UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Testing protocol for evaluating amenability of ores to HPGR crushing for heap leaching Moghadam Zadeh, Sanaz 2017

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

Item Metadata

Download

Media
24-ubc_2017_november_moghadamzadeh_sanaz.pdf [ 9.76MB ]
Metadata
JSON: 24-1.0357256.json
JSON-LD: 24-1.0357256-ld.json
RDF/XML (Pretty): 24-1.0357256-rdf.xml
RDF/JSON: 24-1.0357256-rdf.json
Turtle: 24-1.0357256-turtle.txt
N-Triples: 24-1.0357256-rdf-ntriples.txt
Original Record: 24-1.0357256-source.json
Full Text
24-1.0357256-fulltext.txt
Citation
24-1.0357256.ris

Full Text

  TESTING PROTOCOL FOR EVALUATING AMENABILITY OF ORES TO HPGR CRUSHING FOR HEAP LEACHING   by  SANAZ MOGHADAM-ZADEH  M.Sc. Science and Research University, Tehran, Iran, 2004 B.Sc. Iran University of Science and Technology, Tehran, Iran, 2002  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate and Postdoctoral Studies (Mining Engineering)  The University of British Columbia (Vancouver) October 2017  ©Sanaz Moghadam-Zadeh, 2017 ii   Abstract  Finer particle size distribution (PSD) and generation of more microcracks obtained by the use of high pressure grinding rolls (HPGR) are believed to be the major reasons for metal recovery improvement by heap leaching. Finer PSD increases the surface area of the particles, which are in contact with the leach solution and improves recovery. Microcracks in HPGR materials result in improved penetration of leach solution into the grains and thus metal recovery. The objective of this work is to develop a better understanding of ores crushed by HPGR and cone crusher, in terms of microcracks and PSD, and assess the benefits of using HPGR for heap leach projects.  The test results can be used in the early stages of the project to develop an initial understanding of the benefits of utilizing HPGR in the comminution circuits for heap leaching.  Gold oxide, copper sulfide and copper/gold sulfide samples were used in this research. Laboratory tests including PSD analysis, scanning electron microscope (SEM), slake durability, Helium (He) pycnometry, nitrogen gas adsorption, water absorption and column tests were used to evaluate the changes in the PSD, microcracks morphology, mechanical stability, porosity, exposed surface area, water absorption capacity and unsaturated flow behavior of the samples. Finally, modeling by HYDRUS was conducted to investigate the hydraulic behavior of the crushed ores.  PSD analysis showed finer size distribution for HPGR materials, independent of the ore types. Lower slake durability indices, higher porosities, larger surface areas, higher water absorption values and residual water degree were obtained for HPGR’s materials in comparison with the cone crusher’s. Computer modeling showed lower unsaturated hydraulic conductivity for HPGR samples. Estimation of the microcracks percentage over the total porosity indicated higher values for HPGR samples. Percentage of water inside microcracks and on particle surfaces over the water remaining between particles was iii   higher in most HPGR crushed samples. Finally, a time efficient testing protocol for evaluating the amenability of HPGR to provide ore having the preferred characterization for heap leaching is designed. iv   Lay summary  Due to the increase in energy consumption in recent years, the mining industry needs to replace conventional comminution techniques with energy-efficient methods that can produce improved metal recoveries. One of such energy-efficient methods is use of high-pressure grinding rolls (HPGR) in crushing circuits. Crushers that produce finer sizes and generates more microcracks also reduce the energy consumption in the overall metal recovery operations.   The HPGR’s biggest advantages in mineral processing are its high energy efficiency, finer products and development of microcracking. In addition, because of the recent scarcity of high grade ores and availability of lower grade ores, operations such as heap leaching that extract metals efficiently from low grade ores have been attracting attention and are being further developed.   This study’s main focus is to develop a test protocol to evaluate the advantages of HPGR application for heap leaching. This test protocol includes easy, time efficient and inexpensive laboratory tests for the early stages of industrial project studies.   v   Preface I was responsible for designing the test work, preparing samples and conducting all tests under the supervision of Professor van Zyl. I also analyzed data, conducted computer modeling and developed methods for the design of tests to evaluate samples under the supervision of Dr. van Zyl.  The ores used in this study were crushed and prepared with HPGR by Zorigtkhuu Davaanyam, Stefan Nadolski and Amit Kumar. I prepared the ores crushed by cone crusher. For the Cu ore, I performed HPGR particle size analysis by myself. Zorig provided me with the PSDs data for Au and Cu/Au ores. I prepared samples and conducted tests including slake durability, He pycnometry, gas adsorption, water absorption, and column drainage tests. The SEM images were taken in laboratory outside UBC. Mr. Aaron Hope built the column equipment that Dr. van Zyl and I designed for leach tests. Sally Finora taught me how to use gas adsorption equipment and helped me to understand and evaluate data.   During this study some results and updated data were published in the following paper and posters:   Sanaz Moghadam-Zadeh and Dirk van Zyl (2014), Characterization of ore crushed by HPGR and cone crushers for heap leaching, Proceedings of the Heap Leach Solutions 2014, Lima, Peru (Chapter 2, 3 and 4);  Sanaz Moghadam-Zadeh and Dirk van Zyl (2014), Proof of concept: evaluating microcracks in heap leach ore crushed by high pressure grinding rolls (HPGR) and cone crusher, 143rd SME annual meeting, Salt Lake City, US (poster, Chapter 3 and 4);  Sanaz Moghadam-Zadeh and Dirk van Zyl (2016), PSD and microcracks of HPGR and cone crushed ores for heap leach operations, CIM annual meeting, Vancouver, Canada (poster, Chapter 3 and 4). vi   Also, I presented the following conference paper:  Sanaz Moghadam-Zadeh and Dirk van Zyl (2013), Development and application of simple tests to investigate the feasibility of HPGR, Heap Leach Solutions 2013, Vancouver, Canada (Chapter 2 and 3).            vii   Table of contents  Abstract ........................................................................................................................................... ii  Lay summary ................................................................................................................................. iv  Preface ............................................................................................................................................. v  Table of contents .......................................................................................................................... vii  List of tables .................................................................................................................................. xi  List of figures .............................................................................................................................. xiv  Acknowledgments...................................................................................................................... xviii  Dedications .................................................................................................................................. xix  Chapter 1: Introduction ................................................................................................................. 1  1.1 Background ..................................................................................................................................................... 1 1.2 Objectives ....................................................................................................................................................... 3 1.3 Thesis Outline ................................................................................................................................................. 4 Chapter 2: Background and literature review .............................................................................. 6  2.1 Heap leaching ................................................................................................................................................. 6 2.1.1 Leaching mechanisms ....................................................................................................................... 10 2.1.2 Leaching solution transport within particles...................................................................................... 13 2.1.3 Leaching solution transport through the heap ................................................................................... 15 2.2 HPGR in heap leach circuits ......................................................................................................................... 18 2.2.1    Breakage mechanism .......................................................................................................................... 19 viii   2.2.2 Mineral accessibility and microcracks............................................................................................... 20 Chapter 3: Materials and procedures .......................................................................................... 30  3.1 Materials ....................................................................................................................................................... 30 3.1.1 Sample mineralogy ............................................................................................................................ 31 Au ore ......................................................................................................................................................... 31 Cu ore ......................................................................................................................................................... 31 Cu/Au ore ................................................................................................................................................... 32 3.1.2 Sample preparation ............................................................................................................................ 32 3.1.3 Lab cone crusher ............................................................................................................................... 