the test conditions in their ratio of acidity to mineral buffering capacity. This ratio is high in batch dissolution and low in injection experiments. Lu (2020) conducted batch dissolution experiments which buffered mineral and tailings samples with aqueous solutions at a pH of 3 to 5. In unsaturated conditions during injection, with the pore water in equilibrium with 10% CO2, the pH is usually in the range of 7 to 10 (Assima et al., 2012, 2013a; Harrison et al., 2016). Thus, injection experiments are likely to achieve lower degrees of carbon sequestration than expected from dissolution results. However, one key parameter which differed between the tests was the duration. This parameter may also explain the relative agreement between the two FPK samples. Injection on FPK-1 in PK-E1 was 11 times longer than the batch dissolution, and the sequestered mass of CO2 is closer to the cation 77  dissolution in the batch tests than for FPK-2. Meanwhile, PK-E3, which was conducted 7 times longer than the batch dissolution experiment, amounted to 57% of the expected reactivity. CO2 injection results may converge with the batch dissolution results with increased experimental time. 3.5.2.1.2 Source of Ca While Mg is typically the cation under scrutiny due to its larger availability, Ca should not be discounted, as the noncarbonate contributions in the FPK batch dissolution results were significant (16% and 21% of the cation total). Whether significant Ca is accessed via injection experimental conditions is unclear. Two possible Ca sources exist in the FPK, including smectites (14% and 11 wt.% in FPK-1 and FPK-2) and trace Ca-silicates such as wollastonite (below detection to trace) and diopside (7% and 4 wt.% in FPK-1 and FPK-2). The leached Ca abundance from the batch dissolution experiments, requires an equivalent cation exchange capacity of 2040 \u2013 4860 cmol\/kg of smectite, which is far more than typical literature values of <100 cmol\/kg (Bergaya & Vayer, 1997; Meier, 1999). In comparison, this would require complete dissolution of 0.3 and 0.6 wt.% wollastonite from FPK-1 and FPK-2, respectively. Wollastonite was only identified in the FPK-2 samples in qXRD, while it could have been below detection in the FPK-1 samples, which would be consistent with the greater Ca accessed in the FPK-2 batch dissolution experiment. Wollastonite is known to react with CO2 faster than any Mg-silicate (Huijgen et al., 2006; Kojima et al., 1997). Considering diopside as a Ca source, the amount of leached Ca does not correlate with the diopside abundance and would require 9% and 31% of diopside to dissolve from FPK-1 and FPK-2, respectively, which is unlikely given its slow dissolution kinetics (Oelkers, 1999). Therefore, trace Ca-silicate dissolution from wollastonite may be the most likely Ca source. 78  3.5.2.2 Particle Size Distribution In the coarse-fine blends, the grain size of the fines is small enough for the grains to dissolve to near completion, and therefore the limiting factor is the mineralogy (Harrison et al., 2015). As grain sizes increase, the limiting factor is the high mass to surface area ratio (Assima et al., 2013a; Harrison et al., 2015; Lechat et al., 2016). The reactivity of the coarse material was assessed in comparison to the coarse-fine blend to determine the effect of the grain size distribution on carbon mineralization. In this study, the fine-grained tailings had an average grain size of between 40 and 200 \u00b5m, while the coarse grain sizes had an average size between 1500 and 2000 \u00b5m. Harrison et al. (2015) compared the reactivity of very fine (<53 \u00b5m) and medium (250-500 \u00b5m) grained brucite and found the carbonation extent to decline from 94 to 59% of the brucite as the grain size increased. Likewise, Assima et al. (2013a) investigated grain size distributions up to 2000 \u00b5m and found the coarsest fraction (1800 \u2013 2000 \u00b5m) captured 40% of the CO2 of the finest fraction (<75 \u00b5m). Experiment PS-E10 possessed no fines, with 1.4 wt.% brucite in the coarse phase. Extended reactivity for 1300 hours led to a TIC increase of 0.21 wt.% CO2, equivalent to the reaction of 25% of the brucite to hydromagnesite. Compared to the average experiment with fines (PS-E1 \u2013 PS-E9), which captured 1.41 wt.% CO2, this is one-seventh of the reactivity after ~5 times the injection duration. Accounting for this on a per mass and per brucite basis, the coarse is approximately three times less reactive than the fines. On an equivalent duration basis, this factor would increase to 15 times. However, this is not an apt comparison as the reaction rate is not linear with time. Experiment PK-E3 captured a similar magnitude of carbon as PS-E10. The amount of captured carbon for PK-E3 doubles from 200 hours to 650 hours (Figure 3.6). If the reactivity of PS-E10 is assumed to be similar, then another 50% increase might be expected by 1300 hours. 79  This assumption proposes that the duration increase by a factor of 5 correlates to a reactivity increase by a factor of 3. Therefore, the CPS-2 (mean grain size of 1980 \u00b5m) is ~9 times less reactive than the FPS (mean grain size of 109 \u00b5m). Experiment PK-E4 was run on CPK-2, with a mass fraction below 425 \u00b5m of 18 wt.%, compared to the FPK bearing experiments, which had a fines abundance of ~40 wt.%. Though the CPK-2 contains some fines, there are differences in the PSD of the FPK versus the CPK fines. The average grain size for FPK-1 and FPK-2 are 42 and 182 \u00b5m, while for the sieved fraction of CPK-1 and CPK-2, they are 278 and 324 \u00b5m. This increase in grain size and the lower abundance limits the ability of the CPK fines to coat the coarse grains, making it probable that the coarse did contribute Mg. Therefore, attributing the sequestered CO2 in experiment PK-E4 to the reaction of the CPK-2 fines alone, as done to calculate the degree of lizardite reactivity, is an oversimplification. The CPK-2 in PK-E4 captured approximately one-quarter of the CO2 of the 3:1 CPK-2: FPK-2 blend in PK-E3 in an equivalent amount of time. On a mass (including accounting for the CPK-2 fines in the PK blend) and lizardite abundance basis, the CPK-2 (mean grain size of 1740 \u00b5m) is approximately twelve times less reactive than the FPK-2 (mean grain size of 182 \u00b5m). The coarse phase was an order of magnitude less reactive than the fine fraction for both PS and PK. This finding is significant as this occurs independent of mineralogy. However, this disparity is more significant than determined by Assima et al. (2013a) for similar size fractions of asbestos-bearing PS. They found that the coarsest fraction (1800-2000 \u00b5m) was only 2.5 times less reactive than the finest fraction (<75 \u00b5m)(Assima et al., 2013a). The results shown here indicate that the grain size has a very significant impact on the reactivity to CO2. 80  3.5.2.3 Pore Saturation Water is essential for carbon mineralization as a reactant and as a reaction medium, facilitating the dissolution of CO2 and Mg2+ and the precipitation of Mg-carbonates (Harrison et al., 2016). Harrison et al. (2015, 2016) denoted that the ideal degree of pore saturation lies between 35 and 60 wt.%, while Assima et al. (2013a) found it to be 35 and 50 wt.% at pCO2\u2019s of 2.1% and 14%, respectively. High pore saturations inhibit carbon mineralization as they induce heterogeneous, channelized gas flow during injection (Harrison et al., 2016). However, the degree of pore saturation can be influenced by the moisture content or the degree of compaction. These controls were experimented with by varying both these factors. In the compacted conditions for experiments PS-E2 to E7 (bulk densities of 2.0 \u2013 2.1 g\u00b7cm3), the moisture content needed to achieve 30-40% pore saturation was only 5 wt.%. Harrison et al. (2015) found that brucite carbonation was inhibited at a moisture content of 6.4 wt.%, whereas superior carbonation resulted when the water content was 13 wt.%. However, the results from experiments PS-E2 to E7 show no indication of being inhibited by low moisture contents. This result is consistent with the hydrophilic nature and large surface area of serpentine-dominated tailings, which would hold onto films of water better than the quartz sand of Harrison et al. (2015, 2016). Comparing the degree of brucite carbonation to Harrison et al. (2015) on an equivalent grain size basis, the results are comparable (73% achieved in this study versus an expected 83%). Meanwhile, experiment PS-E8 was performed at a moisture content of 5%, a low degree of compaction (1.4 g\u00b7cm3), and hence a low degree of pore saturation. However, it resulted in similar amounts of sequestered carbon to those at higher bulk densities (Figure 3.10A). Finally, experiment PS-E9 tested the impact of increased moisture content (8 wt.%) at a low degree of 81  compaction (1.6 g\u00b7cm3). However, the results were again consistent with those of previous experiments. The consistent results, regardless of the changes in moisture content and degree of compaction, require explanation. First, water movement and heterogeneity is limited by the experimental set-up and by the well-graded grain size distribution. In this study, the tailings were not compacted into vertical columns as tested by Assima et al. (2013a) and Harrison et al. (2015, 2016). Compaction into horizontal pipes focused on the lateral injection required in tailings storage facilities with a very high lateral to vertical profile. In this orientation, the vertical gradient for gravity to act upon is decreased. Further, the well-graded grain size distribution has a higher residual water saturation as more water is retained due to adsorption to the high grain surface area. These conditions limit drainage, which can lead to saturated conditions at the column base, which impacts gas distribution, and dry conditions at the column top, which inhibits brucite carbonation (Harrison et al., 2016). Second, changes in the grain size distribution, moisture content and degree of compaction all influence the gas-water and water-mineral interfacial areas. The well-graded grain size distribution increases gas flow-path tortuosity, while the water and the reactive fines are concentrated within the porosity of the coarse grains. Enhancing surface area for CO2 dissolution from the gas phase and for mineral dissolution from the reactive fines increases carbon mineralization by promoting supersaturated conditions (Harrison et al., 2017). Maintaining the moisture content and lowering the bulk density may have had no impact because while the gas-water interfacial area was reduced, this may have been compensated for by the increased gas flow-path length. 82  Third, the minimum moisture content required is related to the material\u2019s reactivity and is therefore material-specific. Given that the brucite abundance here is three times less than that studied by Harrison et al. (2015), it is not surprising that low moisture contents do not excessively inhibit brucite carbonation. The higher moisture content in a lower bulk density state may not have had any significant impact because there was already sufficient water to facilitate carbon mineralization. 3.5.2.4 CO2 Supply The impact of CO2 supply was studied by varying the injected flux, injection duration, and the pCO2 content of the gas. Injection of CO2-rich gases increases CO2 dissolution into solution and increases brucite dissolution, as brucite dissolution rates correlate linearly with pCO2 (Harrison et al., 2013a). In unsaturated porous media, the increased gas-water interfacial area allows the pore water to reach saturation with CO2 (Harrison et al., 2016). This result is visible from the undulating CO2 concentrations observed in the outlet sensor data (Figure 3.3 and Figure 3.4). The solubility of CO2 in water increases with decreasing temperature, so diurnal temperature effects lead to undulating CO2 concentrations. This solubility effect on the gas phase concentrations can only occur if the aqueous phase is closer to saturation than the magnitude of change due to temperature fluctuations. Similar effects can be seen in the concentration profiles of Harrison et al. (2015, 2016). Harrison et al. (2016) found that no 13C isotopic kinetic fractionation effect occurred in either DIC or solid carbonate precipitates, which is commonly identified when CO2 dissolution into solution is rate-limiting (Beinlich & Austrheim, 2012; Gras et al., 2017; Harrison et al., 2013a; O\u2019Neil & Barnes, 1971; Wilson et al., 2010, 2014). This similarity indicates that in these experiments, CO2 injection has shifted the rate-limiting process to mineral dissolution. 83  Different gas fluxes and injection durations impacted the amount of carbonation that occurred during the PS carbonation experiments. High gas fluxes maximize reactivity as they oversupply CO2, ensuring that mineral dissolution is always the rate-limiting step. Further, they also ensure that the pCO2 of the gas phase is always equal to that of the injected gas. For example, in PS-E1, the effluent concentrations reach 8% CO2 within 60 hours of injection (Figure 3.3). Whereas with low gas fluxes, the reactivity can draw down the CO2 from the gas phase, with the pCO2 of the gas phase only increasing as the reactivity of the tailings declines. For short-term experiments striving to achieve a high degree of reactivity, higher fluxes are ideal. However, these high fluxes compromise the experimental capture efficiency, as excess CO2 is vented. To compare the PS experiments while accounting for the flux and injection duration, the captured CO2 mass has been plotted against the injected CO2 mass in Figure 3.10A. There is a slight trend between the amounts of carbon captured and injected, which can be understood since higher pCO2\u2019s and more exposure time would lead to more carbonation. However, the results are mostly the same regardless of the injection flux or duration. Therefore, lower fluxes and longer durations can achieve the same results as high fluxes and short timeframes while maintaining an overall high process efficiency (Figure 3.10B). The injection efficiency is dependent on the injected rate and duration, which explains why the relationship does not follow the flow rate, as represented by the symbol (highest to lowest flow rates shown by diamond, square, triangle and circle markers in Figure 3.10). The experiments with the lowest flow rates and injected masses of CO2 captured ~50% of the injected CO2, while those with the highest flow rates and injected masses of CO2 captured <20% of the injected CO2. 100% efficiency could be achieved by tuning the injection rate to match the reactivity of the tailings by monitoring the effluent CO2 concentration (Harrison et al., 2013b; Power et al., 2014). 84   Figure 3.10. A. The captured mass of CO2 from the PS experiment series is plotted versus the mass of injected CO2, with their respective confidence intervals. A line of best fit through the data points and a trend to estimate the early time reactivity are shown. The flow rate magnitude is shown by the diamond, square, triangle and circle symbols, from highest to lowest. B. The efficiency of the captured CO2 mass against the mass of injected CO2 is plotted with an approximate trendline through the data. 85  Since mineral dissolution is dependent on the pCO2, the injected pCO2 directly influences the reactivity of the tailings. All experiments using SFG were done with a CO2 concentration of 10%. However, PK-E2 was run with DFG, which had a measured volumetric concentration of 7.6% CO2. In Figure 3.3, PK-E2 stabilizes at ~6.4% CO2 due to the relatively lower pCO2 of the DFG. In Figure 3.5, PK-E2 proves to be less reactive than PK-E3, even when accounting for mineralogical differences. PK-E2 performs in between the CPK: FPK blend (PK-E3) and the pure CPK (PK-E4) as it has the fines to be more reactive than just the CPK alone. At later times the reactivity rate approaches that of PK-E3. At this point, PK-E3 has declined as readily accessible cations were leached quickly, while for PK-E2, this occurs more slowly due to the lower pCO2. Experiment PK-E2 captured the lowest amount of carbon at only 0.02 wt.% CO2, due to the lower pCO2 and shorter injection duration. The low increase in carbon was also difficult to measure as it was well within the heterogeneity of the inorganic carbon in the unreacted and carbonated samples. Other gas phases present in DFG can affect the carbon mineralization reaction, and these were not tested in the SFG, which balanced the CO2 with inert N2. DFG consists primarily of N2, CO2, O2, H2O, CO, NOx, SOx and hydrocarbons (Legrand et al., 2020). While CO has been confirmed to not impact carbon mineralization (Sarvaramini et al., 2014), the O2 will have detrimental effects as it is known to oxidize Fe2+ to Fe3+ and lead to Fe-hydroxide formation, which passivates mineral surfaces (Assima et al., 2012, 2013a, 2014c, 2014a, 2014d). The other gas of concern is the SOx, as it can dissolve as SO4 and precipitate out as Mg-sulphate minerals, using up some of the labile cation capacity (Hamilton et al., 2020; Mills et al., 2010). It is possible that these phases could have reduced PK reactivity in the DFG injection experiment. 86  3.5.3 Carbonate Stability Carbon sequestration potential is maximized if the accessed cations can be used to their full capacity. Thus, the form of the Mg-carbonate is critical. Characterization of the carbonate form precipitated in the PK experiments was not possible as the abundance was below detection in XRD and TGA and too difficult to locate in SEM. From the characterization of the carbonated PS samples, nesquehonite and hydromagnesite were identified. Hydromagnesite was qualitatively considered the dominant Mg-carbonate in the PS experiments as it was pervasive in SEM, detected in every single thermogravimetry sample and is considered more stable (Canterford et al., 1984; Harrison et al., 2019; Langmuir, 1965). Though nesquehonite was identified in some XRD samples and SEM, assuming all the carbon was captured as hydromagnesite allowed for a defined stoichiometry to determine the amount of Mg dissolved from brucite. Nesquehonite may have been detected in XRD due to its 100-\u00b5m crystal size compared to the 5-\u00b5m scale of the hydromagnesite. Other studies identified nesquehonite in XRD when present at the 100 to 200-\u00b5m scale but could not detect a Mg-carbonate similar to artinite at the 5-\u00b5m scale, as it was deemed amorphous (Harrison et al., 2015, 2016). Quantitative information on the form of the precipitated Mg-carbonate can be obtained from the amount of carbon captured and the mass of water lost to precipitate crystallization. The mean TIC increase for PS experiments E1 \u2013 E9 is 1.41 wt.%, while the mean sequestered mass from the CO2 mass balance was 1.56 wt.% (Table 3.8). Using the two measurements of the sequestered carbon as a range, accounts for uncertainty on the measurement. The mean pore water loss was 1.80 \u00b1 0.19 wt.% H2O (Table 3.10). While information regarding the amount of Mg incorporated into the Mg-carbonate is absent, there are a limited number of defined Mg-carbonate mineral phases. In CO2 injection experimental conditions, phases have been observed to form with 87  MgO: CO2 ratios of 1:1 (nesquehonite, lansfordite), 5:4 (dypingite, hydromagnesite), and 2:1 (pseudo-artinite) (Assima et al., 2013a, 2014c; Hamilton et al., 2020; Harrison et al., 2015, 2016, 2017; Power et al., 2020). These factors were each applied to the CO2 range to determine the MgO content, and the molar abundances of the MgO, CO2 and H2O were normalized. Figure 3.11 shows this plotted, with one standard deviation on the H2O content, on a ternary diagram among other known minerals and experimentally observed Mg-carbonate phases. Nesquehonite and hydromagnesite, which were also observed in carbonated samples in this study, are shown. The calculated composition is consistent with the formation of nesquehonite, protohydromagnesite (Canterford et al., 1984; Davies & Bubela, 1973), and pseudo-artinite (Harrison et al., 2015, 2016), at MgO: CO2 ratios of 1:1, 5:4 and 2:1, respectively. The pore water loss excludes the possibility of lansfordite, dypingite or hydromagnesite from being primary phases. Since the final composition of the precipitated Mg-carbonates was characterized as major hydromagnesite and minor nesquehonite, a phase transformation must have occurred. Hydrated Mg-carbonates are metastable, dissolving and re-precipitating into other phases over time due to changes in temperature, pH, pCO2, and RH (H\u00e4nchen et al., 2008; Harrison et al., 2019, 2021; Hopkinson et al., 2008; H\u00f6velmann et al., 2012; Morgan et al., 2015). For a phase transformation from pseudo-artinite to hydromagnesite or nesquehonite to occur, CO2 would need to be gained after the experiment had completed, which is improbable. Likewise, the transformation from protohydromagnesite to nesquehonite would also require adding CO2. Therefore, the nesquehonite observed weeks and months after experimental completion is a primary phase. Either protohydromagnesite or nesquehonite could have transformed to hydromagnesite over time by losing water, and water and CO2, respectively. However, protohydromagnesite has not been observed to form in carbon mineralization experimental conditions but rather is known as an 88  intermediary phase during the decomposition of nesquehonite to hydromagnesite (Canterford et al., 1984; Davies & Bubela, 1973). Therefore, it is likely that nesquehonite was the primary Mg-carbonate phase that formed during injection, and much of it dissolved and re-precipitated into hydromagnesite, losing water and exsolving CO2.   Figure 3.11. MgO-H2O-CO2 ternary diagram for the possible compositions of the precipitated Mg-carbonates among brucite, magnesite and hydrous Mg-carbonate minerals known to form in CO2 injection experimental conditions. Dotted lines represent a MgO: CO2 ratio of 2, 1.25, and 1. The range in the initial precipitated Mg-carbonate composition from the PS experiment series is presented in red for each ratio, with the extended red lines being the standard deviation on the H2O content. Protohydromagnesite is included due to its proximity to one of the possible precipitated compositions. Dark red indicates the most likely precipitated composition. 89  Hydromagnesite evolving from more hydrated minerals such as nesquehonite is not uncommon (Ballirano et al., 2013; Davies & Bubela, 1973; Hopkinson et al., 2008, 2012; Zarandi et al., 2017b). This composition evolution from nesquehonite to hydromagnesite would have occurred due to the change in environmental conditions from within the experimental set-up (high pCO2 and RH) to ambient conditions (low pCO2 and RH) when the samples were being dried. Nesquehonite stability is enhanced in higher pCO2 and RH conditions preventing the transformation towards hydromagnesite (Gras et al., 2020; Morgan et al., 2015), while dry conditions lead to the transformation to hydromagnesite (Zarandi et al., 2017b). The phase formed has two important impacts on the mineralization of carbon. First, the initial formation of non-porous, crystalline nesquehonite may have led to brucite surface passivation and the incomplete carbonation of brucite (Harrison et al., 2015, 2016; Zarandi et al., 2017b). Second, hydromagnesite is a less efficient carbon sink as it has a MgO: CO2 ratio of 5:4, which sequesters 20% less CO2 per mole of MgO than nesquehonite. This phase transformation means that ~20% of the initially sequestered CO2 was released. Evidence of this exsolution may be observed between the gas-phase mass balance and the TIC analysis. The gas-phase mass balance is made as the injection occurs, whereas the TIC is measured weeks to months after the experiment was completed, potentially after CO2 exsolution has taken place. The gas mass balance was consistently higher than the TIC increase, with this difference being on average ~10%. If TIC had been measured earlier, it is possible that higher readings would have been recorded. While CO2 exsolution may explain the systematic difference, these methods have good agreement regardless.  90  3.6 Implications 3.6.1 Sequestration Magnitude Given the effort required to implement CO2 injection into tailings management practices, it is worth comparing the sequestration rates and capacities with those attained passively in similar deposit types (Figure 3.12). The passive carbonation rate at the Diavik Diamond Mine (NT, Canada) has been assessed to capture 0.15 kt CO2\/Mt tailings\/year in the FPK tailings facility, with the FPK dominating the tailings production (Wilson et al., 2009b). Three estimates have been made at nickel mines or deposits. Mouth Keith Nickel Mine (WA, Australia) passively sequesters 3.62 kt CO2\/Mt tailings each year (Wilson et al., 2014). The Dumont Nickel Project (QC, Canada) has been estimated to capture 1.40 kt CO2\/Mt tailings\/year (Gras et al., 2020), while the Baptiste Nickel Project (BC, Canada) is expected to capture, on average, 0.39 kt CO2\/Mt tailings\/year (Power et al., 2020). Passive sequestration rates are limited by the areal extent of the exposed tailings in the storage facility, the rate of tailings deposition, the water content of the deposited tailings, the form of tailings storage (subaerial versus subaqueous), and the impact of the local climate on the degree of pore saturation and the type of precipitation (Li et al., 2018; Power et al., 2020; Wilson et al., 2014), which all limit the exposure of the upper tailings surface to ambient CO2. Injection of CO2 is not limited by these same constraints, as it enables full access to the tailings mass and occurs in the subsurface. However, CO2 injection should not be considered in opposition to passive carbon mineralization, as these surface processes will still proceed regardless of the injection being done in the subsurface (Hamilton et al., 2021; Power et al., 2014). Further, changes to tailings storage management practices required to make injection feasible, such as co-disposal of coarse and fines, and water management, would also improve passive carbon 91  sequestration (Assima et al., 2014b, 2014c; Gras et al., 2017, 2020; Hamilton et al., 2021; Kandji et al., 2015; Zarandi et al., 2017a).  Figure 3.12. CO2 emission rates and injection (this study) and passive sequestration rates for the Diavik Diamond Mine (NT, Canada), Gahcho Ku\u00e9 Diamond Mine (NT, Canada), Mount Keith Nickel Mine (WA, Australia), Dumont Nickel Project (QC, Canada), and the Baptiste Nickel Project (BC, Canada). Percentages above the symbols are the amount of the total emissions that could be sequestered at the respective mine. The experimental injection sequestration capacities determined in this study were 0.5 \u2013 2.0 kt CO2\/Mt tailings for Gahcho Ku\u00e9, ranging from the CPK to the CPK and FPK mix. For Baptiste, the capacity was 14.1 kt CO2\/Mt tailings. The results from the CPS are not considered as Baptiste is not expected to generate a coarse stream of tailings. All these results were achieved over an injection timeframe of 3-4 weeks. This magnitude is 13 and 8 times greater than the observed passive rate in diamond and nickel mines, respectively. Further, this occurred in one-twelfth the time of the passive rates implying that these enhancements would be even greater with equivalent 92  reaction durations. The PK experiments represent the minimum injection reactivity for an ultramafic deposit, as they lack highly reactive phases such as brucite and possess a relatively low abundance of serpentine and olivine at only 30 \u2013 40 wt.%. Meanwhile, the PS experiments are highly idealized, where very brucite-rich tailings have been explicitly targeted for their reactivity, and as such, represent the maximum reactivity at this abundance of fines. The short time frame for the experimental injection means carbon sequestration potential remains in these materials. The duration of CO2 injection remains a variable due to the climatic conditions of the mine. Cold conditions could prohibit carbon mineralization if the water in the shallow subsurface freezes. Regardless of the specific climatic conditions, the durations tested here are likely to be extended when implemented at the mine scale. For example, Baptiste has above freezing temperatures for five months of the year, and at Gahcho Ku\u00e9, this occurs for four months of the year (De Beers Group, 2019a; Power et al., 2020). At Gahcho Ku\u00e9, the sequestration capacity of 2.0 kt CO2\/Mt PK\/year represents 5.2% of their total annual CO2 emissions and 14% of their point source emissions, which could be captured and injected (De Beers Group, 2019b). If just the CPK alone is used, then the capacity of 0.5 kt CO2\/Mt PK\/year for the two-thirds of the produced PK that is CPK, would amount to 0.9% of total emissions and 2.4% of point source emissions. Area source emissions make up nearly two-thirds of Gahcho Ku\u00e9\u2019s emissions, and for the Baptiste Nickel Project, which intends to run on hydroelectric power (Power et al., 2020), they will be 100% of emissions. For Baptiste to perform CO2 injection, another source of CO2-rich gas would be needed, with potential sources being a nearby pipeline compressor station (Power et al., 2020) or a low purity DAC system (Kelemen et al., 2020). In comparison to Baptiste\u2019s mine emissions, CO2 injection could capture 572% of their emissions. While Baptiste does have a low carbon footprint, this is also due to a very high brucite 93  content in the fines. However, for Baptiste to be carbon neutral, only 2.2 wt.% brucite would be needed in the fines. Since the expected average brucite content is 1.8 wt.%, acquiring 2.2 wt.% from one-quarter of the deposit is very achievable given the computed brucite abundance distribution (Vanderzee et al., 2019). This estimate is similar to that made by Vanderzee et al. (2019), which indicated that Baptiste only needed 30% of the most reactive tailings to make the mine carbon neutral. CO2 injection enables this to be accomplished within weeks.  3.6.2 Large-Scale Implementation It is clear that the injection of CO2-rich gas into mine tailings accelerates passive sequestration rates and can sequester the magnitude of emissions produced by low-carbon mines. While these centimetre-scale permeability and reactivity tests were an essential first step, more must be done to demonstrate injection applicability as an engineered approach to mineralize carbon in mine tailings. Meter-scale experiments are required because the reaction rate and magnitude uncertainties increase with the experimental scale. This study has several implications for such larger-scale studies, from the metre to the mine scale. 3.6.2.1 Characterization Reactivity characterization is critical to identify reactive materials for experimentation and to assess the performance of the carbonation strategy. In this study, this was accomplished by three approaches, whole-rock chemistry characterization, mineralogy characterization, and batch dissolution tests. Whole-rock chemistry provides a useful metric as it enables reactivity comparisons based on the total cation capacity of the sample. However, for reactivity prediction, it is not useful by itself, as the mineralogy controls the dissolution of the cations. Experimentation on pure mineral phases has determined the reactivity of these phases allowing for their reactivity 94  to be predicted. Knowing their reactivity relative to one another allows for simplifications to assess bulk reactivity in mineralogically complex samples. On this basis, the portion of Mg leached from PS brucite was assessed to be 73% of its total, and the reactivity of PK lizardite was 4% of its total. However, the reactivity of a mineral is also dependent on the grain size and leaching approach. While Harrison et al. (2015, 2016) performed injection testing on pure brucite of different grain sizes, enabling a determination directly applicable to further injection testing, mineral dissolution rates are commonly assessed by dissolution tests (Lu, 2020; Paulo et al., 2021). Labile Mg contents are likely to vary based on the particle size distribution of the tailings, as finer grains have higher surface area to mass ratios (Assima et al., 2013a; Harrison et al., 2015; Lechat et al., 2016). Further, dissolution test conditions provide high ratios of acidity to mineral buffering capacity, whereas under injection conditions, the reverse is the case. Batch dissolution experiments allow for a rapid estimate of the maximum reactivity under idealized pH conditions. This study demonstrated that with increased reaction times of an order of magnitude, CO2 injection results approach those found by batch dissolution methods. Batch dissolution testing is an effective characterization tool that can be conducted rapidly compared to injection testing and may also enable an understanding of the longer-term reactivity of mine tailings. This method could allow for sequestration capacities to be predicted for longer durations than are easily tested experimentally. While brucite reacts significantly and lizardite releases some Mg under the time scales of the experiments in this study, the long-term reactivity of the mine tailings under CO2 injection is a critical uncertainty. Longer-term reactivity has not been evaluated in this study and could lead to injection magnitudes closing the gap between those predicted by dissolution testing. 95  3.6.2.2 Physical Design The current study provides implications for how large-scale experiments on CO2 injection may be successfully performed. The inclusion of coarse-grained tailings in the grain size distribution enables the permeability to inject, and the carbonation reaction insignificantly impacts the permeability. The inclusion of fines promotes rapid Mg dissolution, while the coarse grains are expected to insignificantly contribute where fines are present, as the fines coat the coarse grains. This effect indicates that the mineralogy of the coarse is insignificant, which implies that at mine sites without a coarse stream of tailings, waste rock would be suitable and that a form of co-disposal would facilitate injection (Zarandi et al., 2017a). The reactivity of the fines incentivizes strategies to maximize carbon sequestration by isolating highly reactive phases such as brucite-rich zones for fine-grained processing. While coarse streams are an order of magnitude less reactive than their equivalent fine-grained counterparts, at Gahcho Ku\u00e9, the CPK is operationally stacked in subaerial conditions. These storage conditions present an opportunity for large-scale injection deployment with minimal intervention into current tailings management processes, which currently sees the FPK deposited as a slurry. However, the capacity of the CPK to mineralize carbon is minor in the context of the mine\u2019s emissions. The requirement for unsaturated conditions poses a practical problem as mine tailings at active mines are usually saturated or oversaturated. Even dry stack tailings only produce mine wastes at moisture contents of 10 \u2013 20 wt.% (Klohn Crippen Berger, 2017; Oldecop et al., 2017; Watson et al., 2010). This moisture content is generally attributable to being at or just below saturation. Injecting into loosely compacted tailings may facilitate injection by increasing permeability and managing the water content attainable from mineral processing methods. Lowering the degree of compaction could also enable the inclusion of more fines into the grain 96  size distribution. This need for unsaturated conditions may mean that historical tailings storage facilities are more suitable for injection as tailings have been found that are too dry, requiring watering to promote carbon mineralization (Hamilton et al., 2021). While the upper layers of tailings may be kept loosely compacted, compaction will occur from further deposition and the causal consolidation as the stack height increases. For example, at the Baptiste Nickel Project, the height of the tailings in the storage facility is expected to be 60 meters at closure (Power et al., 2020), while the height of the stack pile at the now-closed Thetford Mines is 130 meters (Nowamooz et al., 2018). Compacted mine wastes at moisture contents of above 15% are likely to produce saturated conditions and make successful injection difficult. Therefore, injection is most suitable within upper tailings layers due to the lower saturation and compaction conditions. 3.6.2.3 Monitoring and Verification Monitoring is required during injection to ensure injected CO2 does not leak. In considering the methods used here to monitor and verify carbon sequestration, extensive sampling becomes difficult at large scales, while quantifying through the gas phase may be a possible approach given the success shown in this study and by that of eddy-covariance systems at large scales in other fields, such as forestry, ecology and agriculture (Morin et al., 2018; Rodda et al., 2021; Waldo et al., 2016). The gas-phase monitoring system could also be designed to provide feedback to the injection rate. Tailoring injection rates to match the reactivity of the tailings ensures rapid reactivity and efficient capture. A step-down of the injection rate over time is one approach, but a mechanized process could decrease flow in relation to the CO2 content detected in the exhaust. Monitoring is also necessary after injection has ceased to ensure that CO2 is not leaking out due to carbonate phase transfer. The phase transformation from nesquehonite to hydromagnesite could release 20% of the sequestered CO2. However, changes in experimental 97  conditions in the field are not likely to be as dramatic as in the lab, where samples were removed and dried out completely. While the pCO2 will decrease once injection ceases, water contents may be expected to increase over time as tailings are buried and consolidation occurs. Thus, the conditions in the tailings storage facility may favour nesquehonite stability. Fortunately, hydromagnesite is widely observed in ambient conditions, and as such, does not pose a risk for further CO2 exsolution (Canterford et al., 1984; Harrison et al., 2019; Langmuir, 1965). The stability of the phases formed in the high pCO2 and RH injection environment may mean that higher rates of CO2 can be captured per mole of Mg released than observed in this study. 98  Chapter 4: Metre-Scale Reactivity and Field Pilot Design 4.1 Synopsis Processed kimberlite (PK) mine wastes passively sequester carbon dioxide (CO2) via weathering processes that form magnesium carbonate minerals. These passive rates can be accelerated by increasing CO2 supply through injecting concentrated CO2, such as flue gas or CO2-enriched air. Here, we assess the feasibility of CO2 mineralization by injection into processed kimberlite mine wastes at the metre-scale. One and three-dimensional injection experiments were conducted in the lab at the University of British Columbia and in the field at the Gahcho Ku\u00e9 Diamond Mine (NT, Canada). One-dimensional injection into mixed PK packed inside 6-metre-long pipes examined large-scale reactivity, while three-dimensional injection experiments into pads of PK were performed to assess the design of a conceptual 100-tonne field pilot. Heterogeneity of the original and final carbon content made it difficult to verify by total inorganic carbon analysis the amount of CO2 captured as determined by a gas-phase mass balance. Grain size distribution, moisture content, bulk density, and permeability impact the reactivity and injectivity of PK. Experiments indicate that optimal PK blends could capture 0.10 to 0.23 wt.% CO2, meaning Gahcho Ku\u00e9 could annually sequester up to 16% of their power generation emissions via carbon mineralization at the mine scale.  99  4.2 Introduction Processed kimberlite (PK) from diamond mining is one suitable feedstock for carbon mineralization, as it contains the Mg-hydroxide and silicate minerals that can readily and rapidly provide Mg2+ (Assima et al., 2014a; Chakravarthy et al., 2020; Lu, 2020; Mervine et al., 2018; Moncur & Smith, 2012; Paulo et al., 2021; Rollo & Jamieson, 2006; Stubbs, 2020; Wilson et al., 2009b, 2011). Passive weathering of PK at Diavik Diamond Mine (NT, Canada) occurs naturally and has been documented to actively capture atmospheric carbon dioxide (CO2)(Moncur & Smith, 2012; Wilson et al., 2009b, 2011). Data suggests that carbon mineralization also occurs at the Ekati Diamond Mine (NT, Canada) and Voorspoed Diamond Mine (FS, South Africa)(Rollo & Jamieson, 2006; Stubbs, 2020). Gahcho Ku\u00e9 Diamond Mine (NT, Canada) is another potential candidate to mineralize carbon in its PK due to similar mineralogy. Gahcho Ku\u00e9 is located 280 kilometres northeast of Yellowknife, NT, Canada, and consists of four kimberlite pipes within country felsic granites and gneisses (Johnson & Pilotto, 2020). The kimberlite varies from tuffisitic kimberlite breccia to hypabyssal kimberlite, with the former hosting serpentinized olivine and the latter having more primary olivine (Hetman et al., 2004). Gahcho Ku\u00e9 began mining operations in 2016 and produces 3.2 Mt of PK and 122 kt of CO2 emissions each year (Johnson & Pilotto, 2020). Approximately two-thirds of the PK is produced as coarse processed kimberlite (CPK)(<6 mm, >0.3 mm), which is stacked in unsaturated conditions, while the remaining third is produced as fine processed kimberlite (FPK)(<0.3 mm), which is deposited as a slurry subaqueously (Johnson & Pilotto, 2020). It is possible that PK at Gahcho Ku\u00e9 passively sequesters CO2; however, this has not been established. Carbon sequestration is particularly inhibited by subaqueous containment of the tailings, as occurs at Diavik and Gahcho Ku\u00e9 (Johnson & Pilotto, 2020; Wilson et al., 2009b). However, 100  passive rates can also be limited by the surface area of the tailings storage facility, the rate of tailings deposition, the produced tailings moisture content, and the effect of the local climate on the moisture content and ambient temperature (Li et al., 2018; Power et al., 2020; Wilson et al., 2014). The rate and magnitude of passive carbon mineralization are limited by the CO2 supply rate to the mineral surface (Assima et al., 2013a; Harrison et al., 2013a, 2015; Wilson et al., 2010). If CO2 is supplied at sufficient rates, as occurs during injection, carbon sequestration rates increase until mineral dissolution becomes the new rate-limiting factor (Harrison et al., 2016). Centimetre-scale injection experiments in Chapter 3 have demonstrated an enhanced sequestration rate by over an order of magnitude compared to passive carbon mineralization rates. Successful injection requires specific conditions controlled by the grain size distribution, moisture content, and degree of compaction, which govern the permeability of the medium and the degree of pore saturation (Assima et al., 2012, 2013a; Harrison et al., 2015, 2016). The grain size distribution is controllable by blending coarse and fine tailings streams (Assima et al., 2014b; Hamilton et al., 2021; Zarandi et al., 2017a), while the required unsaturated conditions favour tailings management practices akin to those of filtered or dry stack tailings (Davies, 2011; Klohn Crippen Berger, 2017; Lupo & Hall, 2010; Oldecop et al., 2017; Watson et al., 2010). Under these conditions, metre-scale experiments must demonstrate the sequestration capacity of the centimetre-scale experiments as a step toward industrial-scale pilot projects while overcoming challenges related to reaction heterogeneity and field conditions (Gras et al., 2020; Hamilton et al., 2021; Lechat et al., 2016). Even under ideal conditions, carbon mineralization may be limited by mineral dissolution rates, the formation of passivating layers or minerals, and the reduction in moisture content (Assima et al., 2012, 2013a; Harrison et al., 2015, 2016; Zarandi et al., 2016, 2017b). 101  Two types of CO2 injection experiments were conducted in this study. The first experiment type approximated centimetre-scale pipe injection tests at a much larger scale. Gas, rich in CO2, was injected into one end of 6-meter-long pipes packed with blended coarse processed kimberlite (CPK) and fine processed kimberlite (FPK). The second set of experiments injected CO2 into a metre-scale pad of mixed PK. This set-up represented a larger-scale design and used layer heterogeneities to control the gas flow. Reactivity was assessed by gas-phase CO2 concentration measurements, total inorganic carbon (TIC) analysis on bulk and sieved samples, and total carbon (TC) analysis of bulk samples. To progress the technological readiness of CO2 injection into mine tailings for carbon mineralization, the objectives of this study were to, 1) assess metre-scale PK reactivity, 2) evaluate methods to account for the amount of sequestered carbon, and 3) test the conceptual, experimental design to provide recommendations for industrial-scale field pilots.  102  4.3 Methods 4.3.1 Sample Characterization Samples consisted of four bulk samples, two of CPK and two of FPK, sourced from the Gahcho Ku\u00e9 Diamond Mine (Table 4.1). CPK-1, CPK-2 and FPK-2 were all sourced from the processing circuit, while FPK-1 was dredged from the containment facility. Subsets of these samples were also used in Chapter 3. The four unreacted bulk samples were comprehensively characterized. The initial TIC content was measured by coulometry. Mineralogy was determined by quantitative X-Ray Diffraction (qXRD), thermogravimetric analysis (TGA), and QEMSCAN automated mineralogy (CPK only). Whole-rock chemistry was measured by inductively coupled plasma atomic emission spectroscopy (ICP-AES) and particle size distribution by laser diffraction and sieving. Particle surface area was determined by multipoint BET with N2 adsorption, while the density of the PK was measured by pycnometry. Characterization methods and detailed results are presented in Appendix 3. Table 4.1. List of tailings samples from Gahcho Ku\u00e9 Diamond Mine. Sample Type Source Sample Collection Date CPK-1 Tailings Processing Plant August 2019 FPK-1 Tailings FPK Containment August 2019 CPK-2 Tailings Processing Plant December 2019 FPK-2 Tailings Processing Plant December 2019   103  4.3.2 Experimental Set-Up 4.3.2.1 Six Metre Pipe Experiments The experimental design of the two metre-scale pipe experiments is shown in Figure 4.1. They were similar to the centimetre-scale experiments from Chapter 3. A blend of CPK (<6000 \u00b5m) and FPK (<425 \u00b5m) was produced at a mass ratio of ~3:1 and a moisture content of ~7 wt.%. This mass ratio was selected as a compromise between reactivity and permeability parameters as determined in Chapter 3. The mixed PK was compacted into two 6-metre-long HDPE pipes, with the length of the compacted mass of tailings being ~5.8 metres. Pipe-1 (18.9 cm diameter) was a field experiment run at Gahcho Ku\u00e9 Diamond Mine to increase the scale from the centimetre-scale experiments, while Pipe-2 (16.6 cm diameter) was performed in the lab at the University of British Columbia as a replicate experiment. The biggest difference between the environmental conditions of the experiments was the temperature, as the indoor setting for Pipe-2 experienced less temperature variation. Compaction in Pipe-1 was done in place, with a careful examination that compaction was evenly achieved. For Pipe-2, the pipe was rotated 180 degrees each time it was compacted to ensure the talus slope of the loose material did not affect the compaction. For both pipes, compaction was done every 5 kg. Standard Proctor compaction testing was done on the mixed PK from Pipe-2 to investigate the moisture content-density relationship and understand the relationship to permeability following the methodology of ASTM standard D698-12 (ASTM International, 2012). 104   Figure 4.1. Schematic of the 6-metre pipe injection experimental set-up showing the gas flow control, hydration and injection into the compacted tailings, with the installed sensors shown. Simulated flue gas (SFG; 90 vol.% N2 and 10 vol.% CO2) was bubbled through a hydration flask before injection to replicate the humidity of cooled diesel flue gas and reduce the amount of evaporation from the moist PK. Bosch Sensortec BMP 388 pressure sensors were used to determine the differential pressure needed to calculate the permeability. In addition, Vaisala GMP 221 and 251 CO2 sensors and HMP 110 humidity sensors were used to measure the CO2 concentration, relative humidity and temperature of the influent and effluent gas. Pressure and humidity sensors were positioned at the pipe inlet and outlet, while duplicate CO2 sensors were installed at the inlet, outlet and halfway along the pipe. Gas-phase measurements were recorded every 15 minutes. SFG or air was injected to measure the permeability, while SFG was injected to Hydration FlaskBronkhorstEL-FLOW Prestigemass flow meter\/controllerFLOWMass Flow ControllerBMP388T+PHMP110HMP110VAISALA| GMP251VAISALA| GMP251VAISALA| GMP251BMP388T+PCO2 SensorCompacted Tailings (5.8 m)Pressure SensorHumidity Sensor20 cm10% CO2Pipe-1: Phase 1: 0.72 L min-1 Phase 2: 0.05 \u2013 0.12 L min-1Pipe-2: Phase 1: 0.55 L min-1Phase 2: 0.1 L min-1Pipe-1: 78:22 CPK-1:FPK-1Pipe-2: 75:25 CPK-2:FPK-2105  assess the reactivity. Gas was injected in two phases for both experiments and was controlled by a Bronkhorst EL-Flow Prestige mass flow controller. The two injection phases involved the injection rate being reduced to demonstrate tailoring the injection to the PK reactivity to increase the CO2 capture efficiency. In Pipe-1, phase 1 consisted of a flow rate of 0.72 L min-1, while a leak reduced the flow rate in phase 2 to a flow rate in the range of 0.05 to 0.12 L min-1. In Pipe-2, the two phases were done at a flow rate of 0.55 L min-1 (identical flux as Pipe-1) and 0.1 L min-1. After experimental completion, analytical samples were taken every 50 cm along the length of the pipe and from three horizons (upper, middle and lower) to assess reaction homogeneity. Detailed experimental conditions are presented in Table 4.2.106  Table 4.2. 6-metre pipe injection experimental conditions. Experiment and Sample Fines, <425 \u00b5m (wt.%) Gas Flux (cm\u00b7min-1) Length (cm) Wet Mass (kg) Bulk Density (g\u00b7cm-3) Moisture Content a (wt.%) Duration (hours) Pipe-1  CPK-1: FPK-1b 39 2.54; 0.18 \u2013 0.42 581 264 1.58 6.5 140 Pipe-2 CPK-2: FPK-2c 39 2.54; 0.46 580 236 1.87 6.9 475 a Moisture content measured as the mass of water over the total sample mass. b Pipe-1 experiment blend prepared with 22 wt.% FPK. c Pipe-2 experiment blend prepared with 25 wt.% FPK. 107  4.3.2.2 Pad Experiments Two pads of deposited PK were constructed as metre-scale experiments. The design of these injection experiments more closely matches how injection could work at the mine scale. 4.3.2.2.1 Pad-1 Pad-1 was run in the field at Gahcho Ku\u00e9 Diamond Mine and was constructed within a 112 cm by 96 cm by 100 cm intermediate bulk container (IBC). The pad consisted of a 12 cm thick layer of mixed PK (3:1 mass ratio CPK-1 to FPK-1) layered between two 6 cm layers of FPK-1 (Figure 4.2). The mixed PK layer was the medium into which CO2 was injected, while the FPK layers were meant to be barriers to gas flow. A base layer of CPK was used to elevate the FPK and mixed PK layers to ease construction and was not involved in the reaction.108   Figure 4.2. Schematic of the Pad-1 injection experimental set-up showing the gas flow control, hydration and injection into the mixed PK layer, with embedded CO2 sensors and effluent ports shown. Hydration FlaskBronkhorstEL-FLOW Prestigemass flow meter\/controllerFLOWMass Flow Controller1 234510% CO2Tailings Pad (1.12 m)0.96 m0.24 mMixed CPK: FPK Layer Cross Section0 \u2013 20 hours: 0.08 L min-120 \u2013 132 hours: 0.24 L min-11 23FPK-176:24 CPK-1: FPK-1 MixCPK-1CO2 SensorExhaust Port109  SFG was hydrated and injected into the mixed PK layer and was directed by the pressure gradient to five outlet ports, each 30 degrees apart on the outside of the IBC walls, by opening one port at a time. Six CO2 sensors (3 Vaisala GMP 221 and 3 GMP 251) were embedded in the mixed PK layer, with an additional sensor located in the headspace above the upper FPK layer. CO2 concentrations were recorded every fifteen minutes. The effluent was allowed to exit via the headspace of the pad before this was sealed off to direct the gas to the effluent ports, in the following order: port 3 (opened initially), port 5 (opened at 94 hours), port 2 (opened at 101 hours), port 1 (opened at 118 hours), and port 4 (opened at 125 hours). This order was followed to promote the homogenous distribution of gas flow throughout the mixed PK layer due to the outlet port direction being changed sporadically. SFG was injected with a Bronkhorst EL-Flow Prestige mass flow controller initially at a rate of 0.08 L min-1 for the first 20 hours before increasing to 0.24 L min-1 due to slow plume development. After experimental completion, the pad was dismantled, and grab samples were taken in a three-by-four grid (grid spacing of 30 cm width and 35 cm length) of the upper FPK, the mixed PK layer, and the lower FPK layer. Eight core samples were also taken of the upper FPK layer and the mixed PK layer in a distributed pattern that included some of the desiccation cracks in the upper FPK layer. Experimental conditions are listed in Table 4.3. 110  Table 4.3. Pad injection experimental set-up conditions. Layer Material Fines, <425 \u00b5m (wt.%) Flow Rate (L\u00b7min-1) Height (cm) Wet Mass (kg) Bulk Density (g\u00b7cm-3) Moisture Content a (wt.%) Duration (hours) Pad-1 (112 cm x 96 cm)  0.08; 0.24 64 400   132 Layer 1 (Top) FPK-1 100  6 110 1.7 28.7  Layer 2 CPK-1; FPK-1b 41  12 180 1.4 6.6  Layer 3 FPK-1 100  6 110 1.7 34.0  Layer 4 (Base) CPK-1 23  40 820 1.9 5.9  Pad-2 (115 cm x 95 cm)  1; 0.5; 0.25 55 1050   765 Layer 1 (Top) CPK-2 18  5 100 1.8 1.2  Layer 2 FPK-2 100  8 150 1.7 21.8  Layer 3 CPK-2; FPK-2c 39  30 550 1.7 6.8  Layer 4 (Base) CPK-2 18  12 250 1.9 1.3  a Moisture content measured as the mass of water over the total sample mass. 111  4.3.2.2.2 Pad-2 The Pad-2 design was changed from Pad-1 to improve the experiment\u2019s performance and interpretation of the results. To facilitate the gas transport through the pad, a perforated injection pipe was used to direct gas unilaterally into the base of the pad of PK (Figure 4.3). The pad of PK was constructed inside a 115 cm by 95 cm by 95 cm IBC. A 5.2 cm diameter, 115 cm long, perforated pipe was placed in the bottom corner of the IBC and buried in a 12 cm thick layer of CPK-2. A blend with 33% FPK-2 (representing the produced proportion of CPK and FPK at Gahcho Ku\u00e9) was used for the mixed PK layer. Standard Proctor compaction testing was performed on this material to investigate the moisture content-density relationship. A 30 cm thick layer of the mixed PK was compacted in two stages. This layer was followed by an 8.3 cm layer of FPK-2 and a 5 cm layer of CPK-2. The high permeability of the base layer of CPK was intended to allow gas to flow laterally before rising into the layer of mixed PK. The cap layer of FPK was intended to be an impermeable gas barrier, with the top layer of CPK meant to limit evaporation from the FPK that would compromise the barrier\u2019s integrity. 112   Figure 4.3. Schematic of the Pad-2 injection experimental set-up showing the gas flow control, hydration and injection into the base CPK layer through the perforated pipe. Locations of the CO2 sensors and the effluent ports are shown. Hydration FlaskBronkhorstEL-FLOW Prestigemass flow meter\/controllerFLOWMass Flow ControllerEmbedded Perforated Injection PipeTailings PadMiddle Plane Cross Section0.95 m1.15 m0.55 m10% CO2Phase 1: 1 L min-1Phase 2: 0.5 L min-1Phase 3: 0.25 L min-1 FPK-267:33 CPK-2: FPK-2 MixCPK-2CO2 SensorExhaust Port1232113  Three CO2 sensors (Vaisala GMP 251) were distributed along the perforated injection pipe\u2019s length. Fourteen more sensors (6 Vaisala GMP 221 and 8 GMP 251) were located along the middle vertical plane, orthogonal to the injection pipe, in three layers of five sensors at the base, middle and top of the mixed PK layer. Three exhaust ports, each equipped with a CO2 sensor (GMP 251), allowed gas to flow outwards on the opposite wall of the IBC to the injection pipe, just below the FPK layer. An additional two CO2 sensors (GMP 251) were installed in the headspace of the IBC. CO2 concentrations were logged every fifteen minutes. The effluent exited the top of the pad before this was sealed, which directed the effluent to the exhaust ports. Gas was hydrated and injected, via a Bronkhorst EL-Flow Prestige mass flow controller, in three stages, at 1, 0.5, and 0.25 L min-1 to increase the CO2 capture efficiency as the reactivity declined throughout the experiment. Samples were taken in a three-by-three grid (grid spacing of 32 cm and 35 cm). Samples of the CPK were only taken along the middle vertical plane, orthogonal to the injection pipe. Samples for the mixed layer were taken at three layers, upper, middle and lower, with an additional pair of sample points being taken in between the grid points within the middle vertical plane. Thus, the same vertical plane that had the CO2 sensors was sampled extensively. Detailed experimental conditions are included in Table 4.3.  4.3.3 Analysis of Experimental Results Analytical methods are discussed briefly. Detailed methods may be found in Appendix 3. 4.3.3.1 Permeability Permeability was only assessed on the pipe experiments. Pressure sensors were used to calculate the differential pressure across the pipe length, which is needed to calculate intrinsic permeability, following a method modified from ASTM D4525 (ASTM International, 2013). The 114  pressure was recorded every 2 seconds for 50 to 100 seconds to allow for the pressure to have stabilized. Permeability measurements were made at three flow rates, with the average being the final value. 4.3.3.2 Sequestered CO2 Mass Four different methods were applied to determine the abundance of sequestered CO2, and Table 4.4 highlights the methods used for the respective experiments. Table 4.4. Applied methods to assess the magnitude of sequestered carbon in each experiment. Method Pipe-1 Pipe-2 Pad-1 Pad-2 Gas Mass Balance X X X X Bulk Sample TIC X X X X Sieved Sample TIC X X   Bulk Sample TC  X    4.3.3.2.1 Gas Mass Balance CO2 sensors allowed for the mass of injected and effluent CO2 to be determined, and this was used to perform a CO2 mass balance. The mass was determined from the injected flow rate, which was controlled by a mass flow controller. Knowing the injected and effluent CO2 mass over short intervals permitted the determination of the reactivity rate and throughout the experiment enabled a determination of the captured CO2 mass. CO2 trapped as alkalinity was not considered due to the ease of release from the aqueous phase in the mine tailings environment. Conservative dissolved inorganic carbon (DIC) concentrations of 0.1 M were assumed to discount the alkalinity, and this is equivalent to the pH of the aqueous phase being 8.8. For Pipe-1, a leak occurred after ~50 hours, reducing the injected flux. An estimate of the flow rate was established based on the reactivity at the pipe midpoint (Appendix 3). 115  4.3.3.2.2 TIC Increase Bulk samples of mixed PK from the pipe experiments were split, with one half being pulverized and micronized and the other half sieved through a 425 \u00b5m sieve before being micronized. The pad samples were not sieved; only bulk samples were prepared. The total inorganic carbon (TIC) of the carbonated bulk and sieved samples was assessed by coulometry. By assessing numerous samples, a mean carbonated TIC value was determined, and this had the mean unreacted TIC abundance subtracted from it to determine the TIC increase due to carbon mineralization. 4.3.3.2.3 TC Increase Total Carbon (TC) was performed on bulk samples from the Pipe-2 experiment using a LECO C-230 Carbon Analyzer to assess any contributions from carbon intercalated within smectite minerals. Smectites consist of silicate layers with cations and water between layers (Giesting et al., 2012). This interlayer space can lead to intercalation processes which capture ions and cause the mineral to swell (Giesting et al., 2012; Loring et al., 2012; Michels et al., 2015). CO2 has been found to intercalate without forming carbonic acid or bicarbonate ions (Loring et al., 2012). TIC as measured by coulometry involves acidifying the sample which dissolves the carbonate minerals. Intercalated CO2 which has not complexed, may not be released by acidification. TC heats the samples to release volatiles and would be able to assess this form of carbon capture. 4.3.3.3 Water Mass Balance After the experiments were completed, samples were removed, dried under atmospheric conditions, and weighed before and after drying to calculate the final moisture content. This final 116  moisture content was then compared with the initial moisture content to determine the water mass balance during the experiment. One potential avenue for moisture loss is evaporation. Evaporation was quantified from the pipe experiments using the relative humidity (RH) measurements by determining the carrying capacity of water vapour in the volume of injected and effluent gas. The difference between the influent and effluent allowed the mass of evaporated water to be determined, and this was applied to the overall water mass balance.  117  4.4 Results 4.4.1 Characterization Results The characterization of the PK samples included the mineralogy, whole-rock chemistry, TIC, surface area, grain size, and density. These data were presented in Chapter 3 but have been included below for reference in Tables 4.5, 4.6, and 4.7. Table 4.5. Mean whole-rock chemistry and standard deviations as determined by ICP-AES. Abundance (wt.%) FPK-1 CPK-1 FPK-2 CPK-2 SiO2  47.2 \u00b1 0.5 48.4 \u00b1 0.4 50.1 \u00b1 0.2 45.3 \u00b1 0.4 Al2O3 6.8 \u00b1 0.1 7.0 \u00b1 0.1 7.3 \u00b1 0.0 5.8 \u00b1 0.1 Fe2O3  6.2 \u00b1 0.1 5.9 \u00b1 0.1 6.1 \u00b1 0.1 6.7 \u00b1 0.1 CaO 2.5 \u00b1 0.0 2.4 \u00b1 0.1 3.3 \u00b1 0.0 5.0 \u00b1 0.2 MgO 22.9 \u00b1 0.4 22.6 \u00b1 0.4 20.3 \u00b1 0.1 24.0 \u00b1 0.4 Na2O 0.7 \u00b1 0.0 0.8 \u00b1 0.1 1.1 \u00b1 0.0 1.1 \u00b1 0.0 K2O 2.0 \u00b1 0.1 2.6 \u00b1 0.1 2.9 \u00b1 0.0 2.3 \u00b1 0.0 LOI 10.6 \u00b1 0.1 9.8 \u00b1 0.3 8.6 \u00b1 0.1 9.1 \u00b1 0.1  Table 4.6. TIC, surface area, grain size, and density of PK samples. Sample TIC (wt.% CO2) BET N2 Adsorption (m2\u00b7g-1) Mean Particle Size (\u00b5m) Density (g\u00b7cm-3) FPK-1 0.75 \u00b1 0.08 39.1 42 2.67 CPK-1 0.81 \u00b1 0.13 30.8 1590 2.63 FPK-2 0.71 \u00b1 0.05 - 182 2.73 CPK-2 0.53 \u00b1 0.07 - 1740 2.75 118  Table 4.7. Mean mineralogical abundance and standard deviations as determined by qXRD and TGA. Abundance (wt.%) FPK-1 CPK-1 FPK-2 CPK-2 Brucite a - - - - Serpentine b 37.7 \u00b1 9.4 30.0 \u00b1 2.0 d 27.2 \u00b1 1.5 28.1 \u00b1 1.3 d Forsterite 1.0 \u00b1 0.1 1.5 \u00b1 1.1 d 6.2 \u00b1 0.5 7.8 \u00b1 1.0 d Smectite 14.2 \u00b1 5.5 17.9 \u00b1 4.5 d 10.9 \u00b1 3.1 11.0 \u00b1 1.5 d Magnetite 1.2 \u00b1 0.1 1.1 \u00b1 0.2 d 1.2 \u00b1 0.1 1.4 \u00b1 0.3 d Diopside 7.4 \u00b1 7.0 2.9 \u00b1 0.1 d 3.7 \u00b1 0.9 6.2 \u00b1 1.1 d Phlogopite 10.0 \u00b1 3.5 11.8 \u00b1 0.9 d 11.8 \u00b1 1.2 17.7 \u00b1 1.0 d Clinochlore 6.9 \u00b1 2.7 9.7 \u00b1 2.3 d 7.9 \u00b1 1.3 5.8 \u00b1 0.5 d Calcite 0.5 \u00b1 0.5 0.9 \u00b1 0.1 d 0.9 \u00b1 0.2 0.8 \u00b1 0.2 d Quartz 2.9 \u00b1 1.1 3.8 \u00b1 1.0 d 5.1 \u00b1 0.7 2.1 \u00b1 0.5 d Albite 4.2 \u00b1 1.6 6.0 \u00b1 0.5 d 8.5 \u00b1 1.7 5.7 \u00b1 1.9 d K-Feldspar 6.6 \u00b1 2.0 9.2 \u00b1 1.5 d 12.8 \u00b1 1.3 7.1 \u00b1 0.8 d Talc 6.6 \u00b1 1.4 5.4 \u00b1 0.3 d 3.0 \u00b1 0.1 3.9 \u00b1 0.2 d Dolomite 0.1 \u00b1 0.1 tr. cd 0.8 \u00b1 0.1 0.7 \u00b1 0.2 d Cuspidine 0.6 \u00b1 0.1 tr. cd - 0.1 \u00b1 0.4 d Andradite tr. c tr. cd tr. c 1.1 \u00b1 1.3 d 119  Table 4.7 continued. Abundance (wt.%) FPK-1 CPK-1 FPK-2 CPK-2 Wollastonite - tr. cd tr. c 0.4 \u00b1 0.8 d Grossular - - - tr. Tremolite - - - tr. a Brucite detected and quantified by TGA with a detection limit of 0.3 wt.%. All other minerals were detected and quantified by qXRD. b Dominant serpentine type identified by TGA as lizardite. c Trace abundance. Peak identified in the XRD pattern but not included in the quantitative analysis or identified at an abundance < 1wt.% by QEMSCAN. d Detected by QEMSCAN. Smectite mineral identified as saponite. Other trace minerals include muscovite, biotite, illite, chromite, epidote, Fe-sulphides, melilite, Ti-oxides, apatite, zircon, and barite 120  4.4.2 Standard Proctor Compaction Standard Proctor compaction testing evaluated the relationship between moisture content and dry density under a specific compaction energy. Five compaction tests were done on the 25% FPK mixture at various moisture contents from Pipe-2, while seven tests were measured on the 33% FPK mixture from Pad-2. These data are plotted in Figure 4.4, with the data fitted to determine the maximum dry density (MDD) and the optimum moisture content (OMC), calculated as the mass of water over the mass of solids. The MDD of the 25% FPK mixture was 2.09 g\u00b7cm-3, and the OMC was 11.0 wt.%, whereas, for the 33% FPK mixture, they were 2.10 g\u00b7cm-3 and 10.5 wt.%. While these units have been used as they are the geotechnical convention, the units used throughout this thesis are bulk density and moisture content, defined as the mass of water over the total mass. For the 25% FPK mixture, the MDD is equivalent to a maximum bulk density (MBD) of 2.32 g\u00b7cm-3 and a moisture content of 9.9 wt.%. For the 33% FPK mixture, the MDD is equivalent to a MBD of 2.32 g\u00b7cm-3 and a moisture content of 9.5 wt.%. 121   Figure 4.4. Standard Proctor compaction testing results on PK blends of 25 and 33 wt.% FPK-2 with CPK-2. The maximum dry density (MDD) and the optimum moisture content (OMC) have been determined. The 90% value of the MDD for both mixes and the 60% and 100% pore saturation lines are indicated. A common criterion for compaction is based on the achieved MDD from either the Standard or Modified Proctor test. A lower limit between 90 and 95% of MDD may be used (Lara et al., 2013). This lower limit (90%) has been indicated on the plot for both PK mixtures. For the 25 and 33 wt.% FPK mixtures, it is equivalent to a bulk density range of 2.09 g\u00b7cm-3. Lastly, evaluating the moisture content relationship to dry density, the pore saturation can be evaluated at different dry densities and different moisture contents. The 100% pore saturation line (zero air voids) has been plotted as a solid line, and the 60% pore saturation has been plotted 122  as a dashed line, as 60% pore saturation has been defined as the maximum pore saturation suitable for injection (Harrison et al., 2016). In general, the distribution apex is close to saturation, and the degree of pore saturation decreases as the degree of compaction is lowered.  4.4.3 Pipe Injection Experiments 4.4.3.1 Permeability The permeabilities for the two pipe experiments in comparison to the relationship between permeability and grain size established in Chapter 3 have been plotted in Figure 4.5. For Pipe-1, the initial permeability on the compacted (1.58 g\u00b7cm-3) mixture with 22 wt.% FPK was measured to be 2.5 \u00d7 10-10 m2. The final permeability for Pipe-1 was not measured after injection due to sensor damage incurred during the experiment. Pipe-2 used a 25 wt.% FPK blend and was compacted to a bulk density of 1.87 g\u00b7cm-3. The initial permeability was 6.9 \u00d7 10-11 m2. After injection had been stopped, the final permeability was 4.9 \u00d7 10-11 m2. Comparing the initial and final permeabilities indicates a relative decline of 29%. 123   Figure 4.5. The initial permeability of the 6-metre pipe injection experiments in the context of the permeability of tailings grain size distributions ranging from coarse to fine, at 30% pore saturation. Error bars indicate the standard deviation from triplicate tests and are smaller than the symbol where not indicated. 4.4.3.2 Reaction Progression 4.4.3.2.1 Pipe-1 The results from the gas phase data for Pipe-1 have been plotted against time in Figure 4.6. Due to unforeseen circumstances, the injection of gas in Pipe-1 occurred in two stages. During phase 1 of gas injection, which lasted for the first 52 hours, a flux of 2.54 cm min-1 was injected until sufficient breakthrough occurred. During phase 2, which lasted from 52 to 140 hours, the flux was reduced due to a leak at the pipe inlet, yielding uncertainty on the injected flow rate, with the rate being estimated between 0.18 \u2013 0.42 cm min-1. The CO2 concentration at the inlet exceeded 124  9% within the first 2 hours. After 9 hours, CO2 broke through at the middle of the pipe, and the measured concentration began to rise rapidly. At 44 hours, CO2 breakthrough occurred at the outlet of the pipe. At 52 hours, the outlet concentration reached 2.5%, and the end of phase 1 was reached. At the same time, there was a power outage related to the power generator needing to be serviced, and the leak coincided with this event. During phase 2, the reduced gas flux led to gas flow no longer permeating through the pipe, and the concentration at the end of the pipe declined to zero. However, CO2 concentrations were maintained at the inlet, and at the middle of the pipe, the concentrations gradually increased. Significant diurnal effects are visible in the middle sensor CO2 concentration, relative humidity and gas temperature data for Pipe-1. Daily maximum CO2 concentrations occur in sync with daily maximum temperatures, as seen in Figure 4.6A and C. Outlet gas temperatures increase the most during the daytime, more so than for the inlet and ambient temperatures. The inlet temperatures are typically also higher than recorded for the ambient temperatures; however, all temperatures are relatively consistent at nighttime. The relative humidity (RH) of the gas inside the pipe also varied significantly with temperature, with an inverse relationship (Figure 4.6B). The difference between the injected and effluent RH is typically small (~5%), but the effluent RH (~100%) nearly always exceeds the injected RH (~95%). 125   Figure 4.6. Gas-phase results of the reaction and sequestration of simulated flue gas CO2 over time within the Pipe-1 6-metre injection experiment. A. Gas-phase CO2 concentration profiles overtime at the pipe inlet, midpoint, and outlet. A generator power outage is indicated at 52 hours, coincident with the leak that reduced the injected flowrate. B. Relative humidity (RH) at the inlet and outlet during the experiment. C. Ambient, inlet, and outlet gas temperatures over time. 126  4.4.3.2.2 Pipe-2 Gas-phase data for Pipe-2 is presented in Figure 4.7. Inlet CO2 concentrations rise to ~10% within the first hour, while breakthrough occurs at the middle and outlet sensors at 9 and 30 hours, respectively. Both middle and outlet concentrations then proceed to rise rapidly before steadying off after approximately 100 hours. A 3-hour power outage occurred in the laboratory from 96 to 99 hours and interrupted the gas injection, leading to a decrease in gas concentration for the middle and outlet sensors. However, concentrations recovered after power was restored. At 400 hours, the flow rate was reduced, and it took about 40 hours for concentrations to stabilize again. Inlet gas temperatures were slightly higher than effluent temperatures, with this being relatively consistent over the experiment. The RH was also relatively consistent throughout the experiment but differed between the inlet and outlet, with the injected RH hovering between 95 and 97%, while the effluent RH was consistently about 99 to 100%.127   Figure 4.7. Gas-phase results of the reaction and sequestration of simulated flue gas CO2 over time within the Pipe-2 6-metre injection experiment. A. Gas-phase CO2 concentration profiles over time. A building power outage is indicated at 97 hours, and the planned reduction in flowrate at 400 hours. B. Relative humidity (RH) at the inlet and outlet throughout the experiment. C. Inlet, and outlet gas temperatures over time. 4.4.3.2.3 Instantaneous Reaction Rates Gas concentrations were used to calculate the instantaneous sequestration rates by iteratively determining the difference between the inlet and outlet concentrations during a set time interval (Figure 4.8). For Pipe-1, only the carbonation rate of the first half is shown, as it had 128  consistent flow throughout the experimental duration. Through the first ~50 hours, the two pipe experiments perform similarly when accounting for the difference in the flow rate to sample mass ratio. Both experimental rates then decline. For Pipe-1, the data is messy due to the large diurnal fluctuations. Further, the leak and resultant decrease in injection rate and the uncertainty around the flow rate complicates the interpretation of the data. For Pipe-2, the reaction rate behaves asymptotically, decreasing rapidly at first and then progressively slowing. At 400 hours, the change in flow rate is accompanied by a sharp drop in the reaction rate before the rate recovered.  Figure 4.8. Instantaneous injection (dashed and dash-dotted lines) and carbonation (solid lines) rates for both Pipe-1 and Pipe-2 injection experiments due to carbon mineralization and solubility trapping. The two dash-dotted lines and the two lines with squares indicate the injection and reaction uncertainty during phase 2 injection for the Pipe-1 experiment.129  4.4.3.3 Sequestered Carbon 4.4.3.3.1 Pipe-1 4.4.3.3.1.1 Bulk Sample Total Inorganic Carbon Samples from Pipe-1 were collected for TIC analysis from three horizons at 11 locations, 50 cm apart along the pipe length, totalling 33 samples. Duplicate and some triplicate TIC analyses were performed to improve precision. Figure 4.9A compares a histogram of the TIC content of the 33 carbonated samples with the results of 29 analyses of unreacted CPK and FPK. The unreacted average was weighted for the relative abundances of CPK and FPK to enable clearer comparisons between the data sets. The unreacted CPK (0.81 \u00b1 0.13 wt.% CO2) has more carbon content but not significantly more than the unreacted FPK (0.75 \u00b1 0.08 wt.% CO2). From examining the data distribution, it is clear that the unreacted material has a normal distribution with several outliers with higher carbon content. The carbonated samples display a broader normal distribution, which, again, has a few outliers. This difference between the two distributions demonstrates a clear shift in carbon content, consistent with the sequestration of mineralized carbon. As a result of the mineralized carbon, the mean TIC content increased from 0.80 to 0.90 wt.% CO2. A one-tailed t-test indicates that this difference is significant to the 99.9% confidence level. The captured CO2 was equal to 0.10 \u00b1 0.05 wt.% with a 95% confidence interval (CI). This increase is equivalent to capturing 244 g of CO2 within 247 kg of PK. 0.5 wt.% of the total MgO and 1.8 wt.% of the lizardite Mg (serpentine type identified by TGA) from the fines of the CPK-1 and FPK-1 is required to sequester this amount of carbon at a 5:4 Mg to CO2 ratio. Detailed TIC results are presented in Table 4.8. 130   Figure 4.9. Histograms showing the distribution of TIC measurements from unreacted and reacted samples for bulk (A) and sieved (B) samples from Pipe-1. The solid lines show the average carbonated and unreacted values, while the dashed lines show the mean value in the unreacted FPK and CPK samples. 131  Table 4.8. Sequestered carbon in the 6-metre pipe injection experiments as determined by TIC and a gas phase mass balance. Experiment Initial TIC (wt.% CO2) Final TIC  (wt.% CO2) Sequestered Carbon      (wt.% CO2) MgO Reacted (wt.%)a <425 \u00b5m Lizardite Reacted (wt.%)a Pipe-1 Bulk Sample 0.80 \u00b1 0.10b 0.90 \u00b1 0.09b 0.10 \u00b1 0.05c 0.5 1.8 Pipe-1 Sieved Sampled 0.96 \u00b1 0.07b 1.24 \u00b1 0.04b 0.11 \u00b1 0.01ce 0.6 1.9 Pipe-1 Mass Balance   0.14 \u2013 0.16 0.7 \u2013 0.8 2.5 \u2013 2.8 Pipe-2 Bulk Sample 0.59 \u00b1 0.04b 0.68 \u00b1 0.05b 0.09 \u00b1 0.02c 0.4 2.4 Pipe-2 Sieved Sampled 0.68 \u00b1 0.03b 0.99 \u00b1 0.07b 0.12 \u00b1 0.01ce 0.6 3.1 Pipe-2 Mass Balance   0.23 1.1 5.9 a Assumes the precipitation of hydromagnesite. b One standard deviation. c 95% confidence interval. d Measured on the sample fraction below 425 \u00b5m. e TIC increase from the sieved fraction attributed to the whole sample. 132  4.4.3.3.1.2 Sieved Fines Total Inorganic Carbon Bulk grab samples were sieved below 425 \u00b5m to improve the sequestered carbon mass assessment by TIC. Figure 4.9B shows the distribution of the unreacted versus reacted sieved TIC values. The unreacted samples differ from the bulk samples as the fines of the CPK were sieved out and sampled to determine their TIC (1.22 \u00b1 0.13 wt.% CO2), which was higher than the bulk CPK or the FPK. The result is a bimodal distribution, with the FPK at the low end and the CPK samples at the high end. The mean unreacted TIC value is in the middle of these two extremes and was determined in proportion to the expected abundance of the FPK and CPK fines. The sieved samples of the reacted material yield a normal distribution at higher carbon contents than the unreacted average and much higher than observed for the bulk samples. The increase in carbon observed for the sieved fraction needs to be attributed to the total mass of material used. The sieved fines thus yielded a TIC increase of 0.11 \u00b1 0.01 wt.% CO2 (Table 4.8). This increase is 110% of what the bulk sampling determined and is within the error of that determination. This increase indicates that 271 g of CO2 was captured. Sequestering this carbon would require 0.6 wt.% of the total MgO or 1.9 wt.% of the lizardite Mg from FPK-1 and the CPK-1 fines to be accessed. 4.4.3.3.1.3 Gas Mass Balance A mass balance on the injected gas was also used to determine the amount of captured carbon. The mass balance was accomplished by iteratively examining the volume of injected CO2 and the volume of effluent CO2 over a given time interval. The amount of water held as DIC was discounted due to the instability of this form of carbon capture by assuming a DIC concentration in the remaining water. The mass balance estimated that the amount of captured carbon was between 0.14 to 0.16 wt.% CO2, equivalent to the sequestration of 333 to 385 g CO2 (Figure 4.10). This increase is equivalent to 0.7 \u2013 0.8 wt.% of the MgO or 2.5 \u2013 2.8 wt.% of the Mg from fine-133  grained lizardite being leached. There is uncertainty due to the gas leak, which reduced the flow rate during phase 2 of injection. While the gas phase estimate is within the confidence interval of the bulk sample TIC increase, it does overestimate the value determined by the bulk sample TIC increase and is outside of the interval found for the sieved sample TIC increase.  Figure 4.10. Gas and solid-phase results of the reaction and sequestration of simulated flue gas CO2 over time within the 6-metre pipe injection experiments. Sequestered CO2 mass over time was quantified from gas concentrations discounting the expected DIC (solid and dashed lines) and TIC increases in carbonated versus initial bulk (diamonds) and sieved fine fraction (squares) samples, with a 95% confidence interval indicated. Pipe-1 has uncertainty around the gas phase estimate due to a leak that reduced the flow rate.134  4.4.3.3.1.4 Reaction Homogeneity At the larger scale, heterogeneous reactivity leads to overall lower degrees of sequestration. To evaluate the captured carbon homogeneity, the TIC data for the bulk and sieved samples are presented against the pipe length in Figure 4.11. Symbols indicate the horizon from which the samples were taken. There is no apparent trend for the different horizons. For the bulk samples, there is also no clear trend with distance. There is slightly more reaction at the beginning of the pipe for the sieved samples than towards the end, consistent with the PK being in contact with higher CO2 concentrations. The sieved samples have markedly less heterogeneity around the carbonated TIC values than the bulk samples.135   Figure 4.11. TIC for each reacted sample along the length of Pipe-1 at upper (triangle), middle (square) and lower (circle) horizons. The solid and dashed lines show the mean unreacted and carbonated TIC values for the bulk and sieved samples, respectively.136  4.4.3.3.2 Pipe-2 4.4.3.3.2.1 Bulk Sample Total Inorganic Carbon Thirty-nine samples were collected from three horizons at 13 locations, 50 cm apart, along the length of the pipe. Duplicate or triplicate TIC analyses were performed on each sample. Histograms of the TIC values for the 30 unreacted and 39 carbonated samples are displayed in Figure 4.12A. The unreacted mean was weighted for the respective abundances of CPK and FPK, while samples of unreacted mixed PK contributed based on the number of samples. The unreacted data distribution is bimodal due to the TIC of the FPK (0.71 \u00b1 0.05 wt.% CO2) being higher than the CPK samples (0.53 \u00b1 0.07 wt.% CO2). The unreacted mixed PK samples are in the middle, with the overall mean being slightly lower than the mixed samples. Meanwhile, the carbonated samples are normally distributed with a few high outliers. The TIC of the carbonated samples overlaps significantly with the TIC of the pure FPK samples though there is a clear shift from the TIC of the CPK, which makes up most of the mixture, as shown by the difference in the unreacted and reacted means. The analyses show an increase of 0.09 \u00b1 0.02 wt.% CO2 from unreacted (0.59 wt.% CO2) to reacted (0.68 wt.% CO2) material, with a 95% CI. This increase is statistically significant to a 99.9% confidence level and is equal to the sequestration of 187 g of CO2 within 220 kg of PK. Attributing this reactivity to the abundance of fine lizardite Mg in the CPK-2 and FPK-2, 2.4 wt.% of the Mg was leached. Alternatively, this is equivalent to 0.4 wt.% of the total MgO. Detailed results are shown in Table 4.8. 137   Figure 4.12. Histograms showing the distribution of TIC measurements from unreacted and reacted samples for bulk (A) and sieved (B) samples from Pipe-2. The solid lines show the average carbonated and unreacted values, while the dashed lines show the mean value in the unreacted FPK and CPK samples.138  4.4.3.3.2.2 Sieved Fines Total Inorganic Carbon For the sieved fines, the data for the Pipe-2 samples was clearer to interpret than for Pipe-1 as the TIC of the CPK-2 fines (0.64 \u00b1 0.09 wt.% CO2) was not significantly different from the TIC of the FPK-2 or the sieved mixed PK samples (0.66 \u00b1 0.05 wt.% CO2). Therefore, the distribution shown in Figure 4.12B is normally distributed for both the unreacted and carbonated sample sets. A significant increase in the TIC is observed with nearly no overlap between the sample groups, and this increase is much more significant than observed for the bulk samples. Attributing this increase to the total sample mass yields a TIC increase of 0.12 \u00b1 0.01 wt.% CO2, equivalent to the capture of 264 g of CO2. This increase is 133% higher than found for the bulk sample analysis and outside its confidence interval. If this sequestered CO2 amount is related to the fine lizardite abundance in the CPK-2 and FPK-2, 3.1 wt.% of the lizardite Mg must have been leached. This abundance is also equal to 0.6 wt.% of the total MgO being accessed. 4.4.3.3.2.3 Bulk Sample Total Carbon A subset of unreacted (12) and carbonated (20) samples were analyzed for total carbon to examine whether CO2 was being intercalated within smectite minerals, as TIC would not account for this form of carbon capture. TC was compared against the TIC from the same samples for both unreacted and carbonated samples to see if there was any further increase. Beyond the initial TIC the unreacted samples had an additional 0.07 \u00b1 0.04 wt.% CO2. For the reacted samples, an additional 0.07 \u00b1 0.03 wt.% CO2 was present. These values are effectively the same, with the smaller standard deviation in the carbonated samples being attributable to the larger sample size. Thus, statistically, zero CO2 was captured by smectite intercalation.139  4.4.3.3.2.4 Gas Mass Balance The gas mass balance enabled the determination of the captured CO2 over time, as presented in Figure 4.10 where it is compared to the TIC results. For Pipe-2, the captured carbon was equal to 0.23 wt.% CO2, which is nearly double the value determined from the sieved sample TIC results. This increase is equivalent to the sequestration of 499 g CO2, requiring 1.1 wt.% of the MgO or 5.9 wt.% of the Mg from the fine-grained lizardite to react. The mass balance determined a value higher than found for Pipe-1 and shows a similar trend in the early time results. The TIC increases for both Pipe-1 (0.10 and 0.11 wt.% CO2) and Pipe-2 (0.9 and 0.12 wt.% CO2) are similar. 4.4.3.3.2.5 Reaction Homogeneity The reaction heterogeneity was evaluated by plotting the samples against the pipe length, as presented in Figure 4.13. The different horizons lack any preferential TIC increase, and for the bulk samples, there is no trend of TIC with distance. The sieved samples show slightly higher than average captured carbon at the inlet and slightly less than average at the outlet. Heterogeneity for both sets of samples is approximately the same.140   Figure 4.13. TIC for each sample along the length of Pipe-2 at upper (triangle), middle (square) and lower (circle) horizons. The solid and dashed lines show the mean unreacted and carbonated TIC values for the bulk and sieved samples, respectively.141  4.4.3.4 Water Mass Balance Similar to how the CO2 concentrations were used to perform a mass balance upon the captured CO2, the humidity and temperature measurements were used to determine the mass of water lost to evaporation. This mass balance was accomplished by determining the gas carrying capacity of water vapour into and out of the pipes. Pipe-1 lost 5.9 to 6.5 g of water, with uncertainty on the amount of injected gas due to the leak. Pipe-2 lost 1.7 g of water. These values are insignificant at the kg-scale of the experiments. Pipe-1 had a measured initial moisture content of 6.5 \u00b1 0.6 wt.%, while a final moisture content was measured to be 5.2 \u00b1 0.6 wt.%, though this latter value was only measured on nine samples. Therefore, some moisture loss is unaccounted for in Pipe-1. A likely explanation is that since the final moisture content was measured after samples were shipped from the Northwest Territories to UBC, the samples may have been imperfectly sealed and dried during transit. Pipe-2 had a final measured moisture content of 6.9 \u00b1 0.3 wt.%, versus the measured 6.7 \u00b1 0.4 wt.% on the initial material, meaning no significant change was observed.  4.4.4 Pad Injection Experiments 4.4.4.1 Reaction Progression 4.4.4.1.1 Pad-1 In the Pad-1 experiment, CO2 was injected into the middle of the 12 cm thick mixed PK layer. During the first 70 hours, gas could exit out the top of the pad due to desiccation cracking of the upper FPK cap layer. CO2 concentrations in the headspace progressively rose until approaching 3%. At 70 hours, the top of the IBC was sealed, preventing gas outflow from the top, and this directed gas through the intended ports. Figure 4.14 presents the interpolated CO2 142  concentrations throughout the mixed PK layer. This interpolation was done using the recorded measurements at different time intervals while also providing some boundary constraints. At 1 hour, injection had just begun, with the plume reaching the first sensor. At 90 hours, the plume was linearly directed to the open port 3 with limited lateral progression. Upon opening port 5 at 100 hours, the gas was quickly directed to the right side, with effluent concentrations approaching 6%. Next, with port 2 opened, gas was directed over to the left side though the sensors detected no clear flow path. At 120 hours, port 1 was opened, and again, gas was directed over to the nearby port, with concentrations approaching 5%. Finally, at 140 hours, when port 4 was open, a flow path developed through the middle to the port. Limited plume development occurred in the distal half of the mixed PK layer.143   Figure 4.14. Aerial view of the interpolated CO2 concentrations throughout the mixed PK layer in the Pad-1 injection experiment measured by six embedded sensors (X\u2019s), with the injection (triangle) and effluent (squares or circle) ports shown. 144  4.4.4.1.2 Pad-2 For the Pad-2 experiment, the CO2 concentrations were interpolated from the recorded measurements of 15 embedded sensors to present a plume development. In Figure 4.15, a profile through the base CPK and the mixed PK layer is shown. Gas was injected through a perforated pipe in the lower-left corner to an outlet port in the upper right corner below the cap FPK layer, with limited plume development by 1 hour. Again, gas could exit the top via the headspace, with CO2 being detected here first, at 11 hours. By 50 hours, this remained the only location for effluent detection. At 70 hours, CO2 was detected at the exit ports, with concentrations rising rapidly. At 100 hours, concentrations in the headspace were approximately 4%, whereas concentrations at the effluent ports were ~2%. At 163 hours, phase 1 of injection was ended, and the flow rate was reduced for phase 2. At the same time, the headspace was sealed up to direct gas through to the effluent ports. At 200 hours, the plume showed the same trend, with effluent concentrations having exceeded 4%. By 300 hours, further plume progression had occurred with effluent concentrations at 5%. At 440 hours, the flow rate was reduced for phase 3. Finally, at the end of the experiment (750 hours), the plume had progressed extensively throughout the layers.145   Figure 4.15. Cross-section view of the interpolated CO2 concentration profiles in Pad-2 measured by 15 embedded CO2 sensors (X\u2019s) in the base CPK (indicated by a dashed black line) and the mixed PK layer throughout the experiment. The perforated injection pipe (triangle) and the effluent port (circle) are also indicated. 0 20 40 60 80Length (cm)02040Height (cm)A. CO2 Plume at 1 Hour0510[CO 2] (%)0 20 40 60 80Length (cm)02040Height (cm)B. CO2 Plume at 50 Hours0510[CO 2] (%)0 20 40 60 80Length (cm)02040Height (cm)C. CO2 Plume at 100 Hours0510[CO 2] (%)0 20 40 60 80Length (cm)02040Height (cm)D. CO2 Plume at 200 Hours0510[CO 2] (%)0 20 40 60 80Length (cm)02040Height (cm)E. CO2 Plume at 300 Hours0510[CO 2] (%)0 20 40 60 80Length (cm)02040Height (cm)F. CO2 Plume at 750 HoursCO2 SensorTop of Base CPKInjection PointOutlet0510[CO 2] (%)146  4.4.4.2 Sequestered Carbon 4.4.4.2.1 Pad-1 4.4.4.2.1.1 Bulk Sample Total Inorganic Carbon Figure 4.16A presents the measured TIC values for the reacted mixed PK layer as a histogram compared to the unreacted sample values for Pad-1. The unreacted material shows a normal distribution with a few high-value outliers and similar FPK (0.75 \u00b1 0.08 wt.% CO2) and CPK (0.81 \u00b1 0.13 wt.% CO2) averages. The reacted mixed PK samples (mean of 0.93 \u00b1 0.10 wt.% CO2) are normally distributed with a broader span that overlaps unreacted samples with high amounts of carbon. There is a clear shift from the unreacted to reacted samples of 0.13 \u00b1 0.06 wt.% CO2, sequestering 220 g of CO2. This reactivity requires 0.7 wt.% of the total MgO or 2.3 wt.% of the fine-grained lizardite to have reacted.147   Figure 4.16. Histograms showing the distribution of TIC measurements from unreacted and reacted samples for Pad-1 (A) and Pad-2 (B). The solid lines show the average carbonated and unreacted values, while the dashed lines show the mean value in the unreacted FPK, CPK and mixed PK samples.148  4.4.4.2.1.2 Gas Mass Balance Figure 4.17 summarizes the results for the solid and gas phases. The mass balance presenting the evolution of captured CO2 begins slowly, and the rate of capture increases with the higher flow rate after 20 hours. The rate of capture then progressively slows throughout the experiment. When considering the mixed PK layer to be the source of reaction, the gas phase estimates that 0.09 wt.% CO2 was captured, equivalent to 151 g of CO2. This reactivity demands 0.5 wt.% of the total MgO or 1.6 wt.% of the fine-grained lizardite to have been leached. This increase is within the confidence interval of the TIC increase measured for the mixed PK layer (0.13 \u00b1 0.06 wt.% CO2). However, the FPK layers showed equivalent (lower FPK, 0.13 \u00b1 0.06 wt.% CO2) or higher reactivity (upper FPK, 0.15 \u00b1 0.07 wt.% CO2). Overall, the TIC for the mixed PK and FPK layers would indicate a higher amount of captured carbon than injected. Table 4.9 presents a summary of the gas phase and TIC results.149   Figure 4.17. Gas and solid-phase results of the reaction and sequestration of simulated flue gas CO2 over time within the pad injection experiments. Sequestered CO2 mass over time was quantified from gas concentrations discounting the expected DIC (solid lines) and from TIC increases in carbonated versus initial samples (circles, squares, diamonds and triangles), with a 95% confidence interval indicated.150  Table 4.9. Sequestered carbon in the pad injection experiments as determined by TIC and a gas-phase mass balance. Experiment Initial TIC (wt.% CO2) Final TIC  (wt.% CO2) Sequestered Carbon (wt.% CO2) MgO Reacted     (wt.%) <425\u00b5m Lizardite Reacted (wt.%)a Pad-1 Mass Balance   0.09 0.5 1.6 Pad-1 Mixed PK TIC 0.80 \u00b1 0.10 b 0.93 \u00b1 0.10 b 0.13 \u00b1 0.06 c 0.7 2.3 Pad-1 Cap FPK TIC 0.75 \u00b1 0.08 0.90 \u00b1 0.10 0.15 \u00b1 0.07   Pipe-1 Base FPK TIC 0.75 \u00b1 0.08 0.90 \u00b1 0.05 0.13 \u00b1 0.06   Pad-2 Mass Balance   0.20 1.0 4.4 Pad-2 Mean TIC 0.58 \u00b1 0.04 b 0.68 \u00b1 0.03 b 0.10 \u00b1 0.02 c 0.5 2.2 Pad-2 Mixed PK TIC 0.59 \u00b1 0.04 0.72 \u00b1 0.05 0.14 \u00b1 0.02   Pad-2 Cap FPK TIC 0.71 \u00b1 0.05 0.73 \u00b1 0.04 0.02 \u00b1 0.04   Pad-2 Cap\/Base CPK TIC 0.53 \u00b1 0.07 0.60 \u00b1 0.04 0.07 \u00b1 0.06   a Assumes the precipitation of hydromagnesite. b One standard deviation. c 95% confidence interval. 151  4.4.4.2.1.3 Reaction Homogeneity The distribution of the sampled TIC values is presented in Figure 4.18 as a cross-section through the pad and an aerial view of the mixed PK layer. The first plot shows the general decrease in the TIC increase from the injection point with distance. However, it also shows higher TIC in the confining FPK layers than expected. The aerial view of the mixed layer supports the trend seen in the profile view. Overall, the reaction was relatively evenly distributed with few dead zones. 152   Figure 4.18. TIC increase for each sample (X\u2019s) or sample group as distributed within Pad-1. A. Cross-section view through the middle of the pad, showing the upper FPK, mixed PK and lower FPK layers. B. Aerial view of the mixed PK layer.153  4.4.4.2.2 Pad-2 4.4.4.2.2.1 Bulk Sample Total Inorganic Carbon Figure 4.16B shows the distribution of the sample values as a histogram of the reacted mixed PK samples and sampled unreacted material for Pad-2. While the unreacted material has one prominent normally distributed peak, there are two other modal peaks due to the FPK-2 (0.71 \u00b1 0.05 wt.% CO2) being significantly higher in starting TIC content than the unreacted CPK-2 (0.53 \u00b1 0.07 wt.% CO2) or the mixed PK samples (0.58 \u00b1 0.04 wt.% CO2). Meanwhile, the reacted samples show a normally distributed form that overlaps with the starting FPK but is significantly higher than observed for the bulk of the unreacted samples. This increase for the mixed PK layer was found to be 0.14 \u00b1 0.02 wt.% CO2 (from 0.59 \u00b1 0.04 to 0.72 \u00b1 0.05 wt.% CO2). In the cap FPK and CPK layers, the reaction was insignificant, while there was significant reaction in the base CPK layer due to its close contact with the injected gas. A summary of all the initial, final and TIC increases for the respective layers is presented in Table 4.9. As a whole, Pad-2 yielded a mean TIC increase of 0.10 \u00b1 0.02 wt.% CO2, capturing 975 g of CO2. The amount of Mg needed to sequester this carbon at a 5:4 MgO to CO2 ratio is 0.5 wt.% of the total MgO and 2.2 wt.% of the fine-grained lizardite. 4.4.4.2.2.2 Gas Mass Balance Figure 4.17 shows the gas phase estimate and the increase in the TIC for all the layers. The gas-phase estimate is double (0.20 wt.% CO2) the overall mean TIC increase (0.10 wt.% CO2). The mass balance indicates that 1.98 kg of CO2 was sequestered, requiring 1.0 wt.% of the MgO or 4.4 wt.% of the Mg from the fine-grained lizardite. Three precise slope intervals refer to the three injection phases and their flow rates for the gas phase estimate. Due to the longer injection 154  time and the reactivity being consumed, the sequestration rate decreases over time as the slope of the line decreases. 4.4.4.2.2.3 Reaction Homogeneity For Pad-2, Figure 4.19 shows the distribution of the sampled TIC values throughout the mixed PK layer and the base and cap layers. The cap layers show little to no reaction, whereas the base CPK layer shows the most reactivity closest to the injection point. The mixed PK layer also shows a higher degree of reaction closer to the injection pipe, consistent with the higher concentrations near the injection point, as presented in Figure 4.15. In general, the mixed PK layer is evenly reacted except for the last column of samples, which shows less reaction.155   Figure 4.19. TIC increases for each sample (X\u2019s) or sample group distributed within Pad-2 presented as a cross-section view through the middle of the pad, showing the cap CPK and FPK layers, the mixed PK layer and the base CPK layer. 