33 3.1.4 Lab HPGR ......................................................................................................................................... 34 3.1.5 Particle size distribution .................................................................................................................... 37 3.2 Test procedures ............................................................................................................................................. 37 3.2.1 Scanning electron microscopy ........................................................................................................... 39 3.2.2 Slake durability.................................................................................................................................. 40 3.2.3 He pycnometry .................................................................................................................................. 42 3.2.4 Nitrogen adsorption test .................................................................................................................... 43 3.2.5 Water absorption test ......................................................................................................................... 46 3.2.6 Soil/water characteristic curve analysis ............................................................................................. 47 Column test ................................................................................................................................................ 47 SWCC data modeling ................................................................................................................................. 53 HYDRUS Program .................................................................................................................................... 56 Chapter 4: Results ........................................................................................................................ 57  4.1 PSD of ores ................................................................................................................................................... 57 4.2 Scanning electron microscopy (SEM) .......................................................................................................... 70 4.3 Slake durability ............................................................................................................................................. 74 4.4 He pycnometry .............................................................................................................................................. 78 ix   4.5 Nitrogen adsorption test ................................................................................................................................ 80 4.6 Water absorption ........................................................................................................................................... 86 4.7 Column test ................................................................................................................................................... 88 4.7.1 Analyzing saturated hydraulic conductivity with Kozeny model .................................................... 102 4.7.2 Analyzing hydraulic conductivity with Brooks and Corey model ................................................... 103 4.7.3 Analyzing hydraulic conductivity with HYDRUS program ............................................................ 104 Chapter 5: Discussions and analysis ......................................................................................... 107  5.1 HPGR finer PSDs effect on slake durability indices ................................................................................... 107 5.2 Microcracks generation from HPGR .......................................................................................................... 108 5.2.1 Estimation of microcracks percentage to the total porosity for whole size ores .............................. 110 5.3 Analysis of water inside the microcracks .................................................................................................... 113 5.4 Effect of microcracks on unsaturated hydraulic conductivity ..................................................................... 119 5.5 Suggested testing protocol .......................................................................................................................... 120 5.6 Amenability of samples to be crushed by HPGR for heap leaching ........................................................... 122 5.7 Limitation of the research ........................................................................................................................... 125 Chapter 6: Conclusions, contributions and future work .......................................................... 128  6.1 Contributions .............................................................................................................................................. 132 6.2 Recommendations for future work.............................................................................................................. 134 Bibliography ............................................................................................................................... 136  Appendices .................................................................................................................................. 143  Appendix A: PSD of Cu sample ....................................................................................................................... 143 Appendix B: Slake durability for Cu sample .................................................................................................... 144 Appendix C: Repeatability of slake durability tests .......................................................................................... 145 Appendix D: He Pycnometry for Cu sample .................................................................................................... 146 Appendix E: Repeatability of He pycnometry tests .......................................................................................... 148 x   Appendix F: BET for Cu/Au HPGR 6-under 150 µm sample .......................................................................... 149 Appendix G: Repeatability of gas adsorption test ............................................................................................. 153 Appendix H: Column test and HYDRUS results for Cu sample ...................................................................... 154 Appendix I: Information for mines using HPGR for Heap leaching ................................................................. 168  xi   List of tables Table 2-1 Types of gold ore for heap leaching (Mular et al., 2002) ............................................................. 9 Table 2-2 Five different classes of grain accessibility (Ghorbani et al., 2011) ........................................... 12 Table 2-3 Studies on HPGR crushed materials with column tests .............................................................. 27  Table 3-1 Crushers specifications ............................................................................................................... 35  Table 3-2 HPGR pressures .......................................................................................................................... 35 Table 3-3 Sizes of samples used for the lab tests ........................................................................................ 38 Table 3-4 Measured parameters of the column test .................................................................................... 51  Table 4-1 PSD for the ores used in this study ............................................................................................. 60 Table 4-2 PSD for different ore types crushed with cone crusher and HPGR of this study ....................... 63 Table 4-3 PSD for different ore types processed with cone crusher and HPGR of other studies ............... 68 Table 4-4 Slake durability indices (Id(2) (%)) and related sample weights (W (g)) ..................................... 75 Table 4-5 Porosity (%) of the samples from the He pycnometry test ......................................................... 78 Table 4-6 BET specific surface area (m2/g) for particle assemblage and single particle ............................ 81 Table 4-7 Average microcracks volume and microcracks width from BET test for assemblage ............... 84 Table 4-8 Water absorption results ............................................................................................................. 86 Table 4-9 Moisture content of ores ............................................................................................................. 89 Table 4-10 Initial values in column tests for samples ................................................................................. 90 Table 4-11 Initial results from column tests ............................................................................................... 91  Table 4-12 Saturated water content from SWCC curve for each ore.......................................................... 97 Table 4-13 Residual water content from drainage versus time curves for samples .................................. 102 Table 4-14 Ks estimation from Kozeny model.......................................................................................... 103 xii   Table 4-15 α and n parameters for Brooks and Coreys model .................................................................. 104 Table 4-16 Hydraulic conductivity for ores from HYDRUS inverse modeling ....................................... 104  Table 5-1 Microcracks percentage over total porosity in whole size sample for Cu cone........................ 111 Table 5-2 Microcracks percentage over total porosity in whole size samples .......................................... 112 Table 5-3 Percentage of water in microcracks and on particle surface over the total remaining water .... 115 Table 5-4 Parameters selected as creteria to indicate the ore (as an assemblage) amenability to be crushed by HPGR for heap leaching ...................................................................................................................... 124   Table A-1 Copper sulfide (Cu ore) size distribution data ......................................................................... 143  Table B-1 Slake durability measurement method for Cu sample for assemblage .................................... 144  Table C-1 Results of repeated slake durability tests for Cu samples ........................................................ 145  Table D-1 Pycnometry measurements of Cu samples .............................................................................. 146  Table E-1 Results of He pycnometry tests for repeated tests on three samples ........................................ 148  Table F-1 Multipoint BET for Cu/Au HPGR 6 N/mm2 (under 150 µm) .................................................. 150  Table G-1 Results of gas adsorption tests for repeated tests on five samples ........................................... 153  Table H-1 Initial measurements of column test for Cu cone .................................................................... 154 Table H-2 Column test parameters to plot SWCC for Cu cone (1) .......................................................... 155 xiii   Table H-3 Column test parameters to plot SWCC for Cu cone (2) .......................................................... 156 Table H-4 Column test parameters to plot SWCC for Cu cone (3) .......................................................... 157 Table H-5 Column test parameters to plot SWCC for Cu cone (4) .......................................................... 158 Table H-6 Column test parameters to plot SWCC for Cu cone (5) .......................................................... 159 Table H-7 Column test parameters to plot SWCC for Cu cone (6) .......................................................... 160  Table I-1 List of mines with HPGR in heap leach operations .................................................................. 168     xiv   List of figures  Figure 2-1 Map of mines using heap leach operation ................................................................................... 7 Figure 2-2 HPGR in heap leaching operation (Dhawan et al., 2013) ......................................................... 19 Figure 2-3 Two leach regimes for HPGR and cone crusher (Ghorbani et al., 2013) .................................. 21 Figure 2-4 Separation of particle to shells with different distance from particle core for HPGR products (Ghorbani et al., 2013) ................................................................................................................................ 26   Figure 3-1 Sample preparation flow diagram ............................................................................................. 33 Figure 3-2 Cone crusher laboratory scale ................................................................................................... 34 Figure 3-3 (a) A photograph of HPGR laboratory scale and (b) its schematic diagram (Infomine) ........... 36 Figure 3-4 Slake durability apparatus ......................................................................................................... 41 Figure 3-5 BET plot of Cu/Au sample (under 150 μm size) crushed by cone crusher ............................... 43 Figure 3-6 A photograph of Autosorb IQ Automated Gas Sorption Analyzer ........................................... 45 Figure 3-7 (a) Schematic diagram and (b) a photograph of the column used for SWCC ........................... 48 Figure 3-8 SWCC variables (Zhai & Rahardjo, 2012) ............................................................................... 53   Figure 4-1 PSD for Au (gold oxide) (a), Cu (copper sulphide) (b) and Cu/Au (copper gold sulphide) (c) ores .............................................................................................................................................................. 58  Figure 4-2 Feed and cone PSDs for Au, Cu and Cu/Au ores ...................................................................... 61  Figure 4-3 HPGR PSDs for Au, Cu and Cu/Au ores .................................................................................. 62 Figure 4-4 F(80), F(40) and F(10) of HPGR versus cone crusher (this study) .................................................. 64 Figure 4-5 Cumulative passing sizes of cone and HPGR for two ore categories with different pressure ranges (this study) ....................................................................................................................................... 65 Figure 4-6 PSD of HPGR and cone crusher for different ores (other studies) ............................................ 66 Figure 4-7 F(80), F(40) and F(10) of HPGR versus cone crusher (other studies) ............................................. 69 xv   Figure 4-8 Cumulative passing sizes of cone and HPGR for five ores of other studies and six ores of this study ............................................................................................................................................................ 69  Figure 4-9 Cu feed SEM in 50 and 10 µm scale ......................................................................................... 71  Figure 4-10 Cu cone SEM in 50 and 10 µm scale ...................................................................................... 71 Figure 4-11 Cu HPGR SEM in 50 and 10 µm scale ................................................................................... 71 Figure 4-12 Cu/Au feed SEM in 50 and 10 µm scale ................................................................................. 72 Figure 4-13 Cu/Au cone SEM in 50 and 10 µm scale ................................................................................ 72 Figure 4-14 Cu/Au HPGR 3.5 N/mm2 SEM in 50 and 10 µm scale ........................................................... 72 Figure 4-15 Microcracks in feed, cone and HPGR of Cu samples (50 µm scale) ...................................... 73 Figure 4-16 Microcracks in feed, cone and HPGR of Cu/Au samples (50 µm scale) ................................ 74 Figure 4-17 HPGR to cone slake durability index ratios for Au, Cu and Cu/Au ores ................................ 77 Figure 4-18 HPGR to cone porosity ratios for Au, Cu and Cu/Au ores ..................................................... 79 Figure 4-19 HPGR to cone specific surface area ratios for Au, Cu and Cu/Au ores as a function of particle size .............................................................................................................................................................. 83  Figure 4-20 HPGR to cone microcracks volume ratios for Cu, Au and Cu/Au ores .................................. 85 Figure 4-21 HPGR to cone water absorption ratios for Au, Cu and Cu/Au ores ........................................ 87 Figure 4-22 Phase diagram of Cu cone (Vt-total volume, Va-air volume, Vw-water volume, Vs-solid volume, Mw-water mass, Ms-solid mass, Mt-total mass) ............................................................................ 89 Figure 4-23 HPGR/cone ratios for Sr (%), porosity before adding water (%) and initial added water volume (cm3) in column test ....................................................................................................................... 92 Figure 4-24 SWCC plots for Cu (a), Au (b) and Cu/Au (c) ........................................................................ 94 Figure 4-25 Volumetric water content data for Cu/Au ore at different HPGR pressures ........................... 95 Figure 4-26 Cu cone drainage plot .............................................................................................................. 98  Figure 4-27 Cu HPGR drainage plot........................................................................................................... 98  Figure 4-28 Au cone drainage plot ............................................................................................................. 99  xvi   Figure 4-29 Au HPGR drainage plot .......................................................................................................... 99  Figure 4-30 Cu/Au cone drainage plot ...................................................................................................... 100  Figure 4-31 Cu/Au HPGR 2.5 N/mm2 drainage plot ................................................................................ 100 Figure 4-32 Cu/Au HPGR 3.5 N/mm2 drainage plot ................................................................................ 101 Figure 4-33 Krw from inverse modeling for Cu cone and HPGR .............................................................. 105 Figure 4-34 Krw from inverse modeling for Au cone and HPGR ............................................................. 105 Figure 4-35 Krw from inverse modeling for Cu/Au cone, Cu/Au HPGR 2.5 N/mm2 and Cu/Au HPGR 3.5 N/mm2 ....................................................................................................................................................... 106   Figure 5-1 Microcracks percentage over total porosity in whole size samples ........................................ 112 Figure 5-2 Schematic diagram to show the water in microcracks and on the surface calculation ............ 114 Figure 5-3 Percentage of water remaining after free drainage over the initial added water ..................... 116 Figure 5-4 Percentage of water in microcracks and on surface over water in pores ................................. 117 Figure 5-5 Water content  in microcracks and on surface of particles in terms of the mass of dry ores (%) .................................................................................................................................................................. 118  Figure 5-6 Hydraulic conductivity from inverse modeling with HYDRUS ............................................. 119 Figure 5-7 Suggested test protocol for determination of ore amenability to be crushed by HPGR for heap leaching in scoping evaluation .................................................................................................................. 122   Figure F-1 Isotherm plot for Cu/Au HPGR 6 N/mm2 (under 150 µm) ..................................................... 152  Figure H-1 HYDRUS main page (Cu cone) ............................................................................................. 161 Figure H-2 HYDRUS main process (Cu cone) ......................................................................................... 161 Figure H-3 HYDRUS inverse modeling page (Cu cone).......................................................................... 162 Figure H-4 HYDRUS inverse modeling geometry information (Cu cone) .............................................. 162 xvii   Figure H-5 HYDRUS inverse modeling time information (Cu cone) ...................................................... 163 Figure H-6 HYDRUS inverse modeling iteration criteria (Cu cone) ....................................................... 