156  4.4.4.3 Moisture Content As numerous layers of materials were stacked together in varying moisture content states, capillary action and drainage led to some water movement during the experiments (Table 4.10). In Pad-1, the FPK layers had high moisture contents (28.7 \u2013 34.0 wt.% water) and were observed to lose 4.0 \u2013 10.6 wt.% water. Unfortunately, the initial moisture contents were measured on a small sample size, which does not allow for the potential heterogeneity of the initial moisture content to be evaluated. The mixed PK layer was observed to increase in moisture content by 3.0 wt.% to 9.6 wt.% water. For Pad-2, smaller changes were observed. The moisture content of the FPK decreased the most, from 21.8 wt.% to 18.9 wt.% water. The base CPK increased the most, gaining 2.2 wt.% water throughout the experiment to a final moisture content of 3.5 wt.% water. The CPK cap layer and the top of the mixed PK layer both gained 0.6 wt.% water to 1.8 and 7.4 wt.% water, respectively. The middle of the PK layer remained unchanged, while the bottom of the mixed PK layer lost 0.5 wt.% water to a final moisture content of 6.3 wt.%.157  Table 4.10. Moisture contents of the pad injection experiments before and after injection. Experiment \/ Layer Initial Moisture Contenta (wt.% H2O) Final Moisture Contenta (wt.% H2O) \u2206 Moisture Content (wt.% H2O) Pad-1 FPK Cap 28.7 24.7 \u00b1 1.8 b -4.0 Pad-1 Mixed PK 6.6 9.6 \u00b1 0.4 +3.0 Pipe-1 FPK Base 34.0 23.4 \u00b1 0.8 -10.6 Pad-2 CPK Cap 1.2 1.8 \u00b1 0.2 +0.6 Pad-2 FPK Cap 21.8 \u00b1 2.5 18.9 \u00b1 0.3 -2.9 Pad-2 Mixed PK Top 6.8 \u00b1 0.4 7.4 \u00b1 0.5 +0.6 Pad-2 Mixed PK Middle 6.8 \u00b1 0.4 6.7 \u00b1 0.2 -0.1 Pad-2 Mixed PK Bottom 6.8 \u00b1 0.4 6.3 \u00b1 0.5 -0.5 Pad-2 CPK Base 1.3 3.5 \u00b1 0.2 +2.2 a Moisture content measured as the mass of water over the total sample mass. b One standard deviation. 158  4.5 Discussion 4.5.1 Physical Properties and CO2 Injection The physical properties of tailings play a significant role in carbon mineralization. Principally, the grain size distribution, moisture content, and degree of compaction either impact the sequestration reaction or the pneumatic permeability (Assima et al., 2013a, 2014b; Harrison et al., 2015, 2016; Lechat et al., 2016). Therefore, the relationship between the degree of compaction and the moisture content was evaluated by Standard Proctor compaction testing, while the permeability measurements on the pipe experiments enabled an analysis of the relationship between compaction and permeability. Compaction testing was conducted as the degree of pore saturation is an important parameter governing the distribution of injected gas into porous media (Assima et al., 2013a; Harrison et al., 2015, 2016), and pore saturation is a function of the moisture content and the degree of compaction. From examining Figure 4.4, the relationship between the dry density and the moisture content of the samples follows an inverted parabola. Adding water to dry tailings increases the degree of compaction because grains can slip past each other and pack closer together. As the degree of pore saturation increases, however, there is an inflection point where the pore water begins to resist compaction due to its incompressibility. Both PK mixtures behave similarly to typical silty sands (Budhu, 2011). The degree of compaction is also a critical criterion for tailings management. Compaction increases the strength and stability of the tailings and is applied to structural zones in dry stack tailing storage facilities (Klohn Crippen Berger, 2017; Lupo & Hall, 2010; Ulrich & Coffin, 2013). The acceptable range of moisture content for a required degree of compaction can be determined from this dry density-moisture content relationship. However, the degree of pore saturation is a 159  critical criterion for injecting CO2. The degree of pore saturation was determined from the materials\u2019 specific gravity and is presented in Figure 4.4 as the 60% saturation line. For mediums to be acceptable for injection, they must remain below or to the left of this line (Harrison et al., 2016). Therefore, decreasing the degree of compaction decreases the degree of pore saturation for a given moisture content. The permeabilities measured on both pipe experiments were higher than expected from the relationship established in Chapter 3 for the given abundance of fines alone (Figure 4.5). However, when the degree of compaction is considered, the permeability results are reasonable. Pipe-1 was compacted to a bulk density of 1.58 g\u00b7cm-3 (69% of MBD). In contrast, the 22.5% fines sample from Chapter 3 was compacted to a bulk density of 1.99 g\u00b7cm-3 (87% of MBD). Pipe-2 was compacted to a bulk density of 1.87 g\u00b7cm-3 (81% of MBD) versus 2.01 g\u00b7cm-3 (87% of MBD) for the comparable centimetre-scale experiment. Lowering the degree of compaction can increase the permeability by half to one order of magnitude. A similar effect was found for the PS experiments in Chapter 3, where the loosely compacted samples were also an order of magnitude more permeable than their compacted counterparts. The effect of carbon mineralization on permeability was also studied at the metre scale. The permeability of Pipe-2 decreased by 29% relative throughout the experiment. This decrease is small as far as permeabilities go, as they can vary by orders of magnitude. This accords with the limited change in permeability observed to occur with carbon mineralization reactions in Chapter 3. 160  4.5.2 6-Metre Pipe Injection Performance The 6-metre pipe injection experiments served primarily to investigate any changes that occurred from the ~400-fold scale increase from the centimetre-scale experiments of Chapter 3. Pipe-2 served to replicate the results from Pipe-1, to achieve a longer injection duration and to reduce the diurnal environmental variations observed for Pipe-1 in the field. CO2 concentrations rose rapidly at the inlet but were delayed from rising in the middle and effluent CO2 sensors due to pore gas being flushed, equilibrium being reached with the aqueous phase and carbon mineralization in the PK. Middle and effluent CO2 concentrations were observed to increase rapidly upon CO2 arrival. This rapid breakthrough indicates that the rate of CO2 supply exceeded the reactivity significantly, as observed by the decrease in initial reaction rates in Figure 4.8. Concentrations progressively stabilize as the carbon mineralization rate proves consistent. Once CO2 concentrations stabilized, diurnal fluctuations became more apparent and were most prominently observed in the Pipe-1 middle and outlet sensors. Thus, the instability of CO2 storage as alkalinity due to variable temperature conditions was directly observed (Figure 4.6). The large diurnal variations occur due to the ~20\u00b0C temperature swings at the Gahcho Ku\u00e9 Diamond Mine from overnight ambient lows to solar irradiance heating the black pipe in the daytime. In Pipe-2, more minor fluctuations in the CO2 concentrations are visible due to the lesser observed diurnal temperature fluctuations indoors (Figure 4.7C). The pressurized gas cylinder buffered the ambient temperature change of the injected gas. As temperature increases, the solubility of CO2 in water decreases, meaning that CO2 leaves the aqueous phase and joins the gas phase. This effect is significant because it indicates that the aqueous phase is closer to being in equilibrium with the gas phase than the degree of variation from the temperature change, which implies that the rate of CO2 supply is not rate-limiting (Harrison et al., 2016). 161  Temperature changes also have an impact on RH. As temperature increases, the carrying capacity of water vapour in the air increases and the RH decreases as the capacity increases beyond the amount of vapour in the air. Of concern is that evaporation could inhibit Mg-carbonate formation if it reduces local water contents substantially (Assima et al., 2012, 2013a; Harrison et al., 2015, 2016). However, in these experiments, while evaporation did occur, it was insignificant in magnitude. At ~50 hours, the injected flow rate into Pipe-1 was reduced due to the leak. This accident simulated a staged injection rate, where the injection rate is turned down as the PK reactivity decreases to reduce the amount of CO2 in the effluent. A two-stage approach is suitable, given the bulk and residual reactivity stages identified in Chapter 3. In Pipe-2, the second injection stage was only instigated towards the end of the experiment to simplify interpreting the results. As shown in Figure 4.7, there is a transitionary period between when the flow rate is changed and when the recorded concentrations reach equilibrium. For Pipe-1, the leak\u2019s timing was happenstance in that it occurred just as effluent was beginning to exit the pipe, resulting in Pipe-1 being very efficient in terms of the mass of CO2 captured versus the mass injected. The reduced flow rate in Pipe-1 occurred shortly after the breakthrough at the effluent sensor. While CO2 infiltrated through the whole pipe, significant CO2 in the effluent was only observed for a few hours. For Pipe-2, extensive CO2 was directed throughout the pipe as the effluent sensor reached a final concentration above 9%. This concentration distribution would suggest a well-distributed reaction in the first half of Pipe-1 and throughout Pipe-2. While the concentrations were maintained at the midpoint of Pipe-1, the gas phase data does not indicate to what degree the second half of the pipe experienced carbonation. However, the TIC data supports a well-distributed reaction, regardless of the vertical horizon or distance along either pipe. Sieved 162  TIC samples towards the end of both pipes do have slightly less carbon than on average. To reconcile the gas and solid phase results, the nature of the reaction fronts must be considered. Reaction fronts have been plotted for Pipe-2 at three points in time in Figure 4.20 to typify the expected reactivity for Pipe-1. As CO2 is injected, dispersion acts to smoothen the concentration front. However, during the initial reactivity stage, reactivity counters the dispersion and the progression of the front through the tailings. Further, observed concentrations rise and change rapidly (Figure 4.7), indicating the reaction fronts are sharp, meaning they change quickly over short amounts of distance or time (Figure 4.20; 30 hours). This sharp nature indicates that even though there was no CO2 in the effluent of Pipe-1, the high concentrations observed at the midpoint may have continued to flow into the second half of the Pipe-1. This would explain the homogenous reaction observed from the TIC sampling. The sieved TIC results observed lower reactivity in the second half, which may result from lower pCO2 in the gas phase than at the inlet, rather than from a lack of CO2 exposure. Over time, as the reactivity declines, the concentration front broadens out, which is seen when middle and effluent concentrations increase and approach the concentrations at the inlet (Figure 4.7), as occurs in the residual reactivity stage (Figure 4.20; 300 hours). Decreasing the injection rate reverses this trend, turning the reaction fronts back toward those observed in the beginning, when reactivity was fresh (Figure 4.20; 450 hours). 163   Figure 4.20. Estimated CO2 concentration fronts through Pipe-2 at 30, 300 and 450 hours of injection. Reaction stages were previously defined in Chapter 3: initial reactivity during which reaction rates prevent the breakthrough of CO2, bulk reactivity where the reactivity rate and the effluent concentration decrease and increase, respectively and the residual reactivity where rates stabilize and lead to effluent concentrations reaching steady state. Initial reactivity is visible through 10 and 30 hours for Pipe-1 and Pipe-2, respectively (Figure 4.8). Notably, the rate decline in Pipe-1 during the bulk reactivity stage parallels the slope of that of Pipe-2. Residual reactivity is not observed in Pipe-1. After the injection rate change for Pipe-2, the reaction rate continues to decline, and the stage of residual reactivity is not reached (Figure 4.8). In Chapter 3, experiment PK-E3 was performed on CPK-2 and FPK-2 at the same blend ratio as Pipe-2. For PK-E3, the residual stage was only observed after 400 hours. To directly compare the pipe experiment 164  reactivity to the centimetre-scale experiment PK-E3, the reaction rates normalized to mass have been plotted in Figure 4.21. Pipe-1 and Pipe-2 show similar reactivity through 50 hours and differ from PK-E3 as the relative supply rate for the 6-metre pipes was higher than for PK-E3. After 50 hours, the rates of Pipe-2 and PK-E3 are consistent and follow the same trend as the reactivity declines and more CO2 exits in the effluent. Through 400 hours, the reaction rates of Pipe-2 and PK-E3 are largely similar, with the residual reactivity of PK-E3 being shown to continue. Initial and bulk reactivity stages are consistent across experimental scales. PK-E3 captured 0.23 wt.% CO2 by mass balance with an injection duration of 685 hours, compared to the 0.20 wt.% CO2 accounted for by mass balance in the Pipe-2 experiment after 475 hours.  Figure 4.21. Injection (dashed lines) and carbonation rates (solid lines) for both the 6-metre-long injection experiments Pipe-1 (first half during injection phase 1) and Pipe-2 (injection phase 1) and the 12-centimetre long injection experiment PK-E3 from Chapter 3. 165  4.5.3 Pad Injection Performance The pad injection experiments aimed to test a scalable experimental design for injective carbon mineralization implementation at the mine scale. Pad-1 was meant to test the injection into anisotropic layers and control the gas flow path by opening and closing exhaust ports. Pad-2 improved upon this experiment by switching from a point injection to a linear scheme via a perforated injection pipe. Further, this was no longer one-dimensional through a thin layer, but the gas had to travel horizontally and vertically. In examining the performance of the experiments, the physical performance of the experiment design must also be evaluated. 4.5.3.1 Pad-1 From the results presented for the Pad-1 experiment, the gas phase measurements demonstrate the plume advancement throughout the mixed PK layer. The highest CO2 concentrations occur within the first 40 centimetres from the injection point. Detection of CO2 in the headspace and the formation of desiccation cracking in the upper FPK layer determined the effluent point during the early part of the injection experiment. Highly saturated FPK layers would be relatively impermeable to CO2 infiltration. However, partial reaction of the FPK layers along the planar interface with the mixed PK layer for both the upper and lower confining layers was possible. With desiccation cracking, preferential gas flow could occur through these conduit pathways to escape to the headspace above the pad. This channelized flow would also lead to considerably more reactivity in the FPK along these flow paths (Harrison et al., 2016). With the headspace being sealed off after 70 hours, effluent gas was directed to the side ports, and this led to a complex flow distribution throughout the mixed PK layer that was not captured by the embedded CO2 sensors except for the ports closest to the injection point. Higher data density for the gas phase would have improved the detection of the gas movement patterns. Changing the 166  effluent port locations may have led to preferential flow paths through the mixed PK layer, which may not have been sampled representatively. Considering how the TIC sample results compare to the information provided by the gas-phase, the sampling of the mixed PK layer indicates that the TIC increase is more homogeneous than might be expected from the gas-phase data. While some samples were taken from \u2018dead\u2019 zones with little to no reaction, there was no clear trend with distance from the injection point for the most reacted samples. While the carbonation reaction may have occurred heterogeneously with unrepresentative sampling, the gas phase concentrations made no indication of this. Meanwhile, more homogenous reaction was observed in the FPK layers than expected. Notably, samples taken from zones with desiccation cracks running through them were higher than average. Core samples taken through the whole FPK layers proved indistinguishable from grab samples, removing the possibility of preferential sampling around interfaces where the FPK would have been exposed to CO2. Final moisture contents indicate the FPK layers consistently lost water (Table 4.10). The upper FPK layer could have experienced evaporation and drainage (Figure 4.22). For the lower FPK layer, the water could have been sucked upwards due to capillary action into the relatively drier mixed PK layer, and there was substantial capacity for drainage into the underlying CPK layer. The change in moisture content further complicates interpretations, as this would have increased the FPK\u2019s permeability and exposure to CO2. Reconciling the gas and solid phase results together is difficult. Limited reaction duration means that the expected TIC increase is further within the range of sample heterogeneity. Improving the interpretation and coherency of the gas and solid phase data was a focus of improvement for the second pad experiment. 167   Figure 4.22. Water drainage amongst the layers of Pad-1. 4.5.3.2 Pad-2 For Pad-2, the design of the experiment was influenced by the difficulty in interpreting the results from Pad-1. The perforated pipe injection facilitated interpretations of the results as a single two-dimensional cross-sectional plane of injection, as the injection pipe transported the gas in the third direction. More CO2 sensors were used, along the central plane, to ensure that the CO2 plume development was captured throughout the pad. In the lab setting, longer injection durations were also possible. CO2 was injected in Pad-2 for over five times as long as in Pad-1. Lastly, the FPK cap layer integrity was protected by placing a CPK layer on top to exclude the FPK from interacting with the ambient air, and the headspace volume above the pad was reduced considerably. Examining the gas phase data, Pad-2 differs from Pad-1 in that considerable plume progression is reached throughout the pad pore space. The gas flow through Pad-2 was modeled using Abaqus CAE finite element analysis software (Figure 4.23). The base CPK layer was not observed to enhance horizontal gas flow due to its higher permeability, as was intended. Instead, 168  the flow through the mixed PK layer was similar enough to prevent identifying differential concentration gradients, as there was no direct flow path through the CPK (Figure 4.23). With radial flow out of the perforated pipe, CO2 would have contacted the saturated and impermeable FPK barrier layer before reaching the intended outlet ports. With any weakness in the FPK layer, and potentially with gaps around the FPK layer where it met the wall of the IBC, CO2 would have preferentially leaked into the headspace. While CO2 did leak through the FPK barrier layer, this was eventually stopped by sealing the headspace. Upon sealing, the CO2 was forced towards the outlet ports. By 300 hours, the effluent concentration reached half of what was injected, with 465 hours of reaction remaining. The three outlet ports experienced the same amount of CO2 in the effluent, indicating even gas flow throughout the pad and not just in the middle plane. While concentrations did reach 6% by the end, they reached a maximum of 7% at the end of the second phase of injection, meaning the pad was exposed to sufficient CO2 for carbon mineralization to be well distributed. 169   Figure 4.23. Modeled flow direction (grey) and pressure equipotential lines (blue) of injected gas from the perforated pipe (lower left corner) into Pad-2. During injection, water was able to move between the layers with differing moisture contents (Table 4.10). The altered physical design appears to have helped, as the moisture content of the FPK cap layer proved to have only slightly decreased over the much longer experimental duration than observed for either FPK layer in Pad-1. Pad-2 had slight capillary action from the FPK layer into the cap CPK layer, while there was some drainage from the FPK layer into the top of the mixed PK layer (Figure 4.24). Meanwhile, the middle of the PK layer remained unchanged, while the base of the mixed PK layer lost water as water drained into the relatively drier CPK base layer. The accrued moisture in the base CPK layer would have facilitated reaction in this layer. 170   Figure 4.24. Water drainage and capillary action amongst the layers of Pad-2. TIC analysis on the retrieved samples examined the reaction principally down the middle plane in which the CO2 sensors were installed. Other samples were taken in a grid to ensure the carbonation was distributed evenly throughout the full pad. Little reaction occurred in the FPK and CPK cap layers. With the FPK layer maintaining its moisture content, the desiccation cracks that occurred in Pad-1 likely did not form, but rather, the FPK layer leaked along its edges. This maintained integrity is consistent with a lack of flow through the FPK layer and little identified carbonation (Figure 4.23). CPK at the cap may have been too dry to carbonate. For the mixed PK layer, the amount of carbonation is well distributed with respect to the proximity from the injection point in that it generally decreases with distance (Figure 4.19). The furthest column of samples has the least amount of captured carbon, which is consistent with the distribution of the CO2 gas plume. The shortest flow direction, diagonally from the injection pipe to the outlet, experienced the most reaction (Figure 4.19 and Figure 4.23). Finally, there is some reaction in the CPK base layer, with the most being near the injection point. While considering the heterogeneity involved in the PK sampling, the gas phase and the TIC results qualitatively support the same conclusions. A perforated injection pipe successfully injected CO2 into a thick layer of mixed PK, developing 171  uniformly distributed concentrations and reactivity. As was done for the pipe experiments, the flow rate was progressively stepped down to reduce CO2 in the effluent. The phased flow rates resulted in ~50% of the injected gas being captured, demonstrating progress towards higher capture efficiencies.  4.5.4 Carbon Accounting Accounting for the magnitude of captured carbon and verifying that it has been mineralized is of utmost priority for carbon sequestration. In the centimetre-scale experiments, agreement within 10% relative error was observed between two methods also used in this chapter: sampling representative material before and after carbon mineralization to assess the increase in TIC and performing a mass balance upon the volume of injected and effluent CO2. In examining the larger-scale experiments on processed kimberlite, challenges were experienced due to the lower overall reactivity and differences in results achieved amongst analytical methods. For Pipe-1, the gas phase mass balance accounted the captured carbon to be 0.14 to 0.16 wt.% CO2, while the bulk sample TIC increase was 0.10 wt.% CO2. These results are the beginning of a trend in that the gas mass balance is consistently different from the TIC increase. For Pipe-2, the mass balance found 0.23 wt.% CO2 versus the 0.09 wt.% CO2 of the bulk TIC increase. The pad experiments were similar with 0.09 wt.% CO2 (mass balance) versus 0.13 wt.% CO2 (bulk TIC) for Pad-1 and 0.20 wt.% CO2 (mass balance) versus 0.10 wt.% CO2 (bulk TIC) for Pad-2. Only Pipe-1 had the mass balance within the confidence interval of the TIC increase, while Pad-1 was the only experiment to have the TIC increase be higher than the mass balance. There are larger disparities between the mass balance and TIC increase for both Pipe-2 and Pad-2 as the longer injection durations lead to the mass balance consistently increasing with time. 172  With regards to the TIC increase, two challenges face this method. At larger scales, first, the starting material is more heterogeneous, as the larger sample is sourced from variable ore types and is less homogeneously mixed within the sample. Second, the carbonation reaction will also be more heterogeneous, as factors that affect carbon mineralization, such as the moisture content, mixing, and permeability, will be less homogenous than occurs at smaller scales. This heterogeneity inhibits the ability to ascertain the representative mean initial TIC and the representative mean post-carbonation TIC. When assessing the TIC content for relatively unreactive material, the result is that the TIC increase has significant errors associated with the calculation. For the two pipe experiments, the fine fraction below 425 \u00b5m was sieved out and analyzed independently for TIC to improve this method. 425 \u00b5m was selected as the sieve size, as this leads to a balance between the FPK and CPK fines relative abundance and the ability to include the most strongly reactive fines (Assima, et al., 2013a; Harrison et al., 2015). The analysis on the sieved fraction was done because the initial TIC of the CPK was more heterogeneous than the FPK. It is also known that finer grain sizes are inherently more reactive, and it should be expected that the finer fraction would prove to have a stronger TIC signal than the bulk sample (Assima et al., 2013a; Harrison et al., 2015; Lechat et al., 2016). Removing the coarse fraction would thus remove a diluting agent. Sieved sample results showed a much stronger TIC increase signal for both pipe experiments. When accounting for the stronger signal in relation to the whole sample, the TIC increases were augmented for both pipe experiments. For Pipe-1, the TIC increased from 0.10 to 0.11 wt.% CO2, while for Pipe-2, the increase improved from 0.09 to 0.12 wt.% CO2. Further, there was less heterogeneity for the final sieved TIC values, consistent with Lechat et al. (2016). The reduced confidence interval indicates that this method was successful and that uncertainty due 173  to heterogeneity was reduced. Uncertainties in this method occur primarily over whether a representative grain size is achieved during sieving. This method has an additional issue in that it is possible to miss sequestered carbon in the fraction above 425 \u00b5m (Assima et al., 2013a). After the sieved TIC analysis, the final methodology attempting to quantify the sequestered carbon involved assessing TC on the Pipe-2 experiment bulk samples. This analysis examined whether the smectite minerals contributed via intercalating CO2 in their interlayer space which would not necessarily be determined by coulometry (Giesting et al., 2012; Loring et al., 2012; Michels et al., 2015). Given that the carbon in addition to that measured by TIC was found to be identical before and after CO2 injection, this can be ruled out. Smectite minerals may still be exchanging cations which leads to Mg-carbonate formation, which is detectable by TIC. Pad-1 encountered another problem: more reactivity was observed by TIC increases in the FPK and mixed PK layers than could result from carbonation based on the known mass of CO2 injected, which should not be possible. The most likely reason for this is the larger heterogeneity of the original TIC than accounted for by sampling. The bulk FPK-1 sample was retrieved from the FPK containment facility via dredging. With FPK being deposited as a slurry, sampling vertically through thinly deposited layers would incur a heterogeneous sample, as the deposited tailings would be from different runs of ore from the mine processing circuit. This heterogeneity may not have been captured in the sampling done on the unreacted material. The seemingly homogenous reaction in both cap and base FPK layers may be explained by more initial carbon than was accounted for. Therefore, the contributions to the reactivity from the FPK layers have not been considered. FPK heterogeneity will have a lesser impact on the TIC increase assessed on the mixed PK layer, as the FPK composes only a minority of the sample fraction. 174  For Pad-2, this was not a concern as the FPK was provided as one continuous slurry at the mine. Results proved to be consistent with those found for Pipe-2, in that the TIC increase is only 50% of what might be expected from the gas mass balance. The reactivity quantification is inhibited by the higher focus on the mixed PK layer and the lesser degree of sampling of the other layers. Even with the improved results from analyzing TIC on the sieved fractions, these results do not approach the gas mass balance results. The trend remains that with longer injection durations, there are more significant discrepancies. For Pipe-1, the solid phase determinations found approximately the same result as Pipe-2, regardless of the injection duration differences. Better agreement is found for Pipe-1 with the gas estimate than for Pipe-2. While the solid phase results seem to indicate that after initial reactivity, the carbonation reaction shuts down, the gas phase results show that reactivity continues. Considering the reaction stages identified, the TIC data is consistent with the initial stage and part of the bulk reactivity stage, while the gas phase data shows the bulk reactivity stage continuing into a residual stage of reactivity. This is consistent with what is known of the Mg release rates from serpentine, which decrease asymptotically with time (Daval et al., 2013; Lu, 2020). The progressive decline in PK reactivity rates is observed over time for all experiments in the gas phase. The extended reactivity seen in the gas phase data cannot be explained by a dilution mechanism in the data of the effluent sensor. When the injection rate is changed, the reactivity rate follows suit (Figure 4.7,Figure 4.8Figure 4.15). Additionally, any dilution would be occurring against the pressure gradient which causes flow through the tailings. Granted that TIC heterogeneity has been observed to lead to TIC increases that are too high and too low, compared to mass balance estimates, the mass balance estimate must be considered the superior method in terms of reliability and consistency. Solid and gas phase methods were 175  shown to agree in Chapter 3 where the sample mass can be representatively sampled, and removing heterogeneity from the samples by sieving was shown to produce better agreement between the two methods. Errors around the gas-phase mass balance concern the accuracy of the sensors, the injected gas flow, and determinations made about the CO2 mass in the pore space and the aqueous phase, with all of these being systematic errors in comparison to the random errors associated with assessing TIC increases.  176  4.6 Implications 4.6.1 Magnitude of Carbon Mineralization In Chapter 3, the magnitude of reactivity for the processed serpentinite and the processed kimberlite tailings was considered in the context of mine emissions and passive sequestration rates from similar deposit types. Results from metre-scale experiments must also be considered within the context of the centimetre-scale results. Figure 4.25, initially presented in Chapter 3 as Figure 3.12, has been updated to include the range of sequestered carbon for the pad and pipe experiments. As a point of comparison, Diavik Diamond Mine is the closest mine mineralogically to Gahcho Ku\u00e9 that has had its passive carbonation rate investigated. Diavik has been estimated to capture 0.15 kt CO2\/Mt FPK\/year (Wilson et al., 2009b, 2011). However, it is readily acknowledged that greater ambient sequestration is limited by the subaqueous state of deposition, inhibiting CO2 drawdown from the atmosphere (Wilson et al., 2009b, 2011). In the centimetre-scale experiments, the mixed PK experiment with the longest duration was observed to sequester 2.0 kt CO2\/Mt PK, 13 times higher than observed for Diavik, in approximately one month, rather than over a full year. The two pipe experiments captured 1.1 to 1.2 kt CO2\/Mt PK as determined by the TIC increase and 1.4 to 2.3 kt CO2\/Mt PK as determined by the gas mass balance. Meanwhile, the amount of captured carbon ranged from 1.0 to 1.3 kt CO2\/Mt PK from TIC and 0.9 to 2.0 kt CO2\/Mt PK from the gas mass balance for the pad experiments. For the two field experiments (Pipe-1 and Pad-1), the injection duration was lower, at only ~140 hours. For the lab experiments (Pipe-2 and Pad-2), with injection durations approaching three weeks for Pipe-2 and one month for Pad-2, both sets indicate favourable results compared to the centimetre-scale results for the gas mass balance. The TIC results indicate that 50 to 60% of captured carbon found for the centimetre-scale experiments is sequestered in the metre-scale experiments. Taking the range of 1.0 to 2.3 kt 177  CO2\/Mt PK, this is equivalent to 2.6 to 6% of Gahcho Ku\u00e9\u2019s total CO2 emissions and 7 to 16.2% of their power generation emissions (De Beers Group, 2019b). Power generation emissions only make up approximately one-third of Gahcho Ku\u00e9\u2019s carbon footprint, with the rest being area source emissions.  Figure 4.25. CO2 emissions, injection (this study), and passive sequestration rates for the Diavik Diamond Mine (NT, Canada), Gahcho Ku\u00e9 Diamond Mine (NT, Canada), Mount Keith Nickel Mine (WA, Australia), Dumont Nickel Project (QC, Canada), and the Baptiste Nickel Project (BC, Canada). Percentages are expressed as the amount of sequestration in relation to the emissions from the respective mine. Results from Chapter 3 are presented for Baptiste. Results from Chapter 3 for Gahcho Ku\u00e9 have been replaced with Chapter 4 results, specifically for the TIC increase (filled diamonds) and gas mass balance (unfilled diamonds) for the lab (Pipe-2 and Pad-2) injection experiments.  178  4.6.2 Large-Scale Design Recommendations In this chapter, the focus has been to build carbon mineralization experiments in scale and complexity. Chapter 3 experiments were increased in scale, and configurations that would facilitate expansion to the mine-scale were assessed experimentally. Principally, a large-scale field demonstration would need to incorporate the construction of an engineered tailings pad, the diversion, cooling, compression and injection of flue gas, and the monitoring during and accounting afterward for sequestered carbon. Given the success and difficulties experienced and encountered in this study, there are clear implications concerning how CO2 injection may be deployed in large field trials and at the mine scale. 4.6.2.1 Porous Medium Design While it has been established that grain size distributions of mixed CPK and FPK enable the permeability to inject and the reactivity to capture CO2, mixing of PK and attaining unsaturated conditions in tailings stacks becomes a challenge as experimental scales increase. Co-disposal of mine tailings can occur in several forms, with the two most relevant being layered co-mingling (alternating layers of coarse and fines) and homogenous mixtures as were produced in this study. Layered co-mingling has potential for gas injection, as the coarse layers would enable longitudinal and lateral gas transport to the fine-grained layers, but infiltration of gas into the fines would be limited due to their lower permeability. Homogenous mixtures of coarse and fines have the advantages of increasing strength, removing the need for a tailings dam to contain a slurry, and reducing the total waste volume, expediting mine reclamation (Wickland et al., 2006). Meanwhile, the benefits for dry stack or filtered tailings, required to achieve the unsaturated conditions necessary for injection, include their low water usage, reclamation ease, and small environmental footprint (Davies & Rice, 2001; Oldecop et al., 2017). Disadvantages to both co-disposal and 179  filtered tailings include higher energy and operating costs (Davies & Rice, 2001; Wickland et al., 2006). Filtered tailings are also limited by their ore processing rates; however, many diamond mines are small with relatively low ore processing tonnages (Johnson & Pilotto, 2020; Wilson et al., 2009b, 2011). While Gahcho Ku\u00e9 plans on co-disposing their PK from 2022 onwards (Johnson & Pilotto, 2020), this most likely refers to saturating the CPK and waste rock with the FPK slurry. The Renard Diamond Mine (QC, Canada) is an exemplary case study as their proposed tailings management practices were nearly identical to those espoused here. At Renard, the initial plan was for the coarse (<6 mm) and fine (<1 mm) processed kimberlite to be dewatered by centrifuge, mixed and stacked as dry stack tailings (Godin et al., 2016). However, it appears this plan met challenges related to excess generation of fines, which led to some disposal as a slurry but still 60-65% of their PK being stacked as filtered tailings (Hiyate, 2017). This example indicates that the base conditions favourable for CO2 injection into PK are feasible at the mine scale and could be more broadly adopted. More specific design criteria governing the fines content, moisture content and the degree of compaction are needed to ensure successful injection of CO2. The grain size distribution is an essential design variable. The addition of fines increases the reactivity but decreases the injection permeability. However, this can be managed by lowering the degree of compaction as this has a significant effect on permeability. Initially, a grain size distribution of 25 wt.% FPK was chosen as the permeability of 10-11 m2 was selected as the lower limit. At this permeability, the tailings are still comparable to unconsolidated fine-grained sand (Bear, 1972). However, given that lowering the compaction increases permeability by up to an order of magnitude, it is likely that loosely compacted samples with FPK content of up to 40 wt.% 180  may remain above this permeability threshold. For Gahcho Ku\u00e9, 33 wt.% FPK represents utilizing all of the PK at the produced ratio of CPK to FPK. Guidelines are needed to ensure that PK mixed at the production ratio would maintain pore saturations amenable to CO2 injection based on the moisture content and degree of compaction while also meeting structural integrity requirements. Of critical importance is the management of high moisture contents. A lower degree of compaction permits a higher moisture content without negatively impacting the degree of pore saturation. A high degree of compaction is desired when the structural integrity of the medium is of importance (Ulrich & Coffin, 2013). Structural integrity may only be necessary for some zones of the tailings storage facility, such as the external perimeter (Klohn Crippen Berger, 2017; Lupo & Hall, 2010). Therefore, in cases where compaction must be >90% of the Standard Proctor MDD, the final moisture content (Mw\/Ms) must not exceed 10 wt.% (Figure 4.26). For loosely compacted materials, the final maximum moisture content (Mw\/Ms) should be below 14 wt.% (Figure 4.26). For 14 wt.% to remain below 60% pore saturation the degree of compaction must correspond to 80% of the MDD, which is a probable lower compaction limit given consolidation as the stack height increases. An emphasis is placed on this being the final moisture content, as CPK is more amenable to being drier and drying faster, and these final moisture contents could be achieved by mixing wetter FPK with drier CPK. Even if the material is wetter than ideal, mechanical reworking can promote evaporation before final emplacement in the tailings storage facility (Caldwell & Crystal, 2015; Godin et al., 2016; Oldecop et al., 2017). 181   Figure 4.26. An example of density and moisture content conditions for different locations within the tailings storage facility that enable the promotion of carbon sequestration via CO2 injection. A low state of compaction is desirable for carbon mineralization for including reactive fines and managing high moisture contents. Ultimately, compaction will occur with consolidation (Klohn Crippen Berger, 2017; Lupo & Hall, 2010). However, this may occur after significant carbon mineralization has been achieved by injection in the shallow subsurface. Unsaturated conditions can persist in dry stack tailings up to depths of 10 to 50 metres (Lupo & Hall, 2010). Keeping the material in the unsaturated state is generally preferable as negative pore pressures increase the stability and resistance to liquefaction (Oldecop et al., 2017; Ulrich & Coffin, 2013). In addition, the use of overlying fine-grained layers on coarser layers, as done here in an attempt Structural ZonesInteriorZones182  to trap injected gas, could limit infiltrating precipitation by inducing the capillary barrier effect, thus maintaining the unsaturated conditions (Oldecop et al., 2017). 4.6.2.2 Carbon Mineralization Via Injection The magnitude of carbon sequestration that can be achieved has implications for the source of CO2. With greater storage capacity than point source emissions, the opportunity to incorporate other CO2 sources is available. For example, low purity DAC systems, which require less energy due to the lower achieved pCO2 (Wilcox et al., 2017), could offset mine area source emissions or make the mine carbon negative. However, most mines do not have this excess capacity. For Gahcho Ku\u00e9, the estimated degree of carbon sequestration is a fraction of their power generation emissions. Other emission reduction methods could be coupled with carbon mineralization in mine tailings, with one example being biodiesel to achieve a form of bioenergy with carbon capture and storage. The rate of carbon sequestration that can be achieved has implications for the delivery of CO2, as injection rates should be tailored to the material\u2019s reactivity. The similarity found between the reaction rates of the centimetre and metre-scale experiments demonstrates the ability to predict or forecast the reactivity of larger experiments. The prediction of the reactivity rates allows the injection rate of flue gas to be controlled to match the reactivity of the tailings (Harrison et al., 2013b; Power et al., 2014). Further, it enables predicting the amount of sequestered carbon per mass of tailings and allows for longer-term rates to be forecasted beyond the experimental durations. Demonstrating this for PK indicates that this should also be possible for mine wastes from other deposits and is especially of interest for those containing higher abundances of highly reactive minerals. 183  As the experimental scale increases, the methods used here to verify carbon sequestration prove unsuitable unless adapted. For example, the challenges related to using TIC only increase as the scale increases. In this study, initial and carbonated sample heterogeneity was found to be within the range of the expected reactivity. Higher degrees of reactivity are necessary to enable this approach, as larger carbon increases would have relatively less heterogeneity effects. The abundance of brucite largely governs tailings reactivity. Therefore, the presence and abundance of brucite is a critical control on PK reactivity, and it is recommended that brucite bearing PK be isolated for future experiments to improve experimental success and interpretation. An alternative approach to selecting highly reactive material for larger demonstrations would be to use an isotopic approach to differentiate between the initial and sequestered carbon. Using a CO2 source with a 14C rich signature, and applying radiocarbon detection from the mineral phase, may prove more successful than the approaches used in this study to quantify mineralized carbon at low abundances (Mervine et al., 2014; Oskierski et al., 2013a, 2013b; Wilson et al., 2009a). Such a 14C source could include CO2-enriched air or the exhaust captured from burning biodiesel. Gas-phase estimates seem to be the best alternative to quantifying carbon sequestration in materials with below detection brucite abundances. CO2 mass balance methods were verified in the centimetre-scale experiments, with results matching those from TIC increases. This method could be accomplished at an increased experimental scale by directing effluent to extraction wells, where the flux and concentration could be monitored. This method requires a high degree of sealing to be successful, as the effluent fluxes would need to sum up to the injected flux. However, mass balance methods may be the most applicable at the mine scale as this is already accomplished in industry through eddy-covariance systems, which monitor large areas of the atmosphere for CO2 184  cycling in agriculture and forestry applications (Morin et al., 2018; Rodda et al., 2021; Waldo et al., 2016). These methods installed at the mine site over tailings storage facilities could be deployed to verify that injected carbon is not released. 185  Chapter 5: Conclusion 5.1 Research Objectives This thesis investigated the injection of CO2 into ultramafic mine tailings as a method to accelerate passive carbon mineralization rates and sequester CO2. Ultimately, the objective was to ready this technology for deployment at the mine scale and perhaps more broadly as a geological storage technique. This goal was broken into two sub-objectives, 1) assess dominant reactivity and feasibility controls, and 2) demonstrate and verify the accelerated sequestration of injected carbon across experimental scales.  5.2 Research Outcomes 5.2.1 Porous Medium Conditions For CO2 injection into mine tailings to be technologically feasible, the design of the porous medium is critical. Conditions that limit either reactivity or feasibility have been quantified, and other experimental conditions have been investigated qualitatively and semi-quantitatively. These findings are summarized below. 5.2.1.1 Quantified Parameters The porous medium must be permeable to facilitate sufficient CO2 supply. Permeability was measured on mixtures of coarse and fine tailings at the minimum and maximum degrees of pore saturation (Chapter 3). Up to 40 wt.% fines (<425 \u00b5m) within the grain size distributions results in sufficient permeabilities to inject (>10-11 m2). Carbon mineralization minimally impacted the permeability due to small volume changes in water-saturated pores (Chapters 3 and 4). Previous research determined that the degree of pore saturation must be below 60% for homogenous injection (Harrison et al., 2016). Standard Proctor compaction testing evaluated the 186  moisture-density relationship of PK blends (Chapter 4). Structural zones of tailings requiring >90% maximum dry density must have moisture contents <10 wt.%. Less compacted zones can allow moisture contents to be as high as 14 wt.%. Finally, the abundance of brucite is the most critical parameter for CO2 sequestration. Brucite can be detected at abundances of \u22650.3 wt.% from the TGA derivative sample mass loss (Chapter 2). The background mass loss rate was modelled, allowing the mass loss from brucite to be determined and related to the brucite abundance. Mineral standards with known brucite abundances found relative errors below 50% at low abundances and below 4% at high abundances. 5.2.1.2 Experimental Parameters Centimetre-scale injection experiments investigated the impact of porous medium parameters upon carbon mineralization (Chapter 3). The coarse phase mineralogy was insignificant when fines were present. Pure coarse materials were an order of magnitude less reactive than their fine-grained counterparts. Mineralogy was contrasted between processed serpentinite and kimberlite. Brucite content was the primary control of sequestration rates and magnitudes, with the minimum observed rate of the brucite-bearing PS being equal to the maximum rate of the PK, which had no brucite. 73 wt.% of the fine brucite was estimated to have reacted, while the fine serpentine contributed 2 \u2013 4 wt.% of its Mg.  The minimum necessary moisture content depends on the material\u2019s reactivity and not a specific degree of pore saturation, with moisture contents of 5 wt.% being sufficient. Since the maximum moisture content is relatable to the pore saturation (60%), the degree of saturation can be maintained by lowering compaction. This lower density enables higher moisture contents without negatively impacting injected gas distribution and overall reactivity. 187  Injection enabled the supply of CO2 to exceed the reaction rate of even brucite-rich tailings. The reactivity of brucite and serpentine was asymptotic, reacting rapidly at first and decreasing over time. Delivery of CO2 must follow this pattern to enable maximum reactivity while accomplishing a high degree of capture efficiency. Injection of diesel flue gas successfully captured carbon but was limited by the pCO2 (7.6 vol.% CO2) and the shorter injection duration (188 hours). 5.2.2 CO2 Injection Demonstration and Verification 5.2.2.1 Injection Reaction at Scale Centimetre-scale, one-dimensional pipe injection experiments enabled CO2 injection under idealized porous medium conditions, simplifying data interpretation. The PS precipitates were characterized as mostly hydromagnesite with some nesquehonite. Sequestration as nesquehonite is preferred due to the higher stoichiometric MgO: CO2 ratio and may be achievable in the conditions of the TSF. The precipitate formed in the PK experiments was not determined due to its low abundance. Magnitudes of sequestered carbon were as high as 1.4 wt.% CO2 for PS and 0.2 wt.% CO2 for PK. For the PS, this indicates that the Baptiste Nickel Project could sequester all of their carbon emissions at brucite abundances typical of the deposit. For Gahcho Ku\u00e9 Diamond Mine, a significant portion (14%) of their power generation emissions could be sequestered. Metre-scale, one-dimensional pipe injection experiments on PK served to scale up the centimetre-scale experiments to a length of 6 meters. The degree of compaction was lower than in the centimetre-scale experiments. Experimental complications arose as the first pipe experiment leaked during the experiment; however, the second pipe experiment was successful. Heterogeneity of the initial and reacted carbon made reconciling verification methods difficult, with relative 188  discrepancies of 50%. Nevertheless, metre-scale experiments replicated the reactivity of their centimetre counterparts, capturing 0.1 to 0.2 wt.% CO2. Lastly, metre-scale, three-dimensional pad injection experiments replicated the 6-metre pipe results in a design format amenable to larger-scale implementation. This design injected gas into mixed PK and involved permeable and aquitard layers to direct gas flow. The amount of the FPK in the PK mixture was increased to 33 wt.% for the second pad experiment. For both experiments, maintaining the integrity of the cap layer was a challenge. Interpretation of the first pad experiment was complicated by initial and final carbon heterogeneity and a low degree of data density. Using the perforated injection pipe, the second pad experiment was a significant improvement, with well-distributed reactivity. Assessing the amount of captured carbon was again hampered by heterogeneity. The reaction was still in the range of 0.1 to 0.2 wt.% CO2. This range is attributable to 7 \u2013 16% of Gahcho Ku\u00e9\u2019s power generation emissions. 5.2.2.2 Carbon Verification Four methods were used to determine the magnitude of sequestered carbon at the various experimental scales. The first consisted of a gas mass balance, which measured the volume of injected and effluent CO2, with the difference being attributed to solubility and mineral trapping. All other methods analyzed the solid phase. TIC analysis on bulk samples before and after carbonation was used to assess the amount of mineralized carbon. An alternative approach sieved the fine fraction (<425 \u00b5m) from the bulk sample and analyzed the TIC increase on the fines. Finally, TC, which is measured by a different technique than TIC, was also tested. The gas mass balance and the bulk sample TIC increase were used for all experiments and found good agreement at the centimetre scale for PS and PK experiments. For the metre-scale, the sieved sample TIC increase and the TC analysis attempted to reconcile the disparity between the mass balance and 189  the bulk sample TIC increase. For the bulk sample TIC increase, the small sample mass at the centimetre-scale could be relatively homogenized before sampling. At the metre-scale, random heterogeneity of initial and reacted carbon made it challenging to quantify the mass of carbon captured. Assessments on the sieved fraction improved the analysis in terms of the agreement with the gas mass balance. However, the difference between the sieved sample TIC increase and the mass balance was still significant. TC analysis found no additional sequestered carbon. These difficulties were exacerbated by the PK\u2019s overall low degree of reactivity and the expected TIC increase being within the range of sample heterogeneity. At the metre-scale, gas mass balances were the most reliable method with consistent and systematic errors. Ultimately, the reliability of the gas phase determination is favourable due to the cost associated with representative sampling and chemical analysis at larger scales.  5.3 CO2 Injection Field Pilot Design A large-scale conceptual design for successful CO2 injection into mine tailings was developed from the Pad-2 experiment and is presented in Figure 5.1. Within this diagram, perforated piping is arrayed at the bottom of the tailings layer to transport CO2-bearing gas through the tailings bed and inject the gas into the tailings along the pipe\u2019s length. Tailings arrive from the processing circuit in fine (reactive) and coarse (permeable) fractions. Fine-grained material would ideally be rich in brucite and serpentine, while the mineralogy of the coarse-grained material is insignificant. Materials that are dry enough for high degrees of compaction are isolated for use in structural zones. Wetter materials are deposited in interior zones at lower degrees of compaction. Compaction and moisture contents are managed to ensure the degree of pore saturation remains below 60% to promote homogeneous gas injection. Coarse and fine tailings are deposited together 190  and tilled to mix and produce a well-graded grain size distribution with reactive fines that is permeable for gas transport. Tilling may also be used to promote evaporation when moisture contents are too high. Once the tailings layer has been built up, it is capped with a liner or layer of fines to prevent effluent from escaping. The injection occurs into the finished layer, while the next tailings layer is built up. Flue gas from power generation is cooled, pressurized and injected into the perforated injection pipes. Exhaust is released through ports where the CO2 content and flux of the effluent are measured to monitor the amount and rate of carbon sequestration, with this influencing the injection rate. This design would be suitable for field pilots to demonstrate at scale the capability for CO2 injection to mineralize mine emissions. 191   Figure 5.1. Conceptual design of the large-scale implementation of CO2 injection into mine tailings management.Tailings from processing circuitTransportationDepositionMixing of fine and coarseFlue gas generation, cooling, compression and flow rate measurementPiping delivers injected flue gasLoose, wet mixed tailingsCompacted, dry mixed tailingsGenerator Heat Exchanger Compressor FlowmeterFineCoarseGas BarrierMonitored effluent192  5.4 Future Research Suggestions The next logical step for this research is to deploy this technology at the tonne-scale. However, many factors could be improved, and many challenges must be overcome to ensure experimental success. Critical to improving the magnitude and rate of sequestration involves understanding the long-term reactivity rates and mechanisms to attain higher degrees of reactivity from brucite and serpentine. The integration of flue gas requires the development of a system to cool exhaust, control flow rate and compress the exhaust to inject. Other flue gas components may affect the carbon mineralization reaction, with the most significant being the role of oxygen in reducing iron, leading to the formation of passivating Fe-hydroxides. Whether flue gas behaves similarly to simulated flue gas warrants investigation. The construction of an experiment at such a scale is also of concern, as this requires homogenous mixing of large quantities of tailings to ensure consistent permeability, fines and moisture content. In field conditions, the variable environmental conditions could increase or decrease targeted moisture contents. Moisture contents will also vary with depth as the tailings stack height is increased. The use of liners would facilitate the capping and protection of the mixed tailings from precipitation and contain the injected gas. Uncertainties surrounding the lateral and vertical extent of injection from perforated injection pipes are also of concern, as piping costs could compound if the extent of injection is limited. How injected carbon is verified as being mineralized is also of concern given the difficulties experienced at the metre-scale. Further work at the metre-scale should seek to identify the source of the discrepancies observed between solid and gas phase methods. In future experiments, having materials with brucite would facilitate interpreting the results as the reactivity 193  and the magnitude of the TIC increase would be more significant. Radiocarbon approaches to signal added 14C may be of value if a 14C rich source may be sourced. What is more, if solid phase methods are not preferred, a mass balance on the gas phase would require extensive sensor and data collection equipment and equipment to measure effluent fluxes. A total flux from all effluent ports may not sum to the injected flux leading to an inherent error in the method. Ensuring the complete sealing of a large-scale experiment is unlikely to be successful. Alternatively, at the pilot scale, the addition of eddy-covariance systems could prove successful in providing large-scale monitoring as a step towards accomplishing such a feat at the mine scale. While there is much research remaining to take carbon mineralization via CO2 injection to the mine scale, the research presented in this thesis has made significant contributions. CO2 injection enhances carbon mineralization rates sufficiently to reduce mining emissions and make some mines carbon neutral. Further, this approach may contribute to more diverse projects attempting to mineralize carbon to store carbon securely and mitigate climate change.  194  References Allmann, L. (1966). No Title. Neues Jahrb. Mineral., Monatsh., 161. Andreani, M., Luquot, L., Gouze, P., Godard, M., Hois\u00e9, E., & Gibert, B. (2009). Experimental Study of Carbon Sequestration Reactions Controlled by the Percolation of CO2-Rich Brine through Peridotites. Environmental Science and Technology, 43(4), 1226\u20131231. https:\/\/doi.org\/10.1021\/es8018429 Assima, G. P., Larachi, F., Beaudoin, G., & Molson, J. (2012). CO2 sequestration in chrysotile mining residues-implication of watering and passivation under environmental conditions. Industrial and Engineering Chemistry Research, 51(26), 8726\u20138734. https:\/\/doi.org\/10.1021\/ie202693q Assima, G. P., Larachi, F., Beaudoin, G., & Molson, J. (2013). Dynamics of carbon dioxide uptake in chrysotile mining residues-Effect of mineralogy and liquid saturation. International Journal of Greenhouse Gas Control, 12, 124\u2013135. https:\/\/doi.org\/10.1016\/j.ijggc.2012.10.001 Assima, G. P., Larachi, F., Molson, J., & Beaudoin, G. (2013). Accurate and direct quantification of native brucite in serpentine ores-New methodology and implications for CO2 sequestration by mining residues. Thermochimica Acta, 566, 281\u2013291. https:\/\/doi.org\/10.1016\/j.tca.2013.06.006 Assima, G. P., Larachi, F., Molson, J., & Beaudoin, G. (2014a). Comparative study of five Qu\u00e9bec ultramafic mining residues for use in direct ambient carbon dioxide mineral sequestration. Chemical Engineering Journal, 245, 56\u201364. https:\/\/doi.org\/10.1016\/j.cej.2014.02.010 Assima, G. P., Larachi, F., Molson, J., & Beaudoin, G. (2014b). Emulation of ambient carbon dioxide diffusion and carbonation within nickel mining residues. Minerals Engineering, 59, 195  39\u201344. https:\/\/doi.org\/10.1016\/j.mineng.2013.09.002 Assima, G. P., Larachi, F., Molson, J., & Beaudoin, G. (2014c). Impact of temperature and oxygen availability on the dynamics of ambient CO2 mineral sequestration by nickel mining residues. Chemical Engineering Journal, 240, 394\u2013403. https:\/\/doi.org\/10.1016\/j.cej.2013.12.010 Assima, G. P., Larachi, F., Molson, J., & Beaudoin, G. (2014d). New tools for stimulating dissolution and carbonation of ultramafic mining residues. Canadian Journal of Chemical Engineering, 92(12), 2029\u20132038. https:\/\/doi.org\/10.1002\/cjce.22066 ASTM International. (2012). Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Standard Effort (12,400 ft-lbf\/ft3 (600 kN-m\/m3)). https:\/\/doi.org\/10.1520\/D0698-12E01.1 ASTM International. (2013). Standard Test Method for Permeability of Rocks by Flowing Air (Vols. D4525-08). https:\/\/doi.org\/10.1520\/D4525-13E02.1 Bailey, T. (1955). No Title. Acta Crystallogr., 8, 621. Ballirano, P., De Vito, C., Mignardi, S., & Ferrini, V. (2013). Phase transitions in the Mg-CO2-H2O system and the thermal decomposition of dypingite, Mg5(CO3)4(OH)25H2O: Implications for geosequestration of carbon dioxide. Chemical Geology, 340, 59\u201367. https:\/\/doi.org\/10.1016\/j.chemgeo.2012.12.005 Bariand. (1973). No Title. Mineral. Rec., 4, 18. Bea, S. A., Wilson, S. A., Mayer, K. U., Dipple, G. M., Power, I. M., & Gamazo, P. (2012). Reactive Transport Modeling of Natural Carbon Sequestration in Ultramafic Mine Tailings. Vadose Zone Journal, 11(2). https:\/\/doi.org\/10.2136\/vzj2011.0053 Bear, J. (1972). Dynamics of Fluids in Porous Media. Elsevier. B\u00e9arat, H., McKelvy, M. J., Chizmeshya, A. V. G., Gormley, D., Nunez, R., Carpenter, R. W., 196  Squires, K., & Wolf, G. H. (2006). Carbon Sequestration via Aqueous Olivine Mineral Carbonation: Role of Passivating Layer Formation. Environmental Science and Technology, 40(15), 4802\u20134808. https:\/\/doi.org\/10.1021\/es0523340 Beaudoin, G., H\u00e9bert, R., & Constantin, M. (2008). Spontaneous carbonation of serpentine in milling and mining waste, southern Qu\u00e9bec and Italy. Proceedings of Accelerated Carbonation for Environmental and Materials Engineering (ACEME2008), January, 73\u201382. Beinlich, A., & Austrheim, H. (2012). In situ sequestration of atmospheric CO2 at low temperature and surface cracking of serpentinized peridotite in mine shafts. Chemical Geology, 332\u2013333, 32\u201344. https:\/\/doi.org\/10.1016\/j.chemgeo.2012.09.015 Bergaya, F., & Vayer, M. (1997). CEC of clays: Measurement by adsorption of a copper ethylenediamine complex. Applied Clay Science, 12(3), 275\u2013280. https:\/\/doi.org\/10.1016\/S0169-1317(97)00012-4 Berner, R. A., Lasaga, A. C., & Garrels, R. M. (1983). The carbonate-silicate geochemical cycle and its effect on atmospheric carbon dioxide over the past 100 million years. In American Journal of Science (Vol. 283, Issue 7, pp. 641\u2013683). https:\/\/doi.org\/10.2475\/ajs.283.7.641 Bobicki, E. R., Liu, Q., Xu, Z., & Zeng, H. (2012). Carbon capture and storage using alkaline industrial wastes. Progress in Energy and Combustion Science, 38(2), 302\u2013320. https:\/\/doi.org\/10.1016\/j.pecs.2011.11.002 Boschi, C., Dini, A., Baneschi, I., Bedini, F., Perchiazzi, N., & Cavallo, A. (2017). Brucite-driven CO2 uptake in serpentinized dunites (Ligurian Ophiolites, Montecastelli, Tuscany). Lithos, 288\u2013289, 264\u2013281. https:\/\/doi.org\/10.1016\/j.lithos.2017.07.005 Boudier, F., & Nicolas, A. (1985). Harzburgite and lherzolite subtypes in ophiolitic and oceanic environments. Earth and Planetary Science Letters, 76(1\u20132), 84\u201392. 197  https:\/\/doi.org\/10.1016\/0012-821X(85)90150-5 Brindley, G. (1977). No Title. In ICDD Grant-in-Aid. Britten, R. (2017). Regional metallogeny and genesis of a new deposit type-disseminated awaruite (Ni3Fe) mineralization hosted in the cache creek terrane. Economic Geology, 112(3), 517\u2013550. https:\/\/doi.org\/10.2113\/econgeo.112.3.517 Budhu, M. (2011). Soil Mechanics and Foundations (Third). John Wiley & Sons, Inc. Caldwell, J. A., & Crystal, C. (2015). Filter-pressed tailings facility design, construction, and operating guidelines. Proceedings Tailings and Mine Waste, 8 p. Canterford, J. H., Tsambourakis, G., & Lambert, B. (1984). Some observations on the properties of dypingite, Mg5(CO3)4(OH)2\u00b75H2O, and related minerals. Mineralogical Magazine, 48(September), 437\u2013442. Chakravarthy, C., Chalouati, S., Chai, Y. E., Fantucci, H., & Santos, R. M. (2020). Valorization of kimberlite tailings by carbon capture and utilization (CCU) method. Minerals, 10(7), 1\u201318. https:\/\/doi.org\/10.3390\/min10070611 Cheary, R. W., & Coelho, A. (1992). Fundamental parameters approach to x-ray line-profile fitting. Journal of Applied Crystallography, 25(pt 2), 109\u2013121. https:\/\/doi.org\/10.1107\/S0021889891010804 Chukhrov, F. (1973). No Title. Izv. Akad. Nauk SSSR, 23. Clement, C. R., Skinner, E. M. W., & Scott Smith, B. H. (1984). Kimberlite Redefined. Journal of Geology, 92(2), 223\u2013228. Comodi, P., Mellini, M., Ungaretti, L., & Zanazzi, P. (1991). No Title. Eur. J. Mineral., 3, 485. Daval, D., Hellmann, R., Martinez, I., Gangloff, S., & Guyot, F. (2013). Lizardite serpentine dissolution kinetics as a function of pH and temperature, including effects of elevated pCO2. 198  Chemical Geology, 351, 245\u2013256. https:\/\/doi.org\/10.1016\/j.chemgeo.2013.05.020 Daval, D., Sissmann, O., Menguy, N., Saldi, G. D., Guyot, F., Martinez, I., Corvisier, J., Garcia, B., Machouk, I., Knauss, K. G., & Hellmann, R. (2011). Influence of amorphous silica layer formation on the dissolution rate of olivine at 90\u00b0C and elevated pCO2. Chemical Geology, 284(1\u20132), 193\u2013209. https:\/\/doi.org\/10.1016\/j.chemgeo.2011.02.021 Davies, M. (2011). Filtered dry stacked tailings: the fundamentals. Proceedings Tailings and Mine Waste 2011, Vancouver B.C., 9 p. Davies, M., & Rice, S. (2001). An alternative to conventional tailing management - \u201cdry stack\u201d filtered tailings. In Proceedings Tailings and Mine Waste \u201901, 411\u2013420. Davies, P. J., & Bubela, B. (1973). The transformation of nesquehonite into hydromagnesite. Chemical Geology, 12(4), 289\u2013300. https:\/\/doi.org\/10.1016\/0009-2541(73)90006-5 De Beers Group. (2019a). Gahcho Ku\u00e9 Mine 2018 Annual Water Licence Report (Issue September). De Beers Group. (2019b). Gahcho Ku\u00e9 Mine 2019 Annual Air Quality Report (Issue August). de Coninck, H., Revi, A., Babiker, M., Bertoldi, P., Buckeridge, M., Cartwright, A., Dong, W., Ford, J., Fuss, S., Hourcade, J.-C., Ley, D., Mechler, R., Newman, P., Revokatova, A., Schultz, S., Steg, L., & Sugiyama, T. (2018). Strengthening and implementing the global response. In: Global Warming of 1.5\u00b0C. An IPCC Special Report on the impacts of global warming of 1.5\u00b0C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat  of climate change, sustainable development, and efforts to eradicate poverty. Falini, G., Fermani, S., Gazzano, M., & Ripamonti, A. (1998). Structure and morphology of synthetic magnesium calcite. J. Mater. Chem., 8, 1061. 199  Fanchi, J. R. (2002). Chapter 7 - Measures of Rock-Fluid Interactions. In Shared Earth Modeling (pp. 108\u2013132). Federico, M., Gianfagna, A., & Zonazzi, P. (1988). No Title. Neues Jahrb. Mineral., Monatsh., 495. Finger, L. W., & Hazen, R. M. (1978). Crystal structure and compression of ruby to 46 kbar. J. Appl. Phys., 49, 5823. F\u00f6ldv\u00e1ri, M. (2011). Handbook of thermo-gravimetric system of minerals and its use in geological practice. In Occasional papers of the geological institute of Hungary (Vol. 213). Frost, R. L., Bahfenne, S., Graham, J., & Martens, W. N. (2008). Thermal stability of artinite, dypingite and brugnatellite-Implications for the geosequestration of green house gases. Thermochimica Acta, 475(1\u20132), 39\u201343. https:\/\/doi.org\/10.1016\/j.tca.2008.06.007 Frost, R. L., & Erickson, K. L. (2004). Thermal decomposition of synthetic hydrotalcites reevesite and pyroaurite. Journal of Thermal Analysis and Calorimetry, 76(1), 217\u2013225. https:\/\/doi.org\/10.1023\/B:JTAN.0000027820.58744.bd Giester, G., Lengauer, C. L., & Rieck, B. (2000). The crystal structure of nesquehonite, MgCO3\u00b73H2O, from Lavrion, Greece. Mineral. Petrol., 70, 153\u2013163. Giesting, P., Guggenheim, S., Koster van Groos, A. F., & Busch, A. (2012). Interaction of carbon dioxide with Na-exchanged montmorillonite at pressures to 640bars: Implications for CO2 sequestration. International Journal of Greenhouse Gas Control, 8, 73\u201381. https:\/\/doi.org\/10.1016\/j.ijggc.2012.01.011 Gillery. (n.d.). No Title. In Private Communication. Godin, P., Hopkins, R., & Bedell, P. (2016). Updated Renard Diamond Project Mine Plan and Mineral Reserve Estimate Qu\u00e9bec , Canada NI 43-101 Technical Report. 200  Goodyear, D. (1954). No Title. Mineral. Mag., 30, 306. Gras, A., Beaudoin, G., Molson, J., & Plante, B. (2020). Atmospheric carbon sequestration in ultramafic mining residues and impacts on leachate water chemistry at the Dumont Nickel Project, Quebec, Canada. Chemical Geology, 546(March 2018), 119661. https:\/\/doi.org\/10.1016\/j.chemgeo.2020.119661 Gras, A., Beaudoin, G., Molson, J., Plante, B., Bussi\u00e8re, B., Lemieux, J. M., & Dupont, P. P. (2017). Isotopic evidence of passive mineral carbonation in mine wastes from the Dumont Nickel Project (Abitibi, Quebec). International Journal of Greenhouse Gas Control, 60, 10\u201323. https:\/\/doi.org\/10.1016\/j.ijggc.2017.03.002 Hales, M. C., Frost, R. L., & Martens, W. N. (2008). Thermo-Raman spectroscopy of synthetic nesquehonite \u2013 implication for the geosequestration of greenhouse gases. Journal of Raman Spectroscopy, 39, 1141\u20131149. https:\/\/doi.org\/10.1002\/jrs.1950 Hamilton, J. L., Wilson, S. A., Morgan, B., Harrison, A. L., Turvey, C. C., Paterson, D. J., Dipple, G. M., & Southam, G. (2020). Accelerating mineral carbonation in ultramafic mine tailings via direct CO2 reaction and heap leaching with potential for base metal enrichment and recovery. Economic Geology, 115(2), 303\u2013323. https:\/\/doi.org\/10.5382\/ECONGEO.4710 Hamilton, J. L., Wilson, S. A., Turvey, C. C., Morgan, B., Tait, A. W., McCutcheon, J., Fallon, S. J., & Southam, G. (2021). Carbon accounting of mined landscapes, and deployment of a geochemical treatment system for enhanced weathering at Woodsreef Chrysotile Mine, NSW, Australia. Journal of Geochemical Exploration, 220(May 2020), 106655. https:\/\/doi.org\/10.1016\/j.gexplo.2020.106655 H\u00e4nchen, M., Prigiobbe, V., Baciocchi, R., & Mazzotti, M. (2008). Precipitation in the Mg-carbonate system-effects of temperature and CO2 pressure. Chemical Engineering Science, 201  63(4), 1012\u20131028. https:\/\/doi.org\/10.1016\/j.ces.2007.09.052 Harrison, A. L., B\u00e9n\u00e9zeth, P., Schott, J., Oelkers, E. H., & Mavromatis, V. (2021). Magnesium and carbon isotope fractionation during hydrated Mg-carbonate mineral phase transformations. Geochimica et Cosmochimica Acta, 293, 507\u2013524. https:\/\/doi.org\/10.1016\/j.gca.2020.10.028 Harrison, A. L., Dipple, G. M., Power, I. M., & Mayer, K. U. (2015). Influence of surface passivation and water content on mineral reactions in unsaturated porous media: Implications for brucite carbonation and CO2 sequestration. Geochimica et Cosmochimica Acta, 148, 477\u2013495. https:\/\/doi.org\/10.1016\/j.gca.2014.10.020 Harrison, A. L., Dipple, G. M., Power, I. M., & Mayer, K. U. (2016). The impact of evolving mineral\u2013water\u2013gas interfacial areas on mineral\u2013fluid reaction rates in unsaturated porous media. Chemical Geology, 421, 65\u201380. https:\/\/doi.org\/10.1016\/j.chemgeo.2015.12.005 Harrison, A. L., Dipple, G. M., Song, W., Power, I. M., Mayer, K. U., Beinlich, A., & Sinton, D. (2017). Changes in mineral reactivity driven by pore fluid mobility in partially wetted porous media. Chemical Geology, 463(May), 1\u201311. https:\/\/doi.org\/10.1016\/j.chemgeo.2017.05.003 Harrison, A. L., Mavromatis, V., Oelkers, E. H., & B\u00e9n\u00e9zeth, P. (2019). Solubility of the hydrated Mg-carbonates nesquehonite and dypingite from 5 to 35\u2009\u00b0C: Implications for CO2 storage and the relative stability of Mg-carbonates. Chemical Geology, 504(October 2018), 123\u2013135. https:\/\/doi.org\/10.1016\/j.chemgeo.2018.11.003 Harrison, A. L., Power, I. M., & Dipple, G. M. (2013a). Accelerated Carbonation of Brucite in Mine Tailings for Carbon Sequestration. Environmental Science and Technology, 47(1), 126\u2013134. https:\/\/doi.org\/10.1021\/es3012854 Harrison, A. L., Power, I. M., & Dipple, G. M. (2013b). Strategies for enhancing carbon 202  sequestration in Mg-rich mine tailings. Proceedings of International Mine Water Association, 593\u2013598. Hetman, C. M., Scott Smith, B. H., Paul, J. L., & Winter, F. (2004). Geology of the Gahcho Ku\u00e9 kimberlite pipes, NWT, Canada: Root to diatreme magmatic transition zones. Lithos, 76(1-4 SPEC. ISS.), 51\u201374. https:\/\/doi.org\/10.1016\/j.lithos.2004.03.051 Hitch, M., & Dipple, G. M. (2012). Economic feasibility and sensitivity analysis of integrating industrial-scale mineral carbonation into mining operations. Minerals Engineering, 39, 268\u2013275. https:\/\/doi.org\/10.1016\/j.mineng.2012.07.007 Hiyate, A. (2017). Stornoway\u2019s Renard diamond mine reaches full production. Canadian Mining Journal. http:\/\/www.canadianminingjournal.com\/features\/stornoways-renard-diamond-mine-reaches-full-production\/ Holtz, R. D., & Kovacs, W. D. (1981). An Introduction to Geotechnical Engineering. Prentice Hall. Hopkinson, L., Kristova, P., Rutt, K., & Cressey, G. (2012). Phase transitions in the system MgO-CO2-H2O during CO2 degassing of Mg-bearing solutions. Geochimica et Cosmochimica Acta, 76, 1\u201313. https:\/\/doi.org\/10.1016\/j.gca.2011.10.023 Hopkinson, L., Rutt, K., & Cressey, G. (2008). The transformation of nesquehonite to hydromagnesite in the system CaO-MgO-H2O-CO2: An experimental spectroscopic study. Journal of Geology, 116(4), 387\u2013400. https:\/\/doi.org\/10.1086\/588834 H\u00f6velmann, J., Putnis, C. V., Ruiz-Agudo, E., & Austrheim, H. (2012). Direct nanoscale observations of CO2 sequestration during brucite [Mg(OH)2] dissolution. Environmental Science and Technology, 46(9), 5253\u20135260. https:\/\/doi.org\/10.1021\/es300403n Huijgen, W. J. J., Witkamp, G. J., & Comans, R. N. J. (2006). Mechanisms of aqueous wollastonite 203  carbonation as a possible CO2 sequestration process. Chemical Engineering Science, 61(13), 4242\u20134251. https:\/\/doi.org\/10.1016\/j.ces.2006.01.048 IPCC. (2018). Summary for Policymakers. In Global Warming of 1.5\u00b0C. An IPCC Special Report on the impacts of global warming of 1.5\u00b0C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change,. https:\/\/doi.org\/10.1016\/j.oneear.2019.10.025 Jauffret, G., Morrison, J., & Glasser, F. P. (2015). On the thermal decomposition of nesquehonite. Journal of Thermal Analysis and Calorimetry, 122(2), 601\u2013609. https:\/\/doi.org\/10.1007\/s10973-015-4756-0 Johnson, D. D., & Pilotto, D. (2020). Gahcho Ku\u00e9 Mine Ni 43-101 Technical Report. Johnson, N. C., Thomas, B., Maher, K., Rosenbauer, R. J., Bird, D., & Brown, G. E. (2014). Olivine dissolution and carbonation under conditions relevant for in situ carbon storage. Chemical Geology, 373, 93\u2013105. https:\/\/doi.org\/10.1016\/j.chemgeo.2014.02.026 Jones, S. C. (1997). A technique for faster pulse-decay permeability measurements in tight rocks. SPE Formation Evaluation, 12(1), 19\u201324. https:\/\/doi.org\/10.2118\/28450-pa Kandji, E. H. B., Plante, B., Bussi\u00e8re, B., Awoh, A. S., Ibrahim Sa\u00efb, B. A., & Dupont, P. P. (2015). Prediction of drainage water quality from mining waste with carbon sequestration potential. Proc. (IMWA\/ICARD), Santiago, Chili, 21-24 Avril 2015., 1\u201311. Kandji, E. H. B., Plante, B., Bussi\u00e8re, B., Beaudoin, G., & Dupont, P. P. (2017). Kinetic testing to evaluate the mineral carbonation and metal leaching potential of ultramafic tailings: Case study of the Dumont Nickel Project, Amos, Qu\u00e9bec. Applied Geochemistry, 84, 262\u2013276. https:\/\/doi.org\/10.1016\/j.apgeochem.2017.07.005 Kelemen, P. B., Matter, J., Streit, E. E., Rudge, J. F., Curry, W. B., & Blusztajn, J. (2011). Rates 204  and mechanisms of mineral carbonation in peridotite: Natural processes and recipes for enhanced, in situ CO2 capture and storage. Annual Review of Earth and Planetary Sciences, 39, 545\u2013576. https:\/\/doi.org\/10.1146\/annurev-earth-092010-152509 Kelemen, P. B., McQueen, N., Wilcox, J., Renforth, P., Dipple, G., & Vankeuren, A. P. (2020). Engineered Carbon Mineralization in Ultramafic Rocks for CO2 Removal from Air: Review and New Insights. Chemical Geology. https:\/\/doi.org\/10.1016\/j.chemgeo.2020.119628 Kern, A., & Eysel, W. (1993). No Title. In ICDD Grant-in-Aid. Khoiri, M., & Ou, C. Y. (2013). Evaluation of deformation parameter for deep excavation in sand through case histories. Computers and Geotechnics, 47, 57\u201367. https:\/\/doi.org\/10.1016\/j.compgeo.2012.06.009 Kirkham, M. B. (2005). 7 - Water Movement in Saturated Soil. In Principles of Soil and Plant Water Relations (pp. 85\u2013100). Academic Press. Klohn Crippen Berger. (2017). Study of Tailings Management Technologies MEND Report 2.50.1. Kojima, T., Nagamine, A., Ueno, N., & Uemiya, S. (1997). Absorption and fixation of carbon dioxide by rock weathering. Energy Conversion and Management, 38(SUPPL. 1). https:\/\/doi.org\/10.1016\/s0196-8904(96)00311-1 Lackner, K. S. (2003). A Guide to CO2 Sequestration. Science, 300, 1677\u20131678. https:\/\/doi.org\/10.1126\/science.1079033 Lackner, K. S., Butt, D. P., & Wendt, C. H. (1997). Progress on Binding CO2 In Mineral Substrates. Energy Conversion and Management, 38, 259\u2013264. https:\/\/doi.org\/10.1016\/S0196-8904(96)00279-8 Lackner, K. S., Wendt, C. H., Butt, D. P., Joyce, E. L., & Sharp, D. H. (1995). Carbon dioxide disposal in carbonate minerals. Energy, 20(11), 1153\u20131170. https:\/\/doi.org\/10.1016\/0360-205  5442(95)00071-N Langmuir, D. (1965). Stability of Carbonates in the System MgO-CO2-H2O. The Journal of Geology, 73(5), 730\u2013754. https:\/\/doi.org\/https:\/\/doi.org\/10.1086\/627113 Lara, J. L., Pornillos, E. U., & Munoz, H. E. (2013). Geotechnical-geochemical and operational considerations for the application of dry stacking tailings deposits \u2013 state-of-the-art. Proceedings of the 16th International Seminar on Paste and Thickened Tailings, 251\u2013262. https:\/\/doi.org\/10.36487\/acg_rep\/1363_19_munoz Lechat, K., Lemieux, J. M., Molson, J., Beaudoin, G., & H\u00e9bert, R. (2016). Field evidence of CO2 sequestration by mineral carbonation in ultramafic milling wastes, Thetford Mines, Canada. International Journal of Greenhouse Gas Control, 47, 110\u2013121. https:\/\/doi.org\/10.1016\/j.ijggc.2016.01.036 Legrand, U., Apfel, U. P., Boffito, D. C., & Tavares, J. R. (2020). The effect of flue gas contaminants on the CO2 electroreduction to formic acid. Journal of CO2 Utilization, 42 (August), 101315. https:\/\/doi.org\/10.1016\/j.jcou.2020.101315 Li, J., Hitch, M., Power, I. M., & Pan, Y. (2018). Integrated mineral carbonation of ultramafic mine deposits\u2014A review. Minerals, 8(4), 1\u201318. https:\/\/doi.org\/10.3390\/min8040147 Lin, Y., Zheng, M., Ye, C., & Power, I. M. (2018). Thermogravimetric analysis\u2013mass spectrometry (TGA\u2013MS) of hydromagnesite from Dujiali Lake in Tibet, China. Journal of Thermal Analysis and Calorimetry, 133(3), 1429\u20131437. https:\/\/doi.org\/10.1007\/s10973-018-7197-8 Liu, C., Liu, T., & Wang, D. (2018). Non-isothermal kinetics study on the thermal decomposition of brucite by thermogravimetry. Journal of Thermal Analysis and Calorimetry, 134(3), 2339\u20132347. https:\/\/doi.org\/10.1007\/s10973-018-7654-4 206  Loring, J. S., Schaef, H. T., Turcu, R. V. F., Thompson, C. J., Miller, Q. R. S., Martin, P. F., Hu, J., Hoyt, D. W., Qafoku, O., Ilton, E. S., Felmy, A. R., & Rosso, K. M. (2012). In situ molecular spectroscopic evidence for CO2 intercalation into montmorillonite in supercritical carbon dioxide. Langmuir, 28(18), 7125\u20137128. https:\/\/doi.org\/10.1021\/la301136w Lu, X. (2020). CHARACTERIZATION OF ULTRAMAFIC MINE TAILINGS REACTIVITY FOR CARBON CAPTURE AND STORAGE. University of British Columbia. Lupo, J., & Hall, J. (2010). Dry stack tailings - Design considerations. Tailings and Mine Waste\u201910 - Proceedings of the 14th International Conference on Tailings and Mine Waste, 327\u2013334. https:\/\/doi.org\/10.1201\/b10569-40 Meier, L. P. (1999). Determination of the Cation Exchange Capacity (CEC) of Clay Minerals Using the Complexes of Copper(II) Ion with Triethylenetetramine and Tetraethylenepentamine. Clays and Clay Minerals, 47(3), 386\u2013388. https:\/\/doi.org\/10.1346\/ccmn.1999.0470315 Mervine, E. M., Humphris, S. E., Sims, K. W. W., Kelemen, P. B., & Jenkins, W. J. (2014). Carbonation rates of peridotite in the Samail Ophiolite, Sultanate of Oman, constrained through 14C dating and stable isotopes. Geochimica et Cosmochimica Acta, 126, 371\u2013397. https:\/\/doi.org\/10.1016\/j.gca.2013.11.007 Mervine, E. M., Wilson, S. A., Power, I. M., Dipple, G. M., Turvey, C. C., Hamilton, J. L., Vanderzee, S., Raudsepp, M., Southam, C., Matter, J. M., Kelemen, P. B., Stiefenhofer, J., Miya, Z., & Southam, G. (2018). Potential for offsetting diamond mine carbon emissions through mineral carbonation of processed kimberlite: an assessment of De Beers mine sites in South Africa and Canada. Mineralogy and Petrology, 112, 755\u2013765. https:\/\/doi.org\/10.1007\/s00710-018-0589-4 207  Michels, L., Fossum, J. O., Rozynek, Z., Hemmen, H., Rustenberg, K., Sobas, P. A., Kalantzopoulos, G. N., Knudsen, K. D., Janek, M., Plivelic, T. S., & Da Silva, G. J. (2015). Intercalation and retention of carbon dioxide in a smectite clay promoted by interlayer cations. Scientific Reports, 5, 2\u201310. https:\/\/doi.org\/10.1038\/srep08775 Milidragovic, D., Zagorevski, A., Weis, D., Joyce, N., & Chapman, J. B. (2018). Picrite \u201cIntelligence\u201d from the Middle-Late Triassic Stikine arc: Composition of mantle wedge asthenosphere. Lithos, 308\u2013309(2018), 446\u2013461. https:\/\/doi.org\/10.1016\/j.lithos.2018.03.014 Mills, S. J., Wilson, S. A., Dipple, G. M., & Raudsepp, M. (2010).  The decomposition of konyaite: importance in CO2 fixation in mine tailings . Mineralogical Magazine, 74(5), 903\u2013917. https:\/\/doi.org\/10.1180\/minmag.2010.074.5.903 Moncur, M. C., & Smith, L. J. D. (2012). Processed Kimberlite Porewater Geochemistry from Diavik Diamond Mines, Inc. 9th International Conference on Acid Rock Drainage. https:\/\/doi.org\/10.13140\/2.1.3504.7369 Morgan, B., Wilson, S. A., Madsen, I. C., Gozukara, Y. M., & Habsuda, J. (2015). Increased thermal stability of nesquehonite (MgCO3\u00b73H2O) in the presence of humidity and CO2: Implications for low-temperature CO2 storage. International Journal of Greenhouse Gas Control, 39, 366\u2013376. https:\/\/doi.org\/10.1016\/j.ijggc.2015.05.033 Morin, T. H., Rey-Sanchez, A. C., Vogel, C. S., Matheny, A. M., Kenny, W. T., & Bohrer, G. (2018). Carbon dioxide emissions from an oligotrophic temperate lake: An eddy covariance approach. Ecological Engineering, 114, 25\u201333. https:\/\/doi.org\/https:\/\/doi.org\/10.1016\/j.ecoleng.2017.05.005 Nahdi, K., Rouquerol, F., & Trabelsi Ayadi, M. (2009). Mg(OH)2 dehydroxylation: A kinetic study 208  by controlled rate thermal analysis (CRTA). Solid State Sciences, 11(5), 1028\u20131034. https:\/\/doi.org\/10.1016\/j.solidstatesciences.2009.02.013 National Bureau of Standards. (1956). No Title. In Natl. Bur. Stand. (U. S. ), Circ. 539 (Vol. 6). National Bureau of Standards. (1957). No Title. In Natl. Bur. Stand. (U. S. ), Circ. 539 (Vol. 7). National Bureau of Standards. (1967). No Title. In Natl. Bur. Stand. (U. S. ) Monogr. 25 (Vol. 5). National Bureau of Standards. (1984). No Title. In Natl. Bur. Stand. (U. S. ) Monogr. 25 (Vol. 20). Nield, D. A., & Bejan, A. (2006). Convection in Porous Media (Third). Springer Science+Business Media Inc. Nixon, G. T. (1998). Ni-Cu sulfide mineralization in the Turnagain Alaskan-type complex: A unique magmatic environment. Geological Fieldwork 1997, 1\u201312. Niyazova, S. (2020). FELSIC XENOLITHS ASSIMILATION IN THE RENARD 65 KIMBERLITE PIPE , QU\u00c9BEC , CANADA (Issue December). Nowamooz, A., Dupuis, J. C., Beaudoin, G., Molson, J., Lemieux, J. M., Horswill, M., Fortier, R., Larachi, F., Maldague, X., Constantin, M., Duchesne, J., & Therrien, R. (2018). Atmospheric Carbon Mineralization in an Industrial-Scale Chrysotile Mining Waste Pile. Environmental Science and Technology, 52(14), 8050\u20138057. https:\/\/doi.org\/10.1021\/acs.est.8b01128 O\u2019Hanley, D. S., & Dyar, M. D. (1993). The composition of lizardite 1T and the formation of magnetite in serpentinites. American Mineralogist, 78(3\u20134), 391\u2013404. O\u2019Neil, J. R., & Barnes, I. (1971). C13 and O18 compositions in some fresh-water carbonates associated with ultramafic rocks and serpentinites: western United States. Geochimica et Cosmochimica Acta, 35(7), 687\u2013697. https:\/\/doi.org\/10.1016\/0016-7037(71)90067-6 Oelkers, E. H. (1999). A Comparison of Forsterite and Enstatite Dissolution Rates and Mechanisms. In B. Jamtveit & P. Meakin (Eds.), Growth, Dissolution and Pattern Formation 209  in Geosystems (pp. 253\u2013265). Springer Science+Business Media Inc. Oldecop, L. A., Rodari, G. J., & Mu\u00f1oz, J. J. (2017). Atmosphere Interaction and Capillary Barrier in Filtered Tailings. Geotechnical and Geological Engineering, 35(4), 1803\u20131817. https:\/\/doi.org\/10.1007\/s10706-017-0210-3 Onur, E. M. (2014). Predicting the Permeability of Sandy Soils From Grain Size Distributions. 123. Oskierski, H. C., Dlugogorski, B. Z., & Jacobsen, G. (2013a). Sequestration of atmospheric CO2 in a weathering-derived, serpentinite-hosted magnesite deposit: 14C tracing of carbon sources and age constraints for a refined genetic model. Geochimica et Cosmochimica Acta, 122, 226\u2013246. https:\/\/doi.org\/10.1016\/j.gca.2013.08.029 Oskierski, H. C., Dlugogorski, B. Z., & Jacobsen, G. (2013b). Sequestration of atmospheric CO2 in chrysotile mine tailings of the Woodsreef Asbestos Mine, Australia: Quantitative mineralogy, isotopic fingerprinting and carbonation rates. Chemical Geology, 358, 156\u2013169. https:\/\/doi.org\/10.1016\/j.chemgeo.2013.09.001 Pabst. (1937). No Title. Am. Mineral., 22, 862. Park, A. H. A., & Fan, L. S. (2004). CO2 mineral sequestration: Physically activated dissolution of serpentine and pH swing process. Chemical Engineering Science, 59(22\u201323), 5241\u20135247. https:\/\/doi.org\/10.1016\/j.ces.2004.09.008 Paulo, C., Power, I. M., Stubbs, A. R., Wang, B., Zeyen, N., & Wilson, S. A. (2021). Evaluating feedstocks for carbon dioxide removal by enhanced rock weathering and CO2 mineralization. Applied Geochemistry, 104955. https:\/\/doi.org\/10.1016\/j.apgeochem.2021.104955 Pokrovsky, O. S., Schott, J., & Castillo, A. (2005). Kinetics of brucite dissolution at 25\u00b0C in the presence of organic and inorganic ligands and divalent metals. Geochimica et Cosmochimica 210  Acta, 69(4), 905\u2013918. https:\/\/doi.org\/10.1016\/j.gca.2004.08.011 Power, I. M., Dipple, G. M., Bradshaw, P. M. D., & Harrison, A. L. (2020). Prospects for CO2 mineralization and enhanced weathering of ultramafic mine tailings from the Baptiste nickel deposit in British Columbia, Canada. International Journal of Greenhouse Gas Control, 94(October 2019), 102895. https:\/\/doi.org\/10.1016\/j.ijggc.2019.102895 Power, I. M., Harrison, A. L., Dipple, G. M., Wilson, S. A., Kelemen, P. B., Hitch, M., & Southam, G. (2013). Carbon mineralization: From natural analogues to engineered systems. Reviews in Mineralogy and Geochemistry, 77, 305\u2013360. https:\/\/doi.org\/10.2138\/rmg.2013.77.9 Power, I. M., McCutcheon, J., Harrison, A. L., Wilson, S. A., Dipple, G. M., Kelly, S., Southam, C., & Southam, G. (2014). Strategizing carbon-neutral mines: A case for pilot projects. Minerals, 4, 399\u2013436. https:\/\/doi.org\/10.3390\/min4020399 Pronost, J., Beaudoin, G., Lemieux, J. M., H\u00e9bert, R., Constantin, M., Marcouiller, S., Klein, M., Duchesne, J., Molson, J. W., Larachi, F., & Maldague, X. (2012). CO2-depleted warm air venting from chrysotile milling waste (Thetford Mines, Canada): Evidence for in-situ carbon capture from the atmosphere. Geology, 40(3), 275\u2013278. https:\/\/doi.org\/10.1130\/G32583.1 Pronost, J., Beaudoin, G., Tremblay, J., Larachi, F., Duchesne, J., H\u00e9bert, R., & Constantin, M. (2011). Carbon sequestration kinetic and storage capacity of ultramafic mining waste. Environmental Science and Technology, 45(21), 9413\u20139420. https:\/\/doi.org\/10.1021\/es203063a Rausis, K., \u0106wik, A., & Casanova, I. (2020). Phase evolution during accelerated CO2 mineralization of brucite under concentrated CO2 and simulated flue gas conditions. Journal of CO2 Utilization, 37(November 2019), 122\u2013133. https:\/\/doi.org\/10.1016\/j.jcou.2019.12.007 Rodda, S. R., Thumaty, K. C., Praveen, M., Jha, C. S., & Dadhwal, V. K. (2021). Multi-year eddy 211  covariance measurements of net ecosystem exchange in tropical dry deciduous forest of India. Agricultural and Forest Meteorology, 301\u2013302(108351). https:\/\/doi.org\/https:\/\/doi.org\/10.1016\/j.agrformet.2021.108351 Rollo, H. A., & Jamieson, H. E. (2006). Interaction of diamond mine waste and surface water in the Canadian Arctic. Applied Geochemistry, 21(9), 1522\u20131538. https:\/\/doi.org\/10.1016\/j.apgeochem.2006.05.008 Rosenquist. (1959). No Title. Nor. Geol. Tidsskr., 39, 350. Ross, N. L., & Reeder, R. J. (1992). High-pressure structural study of dolomite and ankerite. Am. Mineral., 77, 412. Saburi, S., Kawahara, A., Henmi, C., Kusachi, I., & Kihara, K. (1977). No Title. Mineral. J., 8, 236. Sarvaramini, A., Assima, G. P., Beaudoin, G., & Larachi, F. (2014). Biomass torrefaction and CO2 capture using mining wastes - A new approach for reducing greenhouse gas emissions of co-firing plants. Fuel, 115, 749\u2013757. https:\/\/doi.org\/10.1016\/j.fuel.2013.07.087 Schott, J., Pokrovsky, O. S., & Oelkers, E. H. (2009). The link between mineral dissolution\/precipitation kinetics and solution chemistry. Reviews in Mineralogy and Geochemistry, 70, 207\u2013258. https:\/\/doi.org\/10.2138\/rmg.2009.70.6 Seifritz, W. (1990). CO2 disposal by means of silicates. Nature, 345, 486. https:\/\/doi.org\/10.1038\/345486b0 Sissmann, O., Brunet, F., Martinez, I., Guyot, F., Verlaguet, A., Pinquier, Y., & Daval, D. (2014). Enhanced Olivine Carbonation within a Basalt as Compared to Single-Phase Experiments: Reevaluating the Potential of CO2 Mineral Sequestration. Environmental Science and Technology, 48(10), 5512\u20135519. https:\/\/doi.org\/10.1021\/es405508a 212  Smith. (1974). No Title. In ICDD Grant-in-Aid. Smith, Y. (1956). No Title. Mineral. Mag, 31, 209. Spratt, H. J., Palmer, S. J., & Frost, R. L. (2008). Thermal decomposition of synthesised layered double hydroxides based upon Mg\/(Fe,Cr) and carbonate. Thermochimica Acta, 479(1\u20132), 1\u20136. https:\/\/doi.org\/10.1016\/j.tca.2008.08.016 Steinthorsdottir, K. (2021). FORMATION AND PRESERVATION OF BRUCITE AND AWARUITE IN A SERPENTINIZED ULTRAMAFITE, CENTRAL BRITISH COLUMBIA: IMPLICATIONS FOR CARBON SEQUESTRATION AND NICKEL RECOVERY. University of British Columbia. Stubbs, A. R. (2020). ENHANCED WEATHERING AND CARBONATION OF KIMBERLITE RESIDUES FROM SOUTH AFRICAN DIAMOND MINES\u2009: IMPLICATIONS FOR CO2 SEQUESTRATION. Trent University. Tanikawa, W., & Shimamoto, T. (2006). Klinkenberg effect for gas permeability and its comparison to water permeability for porous sedimentary rocks. Hydrology and Earth System Sciences Discussions, 3(4), 1315\u20131338. https:\/\/doi.org\/10.5194\/hessd-3-1315-2006 Tidwell, V. C. (2006). 15. Air Permeability Measurements in Porous Media. In Gas Transport in Porous Media (pp. 273\u2013278). Springer. Trittschack, R., Grob\u00e9ty, B., & Brodard, P. (2014). Kinetics of the chrysotile and brucite dehydroxylation reaction: A combined non-isothermal\/isothermal thermogravimetric analysis and high-temperature X-ray powder diffraction study. Physics and Chemistry of Minerals, 41(3), 197\u2013214. https:\/\/doi.org\/10.1007\/s00269-013-0638-9 Tsao, Y. (1964). No Title. Ti Chih Hsueh Pao (J. Geol.), 44, 219. Turenne, M. (2021). FPX Nickel Reports Potential to Achieve Production with Lowest Carbon 213  Footprint in Global Nickel Industry. https:\/\/fpxnickel.com\/2021\/01\/fpx-nickel-reports-potential-to-achieve-production-with-lowest-carbon-footprint-in-global-nickel-industry\/#more-91778 Turvey, C. C., Hamilton, J. L., & Wilson, S. A. (2018). Comparison of Rietveld-compatible structureless fitting analysis methods for accurate quantification of carbon dioxide fixation in ultramafic mine tailings. American Mineralogist, 103(10), 1649\u20131662. https:\/\/doi.org\/10.2138\/am-2018-6515 Turvey, C. C., Wilson, S. A., Hamilton, J. L., & Southam, G. (2017). Field-based accounting of CO2 sequestration in ultramafic mine wastes using portable X-ray diffraction. American Mineralogist, 102(6), 1302\u20131310. https:\/\/doi.org\/10.2138\/am-2017-5953 Turvey, C. C., Wilson, S. A., Hamilton, J. L., Tait, A. W., McCutcheon, J., Beinlich, A., Fallon, S. J., Dipple, G. M., & Southam, G. (2018). Hydrotalcites and hydrated Mg-carbonates as carbon sinks in serpentinite mineral wastes from the Woodsreef chrysotile mine, New South Wales, Australia: Controls on carbonate mineralogy and efficiency of CO2 air capture in mine tailings. International Journal of Greenhouse Gas Control, 79(April 2018), 38\u201360. https:\/\/doi.org\/10.1016\/j.ijggc.2018.09.015 Ulrich, B., & Coffin, J. (2013). Considerations for tailings facility design and operation using filtered tailings. Proceedings of the 16th International Seminar on Paste and Thickened Tailings, 2003, 201\u2013210. https:\/\/doi.org\/10.36487\/acg_rep\/1363_15_ulrich United Nations Framework Convention on Climate Change. (2016). Decisions adopted by the Conference of the Parties. In Report of the Conference of the Parties on its twenty-first session, held in Paris from 30 November to 13 December 2015. Addendum. Part two: Action taken by the Conference of the Parties at its twenty-first session. 214  University of British Columbia. (2018). EOSC 329: Groundwater Hydrology, Lab 2: Darcy\u2019s Law and Hydraulic Conductivity. University of Oxford. (1967). No Title. In Private Communication. van der Merwe, E. M., & Strydom, C. A. (2004). Quantitative thermogravimetric analysis of binary mixtures. Journal of Thermal Analysis and Calorimetry, 76(1), 149\u2013156. https:\/\/doi.org\/10.1023\/b:jtan.0000027814.93703.0c Vanderzee, S. S. S., Dipple, G. M., & Bradshaw, P. M. D. (2019). Targeting Highly Reactive Labile Magnesium in Ultramafic Tailings for Greenhouse-Gas Offsets and Potential Tailings Stabilization at the Baptiste Deposit, Central British Columbia (NTS 093K\/13, 14). Vanderzee, S. S. S., Power, I. M., Dipple, G. M., & Bradshaw, P. M. D. (2018). Carbon Mineralization in Ultramafic Tailings, Central British Columbia: A Case for Stabilizing Mine Waste and Reducing Greenhouse Gas Emissions. Geoscience BC Annual Report, 109\u2013112. Viti, C., & Mellini, M. (1997). No Title. Eur. J. Mineral, 9, 585. Vogelia, J., Reid, D. L., Beckera, M., Broadhursta, J., & Franzidisa, J.-P. (2011). Investigation of the potential for mineral carbonation of PGM tailings in South Africa. Minerals Engineering, 24(12), 1348\u20131356. Waldo, S., Chi, J., Pressley, S. N., O\u2019Keeffe, P., Pan, W. L., Brooks, E. S., Huggins, D. R., St\u00f6ckle, C. O., & Lamb, B. K. (2016). Assessing carbon dynamics at high and low rainfall agricultural sites in the inland Pacific Northwest US using the eddy covariance method. Agricultural and Forest Meteorology, 218\u2013219, 25\u201336. https:\/\/doi.org\/10.1016\/j.agrformet.2015.11.018 Wang, J. A., Novaro, O., Bokhimi, X., L\u00f3pez, T., G\u00f3mez, R., Navarrete, J., Llanos, M. E., & L\u00f3pez-Salinas, E. (1998). Characterizations of the thermal decomposition of brucite prepared by sol-gel technique for synthesis of nanocrystalline MgO. Materials Letters, 35(5\u20136), 317\u2013215  323. https:\/\/doi.org\/10.1016\/S0167-577X(97)00273-5 Watson, A., Corser, P., Garces Pardo, E., Lopez Christian, T., & Vandekeybus, J. (2010). A comparison of alternative tailings disposal methods \u2014 the promises and realities. 