163 Figure H-7 HYDRUS inverse modeling soil hydraulic model (Cu cone) ................................................ 164 Figure H-8 HYDRUS inverse modeling water flow parameters (Cu cone) ............................................. 164 Figure H-9 HYDRUS inverse modeling water flow boundary conditions (Cu cone) .............................. 165 Figure H-10 HYDRUS inverse modeling data (Cu cone) ........................................................................ 165 Figure H-11 HYDRUS inverse modeling profile information (Cu cone) ................................................. 166 Figure H-12 HYDRUS inverse modeling pressure head versus water content (Cu cone) ....................... 166 Figure H-13 HYDRUS inverse modeling pressure head versus hydraulic conductivity (Cu cone) ......... 167 Figure H-14 HYDRUS inverse modeling pressure head versus effective water content (Cu cone) ......... 167    xviii   Acknowledgments  During the past seven years, I was very fortunate to have been kindly supported by many people who helped me completed this experience successfully.   First of all, true kindness, support and precious advice from my supervisor, Professor Dirk van Zyl, were the main reason I could complete my research. I would like to express my deepest appreciation for all the precious comments he provided me during these years. I would like to appreciate guidance and valuable feedback from my supervisory committee, Dr. Bern Klein and Dr. Davide Elmo.  I appreciate the financial support the Mining Department provided me during my research. During the lab experiments, Sally Finora, Pius Lo and Aaron Hope greatly supported me on technical matters. I am so thankful of Zorigtkhuu Davaanyam who provided me the samples and gave me lots of useful advice through the laboratory tests. I also would like to thank Amit Kumar for helping me with laboratory tests during my pregnancy. I really appreciate kind advice from my friend Mehrnoush Javadi who helped me with HYDRUS program.   I was blessed to have my daughter during my PhD program. She brought me joy and happiness through my journey. Having her in my life was a life-changing experience for me. I really appreciate kind advice and support from Nima during the past seven years. I have learned how to be successful from my two brothers.  Without my parents support, I would not have a chance to complete this meaningful step of my life. Since I was a child, their encouragement was the reason I could follow my dreams.   xix   Dedications   To my mother and father, who I learned from true meaning of life. Thank you with love and respect.  To my daughter, Lillian. Thank you to give me the chance to be a mom. I love you my dear.     1   Chapter 1: Introduction This section contains the background of the research as well as the research question and objectives. Also, the outline of the research is explained to clarify the steps to answer the main question and reach the objectives.   1.1 Background Heap leaching is a cost-effective hydrometallurgical process in which relatively large ore particles, usually crushed in one to three stages, are used. In heap leaching, the access of the leach solution to the mineral grains inside the particles directly affects the extent of metal recovery. Finer particle size distribution (PSD) and the presence of microcracks in the particles improve the recovery of the heap leaching due to the increased exposure of the mineral grains to the leach solution (Dunne et al., 1996). Feeds with finer PSD have an increased surface area exposed to the solution and those containing microcracks benefit from facilitated access of the leach solution to the interior of the particles (Ghorbani et al., 2013). It is therefore expected that the performance of heap leaching can be improved by application of a crushing method that produces finer PSD ores and generates more microcracks.   HPGR is an energy-efficient comminution method in mineral processing that has found application in the preparation of ores for heap leaching. HPGR can be used as a secondary or tertiary crusher to replace the cone crusher commonly used in heap leaching operations. It has been shown that the ores crushed by HPGR have a finer PSD and contain more microcracks, in comparison with those crushed by a cone crusher (Daniel, 2007; Kodali et al., 2011; von Michaelis, 2005).  2   Since it is not clear whether the use of HPGR would be advantageous for all types of ores in terms of recovery improvement (due to finer PSD and/or microcracks generation), the potential benefits of HPGR in heap leaching must be evaluated in the early stages of a heap leach project.  PSD is a critical factor affecting recovery in heap leaching. In several studies, the main benefit of HPGR has been described as its ability to produce finer PSD materials. Nevertheless, direct comparative PSD data for HPGR and cone crusher are scarce and the available PSD data for the two techniques are scattered in the literature. Therefore, there is a need to gather and process such data for a better assessment of the benefits of HPGR for heap leaching.  The other benefit of HPGR application, in terms of microcracks generation, has been investigated by several researchers. Various studies used x-ray microtomograpghy (XMT) and other imaging techniques to evaluate microcracks quantitatively and column leach tests to assess the effect of microcracks on recovery (Ghorbani et al., 2013; Kodali et al., 2011). The use of XMT (X-ray micro CT) to study microcracks is associated with some drawbacks including its limited resolution (5 μm for 10 mm sample and 20 μm for 40 mm sample sizes) , particularly for relatively large particles, and ability to process only single particles in high resolution that are not necessarily representative of the whole ore. Also, measuring microcracks quantitatively by imaging methods does not indicate the ability of microcracks to absorb the leach solution inside the particles. Reliable column leach testing is a lengthy process and several months may be needed to obtain results.  This makes the preliminary investigation a time-consuming task. Therefore, there is a need to evaluate or develop testing protocol that could characterize the microcracks and their capacity for solution penetration prior to the column leach test in a more rapid and cost effective manner during the earlier stages of the project.  3   1.2 Objectives This research is proposed to address the lack of a consistent and accepted yet simple and rapid approach to evaluate potentially attainable benefits of using HPGR in a heap leach circuit. The following research question is evaluated:   Can a simple test protocol be used or developed to evaluate the suitability of using HPGR instead of conventional comminution technique (e.g. cone crusher) in a heap leach circuit for a specific ore during project scoping evaluation?  The main objective of this research project is:   Developing of simple test protocol and defining parameters that reliably indicate during scoping studies whether the application of HPGR in heap leach circuits, as a comminution technique, offers advantages over cone crushers.  To achieve the main objective, the following sub-objectives were used to guide the research:   1. Identify the major differences between materials with different mineralogies crushed by the two different techniques in terms of both PSD and microcracks;  2. Establish a method to estimate the quantity of microcracks;  3. Investigate the effects of microcracks on the relative percent of moisture that can enter microcracks; 4. Analyze the effect of microcracks and water penetration/absorption on the unsaturated hydraulic behavior of heap leach ore.  4   Various literature data on PSD for different type of materials were collected and processed to show the differences between PSDs obtained with cone crusher and HPGR. The mechanical stability, porosity, specific surface area, water absorption capacities, residual water content, degree of saturation and unsaturated hydraulic conductivity were chosen as potential experimental parameters to characterize microcracks. SEM, slake durability, He pycnometry, gas adsorption, water absorption and column tests were adopted as experimental procedures to characterize microcracks caused by HPGR and cone crushers. The combination of the data is used to develop or select a simple and expedient laboratory test protocol for initial assessment of the potential benefits of replacing a cone crusher with HPGR in heap leaching operations. This can still be followed by column leach tests.  1.3 Thesis Outline  The first Chapter includes background, objectives and outline of this research. The second Chapter of the dissertation is literature review. In Chapter 2 heap leaching, leaching mechanisms, fluid transport within particles and through the heap are explained. Also, application of HPGR and its breakage mechanism are discussed. The last section includes mineral liberation and microcracks of the HPGR ores.  In Chapter 3, materials used in the research and procedure of each test conducted are explained. Sample mineralogy, sample preparation methods and test procedures are described in this Chapter. The lab tests used in this research are: cone and HPGR crushing, PSD analysis, SEM, slake durability, He pycnometry, nitrogen adsorption, water absorption, and column tests.   In Chapter 4, all results from laboratory testing are presented and analyzed. Further detailed analyses of the unsaturated flow characteristics are preformed using HYDRUS in the modeling. In Chapter 5, the relationship between slake durability results and PSD, a method to estimate the microcrack-areas percentage in crushed ores, the evaluation of the effects of microcracks on solution penetration, the 5   effects of microcracks on unsaturated hydraulic conductivities and limitation of the research are discussed. Chapter 6 presents conclusions, contributions and suggested future research.    6   Chapter 2: Background and literature review This Chapter reviews the literature related to the heap leaching process, leaching mechanisms, leaching solution transport within particles and through the heap. It also reviews HPGR application in ore preparation for heap leaching. The literature survey also discusses the HPGR comminution method in terms of breakage mechanism, mineral liberation and microcrack generation.   2.1 Heap leaching Heap leaching is a hydrometallurgical process for extraction of metals from low-grade ores. In this process, run-of-mine or coarser crushed ores (typically larger particles than those used for agitated leaching) are piled up on a lined pad. The width and length of the heap are typically much larger in comparison with its height. The leaching solution is uniformly applied from the top of the heaps using a variety of sprinkler or dripper techniques. Uniform unsaturated flow, which is the ideal situation, with reaction on particle surfaces and microcracks dissolves minerals. The pregnant leach solution is continuously collected at the bottom of the heap. This pregnant solution is then processed to extract the metal.   Heap leaching does not need costly grinding steps, required for hydrometallurgical mill processes. Other advantages of the heap leaching process are elimination of solid/liquid separation stage and tailings disposal circuits. It has been reported that heap leaching reduces the processing capital costs and overall environmental impacts (Dhawan et al., 2012).  Heap leaching is an efficient method for recovering the metal from large low-grade gold, silver and copper deposits. Note that this technology is also applied to uranium and laterite nickel ores, however the focus of this research is gold, silver and copper ores. The mineralogy of an ore influences the extent of 7   comminution needed, leaching behaviour, reagent consumption, residual mineralogy and equipment performance (Dhawan et al., 2012). Generally, gold and copper ores that are suitable for HPGR application for heap leaching have low clay content (below 20%), low moisture content (below 15%) and high hardness (Baum & Ausburn, 2011). There are several types of gold ores that can be successfully heap-leached. Table 2-1 provides information about these gold ores used for heap leaching (Mular et al., 2002).   A map of mines using heap leach operations can be find in InfoMine website (http://www.infomine.com/intelligence/maps/posters/heapleach/, 15 August 2017) and in Figure 2-1. The list of mines use or study the HPGR application for the heap leach operations is in Appendix I.  Figure 2-1 Map of mines using heap leach operation  8   Hard, siliceous and sulphide-bearing gold ores demonstrated metallurgical benefits in HPGR/heap leach circuits (Baum et al., 1996; Patzelt et al., 1995). These types of ore showed deep micro fracturing and enhanced cyanidation. Iron oxide-rich gold ore also demonstrated better leaching behaviour by application of HPGR (Baum et al., 1996).   9   Table 2-1 Types of gold ore for heap leaching (Mular et al., 2002) Type of ore Geology and mineralization Remarks Process Carlin-type sedimentary ores, invisible gold in pyrite and arsenopyrite Contains shales and limestone, silty carbonates or silicates Gold is very fine -Oxidized ore heap leach recovery is higher than 80% -Unoxidized ore obtains gold in sulphides and containing carbonaceous materials with 10 to 15 % heap leaching recovery -Agitated leaching and heap leaching Low sulphide acid volcanic or intrusive ores Sulphide contains 2 to 3% pyrite Gold is enclosed in pyrite -Oxidized ore leach recovery is 65 to 85% -Unoxidized ore recovery is 45 to 55% -Usually crushed below 12 mm Oxidized massive sulphides Iron oxide-rich ore Gold and silver in iron oxides -Crushed to 75 mm and larger Saprolites/ laterites,  (deeply weathered, ferric compounds) Host ore is volcanic and intrusive in tropical climates with weathering Gold in quartz veinlets -At surface is usually laterite (hard iron oxide nodules), below the laterite is saprolite  -Recovery of 85%  -Run of mine (ROM) or crushing Clay-rich deposits Some Carlin-type deposit and some volcanic-hosted deposit Gold deposition occurred with clay deposition and alteration -Mixture of soft wet clay and hard rock -Single stage impact crusher   Oxide copper ores are suitable for heap leach using sulfuric acid. The suitable copper oxide ores for heap leaching are carbonate minerals like azurite or malachite. Leaching is also beneficial if silicates or sulfates are dominant. This is because of their good solubility in sulfuric acid. If the ore contains plenty of ultra-soluble sulphates, water can be used instead of sulfuric acid for leaching (Watling, 2006). 10    Bio-heap leaching can be used for primary copper sulphides like chalcopyrite. In bacterial leaching, sulphides are oxidized to sulfuric acid, which leaches the copper. Low grade sulphide copper ores like chalcocite, down to the grade of 0.15% , have shown good recovery with bio heap leaching when crushed to minus 6 to 13 mm (Watling, 2006).   2.1.1 Leaching mechanisms The way target minerals are distributed within ore particles affects the leaching efficiency. The valuable grains may exist as exposed or encapsulated forms and with or without access to the leaching solution (Ghorbani et al., 2011).   Depending on the grain accessibility to the leach solution, particles can be categorized into five types. Table 2-2 summarizes these types. For particle groups shown in Table 2-2, there are different leaching mechanisms according to the grain size (Rossi, 1990):  1) The grain size is similar to the particle size (high grade ores):   Chemical leaching reactions occur on the surface  The reaction regime proceeds and causes a shrinkage of the particle  2) The grains are smaller than the particles, but they are accessible to the liquid:   Grains are mostly accessible through the microcracks   The reaction is still controlled by surface-chemical reaction  Since grains are surrounded by gangue minerals, diffusion also controls the process  11   3) The grains are smaller than the previous conditions:   All grains are not accessible from the beginning of the leaching, but still the process is controlled by surface/chemical reaction   Internal grains are exposed to the gangue and other mineral grains  The reaction rate is reduced to a greater extent in this condition  4) Grains are much smaller than particles:   The process is controlled by both chemical surface reaction and diffusion   Gangue minerals decrease the leaching rate as they create longer diffusion paths   12   Table 2-2 Five different classes of grain accessibility (Ghorbani et al., 2011) Condition Schematic diagram 1. Target grains are at the surface of the particle and they are exposed to the leach solution.  2. Grains are inside the particle and they are accessible through pores and cracks.  3. Grains are inside the particle and they are accessible after other grains are leached.  4. Grains are inside the particle, but they are not accessible through the cracks.  5. Grains are inside the particle and there are no cracks and pores.   Except for the type 1 and type 5 ores, leaching solution access to the grains within the ore particles is essential for the leaching process to proceed. Therefore, solution transport phenomenon within the particles significantly affects the leaching kinetics and recovery. Microcracks are critical pathways for direct access of leach solution to grains inside the larger ore particles.  The shrinking core model can explain the process inside a particle. In this model it is assumed that liquid penetrates inside the particles and reacts with specific grains that dissolve into the liquid phase. As a 13   result, the solid core shrinks and the particle size remains virtually unchanged (Tsakiroglou et al., 2017). Since the particles are piled in the heap, the leach solution transport to the particles and between the particles is also critical. Therefore, both leaching solution transport throughout the heap (between particles) and inside particles are the factors determining leaching efficiency in heap leach.  2.1.2 Leaching solution transport within particles Inside a cracked particle, diffusion and capillarity through microcracks are the transport mechanisms. Fluid absorption in porous materials can be studied using the capillary suction theory for cracked single particles (Bioubakhsh, 2011).    The magnitude of the capillary force depends on crack size, surface tension and wettability of the material. If a capillary tube is immersed vertically in a liquid, the liquid level in the tube rises until the upward force due to surface tension is equal to the downward force due to the liquid weight (Jurin’s Law) (Batchelor, 1967). This force balance can be written as:  2 π r γ cos ϑ =  π rଶ hୡ ρ g                                        (1) where r is the tube radius, γ is the surface tension, ϑ is the contact angle between the liquid and tube wall, ρ is the liquid density and g is the gravity acceleration.   Therefore, for a cylindrical capillary tube, the height of the capillary rise, hc, can be calculated as:  hୡ =2 γ cos ϑρ r g                                                                    (2)                       14   Laplace’s law describes the pressure difference (Δ P = ρ g hc in tube) that is due to the surface tension between the liquid and the tube wall and causes the capillary rise:    Δ P =2 γ cos ϑr =  ρ g hୡ                                               (3)  In the case of pure water and a glass tube, the radius of the tube can be related to the pressure according to Laplace’s law, assuming γ for water at room temperature is 7.2 × 10-2 N/m and cos ϑ=1:  Δ P = ଴.ଵସ଺୰                                                                 (4) According to Equation (4), smaller diameter tubes result in more pressure difference and more capillary rise.   Capillary suction in unsaturated conditions creates fluid transport between particles and through porous particles. Assuming that the geometry of the cracks in a porous particle is like parallel tubes, Darcy’s law can describe the average fluid velocity (Adamson, 1967). For a single tube the velocity in the tube is:  V =r γ cos ϑ4 l μ                                                                       (5)  where V is the average fluid velocity (m/s), r is the tube radius, γ is the surface tension, ϑ is the contact angle between the liquid and tube wall, l is the length of the tube (m), and μ is the viscosity of the fluid (Pa. s).   15   Assuming the only driving force for penetration is capillary action (ignoring the gravity force for horizontal tube, (V= dl/dt)), the length of penetration as a function of time is calculated by following integrated equation (Adamson, 1967):   l = ඨr γ cos ϑ  t2 μ                                                                  (6)  Equation 6 shows the liquid penetration depth has a direct relationship with the square root of crack radius, surface tension, contact angle and time, and a reverse relationship with the square root of viscosity. Larger crack opening, liquid surface tension and contact angle and lower viscosity result in deeper penetration. Therefore, for the specific leach solution and ore, the crushers that generate more microcracks with wider crack openings, facilitate the leach solution penetration inside the particles and improve metal recovery.  