499\u2013514. https:\/\/doi.org\/10.36487\/acg_rep\/1008_41_watson Wickland, B. E., Wilson, G. W., Wijewickreme, D., & Klein, B. (2006). Design and evaluation of mixtures of mine waste rock and tailings. Canadian Geotechnical Journal, 43(9), 928\u2013945. https:\/\/doi.org\/10.1139\/T06-058 Wilcox, J., Psarras, P. C., & Liguori, S. (2017). Assessment of reasonable opportunities for direct air capture. Environmental Research Letters, 12(6), 65001. Wilson, S. A., Barker, S. L. L., Dipple, G. M., & Atudorei, V. (2010). Isotopic disequilibrium during uptake of atmospheric CO2 into mine process waters: Implications for CO2 sequestration. Environmental Science and Technology, 44(24), 9522\u20139529. https:\/\/doi.org\/10.1021\/es1021125 Wilson, S. A., Dipple, G. M., Power, I. M., Barker, S. L. L., Fallon, S. J., & Southam, G. (2011). Subarctic weathering of mineral wastes provides a sink for atmospheric CO2. Environmental Science and Technology, 45(18), 7727\u20137736. https:\/\/doi.org\/10.1021\/es202112y Wilson, S. A., Dipple, G. M., Power, I. M., Thom, J. M., Anderson, R. G., Raudsepp, M., Gabites, J. E., & Southam, G. (2009). Carbon dioxide fixation within mine wastes of ultramafic-hosted ore deposits: Examples from the Clinton Creek and Cassiar Chrysotile deposits, Canada. Economic Geology, 104, 95\u2013112. https:\/\/doi.org\/10.2113\/gsecongeo.104.1.95 Wilson, S. A., Harrison, A. L., Dipple, G. M., Power, I. M., Barker, S. L. L., Ulrich Mayer, K., Fallon, S. J., Raudsepp, M., & Southam, G. (2014). Offsetting of CO2 emissions by air capture in mine tailings at the Mount Keith Nickel Mine, Western Australia: Rates, controls and 216  prospects for carbon neutral mining. International Journal of Greenhouse Gas Control, 25, 121\u2013140. https:\/\/doi.org\/10.1016\/j.ijggc.2014.04.002 Wilson, S. A., Raudsepp, M., & Dipple, G. M. (2006). Verifying and quantifying carbon fixation in minerals from serpentine-rich mine tailings using the Rietveld method with X-ray powder diffraction data. American Mineralogist, 91, 1331\u20131341. https:\/\/doi.org\/10.2138\/am.2006.2058 Wilson, S. A., Raudsepp, M., & Dipple, G. M. (2009). Quantifying carbon fixation in trace minerals from processed kimberlite: A comparative study of quantitative methods using X-ray powder diffraction data with applications to the Diavik Diamond Mine, Northwest Territories, Canada. Applied Geochemistry, 24, 2312\u20132331. https:\/\/doi.org\/10.1016\/j.apgeochem.2009.09.018 Zarandi, A. E., Larachi, F., Beaudoin, G., Plante, B., & Sciortino, M. (2016). Multivariate study of the dynamics of CO2 reaction with brucite-rich ultramafic mine tailings. International Journal of Greenhouse Gas Control, 52, 110\u2013119. https:\/\/doi.org\/10.1016\/j.ijggc.2016.06.022 Zarandi, A. E., Larachi, F., Beaudoin, G., Plante, B., & Sciortino, M. (2017a). Ambient mineral carbonation of different lithologies of mafic to ultramafic mining wastes\/tailings \u2013 A comparative study. International Journal of Greenhouse Gas Control, 63(March), 392\u2013400. https:\/\/doi.org\/10.1016\/j.ijggc.2017.06.016 Zarandi, A. E., Larachi, F., Beaudoin, G., Plante, B., & Sciortino, M. (2017b). Nesquehonite as a carbon sink in ambient mineral carbonation of ultramafic mining wastes. Chemical Engineering Journal, 314, 160\u2013168. https:\/\/doi.org\/10.1016\/j.cej.2017.01.003  217  Appendices Appendix 1: Appendix to Chapter 2 A1.1 Detailed Methods A1.1.1 Sample Preparation Sample preparation was required for XRD. Fine-grained samples (<600 \u00b5m) were micronized for seven minutes under anhydrous ethanol using a McCrone Micronizing Mill. Samples were then dried, disaggregated and homogenized using an agate mortar and pestle.  A1.1.2 X-Ray Diffraction Mineralogy of PK and PS samples was determined through qualitative and quantitative X-ray Diffraction (XRD) analysis, which was performed using a Bruker D8 Focus located within the Department of Earth, Ocean and Atmospheric Sciences at the University of British Columbia. Prior to analysis, samples of PK were treated using a Ca2+ cation exchange process for 12 hours using a 1 mol Ca2+ solution as per the method of Wilson et al. (in prep). All samples were prepared as back-loading cavity mounts and were loaded against 400 grit sandpaper to reduce preferred orientation. Patterns were collected on the D8 Focus using a Co X-ray tube operated at 35 kV and 40 mA using a LynxEye 1D position-sensitive detector over a 2\u03b8 range of 3-80\u00b0 with a step size of 0.03\u00b0 and a dwell time of 0.7 s\/step. Qualitative analysis was performed by DIFFRAC.EVA V.4.2 (Bruker AXS) using the ICDD PDF-4+ database. Quantitative analysis was performed using DIFFRAC.TOPAS V.4.2 (Bruker AXS) using the fundamental parameters approach (Cheary & Coelho, 1992). The crystal structure data sources are displayed in Table A1.1. 218  Table A1.1. Sources of the crystal structure data for quantitative XRD analysis. Mineral Source Brucite (National Bureau of Standards, 1956) Serpentine (Viti & Mellini, 1997) Forsterite (National Bureau of Standards, 1984) Smectite (Rosenquist, 1959) Magnetite (National Bureau of Standards, 1967) Diopside (Federico et al., 1988) Phlogopite (Smith, 1956) Clinochlore (Gillery, n.d.) Calcite (Falini et al., 1998) Quartz (Kern & Eysel, 1993) Albite (Goodyear, 1954) K-Feldspar (Bailey, 1955; University of Oxford, 1967)  Talc (Brindley, 1977) Dolomite (Ross & Reeder, 1992) Cuspidine (Saburi et al., 1977) Andradite (Tsao, 1964) Hydromagnesite (Bariand, 1973) Magnesite (National Bureau of Standards, 1957) Ferrihydrite (Chukhrov, 1973) Pyroaurite (Allmann, 1966)  219  A1.2 Detailed Results A1.2.1 X-Ray Diffraction The mineralogy of the processed serpentinite sample, \u2018FPX-CT\u2019 (Baptiste Nickel Project, BC, Canada), was characterized by Steinthorsdottir (2021). The \u2018natural brucite\u2019 sample (Brucite Mine, NV, United States) and the processed kimberlite sample, \u2018FPK-1\u2019 (Gahcho Ku\u00e9 Diamond Mine, NT, Canada), were characterized in this study. The \u2018quartz sand\u2019 sample (Lane Mountain Materials) was characterized by Lu (2020). Quantitative results are presented in Table A1.2, and representative XRD patterns of the samples characterized in this study are included in Figures A1.1 and A1.2.220  Table A1.2. Mean mineralogical abundance and standard deviations as determined by qXRD. Abundance (wt.%) FPX-CT FPK-1 Natural Brucite Quartz Sand Serpentine c 79.8 37.7 \u00b1 9.4 3.8 - Forsterite 8.1 1.0 \u00b1 0.1 - - Smectite -a 14.2 \u00b1 5.5 - - Magnetite 6.3 1.2 \u00b1 0.1 - - Diopside 4.6 7.4 \u00b1 7.0 - - Phlogopite - 10.0 \u00b1 3.5 - 2.4 Clinochlore - 6.9 \u00b1 2.7 - - Calcite - 0.5 \u00b1 0.5 - - Quartz 1.0 2.9 \u00b1 1.1 0.2 97.6 Albite - 4.2 \u00b1 1.6 - - K-Feldspar - 6.6 \u00b1 2.0 - - Talc - 6.6 \u00b1 1.4 - - Dolomite 0.2 0.1 \u00b1 0.1 5.5 - Cuspidine - 0.6 \u00b1 0.1 - - Andradite - tr. b - - Brucite - - 76.2 - Hydromagnesite - - 6.9 - Magnesite - - 5.6 - Ferrihydrite - - 1.5 - Pyroaurite - - 0.3 - a Not detected. b Trace abundance. Peak identified in the XRD pattern but not included in the quantitative analysis. 221   Figure A1.1. Representative XRD pattern of the natural brucite sample.                               Natural Brucite222   Figure A1.2. Representative XRD pattern and Rietveld refinement of the processed kimberlite sample FPK-1.  2Th Degrees8075706560555045403530252015105Counts8,0007,5007,0006,5006,0005,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,500-3,00019GKFPKB2-1-CAT.raw_1 Phlogopite 1M 12.49 %Clinochlore IIb-4 68942 8.81 %Calcite 0.81 %Quartz low 2.09 %Magnetite 1.13 %Microcline intermediate 8.09 %Albite low 5.37 %Talc 1A 7.58 %Dolomite 0.19 %Diopside 2.48 %Forsterite 1.04 %Cuspidine 0.72 %Mellini & Viti Lizardite 1T P31M 31.06 %Ca-Montmorillonite 18.12 %FPK-1223  A1.2.2 Thermogravimetric Analysis Thermogravimetric analysis was performed on each mineral standard. Representative TG and DTG curves are presented for all standards in Figures A1.3 \u2013 A1.19. The linear extrapolation and exponential interpolation are displayed on each plot. Note that the plotted data for these methods extend past the end of the brucite interval and past where the difference in the mass loss was taken.  224   Figure A1.3. Representative TG and DTG curves of PS-STD-1 (left) and PS-STD-2 (right), with the linear extrapolation and exponential interpolation methods shown.225   Figure A1.4. Representative TG and DTG curves of PS-STD-3 (left) and PS-STD-4 (right), with the linear extrapolation and exponential interpolation methods shown.226   Figure A1.5. Representative TG and DTG curves of PS-STD-5, with the linear extrapolation and exponential interpolation methods shown.227   Figure A1.6. Representative TG and DTG curves of PK-STD-1 (left) and PK-STD-2 (right), with the linear extrapolation and exponential interpolation methods shown.228   Figure A1.7. Representative TG and DTG curves of PK-STD-3 (left) and PK-STD-4 (right), with the linear extrapolation and exponential interpolation methods shown.229   Figure A1.8. Representative TG and DTG curves of PK-STD-5 (left) and PK-STD-6 (right), with the linear extrapolation and exponential interpolation methods shown.230   Figure A1.9. Representative TG and DTG curves of PK-STD-7 (left) and PK-STD-8 (right), with the linear extrapolation and exponential interpolation methods shown.231   Figure A1.10. Representative TG and DTG curves of PK-STD-9 (left) and PK-STD-10 (right), with the linear extrapolation and exponential interpolation methods shown.232   Figure A1.11. Representative TG and DTG curves of PK-STD-11 (left) and PK-STD-12 (right), with the linear extrapolation and exponential interpolation methods shown.233   Figure A1.12. Representative TG and DTG curves of PK-STD-13 (left) and PK-STD-14 (right), with the linear extrapolation and exponential interpolation methods shown.234   Figure A1.13. Representative TG and DTG curves of PK-STD-15 (left) and PK-STD-16 (right), with the linear extrapolation and exponential interpolation methods shown.235   Figure A1.14. Representative TG and DTG curves of PK-STD-17 with the linear extrapolation and exponential interpolation methods shown.236   Figure A1.15. Representative TG and DTG curves of BRC-STD-1 (left) and BRC-STD-2 (right), with the exponential interpolation method shown.237   Figure A1.16. Representative TG and DTG curves of BRC-STD-3 (left) and BRC-STD-4 (right), with the exponential interpolation method shown.238   Figure A1.17. Representative TG and DTG curves of BRC-STD-5 (left) and BRC-STD-6 (right), with the exponential interpolation method shown.239   Figure A1.18. Representative TG and DTG curves of BRC-STD-7 (left) and BRC-STD-8 (right), with the exponential interpolation method shown.240   Figure A1.19. Representative TG and DTG curves of BRC-STD-9, with the exponential interpolation method shown. 241  Appendix 2: Appendix to Chapter 3 A2.1 Detailed Experimental Set-Up A2.1.1 Pneumatic Permeability ASTM D4525 details the standard method for determining the permeability of rock by injecting air (ASTM International, 2013). A modified method has been used to determine the permeability of the tailings samples, which differ in that being aggregates, their permeabilities are orders of magnitude higher than typically observed in rock cores. Tailings samples were weighed dry using a Mettler Toledo ML3002E scale. In the case where coarse and fine tailings were mixed, the respective proportions were weighed and combined. A mass of deionized water was added to reach the desired moisture content, and the wet sample was homogeneously mixed. The moist tailings sample was compacted in ~2 cm stages to the desired degree of compaction. The sample holder was a polycarbonate column manufactured by W.A. Hammond Drierite. It had an internal diameter of 5.65 cm, and the sample lengths varied from 15 to 20 cm. The diameter exceeded the largest grain size (<6 mm) by an order of magnitude, and the length of the sample exceeded the diameter by more than a factor of 2. Bosch Sensortec BMP280 or BMP388 pressure sensors were installed on either end of the column, enabling the measurement of the inlet and outlet absolute pressures. The sample holder was sealed, and the air was injected under laminar flow conditions into the horizontal sample holder by either a calibrated valve (WEONE LZM-6 flow meter), a Matheson FM-1050 series rotameter flowmeter, or a Bronkhorst EL-FLOW Prestige mass flow controller. Three flow rates were injected. Pressures were recorded every two seconds during each test by a Raspberry Pi 3B+. Test durations were typically 50 to 100 seconds, with sufficient time being given for pressures to stabilize. 242  A2.1.2 CO2 Injection Tailings samples were weighed dry using a Mettler Toledo ML3002E scale. Coarse and fine tailings were weighed in their proportions and combined. Deionized water was added to achieve the desired moisture content, and the wet tailings were homogeneously mixed. The wet tailings were consistently compacted every ~2 cm. The sample mass was known by the amount of mass introduced into the sample holder. The sample holder consisted of 2 or 4-inch diameter schedule 40 PVC piping, with an internal diameter of 5.2 or 10.1 cm. Sample lengths ranged from 12 to 34 cm, and sample masses ranged from 500 g to 2 kg. Bosch Sensortec BMP280 or BMP388 pressure sensors were installed in the PVC cap on each side of the pipe. Vaisala GMP251 CO2 sensors were installed into the wall of the pipe on either side of the tailings sample. This required excess pipe to allow for enough room on either side of the tailings sample to install the CO2 sensors. Silicone was used to seal the caps and sensors and the column mass was weighed prior to gas injection. A permeability test was done before and after CO2 injection. SFG flow was controlled by either a Matheson FM-1050 series rotameter flowmeter, a Masterflex variable area flowmeter (model RK-03227-06), or a Masterflex L\/S precision standard pump with an Easy-Load II pump head. SFG was hydrated by bubbling through water within an Erlenmeyer flask prior to flow through the tailings sample. When the Masterflex peristaltic pump was used, gas was withdrawn from the Erlenmeyer flask, which was oversupplied with gas by a mass flow controller (Matheson FM-1050 series rotameter flowmeter, or Masterflex variable area flowmeter (model RK-03227-06)). DFG was supplied by the Masterflex pump directly from the Tedlar gas sampling bag. Vaisala MI70 Measurement Indicators were used to record the CO2 concentrations at 5 or 15-minute intervals. Upon experimental completion, the column was weighed, and the sample mass was removed and dried under ambient conditions. 243  A2.2 Detailed Methods A2.2.1 Sample Characterization A2.2.1.1 Sample Preparation Sample preparation was required for XRD, TGA and TIC analysis. Coarse-grained samples (>600 \u00b5m) were pulverized with a ring mill for 30 seconds. Pulverized and fine-grained samples (<600 \u00b5m) were micronized for seven minutes under anhydrous ethanol using a McCrone Micronizing Mill. Samples were then dried, disaggregated and homogenized using an agate mortar and pestle. A2.2.1.2 X-Ray Diffraction Mineralogy of PK and PS samples was determined through qualitative and quantitative X-ray Diffraction (XRD) analysis, which was performed using a Bruker D8 Focus located within the Department of Earth, Ocean and Atmospheric Sciences at the University of British Columbia. Prior to analysis, samples of PK were treated using a Ca2+ cation exchange process for 12 hours using a 1 mol Ca2+ solution as per the method of Wilson et al. (in prep). All samples were prepared as back-loading cavity mounts and were loaded against 400 grit sandpaper to reduce preferred orientation. Patterns were collected on the D8 Focus using a Co X-ray tube operated at 35 kV and 40 mA using a LynxEye 1D position-sensitive detector over a 2\u03b8 range of 3-80\u00b0 with a step size of 0.03\u00b0 and a dwell time of 0.7 s\/step. Qualitative analysis was performed by DIFFRAC.EVA V.4.2 (Bruker AXS) using the ICDD PDF-4+ database. Quantitative analysis was performed using DIFFRAC.TOPAS V.4.2 (Bruker AXS) using the fundamental parameters approach (Cheary & Coelho, 1992). The crystal structure data sources are displayed in Table A2.1.244  Table A2.1. Crystal structure data sources for the quantitative XRD analysis. Mineral Source Brucite (National Bureau of Standards, 1956) Serpentine (Viti & Mellini, 1997) Forsterite (National Bureau of Standards, 1984) Smectite (Rosenquist, 1959) Magnetite (National Bureau of Standards, 1967) Diopside (Federico et al., 1988) Phlogopite (Smith, 1956) Clinochlore (Gillery, n.d.) Calcite (Falini et al., 1998) Quartz (Kern & Eysel, 1993) Albite (Goodyear, 1954) K-Feldspar (Bailey, 1955; University of Oxford, 1967)  Talc (Brindley, 1977) Dolomite (Ross & Reeder, 1992) Cuspidine (Saburi et al., 1977) Andradite (Tsao, 1964) Wollastonite (Smith, 1974) Grossular (Pabst, 1937) Tremolite (Comodi et al., 1991) Nesquehonite (Giester et al., 2000) Hydromagnesite (Bariand, 1973) Magnesite (National Bureau of Standards, 1957) Ferrihydrite (Chukhrov, 1973) Pyroaurite (Allmann, 1966) Corundum (Finger & Hazen, 1978)  245  A2.2.1.3 Thermogravimetric Analysis A Perkin Elmer TGA 4000 with a Polyscience chiller and an AS 6000 autosampler performed the thermo-gravimetric analysis at the University of British Columbia. N2 was used as an inert carrier gas and was pumped at a rate of 19.8 mL min-1. Samples (~50 mg) were heated from 20 or 100\u00b0C to 900\u00b0C at a rate of 10\u00b0C per minute. Recordings of the temperature and sample mass were taken every second. TG curves of mass versus temperature were plotted as wt.%. DTG curves were plotted as every 50th data point to smoothen out the curves. Brucite is qualitatively identified due to a signature mass loss in the temperature interval from 300 to 450\u00b0C. Brucite decomposition releases one mole of water per mole of brucite according to Equation A2.1. Equation A2.1. !\"($%)! \t\u2192 \t!\"$\t +\t%!$(#) Brucite abundance was quantified by using an exponential function to model the background mass loss from other minerals. This improves the determination of the mass loss attributable to brucite. The derivative mass loss data before and after the brucite interval was used to fit the exponential function of the form shown in Equation A2.2. Equation A2.2. * +%&.%\u00b0* , = .+,(\u00b0*) + \/-,(\u00b0*) Where \u2018y\u2019 is the derivative mass loss, \u2018x\u2019 is the temperature, and \u2018a\u2019, \u2018b\u2019, \u2018c\u2019, and \u2018d\u2019, are constants determined by the \u2018fit\u2019 function in MATLAB R2020a (MathWorks). The fit for the DTG curve was then projected back onto the TG curve. This fitted function matches the slope of the provided data and enables the quantification of mass loss attributable to brucite. This calculation is shown in Equation A2.3. 246  Equation A2.3. 012\/345.,\/(64.%) = 9. :; \u00b7 +=.,\/(64.%) \u2212=012(64.%), Where \u2018BruciteExp\u2019 is the exponential method\u2019s determined brucite abundance, \u2018mExp\u2019 the exponentially projected mass in wt.% if there had been no brucite in the sample, \u2018m450\u2019 the sample mass in wt.% at the end of the characteristic brucite mass loss interval (~450\u00b0C), and \u20183.24\u2019 the stoichiometric factor between water and brucite. A2.2.1.4 Total Inorganic Carbon To measure the total inorganic carbon (TIC) content of the raw and carbonated samples, a CM5130 acidification module with a Model 5014 CO2 Coulometer from UIC Inc. was used. Micronized and homogenized samples were acidified, releasing gaseous CO2, which was measured using a photodetector to determine the colour change of a colorimetric pH indicator. The instrument was tested with calcium carbonate standards to ensure calibration prior to samples being analyzed. All data are expressed as wt.% CO2. This method has a detection limit better than 0.37 wt.% CO2 and measurements were repeatable within 0.03 wt.% CO2. A2.2.1.5 Major Oxide Composition Inductively coupled plasma-atomic emission spectroscopy (ICP-AES) was used for PK samples and CPS samples to determine the major oxide composition, with detection limits of 0.01 wt.%. For the FPS, this was done on duplicate samples and was determined by X-ray fluorescence (XRF) spectroscopy, again with detection limits of 0.01 wt.%. A2.2.1.6 Particle Size Distribution The particle size distribution of the FPS and the FPK was determined using a Malvern Mastersizer 2000 Laser Diffraction Particle Size Analyzer. A suspension of 5% solids in distilled water was prepared and mixed before an aliquot was withdrawn for analysis. Ultrasound treatment 247  of 60 seconds was applied to disaggregate particle clusters. The particle size distribution for the CPS and CPK fraction above 425 \u00b5m were found by mechanical sieving. A2.2.1.7 Surface Area Multipoint BET with N2 adsorption was performed with a Quantachrome Autosorb-1 surface area analyzer on the original sample materials to determine their respective grain surface areas. A2.2.1.8 Scanning Electron Microscopy Scanning electron microscopy (SEM) samples were prepared as polished thin sections, and aggregates of carbonated tailings. Thin sections were coated with C. Aggregates were coated with Pt. Elemental analysis was performed by an EDAX Team Pegasus system for energy dispersive X-ray spectroscopy (EDS) operating at 10.0 kV. High-resolution images were taken using an FEI Helios NanoLab 650 FIB SEM at 10.0 kV. A2.2.1.9 Quantitative Evaluation of Materials by SEM Samples of CPK-1 and CPK-2 were analyzed using Quantitative Evaluation of Materials by Scanning Electron Microscopy (QEMSCAN). The samples were prepared as 3 g samples embedded in epoxy on a 30 mm mould. The samples were cut through and polished with a carbon coating. Maps were produced from back-scatter electron signals and electron dispersive X-ray spectroscopy on individual points taken with a stepping interval of 10 microns. The accelerating voltage was 20 kV, and the beam current was 10 nA. The QEMSCAN image analysis software iDiscover combined individual points into mineralogical maps based on elemental concentrations. 248  A2.2.2 Pneumatic Permeability Pressure recordings were made for three flowrates per sample. During periods where the pressure had stabilized, the average inlet and outlet pressures were calculated. The coefficient of permeability was then calculated using Equation A2.4. Equation A2.4. ?#(=!) = !34!\"56(78\u00b7:);(<)7#(78)=><#?47$#>78#?@7##>78#?5 Where \u2018Kg\u2019 is the intrinsic gas permeability, \u2018Q\u2019 is the volumetric flowrate, \u2018\u00b5\u2019 is the fluid viscosity, \u2018L\u2019 is the sample length, \u2018A\u2019 is the cross-sectional area, \u2018P1\u2019 is the inlet pressure, and \u2018P2\u2019 is the outlet pressure. The mean permeability for the three tests was taken to be the final result for the gas permeability.  A2.2.3 CO2 Mass Balance Recorded CO2 concentrations were used to perform a mass balance on the amount of sequestered CO2 in the porosity, in the aqueous phase and as mineralized CO2. This was accomplished by iteratively determining the difference between the injected volume of CO2 (Equation A2.5) and the effluent volume of CO2 (Equation A2.6) for a specified time step (Equation A2.7), which yielded the carbonation rate (Equation A2.8). The sequestered mass was then found by summing the product of the carbonation rate with the time step duration (Equation A2.9). Equation A2.5. @ABCDE&D-(=A) = B+<;<FB, \u00b7 \u22064(=3D) \u00b7 [F$!]ABGD&(%) Equation A2.6. @.HHGIDB&(=A) = B+<;<FB, \u00b7 \u22064(=3D) \u00b7 H J@[*L#]%&'()(%)J@[*L#]*+)'()(%) \u2212 IJ \u2212 [F$!]ABGD&(%)KL 249  Equation A2.7. F$!*8\/&IND-\t(%) = J \u2212 O,--'+(&)(<;)O%&.(\/)(0(<;) Equation A2.8. F$!*8\/P8&D \t+<#<FB, = F$!*8\/&IND-(%) \u00b7 B+<;<FB, \u00b7 ==*L#(<#<QG) \u00b7 7(8&<)P41!\u00b73)14\u00b715' 5\u00b7R(S) Equation A2.9. F$!T8::U8G8BED\t(=\") = \u2211F$!*8\/P8&D +<#<FB, \u00b7 \u22064(=3D) Where \u2018V!\"#$%&$'\u2019 is the injected volume of CO2, \u2018Q\u2019 is the injected flow rate, \u2018\u2206t\u2019 is the differential time between CO2 sensor measurements, \u2018[CO(]!\")$&\u2019 is the inlet CO2 concentration, \u2018V*++),$\"&\u2019 is the effluent volume of CO2, \u2018[CO(]-,&)$&\u2019 is the outlet CO2 concentration, \u2018CO(.\/01\/&$\u2019 is the carbonation rate, \u2018mm.-V\u2019 is the molar mass of CO2, \u2018P\u2019 is the average ambient pressure, \u2018R\u2019 is the ideal gas constant, and \u2018T\u2019 is the average ambient temperature. A2.2.3.1 Correcting Mass Balance to Mineralized Carbon The gas mass balance accounts for the total mass of sequestered CO2, including that in the porosity and aqueous phase. Porosity held CO2 can be discounted by determining the sample porosity using the sample density and the known volume of water, and then calculating the volume of CO2 in the pore space given the known concentration and converting the volume to mass (Equation A2.10). Equation A2.10. F$!7QNQ:F&W\t(=\") = H+A(\/=) \u00b7 N(\/=!), \u2212 O%H(=A) \u2212 <\"(#)X\"(#\u00b7E<67L \u00b7 [F$!](%) \u00b7 <<8*#(1915')\u00b77(8&<)P41!\u00b73)14\u00b715' 5\u00b7R(S)  Where \u2018CO(234356&7\u2019 is the mass of CO2 held in the pore space, \u2018L\u2019 the length of the sample, \u2018A\u2019 the cross-sectional area of the sample, \u2018v8+\u2019 the final mass of water, \u2018m5\u2019 the mass of initial dry solids, \u2018\u03c15\u2019 the sample density, \u2018[CO(]\u2019 the concentration of CO2 in the injected gas, \u2018mm.-V\u2019 the molar mass of CO2, \u2018P\u2019 the ambient pressure, \u2018R\u2019 the ideal gas constant, and \u2018T\u2019 the gas temperature. 250  The CO2 held as alkalinity is not considered a stable form of carbon sequestration in mine tailings as the DIC can be released to the atmosphere due to changes in temperature or pH. Due to the limited reactivity of the tailings, it has been conservatively assumed that the pore water equilibrated with the CO2 at a pH of 8.8 (PK) or 9.0 (PS). At 25\u00baC, this means the water has a DIC of 0.01 to 0.02 M. Knowing the final moisture content of the tailings enables the mass of aqueous CO2 to be calculated (Equation A2.11). Equation A2.11. F$!(8Y)(=\") = PQF(!) \u00b7 @%H(A) \u00b7 ==*L#(<#<QG) Where \u2018CO((\/:)\u2019 is the CO2 mass in the aqueous phase and \u2018DIC\u2019 is the dissolved inorganic carbon concentration in the aqueous phase. Then the mass balance (CO(<\/55=\/)\/\"%$) can be corrected to account for only the mineralized carbon (CO(<6\"$4\/)6>$') as shown in Equation A2.12. Equation A2.12. F$!TFBDN8GFZD-(=\") = F$!T8::U8G8BED(=\") \u2212 F$!7QNQ:F&W(=\") \u2212 F$!(8Y)(=\") A2.2.3.2 Mass Balance Error Calculation The error envelope around the gas mass balance was calculated by using the errors stipulated for the Vaisala GMP 251 CO2 sensors (\u00b1 0.2 vol.% if < 8 vol.% and \u00b1 0.4 vol.% if > 8 vol.%), and the flow rate from the mass flow controllers (27 \u00b1 1 mL min-1, 20 \u00b1 1 mL min-1, 12 \u00b1 0.5 mL min-1, 5.5 \u00b1 0.2 mL min-1, 1 \u00b1 0.1 mL min-1, and 0.9 \u00b1 0.1 mL min-1). This was done using the same approach to calculate the sequestered mass of CO2, as shown in Equations A2.5 \u2013 A2.9. The difference was that the inlet and outlet concentrations and the flow rate was adjusted based on the manufacturer\u2019s uncertainty or based on the uncertainty experienced during calibration of the mass flow controllers. 251  To calculate the maximum amount of sequestered carbon, the positive uncertainty was added to the injected flow rate and each inlet concentration, while the negative error was added to the outlet concentrations (Equations A2.13 \u2013 A2.15). These inputs were then used to calculate the mass of sequestered CO2 over time. Equation A2.13. [F$!]ABGD&@T8,(ORS.%) = [F$!]ABGD&(ORS.%) + T!U:VJ.NNQN(ORS.%) Equation A2.14. [F$!]LI&GD&@T8,(ORS.%) = [F$!]LI&GD&(ORS.%) \u2212 T!U:VJ.NNQN(ORS.%) Equation A2.15. WSR6\tX.45T8,(AU!) = WSR6\tX.45(AU!) + YA\tWSR6\tU15Z43\"5.NNQN(AU!) To calculate the minimum amount of sequestered carbon, the negative uncertainty was added to the injected flow rate and each inlet concentration, while the positive error was added to the outlet concentrations (Equations A2.16 \u2013 A2.18). These inputs were then used to calculate the mass of sequestered CO2 over time. Equation A2.16. [F$!]ABGD&@T8,(ORS.%) = [F$!]ABGD&(ORS.%) \u2212 T!U:VJ.NNQN(ORS.%) Equation A2.17. [F$!]LI&GD&@T8,(ORS.%) = [F$!]LI&GD&(ORS.%) + T!U:VJ.NNQN(ORS.%) Equation A2.18. WSR6\tX.45T8,(AU!) = WSR6\tX.45(AU!) \u2212 YA\tWSR6\tU15Z43\"5.NNQN(AU!)  A2.2.4 Total Inorganic Carbon Increase To assess the increase in TIC due to carbon mineralization, the initial material was representatively sampled to establish an average value and standard deviation. After CO2 injection, the TIC measurements determined the mean and standard deviation for the carbonated material. 252  Initial, unreacted TIC was subtracted from the final, reacted TIC to determine the TIC increase, as shown in Equation A2.19. Equation A2.19. TICIncrease (wt.%) = TICReacted (wt.%) \u2013 TICUnreacted (wt.%) To ensure that the TIC increase was statistically significant, a one-tailed t-test was performed on the difference in the means to determine the degree of confidence to which the means are statistically different. For the difference between the means, that is, the TIC increase, a 95% confidence interval (CI) was calculated based on the Standard Error (SE) for the TIC increase, and the T-value based on the degrees of freedom and the selected 95% degree of confidence, as shown in Equation A2.20. Equation A2.20. CI = SE\u00b7T-value  A2.2.5 Evaporative Moisture Loss The expected evaporation during the reactivity tests was calculated by assuming an average difference in relative humidity across the sample of 10% and an average temperature of 20\u00b0C, with the amount of injected gas being known based on the flow rate and injection duration. The difference in relative humidity across the sample was presumed to be small due to the hydration of the gas prior to injection. Based on the saturation vapour pressure of water in the air, the evaporative losses could be calculated, as shown in Equation A2.21. Equation A2.21. 6D = \u2206P\\\u00b7\/:\"3)\u00b7<<:P\u00b7R\u00b7J222 \u00b7 ORS#8:;&. 253  Where \u2018\u2206RH\u2019 is the difference in relative humidity across the sample (%), \u2018p8]^_\u2019 is the saturation vapour pressure (Pa), \u2018mm8\u2019 is the molar mass of water (g\u00b7mol-1), and \u2018vol?\/5`ab\u2019 is the total volume of injected gas over the course of the experiment (L). 254  A2.3 Detailed Results A2.3.1 Sample Characterization A2.3.1.1 X-Ray Diffraction Representative diffractograms of each unreacted sample are presented in Figures A2.1 to A2.7.  Figure A2.1. Representative XRD pattern and Rietveld refinement of unreacted FPS. 2Th Degrees8075706560555045403530252015105Counts40,00035,00030,00025,00020,00015,00010,0005,0000-5,000-10,000-15,00020DCR EW FPCOMP S2.raw_1 Brucite 10.05 %Magnetite 0.72 %Forsterite 3.80 %Diopside 1.65 %Mellini & Viti Lizardite 1T P31M 83.78 %FPS255   Figure A2.2. Representative XRD pattern and Rietveld refinement of unreacted CPS-1.  2Th Degrees8075706560555045403530252015105Counts19,00018,00017,00016,00015,00014,00013,00012,00011,00010,0009,0008,0007,0006,0005,0004,0003,0002,0001,0000-1,000-2,000-3,000-4,000-5,000-6,000-7,000-8,00019FPX-LBC-1.raw_1 Diopside 1.86 %Magnetite 6.82 %Forsterite 6.97 %Mellini & Viti Lizardite 1T P31M 84.36 %C S-1256   Figure A2.3. Representative XRD pattern and Rietveld refinement of unreacted CPS-2. 2Th Degrees8075706560555045403530252015105Counts15,00014,00013,00012,00011,00010,0009,0008,0007,0006,0005,0004,0003,0002,0001,0000-1,000-2,000-3,000-4,000-5,000-6,00019FPX-MBC-1.raw_1 Diopside 1.87 %Brucite 3.30 %Magnetite 5.36 %Forsterite 5.27 %Mellini & Viti Lizardite 1T P31M 84.22 %CPS-2257   Figure A2.4. Representative XRD pattern and Rietveld refinement of unreacted FPK-1.  2Th Degrees8075706560555045403530252015105Counts8,0007,5007,0006,5006,0005,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,500-3,00019GKFPKB2-1-CAT.raw_1 Phlogopite 1M 12.49 %Clinochlore IIb-4 68942 8.81 %Calcite 0.81 %Quartz low 2.09 %Magnetite 1.13 %Microcline intermediate 8.09 %Albite low 5.37 %Talc 1A 7.58 %Dolomite 0.19 %Diopside 2.48 %Forsterite 1.04 %Cuspidine 0.72 %Mellini & Viti Lizardite 1T P31M 31.06 %Ca-Montmorillonite 18.12 %FPK-1258   Figure A2.5. Representative XRD pattern and Rietveld refinement of unreacted CPK-1. 2Th Degrees8075706560555045403530252015105Counts7,0006,5006,0005,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,50019 GK CPK B3-1 CAT.raw_1 Phlogopite 1M 12.05 %Clinochlore IIb-4 68942 11.03 %Calcite 0.95 %Quartz low 3.22 %Magnetite 1.31 %Microcline intermediate 10.94 %Albite low 6.55 %Talc 1A 5.57 %Diopside 2.96 %Forsterite 2.71 %Mellini & Viti Lizardite 1T P31M 27.65 %Ca-Montmorillonite 15.08 %CPK-1259   Figure A2.6. Representative XRD pattern and Rietveld refinement of unreacted FPK-2.  2Th Degrees8075706560555045403530252015105Counts8,5008,0007,5007,0006,5006,0005,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,500-3,000IBC 2-1 FPK CAT 20GKBS.raw_1 Phlogopite 1M 10.71 %Clinochlore IIb-4 68942 7.50 %Calcite 1.29 %Quartz low 5.33 %Magnetite 1.25 %Albite low 9.67 %Microcline intermediate 12.84 %Talc 1A 2.91 %Dolomite 0.76 %Diopside 4.49 %Forsterite 6.04 %Mellini & Viti Lizardite 1T P31M 27.54 %Ca-Montmorillonite 9.69 %FPK-2260   Figure A2.7. Representative XRD pattern and Rietveld refinement of unreacted CPK-2. 2Th Degrees8075706560555045403530252015105Counts5,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,50020gkbs-cpk-lscinj-pre-1-cat.raw_1 Phlogopite 1M 17.07 %Clinochlore IIb-4 68942 6.11 %Calcite 0.56 %Quartz low 3.00 %Magnetite 1.48 %Albite low 7.24 %Orthoclase 6.29 %Talc 1A 3.93 %Dolomite 0.33 %Andradite 1.71 %Cuspidine 1.07 %Diopside 5.58 %Forsterite 9.74 %Mellini & Viti Lizardite 1T P31M 25.92 %Ca-Montmorillonite 9.97 %CPK-2261  A2.3.1.2 Particle Size Distribution The particle size distribution for all unreacted samples is plotted in Figures A2.8 and A2.9.  Figure A2.8. Particle size distributions for FPS, FPK-1, FPK-2, and CPK-1 and CPK-2 sieved <425 \u00b5m, as determined by a Malvern Mastersizer 2000 Laser Diffraction Particle Size Analyzer.  Figure A2.9. Particle size distributions from mechanical sieving for CPS-1, CPS-2, CPK-1 and CPK-2 >53 \u00b5m. 262  A2.3.1.3 Quantitative Evaluation of Materials by SEM The results of the QEMSCAN characterization for samples CPK-1 and CPK-2 are presented in Table A2.2. Table A2.2. QEMSCAN mineralogy abundances. Mineral Abundance (wt.%) CPK-1 CPK-2 Quartz 3.94 5.15 Alkali Feldspar 2.53 1.44 K Feldspar 12.78 6.78 Albite 1.80 0.64 Calcic Plagioclase 1.33 0.02 Muscovite 0.12 0.45 Biotite 0.51 2.34 Phlogopite 15.00 15.83 Illite 0.32 0.76 Chlorite 0.87 1.43 Saponite 4.64 5.28 Diopside 4.79 16.38 Olivine 1.25 2.66 Talc 12.45 2.85 Serpentine 14.27 14.27 Serpentine (Mg-rich) 2.55 3.14 Fe Serpentine 15.41 8.49 Al Serpentine 1.36 4.87 Calcite 0.23 1.73 Dolomite 0.36 0.06 Fe Oxides 0.39 0.33 Chromite 0.13 0.17 Epidote 0.09 0.14 263  Table A2.2 continued. Mineral Abundance (wt.%) CPK-1 CPK-2 Fe Sulphides 0.13 0.03 Wollastonite 0.37 1.41 Andradite 0.02 0.21 Cuspidine 0.01 0.59 Melilite 0.02 1.69 Ti Oxides 0.27 0.29 Apatite 2.01 0.46 Zircon 0.03 0.03 Barite 0.04 0.09 Undifferentiated 0.01 0.01 Total 100.00 100.00  A2.3.1.4 Thermogravimetric Analysis Representative TG and DTG curves have been plotted for each unreacted sample in Figure A2.10.264   Figure A2.10. Representative TG and DTG curves of unreacted FPS, CPS-1 and CPS-2 (left) and FPK-1, CPK-1, FPK-2 and CPK-2 (right). Brucite is indicated by a distinct peak from 300 to 425\u00b0C. In the left plot, peaks at 600-750\u00b0C are from lizardite or antigorite dihydroxylation. In the right plot, peaks at 60, 575-675, and 800\u00b0C are due to adsorbed water, clinochlore, serpentine, and smectite dehydroxylation, and talc dehydroxylation, respectively.265  A2.3.2 Pneumatic Permeability The test conditions and results from each permeability measurement for the fine to coarse grain size distribution at 30% pore saturation are presented in Table A2.3. Conditions and results for the distribution at 60% pore saturation are provided in Table A2.4, and the same is included for the asbestos-bearing serpentinite samples in Table A2.5. The relative change in permeability after CO2 injection was found as shown in Equation A2.22. Equation A2.22. \u2206\"#$%#&'()(*+\t(%) = !\"#$#%&\t()*+)%,#&#$-\t.+!\/01#\"%&\t()*+)%,#&#$-\t.+!\/!\"#$#%&\t()*+)%,#&#$-\t.+!\/  266  Table A2.3. Permeability test conditions and results on mixed CPK-Perm and FPS at ~30% pore saturation, using air, and with an area of 25 cm2. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length(cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 0     17.9 5.69 1022.9 1020.1 1017.5 4.4E-10 5.7E-10 8.31 1027.5 1023.1 1017.0 4.0E-10     17.1 8.31 1017.7 1015.1 1009.4 6.7E-10 5.69 1013.9 1012.3 1009.4 7.5E-10 1 28 28 6.2 1.75 17.5 5.69 1007.9 1005.6 1002.8 5.3E-10 5.7E-10 8.31 1011.8 1008.3 1002.8 5.0E-10 2.6 1004.3 1003.5 1002.8 7.0E-10 3 29 28 6.4 1.75 17.7 2.6 1005.5 1003.1 1002.6 2.3E-10 3.6E-10 5.69 1011.7 1009.0 1006.0 4.5E-10 8.31 1016.6 1012.2 1006.0 4.0E-10 5 30 25 5.8 1.83 17.3 5.69 1014.9 1009.9 1007.0 2.4E-10 2.5E-10 8.31 1021.0 1012.9 1007.0 2.1E-10 2.6 1009.7 1007.9 1007.0 3.0E-10 267  Table A2.3 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 7.5 29 25 5.5 1.84 17.3 2.6 998.0 993.2 992.9 1.1E-10 2.8E-10 5.69 1000.6 997.6 992.0 3.9E-10 8.31 1010.4 1005.3 992.0 3.3E-10 10 29 24 5.0 1.89 17.5 5.69 1013.3 1005.1 1003.3 1.5E-10 2.0E-10 8.31 1020.5 1012.0 1003.3 2.1E-10 2.6 1006.4 1004.2 1003.3 2.5E-10 12.5 28 23 4.8 1.90 17.7 5.69 994.4 989.7 987.1 2.6E-10 2.4E-10 8.31 1000.2 992.2 987.1 2.2E-10 15 28 22 4.2 1.96 17.8 0.369 999.9 999.3 998.9 1.2E-10 1.1E-10 5.69 1013.2 1001.8 999.1 1.1E-10 8.31 1023.1 1005.0 999.1 9.8E-11 2.6 1004.6 999.8 999.1 1.2E-10 17.5 29 22 4.5 1.94 18.3 5.69 1012.1 1003.0 1000.5 1.4E-10 1.3E-10 8.31 1019.9 1006.1 1000.5 1.3E-10 268  Table A2.3 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 20 25 23 4.0 1.93 19.2 0.143 995.1 994.1 994.4 3.2E-11 3.5E-11 0.233 995.6 994.1 994.4 3.6E-11 0.369 996.5 994.2 994.4 3.6E-11 5.69 1032.6 997.6 994.0 3.7E-11 8.31 1056.0 1002.0 994.0 3.4E-11 2.6 1011.5 994.7 994.0 3.6E-11 22.5 26 24 4.2 1.91 19.6 0.143 1016.4 1015.5 1014.8 4.1E-11 4.5E-11 0.233 1016.8 1015.6 1014.8 4.6E-11 25 24 4.2 1.90 19.7 0.369 1017.7 1015.7 1014.8 4.4E-11 0.143 1017.2 1016.5 1015.8 4.7E-11 0.233 1017.7 1016.5 1015.8 4.9E-11 25 26 23 4.2 1.93 19.6 0.143 1025.0 1023.9 1023.0 3.3E-11 2.7E-11 0.223 1025.6 1023.9 1023.0 3.1E-11 27 23 4.4 1.93 18.7 0.369 1026.7 1023.9 1023.0 3.1E-11 0.143 1025.8 1024.2 1023.6 2.0E-11 0.223 1026.7 1024.2 1023.6 2.0E-11 269  Table A2.3 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 27.5 27 22 4.3 1.94 19.1 0.143 1009.2 1006.3 1005.6 1.1E-11 1.1E-11 0.223 1011.0 1006.2 1005.6 1.1E-11 27 23 4.3 1.93 19.2 0.369 1013.9 1006.2 1005.6 1.1E-11 0.143 1009.1 1006.1 1005.4 1.1E-11 0.223 1010.9 1006.1 1005.4 1.1E-11 30 29 21 4.2 1.98 18.9 0.369 1053.5 1006.4 1006.3 1.7E-12 2.1E-12 0.223 1024.5 1006.4 1006.3 2.8E-12 0.068 1014.9 1006.3 1006.3 1.8E-12 35 27 24 4.6 1.90 18.8 0.068 1026.1 1015.4 1015.0 1.4E-12 1.3E-12 0.233 1052.9 1015.4 1015.0 1.4E-12 0.02 1018.1 1015.5 1015.0 1.8E-12 27 23 4.5 1.91 19.0 0.068 1029.1 1015.8 1015.6 1.2E-12 0.143 1044.1 1015.8 1015.6 1.1E-12 0.233 1061.1 1015.8 1015.6 1.2E-12 270  Table A2.3 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 40 29 23 4.8 1.92 19.5 0.143 1093.8 1022.5 1023.3 4.6E-13 3.9E-13 0.068 1057.3 1022.5 1023.3 4.5E-13 0.02 1032.7 1022.5 1023.3 4.6E-13 30 22 4.7 1.94 19.3 0.068 1073.1 1022.8 1023.5 3.1E-13 0.02 1037.3 1022.8 1023.5 3.2E-13 0.025 1040.9 1022.8 1023.5 3.2E-13 45 32 22 5.2 1.93 19.1 0.068 1090.5 1021.2 1021.9 2.2E-13 2.1E-13 0.02 1041.3 1021.2 1021.9 2.3E-13 0.025 1047.0 1021.2 1021.9 2.2E-13 31 22 5.4 1.91 18.8 0.068 1096.0 1020.9 1021.6 2.0E-13 0.02 1041.9 1020.9 1021.6 2.1E-13 0.025 1048.6 1021.0 1021.6 2.0E-13 271  Table A2.3 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 50 34 23 6.4 1.85 20.1 0.068 1028.9 1004.0 1004.8 6.6E-13 4.8E-13 0.02 1010.5 1004.1 1004.8 7.5E-13 36 22 6.3 1.90 19.8 0.068 1052.6 1004.0 1004.8 3.3E-13 0.02 1017.8 1004.0 1004.8 3.4E-13 0.025 1022.3 1004.0 1004.8 3.2E-13 70 36 24 7.3 1.82 18.5 0.143 1089.0 1016.0 1015.9 4.2E-13 4.2E-13 0.068 1052.5 1015.9 1015.9 4.1E-13 0.02 1026.3 1015.9 1015.9 4.3E-13 100     16.2 0.02 1056.4 984.7 985.1 5.5E-14 4.8E-14 0.012 1031.3 984.6 985.1 5.1E-14 0.025 1080.5 984.7 985.1 5.2E-14     16.2 0.02 1055.1 984.6 984.6 5.6E-14 0.012 1030.9 984.6 984.6 5.1E-14 0.025 1076.2 984.6 984.6 5.4E-14     15.6 0.012 1054.7 996.2 996.2 3.9E-14 0.02 1083.9 996.1 996.2 4.3E-14 272  Table A2.4. Permeability test conditions and results on mixed CPK-Perm and FPS at ~60% pore saturation, using air, and with an area of 25 cm2. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 7.5 56 15 10.4 1.93 17.3 0.143 1002.3 994.9 993.7 4.0E-12 3.3E-12 0.223 1010.2 995.0 993.7 3.0E-12 0.068 997.2 994.9 993.7 6.4E-12 0.068 1000.6 995.0 994.3 2.5E-12 0.143 1003.2 995.4 994.3 3.8E-12 0.223 1005.1 995.8 994.3 5.0E-12 56 16 10.5 1.92 17.3 0.068 1003.9 995.3 995.6 1.7E-12 0.143 1007.8 995.3 995.6 2.4E-12 0.02 996.9 995.3 995.6 2.5E-12 0.068 1002.6 994.5 995.6 1.8E-12 10 55 16 9.8 1.94 17.3 0.068 1021.6 1009.4 997.3 1.1E-12 3.0E-12 0.143 1048.3 1031.2 997.3 1.7E-12 0.233 1088.0 1066.2 997.3 2.1E-12 0.143 1004.7 996.8 998.2 3.8E-12 0.223 1010.2 996.8 998.2 3.4E-12 0.369 1021.3 996.8 998.2 3.1E-12 273  Table A2.4 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 10 61 12 9.9 2.01 15.9 0.369 1033.3 1009.1 1008.2 2.9E-12 3.0E-12 0.223 1022.3 1009.1 1008.2 3.3E-12 0.143 1016.9 1009.1 1008.2 3.5E-12 57 14 9.9 1.96 16.2 0.369 1031.3 1008.6 1007.8 3.2E-12 0.223 1020.6 1008.4 1007.8 3.6E-12 0.143 1014.5 1008.4 1007.8 4.6E-12 15 59 13 8.9 2.04 16.2 0.369 1010.0 1005.4 1004.1 1.6E-11 2.2E-11 0.223 1007.9 1005.4 1004.1 1.7E-11 0.143 1006.8 1005.4 1004.1 2.0E-11 58 13 8.9 2.03 16.2 0.369 1007.9 1004.8 1003.6 2.3E-11 0.223 1006.5 1004.7 1003.6 2.5E-11 0.143 1005.7 1004.7 1003.6 2.8E-11 21 60 11 7.6 2.12 16.4 0.369 1022.2 1018.3 1017.5 1.9E-11 1.4E-11 0.223 1020.5 1017.5 1017.5 1.4E-11 0.143 1019.4 1016.6 1017.5 1.0E-11  274  Table A2.4 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 21 59 11 7.6 2.11 16.4 0.369 1023.4 1018.6 1017.7 1.5E-11 1.4E-11 0.223 1021.3 1018.6 1017.7 1.6E-11 0.143 1020.2 1016.8 1017.7 8.4E-12 25 56 12 7.3 2.10 17.3 0.369 1008.4 1006.5 1005.6 4.0E-11 4.9E-11 0.223 1007.4 1006.5 1005.6 5.2E-11 0.143 1006.9 1006.5 1005.6 7.5E-11 55 13 7.3 2.09 17.4 0.369 1008.4 1006.4 1005.5 3.8E-11 0.223 1007.5 1006.4 1005.5 4.1E-11 0.143 1007.0 1006.3 1005.5 4.6E-11 30 64 9 7.6 2.16 19.5 0.143 1021.5 1004.0 1004.2 1.9E-12 1.8E-12 0.223 1034.9 1004.1 1004.2 1.7E-12 0.369 1060.7 1004.1 1004.2 1.5E-12 0.068 1011.7 1004.1 1004.2 2.1E-12 275  Table A2.4 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 35 63 10 8.0 2.13 19.2 0.068 1004.6 1002.0 1001.2 6.1E-12 4.6E-12 0.143 1007.4 1002.0 1001.2 6.2E-12 0.223 1011.0 1002.1 1001.2 5.8E-12 0.369 1017.0 1002.1 1001.2 5.7E-12 64 10 8.0 2.14 19.3 0.068 1004.9 999.2 1000.3 2.8E-12 0.143 1008.7 999.2 1000.3 3.5E-12 0.223 1013.4 999.2 1000.3 3.7E-12 0.369 1021.2 999.3 1000.3 3.9E-12 65 9 7.8 2.15 19.5 0.223 1014.7 1003.7 1004.5 4.7E-12 0.143 1011.2 1003.7 1004.5 4.5E-12 0.068 1008.3 1003.7 1004.5 3.5E-12 40 62 11 8.6 2.08 19.5 0.068 1013.9 1009.4 1009.5 3.5E-12 3.5E-12 0.143 1018.7 1009.4 1009.5 3.6E-12 0.223 1024.6 1009.4 1009.5 3.4E-12 0.369 1034.1 1009.4 1009.5 3.5E-12 276  Table A2.4 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 40 61 11 8.4 2.09 19.9 0.068 1016.2 1009.2 1009.4 2.3E-12 3.5E-12 0.143 1023.6 1009.2 1009.4 2.4E-12 0.223 1032.2 1009.2 1009.4 2.3E-12 0.369 1046.6 1009.2 1009.4 2.3E-12 64 10 8.5 2.11 19.4 0.233 1011.4 992.3 992.4 2.8E-12 0.143 1004.0 992.3 992.4 2.9E-12 0.068 998.0 992.3 992.4 2.8E-12 62 11 8.3 2.10 19.9 0.223 1012.2 992.2 992.4 2.7E-12 0.143 1004.7 992.2 992.4 2.7E-12 0.068 998.3 992.2 992.4 2.7E-12 61 11 8.5 2.08 19.7 0.223 1013.7 1004.1 1004.2 5.5E-12 0.143 1010.0 1004.1 1004.2 5.8E-12 0.068 1007.0 1004.1 1004.2 5.6E-12 61 11 8.3 2.09 20.1 0.223 1016.6 1004.2 1004.3 4.3E-12 0.143 1011.8 1004.1 1004.3 4.5E-12 0.068 1007.8 1004.2 1004.3 4.5E-12 277  Table A2.4 continued. Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 50 64 12 10.4 2.01 20.0 0.223 1053.2 1013.1 1013.2 1.3E-12 1.3E-12 0.143 1037.9 1013.0 1013.2 1.4E-12 0.068 1024.9 1012.8 1013.2 1.3E-12 60 63 12 10.2 2.01 18.5 0.223 1023.5 1004.4 1004.4 2.6E-12 1.6E-12 0.143 1016.2 1004.4 1004.4 2.7E-12 0.068 1010.3 1004.4 1004.4 2.6E-12 63 12 10.0 2.02 18.9 0.223 1064.7 1005.3 1005.3 8.3E-13 0.143 1042.0 1005.3 1005.3 8.7E-13 0.068 1023.5 1005.3 1005.3 8.4E-13 61 13 10.2 1.99 18.7 0.369 1061.8 1005.9 1005.9 1.5E-12 0.223 1040.2 1005.9 1005.9 1.4E-12 0.143 1027.5 1005.9 1005.9 1.5E-12 0.068 1016.3 1006.0 1005.9 1.5E-12 278  Table A2.4 continued. Fines (wt.%) Pore Saturation (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Porosity (%) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean 70 59 14 10.4 1.95 19.2 0.369 1076.2 1016.9 1016.8 1.4E-12 1.7E-12 0.223 1054.3 1016.8 1016.8 1.4E-12 0.143 1040.0 1016.8 1016.8 1.4E-12 56 16 10.8 1.91 19.0 0.369 1057.9 1016.2 1016.3 2.0E-12 0.223 1042.5 1016.1 1016.3 1.9E-12 0.143 1032.5 1016.1 1016.3 2.0E-12 279  Table A2.5. Permeability test conditions on asbestos-bearing serpentinite at ~30% pore saturation, using air, and with a cross-sectional area of 25 cm2. Test Pore Saturation (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Porosity (%) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean PSA-1 37 7.0 1.87 19.0 23 0.5 1013.6 1012 1012 7.2E-11 6.7E-11 1.5 1017.6 1012.5 6.7E-11 2.5 1022.3 1013.1 6.2E-11 PSA-2 34 7.0 1.82 19.9 25 0.5 1035 1010 1010 4.8E-12 4.8E-12 0.4 1029.8 1010 4.8E-12 0.3 1024.7 1010 4.9E-12 PSA-3 34 7.0 1.82 18.9 25 0.5 1014 1009 1008.9 2.3E-11 2.1E-11 0.75 1017.1 1009 2.1E-11 1 1020.4 1009.1 2.0E-11 PSA-4 32 7.0 1.78 19.9 26 0.5 1025.6 1004.5 1004.5 5.6E-12 5.8E-12 0.3 1016.9 1004.4 5.7E-12 0.1 1008.3 1004.3 6.0E-12  280  A2.3.3 CO2 Injection Detailed Results A2.3.3.1 Concentration Profiles The concentration profiles for the experiments that were not presented in Chapter 3 are presented in Figures A2.11 \u2013 A2.13.281   Figure A2.11. Gas-phase CO2 concentrations over time for the processed serpentinite injection experiments PS-E2, PS-E3 and PS-E4. Data loss required interpolating using an exponential function.282   Figure A2.12. Gas-phase CO2 concentrations over time for the processed serpentinite injection experiments PS-E5, PS-E6 and PS-E7. Data loss required interpolating using an exponential function.283   Figure A2.13. Gas-phase CO2 concentrations over time for the processed serpentinite injection experiments PS-E8 and PS-E10 and processed kimberlite injection experiment PK-E1. Data from experiments PS-E10 and PK-E1 were not used to calculate reactivity rates or the sequestered mass of CO2 as the disparity between the reactivity and the injected gas flux yielded significant uncertainty on the estimate.284  A2.3.3.2 Carbon Mineralization Rates From the concentration data, the reaction rates over time were determined and are displayed in Figures A2.14 and A2.15.  Figure A2.14. Rates of CO2 sequestration over time within injection experiments PS-E2, E3 and E4. 285   Figure A2.15. Rates of CO2 sequestration over time within injection experiments PS-E5, E6, E7 and E8.286  A2.3.3.3 Mass of Sequestered CO2 By summing up the mass of sequestered carbon over time, a cumulative mass balance enables an estimate of the sequestered mass of CO2. This is plotted in Figures A2.16 \u2013 A2.19, and the increase from the TIC measurements is also shown. The gas mass balance overestimates the sequestered mass from the TIC increase by ~10%, which is good agreement between methods.  Figure A2.16. Sequestered CO2 mass over time in experiments PS-E2 and E3 as quantified from a gas mass balance and TIC increases in carbonated versus initial samples, with a 95% confidence interval indicated. 287   Figure A2.17. Sequestered CO2 mass over time in experiments PS-E4 and E5 as quantified from a gas mass balance and TIC increases in carbonated versus initial samples, with a 95% confidence interval indicated. 288   Figure A2.18. Sequestered CO2 mass over time in experiments PS-E6 and E7 as quantified from a gas mass balance and TIC increases in carbonated versus initial samples, with a 95% confidence interval indicated. 289   Figure A2.19. Sequestered CO2 mass over time in experiment PS-E8 as quantified from a gas mass balance and TIC increases in carbonated versus initial samples, with a 95% confidence interval indicated.  A2.3.3.4 CO2 Concentration Data The CO2 concentration data required to produce the concentration profiles, calculate the reaction rate and sequestered mass are presented for each experiment in Tables A2.6 \u2013 A2.9.290  Table A2.6. CO2 concentrations for experiments PS-E1, E5 and E6 at 5-hour intervals. Time (Hours) PS-E1 (vol.% CO2) PS-E5 (vol.% CO2) PS-E6 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet 0 0.0 0.0 0.0 0.0 0.0 0.0 5 9.1 3.7 9.1 2.8 8.0 0.0 10 9.2 4.4 9.2 4.0 8.6 0.2 15 9.2 5.5 9.3 4.9 8.8 0.9 20 9.2 5.9 9.3 5.5 8.6 0.9 25 9.3 6.3 9.2 5.6 8.6 1.0 30 9.3 6.6 9.1 5.8 8.6 1.4 35 9.3 6.9 9.1 5.9 8.8 2.5 40 9.3 7.5 9.2 6.0 8.9 3.2 45 9.3 7.0 9.2 6.3 8.8 2.6 50 9.4 7.0 9.1 6.4 8.8 2.7 55 9.4 6.9 9.2 6.5 8.9 3.0 60 9.3 7.4 9.2 6.6 9.0 3.9 65 9.2 8.1 9.2 6.9 9.1 4.3 70 9.3 7.7 9.3 7.2 9.0 3.8 75 9.3 7.3 9.4 7.4 9.1 3.7 80 9.4 7.4 9.4 7.6 9.1 4.2 85 9.4 8.4 9.4 7.7 9.2 5.4 90 9.3 8.3 9.4 7.9 9.3 6.0 95 9.3 7.8 9.4 8.0 9.2 5.5 100 9.2 7.4 9.3 8.1 9.1 5.4 105 9.3 7.6 9.3 8.2 9.2 6.3 110 9.3 8.9 9.3 8.3 9.3 6.9 115 9.2 8.3 9.4 8.6 9.3 7.0 120 9.3 7.9 9.3 8.5 9.2 6.8 125 9.3 7.7 - - 9.2 6.8 291  Table A2.6 continued. Time (Hours) PS-E1 (vol.% CO2) PS-E5 (vol.% CO2) PS-E6 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet 130 9.2 8.0 - - 9.3 7.5 135 9.3 9.1 - - 9.4 7.9 140 9.2 8.2 - - 9.4 7.9 145 9.2 8.0 - - 9.3 7.7 150 9.3 8.0 - - 9.2 7.8 155 9.3 8.3 - - 9.3 8.1 160 9.3 9.0 - - 9.4 8.4 165 9.3 8.5 9.4 9.0 9.4 8.4 170 9.3 7.9 9.4 9.0 9.3 8.4 175 9.3 8.3 9.4 8.9 9.4 8.6 180 9.4 8.6 9.4 9.0   185 9.3 9.1 9.5 9.2   190 9.2 9.0     292  Table A2.7. CO2 concentrations for experiments PS-E2, E4, E8 and E9 at 5-hour intervals. Time (Hours) PS-E2  (vol.% CO2) PS-E4  (vol.% CO2) PS-E8  (vol.% CO2) PS-E9  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5 7.8 0.1 9.6 0.0 4.2 0.0 4.4 0.0 10 8.5 0.7 0.0 0.0 4.6 0.0 5.0 0.0 15 8.6 1.4 0.0 0.0 4.8 0.0 5.4 0.0 20 8.3 1.1 0.0 0.2 5.1 0.0 5.7 0.0 25 8.4 1.4 9.7 0.3 5.2 0.0 5.6 0.0 30 8.9 3.0 9.7 0.7 5.1 0.0 5.7 0.0 35 9.1 4.5 9.7 1.0 5.2 0.0 6.0 0.0 40 8.9 4.5 9.7 1.9 5.4 0.0 6.1 0.0 45 8.8 3.7 9.7 2.3 5.7 0.1 6.3 0.0 50 8.8 3.5 9.7 2.1 5.9 0.2 6.0 0.0 55 8.9 4.5 9.7 2.3 6.0 0.2 6.1 0.0 60 8.9 4.7 9.7 2.7 6.1 0.3 6.4 0.0 65 8.7 3.9 9.7 3.1 6.2 0.4 6.6 0.1 70 8.7 3.8 9.7 3.1 6.4 0.5 6.6 0.1 75 8.7 3.4 9.7 3.3 6.5 0.6 6.6 0.1 80 8.9 4.5 0.0 0.0 6.6 0.8 6.7 0.1 85 9.0 5.0 0.0 0.0 6.7 1.1 6.9 0.1 90 8.9 4.3 0.0 0.0 6.9 1.4 7.2 0.2 95 8.9 4.6 9.8 5.0 7.0 1.8 7.1 0.2 100 8.9 4.7 9.8 4.9 7.0 1.9 7.1 0.2 105 9.1 5.7 0.0 5.1 7.1 2.0 7.1 0.3 110 9.2 6.3 0.0 5.8 7.2 2.3 7.3 0.5 115 9.1 5.9 0.0 6.6 7.4 2.7 7.6 0.8 120 9.0 5.7 0.0 6.4 7.5 2.9 7.5 0.8 125 9.1 6.0 0.0 6.5 7.6 3.0 7.5 1.0 293  Table A2.7 continued. Time (Hours) PS-E2  (vol.% CO2) PS-E4  (vol.% CO2) PS-E8  (vol.% CO2) PS-E9  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 130 9.3 7.0 0.0 7.0 7.6 3.2 7.6 1.3 135 9.3 7.3 0.0 7.4 7.7 3.4 7.8 1.8 140 9.2 6.9 0.0 7.7 7.9 3.8 8.1 2.3 145 9.1 6.8 10.0 7.6 7.9 3.9 8.0 2.4 150 9.1 7.0 9.9 7.4 7.9 4.1 8.0 2.7 155 9.2 7.6 10.0 7.9 7.9 4.2 8.1 2.9 160 9.3 7.9 10.0 8.2 8.0 4.5 8.2 3.4 165 9.2 7.6 10.0 8.6 8.2 5.0 8.5 4.1 170 9.1 7.5 10.0 8.6 8.2 5.0 8.4 4.0 175 9.2 7.7 9.9 8.4 8.2 5.1 8.4 4.2 180 9.4 8.2 10.0 8.5 8.3 5.2 8.4 4.4 185 9.3 8.3 10.0 9.2 8.5 5.7 8.6 4.8 190 9.2 8.0 9.9 8.9 8.6 6.2 8.8 5.5 195 9.2 8.0 9.9 8.8 8.6 6.0 8.7 5.4 200 9.2 8.1 0.0 0.0 8.5 5.9 8.6 5.2 205 9.3 8.4 0.0 0.0 8.6 6.2 8.7 5.7 210 9.2 8.4 0.0 0.0 8.8 6.8 8.9 6.1 215 9.2 8.3 9.7 9.0 8.9 7.0 9.0 6.5 220 9.1 8.2 9.7 9.0 8.8 6.9 8.8 6.3 225 9.2 8.3 9.7 8.8 8.8 6.7 8.8 6.2 230 9.2 8.5 9.7 9.0 8.9 7.0 8.9 6.5 235 9.1 8.5 9.7 9.3 9.1 7.6 9.1 7.0 240 9.1 8.4 9.7 9.2 9.1 7.7 9.1 7.1 245 9.1 8.3 9.7 9.1     250 9.2 8.6 9.7 9.2      294  Table A2.7 continued. Time (Hours) PS-E2  (vol.% CO2) PS-E4  (vol.% CO2) PS-E8  (vol.% CO2) PS-E9  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 255 9.3 8.9 9.8 9.3     260 9.2 8.8 9.9 9.6     265 9.2 8.7       270 9.3 8.7       275 9.3 8.9       280 9.3 9.0       295  Table A2.8. CO2 concentrations for experiments PS-E3, E7, E10 and PK-E1 at 5 or 10-hour intervals. Time (Hours) PS-E3  (vol.% CO2) PS-E7  (vol.% CO2) PS-E10  (vol.% CO2) PK-E1  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5 5.8 0.0 6.0 0.0 9.1 7.1 8.7 6.6 10 5.9 0.0 7.0 0.0 9.2 7.8 9.1 8.1 15 6.1 0.0 7.7 0.1 9.2 8.0 9.2 8.6 20 7.2 0.0 7.6 0.1 9.3 8.1 9.3 8.7 25 7.4 0.0 7.5 0.1 9.4 8.9 9.4 8.7 30 7.1 0.0 7.6 0.1 9.3 8.8 9.3 8.5 35 7.2 0.1 8.0 0.2 9.3 8.8 9.3 8.5 40 7.6 0.1 8.3 0.6 9.4 9.0 9.4 8.4 45 7.9 0.2 7.9 0.6 9.4 9.0 9.4 8.3 50 7.9 0.2 7.7 0.4 9.4 9.0 9.4 8.7 55 7.5 0.2 7.9 0.4 9.4 9.0 9.2 8.7 60 7.5 0.2 8.3 0.8 9.4 9.1 9.3 8.8 65 8.0 0.4 8.5 1.3 9.5 9.2 9.4 8.9 70 8.1 0.6 7.9 1.0 9.4 9.1 9.4 8.9 75 8.0 0.5 8.0 0.8 9.4 9.0 9.4 8.9 80 7.7 0.4 8.2 0.9 9.4 9.0 9.4 9.0 85 7.8 0.5 8.5 1.5 9.4 9.1 9.5 9.0 90 8.2 0.8 8.5 2.1 9.5 9.2 9.4 9.1 95 8.1 0.7 8.2 1.6 9.5 9.2 9.3 9.0 100 8.1 0.7 8.2 1.6 9.5 9.2 9.3 8.9 105 7.9 0.7 8.4 1.8 9.5 9.1 9.3 9.0 110 8.1 0.9 8.7 2.6 9.5 9.1 9.4 9.0 115 8.3 1.2 8.8 3.3 9.5 9.2 9.4 9.1 120 8.3 1.2 8.7 3.4 9.5 9.2 9.5 9.2 125 8.3 1.3 8.7 3.3 9.5 9.1 9.5 9.1 296  Table A2.8 continued. Time (Hours) PS-E3  (vol.% CO2) PS-E7  (vol.% CO2) PS-E10  (vol.% CO2) PK-E1  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 130 8.3 1.6 9.0 3.9 9.5 9.1 9.5 9.1 135 8.4 1.9 9.1 4.9 9.6 9.2 9.4 9.2 140 8.6 2.4 9.1 5.3 9.5 9.2 9.4 9.1 145 8.6 2.5 9.0 5.0 9.4 9.1 9.4 9.1 150 8.7 2.8 9.0 4.8 9.3 8.9 9.5 9.1 155 8.6 2.9 9.1 5.2 9.4 9.0 9.6 9.3 160 8.8 3.6 9.2 5.9 9.4 9.1 9.5 9.4 165 9.0 4.2 9.2 6.1 9.3 9.0 9.4 9.2 170 9.0 4.4 9.1 5.9 9.2 8.9 9.4 9.1 175 8.9 4.5 9.1 5.8 9.2 8.8 9.5 9.1 180 8.9 4.7 9.2 6.3 9.2 8.9 9.4 9.2 185 9.0 5.2 9.3 6.8 9.4 9.1 9.4 9.2 190 9.1 5.7 9.3 7.0 9.4 9.1 9.3 9.1 195 9.1 5.8 9.3 7.0 9.4 9.0 9.3 9.0 200 9.0 5.6 9.3 7.1 9.5 9.1 9.3 9.0 205 9.1 6.0 9.3 7.2 9.5 9.1 9.4 9.1 210 9.2 6.5 9.3 7.5 9.5 9.2 9.5 9.3 215 9.1 6.7 9.3 7.6 9.5 9.1 9.5 9.3 220 9.1 6.5 9.3 7.6 9.5 9.1 9.6 9.3 225 9.1 6.5 9.3 7.6 9.5 9.2 9.6 9.4 230 9.1 6.9 9.3 7.8 9.6 9.2 9.6 9.4 235 9.2 7.3 9.3 8.0 9.5 9.2 9.6 9.4 240 9.2 7.4 9.3 8.0 9.6 9.2 9.5 9.3 245 9.1 7.3 9.3 8.0 9.6 9.2 9.6 9.3 250 9.0 7.0 9.3 8.0 9.6 9.2 9.6 9.4 297  Table A2.8 continued. Time (Hours) PS-E3  (vol.% CO2) PS-E7  (vol.% CO2) PS-E10  (vol.% CO2) PK-E1  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 255 9.1 7.4 9.4 8.2 9.6 9.3 9.6 9.5 260 9.2 7.8 9.4 8.4 9.6 9.3 9.7 9.5 265 9.1 7.7 9.4 8.5 9.5 9.2 9.6 9.4 270 9.1 7.6 9.3 8.3 9.4 9.0 9.7 9.3 275 9.0 7.5 9.3 8.1 9.4 9.1 9.7 9.4 280 - - 9.3 8.3 9.5 9.2 9.7 9.5 285 9.2 8.3 9.3 8.6 9.4 9.2 9.6 9.4 290 9.2 8.2 - - 9.4 9.1 9.5 9.3 295 9.1 7.9 9.3 - 9.4 9.0 9.5 9.2 300 9.1 7.9 9.3 8.3 9.4 9.1 9.5 9.2 305 9.2 8.3 9.4 8.5 9.5 9.2 9.5 9.4 310 9.2 8.4 9.4 8.7 9.4 9.1 9.4 9.3 315 9.3 8.4 9.4 8.7 9.3 9.1 9.5 9.2 320 9.2 8.1 9.4 8.7 9.4 9.0 9.5 9.2 325 9.2 8.1 9.4 8.5 9.4 9.1 9.5 9.3 330 9.3 8.6 9.4 8.7 9.4 9.1 9.5 9.4 335 9.3 8.6 9.5 9.0 9.2 9.0 9.5 9.2 340 9.3 8.6 9.4 8.9 9.2 8.9 9.5 9.2 345 9.2 8.3 9.4 8.7 9.2 8.9 9.5 9.3 350 9.3 8.4 9.4 8.6 9.3 9.0 9.5 9.3 355 9.4 8.9 9.4 8.8 9.5 9.2 9.4 9.2 360 9.3 8.8 9.4 9.0 9.5 9.2 9.1 9.0 365 9.3 8.5 9.3 8.9 9.6 9.2 9.3 9.0 370 9.3 8.5 - - 9.6 9.3 9.4 9.1 375 9.3 8.6 - - 9.7 9.3 9.5 9.3  298  Table A2.8 continued. Time (Hours) PS-E3  (vol.% CO2) PS-E7  (vol.% CO2) PS-E10  (vol.% CO2) PK-E1  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 380 9.4 9.0 9.4 9.0 9.7 9.4 9.6 9.4 385 9.3 8.8 9.4 9.0 9.6 9.3 9.6 9.3 390 9.3 8.6 9.3 8.8 9.6 9.3 9.7 9.5 395 9.2 8.5 9.4 8.8 9.6 9.3 9.7 9.6 400 9.3 8.8 9.4 8.9 9.7 9.4 9.7 9.6 405 9.3 8.9 9.5 9.1 9.7 9.4 9.7 9.5 410 9.2 8.8 9.4 9.1 9.6 9.3 9.7 9.4 415 9.2 8.7 9.4 8.9 9.5 9.2 9.7 9.5 420 9.2 8.6 9.4 8.9 9.5 9.2 9.7 9.5 425 9.3 8.7 9.4 9.0 9.5 9.3 9.7 9.6 430 9.3 8.9   9.5 9.3 9.6 9.5 435 9.3 8.9   9.5 9.2 9.7 9.4 440 9.2 8.6   9.5 9.1 9.6 9.4 445 9.2 8.6   9.5 9.3 9.6 9.4 450 9.3 8.9   9.5 9.1 9.5 9.4 455 9.3 9.0   9.5 9.3 9.5 9.4 460 9.3 8.9   9.4 9.1 9.6 9.2 465 9.2 8.7   9.4 9.1 9.6 9.4 470 9.2 8.7   9.4 9.0 9.6 9.5 475 9.3 8.9   9.4 9.1 9.6 9.5 480 9.3 9.0   9.3 9.1 9.5 9.3 485     9.3 9.0 9.5 9.2 490     9.3 9.0 9.5 9.3 495     9.4 9.1 9.4 9.3 500     9.5 9.3 9.3 9.2  299  Table A2.8 continued. Time (Hours) PS-E3  (vol.% CO2) PS-E7  (vol.% CO2) PS-E10  (vol.% CO2) PK-E1  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 510     9.4 9.0 9.4 9.1 520     9.4 9.2 9.5 9.4 530     9.3 9.0 9.5 9.2 540     9.4 9.2 9.5 9.3 550     9.5 9.2 9.4 9.2 560     9.5 9.2 9.5 9.3 570     9.6 9.4 9.5 9.4 580     9.6 9.3 9.6 9.3 590     9.6 9.3 9.7 9.6 600     9.5 9.3 9.6 9.4 610     9.6 9.3 9.7 9.4 620     9.6 9.4 9.5 9.4 630     9.5 9.2 9.6 9.2 640     9.6 9.4 9.6 9.5 650     9.5 9.3 9.6 9.4 660     9.5 9.2 9.6 9.4 670     9.5 9.3 9.5 9.4 680     9.4 9.1 9.6 9.4 690     9.5 9.4 9.6 9.4 700     9.4 9.1 9.5 9.2 710     9.3 9.0 9.5 9.4 720     9.4 9.2 9.5 9.2 730     9.6 9.3 9.4 9.2 740     9.7 9.5 9.4 9.3 750     9.7 9.4 9.6 9.4  300  Table A2.8 continued. Time (Hours) PS-E3 (vol.% CO2) PS-E7 (vol.% CO2) PS-E10 (vol.% CO2) PK-E1 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 760     9.7 9.5 9.8 9.7 770     9.6 9.4 9.7 9.5 780     9.6 9.4 9.8 9.6 790     9.7 9.5 9.7 9.6 800     9.6 9.3 9.7 9.5 810     9.6 9.5 9.7 9.6 820     9.5 9.3 9.7 9.4 830     9.4 9.2 9.7 9.6 840     9.3 9.1 9.6 9.4 850     9.3 9.0 9.6 9.3 860     9.4 9.3 9.4 9.2 870     9.5 9.2 9.4 9.1 880     9.6 9.4 9.5 9.4 890     9.6 9.3 9.5 9.2 900     9.6 9.3 9.7 9.5 910     9.5 9.3 9.6 9.4 920     9.4 9.2 9.6 9.4 930     9.5 9.4 9.6 9.4 940     9.4 9.2 9.5 9.2 950     9.5 9.2 9.6 9.4 960     9.5 9.3 9.3 9.2 970     9.4 9.1 9.5 9.2 980     9.4 9.2 9.6 9.4 990     9.3 9.1 9.5 9.1 1000     9.4 9.2 9.4 9.3  301  Table A2.8 continued. Time (Hours) PS-E3 (vol.% CO2) PS-E7 (vol.% CO2) PS-E10 (vol.% CO2) PK-E1 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 1010     9.3 9.1 9.4 9.1 1020     9.3 9.1 9.5 9.2 1030     9.4 9.2 9.4 9.1 1040     9.4 9.2 9.4 9.0 1050     9.5 9.4 9.4 9.3 1060     9.5 9.2 9.4 9.1 1070     9.5 9.3 9.5 9.3 1080     9.5 9.3 9.5 9.1 1090     9.5 9.2   1100     9.5 9.3   1110     9.3 9.1   1120     9.4 9.2   1130     9.5 9.3   1140     9.5 9.2   1150     9.5 9.3   1160     9.2 8.9   1170     9.2 9.0   1180     9.1 8.8   1190     9.0 8.7   1200     9.2 8.9   1210     8.9 8.6   1220     9.1 8.9   1230     8.8 8.5   1240     8.8 8.6   1250     9.3 9.0    302  Table A2.8 continued. Time (Hours) PS-E3  (vol.% CO2) PS-E7  (vol.% CO2) PS-E10  (vol.% CO2) PK-E1  (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet Inlet Outlet 1260     9.1 8.8   1270     9.1 8.9   1280     8.7 8.5   1290     8.7 8.5   303  Table A2.9. CO2 concentration results from experiments PK-E2, E3 and E4 at 5 or 10-hour intervals. Time (Hours) PK-E2 (vol.% CO2) PK-E3 (vol.% CO2) PK-E4 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet 0 0.0 0.0 0.0 0.0 0.0 0.0 5 2.8 0.0 4.0 0.0 1.9 0.0 10 3.8 0.2 5.5 0.0 2.5 0.1 15 4.4 0.8 6.2 0.0 2.9 0.4 20 4.9 1.6 6.7 0.2 3.2 0.8 25 5.2 2.3 7.1 0.6 3.6 1.4 30 5.4 2.7 7.2 1.1 4.0 1.9 35 5.6 3.1 7.3 1.7 4.3 2.2 40 5.8 3.3 7.4 2.2 4.6 2.6 45 5.9 3.7 7.6 2.8 4.9 3.0 50 5.9 3.8 7.7 3.4 5.2 3.4 55 5.9 3.9 7.8 3.8 5.2 3.6 60 5.9 4.0 7.8 4.0 5.3 3.7 65 6.0 4.1 7.9 4.3 5.5 3.9 70 6.1 4.3 8.0 4.6 5.8 4.2 75 6.1 4.2 8.1 5.0 6.0 4.5 80 6.1 4.3 8.2 5.3 6.0 4.5 85 6.2 4.4 8.2 5.3 5.9 4.5 90 6.3 4.5 8.2 5.4 6.0 4.6 95 6.3 4.5 8.3 5.7 6.1 4.8 100 6.3 4.5 8.4 5.9 6.2 4.9 105 6.2 4.6 8.3 6.0 6.3 5.0 110 6.3 4.6 8.2 5.8 6.4 5.0 115 6.3 4.6 8.3 5.9 6.5 5.2 120 6.2 4.6 8.3 6.1 6.8 5.5 125 6.2 4.6 8.3 6.3 6.9 5.7  304  Table A2.9 continued. Time (Hours) PK-E2 (vol.% CO2) PK-E3 (vol.% CO2) PK-E4 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet 130 6.2 4.6 8.2 6.2 6.9 5.7 135 6.2 4.6 8.1 5.9 6.9 5.7 140 6.3 4.7 8.2 6.0 7.0 5.8 145 6.2 4.7 8.3 6.3 7.2 6.1 150 6.2 4.6 8.2 6.2 7.2 6.1 155 6.2 4.6 8.2 6.2 7.1 6.0 160 6.3 4.7 8.2 6.1 7.0 5.9 165 6.3 4.8 8.2 6.2 7.0 5.9 170 6.3 4.8 8.3 6.4 7.0 5.9 175 6.3 4.7 8.3 6.5 6.9 5.8 180 6.2 4.7 8.2 6.4 6.8 5.7 185 6.2 4.8 8.2 6.4 6.8 5.7 190   8.2 6.4 6.9 5.8 195   8.2 6.6 7.0 6.0 200   8.2 6.5 6.9 5.9 205   8.2 6.5 6.9 5.8 210   8.2 6.6 6.8 5.8 215   7.8 6.4 7.0 6.0 220   7.8 6.1 7.1 6.1 225   8.0 6.2 7.0 6.0 230   8.2 6.4 6.9 5.9 235   8.3 6.7 7.0 6.0 240   8.3 6.9 7.2 6.2 245   8.2 6.8 7.2 6.2 250   8.2 6.7 7.1 6.2 255   8.3 6.7 7.1 6.1  305  Table A2.9 continued. Time (Hours) PK-E2 (vol.% CO2) PK-E3 (vol.% CO2) PK-E4 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet 260   8.5 7.2 7.1 6.1 265   8.5 7.3 7.3 6.3 270   8.4 7.1 7.3 6.3 275   8.4 6.8 7.2 6.2 280   8.4 6.9 7.1 6.1 285   8.5 7.2 7.1 6.2 290   8.5 7.4 7.4 6.5 295   8.4 7.2 7.4 6.5 300   8.4 7.1 7.3 6.3 305   8.5 7.2 7.2 6.2 310   8.5 7.3 7.3 6.4 315   8.5 7.3 7.5 6.6 320   8.4 7.2 7.5 6.6 325   8.4 7.2 7.3 6.3 330   8.5 7.4 7.2 6.3 335   8.6 7.5 7.2 6.3 340   8.5 7.4 7.2 6.4 345   8.5 7.2 7.3 6.4 350   8.5 7.3 7.2 6.3 355   8.6 7.4 7.2 6.3 360   8.6 7.5 7.4 6.5 365   8.6 7.4 7.5 6.7 370   8.5 7.3 7.4 6.5 375   8.6 7.4 7.4 6.4 380   8.7 7.6 7.4 6.5 385   8.7 7.7 7.7 6.8  306  Table A2.9 continued. Time (Hours) PK-E2 (vol.% CO2) PK-E3 (vol.% CO2) PK-E4 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet 390   8.6 7.6 7.7 6.8 395   8.6 7.5 7.5 6.6 400   8.6 7.5 7.5 6.6 405   8.8 7.9 7.5 6.6 410   8.7 7.9 7.7 6.8 415   8.5 7.5 7.7 6.8 420   8.5 7.4 7.6 6.7 425   8.6 7.5 7.5 6.6 430   8.7 7.9 7.5 6.7 435   8.6 7.9 7.6 6.9 440   8.5 7.5 7.6 6.8 445   8.5 7.3 7.5 6.6 450   8.5 7.4 7.4 6.6 455   8.7 7.8 7.6 6.8 460   8.7 7.9 7.6 6.9 465   8.5 7.6 7.5 6.8 470   8.5 7.5 7.4 6.5 475   8.6 7.6 7.3 6.5 480   8.7 8.0 7.4 6.6 485   8.7 7.9 7.5 6.7 490   8.5 7.5 7.4 6.7 495   8.5 7.4 7.3 6.5 500   8.2 6.6 7.3 6.5 510   8.7 7.9 7.6 6.8 520   8.6 7.6 7.5 6.6 530   8.8 7.9 7.7 7.0  307  Table A2.9 continued. Time (Hours) PK-E2 (vol.% CO2) PK-E3 (vol.% CO2) PK-E4 (vol.% CO2)  Inlet Outlet Inlet Outlet Inlet Outlet 540   8.6 7.7 7.6 6.7 550   8.6 7.6 7.5 6.7 560   8.7 7.8 7.7 7.0 570   8.6 7.6 7.5 6.7 580   8.8 8.1 7.6 7.0 590   8.6 7.6 7.3 6.6 600   8.8 7.9 7.3 6.5 610   8.7 8.0 7.1 6.4 620   8.7 7.7 7.0 6.3 630   - - 7.1 6.4 640   8.8 8.1 7.1 6.4 650   8.7 7.9 7.3 6.6 660   8.7 7.9 7.3 6.5 670   8.7 8.0 7.2 6.4 680   8.6 7.7   308  A2.3.3.5 Precipitate Characterization A2.3.3.5.1 X-Ray Diffraction Diffractograms of reacted samples from the PS experiments when nesquehonite was detected are provided in Figures A2.20 \u2013 A2.23.  Figure A2.20. Representative XRD pattern and Rietveld refinement of a reacted sample from experiment PS-E1.  2Th Degrees8075706560555045403530252015105Counts16,00015,00014,00013,00012,00011,00010,0009,0008,0007,0006,0005,0004,0003,0002,0001,0000-1,000-2,000-3,000-4,000-5,00019FPX-EW-M05-01.raw_1 Nesquehonite 3.03 %Brucite 1.91 %Magnetite 3.64 %Forsterite 4.06 %Diopside 1.05 %Mellini & Viti Lizardite 1T P31M 86.31 %PS-E1309   Figure A2.21. Representative XRD pattern and Rietveld refinement of a reacted sample from experiment PS-E2. 2Th Degrees8075706560555045403530252015105Counts20,00019,00018,00017,00016,00015,00014,00013,00012,00011,00010,0009,0008,0007,0006,0005,0004,0003,0002,0001,0000-1,000-2,000-3,000-4,000-5,000-6,000-7,000-8,000-9,000M22-1.raw_1 Brucite 1.41 %Magnetite 7.90 %Forsterite 6.23 %Diopside 1.60 %Nesquehonite 1.19 %Mellini & Viti Lizardite 1T P31M 81.67 %PS-E2310   Figure A2.22. Representative XRD pattern and Rietveld refinement of a reacted sample from experiment PS-E6.  2Th Degrees8075706560555045403530252015105Counts11,00010,0009,0008,0007,0006,0005,0004,0003,0002,0001,0000-1,000-2,000-3,000-4,000J11-1-2 SPIKED.raw_1 Nesquehonite 0.70 %Brucite 1.74 %Magnetite 4.32 %Forsterite 3.15 %Corundum 25.79 %Diopside 1.10 %Mellini & Viti Lizardite 1T P31M 63.21 %PS-E6311   Figure A2.23. Representative XRD pattern and Rietveld refinement of a reacted sample from experiment PS-E10. A2.3.3.5.2 Thermogravimetric Analysis Results from the characterization of all PS experiments, showing the detection of hydromagnesite, are included in Figures A2.24 and A2.25.  2Th Degrees8075706560555045403530252015105Counts11,00010,50010,0009,5009,0008,5008,0007,5007,0006,5006,0005,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,500-3,000-3,50019FPX-MBC-LSCINJ-02.raw_1 Nesquehonite 0.26 %Brucite 3.19 %Magnetite 6.48 %Forsterite 5.21 %Diopside 1.59 %Mellini & Viti Lizardite 1T P31M 83.26 %PS-E10312   Figure A2.24. Representative TG and DTG curves of the unreacted CPS-1: FPS mix and reacted samples from experiments PS-E1, E2, E3 and E4 (left). Representative TG and DTG curves of the unreacted CPS-2: FPS mix and reacted samples from experiments PS-E5, E6, and E7 (right). In the initial sample, the peaks refer to the dehydroxylation of brucite (330 \u2013 410\u00b0C) and dehydroxylation of serpentine (600 \u2013 750\u00b0C). For the carbonated sample, the mass loss ranges are due to the loss of adsorbed water (60 \u2013 130\u00b0C), dehydration of hydromagnesite (150 \u2013 230\u00b0C), dehydroxylation of brucite and hydromagnesite and decarbonation of hydromagnesite (340 \u2013 460\u00b0C), and the dehydroxylation of serpentine (600 \u2013 750\u00b0C).313   Figure A2.25. Representative TG and DTG curves of the unreacted CPS-2: FPS mix and reacted samples from experiments PS-E8 and E9 (left). Representative TG and DTG curves of unreacted CPS-2 and reacted sample from experiment PS-E10 (right). In the initial sample, the peaks refer to the dehydroxylation of brucite (330 \u2013 410\u00b0C) and dehydroxylation of serpentine (600 \u2013 750\u00b0C). For the carbonated sample, the mass loss ranges are due to the loss of adsorbed water (60 \u2013 130\u00b0C), dehydration of hydromagnesite (150 \u2013 230\u00b0C), dehydroxylation of brucite and hydromagnesite and decarbonation of hydromagnesite (340 \u2013 460\u00b0C), and the dehydroxylation of serpentine (600 \u2013 750\u00b0C). 314  A2.3.3.5.3 Scanning Electron Microscopy Energy dispersive spectra of the hydromagnesite rosettes and the prismatic nesquehonite needles were taken and are presented in Figures A2.26 and A2.27. Both spectra show minor to no silica and a carbon signal typical of Mg-carbonates. Pt was used to coat aggregate samples.  Figure A2.26. Energy dispersive spectra for a representative hydromagnesite rosette. 0 0.5 1 1.5 2 2.5 3Energy (keV)0200400600800100012001400160018002000Intensity (Counts)OCMgSiPt315   Figure A2.27. Energy dispersive spectra for a representative nesquehonite needle.0 0.5 1 1.5 2 2.5 3Energy (keV)050010001500200025003000350040004500Intensity (Counts)OCMgPt316  A2.3.3.6 Reaction Efficiency The sequestered mass of CO2 was compared against three metrics, the percentage of MgO reacted, the percentage of reactivity from the dominant reactive minerals and against the results from the batch dissolution tests. A2.3.3.6.1 Batch Dissolution Test Comparison The results for the comparison against the batch dissolution results were presented in Chapter 3. The calculation process for these results is shown through Equations A2.23 \u2013 A2.30. The initial FPK TIC (TIC!\"#$#%&) is converted from units of wt.% or g CO2 per g of sample to units of mmol per gram. Equation A2.23. !\"#!\"#$#%& $''(&) % = !\"#!\"#$#%& $ )\t+,!)\t-%'.&\/% \u00b7 011.3345\"\t$%!&&'(6 The leached Ca (Ca'(%)*(+) is corrected to account for Ca leached from calcite during batch dissolution, as this process would release CO2. Where no significant Ca was leached, Mg is assumed to be the responsible cation, and the abundance of Mg is corrected. This correction is done since the initial TIC was measured and assumes a 1:1 ratio of Ca or Mg to CO2 and yields the leached Ca or Mg from noncarbonate sources (Ca,-\")%.\/-\"%$(;\tMg,-\")%.\/-\"%$(). Equation A2.24. #(7(\"8%9:(\"%$\/ $''(&) % = #(;\/%8<\/= $''(&) % \u2212 !\"#!\"#$#%& $''(&) % Equation A2.25. *+7(\"8%9:(\"%$\/ $''(&) % = *+;\/%8<\/= $''(&) % \u2212 !\"#!\"#$#%& $''(&) % The total leached divalent cations during batch dissolution is found by summing the noncarbonate Mg and the noncarbonate Ca. Equation A2.26. !,-(.\t012(.34-\t#(-1,45(!0#) $''(&) % = *+7(\"8%9:(\"%$\/ $''(&) % + #(7(\"8%9:(\"%$\/ $''(&) % 317  The equivalent TIC increase from the cations leached in the batch dissolution test (Batch\tTIC!\").(%0() can be found by taking the total leached divalent cations and converting them to the equivalent wt.% of CO2 that would be captured by this cation abundance. Ca was assumed to form calcite (1:1 ratio of Ca to CO2), whereas Mg was assumed to form hydromagnesite (5:4 ratio of Mg to CO2). Equation A2.27. !\"#$%\t'()!\"#$%&'%(+#.%\t).() = )*&!\"#$%&'\"#%()+**\"+, ,-..(0\u00b723!\"#$%&'\"#%()+**\"+, ,4\u00b7.(678)\u00b7::.;;<+,\t.\/0**\"+,\u00b7.;;(=>.%).;;;(6678)  Since the batch dissolution test was performed on the fine-grained sample alone, while the injection test was done on a coarse-fine mix, the TIC increase from injected and mineralized CO2 (TIC!\").(%0() needed to be attributed to the reactivity of the fines (FPS\tInjection\tTIC!\").(%0(; FPK\tInjection\tTIC!\").(%0(). This was accomplished by first attributing the reactivity to the proportion (25 wt.%) of fines in the injection mixture. For the FPK, another factor was required to account for the percentage of fines from the FPK (FPK\tFines) versus the total, which included fines from the CPK (Total\tFines). Equation A2.28. 9:;\t\"4<3=-1,4\t!\"#!\"89\/%>\/(>-.%\t#A?) = !\"#!\"89\/%>\/(>-.%\t#A?) \u00b7 0()\t-%'.&\/)3.?B()\tCD-)  Equation A2.29. 9:B\t\"4<3=-1,4\t!\"#!\"89\/%>\/(>-.%\t#A?) = !\"#!\"89\/%>\/(>-.%\t#A?) \u00b7 CDE\tC#\"\/>F($%&\tC#\"\/> (%) \u00b7 0()\t-%'.&\/)3.?B()\tCDE) The reactivity from the injection and batch tests were then compared. Equation A2.30. !\"G\/8$#(\"H%$8< (%) = CDE\/CD-\t!\"G\/8$#(\"\tF!+)*+,-.\/-(J$.%\t+,!)H%$8<\tF!+)*+,-.\/-(J$.%\t+,!)  318  A2.3.3.6.2 MgO Reactivity The total MgO content was determined from the whole-rock chemistry. The mass of sequestered CO2 was related to the MgO content by assuming the formation of hydromagnesite, which consists of MgO and CO2 at a 1.25 to 1 ratio. The calculation process used to make this comparison is shown in Equations A2.31 \u2013 A2.34. Table A2.10 shows the results of the calculation process for the comparison against the MgO content. The total MgO content (MgO1-$%&) is found from the fraction of each sample type in the whole sample (2\t4-%.0(2\t5%67&( and 2\t8#\"(2\t5%67&() and from the whole-rock chemistry analysis of each sample type (MgO4-%.0( and MgO8#\"(). Equation A2.31. *+AF($%&(>-.%) = *+A+(%9>\/(>-.%) \u00b7 )\t+(%9>\/)\t-%'.&\/ (%) +*+AC#\"\/(>-.%) \u00b7 )\tC#\"\/)\t-%'.&\/ (%) The MgO total was then converted to molar percent. Equation A2.32. *+AF($%&(C,..%) = *+AF($%& $ )\tL),)\t-%'.&\/% \u00b7 013.M31 $ '(&)\tL),% The TIC increase was converted to an equivalent MgO content based on the formation of hydromagnesite (TIC!\").(%0(\tHmg\tMgO). This was done by converting the TIC increase to moles and then multiplying by the stoichiometric ratio of MgO to CO2, which is 1.25. Equation A2.33. !\"#!\"89\/%>\/\tDC+\t*+A(C,..%) = !\"#!\"89\/%>\/ $ )\t+,!)\t-%'.&\/% \u00b7 011.334 $ '(&)\t+,!% \u00b7 E. FG $'(&\tN')\tL),'(&\tN')\t+,!% The TIC increase MgO was then compared to the total MgO content. Equation A2.34. ;\/%8<\/=\tN')\tL),F($%&\tL), \t(%) = F!+)*+,-.\/-\tN')\tL),('(&.%)L),0'1.(('(&.%)   319  Table A2.10. Determination of the degree of MgO accessed to sequester carbon in each of the centimetre-scale injection experiments. Experiment CPS\/CPK MgO (wt.%) a FPS\/FPK MgO (wt.%) a Total MgO (wt.%) b Total MgO (mol.%) TIC Increase (wt.% CO2) TIC Increase as Hmg MgO (mol.%) c Leached Hmg MgO \/ Total MgO (%) c PS-E1 38.1 41.9 39.1 0.97 1.430 \u00b1 0.217 d 0.041 4.2 \u00b1 0.6 d PS-E2 38.1 41.9 39.1 0.97 1.328 \u00b1 0.241 0.038 3.9 \u00b1 0.7 PS-E3 38.1 41.9 39.1 0.97 1.439 \u00b1 0.230 0.041 4.2 \u00b1 0.7 PS-E4 38.1 41.9 39.1 0.97 1.534 \u00b1 0.139 0.044 4.5 \u00b1 0.4 PS-E5 38.2 41.9 39.1 0.97 1.407 \u00b1 0.200 0.040 4.1 \u00b1 0.6 PS-E6 38.2 41.9 39.1 0.97 1.364 \u00b1 0.125 0.039 4.0 \u00b1 0.4 PS-E7 38.2 41.9 39.1 0.97 1.422 \u00b1 0.078 0.040 4.2 \u00b1 0.2 PS-E8 38.2 41.9 39.1 0.97 1.348 \u00b1 0.124 0.038 3.9 \u00b1 0.4 PS-E9 38.2 41.9 39.1 0.97 1.393 \u00b1 0.044 0.040 4.1 \u00b1 0.1 PS-E10 38.2 - 38.2 0.95 0.210 \u00b1 0.031 0.0060 0.63 \u00b1 0.09 PK-E1 22.6 22.9 22.7 0.56 0.188 \u00b1 0.091 0.0053 0.95 \u00b1 0.46 PK-E2 22.6 22.9 22.7 0.56 0.017 \u00b1 0.020 0.0005 0.09 \u00b1 0.10 PK-E3 24.0 20.3 23.1 0.57 0.207 \u00b1 0.052 0.0059 1.0 \u00b1 0.3 PK-E4 24.0 - 24.0 0.60 0.049 \u00b1 0.059 0.0014 0.23 \u00b1 0.28 a Value from whole-rock chemistry, determined by XRF or ICP-AES.  b Coarse to fine mixtures done at a mass ratio of 3 to 1. c Hydromagnesite (Hmg) has an MgO to CO2 ratio of 1.25 to 1.  d 95% confidence interval. 320  A2.3.3.6.3 Mineral Reactivity The abundance of brucite and lizardite was determined by TGA and qXRD, respectively. Knowing the stoichiometry of these minerals enabled the determination of their Mg abundance, and the calculation process relating the reactivity to this abundance (brucite for PS, lizardite for PK) is shown in Equations A2.35 \u2013 A2.38. Table A2.11 shows the results of the calculation process for the comparison against the brucite and lizardite reactivity. The total Mg content from brucite (Total\tMg!\"#$%&') was found by multiplying the brucite abundance in the whole sample by the percentage of Mg in brucite based on its stoichiometry. The CPS-2 was not considered to contribute apart from in experiment PS-E10. Equation A2.35. !\"#$%\t'(!\"#$%&'(*#.%) = \/0123#4(*#.%) \u00b7 (\t*+,\/.+,\/0(\t,1234' (*#.%) \u00b7 6. 789:; (\t5((\t!\"#$%&'< The total Mg content from fine-grained lizardite (Total\tMg(%)*\"+%&') was found by multiplying the lizardite abundance by the percentage of Mg in lizardite based on its stoichiometry. Equation A2.36. !\"#$%\t'(6%71\"8%&'(*#.%) = =3>$0?3#4(*#.%) \u00b7 (\t*+9:.+9\t*%;'<(\t,1234' (*#.%) \u00b7 6. @9A8; (\t5((\t6%71\"8%&'< The TIC increase was converted to an equivalent Mg content based on the formation of hydromagnesite (TIC,-$\"'*.'\tHmg\tMg). This was done by converting the TIC increase to moles and then multiplying by the stoichiometric ratio of MgO to CO2, which is 1.25. Equation A2.37. !BC=;$\"'1<'\tDE(\t'((E\"%.%) = !BC=;$\"'1<' ; (\t.>!(\t,1234'< \u00b7 ?@@.BBC ; 2D4(\t.>!< \u00b7 8. @F;2D4\tE2(\t5(2D4\tE2(\t.>!< The TIC increase Mg was then compared to the total brucite or lizardite Mg content. Equation A2.38. 6'1$F'8\tE2(\t5(GD&14\t5( \t(%) = G=.\"#$%&'(&\tE2(\t5((2D4.%)GD&14\t5()*#&%'+(2D4.%)321  Table A2.11. Determination of the degree of Mg accessed from brucite (PS) or lizardite (PK) to sequester carbon in the centimetre-scale injection experiments. Experiment Mg Source Portion (wt.%) Abundance (wt.%) a Total Mg (mol.%) TIC Increase   (wt.% CO2) TIC Increase as Hmg Mg (mol.%) b Leached Hmg Mg \/ Total Mg (%) b PS-E1 FPS Brucite 25 12.8 0.055 1.430 \u00b1 0.217 c 0.041 74.0 \u00b1 11.2 c PS-E2 FPS Brucite 25 12.8 0.055 1.328 \u00b1 0.241 0.038 68.7 \u00b1 12.5 PS-E3 FPS Brucite 25 12.8 0.055 1.439 \u00b1 0.230 0.041 74.5 \u00b1 11.9 PS-E4 FPS Brucite 25 12.8 0.055 1.534 \u00b1 0.139 0.044 79.4 \u00b1 7.2 PS-E5 FPS Brucite 25 12.8 0.055 1.407 \u00b1 0.200 0.040 72.8 \u00b1 10.4 PS-E6 FPS Brucite 25 12.8 0.055 1.364 \u00b1 0.125 0.039 70.6 \u00b1 6.5 PS-E7 FPS Brucite 25 12.8 0.055 1.422 \u00b1 0.078 0.040 73.6 \u00b1 4.0 PS-E8 FPS Brucite 25 12.8 0.055 1.348 \u00b1 0.124 0.038 69.8 \u00b1 6.4 PS-E9 FPS Brucite 25 12.8 0.055 1.393 \u00b1 0.044 0.040 72.1 \u00b1 2.3 PS-E10 CPS Brucite 100 1.4 0.024 0.210 \u00b1 0.031 0.0060 24.8 \u00b1 3.7 PK-E1 <425\u00b5m Lizardite 42 d 34.6 0.16 0.188 \u00b1 0.091 0.0053 3.1 \u00b1 1.5 PK-E2 <425\u00b5m Lizardite 42 d 34.6 0.16 0.017 \u00b1 0.020 0.0005 0.3 \u00b1 0.3 PK-E3 <425\u00b5m Lizardite 39 d 27.5 0.11 0.207 \u00b1 0.052 0.0059 4.6 \u00b1 1.2 PK-E4 <425\u00b5m Lizardite 18 d 28.1 0.055 0.049 \u00b1 0.059 0.0014 2.5 \u00b1 3.1 a Brucite abundance determined by TGA and lizardite abundance by qXRD.  b Hydromagnesite (Hmg) has an MgO to CO2 ratio of 1.25 to 1. c 95% confidence interval.  d Abundance of FPK (if present) plus the abundance of CPK fines. 322  A2.3.3.7 Water Mass Balance The water mass loss could be found by subtracting the final moisture content from the initial moisture content. Accounting for the evaporated water was accomplished using Equation A2.21. Table A2.12 presents this calculation, including the injected gas volume and the evaporated mass of water. The evaporated water mass was converted to wt.% H2O by dividing by the dry sample mass. Table A2.12. Determination of the evaporated water from each injection experiment. Experiment Injected Gas Volume (L) Evaporated Water (g) Evaporated Water  (wt.% H2O) PS-E1 306.4 0.5 0.07 PS-E2 138.0 0.2 0.07 PS-E3 529.9 0.9 0.05 PS-E4 407.1 0.7 0.04 PS-E5 308.4 0.5 0.07 PS-E6 132.0 0.2 0.03 PS-E7 539.9 0.9 0.05 PS-E8 79.5 0.1 0.02 PS-E9 79.5 0.1 0.03 PS-E10 943.0 1.6 0.27 PK-E1 618.4 1.1 0.16 PK-E2 11.3 0.02 0.004 PK-E3 41.1 0.07 0.01 PK-E4 40.6 0.07 0.01 323  A2.4 Discussion A2.4.1 FPS Labile Mg The expected lability of the brucite and serpentine Mg from the FPS was calculated to compare the amounts and to justify the assumption that reactivity would be dominantly from brucite. This is shown in Equations A2.39 and A2.40, and the inputs and outputs of the equation are presented in Table A2.13. Equation A2.39. !\"#$%&\t()(+,.%) = 1234$,&(+,.%) \u00b7 !\t#!!\t$%&'()* (+,.%) \u00b7 +,,\t-.\/(0*\t#!+,,\t12).0\t#! (+,.%) Equation A2.40. !\"#$%&\t()(+,.%) = 6&27&8,$8&(+,.%) \u00b7 !\t#!!\t3*%4*5)(5* (+,.%) \u00b7 6\t-.\/(0*\t#!+,,\t12).0\t#! (+,.%) Table A2.13. Labile Mg sources in the FPS. Mineral Abundance (wt.%) Mg Abundance (wt.%) Mg Lability (wt.%) Labile Mg (wt.%) Brucite 12.8 5.3 100 5.3 Serpentine a 80.9 21.3 4 0.85 a Assumed to be lizardite to define the stoichiometry.   A2.4.2 Calcium Source The abundance of calcium leached from the batch dissolution results, beyond that attributable to calcite, was related to its potential sources, either cation exchange from smectites or from the dissolution of Ca-silicates. A2.4.2.1 Smectite Cation Exchange The cation exchange capacity (CEC) required to contribute the noncarbonate Ca (Ca!\"#$%&'\"#%()) was calculated by relating the amount of calcium to the smectite abundance. This 324  calculation is shown in Equation A2.41, and the results for both FPK-1 and FPK-2 are presented in Table A2.14. Equation A2.41. 9:9 ;'7208! < = 9\"925'.%\/25.)* ;'7208! < \u00b7 +37*')()*(;).%) Table A2.14. Determination of the cation exchange capacity necessary to release the leached Ca abundance. Sample Leached Noncarbonate Ca2+ (cmol\u00b7kg-1) qXRD Smectite Abundance (wt.%) Required Cation Exchange Capacity (cmol\u00b7kg-1) FPK-1 290 14.2 2040 FPK-2 530 10.9 4860  A2.4.2.2 Ca-Silicate Dissolution The amount of Ca-silicate minerals, wollastonite and diopside, needed to dissolve to release sufficient calcium was calculated based on the mineral stoichiometry, as shown in Equations A2.42 and A2.43, and in Tables A2.15 and A2.16. Equation A2.42. =$>>?%@&A\tB?%%\">,?8$,&(+,.%) = ?@.!\"#$%&'\"#%()A**\"+, B\u00b7 -(*\"+)-000(**\"+)\u00b76,.,DEA,\t2%*\"+B\u00b7+,,(;).%)F,.G6HA 3(.%\t2%3(.%\t6\"++%7(\"#8()B  Equation A2.43. =$>>?%@&A\t=$?7>$A&(+,.%) = ?@.!\"#$%&'\"#%()A**\"+, B\u00b7 -(*\"+)-000(**\"+)\u00b76,.,DEA,\t2%*\"+B\u00b7+,,(;).%)F,.+EH+A 3(.%\t2%3(.%\t98\":78;)B  Table A2.15. Determination of the wollastonite dissolution needed to release the noncarbonate Ca abundance. Sample Leached Noncarbonate Ca2+ (mmol\u00b7g-1) Equivalent Ca (wt.%) Required Wollastonite Abundance (wt.%) qXRD Wollastonite Abundance (wt.%) FPK-1 0.029 0.116 0.34 b.d. FPK-2 0.053 0.212 0.61 Trace 325  Table A2.16. Determination of the diopside dissolution needed to release the noncarbonate Ca abundance. Sample Leached Noncarbonate Ca2+ (mmol\u00b7g-1) Equivalent Ca (wt.%) Required Diopside Abundance (wt.%) qXRD Diopside Abundance (wt.%) FPK-1 0.029 0.116 0.63 7.4 \u00b1 7.0 FPK-2 0.053 0.212 1.15 3.7 \u00b1 0.9   A2.4.3 Particle Size Distribution To compare the reactivity of the FPS against that of the CPS, the TIC increase was normalized to the sample mass fraction and the brucite abundance. The equation for this is shown in Equation A2.44, and the results are in Table A2.17. Equation A2.44. C?2D\"%$E&A\tFG9\tG842&\">& = 1I@<#$&)%7)(;).%)3.740*J%.')(25(%)\u00b7$%&'()*(;).%) Table A2.17. PS injection experiment TIC increase normalized to sample mass and brucite abundance. Sample Experiment TIC Increase (wt.% CO2) Fraction of Sample Mass Brucite Abundance (wt.%) Normalized TIC Increase FPS PS-E1 to E9 1.41 0.25 12.8 0.44 CPS-2 PS-E10 0.21 1 1.4 0.15  To compare the reactivity of the FPK against that of the CPK, the TIC increase was normalized to the sample mass fraction and the lizardite abundance. Since the CPK had some fines, the reactivity of the mixed FPK and CPK experiment was attributed to the FPK based on its mass 326  proportion of fines. The equation for this is shown in Equation A2.45, and the results are in Table A2.18. Equation A2.45. C?2D\"%$E&A\tFG9\tG842&\">& = 1I@<#$&)%7)(;).%)\u00b7J(5*KL*%'*5).!*(%)3.740*J%.')(25(%)\u00b7-(M.%N()*(;).%)  Table A2.18. PK injection experiment TIC increase normalized to sample mass and lizardite abundance. Sample Experiment TIC Increase (wt.% CO2) Percentage of Fines (wt.%) Fraction of Sample Mass (%) Lizardite Abundance (wt.%) Normalized TIC Increase FPK-2 PK-E3 0.207 64.9 25 27.2 0.020 CPK-2 PK-E4 0.049 100 100 28.1 0.0017   A2.4.4 CO2 Supply To compare the TIC increases of PS-E1 to E9 to the total volume of injected CO2 (Injected\tV*+O), this volume was calculated from the injected flow rate and the experimental duration as shown in Equation A2.46 and Table A2.19. This volume was then normalized to the sample mass before plotting against the TIC increase. Equation A2.46. G8H&4,&A\tI@P=(!) = J%?+\tK\",& ;7-\t@P=7(5 < \u00b7 =32\",$?8(L2) \u00b7 MN ;7(5Q% < \u00b7 +\t-+,,,\t7- 327  Table A2.19. Calculated mass of CO2 injected and experimental CO2 capture efficiency. Experiment Flow Rate (mL\u00b7min-1) Duration (hours) CO2 Injected (g) TIC Increase (wt.% CO2) Sample Mass (kg) CO2 Captured (g) CO2 Capture Efficiency (%) PS-E1 27 191 56.6 1.43 0.674 9.6 17 PS-E2 12 285 37.5 1.33 0.673 9.0 24 PS-E3 20 484 106.2 1.44 1.925 27.7 26 PS-E4 27 266 78.8 1.53 1.811 27.7 35 PS-E5 27 189 56.0 1.41 0.674 9.5 17 PS-E6 12 177 23.3 1.36 0.674 9.2 39 PS-E7 20 428 93.9 1.42 1.926 27.3 29 PS-E8 5.5 241 14.5 1.35 0.533 7.2 49 PS-E9 5.5 241 14.5 1.39 0.509 7.1 49 PS-E10 12 1299 171.1 0.21 0.598 1.3 1 PK-E1 9.5 1085 113.1 0.19 0.674 1.3 1 PK-E2 1 188 1.6 0.02 0.539 0.1 7 PK-E3 1 685 7.5 0.21 0.517 1.1 14 PK-E4 0.9 677 6.7 0.05 0.589 0.3 4   A2.4.5 Carbonate Stability The composition of the precipitated phase in the PS injection experiments was evaluated by using the known abundance of sequestered CO2 from the TIC increase and the mass balance, and the mass of pore water lost to hydrous Mg-carbonate precipitation. The abundances from the TIC increase and mass balance are used as a range on the amount of CO2.  Table A2.20 presents these results and converts the abundance in wt.% to a molar abundance (Equations A2.47 \u2013 A2.49). 328  Equation A2.47. FG9\tG842&\">&\t(D?%.%) = FG9\tG842&\">& ; !\t@P=!\t3.740*< \u00b7 +\t72066.,,R\t!\t@P= Equation A2.48. (\">>\t1\"%\"84&\t(D?%.%) = (\">>\t1\"%\"84& ; !\t@P=!\t3.740*< \u00b7 +\t72066.,,R\t!\t@P= Equation A2.49. B\",&2\t!?>>\t(D?%.%) = B\",&2\t!?>> ; !\tS=P!\t3.740*< \u00b7 +\t720+E.,+H\t!\tS=P From the known CO2 abundance, the abundance of MgO is determined by applying one of three stoichiometries, either a 2:1, 1:1 or 5:4 ratio (Equations A2.50 \u2013 A2.55). This is presented in Table A2.21. Equation A2.50. ()O\t(D?%.%) = P \u00b7 FG9\tG842&\">&(D?%.%) Equation A2.51. ()O\t(D?%.%) = P \u00b7 (\">>\t1\"%\"84&(D?%.%) Equation A2.52. ()O\t(D?%.%) = Q \u00b7 FG9\tG842&\">&(D?%.%) Equation A2.53. ()O\t(D?%.%) = Q \u00b7 (\">>\t1\"%\"84&(D?%.%) Equation A2.54. ()O\t(D?%.%) = Q. PR \u00b7 FG9\tG842&\">&(D?%.%) Equation A2.55. ()O\t(D?%.%) = Q. PR \u00b7 (\">>\t1\"%\"84&(D?%.%) Additionally, the molar abundances have been normalized (Equations A2.56 \u2013 A2.58). Equation A2.56. ()O\t(%) = #!P(720.%)#!P(720.%)T@P=(720.%)TS=P(720.%) Equation A2.57. 9OU\t(%) = @P=(720.%)#!P(720.%)T@P=(720.%)TS=P(720.%) Equation A2.58. SUO\t(%) = S=P(720.%)#!P(720.%)T@P=(720.%)TS=P(720.%) 329  Table A2.20. Abundances of MgO, CO2 and H2O in the reaction product from experiments PS-E1 to E9. Experiment TIC Increase (wt.% CO2) TIC Increase (mol) Mass Balance (wt.% CO2) CO2 (mol) Water Loss (wt.% H2O) H2O (mol) PS-E1 1.430 0.032 1.61 0.036 1.8 0.10 PS-E2 1.328 0.030 1.65 0.038 2.1 0.12 PS-E3 1.439 0.033 1.84 0.042 - - PS-E4 1.534 0.035 1.68 0.038 1.7 0.10 PS-E5 1.407 0.032 1.67 0.038 1.8 0.10 PS-E6 1.364 0.031 1.46 0.033 1.5 0.08 PS-E7 1.422 0.032 1.54 0.035 1.6 0.09 PS-E8 1.348 0.031 1.13 0.026 1.8 0.10 PS-E9 1.393 0.032 1.45 0.033 2.0 0.11 Average 1.41 0.032 1.56 0.035 1.8 0.10 Std 0.06 0.001 0.20 0.005 0.2 0.01 330  Table A2.21. Molar and normalized abundances of MgO, CO2 and H2O in the reaction product from experiments PS-E1 to E9 for each possible precipitate stoichiometry. The difference between the TIC increase and the mass balance provides a range on the composition.  Molar Abundance Normalized Abundance  MgO CO2 H2O MgO CO2 H2O TIC Increase 2:1 MgO:CO2 0.064 \u00b1 0.003 0.032 \u00b1 0.001 0.10 \u00b1 0.01 0.33 \u00b1 0.01 0.16 \u00b1 0.01 0.51 \u00b1 0.05 Mass Balance 2:1 MgO:CO2 0.071 \u00b1 0.009 0.035 \u00b1 0.005 0.10 \u00b1 0.01 0.34 \u00b1 0.04 0.17 \u00b1 0.02 0.48 \u00b1 0.05 TIC Increase 1:1 MgO:CO2 0.032 \u00b1 0.001 0.032 \u00b1 0.001 0.10 \u00b1 0.01 0.20 \u00b1 0.01 0.20 \u00b1 0.01 0.61 \u00b1 0.06 Mass Balance 1:1 MgO:CO2 0.035 \u00b1 0.005 0.035 \u00b1 0.005 0.10 \u00b1 0.01 0.21 \u00b1 0.03 0.21 \u00b1 0.03 0.59 \u00b1 0.06 TIC Increase 5:4 MgO:CO2 0.040 \u00b1 0.002 0.032 \u00b1 0.001 0.10 \u00b1 0.01 0.23 \u00b1 0.01 0.19 \u00b1 0.01 0.58 \u00b1 0.06 Mass Balance 5:4 MgO:CO2 0.044 \u00b1 0.006 0.035 \u00b1 0.005 0.10 \u00b1 0.01 0.25 \u00b1 0.03 0.20 \u00b1 0.03 0.56 \u00b1 0.06 331  A2.5 Implications A2.5.1 Sequestration Magnitude Injection rates measured in this study were compared to known or estimated passive sequestration rates and to emission rates at several diamond and nickel mines. To this end, the data used in these calculations and comparisons are presented in Table A2.22. Table A2.22. Data pertaining to the tailings production, emission and passive and injection sequestration rates at several mines and deposits. Mine\/Deposit Tailings Production (Mt\/yr) Emissions (kt CO2\/yr) Passive CO2 Sequestration (kt CO2\/yr) Injection CO2 Sequestration (kt CO2\/yr) Coarse + Fine Coarse Diavik 2 150 a 0.3 (0.2%) - - Gahcho Ku\u00e9 3.2 124a \/ 45.9b - 6.4 (5.2c\/13.9d%) 1.1 (0.9c\/2.4d%) Mount Keith 11 370 a 40 (10.8%) - - Dumont 15 137 a 21 (15.3%) - - Baptiste 43.8 108 a 17 (15.7%) 618 (572c%) - a Total CO2 emissions. b Power generation emissions from point sources make up approximately one-third of Gahcho Ku\u00e9\u2019s emissions. c Percent of total mine emissions. d Percent of mine power generation emissions.  332  Appendix 3: Appendix to Chapter 4 A3.1 Detailed Experimental Set-Up A3.1.1 Pipe-1 CPK (<6000 \u00b5m) and FPK (<425 \u00b5m) were dried under ambient conditions until the moisture contents, when mixed, would achieve the targeted moisture content of ~7 wt.%. CPK and FPK were manually blended at a 78:22 ratio and inserted into a 6-metre long HDPE pipe with a diameter of 18.9 cm. The mixed PK was compacted using two Wooster Sherlock 12-foot long extension poles, with a wooden plate attached to the end of each pole. Compaction was done with the pipe in place after every 5 kg of PK was inserted, with a careful examination that compaction was evenly achieved top to bottom. A total of 264 kg of moist, mixed PK was introduced into the pipe. Holes for the gas inlet, gas outlet, and sensor installation were drilled. Swagelock connectors, Vaisala GMP 251, GMP 221 and HMP 110 and Bosch Sensortec BMP280 pressure sensors were installed into the holes in the pipes, with the sensors being located inside the pipe. One GMP 251 and 221 were installed at the inlet, middle and outlet, while one HMP 110 and two BMP280 pressure sensors were installed at the inlet and outlet. Sensors were sealed with a combination of epoxy and silicone. The pipe face was capped with high-density EvaCell20 foam and was sealed with layers of epoxy and silicone. Data from the GMP 251 and HMP 110 sensors were logged at 15-minute intervals to a Vaisala measurement indicator MI70, while the GMP 221 sensors were logged every minute using a LabJack U3-HV. BMP 280 pressure readings were recorded using a Raspberry Pi 3B+ at 2-second increments during the permeability tests. 333  From the compressed gas cylinder, simulated flue gas (90 vol.% N2 and 10 vol.% CO2) flow was controlled by a Bronkhorst EL-Flow Prestige mass flow controller at 0.72 LPM. The gas was directed through water within a Rubbermaid jug to humidify the gas. SFG was injected to measure permeability and reactivity. After experimental completion, samples of ~500 g of tailings were taken every 50 cm, starting at 60 cm from the inlet. Samples were taken from the top, middle and bottom horizons within the pipe.  A3.1.2 Pipe-2 CPK (<6000 \u00b5m) and FPK (<425 \u00b5m) were dried under ambient conditions until the moisture contents, when mixed, would achieve the targeted moisture content of ~7 wt.%. CPK and FPK were mechanically blended at a 75:25 ratio with a QEP Thinset mortar power mixer and inserted into a 6-metre long HDPE pipe with a diameter of 16.6 cm. The mixed PK was compacted using two Wooster Sherlock 12-foot long extension poles, with a wooden plate attached to the end of each pole. Compaction was done with the pipe in place after every 5 kg of PK was inserted, with a careful examination that compaction was evenly achieved top to bottom. The pipe was rotated 180 degrees during compaction to ensure compaction occurred evenly. A total of 236 kg of moist, mixed PK was introduced into the pipe. Holes for the gas inlet, gas outlet, and sensor installation were drilled. Swagelock connectors, Vaisala GMP 251 and HMP 110 and Bosch Sensortec BMP 388 pressure sensors were installed into the holes in the pipes, with the sensor detector being located inside the pipe. Two GMP 251 probes were installed at the inlet, middle and outlet, while one HMP 110 and two BMP 388 pressure sensors were installed at the inlet and outlet. Sensors were sealed with a combination 334  of epoxy and silicone. The pipe face was capped with high-density EvaCell20 foam and was sealed with layers of epoxy and silicone. Data from the GMP 251 and HMP 110 sensors were logged at 15-minute intervals to a Vaisala measurement indicator MI70. BMP 388 pressure readings were recorded using a Raspberry Pi 3B+ at 2-second increments during the permeability tests. From the compressed gas cylinder, air flow was controlled by a Bronkhorst EL-Flow Prestige mass flow controller for the permeability tests. Simulated flue gas (90 vol.% N2 and 10 vol.% CO2) flow was controlled by a Bronkhorst EL-Flow Prestige mass flow controller at 0.55 and 0.1 LPM. The gas was directed through water within an Erlenmeyer flask to humidify the gas. After experimental completion, samples of ~500 g of tailings were taken every 50 cm, starting at the inlet. Samples were also taken at the outlet. Samples were taken from the top, middle and bottom horizons within the pipe.  A3.1.3 Pad-1 A 40 cm thick base layer of 820 kg of moist CPK was compacted into the bottom of an intermediate bulk container (IBC) with dimensions of 112 cm by 96cm by 100 cm. This was done to elevate the rest of the pad and facilitate construction and was not involved in the reaction. The next layer consisted of 110 kg of saturated FPK which formed a of 6 cm thick lower aquitard. CPK (<6000 \u00b5m) and FPK (<425 \u00b5m) were dried under ambient conditions until the moisture contents, when mixed, would achieve the targeted moisture content of ~7 wt.%. CPK and FPK were manually blended at a ratio of 76:24 and were lightly compacted to produce a permeable layer, consisting of 180 kg of moist tailings, which formed a 12 cm thick layer. Compaction was done in stages to produce even compaction throughout the layer. Injection of CO2 took place into this layer. 335  Effluent ports were drilled into the side of the IBC, every 30 degrees to be distributed radially around the injection port, in the middle of the mixed PK layer. Lastly, an additional 6 cm layer of 110 kg of saturated FPK was spread out on top of the mixed PK layer to act as an upper aquitard. The top of the IBC was sealed up, which left a headspace of 36 cm in height. 4 Vaisala GMP 251 and 2 GMP 221 CO2 sensors were embedded into the middle of the mixed PK layer. One GMP 221 sensor was located in the middle of the headspace above the pad. The locations of these sensors are shown in Figure A3.1. Effluent was flowed from the effluent ports through to a GMP 251 CO2 sensor where the effluent concentrations were recorded. This sensor was moved around to each open effluent port. Injection, effluent, and sensor ports were sealed using epoxy. The top of the IBC was sealed with epoxy. Initially this was done imperfectly which led to gas flow out through the headspace. Additional application of epoxy led to the headspace being sealed off. Gas leaks were monitored for using a K30 FS Fast Response 10000 ppm CO2 sensor. Data from the GMP 251 and HMP 110 sensors were logged at 15-minute intervals to a Vaisala measurement indicator MI70, while the GMP 221 sensors were logged every minute using a LabJack U3-HV. 336   Figure A3.1. Diagram showing the locations of embedded CO2 sensors placed in Pad-1 in the respective layers. From the compressed gas cylinder, simulated flue gas (90 vol.% N2 and 10 vol.% CO2) flow was controlled by a Bronkhorst EL-Flow Prestige mass flow controller at 0.08 and 0.24 LPM. The gas was directed through water within a Rubbermaid jug to humidify the gas. SFG was directed by the pressure gradient to five outlet ports, each 30 degrees apart on the outside of the IBC walls in the middle of the mixed PK layer. One port was opened at a time in the following order: port 3, port 5, port 2, port 1, and port 4. After experimental completion, samples of ~500 g of tailings were taken in a three-by-four grid (grid spacing of 30 cm width and 35 cm length) of the upper FPK, the mixed layer, and the lower FPK layer. Eight core samples were also taken of the upper FPK layer and the mixed layer in a distributed pattern that included some of the desiccation cracks in the upper FPK layer. Sample locations are shown in Figure A3.2.337   Figure A3.2. Diagram showing the locations of samples taken from Pad-1 from the respective layers. 338  A3.1.4 Pad-2 A 12 cm thick base layer of 250 kg of moist CPK was compacted into the bottom of an IBC with dimensions of 115 cm by 95 cm by 95 cm. This was done to attempt to distribute gas laterally through this permeable layer. To form the next layer, CPK (<6000 \u00b5m) and FPK (<425 \u00b5m) were dried under ambient conditions until the moisture contents, when mixed, would achieve the targeted moisture content of ~7 wt.%. CPK and FPK were mechanically blended with a QEP Thinset mortar power mixer at a ratio of 67:33 and were lightly compacted to produce a permeable layer, consisting of 550 kg of moist tailings, which formed a 30 cm thick layer. Compaction was done in two stages to produce even compaction throughout the layer. The next layer consisted of 150 kg of saturated FPK which formed an 8.3 cm thick upper aquitard. Lastly, an additional 5 cm layer of CPK was spread out on top of the FPK layer to prevent interaction with the atmosphere. A polyethylene liner was placed on top of the CPK cap layer and was sealed along the walls of the IBC with Lepage All Plastics Super Glue and silicone. From the compressed gas cylinder, simulated flue gas (90 vol.% N2 and 10 vol.% CO2) flow was controlled by a Bronkhorst EL-Flow Prestige mass flow controller at 1, 0.5 and 0.25 LPM. The gas was directed through water within an Erlenmeyer flask to humidify the gas. SFG was directed into the tailings pad by a perforated injection pipe. The 5.2 cm diameter, 115 cm long, perforated pipe was placed in the bottom corner of the IBC within the base layer of CPK-2. The pipe was perforated with 3\/8-inch diameter holes, spaced one inch apart. SFG was directed by the pressure gradient to three outlet ports in the top of the mixed PK layer in the opposite corner from the injection pipe. Three Vaisala GMP 251 CO2 sensors were placed along the length of the injection pipe. One of the sensors was placed in the middle. The cross-sectional plane through the middle of the 339  injection pipe had an additional four GMP 251 sensors embedded in the upper portion of the base CPK layer. Midway through the mixed PK layer two GMP 251 and three GMP 221 sensors were placed along this middle profile. At the top of the mixed PK layer two more GMP 251 and three GMP 221 sensors were embedded. The last sensor was installed directly against one of the effluent ports. The other two effluent ports also had a GMP 251 sensor placed next to them to record the effluent concentrations. Lastly, two GMP 251 sensors were placed above the cap CPK layer in the headspace. All sensor locations are shown in Figure A3.3. The injection pipe, effluent ports and the sensor ports were all sealed with silicone. Gas leaks were monitored for using a K30 FS Fast Response 10000 ppm CO2 sensor. Data from the GMP 251 and HMP 110 sensors were logged at 15-minute intervals to a Vaisala measurement indicator MI70, while the GMP 221 sensors were logged every minute using a LabJack U3-HV. 340   Figure A3.3. Diagram showing the locations of embedded CO2 sensors placed in Pad-2 in the respective layers.  341  After experimental completion, ~500 g tailings samples were taken in a three-by-four grid (grid spacing of 32 cm width and 35 cm length). CPK samples were only taken along the middle vertical plane, orthogonal to the injection pipe. Mixed PK samples were taken at three layers, upper, middle and lower, with an additional pair of sample points between the grid points within the middle vertical plane. Thus, the same vertical plane that had the CO2 sensors was sampled extensively. Sample locations are shown in Figure A3.4. 