2.1.3 Leaching solution transport through the heap Capillary pressure and gravity are the forces that lead to unsaturated fluid flow between particles in porous media. In a porous medium, capillarity accounts for the fluid transport especially in unsaturated condition close to the surface (Martys & Ferraris, 1997). Unsaturated flow models are used to evaluate the fluid flow in the particle assemblage.  The extent of fluid absorption, resulting from capillary action, depends on the pore structure and fluid properties in the medium. Water undergoes capillary condensation at humidity levels well below bulk saturation in the confines of the pore (Adamson, 1967).  16   Richard’s equation (Richards, 1931), Equation 9, is obtained from Darcy’s equation and the law of conservation of mass, and can be used to model one-dimensional capillary rise for unsteady flow in unsaturated condition in porous media. Darcy’s velocity for unsaturated state is:  v = −Kdhdx                                                                                (7)                                      To model the unsteady-state flow, the law of conservation of mass in unsteady-state flow is used:  ∂ϑ∂t=  − ∂v∂x                                                                              (8)                where ϑ is volumetric water content. Combining Equations 7 and 8 results in Equation 9. Soil matric suction is described in the form of capillary forces. In an unsaturated porous medium, the fluid is subject to suction. Therefore, the fluid flow is a function of the matric suction.  ∂ϑ∂t=  ∂∂x ሾ K (Ψ)׏ hሿ                                                           (9)      where Ψ is the suction and ׏ h is the hydraulic head gradient that may contain both suction and gravitational forces. If the gravitational force is negligible, for one-dimensional flow the equation is:  ∂ϑ∂t=  ∂∂x ൤ K (Ψ) ∂Ψ∂x ൨                                                        (10)                The hydraulic conductivity and suction relationship demonstrates hysteresis (different values for absorption and desorption), but the hydraulic conductivity and volumetric water content relationship is less affected by hysteresis (Bioubakhsh, 2011). Therefore:  17   ∂ϑ∂t=  ∂∂x ൤ K (ϑ) ∂Ψ∂x ൨                                                          (11)        Equation 11 can be written in the form of a diffusion equation to simplify the mathematical and experimental treatment of unsaturated flow process.  ∂ϑ∂t=  ∂∂x ൤K (ϑ)c(ϑ) ∂ϑ∂x൨                                                            (12)       where c (ϑ) is the specific water capacity (m -1) and c(θ) = dθ/ dΨ.   Therefore, the following equations can describe the capillary suction that results in fluid absorption in unsaturated porous media (Hall, 1994).  ∂ϑ∂t=  ∂∂x ൤D (ϑ) ∂ϑ∂x൨                                                           (13)                  D (ϑ) =K (ϑ)c (ϑ)= K (ϑ)dΨdϑ                                                 (14)                      where x is the depth (m), ϑ is the volumetric water content (L3/L3), D (ϑ) is the capillary diffusivity (water diffusivity/unsaturated hydraulic diffusivity) (m2/s), K(ϑ) is unsaturated hydraulic conductivity (m/s) and Ψ is the capillary potential (m).   Capillary diffusivity depends on the material and fluid properties and pore structure and how much the fluid can be transported through the material by capillary action (Martys & Ferraris, 1997). According to Equations 6 and 14, the factors that influence the fluid flow within particles and through the heap are the size of cracks and the hydraulic conductivity, respectively.   18   2.2 HPGR in heap leach circuits High pressure grinding roll (HPGR) machines have two counter rolls, one of which is fixed and only turns around its axis and the other one floats and rotates in an opposite direction with the fixed one. The crushing force is applied to a bed of particles fed between the rolls by the floating roll.  HPGR was first used in the cement industry and subsequently in diamond processing. In the 1980’s HPGR found applications in the iron ore industry and hard rock industries including gold, copper and platinum mining (Batterham, 2011). Recently, HPGR application in the industry has increased due to several metallurgical benefits. Numerous studies have been conducted to evaluate the potential benefits of applying HPGR to prepare feed for leaching (Dhawan et al., 2012; Ghorbani et al., 2012; Kodali et al., 2011; Nwaila, 2014).   The ore used for heap leach processing may be run-of-mine or crushed ore. In order to use HPGR for direct feed preparation for the heap leaching operation, the circuit shown in Figure 2-2 may be adopted (Dhawan et al., 2013). In this circuit, HPGR is used as a secondary or tertiary crusher instead of a cone crusher. In case of generation of excessive amounts of fines that interfere with heap leach operation agglomeration may be used.   19    Figure 2-2 HPGR in heap leaching operation (Dhawan et al., 2013)   2.2.1    Breakage mechanism  In comminution, particle breakage occurs when tensile, compressive or shear stresses on the particle exceed forces of inter-atomic bonds in the grain. Inter-atomic bonds consist of physical and chemical bonds in crystalline arrangement in minerals. The non-uniform structure of the inter-particle bonds generates internal stresses, and the breakage happens when the stress levels exceed the atomic bond forces (Wills & Napier-Munn, 2011).   The stress concentration in a particle depends on both mechanical properties of the minerals and pre-existing internal cracks in the matrix. There is a critical crack length, beyond which the atomic bonds in grains are broken. By breakage of the atomic bonds, the crack length and stress concentration increase and the fracture occurs (Wills & Napier-Munn, 2011).  20   Types of input forces leading to particle breakage in a particular machine depends on the rock strength and behavior of the ore and the type of loading. In HPGR, ores are crushed by compressive forces applied to the bed of particles rather than impact and abrasion, as in conventional methods. Breakage by compression generates two different product size ranges. Coarse size particles are the result of induced tensile failure and fine size particles are the result of compressive failure close to the points of loading or by shear of the projections (Fandrich, 1999).   The bed compression breakage occurring in HPGR is more efficient than other types of forces occurring in other grinding equipment such as abrasion and chipping (Fandrich, 1999). The breakage force is transferred via particles contacting each other (Mular et al., 2002; Schonert, 1988; Solomon et al., 2010). The breakage force results in generation of fractures along grain boundaries leading to improvements in mineral liberation.   2.2.2 Mineral accessibility and microcracks  HPGR is believed to produce higher densities of microcracks with longer depth than other comminution techniques (Ghorbani et al., 2013). Microcracks improve leach solution transport to mineral grains inside ore particles, accelerate leaching kinetics and improve metal recovery (Dhawan et al., 2013). Extraction of valuable minerals does not necessarily need the mineral grains to be completely liberated, and partial mineral exposure to the leach solution can provide a suitable condition for leaching.  Figure 2-3 schematically shows a more rapid and deeper solution penetration for an ore particle crushed by HPGR through the crack network (Ghorbani et al., 2013) compared to the one by cone crusher. In the particles without deep microcracks, the target grains close to the particle centre may not be leached.   21   Figure 2-3 Two leach regimes for HPGR and cone crusher (Ghorbani et al., 2013)    Two factors that influence the breakage and mineral liberation in the HPGR are the ore hardness and orientation of particles respective to the force field (Schonert, 1991). Hilden indicated that harder grains may not break even with applying high pressures (Hilden, 2005). A portion of particles may not be exposed to the differential stress field and remain intact (Daniel, 2007).   Knecht (1994) investigated the oxide and refractory gold ores crushed with HPGR by SEM, and showed that deep microcracks (up to 150 µm) were generated by HPGR. These hard silicified materials were cracked open by HPGR and had improved liberation of the inside grains through microcracks. Knecht and 22   Baum observed 18% improvement in gold extraction in comparison with conventionally crushed materials. The leach time for HPGR samples was 50% less (Knecht & Baum, 1994).   Watson (1994) also investigated the effects of mineralogy on the generation of microcracks. Images showed softer minerals have less visible microcracks in comparison with the harder and brittle minerals such as quartz, feldspar and pyrite. However, when brittle grains were surrounded with softer materials, no microcracks existed in brittle particles. The reason could be the protection happened because of the soft minerals (Watson & Brooks, 1994).  Stephenson (1997) investigated the length of microcracks induced by HPGR in comparison with the conventional crushers. He showed that the density of microcracks for two different gold ores with the size of 300 and 425 µm are enhanced as the result of crushing by HPGR. His analysis was based on the statistical analysis on length of cracks measured by computer mouse for 50 samples (Stephenson, 1997). Stephenson also used Bond work index (BWI) as an ore property to measure the strength of the HPGR  crushed ores. He found up to 5% reduction in Bond work index (BWI) for HPGR materials in comparison with the conventional crushing. Stephenson concluded that the reduction in the strength of rocks by HPGR is because of the microcracks.   However, Daniel (2007) stated that when the feed size and closing screen size are different from the standard Bond Work Index test, the BWI cannot be defined as an ore property. Daniel explained that in Stephenson’s case, BWI can be an indicator of the energy requirement to do the specific work in the ball mill by applying the modified work index (Daniel, 2007).  Stephenson (1997) completed test work on the six different types of ore and obtained different results on the ore strength. The ores used were quartz, marble, pyritic gold ore, siliceous gold ore, lamproite and 23   bauxite. All ores except lamproite and bauxite showed lower strength after HPGR. It was concluded that the reason was the higher amount of voids (porosity) in these two ores. Voids in the ores structure may stop the propagation of microcracks (Daniel, 2007). Daniel also mentioned that microcracks follow the grain boundary path inside the ores (intergranular cracking) (Stephenson, 1997).   