342   Figure A3.4. Diagram showing the locations of samples taken from Pad-2 from the respective layers. 343  A3.2 Detailed Methods A3.2.1 Sample Characterization A3.2.1.1 Sample Preparation Sample preparation was required for XRD, TGA and TIC analysis. Coarse-grained samples (>600 \u00b5m) were pulverized with a ring mill for 30 seconds. Pulverized and fine-grained samples (<600 \u00b5m) were micronized for seven minutes under anhydrous ethanol using a McCrone Micronizing Mill. Samples were then dried, disaggregated, and homogenized using an agate mortar and pestle. A3.2.1.2 X-Ray Diffraction Mineralogy of PK samples was determined through qualitative and quantitative X-ray Diffraction (XRD) analysis, which was performed using a Bruker D8 Focus located within the Department of Earth, Ocean, and Atmospheric Sciences at the University of British Columbia. Prior to analysis, samples of PK were treated using a Ca2+ cation exchange process for 12 hours using a 1 mol Ca2+ solution as per the method of Wilson et al. (in prep). All samples were prepared as back-loading cavity mounts and were loaded against 400 grit sandpaper to reduce preferred orientation. Patterns were collected on the D8 Focus using a Co X-ray tube operated at 35 kV and 40 mA using a LynxEye 1D position-sensitive detector over a 2\u03b8 range of 3-80\u00b0 with a step size of 0.03\u00b0 and a dwell time of 0.7 s\/step. Qualitative analysis was performed by DIFFRAC.EVA V.4.2 (Bruker AXS) using the ICDD PDF-4+ database. Quantitative analysis was performed using DIFFRAC.TOPAS V.4.2 (Bruker AXS) using the fundamental parameters approach (Cheary & Coelho, 1992). The crystal structure data sources are displayed in Table A3.1.344  Table A3.1. Crystal structure data sources for the quantitative XRD analysis. Mineral Source Brucite (National Bureau of Standards, 1956) Serpentine (Viti & Mellini, 1997) Forsterite (National Bureau of Standards, 1984) Smectite (Rosenquist, 1959) Magnetite (National Bureau of Standards, 1967) Diopside (Federico et al., 1988) Phlogopite (Smith, 1956) Clinochlore (Gillery, n.d.) Calcite (Falini et al., 1998) Quartz (Kern & Eysel, 1993) Albite (Goodyear, 1954) K-Feldspar (Bailey, 1955; University of Oxford, 1967)  Talc (Brindley, 1977) Dolomite (Ross & Reeder, 1992) Cuspidine (Saburi et al., 1977) Andradite (Tsao, 1964) Wollastonite (Smith, 1974) Grossular (Pabst, 1937) Tremolite (Comodi et al., 1991)  A3.2.1.3 Thermogravimetric Analysis A Perkin Elmer TGA 4000 with a Polyscience chiller and an AS 6000 autosampler performed the thermo-gravimetric analysis at the University of British Columbia. N2 was used as an inert carrier gas and was pumped at a rate of 19.8 mL min-1. Samples (~50 mg) were heated from 20 or 100\u00b0C to 900\u00b0C at a rate of 10\u00b0C per minute. Recordings of the temperature and sample 345  mass were taken every second. TG curves of mass versus temperature were plotted as wt.%. DTG curves were plotted as every 50th data point to smoothen out the curves. Brucite is qualitatively identified due to a signature mass loss in the temperature interval from 300 to 450\u00b0C. Brucite decomposition releases one mole of water per mole of brucite according to Equation A3.1. Equation A3.1. !\"($%)! \t\u2192 \t!\"$\t +\t%!$(#) Brucite abundance was quantified by using an exponential function to model the background mass loss from other minerals. This improves the determination of the mass loss attributable to brucite. The derivative mass loss data before and after the brucite interval was used to fit the exponential function of the form shown in Equation A3.2. Equation A3.2. * +%&.%\u00b0* , = .+,(\u00b0*) + \/-,(\u00b0*) Where \u2018y\u2019 is the derivative mass loss, \u2018x\u2019 is the temperature, and \u2018a\u2019, \u2018b\u2019, \u2018c\u2019, and \u2018d\u2019, are constants determined by the \u2018fit\u2019 function in MATLAB R2020a (MathWorks). The fit for the DTG curve was then projected back onto the TG curve. This fitted function matches the slope of the provided data and enables the quantification of mass loss attributable to brucite. This is calculated by Equation A3.3. Equation A3.3. 012\/345.,\/(64.%) = 9. :; \u00b7 +=.,\/(64.%) \u2212=012(64.%), Where \u2018BruciteExp\u2019 is the exponential method\u2019s determined brucite abundance, \u2018mExp\u2019 the exponentially projected mass in wt.% if there had been no brucite in the sample, \u2018m450\u2019 the sample 346  mass in wt.% at the end of the characteristic brucite mass loss interval (~450\u00b0C), and \u20183.24\u2019 the stoichiometric factor between water and brucite. A3.2.1.4 Total Inorganic Carbon To measure the total inorganic carbon (TIC) content of the raw and carbonated samples, a CM5130 acidification module with a Model 5014 CO2 Coulometer from UIC Inc. at the University of British Columbia was used. Micronized and homogenized samples were acidified, releasing gaseous CO2, which was measured using a photodetector to determine the colour change of a colorimetric pH indicator. The instrument was tested with calcium carbonate standards to ensure calibration prior to samples being analyzed. All data are expressed as wt.% CO2. This method has a detection limit better than 0.37 wt.% CO2 and measurements were repeatable within 0.03 wt.% CO2. A3.2.1.5 Total Carbon Total carbon analysis was done at ALS Geochemistry in North Vancouver. Samples were heated in a LECO furnace to ~1350\u00baC, with oxygen flowing through the sample. Carbon dioxide is released from the sample and was measured by infrared spectroscopy. This method has a detection limit of 0.01 wt.% C. A3.2.1.6 Major Oxide Composition Inductively coupled plasma-atomic emission spectroscopy (ICP-AES) was used for PK samples to determine the major oxide composition, with a detection limit of 0.01 wt.%. A3.2.1.7 Particle Size Distribution The particle size distribution of the FPK was determined using a Malvern Mastersizer 2000 Laser Diffraction Particle Size Analyzer. A suspension of 5% solids in distilled water was prepared and mixed before an aliquot was withdrawn for analysis. Ultrasound treatment of 60 seconds was 347  applied to disaggregate particle clusters. The particle size distribution for the CPK fraction above 425 \u00b5m was found by mechanical sieving. A3.2.1.8 Surface Area Multipoint BET with N2 adsorption was performed with a Quantachrome Autosorb-1 surface area analyzer on the original sample materials to determine their respective grain surface areas. A3.2.1.9 Quantitative Evaluation of Materials by SEM Samples of CPK-1 and CPK-2 were analyzed using Quantitative Evaluation of Materials by Scanning Electron Microscopy (QEMSCAN). The samples were prepared as 3 g samples embedded in epoxy on a 30 mm mould. The samples were cut through and polished with a carbon coating. Maps were produced from back-scatter electron signals and electron dispersive X-ray spectroscopy on individual points taken with a stepping interval of 10 microns. The accelerating voltage was 20 kV, and the beam current was 10 nA. The QEMSCAN image analysis software iDiscover was used to combine the individual points into mineralogical maps based on elemental concentrations.  A3.2.2 Standard Proctor Compaction ASTM standard D698 was followed to determine the relationship between the moisture content and the PK\u2019s dry density (ASTM International, 2012). Method C was followed, which used the 6-in mould, with no oversize fraction. The mould was filled in three layers, with each layer receiving 56 blows from a rammer exerting 600 kN-m\/m3, in a specified pattern. Specific tolerances on the height of the material in the mould governed an acceptable compaction test. Five tests were conducted for the 25 wt.% FPK mixture, while seven tests were done for the 33 wt.% 348  FPK mixture. Each test was done at a different moisture content to establish the relationship between the moisture content and dry density. Dry densities were determined from the final sample mass in the mould after compaction, with the volume being known. Moisture contents were plotted against the dry density. A curvilinear relationship was fitted through these data points, with the apex determining the maximum dry density and the optimum moisture content. The 100% saturation line was calculated from Equation A3.4 and plotted. The 60% saturation line can be determined by applying a factor of 0.6 to the saturation line. On the wet-side of optimum, the trend line through the experimental data points matches the slope of the saturation line. Equation A3.4. 634&(%) = 5!6\"#$%7\u22199&:5'6\"#$%75'6\"#$%7\u22199& \u2219 @AA Where w!\"# is the water content at saturation, \u03b3$ is the unit weight of water, \u03b3% is the unit weight of the tailings, and G& is the specific gravity of the tailings.  A3.2.3 Pneumatic Permeability Pressure recordings were made for one flow rate into Pipe-1 and three flow rates for Pipe-2. During periods where the pressure had stabilized, the average inlet and outlet pressures were calculated. The coefficient of permeability was then calculated using Equation A3.5. Equation A3.5. B#(=!) = !;6(&7<(=4\u00b7?)@(A)=)(=4)BCA)D6=*)C=4)D:=))C=4)D7 Where \u2018Kg\u2019 is the intrinsic gas permeability, \u2018Q\u2019 is the volumetric flowrate, \u2018\u00b5\u2019 is the fluid viscosity, \u2018L\u2019 is the sample length, \u2018A\u2019 is the cross-sectional area, \u2018P1\u2019 is the inlet pressure, and 349  \u2018P2\u2019 is the outlet pressure. The mean permeability for the three tests was taken to be the final result for the gas permeability.  A3.2.4 CO2 Mass Balance Recorded CO2 concentrations were used to perform a mass balance on the amount of sequestered CO2 in the porosity, in the aqueous phase and as mineralized CO2. This was accomplished by iteratively determining the difference between the injected volume of CO2 (Equation A3.6) and the effluent volume of CO2 (Equation A3.7) for a specified time step (Equation A3.8), which yielded the carbonation rate (Equation A3.9). The sequestered mass was then found by summing the product of the carbonation rate with the time step duration (Equation A3.10). Equation A3.6. CEFGHI&H-(=D) = E+A@AJF, \u00b7 \u22064(=3G) \u00b7 [I$!]EFKH&(%) Equation A3.7. C.LLKMHF&(=D) = E+A@AJF, \u00b7 \u22064(=3G) \u00b7 K N:[*P)]+,-.\/(%)N:[*P)]01\/-.\/(%) \u2212 L@ \u2212 [I$!]EFKH&(%)MN Equation A3.8. I$!*4\/&MRH-\t(%) = @ \u2212 S233-1.,\/(A@)S+,4.5\/.'(A@) Equation A3.9. I$!*4\/T4&H \t+A#AJF, = I$!*4\/&MRH-(%) \u00b7 E+A@AJF, \u00b7 ==*P)(A#AUK) \u00b7 =(4&A)T6$(\u00b77\/$8\u00b7$9- 7\u00b7V(W) Equation A3.10. I$!X4??Y4K4FIH\t(=\") = \u2211I$!*4\/T4&H +A#AJF, \u00b7 \u22064(=3G) Where \u2018V'()*+#*%\u2019 is the injected volume of CO2, \u2018Q\u2019 is the injected flow rate, \u2018\u2206t\u2019 is the differential time between CO2 sensor measurements, \u2018[CO,]'(-*#\u2019 is the inlet CO2 concentration, \u2018V.\/\/-0*(#\u2019 is the effluent volume of CO2, \u2018[CO,]10#-*#\u2019 is the outlet CO2 concentration, 350  \u2018CO,2\"34\"#*\u2019 is the carbonation rate, \u2018mm21Z\u2019 is the molar mass of CO2, \u2018P\u2019 is the average ambient pressure, \u2018R\u2019 is the ideal gas constant, and \u2018T\u2019 is the average ambient temperature. A3.2.4.1 Correcting Mass Balance to Mineralized Carbon The gas mass balance accounts for the CO2 in the porosity and aqueous phase. Porosity held CO2 can be discounted by determining the sample porosity using the sample density and the known volume of water, and then calculating the volume of CO2 in the pore space from the known concentration and the known total volume and by converting the volume to mass (Equation A3.11). Equation A3.11. I$!=URU?J&[\t(=\") = PCV \u2212 Q%L(=D) \u2212 A&(#)\\&C#\u00b7IA:%DR \u00b7 [I$!](%) \u00b7 AA;0)($<$9-)\u00b7=(4&A)T6$(\u00b77\/$8\u00b7$9- 7\u00b7V(W)  Where \u2018CO,5676&8#9\u2019 is the mass of CO2 held in the pore space, \u20181:\u2019 is the volume of the sample container, \u2018v$\/\u2019 is the final mass of water, \u2018m&\u2019 is the mass of initial dry solids, \u2018\u03c1&\u2019 is the sample density, \u2018[CO,]\u2019 is the concentration of CO2 in the injected gas, \u2018mm21Z\u2019 is the molar mass of CO2, \u2018P\u2019 is the ambient pressure, \u2018R\u2019 is the ideal gas constant, and \u2018T\u2019 is the gas temperature. The CO2 held as alkalinity is not considered as a stable form of carbon sequestration in mine tailings as the DIC can be released to the atmosphere due to changes in temperature or pH. Due to the limited reactivity of the tailings, it has been conservatively assumed that the pore water equilibrated with the CO2 at a pH of 8.8. At 25\u00baC, this means the water has a DIC of 0.01 M. Knowing the final moisture content of the tailings enables the mass of aqueous CO2 to be calculated (Equation A3.12). Equation A3.12. I$!(4])\t(=\") = STI(!) \u00b7 C%L(D) \u00b7 ==*P)(A#AUK) Where \u2018CO,(\"<)\u2019 is the CO2 mass in the aqueous phase and \u2018DIC\u2019 is the dissolved inorganic carbon concentration in the aqueous phase. 351  Then the mass balance (CO,>\"&&?\"-\"(+*) can be corrected to account for only the mineralized carbon (CO,>8(*7\"-8@*%) as shown in Equation A3.13. Equation A3.13. I$!XJFHR4KJ^H-\t(=\") = I$!X4??Y4K4FIH(=\") \u2212 I$!=URU?J&[(=\") \u2212 I$!(4])(=\") A3.2.4.2 Mass Balance Error Calculation The error envelope around the gas mass balance was calculated by using the errors stipulated for the Vaisala GMP 251 CO2 sensors (\u00b1 0.2 vol.% if < 8 vol.% and \u00b1 0.4 vol.% if > 8 vol.%), and the flow rate from the Bronkhorst EL-Flow Prestige mass flow controller (\u00b1 0.5% of reading plus \u00b1 0.2% of full scale). This was done using the same approach to calculate the sequestered mass of CO2, as shown in Equations A3.6 \u2013 A3.10. The difference was that the inlet and outlet concentrations and the flow rate was adjusted based on the manufacturer\u2019s uncertainty. To calculate the maximum amount of sequestered carbon, the positive uncertainty was added to the injected flow rate and each inlet concentration, while the negative error was added to the outlet concentrations (Equations A3.14 \u2013 A3.16). These inputs were then used to calculate the mass of sequestered CO2 over time. Equation A3.14. [I$!]EFKH&:X4,(QUV.%) = [I$!]EFKH&(QUV.%) + W!X:Y@.RRUR(QUV.%) Equation A3.15. [I$!]PM&KH&:X4,(QUV.%) = [I$!]PM&KH&(QUV.%) \u2212 W!X:Y@.RRUR(QUV.%) Equation A3.16. ZVU6\t[.45X4,(DX!) = ZVU6\t[.45(DX!) + \\D\tZVU6\tX15]43\"5.RRUR(DX!) To calculate the minimum amount of sequestered carbon, the negative uncertainty was added to the injected flow rate and each inlet concentration, while the positive error was added to the outlet concentrations (Equations A3.17 \u2013 A3.19). These inputs were then used to calculate the mass of sequestered CO2 over time. 352  Equation A3.17. [I$!]EFKH&:X4,(QUV.%) = [I$!]EFKH&(QUV.%) \u2212 W!X:Y@.RRUR(QUV.%) Equation A3.18. [I$!]PM&KH&:X4,(QUV.%) = [I$!]PM&KH&(QUV.%) + W!X:Y@.RRUR(QUV.%) Equation A3.19. ZVU6\t[.45X4,(DX!) = ZVU6\t[.45(DX!) \u2212 \\D\tZVU6\tX15]43\"5.RRUR(DX!)  A3.2.5 Total Inorganic Carbon Increase To assess the increase in TIC due to carbon mineralization, the initial material was representatively sampled to establish an average value and standard deviation. After CO2 injection, the TIC measurements determined the mean and standard deviation for the carbonated material. Initial, unreacted TIC was subtracted from the final, reacted TIC to determine the TIC increase, as shown in Equation A3.20. Equation A3.20. TICIncrease (wt.%) = TICReacted (wt.%) \u2013 TICUnreacted (wt.%) To ensure that the TIC increase was statistically significant, a one-tailed t-test was performed on the difference in the means to determine the degree of confidence to which the means are statistically different. For the difference between the means, that is, the TIC increase, a 95% confidence interval (CI) was calculated based on the Standard Error (SE) for the TIC increase, and the T-value based on the degrees of freedom and the selected 95% degree of confidence, as shown in Equation A3.21. Equation A3.21. CI = SE\u00b7T-value A3.2.5.1 Sieved Sample TIC Increase To improve the TIC increase analysis, bulk samples had a fraction sieved below 425 \u00b5m. TIC was conducted on this sample fraction for Pipe-1 and Pipe-2. While this found the amount of 353  carbon captured in the fines, this mass needed to be attributed to the entire sample mass used. This was accomplished through Equation A3.22. Equation A3.22. ^TI3JH_H- + #\t*P)#\t34A\/KH, = ^TI3JH_H- + #\t*P)#\taR4I&JUF\tb0!1\t<A, \u00b7 aJFH\t34A\/KH\tX4??\t(#\taR4I&JUF\tb0!1\t<A)VU&4K\t34A\/KH\tX4??\t(#\t34A\/KH)   A3.2.6 Total Carbon TC was assessed to determine if any carbon was being sequestered by smectite minerals. To determine the contribution from smectite intercalation, the TIC was subtracted from the TC, as shown in Equation A3.23. Equation A3.23. I$!3AHI&J&H(64.%) = ^I\t(64.%) \u2212 ^TI\t(64.%)  A3.2.7 Evaporative Moisture Loss The amount of pore water evaporation during the pipe injection tests was determined from the difference in humidity between the inlet and outlet and the recorded temperatures. Based on the saturation vapour pressure of water in air, the evaporative losses could be calculated, as shown in Equation A3.24. Equation A3.24. 6H = \u2206Td\u00b7\/!&7\/\u00b7AA!T\u00b7V\u00b7N222 \u00b7 QUV#4?=,4 Where \u2018\u2206RH\u2019 is the difference in relative humidity across the sample (%), \u2018p$efg\u2019 is the saturation vapour pressure (Pa), \u2018mm$\u2019 is the molar mass of water (g\u00b7mol-1), and \u2018volA\"&hij\u2019 is the total volume of injected gas over the course of the experiment (L). For the pad injection experiments, the evaporative moisture losses were presumed to be insignificant after the findings for the pipe injection experiments were negligible.354  A3.3 Detailed Results A3.3.1 Sample Characterization A3.3.1.1 X-Ray Diffraction Representative diffractograms of each unreacted sample are presented in Figures A3.5 \u2013 A3.8.  Figure A3.5. Representative XRD pattern and Rietveld refinement of unreacted FPK-1.  2Th Degrees8075706560555045403530252015105Counts8,0007,5007,0006,5006,0005,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,500-3,00019GKFPKB2-1-CAT.raw_1 Phlogopite 1M 12.49 %Clinochlore IIb-4 68942 8.81 %Calcite 0.81 %Quartz low 2.09 %Magnetite 1.13 %Microcline intermediate 8.09 %Albite low 5.37 %Talc 1A 7.58 %Dolomite 0.19 %Diopside 2.48 %Forsterite 1.04 %Cuspidine 0.72 %Mellini & Viti Lizardite 1T P31M 31.06 %Ca-Montmorillonite 18.12 %FPK-1355   Figure A3.6. Representative XRD pattern and Rietveld refinement of unreacted CPK-1.  2Th Degrees8075706560555045403530252015105Counts7,0006,5006,0005,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,50019 GK CPK B3-1 CAT.raw_1 Phlogopite 1M 12.05 %Clinochlore IIb-4 68942 11.03 %Calcite 0.95 %Quartz low 3.22 %Magnetite 1.31 %Microcline intermediate 10.94 %Albite low 6.55 %Talc 1A 5.57 %Diopside 2.96 %Forsterite 2.71 %Mellini & Viti Lizardite 1T P31M 27.65 %Ca-Montmorillonite 15.08 %CPK-1356   Figure A3.7. Representative XRD pattern and Rietveld refinement of unreacted FPK-2.  2Th Degrees8075706560555045403530252015105Counts8,5008,0007,5007,0006,5006,0005,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,500-3,000IBC 2-1 FPK CAT 20GKBS.raw_1 Phlogopite 1M 10.71 %Clinochlore IIb-4 68942 7.50 %Calcite 1.29 %Quartz low 5.33 %Magnetite 1.25 %Albite low 9.67 %Microcline intermediate 12.84 %Talc 1A 2.91 %Dolomite 0.76 %Diopside 4.49 %Forsterite 6.04 %Mellini & Viti Lizardite 1T P31M 27.54 %Ca-Montmorillonite 9.69 %FPK-2357   Figure A3.8. Representative XRD pattern and Rietveld refinement of unreacted CPK-2. 2Th Degrees8075706560555045403530252015105Counts5,5005,0004,5004,0003,5003,0002,5002,0001,5001,0005000-500-1,000-1,500-2,000-2,50020gkbs-cpk-lscinj-pre-1-cat.raw_1 Phlogopite 1M 17.07 %Clinochlore IIb-4 68942 6.11 %Calcite 0.56 %Quartz low 3.00 %Magnetite 1.48 %Albite low 7.24 %Orthoclase 6.29 %Talc 1A 3.93 %Dolomite 0.33 %Andradite 1.71 %Cuspidine 1.07 %Diopside 5.58 %Forsterite 9.74 %Mellini & Viti Lizardite 1T P31M 25.92 %Ca-Montmorillonite 9.97 %CPK-2358  A3.3.1.2 Particle Size Distribution The particle size distribution for all unreacted PK samples is plotted in Figures A3.9 and A3.10.  Figure A3.9. Particle size distributions for FPK-1, FPK-2, and CPK-1 and CPK-2 sieved <425 \u00b5m, as determined by a Malvern Mastersizer 2000 Laser Diffraction Particle Size Analyzer.  Figure A3.10. Particle size distributions from mechanical sieving for CPK-1 and CPK-2 >53 \u00b5m. 359  A3.3.1.3 Quantitative Evaluation of Materials by SEM The results of the QEMSCAN characterization for samples CPK-1 and CPK-2 are presented in Table A3.2. Table A3.2. QEMSCAN mineralogy abundances. Mineral Abundance (wt.%) CPK-1 CPK-2 Quartz 3.94 5.15 Alkali Feldspar 2.53 1.44 K Feldspar 12.78 6.78 Albite 1.80 0.64 Calcic Plagioclase 1.33 0.02 Muscovite 0.12 0.45 Biotite 0.51 2.34 Phlogopite 15.00 15.83 Illite 0.32 0.76 Chlorite 0.87 1.43 Saponite 4.64 5.28 Diopside 4.79 16.38 Olivine 1.25 2.66 Talc 12.45 2.85 Serpentine 14.27 14.27 Serpentine (Mg-rich) 2.55 3.14 Fe Serpentine 15.41 8.49 Al Serpentine 1.36 4.87 Calcite 0.23 1.73 Dolomite 0.36 0.06 Fe Oxides 0.39 0.33 Chromite 0.13 0.17 Epidote 0.09 0.14 360  Table A3.2 continued. Mineral Abundance (wt.%) CPK-1 CPK-2 Fe Sulphides 0.13 0.03 Wollastonite 0.37 1.41 Andradite 0.02 0.21 Cuspidine 0.01 0.59 Melilite 0.02 1.69 Ti Oxides 0.27 0.29 Apatite 2.01 0.46 Zircon 0.03 0.03 Barite 0.04 0.09 Undifferentiated 0.01 0.01 Total 100.00 100.00  A3.3.1.4 Thermogravimetric Analysis Representative TG and DTG curves have been plotted for each unreacted sample in Figure A3.11. 361   Figure A3.11. Representative TG and DTG curves of unreacted FPK-1, CPK-1, FPK-2 and CPK-2. Brucite would be indicated by a distinct peak from 300 to 425\u00b0C but is absent. Peaks at 60, 575-675, and 800\u00b0C are due to adsorbed water, clinochlore, serpentine, and smectite dehydroxylation, and talc dehydroxylation, respectively.362  A3.3.2 Standard Proctor Compaction The results of the Standard Proctor compaction analysis are presented in Table A3.3 and were used to plot the moisture content-dry density relationship. Table A3.3. Standard Proctor compaction test results for 25 and 33 wt.% FPK-2 mixtures with CPK-2. 25 wt.% FPK 33 wt.% FPK Moisture Content (wt.%)a Dry Density (g\u00b7m-3) Moisture Content (wt.%)a Dry Density (g\u00b7m-3) 2.0 1.94 4.1 1.90 2.0 1.94 4.4 1.94 8.1 2.04 8.4 2.04 10.5 2.08 8.5 2.04 12.2 2.04 10.0 2.10   11.0 2.10   13.2 1.97 a Moisture content measured as the mass of water over the mass of solids.  To convert from the maximum dry density (MDD) to the maximum bulk density (MBD), the following calculation was made as shown in Equation A3.25. Equation A3.25. !\"#(% \u22c5 '(!\") = !##(% \u22c5 '(!\") \u22c5 +, + #!$\"#\"$ %(%))** . Where \u2018Mc\u2019 is the moisture content, \u2018Mw\u2019 is the mass of water, and \u2018Ms\u2019 is the mass of solids. To convert the moisture content to the units of the water mass over the total mass (MT), Equation A3.26 was used. Equation A3.26. !+ \/###%0 (%) = #!$\"#\"$ %(%),)-\"!&\"#\"$ '(%)+,, . 363  A3.3.3 Pipe Experiment Permeability Permeability test conditions and results are presented in Table A3.4. The relative change in permeability after CO2 injection was calculated as shown in Equation A3.27. Equation A3.27. \u2206345(467898:;\t(%) = \/012134\t678973:1412;\t<9-=!>1034\t678973:1412;\t<9-=\/012134\t678973:1412;\t<9-=   364  Table A3.4. Permeability test conditions and results for Pipe-1 (283 cm2 area; SFG) and Pipe-2 (216 cm2 area; air). Fines (wt.%) Pore Saturation (%) Porosity (%) Moisture Content (%) Bulk Density (g\u00b7cm3) Length (cm) Flow Rate (L\u00b7min-3) Pressure (hPa) Permeability (m2) Inlet Outlet Ambient  Mean Pipe-1 Initial 24 34 6.5 1.58 581 0.72 969.9 968.1 968.1 2.5E-10 2.5E-10 Pipe-2 Initial 37 22 6.9 1.87 580 1 1023.1 1011.1 1006.6 6.7E-11 6.9E-11 0.75 1018.7 1009.8 1006.6 6.8E-11 0.5 1014.2 1008.6 1006.6 7.2E-11 Pipe-2 Final 37 22 6.9 1.87 580 1 1037.3 1021.2 1011.9 4.9E-11 4.9E-11 0.75 1030.7 1018.5 1011.9 4.9E-11 0.5 1024.4 1016.1 1011.9 4.9E-11 365  A3.3.4 Pipe and Pad Experiment CO2 Injection A3.3.4.1 Pipe-1 Flow Rate Analysis Pipe-1 suffered from a leak at the inlet after a power outage had occurred. CO2 was still injected into the pipe, as evidenced by the concentrations at the inlet being maintained and the concentrations at the middle continuing to increase, albeit at a lower rate than before the leak. However, the exact flow rate was no longer known as a portion of the gas was going into the atmosphere. This required the flow rate to be estimated. From Pipe-2, the behaviour of the CO2 concentrations when the flow rate is suddenly decreased is known (Figure 4.7 and 4.8). For Pipe-2, at 400 hours, the flow rate was reduced from 550 to 100 mL min-1. The result is that time is required for the concentrations to stabilize relative to the change in flow rate. While the flow rate changes instantly, the concentration changes gradually in response. Examining the carbonation rate (Figure 4.8), the rate stabilizes and follows the same trend as before the flow rate change. Pipe-1 does not follow the expected behaviour seen in Pipe-2. This is evidently due to the power outage that occurred directly before the leak commenced. Due to the power outage, concentrations had already decreased significantly. From Pipe-2, the behaviour following a power outage is also known. While concentrations dropped during the outage, they quickly recovered once gas flow was resumed. However, little to no recovery is observed for Pipe-1. Rather it appears that the concentrations merely continue to follow typical trends, such as the diurnal fluctuations. What this indicates is that the power outage and the resultant concentration decrease prepared the system for the flow rate change due to the leak. Instead of having a transition stage between the new and old flow rate, this transition occurred during the power outage. Rather than observing 366  concentrations recover once flow resumed, the concentrations in the pipe were already consistent with the new post-leak flow rate. To estimate the injected flow rate during the second phase of injection, the reactivity of Pipe-2 during the same time interval (52 \u2013 140 hours) was used as a proxy of what might have been expected from Pipe-1. The PK used in Pipe-1 is broadly the same as that used in Pipe-2. Further, from contrasting the centimetre-scale experiment PK-E3 with Pipe-2 (Figure 4.20), it can be seen that the reactivity rate of PK is transferable between experiments. Therefore, the maximum and minimum flow rates were selected based on the maximum and minimum reaction rates observed for Pipe-2 during the same time interval. These rates are equivalent to flow rates of 120 and 50 mL min-1, respectively. Pipe-1 did not see any CO2 in the effluent after the leak. Therefore, the injected rate cannot have exceeded the reaction rate. A range has been selected, rather than merely choosing the minimum reaction rate that would prevent CO2 in the effluent, as it is possible for the leak to have increased with time, meaning that the injection rate may not have been constant during phase 2. This range is, therefore, an upper bound on the possible injected flow rate. A3.3.4.2 Gas Phase Data The gas phase measurements for CO2 concentrations, relative humidity and temperatures for Pipe-1 and Pipe-2 are presented in 5 and 10-hour intervals in Tables A3.5 and A3.6, respectively.  CO2 concentrations in 5 and 10-hour intervals for Pad-1 and Pad-2 are presented in Tables A3.7 and A3.8 \u2013 A3.10, respectively. Sensor locations are shown in Figures A3.1 and A3.3.  367  Table A3.5. Pipe-1 gas CO2 concentrations, relative humidity\u2019s, and temperatures at 5-hour intervals. Time (Hours) [CO2] (vol.%) RH (%) Temperature (\u00baC) Inlet Middle Outlet Inlet Outlet Inlet Outlet Ambient 0 0.0 0.0 0.0 78.4 100.0 17.7 14.8 12.1 5 9.0 0.0 0.0 67.5 87.2 22.3 22.1 11.3 10 9.1 0.1 0.0 93.4 100.0 12.5 13.8 8.5 15 9.1 1.7 0.0 96.5 100.0 8.9 8.1 7.3 20 9.2 4.4 0.0 94.9 100.0 9.1 10.5 11.4 25 9.2 6.0 0.0 90.3 96.6 13.5 18.3 15.6 30 9.1 6.4 0.0 82.9 86.0 18.1 23.1 12.4 35 9.1 5.9 0.0 96.9 100.0 11.1 12.1 8.9 40 9.2 7.3 0.0 97.0 100.0 8.5 9.0 10.0 45 9.1 8.0 0.1 87.2 100.0 13.0 16.9 15.9 50 9.1 8.7 2.4 79.7 76.5 20.5 25.9 18.1 55 8.9 7.3 1.4 72.9 92.3 23.3 24.4 12.0 60 9.0 7.2 0.7 96.2 100.0 13.2 13.3 9.5 65 9.1 7.5 0.4 96.3 100.0 10.6 10.8 9.2 70 9.2 8.0 0.3 92.5 100.0 11.1 18.9 11.3 75 9.2 8.5 0.4 90.3 75.9 14.3 27.9 12.0 80 9.1 7.9 0.3 94.1 100.0 12.7 16.0 9.7 85 9.1 7.5 0.2 95.9 100.0 10.0 10.3 8.0 90 9.1 7.6 0.1 95.8 100.0 9.2 9.2 10.7 95 9.0 8.3 0.1 83.8 100.0 16.1 20.1 14.1 100 0.0 9.0 0.2 78.6 86.3 21.1 23.6 14.4 105 9.0 8.0 0.1 96.4 - 14.4 - 11.3 110 9.0 7.2 0.0 96.1 - 11.6 - 11.0 115 9.0 7.5 0.0 95.5 - 11.9 - 12.6 120 9.0 7.5 0.0 89.6 - 15.7 - 14.9 125 9.0 8.4 0.1 89.4 - 15.6 - 13.1 368  Table A3.5 continued. Time (Hours) [CO2] (vol.%) RH (%) Temperature (\u00baC) Inlet Middle Outlet Inlet Outlet Inlet Outlet Ambient 130 9.0 8.5 0.0 84.9 - 18.0 - 10.7 135 9.1 8.6 0.0 90.6 - 17.3 - 10.3 140 9.1 8.7 0.0 91.9 - 16.1 - 12.1 369  Table A3.6. Pipe-2 gas CO2 concentrations, relative humidity\u2019s, and temperatures at 5-hour intervals. Time (Hours)          [CO2] (vol.%)       RH (%) Temperature (\u00baC) Inlet Middle Outlet Inlet Outlet Inlet Outlet 0 0.0 0.0 0.0 97.5 100.0 20.9 20.6 5 9.7 0.0 0.0 96.4 100.0 21.1 20.6 10 9.7 0.9 0.0 95.9 100.0 21.4 20.8 15 9.7 5.3 0.0 95.6 100.0 21.5 21.0 20 9.7 6.5 0.0 95.9 100.0 21.6 21.0 25 9.7 7.1 0.0 95.8 100.0 21.7 21.2 30 9.8 7.5 0.1 95.7 100.0 21.7 21.1 35 9.8 7.8 2.8 95.8 100.0 21.6 21.0 40 9.8 8.0 4.5 95.8 100.0 21.4 20.7 45 9.8 8.3 5.5 96.0 99.9 22.1 21.5 50 9.8 8.4 6.0 94.8 99.0 22.3 22.0 55 9.8 8.4 6.4 95.1 99.3 22.3 22.1 60 9.8 8.5 6.6 95.7 100.0 22.0 21.5 65 9.8 8.6 6.9 95.4 100.0 22.0 21.4 70 9.8 8.7 7.2 95.3 99.3 22.3 22.0 75 9.8 8.6 7.2 95.7 100.0 21.9 21.3 80 9.8 8.7 7.3 95.8 100.0 21.6 20.9 85 9.7 8.7 7.5 95.4 100.0 21.5 20.8 90 9.8 8.8 7.7 95.2 99.6 21.7 21.2 95 9.4 8.1 7.2 96.2 99.5 21.7 21.3 100 9.7 8.7 7.1 95.7 100.0 21.3 20.4 105 9.7 8.8 7.7 95.5 100.0 20.9 19.9 110 9.8 8.9 8.0 95.2 99.7 21.3 20.7 115 9.7 8.9 8.1 95.2 99.4 21.5 20.9 120 9.7 8.9 8.1 95.5 99.7 21.4 20.7 125 9.7 9.0 8.2 94.9 98.8 21.7 21.3 370  Table A3.6 continued. Time (Hours)          [CO2] (vol.%)       RH (%) Temperature (\u00baC) Inlet Middle Outlet Inlet Outlet Inlet Outlet 130 9.8 9.0 8.2 95.6 99.7 21.4 20.8 135 9.8 9.0 8.3 95.3 99.6 21.4 20.7 140 9.8 9.0 8.3 95.3 99.5 21.4 20.7 145 9.7 9.0 8.3 95.2 99.6 21.4 20.7 150 9.7 9.1 8.4 95.0 98.9 21.6 21.2 155 9.7 9.0 8.4 95.3 99.7 21.3 20.7 160 9.8 9.1 8.4 95.3 99.6 21.2 20.7 165 9.8 9.1 8.4 95.2 99.5 21.2 20.6 170 9.8 9.1 8.5 95.3 99.4 21.3 20.7 175 9.8 9.1 8.5 95.4 99.5 21.2 20.6 180 9.8 9.1 8.5 95.1 99.3 21.2 20.7 185 9.8 9.1 8.6 95.2 99.3 21.2 20.7 190 9.8 9.2 8.6 95.2 99.3 21.2 20.7 195 9.8 9.2 8.6 95.3 99.3 21.2 20.7 200 9.7 9.1 8.6 95.3 99.4 21.2 20.7 205 9.8 9.2 8.6 95.2 99.4 21.2 20.6 210 9.8 9.2 8.7 95.3 99.3 21.1 20.6 215 9.8 9.2 8.7 95.4 99.3 21.2 20.7 220 9.8 9.3 8.7 95.2 99.4 21.1 20.6 225 9.8 9.3 8.8 95.3 99.3 21.1 20.6 230 9.8 9.3 8.8 95.2 99.3 21.1 20.6 235 9.9 9.3 8.8 95.2 99.3 21.2 20.6 240 9.9 9.4 8.9 95.0 98.9 21.4 20.9 245 9.9 9.3 8.9 95.4 99.9 21.0 20.3 250 9.9 9.3 8.9 95.4 100.0 20.8 19.9  371  Table A3.6 continued. Time (Hours)          [CO2] (vol.%)       RH (%) Temperature (\u00baC) Inlet Middle Outlet Inlet Outlet Inlet Outlet 255 9.9 9.3 8.9 95.2 100.0 20.6 19.8 260 9.9 9.4 9.0 94.6 98.9 21.1 20.7 265 10.0 9.5 9.1 94.4 98.5 21.5 21.3 270 9.9 9.5 9.1 95.1 99.1 21.5 21.2 275 9.9 9.4 9.0 95.6 99.9 21.1 20.6 280 9.9 9.4 9.0 95.2 99.8 21.1 20.5 285 9.9 9.4 9.0 95.0 99.6 20.9 20.4 290 9.9 9.4 9.0 95.1 99.7 20.9 20.3 295 9.9 9.4 9.0 94.9 99.6 20.9 20.3 300 9.8 9.4 9.0 94.9 99.6 20.9 20.3 305 9.8 9.4 9.0 95.1 99.3 20.9 20.4 310 9.8 9.4 9.0 96.0 99.5 20.7 20.2 315 9.8 9.3 9.0 96.4 99.7 20.7 20.2 320 9.8 9.3 9.0 96.5 99.8 20.6 20.2 325 9.8 9.3 9.0 96.5 99.8 20.6 20.2 330 9.8 9.3 8.9 96.6 100.0 20.5 19.9 335 9.8 9.4 9.1 96.4 99.8 20.7 20.3 340 9.8 9.4 9.1 96.5 99.7 20.8 20.4 345 9.8 9.3 9.0 96.7 99.9 20.7 20.3 350 9.8 9.3 9.0 96.9 100.0 20.3 19.4 355 9.8 9.4 9.1 96.2 100.0 20.5 20.0 360 9.8 9.4 9.2 96.2 99.8 21.2 20.8 365 9.8 9.4 9.1 96.7 100.0 20.9 20.3 370 9.8 9.3 9.1 96.7 100.0 20.7 20.3 375 9.8 9.3 9.0 96.8 100.0 20.4 19.7 380 9.8 9.3 9.0 96.5 100.0 20.6 20.1  372  Table A3.6 continued. Time (Hours)          [CO2] (vol.%)       RH (%) Temperature (\u00baC) Inlet Middle Outlet Inlet Outlet Inlet Outlet 385 9.7 9.4 9.2 96.3 99.6 21.2 20.9 390 9.8 9.3 9.1 96.7 100.0 20.9 20.3 395 9.8 9.4 9.1 96.5 100.0 20.8 20.3 400 9.8 9.4 9.2 96.3 100.0 20.8 20.3 405 9.7 9.3 9.1 96.5 99.9 21.1 20.6 410 9.6 8.8 8.5 96.3 99.9 21.2 20.6 415 9.6 8.4 8.0 96.4 100.0 21.1 20.5 420 9.6 8.2 7.6 96.4 100.0 21.0 20.4 425 9.6 8.1 7.2 96.2 100.0 21.0 20.5 430 9.6 8.1 7.1 96.3 99.9 21.0 20.5 435 9.6 8.1 6.9 96.3 100.0 21.1 20.5 440 9.6 8.1 6.8 96.7 100.0 20.6 19.8 445 9.6 8.0 6.7 96.1 100.0 20.8 20.3 450 9.6 8.1 6.8 96.2 99.9 21.0 20.4 455 9.6 8.2 6.8 96.3 100.0 21.0 20.4 460 9.6 8.2 6.8 96.3 100.0 20.9 20.4 465 9.6 8.2 6.9 96.4 100.0 20.9 20.4 470 9.6 8.3 6.9 96.4 99.9 21.0 20.5 475 9.6 8.3 6.9 96.5 99.8 21.1 20.7  373  Table A3.7. Pad-1 gas CO2 concentrations at 5-hour intervals. Time (Hours) [CO2] (vol.%) Inlet Outlet S1 S2 S3 S4 S5 Headspace 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5 3.0 0.0 - 0.0 0.0 0.0 - - 10 3.7 0.0 - 0.0 0.0 0.0 - - 15 4.2 0.0 - 0.0 0.0 0.0 - - 20 4.5 0.0 0.7 0.0 0.0 0.0 0.7 0.1 25 7.2 0.0 2.1 0.0 0.0 0.0 2.3 0.1 30 7.6 0.0 3.0 0.0 0.0 0.0 3.6 0.5 35 7.8 0.0 3.5 0.0 0.0 0.0 4.2 0.9 40 8.0 0.0 3.8 0.0 0.0 0.0 4.8 1.2 45 8.1 0.0 4.2 0.0 0.0 0.0 5.1 1.4 50 8.2 0.1 4.5 0.0 0.1 0.0 5.5 1.7 55 8.3 0.0 4.8 0.0 0.2 0.0 5.8 1.9 60 8.3 0.0 4.9 0.0 0.3 0.0 6.0 2.1 65 8.4 0.0 5.2 0.1 0.4 0.1 6.0 2.3 70 8.5 0.0 5.2 0.1 0.5 0.1 6.5 2.5 75 8.5 0.9 5.5 0.2 0.7 0.2 6.5 2.6 80 8.4 1.2 - 0.2 0.8 0.3 6.3 2.9 85 8.4 1.5 - 0.3 0.9 0.4 6.3 3.1 90 8.4 1.8 - 0.3 1.0 0.5 6.5 3.1 95 8.4 4.9 5.5 0.4 1.1 0.6 7.0 3.1 100 8.4 4.8 5.3 0.5 1.1 0.6 6.6 3.2 105 8.5 2.3 - 0.6 1.2 0.7 6.6 3.2 110 8.6 2.6 - 0.7 1.4 0.8 6.7 3.5 115 8.6 2.8 6.1 0.9 1.5 0.9 6.8 3.4 120 8.7 3.6 6.7 0.9 1.7 1.0 7.1 3.5 125 8.7 1.9 6.9 1.0 1.9 1.2 7.1 3.5 374  Table A3.7 continued. Time (Hours) [CO2] (vol.%) Inlet Outlet S1 S2 S3 S4 S5 Headspace 130 8.6 2.7 - 1.1 2.3 1.3 7.0 3.8 135 8.6 3.0 - 1.1 2.5 1.4 7.1 3.8 140 8.6 3.3 6.0 1.2 2.7 1.5 - - 375  Table A3.8. Pad-2 gas CO2 concentrations at 5- or 10-hour intervals. Time (Hours) [CO2] (vol.%) Inlet Outlet 1-1 1-2 1-3 1-4 1-5 1-6 1-7 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5 9.3 0.0 9.6 9.1 8.6 1.6 0.0 0.0 0.0 10 9.5 0.0 9.6 9.2 9.0 3.5 0.2 0.0 0.0 15 9.6 0.1 9.7 9.3 9.2 4.7 0.8 0.0 0.0 20 9.7 0.2 9.8 9.5 9.3 5.5 1.7 0.0 0.0 25 9.7 0.3 9.7 9.5 9.3 6.1 2.5 0.0 0.0 30 9.7 0.5 9.8 9.5 9.4 6.5 3.2 0.2 0.0 35 9.7 0.6 9.7 9.5 9.3 6.8 3.8 0.5 0.0 40 9.7 0.8 9.7 9.5 9.4 7.1 4.3 0.9 0.0 45 9.7 1.0 9.7 9.5 9.4 7.3 4.8 1.4 0.0 50 9.7 1.2 9.6 9.5 9.3 7.6 5.5 2.2 0.1 55 9.7 1.5 9.7 9.5 9.3 7.8 5.8 2.7 0.2 60 9.7 1.7 9.7 9.5 9.4 7.9 6.0 3.1 0.4 65 9.7 1.8 9.7 9.5 9.4 8.0 6.3 3.6 0.6 70 9.8 1.9 9.7 9.5 9.5 8.3 6.9 4.5 1.2 75 9.7 2.1 9.6 9.5 9.3 8.3 7.0 4.9 1.7 80 9.7 2.3 9.6 9.5 9.4 8.3 7.0 5.0 1.9 85 9.7 2.5 9.6 9.4 9.3 8.3 7.1 5.1 2.2 90 9.7 2.9 9.6 9.5 9.4 8.3 7.1 5.2 2.4 95 9.7 3.2 9.6 9.4 9.4 8.4 7.5 5.7 3.1 100 9.8 3.5 9.7 9.6 9.5 8.5 7.6 5.9 3.5 105 9.8 3.6 9.7 9.5 9.5 8.4 7.4 5.7 3.5 110 9.8 3.7 9.7 9.6 9.5 8.5 7.4 5.7 3.5 115 9.8 4.0 9.8 9.6 9.5 8.6 7.5 5.8 3.7 120 9.8 4.2 9.7 9.6 9.5 8.6 7.6 5.9 3.9 125 9.8 4.4 9.7 9.5 9.5 8.6 7.6 6.1 4.1 376  Table A3.8 continued. Time (Hours) [CO2] (vol.%) Inlet Outlet 1-1 1-2 1-3 1-4 1-5 1-6 1-7 130 9.8 4.6 9.7 9.5 9.5 8.6 7.7 6.2 4.3 135 9.8 4.8 9.7 9.6 9.5 8.6 7.8 6.3 4.5 140 9.8 5.0 9.7 9.5 9.5 8.6 7.8 6.4 4.8 145 9.8 5.1 9.7 9.5 9.4 8.7 7.9 6.5 4.9 150 9.8 5.3 9.7 9.6 9.5 8.6 7.9 6.6 5.0 155 9.8 5.4 9.7 9.6 9.5 8.7 7.9 6.7 5.2 160 9.8 5.6 9.7 9.6 9.5 8.8 8.0 6.8 5.3 165 9.6 4.9 9.6 9.3 9.2 8.1 7.3 6.1 4.8 170 9.5 4.5 9.6 9.3 9.1 8.0 7.2 6.0 4.7 175 9.5 4.3 9.6 9.3 9.1 8.0 7.1 5.9 4.6 180 9.6 4.1 9.6 9.3 9.1 7.9 7.1 5.8 4.5 185 9.6 4.1 9.7 9.4 9.1 8.0 7.1 5.8 4.5 190 9.6 4.1 9.7 9.4 9.1 8.0 7.1 5.9 4.6 195 9.6 4.2 9.6 9.4 9.1 8.0 7.2 5.9 4.7 200 9.6 4.2 9.7 9.4 9.1 8.1 7.2 6.0 4.7 205 9.6 4.3 9.7 9.4 9.1 8.1 7.3 6.1 4.8 210 9.6 4.3 9.7 9.5 9.1 8.1 7.3 6.1 4.9 215 9.6 4.4 9.7 9.5 9.2 8.1 7.3 6.2 5.0 220 9.6 4.5 9.6 9.4 9.1 8.1 7.4 6.2 5.0 225 9.6 4.5 9.7 9.5 9.1 8.2 7.4 6.3 5.1 230 9.6 4.6 9.7 9.5 9.2 8.2 7.4 6.3 5.2 235 9.6 4.6 9.8 9.5 9.1 8.1 7.3 6.2 5.2 240 9.7 4.7 9.7 9.6 9.2 8.2 7.5 6.3 5.2 245 9.7 4.7 9.7 9.5 9.2 8.3 7.6 6.4 5.3 250 9.7 4.7 9.8 9.6 9.3 8.3 7.6 6.5 5.3 255 9.8 4.8 9.8 9.6 9.3 8.3 7.6 6.5 5.3  377  Table A3.8 continued. Time (Hours) [CO2] (vol.%) Inlet Outlet 1-1 1-2 1-3 1-4 1-5 1-6 1-7 260 9.8 4.8 9.8 9.6 9.3 8.3 7.6 6.5 5.4 265 9.7 4.9 9.8 9.5 9.3 8.4 7.7 6.6 5.4 270 9.7 4.9 9.7 9.6 9.2 8.4 7.7 6.6 5.4 275 9.7 5.1 9.7 9.6 9.3 8.4 7.7 6.7 5.7 280 9.8 5.2 9.8 9.6 9.3 8.4 7.8 6.8 5.9 285 9.8 5.4 9.8 9.7 9.3 8.5 7.8 6.8 5.9 290 9.8 5.5 9.8 9.6 9.3 8.5 7.8 6.8 5.9 295 9.8 5.6 9.8 9.6 9.2 8.5 7.8 6.8 5.8 300 9.8 5.7 9.9 9.7 9.3 8.5 7.8 6.8 5.8 305 9.8 5.8 9.9 9.7 9.3 8.5 7.9 6.9 5.8 310 9.9 5.8 9.9 9.7 9.3 8.5 7.9 6.9 5.8 315 9.8 5.9 9.8 9.7 9.3 8.5 7.9 6.9 5.9 320 9.8 5.9 9.8 9.6 9.3 8.5 7.9 6.9 5.9 325 9.8 6.0 9.9 9.7 9.3 8.5 7.9 6.9 6.0 330 9.8 6.0 9.9 9.7 9.3 8.6 7.9 7.0 6.0 335 9.8 6.1 9.8 9.7 9.4 8.6 8.0 7.0 6.0 340 9.8 6.1 9.8 9.7 9.3 8.6 8.0 7.1 6.1 345 9.9 6.2 9.9 9.8 9.3 8.6 8.0 7.1 6.1 350 9.8 6.2 9.9 9.7 9.3 8.6 8.1 7.1 6.1 355 9.8 6.2 9.9 9.7 9.4 8.6 8.1 7.1 6.2 360 9.8 6.3 9.8 9.7 9.3 8.6 8.1 7.1 6.2 365 9.8 6.3 9.8 9.7 9.3 8.7 8.1 7.2 6.3 370 9.8 6.4 9.8 9.6 9.3 8.6 8.1 7.2 6.3 375 9.8 6.4 9.8 9.7 9.3 8.7 8.1 7.3 6.4 380 9.9 6.5 9.8 9.7 9.4 8.7 8.2 7.3 6.4 385 9.8 6.5 9.8 9.7 9.4 8.7 8.2 7.4 6.5  378  Table A3.8 continued. Time (Hours) [CO2] (vol.%) Inlet Outlet 1-1 1-2 1-3 1-4 1-5 1-6 1-7 390 9.9 6.6 9.9 9.8 9.4 8.8 8.2 7.4 6.5 395 9.9 6.6 9.9 9.7 9.4 8.8 8.3 7.4 6.5 400 9.9 6.7 9.9 9.7 9.4 8.8 8.3 7.5 6.6 405 9.9 6.7 9.9 9.8 9.4 8.8 8.3 7.5 6.6 410 9.8 6.7 9.8 9.7 9.4 8.8 8.3 7.5 6.6 415 9.8 6.8 9.8 9.7 9.4 8.8 8.3 7.5 6.7 420 9.8 6.8 9.8 9.6 9.4 8.8 8.3 7.5 6.7 425 9.8 6.8 9.8 9.6 9.4 8.8 8.4 7.6 6.8 430 9.8 6.8 9.8 9.6 9.3 8.8 8.4 7.6 6.7 435 9.5 6.7 9.6 9.3 8.9 8.4 8.0 7.3 6.6 440 9.4 6.4 9.7 9.3 8.7 8.2 7.7 7.0 6.4 445 9.4 6.1 9.7 9.2 8.6 8.0 7.6 6.8 6.2 450 9.4 5.9 9.7 9.3 8.6 7.9 7.5 6.7 6.0 455 9.4 5.7 9.7 9.3 8.6 7.9 7.4 6.5 5.8 460 9.4 5.5 9.7 9.2 8.6 7.8 7.3 6.5 5.7 465 9.4 5.4 9.7 9.2 8.5 7.8 7.2 6.4 5.6 470 9.4 5.3 9.7 9.2 8.5 7.8 7.2 6.3 5.6 475 9.3 5.3 9.8 9.2 8.4 7.7 7.1 6.2 5.4 480 9.3 5.2 9.7 9.2 8.4 7.6 7.0 6.1 5.3 485 9.3 5.1 9.7 9.1 8.3 7.6 7.0 6.0 5.2 490 9.2 5.1 9.7 9.1 8.3 7.6 7.0 6.0 5.1 495 9.2 5.1 9.7 9.1 8.3 7.6 7.0 5.9 5.1 500 9.3 5.1 9.8 9.1 8.3 7.6 7.0 5.9 5.1 510 9.3 5.0 9.8 9.2 8.3 7.6 6.9 5.9 5.0 520 9.3 5.0 9.8 9.2 8.4 7.6 7.0 5.9 5.0 530 9.3 5.0 9.7 9.2 8.6 7.7 7.0 6.0 5.2  379  Table A3.8 continued. Time (Hours) [CO2] (vol.%) Inlet Outlet 1-1 1-2 1-3 1-4 1-5 1-6 1-7 540 9.3 5.0 9.6 9.1 8.6 7.8 7.1 6.2 5.3 550 9.3 4.9 9.6 9.1 8.6 7.8 7.1 6.2 5.3 560 9.3 4.9 9.5 9.1 8.6 7.8 7.1 6.2 5.3 570 9.3 4.9 9.6 9.1 8.6 7.8 7.2 6.2 5.3 580 9.3 4.9 9.5 9.1 8.6 7.8 7.2 6.2 5.3 590 9.3 4.9 9.6 9.1 8.6 7.8 7.2 6.3 5.4 600 9.3 4.9 9.5 9.1 8.6 7.9 7.2 6.3 5.4 610 9.3 5.0 9.6 9.1 8.6 7.9 7.3 6.4 5.6 620 9.3 5.2 9.6 9.1 8.6 7.9 7.4 6.5 5.7 630 9.3 5.2 9.6 9.0 8.6 7.9 7.4 6.6 5.8 640 9.3 5.3 9.6 9.1 8.7 8.0 7.4 6.6 5.8 650 9.3 5.4 9.5 9.1 8.6 8.0 7.5 6.7 5.9 660 9.3 5.5 9.6 9.1 8.7 8.0 7.5 6.8 6.0 670 9.3 5.5 9.6 9.1 8.7 8.1 7.6 6.8 5.9 680 9.4 5.6 9.6 9.2 8.7 8.1 7.6 6.9 6.1 690 9.4 5.6 9.6 9.2 8.7 8.1 7.6 6.9 6.1 700 9.4 5.5 9.7 9.2 8.8 8.1 7.6 6.8 6.0 710 9.4 5.6 9.6 9.1 8.8 8.1 7.7 6.9 6.1 720 9.3 5.7 9.5 9.1 8.7 8.1 7.7 6.9 6.2 730 9.4 5.7 9.6 9.2 8.8 8.2 7.7 7.0 6.2 740 9.4 5.8 9.6 9.1 8.8 8.2 7.7 7.0 6.3 750 9.4 5.8 9.6 9.1 8.8 8.2 7.7 7.0 6.3 760 9.3 5.9 9.5 9.1 8.8 8.2 7.8 7.0 6.3 380  Table A3.9. Pad-2 gas CO2 concentrations at 5- or 10-hour intervals. Time (Hours) [CO2] (vol.%) 2-1 2-2 2-3 2-4 2-5 HS-1 HS-2 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5 1.6 0.0 0.0 0.0 0.0 0.0 0.0 10 3.6 0.7 0.0 0.0 0.0 0.0 0.0 15 4.7 1.8 0.0 0.0 0.0 0.1 0.0 20 5.5 2.7 0.0 0.0 0.0 0.3 0.1 25 6.0 0.0 0.0 0.0 0.0 0.4 0.2 30 6.5 4.0 0.6 0.0 0.0 0.5 0.4 35 6.9 4.5 1.2 0.0 0.0 0.7 0.5 40 7.2 5.0 1.8 0.0 0.0 0.9 0.7 45 7.4 5.5 2.3 0.0 0.0 1.1 0.9 50 7.6 5.9 3.1 0.0 0.0 1.3 1.1 55 7.8 6.1 3.5 0.3 0.0 1.5 1.4 60 7.9 6.5 4.0 0.9 0.0 1.7 1.6 65 8.0 6.7 4.4 1.4 0.0 1.9 1.7 70 8.2 7.1 5.2 2.3 0.0 1.9 1.8 75 8.2 7.3 5.5 2.9 0.0 2.1 2.0 80 8.3 7.4 5.7 3.2 0.1 2.7 2.6 85 8.3 7.5 5.8 3.5 0.3 3.1 2.9 90 8.4 7.5 5.9 3.7 0.6 3.4 3.3 95 8.5 7.7 6.3 4.4 1.3 3.6 3.6 100 8.6 7.9 6.5 4.8 2.0 4.0 4.0 105 8.6 7.8 6.5 4.8 2.3 4.3 4.3 110 8.6 7.9 6.5 4.6 2.4 4.5 4.5 115 8.7 7.9 6.6 4.8 2.7 4.8 4.8 120 8.7 8.0 6.7 5.0 2.9 5.1 5.0 125 8.7 8.1 6.9 5.2 3.2 5.3 5.2 381  Table A3.9 continued. Time (Hours) [CO2] (vol.%) 2-1 2-2 2-3 2-4 2-5 HS-1 HS-2 130 8.7 8.1 6.9 5.4 3.5 5.4 5.4 135 8.7 8.2 7.0 5.6 3.8 5.6 5.5 140 8.8 8.1 7.1 5.7 4.1 5.7 5.6 145 8.8 8.3 7.2 5.8 4.3 5.9 5.8 150 8.8 8.3 7.3 6.0 4.5 6.0 5.9 155 8.9 8.3 7.4 6.1 4.7 6.1 6.0 160 8.9 8.4 7.4 6.3 4.9 6.2 6.1 165 8.5 7.9 6.9 5.7 4.6 5.1 4.9 170 8.3 7.6 6.6 5.5 4.4 4.5 4.1 175 8.3 7.6 6.6 5.4 4.3 4.2 3.7 180 8.2 7.6 6.5 5.4 4.3 4.0 3.4 185 8.3 7.6 6.5 5.4 4.3 3.8 3.2 190 8.3 7.6 6.5 5.4 4.3 3.8 3.2 195 8.3 7.7 6.6 5.5 4.4 3.8 3.2 200 8.4 7.7 6.7 5.6 4.5 3.8 3.2 205 8.4 7.7 6.7 5.6 4.6 3.8 3.2 210 8.4 7.7 6.8 5.7 4.7 3.7 3.2 215 8.4 7.8 6.8 5.6 4.7 3.8 3.2 220 8.4 7.8 6.8 5.8 4.8 3.9 3.3 225 8.4 7.9 6.9 5.9 4.9 3.9 3.3 230 8.5 7.8 7.0 6.0 5.0 4.0 3.4 235 8.4 7.8 6.8 5.9 5.1 4.0 3.4 240 8.5 7.9 7.0 6.0 5.1 4.0 3.5 245 8.5 7.9 7.0 6.1 5.2 4.0 3.4 250 8.6 8.0 7.1 6.2 5.1 4.0 3.5 255 8.6 8.1 7.2 6.2 5.1 4.0 3.5  382  Table A3.9 continued. Time (Hours) [CO2] (vol.%) 2-1 2-2 2-3 2-4 2-5 HS-1 HS-2 260 8.6 8.2 7.2 6.2 5.1 4.1 3.5 265 8.7 8.1 7.2 6.2 5.2 4.1 3.5 270 8.7 8.1 7.2 6.4 5.2 4.2 3.6 275 8.7 8.2 7.4 6.5 5.6 4.2 3.6 280 8.8 8.2 7.5 6.6 5.9 4.7 4.0 285 8.8 8.3 7.4 6.7 5.9 5.3 4.5 290 8.9 8.3 7.5 6.7 5.9 5.7 4.9 295 8.9 8.2 7.5 6.7 5.8 5.8 5.0 300 8.9 8.4 7.5 6.6 5.8 5.9 5.1 305 8.9 8.3 7.5 6.7 5.8 5.9 5.2 310 8.9 8.4 7.5 6.6 5.8 6.0 5.2 315 8.9 8.4 7.5 6.6 5.8 6.1 5.3 320 8.9 8.4 7.6 6.8 5.8 6.1 5.3 325 8.9 8.4 7.5 6.9 5.9 6.3 5.5 330 8.9 8.4 7.6 6.7 5.9 6.2 5.5 335 9.0 8.4 7.7 6.9 6.0 6.3 5.5 340 9.0 8.5 7.6 6.9 6.0 6.3 5.6 345 9.0 8.5 7.7 6.9 6.1 6.3 5.6 350 9.0 8.4 7.7 7.0 6.1 6.4 5.6 355 9.0 8.5 7.7 6.9 6.1 6.4 5.7 360 9.0 8.6 7.8 7.0 6.2 6.4 5.7 365 9.0 8.6 7.8 7.2 6.3 6.5 5.8 370 9.0 8.6 7.9 7.2 6.3 6.5 5.8 375 9.0 8.5 7.9 7.3 6.4 6.6 5.9 380 9.1 8.7 7.9 7.3 6.4 6.6 5.9 385 9.1 8.7 7.9 7.3 6.5 6.7 6.0  383  Table A3.9 continued. Time (Hours) [CO2] (vol.%) 2-1 2-2 2-3 2-4 2-5 HS-1 HS-2 390 9.1 8.8 7.9 7.4 6.5 6.7 6.0 395 9.1 8.7 8.1 7.3 6.6 6.7 6.1 400 9.1 8.7 8.1 7.4 6.6 6.7 6.1 405 9.1 8.7 8.1 7.4 6.7 6.8 6.1 410 9.1 8.7 8.1 7.5 6.7 6.8 6.2 415 9.1 8.7 8.1 7.4 6.7 6.8 6.2 420 9.1 8.8 8.1 7.5 6.7 6.9 6.2 425 9.1 8.7 8.1 7.5 6.8 6.9 6.3 430 9.1 8.7 8.1 7.6 6.8 6.9 6.3 435 8.8 8.6 7.9 7.3 6.7 6.6 6.1 440 8.6 8.2 7.6 7.0 6.5 6.2 5.7 445 8.5 8.0 7.4 6.8 6.3 5.9 5.4 450 8.5 7.9 7.2 6.7 6.1 5.6 5.2 455 8.4 7.9 7.1 6.4 6.0 5.4 4.9 460 8.4 7.8 7.0 6.4 5.8 5.3 4.8 465 8.4 7.8 7.0 6.3 5.8 5.2 4.6 470 8.3 7.7 6.9 6.4 5.7 5.1 4.5 475 8.3 7.7 6.8 6.1 5.5 5.0 4.4 480 8.2 7.6 6.7 6.0 5.3 5.0 4.4 485 8.2 7.6 6.6 5.8 5.3 5.0 4.4 490 8.2 7.6 6.5 5.9 5.2 5.0 4.4 495 8.2 7.6 6.6 5.9 5.1 5.0 4.4 500 8.2 7.5 6.5 5.8 5.1 5.0 4.4 510 8.2 7.6 6.5 5.7 5.0 5.0 4.4 520 8.3 7.6 6.6 5.7 5.0 5.0 4.4 530 8.3 7.7 6.6 5.8 5.2 4.8 4.3  384  Table A3.9 continued. Time (Hours) [CO2] (vol.%) 2-1 2-2 2-3 2-4 2-5 HS-1 HS-2 540 8.3 7.7 6.8 6.0 5.3 4.5 4.0 550 8.4 7.7 6.8 6.0 5.3 4.3 3.7 560 8.3 7.7 6.9 6.0 5.3 4.1 3.6 570 8.4 7.8 6.9 6.1 5.3 4.0 3.5 580 8.4 7.8 6.9 6.0 5.3 4.0 3.5 590 8.4 7.8 6.9 6.1 5.4 4.0 3.4 600 8.4 7.8 7.0 6.2 5.4 4.0 3.4 610 8.4 7.8 7.0 6.2 5.7 4.1 3.5 620 8.4 7.9 7.1 6.4 5.8 4.3 3.7 630 8.4 7.8 7.1 6.4 5.8 4.4 3.8 640 8.5 7.9 7.2 6.5 5.9 4.4 3.8 650 8.5 8.1 7.2 6.6 6.0 4.5 3.9 660 8.5 8.1 7.3 6.7 6.1 4.6 4.0 670 8.6 8.1 7.4 6.8 6.1 4.7 4.1 680 8.6 8.1 7.4 6.8 6.2 4.7 4.1 690 8.6 8.2 7.4 6.8 6.2 4.7 4.1 700 8.6 8.1 7.4 6.8 6.1 4.6 4.1 710 8.6 8.1 7.5 6.8 6.2 4.7 4.1 720 8.6 8.1 7.4 6.9 6.3 4.7 4.1 730 8.6 8.2 7.5 6.9 6.4 4.8 4.2 740 8.6 8.2 7.5 6.9 6.4 5.0 4.4 750 8.6 8.1 7.5 6.9 6.4 5.0 4.4 760 8.6 8.1 7.5 7.0 6.5 5.0 4.4 385  Table A3.10. Pad-2 gas CO2 concentrations at 5- or 10-hour intervals. Time (Hours) [CO2] (vol.%) 3-1 3-2 3-3 3-4 3-5 3-6 3-7 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 10 0.0 0.0 0.0 0.0 0.0 0.0 0.0 15 0.0 0.0 0.0 0.0 0.0 0.0 0.0 20 0.3 0.0 0.0 0.0 0.3 0.3 0.3 25 1.0 0.0 0.0 0.0 1.0 1.0 1.0 30 2.1 0.0 0.0 0.0 2.1 2.1 2.1 35 3.0 0.0 0.0 0.0 3.0 3.0 3.0 40 3.8 0.5 0.0 0.0 3.8 3.8 3.8 45 4.5 1.6 0.0 0.0 4.5 4.5 4.5 50 5.0 2.5 0.0 0.0 5.0 5.0 5.0 55 5.5 3.1 0.0 0.0 5.5 5.5 5.5 60 5.9 3.6 0.3 0.0 5.9 5.9 5.9 65 6.2 3.9 1.0 0.0 6.2 6.2 6.2 70 6.4 4.8 2.0 0.0 6.4 6.4 6.4 75 6.7 5.1 2.8 0.0 6.7 6.7 6.7 80 6.8 5.2 3.2 0.0 6.8 6.8 6.8 85 6.9 5.1 3.6 0.4 6.9 6.9 6.9 90 7.1 5.6 4.0 1.0 7.1 7.1 7.1 95 7.2 5.5 4.4 2.0 7.2 7.2 7.2 100 7.3 6.1 4.7 2.7 7.3 7.3 7.3 105 7.4 6.0 4.9 3.0 7.4 7.4 7.4 110 7.4 6.2 4.9 3.1 7.4 7.4 7.4 115 7.6 6.5 5.1 3.4 7.6 7.6 7.6 120 7.7 6.3 5.3 3.7 7.7 7.7 7.7 125 7.7 6.5 5.4 4.0 7.7 7.7 7.7 386  Table A3.10 continued. Time (Hours) [CO2] (vol.%) 3-1 3-2 3-3 3-4 3-5 3-6 3-7 130 7.7 6.3 5.6 4.2 7.7 7.7 7.7 135 7.8 6.5 5.8 4.5 7.8 7.8 7.8 140 7.9 6.7 5.8 4.7 7.9 7.9 7.9 145 7.9 6.7 6.0 4.9 7.9 7.9 7.9 150 7.9 6.1 6.1 5.0 7.9 7.9 7.9 155 8.0 6.4 6.2 5.3 8.0 8.0 8.0 160 8.0 7.3 6.3 5.3 8.0 8.0 8.0 165 7.8 6.7 6.1 5.1 7.8 7.8 7.8 170 7.5 6.3 5.7 4.9 7.5 7.5 7.5 175 7.4 6.3 5.6 4.8 7.4 7.4 7.4 180 7.4 6.2 5.6 4.8 7.4 7.4 7.4 185 7.4 6.1 5.6 4.8 7.4 7.4 7.4 190 7.5 6.2 5.6 4.8 7.5 7.5 7.5 195 7.5 6.3 5.7 4.9 7.5 7.5 7.5 200 7.5 6.3 5.7 5.0 7.5 7.5 7.5 205 7.5 6.7 5.8 5.1 7.5 7.5 7.5 210 7.6 6.2 5.9 5.1 7.6 7.6 7.6 215 7.7 6.6 5.9 5.3 7.7 7.7 7.7 220 7.7 6.2 6.0 5.3 7.7 7.7 7.7 225 7.7 6.5 6.1 5.3 7.7 7.7 7.7 230 7.7 6.7 6.1 5.4 7.7 7.7 7.7 235 7.7 6.7 6.1 5.5 7.7 7.7 7.7 240 7.7 6.7 6.1 5.5 7.7 7.7 7.7 245 7.8 6.5 6.2 5.5 7.8 7.8 7.8 250 7.9 6.7 6.3 5.5 7.9 7.9 7.9 255 7.9 6.9 6.3 5.5 7.9 7.9 7.9  387  Table A3.10 continued. Time (Hours) [CO2] (vol.%) 3-1 3-2 3-3 3-4 3-5 3-6 3-7 260 8.0 6.6 6.4 5.6 8.0 8.0 8.0 265 8.0 6.8 6.4 5.7 8.0 8.0 8.0 270 8.0 6.7 6.4 5.6 8.0 8.0 8.0 275 8.0 6.7 6.5 5.9 8.0 8.0 8.0 280 8.1 7.1 6.7 6.2 8.1 8.1 8.1 285 8.2 6.6 6.7 6.2 8.2 8.2 8.2 290 8.2 7.2 6.7 6.2 8.2 8.2 8.2 295 8.2 6.7 6.8 6.2 8.2 8.2 8.2 300 8.2 7.3 6.7 6.2 8.2 8.2 8.2 305 8.3 7.0 6.8 6.2 8.3 8.3 8.3 310 8.2 6.7 6.8 6.3 8.2 8.2 8.2 315 8.2 7.0 6.9 6.3 8.2 8.2 8.2 320 8.2 7.1 6.8 6.3 8.2 8.2 8.2 325 8.3 6.5 6.9 6.4 8.3 8.3 8.3 330 8.3 7.0 7.0 6.3 8.3 8.3 8.3 335 8.4 7.3 7.0 6.4 8.4 8.4 8.4 340 8.4 7.2 7.0 6.5 8.4 8.4 8.4 345 8.4 6.9 7.0 6.6 8.4 8.4 8.4 350 8.4 7.1 7.1 6.6 8.4 8.4 8.4 355 8.4 7.3 7.1 6.6 8.4 8.4 8.4 360 8.4 7.0 7.1 6.6 8.4 8.4 8.4 365 8.4 7.2 7.2 6.7 8.4 8.4 8.4 370 8.4 7.1 7.2 6.8 8.4 8.4 8.4 375 8.5 7.5 7.2 6.8 8.5 8.5 8.5 380 8.5 7.3 7.3 6.8 8.5 8.5 8.5 385 8.5 7.2 7.3 6.9 8.5 8.5 8.5  388  Table A3.10 continued. Time (Hours) [CO2] (vol.%) 3-1 3-2 3-3 3-4 3-5 3-6 3-7 390 8.5 7.1 7.3 6.9 8.5 8.5 8.5 395 8.6 7.3 7.4 7.0 8.6 8.6 8.6 400 8.6 7.6 7.4 7.1 8.6 8.6 8.6 405 8.6 7.3 7.5 7.1 8.6 8.6 8.6 410 8.6 8.0 7.4 7.0 8.6 8.6 8.6 415 8.6 7.7 7.5 7.2 8.6 8.6 8.6 420 8.6 7.5 7.5 7.1 8.6 8.6 8.6 425 8.6 7.7 7.5 7.2 8.6 8.6 8.6 430 8.6 6.9 7.5 7.1 8.6 8.6 8.6 435 8.4 7.5 7.4 7.1 8.4 8.4 8.4 440 8.2 6.8 7.2 6.8 8.2 8.2 8.2 445 8.0 7.0 6.9 6.6 8.0 8.0 8.0 450 7.9 7.1 6.7 6.4 7.9 7.9 7.9 455 7.9 6.7 6.6 6.3 7.9 7.9 7.9 460 7.8 6.9 6.6 6.2 7.8 7.8 7.8 465 7.8 6.7 6.5 6.0 7.8 7.8 7.8 470 7.8 7.3 6.5 6.0 7.8 7.8 7.8 475 7.8 6.7 6.4 5.8 7.8 7.8 7.8 480 7.7 6.7 6.3 5.7 7.7 7.7 7.7 485 7.7 6.6 6.2 5.6 7.7 7.7 7.7 490 7.6 6.5 6.1 5.6 7.6 7.6 7.6 495 7.6 6.7 6.2 5.5 7.6 7.6 7.6 500 7.6 6.3 6.1 5.5 7.6 7.6 7.6 510 7.6 6.7 6.1 5.4 4.8 4.9 5.4 520 7.7 6.6 6.1 5.5 4.7 4.9 5.5 530 7.8 6.5 6.2 5.6 5.0 4.9 5.4  389  Table A3.10 continued. Time (Hours) [CO2] (vol.%) 3-1 3-2 3-3 3-4 3-5 3-6 3-7 540 7.8 6.9 6.3 5.7 5.4 4.9 5.2 550 7.8 6.6 6.3 5.7 5.6 4.9 5.1 560 7.8 6.7 6.3 5.7 5.6 4.9 5.1 570 7.9 6.8 6.5 5.7 5.7 4.9 5.1 580 7.8 7.1 6.3 5.8 5.7 4.9 5.0 590 7.9 7.0 6.4 5.8 5.7 4.9 5.1 600 7.9 7.1 6.4 5.8 5.7 4.9 5.1 610 7.9 7.1 6.5 6.0 5.8 5.3 5.2 620 8.0 6.8 6.5 6.1 5.8 5.4 5.3 630 7.9 6.8 6.5 6.2 5.9 5.5 5.4 640 8.0 7.0 6.7 6.2 6.0 5.6 5.5 650 8.0 6.7 6.8 6.3 6.0 5.6 5.6 660 8.1 7.1 6.8 6.4 6.1 5.7 5.7 670 8.1 7.3 6.9 6.5 6.2 5.7 5.8 680 8.1 7.0 6.8 6.5 6.2 5.9 5.8 690 8.2 6.9 6.9 6.5 6.3 5.9 5.8 700 8.2 7.2 6.9 6.5 6.3 5.7 5.8 710 8.2 7.5 6.9 6.6 6.3 5.9 5.8 720 8.1 7.2 7.0 6.6 6.3 5.9 5.9 730 8.2 7.3 7.0 6.6 6.4 6.0 5.9 740 8.2 7.0 7.0 6.7 6.4 6.1 6.0 750 8.2 7.3 7.1 6.7 6.4 6.1 6.0 760 8.2 7.0 7.1 6.7 6.5 6.1 6.1 390  A3.3.5 Solid Phase Analysis The results from the analysis of the TIC and TC of unreacted and reacted individual samples of starting material and samples from both pipe and pad experiments is presented in Tables A3.11 through A3.14. Coordinates for samples taken from the pad experiments are presented in Figures A3.2 and A3.4. Table A3.11. Unreacted TIC analyses on CPK-1 and FPK-1. CPK-1 TIC (wt.% CO2) CPK-1 Sieved TIC (wt.% CO2) FPK-1 TIC (wt.% CO2) 0.631 0.780 0.749 0.527 0.502 0.693 0.586 0.640 0.715 0.563 0.663 0.786 0.558 0.697 0.740 0.430 0.598 0.749 0.404 0.590 0.709 0.531  0.727 0.485  0.631 0.588  0.637 391  Table A3.12. Unreacted TIC and TC analyses on CPK-2 FPK-2, and mixtures of the two. TIC (wt.% CO2) TC (wt.% CO2) CPK-2 CPK-2 Sieved FPK-2 25 wt.% FPK Mix 25 wt.% FPK Mix Sieved 33 wt.% FPK Mix 25 wt.% FPK Mix 0.631 0.780 0.749 - 0.645 0.486 0.81 0.527 0.502 0.693 0.596 0.722 0.575 0.70 0.586 0.640 0.715 0.638 0.678 0.650 0.73 0.563 0.663 0.786 - 0.685 0.601 0.70 0.558 0.697 0.740 0.598 0.663 0.586 0.73 0.430 0.598 0.749 0.582 0.668 0.558 0.62 0.404 0.590 0.709 - 0.540 0.636 0.73 0.531  0.727 0.683 0.673 0.549 0.70 0.485  0.631 0.670 0.685 0.593 0.73 0.588  0.637 - 0.703 0.552 0.73    0.536 0.627 0.541 0.59    0.672 0.620 0.569 0.70 392  Table A3.13. Reacted TIC and TC analyses from Pipe-1 and Pipe-2, at the sampled horizon and distance. Pipe-1 Pipe-2 Distance (cm) Bulk TIC (wt.% CO2) Sieved TIC (wt.% CO2) Distance (cm) Bulk TIC (wt.% CO2) Sieved TIC (wt.% CO2) Bulk TC (wt.% CO2) 60T 0.883 1.255 0T 0.678 0.838 0.77 60M 0.775 1.284 0M 0.674 0.963 0.73 60B 1.010 1.281 0B 0.736 1.169  110T 0.853 1.298 50T 0.644 0.997  110M 0.886 1.294 50M 0.631 1.303 0.70 110B 0.992 1.275 50B 0.650 1.014  170T 0.851 1.264 100T 0.584 0.982 0.62 170M 1.213 1.278 100M 0.819 0.944  170B 0.952 1.213 100B 0.617 0.975 0.70 210T 0.933 1.245 150T 0.680 1.025 0.70 210M 0.896 1.210 150M 0.635 0.986  210B 0.872 1.253 150B 0.816 0.996  260T 0.927 1.235 200T 0.652 0.960  260M 0.905 1.248 200M 0.652 1.008 0.77 260B 0.839 1.224 200B 0.669 0.958 0.73 310T 0.799 1.219 250T 0.682 0.968  310M 0.840 1.219 250M 0.651 1.033  310B 0.870 1.178 250B 0.699 0.992 0.73 360T 0.797 1.202 300T 0.631 0.998 0.66 360M 0.971 1.262 300M 0.671 1.014 0.77 360B 0.928 1.244 300B 0.708 0.945  410T 0.796 1.241 350T 0.656 0.982  410M 0.936 1.214 350M 0.680 1.001 0.77 410B 1.052 1.233 350B 0.669 0.984  393  Table A3.13 continued. Pipe-1 Pipe-2 Distance  (cm) Bulk TIC  (wt.% CO2) Sieved TIC (wt.% CO2) Distance  (cm) Bulk TIC  (wt.% CO2) Sieved TIC (wt.% CO2) Bulk TC (wt.% CO2) 460T 0.967 1.254 400T 0.709 0.971 0.81 460M 0.886 1.183 400M 0.678 0.981  460B 0.793 1.220 400B 0.730 1.015 0.81 510T 0.778 1.242 450T 0.636 0.953 0.70 510M 0.885 1.191 450M 0.615 0.984  510B 0.881 1.308 450B 0.631 0.986  560T 0.926 1.183 500T 0.671 0.998  560M 0.893 1.216 500M 0.670 0.994 0.73 560B 0.862 1.169 500B 0.683 0.939 0.73    550T 0.799 0.921     550M 0.628 0.962     550B 0.674 0.945 0.73    580T 0.638 0.972 0.70    580M 0.732 0.922 0.84    580B 0.756 0.986  T, M and B signify top, middle and bottom horizons, respectively.  394  Table A3.14. Reacted TIC analyses from Pad-1 and Pad-2, at the sampled location. Pad-1 Pad-2 Layer Length (cm) Width (cm) Bulk TIC (wt.% CO2) Layer Length (cm) Width (cm) Bulk TIC (wt.% CO2) Cap FPK 13 15 0.751 Cap CPK 15 57.5 0.583 13 45 0.795 47.5 57.5 0.629 13 75 0.824 80 57.5 0.522 13 105 0.747 - - 0.440 48 15 0.938 Cap FPK 15 20 0.732 48 45 0.799 15 57.5 0.690 48 75 0.923 15 90 0.782 48 105 0.869 47.5 20 0.729 83 15 0.857 47.5 57.5 0.683 83 45 0.865 47.5 90 0.751 83 75 0.806 80 20 0.779 83 105 0.861 80 57.5 0.707 48 15 0.917 80 90 0.752 48 102 1.028 Mixed PK Top 15 20 0.681 68 60 1.092 15 57.5 0.722 73 25 0.874 15 90 0.778 86 84 0.987 32.5 57.5 0.733 23 25 0.919 47.5 20 0.760 10 84 1.078 47.5 57.5 0.767 - - 1.014 47.5 90 0.778 Mixed PK 13 15 1.210 62.5 57.5 0.743 13 45 0.924 80 20 0.748 13 75 0.837 80 57.5 0.673 13 105 0.920 80 90 0.767  395  Table A3.14 continued. Pad-1 Pad-2 Layer Length (cm) Width (cm) Bulk TIC (wt.% CO2) Layer Length (cm) Width (cm) Bulk TIC (wt.% CO2) Mixed PK 48 15 0.945 Mixed PK Middle 15 20 0.722 48 45 0.942 15 57.5 0.778 48 75 0.802 15 90 0.733 48 105 0.889 32.5 57.5 0.760 83 15 0.941 47.5 20 0.767 83 45 0.926 47.5 57.5 0.778 83 75 0.961 47.5 90 0.743 83 105 0.912 62.5 57.5 0.748 48 15 1.068 80 20 0.673 48 102 0.968 80 57.5 0.767 68 60 0.928 80 90 0.739 73 25 0.754 Mixed PK Bottom 15 20 0.778 28 60 0.859 15 57.5 0.733 23 25 0.825 15 90 0.760 10 84 1.014 32.5 57.5 0.767 Base FPK 13 15 0.868 47.5 20 0.778 13 45 0.861 47.5 57.5 0.743 13 75 0.810 47.5 90 0.748 13 105 0.949 62.5 57.5 0.673 48 15 0.971 80 20 0.767 48 45 0.894 80 57.5 0.739 48 75 0.938 80 90 0.753 48 105 0.927 Base CPK 15 57.5 0.662 83 15 0.806 47.5 57.5 0.560 83 45 0.828 80 57.5 0.637 396  Table A3.14 continued. Pad-1 Layer Length (cm) Width (cm) Bulk TIC (wt.% CO2)  83 75 0.832 83 105 0.857 - - 0.883   A3.3.6 Reaction Efficiency The sequestered mass of CO2 was compared against two metrics, the percentage of MgO reacted and the percentage of reactivity from the dominant reactive minerals. A3.3.6.1 MgO Reactivity The total MgO content was determined from the whole-rock chemistry. The mass of sequestered CO2 was related to the MgO content by assuming the formation of hydromagnesite, which consists of MgO and CO2 at a 1.25 to 1 ratio. The calculation process used to make this comparison is shown in Equations A3.28 \u2013 A3.31. Table A3.15 shows the results of the calculation process for the comparison against the MgO content. The total MgO content (MgO!\"#$%) is found from the fraction of each sample type in the whole sample $&\t(\"$)*+&\t,$-.%+% and $ &\t\/01+&\t,$-.%+%, and from the whole-rock chemistry analysis of each sample type (MgO(\"$)*+ and MgO\/01+). Equation A3.28. !\"#!\"#$%(%&.%) = !\"#&\"$'()(%&.%) \u00b7 *\t&\"$'()*\t,$-.%) (%) +!\"#\/01)(%&.%) \u00b7 *\t\/01)*\t,$-.%) (%) The MgO total was then converted to molar percent. 397  Equation A3.29. !\"#!\"#$%(-.\/.%) = !\"#!\"#$% 0 *\t2*3*\t,$-.%)1 \u00b7 456.865 0 -\"%*\t2*31 The TIC increase was converted to an equivalent MgO content based on the formation of hydromagnesite (TIC213)+$*+\tHmg\tMgO). This was done by converting the TIC increase to moles and then multiplying by the stoichiometric ratio of MgO to CO2, which is 1.25. Equation A3.30. 23491:')$()\t6-\"\t!\"#(-.\/.%) = 23491:')$() 0 *&3!*,$-.%)1 \u00b7 455.66; 0 -\"%*\t&3!1 \u00b7 7. 890-\"%\t<-*\t2*3-\"%\t<-*\t&3!1 The TIC increase MgO was then compared to the total MgO content. Equation A3.31. =)$:>)?\t<-*\t2*3!\"#$%\t2*3 \t(%) = !9&\"#$%&'(&\t<-*\t2*3(-\"%.%)2*3)*+',(-\"%.%)   398  Table A3.15. Determination of the degree of MgO accessed to sequester carbon in each of the metre-scale injection experiments. Experiment and Method CPK MgO (wt.%) a FPK MgO (wt.%) a Total MgO (wt.%) Total MgO (mol.%) Sequestered Carbon  (wt.% CO2) Sequestered Carbon as Hmg MgO (mol.%) b Leached Hmg MgO \/ Total MgO (%) c Pipe-1 Bulk Sample 22.6 22.9 22.7 d 0.56 0.10 \u00b1 0.05e 0.0028 0.5 Pipe-1 Sieved Sample c 0.11 \u00b1 0.01e 0.0031 0.6 Pipe-1 Mass Balance 0.14 \u2013 0.16 0.0040 \u2013 0.0045 0.7 \u2013 0.8 Pipe-2 Bulk Sample 24.0 20.3 23.1 d 0.57 0.09 \u00b1 0.02e 0.0026 0.4 Pipe-2 Sieved Sample c 0.12 \u00b1 0.01e 0.0034 0.6 Pipe-2 Mass Balance 0.23 0.0065 1.1 Pad-1 Mass Balance 22.6 22.9 22.7 d 0.56 0.09 0.0026 0.5 Pad-1 Mixed PK TIC 0.13 \u00b1 0.06 e 0.0037 0.7 Pad-2 Mass Balance 24.0 20.3 22.8 d 0.57 0.20 0.0057 1.0 Pad-2 Mean TIC 0.10 \u00b1 0.02 e 0.0028 0.5 a Value from whole-rock chemistry, determined by ICP-AES. b Hydromagnesite (Hmg) has an MgO to CO2 ratio of 1.25 to 1. c Measured on the sample fraction below 425 \u00b5m and attributed to the whole sample. d Coarse to fine mixtures done at a mass ratio of 78:22 (Pipe-1), 75:25 (Pipe-2), 76:24 (Pad-1), and 67:33 (Pad-2). e 95% confidence interval.  399  A3.3.6.2 Mineral Reactivity The abundance of lizardite was determined by qXRD. Knowing the stoichiometry enabled the determination of the Mg abundance, and the calculation process relating the reactivity to this abundance is shown in Equations A3.32 \u2013 A3.34. Table A3.16 shows the results of the calculation process for the comparison against the lizardite reactivity. The total Mg content from lizardite (Total\tMg!\"#$%&\"'() was found by multiplying the lizardite abundance in the whole sample by the percentage of Mg in lizardite based on its stoichiometry. Equation A3.32. !\"#$%\t'(!\"#$%&\"'( = *+,$-.+#\/(1#.%) \u00b7 )\t+,-.\/,-\t+\"0(1)\t2$345( (%) \u00b7 6. 789: ; )\t6))\t!\"#$%&\"'(< The TIC increase was converted to an equivalent Mg content based on the formation of hydromagnesite (TIC)*+%($,(\tHmg\tMg). This was done by converting the TIC increase to moles and then multiplying by the stoichiometric ratio of MgO to CO2, which is 1.25. Equation A3.33. !=>708%($1(\t?@(\t'((@\"%.%) = !=>708%($1( ; )\/9!)2$345(< \u00b7 :;;.==> ; 3?5)\t\/9!< \u00b7 :. 7A ;3?5\t@3)\t6)3?5\t@3)\t\/9!< The TIC increase Mg was then compared to the total lizardite Mg content available. Equation A3.34. !($8A(&\t@3)\t6)B?'$5\t6) \t(%) = B7\/\"#$%&'(&\t@3)\t6)(3?5.%)B?'$5\t6))*+'%,*-&(3?5.%)   400  Table A3.16. Determination of the degree of Mg accessed from lizardite to sequester carbon in the metre-scale injection experiments. Experiment Mg Source Portion (wt.%) a Abundance (wt.%) b Total Mg (mol.%) Sequestered Carbon (wt.% CO2) Sequestered Carbon as Hmg Mg (mol.%) b Leached Hmg Mg \/ Total Mg (%) b Pipe-1 Bulk Sample <425\u00b5m Lizardite 43 34.9 0.16 0.10 \u00b1 0.05e 0.0028 1.8 Pipe-1 Sieved Sample d 0.11 \u00b1 0.01e 0.0031 1.9 Pipe-1 Mass Balance 0.14 \u2013 0.16 0.0040 \u2013 0.0045 2.5 \u2013 2.8 Pipe-2 Bulk Sample <425\u00b5m Lizardite 39 27.6 0.11 0.09 \u00b1 0.02e 0.0026 2.4 Pipe-2 Sieved Sample d 0.12 \u00b1 0.01e 0.0034 3.1 Pipe-2 Mass Balance 0.23 0.0065 5.9 Pad-1 Mass Balance <425\u00b5m Lizardite 42 34.6 0.16 0.09 0.0026 1.6 Pad-1 Mixed PK TIC 0.13 \u00b1 0.06e 0.0037 2.3 Pad-2 Mass Balance <425\u00b5m Lizardite 45 27.4 0.13 0.20 0.0057 4.4 Pad-2 Mean TIC 0.10 \u00b1 0.02e 0.0028 2.2 a Abundance of FPK (if present) plus the abundance of CPK fines. Coarse to fine mixtures done at a mass ratio of 78:22 (Pipe-1), 75:25 (Pipe-2), 76:24 (Pad-1), and 67:33 (Pad-2). CPK-1 had 23 wt.% fines, while CPK-2 had 18 wt.% fines. b Lizardite abundance determined by qXRD. c Hydromagnesite (Hmg) has an MgO to CO2 ratio of 1.25 to 1. d Measured on the sample fraction below 425 \u00b5m and attributed to the whole sample. e 95% confidence interval.  401  A3.4 Discussion A3.4.1 Reaction Efficiency To examine the overall efficiency of the injection experiments, the total mass of injected CO2 (CO!\"#$%&'%() was calculated from the injected flow rate and the experimental duration as shown in Equation A3.35. This mass was then compared against the captured CO2 mass (CO!)*+',-%(), as determined from the gas phase mass balance, which is shown in Equations A3.36 and A3.37. The results of these calculations are shown in Table A3.17. Equation A3.35. !\"!\"#$%&'%(($) = '()*\t,-.\/ 0)*\t,-!).# 1 \u00b7 345-.6)7(85) \u00b7 9:0).#\/0 1 \u00b7 1\t*1222\t)* \u00b7 ;. =>? 03\t,-!*\t,-!1 Equation A3.36. !\"!,45'60%(($) = @-AA\tB-(-7C\/(*..%\t!\"!) \u00b7 E-FG(\/\t@-AA\t($) Equation A3.37. !\"!\t!-G.45\/\tHII6C6\/7CJ\t(%) = ,-!\"#$%&'()(3),-!*+,(-%()(3)  Table A3.17. Calculated mass of CO2 injected and experimental CO2 capture efficiency. Experiment Flow Rate (mL\u00b7min-1) Duration (hours) CO2 Injected (g) Mass Balance (wt.% CO2) Sample Mass (kg) CO2 Captured (g) CO2 Capture Efficiency (%) Pipe-1 720; 50 \u2013 120 140 459 \u2013 527 0.14 \u2013 0.16 247 346 \u2013 395 75 Pipe-2 550; 100 475 2496 0.23 220 506 20 Pad-1 80; 240 132 313 0.09 180 162 52 Pad-2 1000; 500; 250 765 4201 0.20 1050 2100 50  402  A3.4.2 Pad-2 Gas Flow Simulation The input parameters for the simple steady state model made using Abaqus CAE finite element analysis software are provided in Table A3.18. Typical values for Poisson\u2019s ratio and Young\u2019s Modulus were used (Budhu, 2011; Khoiri & Ou, 2013). Permeability and density values were from characterization results in this study. Void ratios were chosen as a simplified parameter for loosely compacted material. Table A3.18. Input parameters for the gas flow model. Parameter Base CPK Mixed PK Cap FPK Cap CPK Density (kg\u00b7m-3) 2700 2700 2700 2700 Young\u2019s Modulus (MPa) 70 70 70 70 Poisson\u2019s Ratio 0.3 0.3 0.3 0.3 Permeability (m\/s) 5.0E-4 2.0E-4 5.0E-7 5.0E-4 Void Ratio 1 1 1 1   403  A3.5 Implications A3.5.1 Sequestration Magnitude Injection rates measured in this study were compared to known or estimated passive sequestration rates and to emission rates at several diamond and nickel mines. To this end, the data used in these calculations and comparisons are presented in Table A3.19. Table A3.19. Data pertaining to tailings production, emission rates and passive and injection sequestration rates at several mines and deposits. Mine\/Deposit Tailings Production (Mt\/yr) Emissions (kt CO2\/yr) Passive CO2 Sequestration (kt CO2\/yr) Injection CO2 Sequestration (kt CO2\/yr) Coarse + Fine Diavik 2 150 a 0.3 (0.2%c) - Gahcho Ku\u00e9 3.2 124a | 45.9b - TIC: 3.2 \u2013 3.8 (5.2% \u2013 6.0%c | 7.0% \u2013 8.4%d) MB: 6.4 \u2013 7.4 (2.6% \u2013 3.1%c | 14% \u2013 16.2%d) Mount Keith 11 370 a 40 (10.8%c) - Dumont 15 137 a 21 (15.3%c) - Baptiste 43.8 108 a 17 (15.7%c) 618 (572%) a Total CO2 emissions. b Power generation emissions from point sources make up approximately one-third of Gahcho Ku\u00e9\u2019s emissions. c Percent of total mine emissions. d Percent of mine power generation emissions. 404  Appendix 4: Pneumatic Permeameter Development A4.1 Introduction Permeability is an important physical characteristic that is involved in the transport of fluids through porous media. It is defined as the measurement of the ability of a medium to transmit a fluid. Permeability is an important consideration in the context of groundwater and contaminant transport, fossil fuels migration, oxygen deliverability in soils, and subsurface remediation methods such as in-situ air sparging. In the context of enhanced weathering or carbon mineralization, the ability to efficiently supply CO2 enriched gasses into ultramafic mine tailings is controlled by the medium\u2019s permeability. Under industrial settings, at high rates of gas flow, this becomes an important factor due to energy demands required to maintain the necessary pressure gradient. Additionally, regarding carbon mineralization reaction kinetics, insufficient CO2 supply can be a rate-limiting factor. Sufficient supply at economic pressure gradients thus necessitates understanding how the mine tailings conduct fluids. Given that finer-grained materials are inherently more reactive due to their larger surface area but also pose gas delivery challenges due to their lower permeabilities, a designed grain size distribution is therefore essential to meet reactivity and gas delivery requirements. To accomplish this, this report will seek to develop an experimental method capable of determining the pneumatic permeability of various mediums and how they vary under different conditions.405  A4.2 Literature Review A comprehensive literature review has been conducted to synthesize the necessary background information related to the permeability of porous media, as well as the methods through which it is measured.  A4.2.1 Darcy\u2019s Law In 1856, Henri Darcy established that the flow rate of a fluid through a porous medium was proportional to the pressure difference across the sample length (Nield & Bejan, 2006). He accomplished this through the use of water flowing through a sand column, but his equation was quickly applied to other fluids in other media. Darcy\u2019s law can be written as shown in Equation A4.1 (Nield & Bejan, 2006). Equation A4.1. K =\t\u2212 9:;* \u2206N Where \u2018Q\u2019 is the volumetric flow rate (cm3\u00b7s-1), \u2018k\u2019 is the permeability (D), \u2018A\u2019 is the cross-sectional area of the medium (cm2), \u2018\u00b5\u2019 is the fluid viscosity (cP), \u2018L\u2019 is the length of the porous medium (cm), and \u2018DP\u2019 is the differential pressure across the medium (atm). A4.2.1.1 Reynold\u2019s Number and Laminar Flow For Darcy\u2019s law to apply, the fluid flow must be laminar, which is defined as the fluid moving in independent parallel layers. The Reynold\u2019s number is used to classify fluid flow regimes and is defined as the ratio of the fluid inertial forces to the viscous forces (Fanchi, 2002). For porous media, it is defined as shown in Equation A4.2 (Budhu, 2011). Equation A4.2. ,\/ = <=(.\/;  406  Where \u2018Re\u2019 is the Reynolds number (dimensionless), \u2018r\u2019 is the fluid density (g\u00b7cm-3), \u2018n\u2019 is the specific discharge (cm\u00b7s-1), and \u2018\u00b5\u2019 is the fluid viscosity (g\u00b7cm-1\u00b7s-1). \u2018d30\u2019 (cm) is the effective grain size, for which 30% of the medium\u2019s mass passes through the corresponding sieve.  Calculating the Reynold\u2019s number allows for the flow regime to be determined. Laminar flow is classified as a Reynold\u2019s number of less than 1 (Fanchi, 2002). However, from experimental approaches, it has been found that laminar conditions still exist when the Reynold\u2019s number remains below 10, with a transitional zone of non-linear laminar flow existing before pure turbulent flow is first seen at Reynold\u2019s numbers as large as 2100 (Budhu, 2011; Holtz & Kovacs, 1981; Nield & Bejan, 2006).  A4.2.2 Pneumatic and Hydraulic Permeability Relationships In the literature, it is very common for permeability to be assessed in two cases, for groundwater flow and for hydrocarbon migration and trapping. The first considers water flow through a wide range of permeability values, while the second focuses on other fluids and gasses, but generally at very low permeabilities. In groundwater systems, the hydraulic conductivity, or the coefficient of permeability, is measured in units of length over time (Budhu, 2011). It takes into account the intrinsic permeability but also the fluid density and viscosity (Kirkham, 2005). Since hydraulic permeameter tests allow for the hydraulic conductivity to be determined, it is necessary to correlate this with the intrinsic permeability found in pneumatic tests. This can be accomplished through Equation A4.3 (Kirkham, 2005). Equation A4.3. O = 9\u00b7<\u00b73;  407  Where \u2018K\u2019 is the hydraulic conductivity (cm\u00b7s-1), \u2018k\u2019 is the intrinsic permeability (cm2), \u2018r\u2019 is the fluid density (g\u00b7cm-3), \u2018g\u2019 is the gravitational acceleration, and \u2018\u00b5\u2019 is the viscosity of the fluid (g\u00b7cm-1\u00b7s-1). For converting between water and air, this conversion factor is approximately 105. In extreme cases, a pneumatic permeability value will not be directly proportional to a hydraulic permeability value. While permeability is a material-dependent property that does not depend upon the fluid, at very low permeabilities (less than 10-18 m2), the Klinkenberg effect occurs, resulting in transmission rates being dependent upon the gas in use and the mean pressure across the sample (Tanikawa & Shimamoto, 2006; Tidwell, 2006). Liquid flow through a pore is represented with the maximum velocity at the centre of the pore and zero velocity at the walls due to frictional forces (Tidwell, 2006). Low pressures can negate this frictional effect in gasses, allowing them to be transmitted faster, producing permeability overestimates (Tidwell, 2006). By varying the mean pressure, the true intrinsic permeability can be determined from pneumatic measurements since the measured permeability will increase linearly with the increase in mean pressure (Tanikawa & Shimamoto, 2006; Tidwell, 2006). However, for higher pressures and higher permeabilities, the measured pneumatic permeability will trend towards the true value. The Klinkenberg effect will not be considered in this study as permeability values less than 10-18 m2 will not be experienced by the unconsolidated materials in use.  A4.2.3 Effect of Physical Properties on Permeability Since permeability is a material-dependent property, other physical characteristics are related to it. Many of these properties relate to both porosity and permeability; however, the relationship between porosity and permeability is complex. The following physical characteristics 408  and their relationship with permeability are listed (Budhu, 2011; Holtz & Kovacs, 1981; Onur, 2014): \u00a7 Grain size: larger grain sizes have correspondingly larger pore spaces, and therefore have larger permeabilities in comparison to finer-grained materials. \u00a7 Grain shape: angular particles interlock together better than spherical grains, thus increasing flow tortuosity and reducing pore spaces and permeabilities. \u00a7 Grain size distribution: poorly sorted materials allow for finer grain sizes to take up space within the pore space of the larger-sized particles, thus reducing permeability. \u00a7 Compaction: grains are able to be moved closer to one another under compaction forces, reducing porosity and permeability. The degree of compaction is equivalent to the density of the material or it\u2019s void ratio (ratio of the volume of the solids to the volume of the voids). \u00a7 Water saturation: water blocks off the movement of gas through finer pores. Further, water saturation also affects the degree of compaction, increasing or decreasing it depending on the degree of saturation. \u00a7 Entrapped gasses: residual gas saturation takes up pore space, reducing permeability. Grain size distribution is one of the most important factors affecting permeability and is commonly characterized by two indices, the coefficient of uniformity (Cu) and the coefficient of curvature (Cc). These coefficients are defined as follows: Equation A4.4. !6 = ?0\/?1\/ Equation A4.5. !& = ?.\/!?1\/\u00b7?0\/ 409  Where D10, D30, and D60 are the sieve sizes through which 10%, 30%, and 60% of the material\u2019s mass passes through. The value of the coefficients is related to the degree of sorting, with the coefficient of curvature having the advantage that it takes into account three different sizes, reducing the likelihood of misidentifying a gap graded distribution. Stemming from the relationship of permeability to these various physical properties, many attempts have been made to conceptualize this empirical relationship for homogenous and coarse-grained soils. These efforts have been the result of difficulties incurred during laboratory and field testing. Testing permeability in the lab setting is often not representative of the corresponding field site, while field pumping tests are expensive, time-consuming and generally limited solely to horizontal permeability measurements.  There are many examples of empirical relationships between physical properties and the permeability of a material. Hazen (1930) established the proportional relationship between the hydraulic conductivity and the effective grain size, which was set at D10, with a constant that could vary between 0.4 and 1.2 (Budhu, 2011; Holtz & Kovacs, 1981). This method allowed for fast determinations to be made; however, it was limited to being reliable for clean sands. The Kozeny-Carman (1956) equation is another example suitable for sand-sized mediums and incorporates the fluid viscosity and the medium porosity (Holtz & Kovacs, 1981). Taylor (1948) developed a relationship between hydraulic conductivity and a material\u2019s void ratio (Budhu, 2011). Many other examples exist; however, most are limited to coarser materials or well-graded soils. Silts and clays require different relationships, with correlations to the void ratio having been shown to be poor (Holtz & Kovacs, 1981). An attempt to establish this empirical relationship while making use of the entire grain size distribution and the material\u2019s density has shown some success and is based 410  on applying different weights to corresponding size fractions and evaluating the material\u2019s permeability at different degrees of its maximum dry density (Onur, 2014). A4.2.4 Permeability Measurement Techniques Permeability measurement techniques differ depending upon the fluid in use. A4.2.4.1 Hydraulic Tests There are two main methods through which the hydraulic conductivity of a material is determined, and they differ based upon the soils types they are applicable for, with the constant head test being for coarse materials and the falling head test being for fine-grained materials (Budhu, 2011). Further, the constant head test is conducted under steady-state conditions, while the falling head test is not. A4.2.4.1.1 Constant Head Permeameter A constant head permeameter determines the hydraulic conductivity of a medium by allowing water to flow through the material under constant pressure (Budhu, 2011). The volume (V, L) of the water flowing out of the medium is then collected over a set amount of time (\u2206t, s). The change in head (\u2206h, cm) observed within a monometer is observed from before to during water flow. The hydraulic conductivity can then be determined through Equation A4.6 (Budhu, 2011). Equation A4.6. O = @*\u2206':\u2206\/ A4.2.4.1.2 Falling Head Permeameter During a falling head test, a compacted sample has water flowed through it. As water flows through the sample, the water pressure decreases (Budhu, 2011). The rate of head loss (h1 and h2, cm) over time (\u2206t, s) is recorded. Based on knowing that the water inflow (through a pipe with cross-sectional area, a, cm2) and outflow are equal, it is possible to integrate over the range of two 411  head measurements. The hydraulic conductivity can then be solved by using Equation A4.7 (Budhu, 2011). Equation A4.7. O = 4*:(\u2206') (7 0\/1\/!1 A4.2.4.2 Pneumatic Tests Similarly, there are two main methods through which the pneumatic permeability may be determined. Again, this is related to the steady and unsteady state nature of the two tests as well as how permeable the material is. A4.2.4.2.1 Steady State In the steady-state method, the pressure on the inflow and outflow end of a cylindrical sample is measured (ASTM International, 2013). The rate at which the gas is introduced is kept constant. Then the permeability can be calculated using Equation A4.8 (ASTM International, 2013). Equation A4.8. KB#%2 = :9;* \u00b7 B1!CB!!!\u00b7B!  Where \u2018P1\u2019 is the inlet pressure (atm) and \u2018P2\u2019 is the outlet pressure (atm). This equation is very similar to Equation A4.1; however, Equation A4.8 accounts for the changes in the flow rate of the gas due to its compressibility. A4.2.4.2.2 Pulse Decay Pulse decay methods are effective for determining the permeability of materials when the permeability is less than 10-16 m2 (Jones, 1997). This method involves reservoirs being placed at both ends of the sample and them being equally pressurized. One of the reservoirs is then pressurized further and allowed to equalize through the sample with the other reservoir. The rate at which this equalization occurs is related to the permeability of the sample (Jones, 1997). The 412  pulse decay method will not be required for this study since permeability values are not expected to be less than 10-16 m2.413  A4.3 Methodology To be able to measure the pneumatic permeability, a custom-made permeameter was developed, and a constant head permeameter was used to validate the pneumatic results.  A4.3.1 Pneumatic Permeameter A polycarbonate column, manufactured by W.A. Hammond Drierite, was used as the vessel for the apparatus. Two Bosch Sensortec BMP280 barometric pressure and temperature sensors were mounted at either end within the column, with wires protruding through, and epoxy and silicone administered to the base of the wires to seal the column. The BMP280 pressure sensors were wired to a Raspberry Pi 3B+, which was used to read the sensors, compensate the temperature measurements, and output the readings. The set-up is shown in Figure A4.1.  Figure A4.1. Pneumatic permeameter set-up. 414  A4.3.2 Constant Head Permeameter A hydraulic permeameter was used to validate the pneumatic permeability results. The set-up for the constant head permeameter is shown below in Figure A4.2 and is comprised of a column with a manometer and an elevated flask to supply water at a constant pressure.  Figure A4.2. Constant head permeameter diagram (University of British Columbia, 2018).  A4.3.3 Procedure Four types of materials were tested to measure their permeabilities: (1) coarse mine tailings from Gahcho Ku\u00e9 Diamond Mine (GK), (2) medium-grained, well-sorted quartz sand (Lane Mountain Materials), (3) fine-grained GK mine tailings, and (4) another sample of fine-grained mine tailings from the Baptiste Nickel Project. All four materials were tested for their pneumatic Note that the piezometer tube is placed in the wall of the permeameter immediately above the black porous stone so that test results are not affected by the hydraulic conductivity of the stone. The elevation of the water level in the piezometer will probably be slightly lower than that in the constant head tank due to head loss that occurs as water flows through the tubing and across the porous stone.    2. Mount the constant head tank so that the hydraulic gradient is slightly less than one (i.e. \u0394h is approximately 0.8 of L).  3. Collect the discharge in the graduated cylinder while timing the flow with the stop watch. Once you are confident that the flow is constant, make two different measurements. Record the volume of water, elapsed time, \u0394h, and L in the table at the end of the lab. Calculate Q from these results.  4. Mount the constant-head tank so that the hydraulic gradient is about half the value used in step 3 and record the \u0394h and L and calculate Q in this case.  5. Move the tank so that the gradient is greater than 1. Record any disturbance you may see in the sediments which might indicate turbulent (i.e. non-laminar) flow (e.g. piping).      Figure 1: The constant head apparatus used in this lab.  Water flows out of the tap to the constant head reservoir and then into the permeameter.   sink graduated cylinder tap stand standpipe ruler L h ' ponded  water air vent constant head control Sample 415  permeability, while only the coarse tailings and the quartz sand were tested using the constant head permeameter. A4.3.3.1 Pneumatic Test For the pneumatic tests, the material was introduced into the polycarbonate column in three states, loose, compacted, and partially (~30%) saturated. The loose state was achieved without any compaction save for the force exerted by a spring on the top of the column to keep the material immobile during the test. The compacted state was attained through successive blows administered to small (~2 cm) layers of material until the column was filled. The partially saturated state required a volume of water that would take up approximately 30% of the pore space to be well mixed with the material prior to insertion into the column. The water volume was calculated using the density of the material and the compacted volume and mass of the material within the column. The partially saturated material was then compacted in a similar fashion to the dry, compacted state. These three states were used to determine the variability of the permeability due to compaction and partial saturation. Partial saturation is required for carbon mineralization to occur as the produced minerals are hydrated magnesium carbonates. Upon implementation into the column, the column was then laid on its side to ensure no difference in elevation head during the test. The length of the sample was recorded, and the cross-sectional area was determined. The BMP280 sensors were then wired to the Raspberry Pi, and the inlet end of the column was connected to the flow meter, in line with the compressed air cylinder. The flow meters used were a Matheson Gas Products E100 REVB flow meter for the low flow rates and a WEONE LZM-6 flow meter for the high flow rates. The sensors were started, and the data was recorded for at least 100 seconds prior to gas flow. These pre-readings were used to establish the current differential pressure between the two sensors. These differences were then 416  averaged, and this was used to correct the data. Subsequently, the gas cylinder was opened, and gas was permitted to flow into the flow meter. The gas coming from the tank was maintained at 20 PSI (138 kPa). The flow meter was then set to a pre-determined flow rate. The flow rate was maintained, and the gas flow was allowed to run for a sufficient duration. From the output pressure data, a graph was produced of the differential pressure against time. Segments where the differential pressure was constant were then used to determine an average value for the differential pressure at that flow rate. Permeability values were then calculated using Equation A4.1. An alternative method would have been to use Equation A4.8. Under the small differential pressures used in these tests (<100 hPa) and the short length of the sample (<20 cm), these equations yield nearly identical results. A4.3.3.1.1 Reynold\u2019s Number Verification To ensure that the gas flow was laminar during the test, the Reynold\u2019s number was calculated for the highest flow rates (8.3, 5.7, and 2.6 LPM) and for the coarsest material (the coarse tailings) since this set-up had the highest potential to produce turbulent flow. From the particle size distribution of the coarse tailings, the d30 grain size was found to be 0.13 cm. The calculated Reynold\u2019s numbers for the three flow rates are displayed below in Table A4.1.417  Table A4.1. Reynold's number calculations for the coarse tailings at high rates of gas flow. Air Density (g\u00b7cm-3) (20\u00b0C) Volumetric Flow Rate (LPM) Specific Discharge (cm\u00b7s-1) d30 (cm)   Air Dynamic Viscosity (20\u00b0C) (g\u00b7cm-1\u00b7s-1)  Reynold\u2019s Number  0.0012041 8.31 5.540 0.13 0.0001813 4.77 0.0012041 5.69 3.795 0.13 0.0001813 3.27 0.0012041 2.60 1.730 0.13 0.0001813 1.49  As can be seen, the Reynold\u2019s numbers do exceed the laminar flow zone under the specification that it may not exceed 1. However, the values lie in the transitional zone under which a Reynold\u2019s number of 1 to 10 can still be considered laminar flow. Note also that the value for the 2.6 LPM flow rate is only barely over the boundary of 1. Further evidence that laminar flow is present under these conditions is that for the various flow rates, the calculated permeabilities are consistent. A4.3.3.2 Hydraulic Test To compare the results obtained using the developed pneumatic permeameter, constant head tests were conducted on the coarse tailings and the quartz sand. The materials were implemented into the permeameter column in two states, loose and compacted as accomplished for the pneumatic case. This was done to obtain similar degrees of compaction between the pneumatic and hydraulic tests. Following implementation into the column, the length of the sample was measured along with the diameter. Then the material was fully saturated. Under a constant pressure head, the flow of water through the sample was collected over a measured time. Equation A4.4 was then used to calculate the hydraulic conductivity, and Equation A4.3 to convert this to the intrinsic permeability.418  A4.4 Results The summarized results obtained from the investigation of the permeability of the four materials are displayed below in Tables A4.2 \u2013 A4.5. The calculations for each trial are included in the supplementary material. The weighted mean permeabilities were calculated using the inverse of the calculated permeability errors as the weightings for each respective test. Table A4.2. GK coarse-grained tailings permeability results. Apparatus and Condition Weighted Mean Permeability (m2) Pneumatic, Loose 2.2E-9 Pneumatic, Compacted 1.1E-9 Pneumatic, 30% Saturated and Compacted 5.3E-10 Hydraulic, Loose 7.3E-10 Hydraulic, Compacted 3.4E-10  Table A4.3. Medium-grained quartz sand permeability results. Apparatus and Condition Weighted Mean Permeability (m2) Pneumatic, Loose 4.0E-10 Pneumatic, Compacted 1.9E-10 Pneumatic, 30% Saturated and Compacted 9.9E-11 Hydraulic, Loose 1.8E-10 Hydraulic, Compacted 1.4E-10 419  Table A4.4. GK fine-grained tailings permeability results. Apparatus and Condition Weighted Mean Permeability (m2) Pneumatic, Loose 1.2E-12 Pneumatic, Compacted 1.2E-13 Pneumatic, 30% Saturated and Compacted 8.8E-14  Table A4.5. FPX fine-grained tailings permeability results. Apparatus and Condition Weighted Mean Permeability (m2) Pneumatic, Loose 1.2E-12 1) Pneumatic, Compacted 2.9E-14 1) Pneumatic, 30% Saturated and Compacted 5.3E-14 2) Pneumatic, Compacted 4.5E-14 2) Pneumatic, 30% Saturated and Compacted 4.2E-14  From the values in Tables A4.2 and A4.3, the determined permeability from the equivalent hydraulic and pneumatic tests have been compared below in Figure A4.3. 420   Figure A4.3. Hydraulic versus pneumatic permeability comparison for the loose and compacted GK tailings and medium-coarse sand.421  A4.5 Discussion From the tabulated permeability values in Tables A4.2 and A4.3, a comparison of the measurements obtained through pneumatic and hydraulic methods can be made. For both the coarse tailings and the quartz sand, the hydraulic method estimated lower permeabilities by approximately a factor of three for the coarse tailings and two for the quartz sand. The only exception was for the compacted quartz sand, where the pneumatic and hydraulic values were similar. Figure A4.3 shows this overall trend. The consistency between the tests is demonstrated by the linear relationship of the data. In terms of permeabilities, which vary by orders of magnitude, a difference of a factor of three is relatively close. It is also important to account for differences between the tests that could explain this discrepancy. Of note is the role of trapped residual gas saturation within the pore space of the medium during the hydraulic test. It is nearly impossible to saturate a medium entirely, and the trapped gas would act to decrease the permeability, with this being a known issue of using a constant head permeameter (Budhu, 2011; Holtz & Kovacs, 1981). Whether this could account for the difference between the two tests is uncertain. Slight differences in compaction were also possible but were not expected to result in significant differences. Migration of fines during the tests could also be a factor impacting the validity of both results (Holtz & Kovacs, 1981). From examining the effect of compaction, it is clear that compaction decreases the permeability in fine-grained materials far more significantly than in coarser mediums. In the coarse tailings and in the sand, compaction decreased permeability at most by a factor of two. In the two fine-grained tailings, the permeability was decreased by an order of magnitude in the GK tailings and nearly two orders in the FPX tailings.  422  Finally, water saturation at approximately 30% decreased the permeability in the coarse tailings, sand and GK fine tailings by approximately a factor of two in comparison to the dryly compacted case. In the FPX tailings, the first trial saw the permeability actually increase upon being partially saturated. Understandably this is an unexpected result. Upon a second attempt, the permeability slightly decreased when partially saturated, though nowhere near as significantly as achieved for the GK fine-grained tailings. The most likely explanation could be that permeabilities in the 10-14 m2 range are the lower limit of what this pneumatic permeameter is capable of accurately measuring.423  A4.6 Conclusion In conclusion, a pneumatic permeameter was developed and was applied to measure the permeability of four distinct materials. Two of these were compared to results obtained using a constant head permeameter, and the results were comparable given slight differences that could arise between the different tests. It is speculated that trapped residual gas may be responsible for the hydraulic tests producing slightly lower permeability values than for the pneumatic case. The impact of the condition of the materials was also examined, namely the effect of compaction and water saturation on the permeability. Compaction affected the fine-grained materials significantly more than the coarser materials and collectively decreased the permeability by factors from 2 to nearly 100. Partial water saturation was confirmed to decrease the permeability of all materials, though the effect was in some cases less than significant. Overall, the pneumatic permeameter performed well, accurately and consistently determining the permeability of an array of materials. An obvious next step for this study is to experiment with various grain size distributions of the coarse and fine-grained materials and examine the effect this has on the permeability of the mixture.424  A4.7 Supplementary Material A4.7.1 Pneumatic Permeability Calculations The permeability calculations, using Equation A4.1, for each of the materials have been tabulated below in Tables A4.6 \u2013 A4.9. 425  Table A4.6. GK coarse-grained tailings pneumatic permeability calculations. Trial # Condition   Gas Flow Rate (L\u00b7min-1) Sample Length (cm) Differential Pressure (hPa) Permeability (m2)  Permeability Error (%) 1 Loose 8.31 18.6 0.92 2.0E-9 \u00b1 19 2 Loose 8.31 18.6 0.85 2.2E-9 \u00b1 21 3 Loose 8.31 18.6 0.86 2.2E-9 \u00b1 20 4 Loose 8.31 18.6 0.90 2.1E-9 \u00b1 19 5 Loose 8.31 18.6 0.93 2.0E-9 \u00b1 19 6 Loose 8.31 18.0 0.83 2.2E-9 \u00b1 21 6 Loose 5.69 18.0 0.40 3.1E-9 \u00b1 43 7 Loose 5.69 18.6 0.61 2.1E-9 \u00b1 28 8 Loose 5.69 18.6 0.63 2.0E-9 \u00b1 27 9 Loose 5.69 18.6 0.58 2.2E-9 \u00b1 30 10 Loose 2.6 18.6 0.25 2.3E-9 \u00b1 68 11 Compacted 8.31 16.8 1.68 1.0E-9 \u00b1 11 11 Compacted 5.69 16.8 1.00 1.2E-9 \u00b1 17 12 Compacted 8.31 16.5 1.61 1.0E-9 \u00b1 12 12 Compacted 5.69 16.5 0.98 1.2E-9 \u00b1 18 13 30% Saturated, Compacted 5.69 17.9 2.78 4.4E-10 \u00b1 7 14 30% Saturated, Compacted 8.31 17.9 4.47 4.0E-10 \u00b1 6 15 30% Saturated, Compacted 8.31 17.1 2.57 6.7E-10 \u00b1 8 15 30% Saturated, Compacted 5.69 17.1 1.57 7.5E-10 \u00b1 11  426  Table A4.7. Medium-grained quartz sand pneumatic permeability calculations. Trial # Condition   Gas Flow Rate (L\u00b7min-1) Sample Length (cm) Differential Pressure (hPa) Permeability (m2)  Permeability Error (%) 1 Loose 5.69 18.2 3.26 3.8E-10 \u00b1 6 2 Loose 8.31 18.2 4.75 3.9E-10 \u00b1 6 3 Loose 5.69 18.2 2.95 4.2E-10 \u00b1 7 4 Compacted 8.31 17.7 9.37 1.9E-10 \u00b1 5 4 Compacted 5.69 17.7 6.53 1.9E-10 \u00b1 4 5 30% Saturated, Compacted 5.69 19.0 13.24 9.8E-11 \u00b1 4 5 30% Saturated, Compacted 8.31 19.0 20.22 9.4E-11 \u00b1 5 5 30% Saturated, Compacted 2.6 19.0 5.62 1.1E-10 \u00b1 5 427  Table A4.8. GK fine-grained tailings pneumatic permeability calculations. Trial # Condition   Gas Flow Rate (L\u00b7min-1) Sample Length (cm) Differential Pressure (hPa) Permeability (m2)  Permeability Error (%) 1 Lightly Compacted 0.020 17.8 5.28 8.2E-13 \u00b1 6 2 Lightly Compacted 0.020 17.8 5.39 8.0E-13 \u00b1 6 2 Lightly Compacted 0.068 17.8 20.13 7.3E-13 \u00b1 3 3 Loose 0.020 17.4 3.3 1.3E-12 \u00b1 7 3 Loose 0.068 17.4 12.09 1.2E-12 \u00b1 3 4 Loose 0.020 17.8 3.81 1.1E-12 \u00b1 7 4 Loose 0.068 17.8 13.84 1.1E-12 \u00b1 3 5 Compacted 0.025 15.7 40.06 1.2E-13 \u00b1 4 6 Compacted 0.020 15.7 29.57 1.3E-13 \u00b1 5 6 Compacted 0.025 15.7 39.2 1.2E-13 \u00b1 4 7 30% Saturated, Compacted 0.020 13 35.38 9.0E-14 \u00b1 5 7 30% Saturated, Compacted 0.025 13 46.09 8.6E-14 \u00b1 4 428  Table A4.9. FPX fine-grained tailings pneumatic permeability calculations. Trial # Condition   Gas Flow Rate (L\u00b7min-1) Sample Length (cm) Differential Pressure (hPa) Permeability (m2)  Permeability Error (%) 1 Lightly Compacted 0.020 18.2 15.50 2.8E-13 \u00b1 5 2 Lightly Compacted 0.020 18.5 18.03 2.5E-13 \u00b1 5 3 Lightly Compacted 0.020 18.5 18.25 2.5E-13 \u00b1 5 4 Loose 0.020 17.8 3.55 1.2E-12 \u00b1 7 4 Loose 0.068 17.8 13.03 1.1E-12 \u00b1 3 5 Compacted 0.012 17.6 87.75 2.9E-14 \u00b1 9 6 Compacted 0.012 17.6 88.39 2.9E-14 \u00b1 9 7 Compacted 0.012 17.4 57.29 4.4E-14 \u00b1 9 7 Compacted 0.020 17.4 88.17 4.8E-14 \u00b1 5 8 Compacted 0.012 17.4 62.77 4.0E-14 \u00b1 9 8 Compacted 0.02 17.4 95.16 4.4E-14 \u00b1 5 9 30% Saturated, Compacted 0.02 16.2 71.17 5.5E-14 \u00b1 5 9 30% Saturated, Compacted 0.012 16.2 46.50 5.1E-14 \u00b1 9 9 30% Saturated, Compacted 0.025 16.2 94.50 5.2E-14 \u00b1 4 10 30% Saturated, Compacted 0.02 16.2 70.64 5.6E-14 \u00b1 5 10 30% Saturated, Compacted 0.012 16.2 46.31 5.1E-14 \u00b1 9  429  Table A4.9 continued. Trial # Condition   Gas Flow Rate (L\u00b7min-1) Sample Length (cm) Differential Pressure (hPa) Permeability (m2)  Permeability Error (%) 10 30% Saturated, Compacted 0.025 16.2 91.73 5.4E-14 \u00b1 4 11 30% Saturated, Compacted 0.012 15.6 58.42 3.9E-14 \u00b1 9 11 30% Saturated, Compacted 0.02 15.6 87.70 4.3E-14 \u00b1 5   A4.7.2 Hydraulic Permeability Calculations The permeability calculations, using Equations A4.3 and A4.4, for each of the materials have been tabulated below in Tables A4.10 and A4.11.430  Table A4.10. GK coarse-grained tailings hydraulic permeability calculations. Condition Head (cm)  Volume (mL)  Time (s)   Hydraulic Conductivity (cm\u00b7s-1) Permeability (m2)  Permeability Error (%)  Compacted 3.1 85 30.6 3.56E-01 3.6E-10 \u00b1 2 Compacted 3.35 86.5 30.8 3.33E-01 3.3E-10 \u00b1 2 Compacted 3.4 91.5 33.5 3.19E-01 3.2E-10 \u00b1 2 Compacted 3.85 91 26.2 3.59E-01 3.6E-10 \u00b1 2 Compacted 3.9 90 27.2 3.37E-01 3.4E-10 \u00b1 2 Compacted 3.9 91 27.9 3.33E-01 3.3E-10 \u00b1 2 Compacted 3.9 94 29.9 3.21E-01 3.2E-10 \u00b1 2 Loose 3.1 88 17.4 7.02E-01 7.0E-10 \u00b1 2 Loose 3.1 91 18.6 6.79E-01 6.8E-10 \u00b1 2 Loose 3.1 92 19 6.72E-01 6.7E-10 \u00b1 2 Loose 3.25 91 19.4 6.21E-01 6.2E-10 \u00b1 2 Loose 3.5 85 18.1 5.77E-01 5.8E-10 \u00b1 2 Loose 2.9 90 17.2 7.17E-01 7.2E-10 \u00b1 2 Loose 2 92 21 8.71E-01 8.7E-10 \u00b1 3 Loose 2 94 21.9 8.53E-01 8.5E-10 \u00b1 3 Loose 2.1 88 20.2 8.25E-01 8.3E-10 \u00b1 3 Loose 2.05 90 19.2 9.09E-01 9.1E-10 \u00b1 3 Loose 2.05 89.5 18.5 9.38E-01 9.34E-10 \u00b1 3 Loose 2 85 17.1 9.88E-01 9.9E-10 \u00b1 3 Loose 2.05 92.5 18.2 9.86E-01 9.9E-10 \u00b1 3 Loose 2.05 92 19 9.39E-01 9.4E-10 \u00b1 3 Loose 1.8 92 36.6 5.55E-01 5.6E-10 \u00b1 3 Loose 2.7 90 19.6 6.76E-01 6.8E-10 \u00b1 2  431  Table A4.10 continued. Condition Head (cm)  Volume (mL)  Time (s)   Hydraulic Conductivity (cm\u00b7s-1) Permeability (m2)  Permeability Error (%)  Loose 3.5 92 14.3 7.31E-01 7.3E-10 \u00b1 2 Loose 3.5 92.5 14.6 7.20E-01 7.2E-10 \u00b1 2 Loose 3.6 91 14.5 6.93E-01 6.9E-10 \u00b1 2 Loose 2.6 92 22.7 6.20E-01 6.2E-10 \u00b1 2 Loose 2.05 94.5 32.2 5.69E-01 5.7E-10 \u00b1 3 432  Table A4.11. Medium-grained quartz sand hydraulic permeability calculations. Condition Head (cm)  Volume (mL)  Time (s)   Hydraulic Conductivity (cm\u00b7s-1) Permeability (m2)  Permeability Error (%)  Compacted 5.7 19.5 9 1.50E-01 1.5E-10 \u00b1 3 Compacted 2.5 23.5 32 1.37E-01 1.4E-10 \u00b1 2 Compacted 2.3 23.5 36.2 1.32E-01 1.3E-10 \u00b1 3 Compacted 4.8 23.5 15.4 1.48E-01 1.5E-10 \u00b1 2 Compacted 4.9 23.5 16.4 1.36E-01 1.4E-10 \u00b1 2 Compacted 8.8 21.5 9.4 1.21E-01 1.2E-10 \u00b1 3 Compacted 8.9 22.5 10 1.18E-01 1.2E-10 \u00b1 2 Compacted 9.5 23.5 10.2 1.13E-01 1.1E-10 \u00b1 2 Compacted 9.4 84 28.4 1.64E-01 1.6E-10 \u00b1 1 Compacted 10.2 87.5 29.2 1.53E-01 1.5E-10 \u00b1 1 Compacted 10.75 92.5 30.8 1.46E-01 1.5E-10 \u00b1 1 Compacted 10.2 89 32.5 1.40E-01 1.4E-10 \u00b1 1 Loose 9.15 96 27.9 2.05E-01 2.1E-10 \u00b1 1 Loose 9.05 95.5 29 1.98E-01 2.0E-10 \u00b1 1 Loose 9.1 93.5 29.6 1.89E-01 1.9E-10 \u00b1 1 Loose 8.85 96 32.3 1.83E-01 1.8E-10 \u00b1 1 Loose 8.75 96 33.4 1.79E-01 1.8E-10 \u00b1 1 Loose 8.65 96.5 34.7 1.75E-01 1.8E-10 \u00b1 1 Loose 5.4 52 30.3 1.73E-01 1.7E-10 \u00b1 1 Loose 5.1 60 37.4 1.71E-01 1.7E-10 \u00b1 1 Loose 5 55 35.1 1.71E-01 1.7E-10 \u00b1 1  ","@language":"en"}],"Genre":[{"@value":"Thesis\/Dissertation","@language":"en"}],"GraduationDate":[{"@value":"2021-11","@language":"en"}],"IsShownAt":[{"@value":"10.14288\/1.0402341","@language":"en"}],"Language":[{"@value":"eng","@language":"en"}],"Program":[{"@value":"Geological Sciences","@language":"en"}],"Provider":[{"@value":"Vancouver : University of British Columbia Library","@language":"en"}],"Publisher":[{"@value":"University of British Columbia","@language":"en"}],"Rights":[{"@value":"Attribution-NonCommercial-NoDerivatives 4.0 International","@language":"*"}],"RightsURI":[{"@value":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/","@language":"*"}],"ScholarlyLevel":[{"@value":"Graduate","@language":"en"}],"Supervisor":[{"@value":"Dipple, Gregory","@language":"en"}],"Title":[{"@value":"Carbon mineralization in ultramafic mine tailings via CO\u2082 injection","@language":"en"}],"Type":[{"@value":"Text","@language":"en"}],"URI":[{"@value":"http:\/\/hdl.handle.net\/2429\/79796","@language":"en"}],"SortDate":[{"@value":"2021-12-31 AD","@language":"en"}],"@id":"doi:10.14288\/1.0402341"}