Bettersby (1993) also suggested the possibility of the breakage in HPGR materials along the grain boundaries in chalcopyrite as observed with SEM (Bettersby et al., 1993). Klingmann (2005) studied the leaching of two different gold samples from the Golden Queen Mining Co. Ltd.’s Soledad Mountain Gold project. Column leach tests were conducted for two different comminution circuits. One circuit used three stages of crushing with HPGR as a tertiary crusher. The other circuit used four stages of crushing with a vertical shaft impact crusher. Although, the four-stage crushing circuit products were finer, 7.7% to 10.7% additional recovery improvement was observed when HPGR was used (Klingmann, 2005).   In a study, Nwaila (2014) investigated the application of HPGR in crushing gold ore for heap leaching. The approach was mainly concerned with mineralogical aspects. Mineralogical studies were done by optical microscopy, Quantitative Scanning Electron Microscopy (QEMSCAN), X-Ray and Neutron Computed Tomography (X-ray CT/XRCT and NCT), Quantitative X-Ray Diffraction (QXRD) and X-Ray Fluorescence (XRF) (Nwaila, 2014).   Nwaila used three pressures of HPGR (6, 9 and 12 N/mm2) to crush the samples. Nwaila used two different size ranges, 16-11.2 mm and 5.6-4 mm, for the analysis. Finally, column leach tests were performed to measure the gold recovery and suitability of HPGR for heap leaching. The level of microcracks network, as measured by X-ray CT (XMT), was less than 1 vol. %. The microcracks width and porosity percentage in small range of particle sizes were higher in comparison with larger particle size range. By increasing the HPGR pressure, the microcracks connectivity increased.  24    However, considering the connectivity and volume percentage of microcracks, the pressure of 9 N/mm2 was considered optimal. The column leach test results showed 61% recovery for smaller particles and 30% for larger particles. Nwaila showed that this ore is amenable to HPGR crushing for heap leaching due to microcracks generation that improved the metal recovery. However, in this study, the effect of microcracks on solution penetration was not measure precisely (Nwaila, 2014).  In a study by Kodali (2011), different comminution methods including jaw crusher and HPGR as well as column leaching tests were conducted for copper oxide and copper sulphide ores. The extent of particle damage and grain exposure for the ores crushed by HPGR were analyzed and compared to those for the ores crushed by jaw crusher by means of XMT.  The results showed a higher degree of grain exposure especially for the sulphide ores with application of HPGR. The results of column leach tests showed improvement in the recovery for copper oxide ores only and no improvement was observed for copper sulphide ores. The authors attributed this observation to the high grade of copper in the copper sulphide ore that leads to a high recovery for both samples regardless of the comminution technique used (Kodali et al., 2011).  Ghorbani (2012) studied the relation between leaching recovery and porosity and fineness for a sphalerite ore. The sphalerite ore was prepared by HPGR at various pressures as well as cone crusher to obtain three relatively large particle sizes of 25-23, 16-14 and 6.75-5.25 mm. The ores were leached in a circulating fluid, fixed-bed reactor. Eight bioreactors operated for 11 months (Ghorbani et al., 2012).   The extraction increased with reduction in particle size and was higher for the ore prepared by HPGR at the pressure of 9.5 N/mm2. The highest metal extraction obtained at 9.5 N/mm2 is due to the highest 25   density of microcracks at this optimal pressure. The ores crushed by HPGR at all particle sizes showed between 10 to 15% improvement in extraction.   Measurements of porosity by different methods such as X-ray micro tomography (XMT) and mercury intrusion porosimetry for crushed ores showed higher porosity levels when HPGR was used. The porosity for one particle prepared by HPGR, as measured by XMT, Quantitative Evaluation of Minerals by Scanning Electron Microscopy (QEMSCAN) and mercury intrusion porosimetry, were 0.75, 1.07 and 0.55%, respectively (Ghorbani et al., 2011).  Mercury porosimetry showed the lowest fraction of the porosity of the particle, 0.55%, among the techniques. This may be related to the effect of pore connectivity and exclusion of closed pores. For porosity measurements using imaging techniques, the connectivity of the cracks does not affect the results (Ghorbani et al., 2011). Ghorbani (2013) also investigated the interior progression of leaching in large sphalerite particles by means of XMT. Samples of different size fractions prepared by HPGR and cone crusher were studied at different times during leaching. To analyze the leaching progress for an individual particle, each particle image was divided to a succession of shells with different distances from the center as shown in Figure 2-4 (Ghorbani et al., 2013).   26    Figure 2-4 Separation of particle to shells with different distance from particle core for HPGR products (Ghorbani et al., 2013)    To analyze the leaching progress, the concentration of Zn was determined using VGStudio MAX image analysis software for each shell that was related to a distance at different times. Based on shell size that is different for each zone, the Zn grade was calculated. The samples prepared using HPGR exhibited deeper leach solution penetration. Samples prepared by cone crusher, on the other hand, exhibited almost no change in the particle core. This observation was attributed to the presence of microcracks in the samples prepared by HPGR (Ghorbani et al., 2013).   Table 2-3 briefly explains the studies that used HPGR materials for column test to measure the recoveries. For some of them, test works to investigate the microcracks have been conducted.   27   Table 2-3 Studies on HPGR crushed materials with column tests Ore type Number of samples Crusher type Lab test Leach time (Days) Recovery% Microcracks Reference Au  11 HPGR after cone Column test 70-130 71-90 NA (County, 2015) Zn 6 Cone, HPGR (4.5, 9.5 and 12 N/mm2) Column test, XMT, QEMSCAN 330 HPGR:  58-80 Cone: 45-65 XMT, QEMSCAN and porosimetry, porosities were 0.75, 1.07 and 0.55% (Ghorbani et al., 2012) Au 2  HPGR (6, 9 and 12 N/mm2) Column test, XMT, QEMSCAN 14 HPGR (5.6-4 mm): 60.7 HPGR: (16-11.2 mm) 30 0.1-0.9% porosity (large sample 20-670 µm width) (Small 40-150 µm width) (Nwaila, 2014) Cu 40 HPGR, Jaw Column test, XMT 10  HPGR: 70  Jaw: 65 Crack surface area 10 mm2/mm3 (Kodali et al., 2011) Cu 2 HPGR, Cone Column test, mercury porosimetry 150  HPGR: 74 Cone: 66 Porosity 4-10% (Baum & Ausburn, 2011) Au NA HPGR, vertical shaft impact (VSI) Column test 70  HPGR: 79-82 VSI: 58-73 NA (Patzelt et al., 1995) Au NA HPGR, Cone Column tests, SEM NA 18% recovery improvement with HPGR Up to 150 µm microcracks depth  (Knecht & Baum, 1994)    28   Summary The mentioned studies showed that the application of HPGR in comminution circuits improves leaching recoveries (by column leaching tests) mainly through generation of microcracks (investigated by XMT), enhancing the mineral grain exposure. However, in these studies the solution transport phenomena into the particles through microcracks and its relation to the ore properties were not addressed. In addition, the porous media unsaturated hydraulic behavior of HPGR crushed ore that impacts the metal recovery in heap leaching is not fully understood. Therefore, it is necessary to understand the mechanisms of the penetration of solution into the microcracks and micro pores in heap leaching through easy and rapid test procedures.  In industrial laboratories the investigation of potential benefits of HPGR for heap leaching is mainly by measuring the recovery improvement. In the laboratory different types of crushing including HPGR and conventional methods, PSD analysis, SEM and column leach tests are used to evaluate the effects of microcracks on the improved recovery obtained by deeper penetration of leach solution into the particles and accessibility of mineral grains through microcracks and pores.   These investigations are conducted by considering the finer PSD by HPGR, analyzing the extent of mineral microcracks through SEM imaging and measuring the improved recoveries in column leach tests. The analysis using SEM images is, however, limited and relatively small single particles are needed to detect microcracks. Also, single particles are not necessarily representatives of the entire ore sample and they are damaged during the sample preparation.   In addition, liberation of the minerals inside particles and microcracks existence, do not necessarily infer their accessibility to the leach solution, and liberated grains that are not exposed to the leach solution will 29   not leach. Also, column leach tests need two to six months to measure the recovery improvement and the cost of the operations are relatively high for the early engineering studies.     30   Chapter 3: Materials and procedures A simple, cost effective and rapid alternative test protocol capable of evaluating the benefits of the application of HPGR in heap leaching will greatly facilitate the screening evaluation of HPGR applications. Investigation of the potential benefits of HPGR for gold and copper ores considering the generation of microcracks, finer PSD and its effect on leach solution flow is proposed. Laboratory testing procedures were adapted from several disciplines including mineral processing, soil science, powder science and aggregate testing.     To answer the research question, a series of tests were designed to evaluate the extent of microcracks generation and its effects on the solution penetration and absorption into single particles and an assemblage of particle. This chapter provides a description of the materials and preparation methods used in this study. It also describes the test procedures and methods in detail.   3.1 Materials Three types of ores including gold, copper and copper/gold were used in this study. Gold ore was a gold oxide sample (hereinafter called Au), which was the product of the industrial cone crusher from a mine in Nevada, United States.   The copper ore was a copper sulphide sample from a porphyry copper mine (hereinafter called Cu) in central British Columbia, Canada. This ore was previously crushed with gyratory and jaw crushers. The copper/gold ore was a sulphide sample (hereinafter called Cu/Au) from a mine in British Columbia, Canada. The sample of this ore was a gyratory crusher product.   31   There was no control over the particle size of the delivered ores as they were provided by the mines and considered the appropriate size of samples to be crushed by HPGR. After crushing the samples by cone crusher and HPGR and measuring the PSDs in the lab, subsamples were prepared by riffle and rotary sample splitter and for further laboratory tests.   3.1.1 Sample mineralogy Mineralogy information of samples used in this research were obtained from the mine websites. Mine identification cannot be made due to confidentiality related to the HPGR testing agreements on other projects. Au ore The Au sample was obtained from a gold mine from Nevada, US. The mine was an open pit conventional heap leach operation. This oxide gold ore was from a gold-quartz district in an uplifted metamorphic core complex, which was in contact with granitoids and Precambrian metamorphic rocks. The gold was in the form of native and electrum intergrowths with quartz. It was associated with micaceous or carbonates, goethite, pyrite, and occasionally with sphalerite, galena, anglesite/cerrusite and pyrite.   The average grade of the ore was previously determined as 0.061 oz/ton gold. The size of gold in ore was from 1 to 700 µm with the average of 5 to 50 µm. Effective crushing was required for this type of ore because of the small size of gold grains, which were intergrowths within quartz. At the mine site, the ore was crushed with jaw and cone crusher to minus 64 mm for conventional heap leaching.  Cu ore The Cu sample was from an alkalic porphyry copper deposit from a sedimentary–volcanic complex in British Columbia. Mineralization consists of structurally controlled, multidirectional veins and vein stockworks. Copper was associated with chalcopyrite, bornite, chalcocite and pyrite in altered rocks. The 32   Cu ore also contained hematite–magnetite–chalcopyrite replacements and/or veins. Bornite–chalcocite–chalcopyrite associated with pegmatite type veins and magnetite breccias. The copper grade was 0.18%.  Cu/Au ore The sample was from a Cu/Au alkalic porphyry deposit in British Columbia. The copper was associated with quartz, pyrite, chalcopyrite, carbonate, chlorite and tourmaline. This Early Jurassic Cu/Au deposit had sulphide mineralogy and provided a silica-saturated alkalic porphyry system. The grade of this ore was 0.2% copper and 0.014 oz/ton gold. The mineral processing site in the mine used pebble crushing, SAG and ball mill, Isa mill and flotation circuit.  3.1.2 Sample preparation Samples were first crushed using a lab-scale jaw crusher and then by cone crusher and HPGR. For each HPGR test, 242 kg jaw crushed sample was used. The reason for using this mass of samples was the pilot HPGR equipment capability to crush 35-45 ton/hr sample. The total time to crush each sample is 45 seconds. The first 5 to 6 seconds crushing is for stabilizing. Therefore, the mass needed is somewhere between 250 to 300 kg. By screening, the oversize material (larger than 32 mm because of the machine limitation) were separated and returned to the gyratory crusher. Materials under 32 mm were used for HPGR. A subsample for PSD analysis for HPGR and cone crusher products were separated with a rotary splitter and riffle. Figure 3-1 shows the sample preparation flow diagram.    33   Figure 3-1 Sample preparation flow diagram   3.1.3 Lab cone crusher  In cone crushers, an electric motor drives a countershaft via a belt or gear. The countershaft is connected to a tapered pinion, which rotates a gear. An eccentric assembly with an offset tapered bore holds the main shaft and provides an eccentric rotation to the main shaft and, thus, the head  (Revnivtsev et al., 1984). Cone crushers are used for intermediate or fine crushing, and/or to obtain a product with good cubical shape. Figure 3-2 shows a picture of the cone crusher used for the lab testing in this study.    34   Figure 3-2 Cone crusher laboratory scale   3.1.4 Lab HPGR Table 3-1 lists the specifications for the HPGR and cone crusher used to prepare the samples for this research. Figure 3-3 (a) and Figure 3-3 (b) show a photograph and schematic diagram of the HPGR equipment. Pressures between 1.5 and 8.5 N/mm2 could be used with this equipment.    35   Table 3-1 Crushers specifications HPGR Manufacturer Koppern Germany Roll diameter 750 mm Roll width 220 mm Capacity 35 – 45 tph Maximum specific pressing force 8.5 N/mm2 Cone crusher  Open side setting 38 mm Closed side setting 19 mm  Table 3-2 lists the HPGR pressures used for each sample. These samples were previously prepared for a series of other tests and different HPGR pressures were selected for each ore types. Roll speed was fixed at 19.1 rotations per minute (rpm). The total process time was 45 seconds. The product obtained in the first 5 seconds was separated as waste material. The center and edge products, obtained between 5 to 25 seconds, were collected.   Table 3-2 HPGR pressures Sample  HPGR pressure (N/mm2) Gold ore (Au ore) 4.5 Copper ore (Cu ore) 2.75 Gold/copper ore (Cu/Au ore) 2.5, 3.5, 4.5 and 6    36   Figure 3-3 (a) A photograph of HPGR laboratory scale and (b) its schematic diagram (Infomine)      (a) (b) 37   3.1.5 Particle size distribution Screening in mineral processing is an easy yet versatile technique used to separate mineral particles based on their sizes. Screens are simply metallic plates with openings of uniform size. Particle size analysis was completed using dry sieve analysis. A sample was weighed and then screened using sieve series. The sieve series sizes used were 32, 25, 19, 16, 12.5, 8, 5.6, 4, 2.8, 2, 1.4, 1, 0.71, 0.5, 0.355, 0.25, 0.18, 0.125, 0.09, 0.063 and 0.045 mm. Weights retained on each screen were measured and size versus percent passing was plotted, the results are presented in Chapter 4.   3.2 Test procedures  Laboratory tests carried out in this study and their main focus are listed below.   SEM was used for sample (single particle) morphology analysis;   The slake durability test was used to evaluate the changes, due to crushing, in the mechanical stability of the samples;   He pycnometry test was used to measure porosity of fine materials;  Gas adsorption test (BET) was used to measure the surface area, microcracks width and microcracks volume of the crushed samples;  The water absorption test was used for absorbed water percentage measurement;   Column tests were used for the soil water characteristic curve (SWCC) analysis to obtain the suction properties for modeling with HYDRUS program.   For each test, feed, cone and HPGR materials were prepared in different size ranges and specific sample sizes were selected. For the slake durability test, the sample sizes were based on the procedure limitation for the smallest size (2 mm). The other size ranges chosen were based on the sample sizes usually used for heap leaching, up to 16 mm.  38   He pycnometry equipment only measures porosity of fine samples. Therefore, three size ranges of fine samples were selected. In BET test, fine sample ranges prepared with the same size as He pycnometry test were used.  In addition, single particles with the size of 7 mm was used because of the equipment limitation for larger sizes. The same single 7 mm size was also used to obtain SEM images.   Water absorption sample size was based on the ASTM procedure (ASTM C127, 2012). In the column test the whole size sample used for its similarity to the real condition in heap leach operations. Table 3-3 shows the size ranges prepared for different tests.   Table 3-3 Sizes of samples used for the lab tests  SEM Slake durability He pycnometry Nitrogen adsorption Water absorption Column test Feed Single particle of 7 mm Two size ranges between 2 to 32 mm None Single particle of 7 mm None None Cone material Single particle of 7 mm Four size ranges between 2 to 16 mm, also the whole size sample  Three assemblage size ranges under 600 µm Single particle of 7 mm, three assemblage size ranges under 600 µm One kilogram sample of 3.35 to 16 mm Whole size sample HPGR material Single particle of 7 mm Four size ranges between 2 to 16 mm, also the whole size sample Three assemblage size ranges under 600 µm Single particle of 7 mm, three assemblage size ranges under 600 µm One kilogram sample of 3.35 to 16 mm Whole size sample  39   The time frames for each test were as below.  Slake durability: 3 days   He pycnometry: 2 hours  SEM: 2 hours  Water absorption: 2 days  Nitrogen absorption: 2 days  Column tests: 7 days   3.2.1 Scanning electron microscopy  SEM equipped with energy dispersive X-ray spectrometer (EDX) is a useful tool for high-resolution (nano-scale) imaging and elemental analysis (~1 at% accuracy) of various mineralogical samples. SEM can be used to observe the crack networks of polished samples in two dimensions only (Videla et al., 2007).  The SEM was used to search for microcracks in HPGR products in comparison with conventional methods. However, in this method the microcracks can be the result of mounting the sample in resin and polishing action.   Stephenson (1997) used SEM and bottle roll tests to show the microcracks and improved recovery of 2-10% for HPGR samples (Stephenson, 1997). Battersby (1992) also used SEM for HPGR crushed copper ore and showed the possibility of preferential liberation of chalcopyride with grain boundary breakage. However, he did not use any other crusher products for the comparison (Battersby et al., 1992).  Samples of Cu and Cu/Au ores were mounted and polished using conventional metallography methods. 40   Cutting, grinding and polishing can damage brittle samples and cause pre-existing cracks to propagate and fragment the sample. SEM imaging was conducted for feed, cone and HPGR samples with 7 mm sizes to observe crack networks for the samples.   3.2.2 Slake durability  The density of microcracks induced by HPGR, their morphology and geometry are linked to the mechanical properties of the ore. In other words, mechanical properties of an ore should be an indicator of its response to HPGR in terms of microcracks formation. Rock strength is one of the main factors influencing the microcracks generation in HPGR. In porous rocks, the mechanical properties depend on the amount of porosity (Dunn et al., 1973; Hatzor & Palchik, 1997; Hoshino, 1974).  For instance, the point load strength and Brazilian strength of porous Adulam chalks decreases as the porosity increases (Palchik & Hatzor, 2004). For porous Adulam chalks, the uniaxial compressive strength over the point load strength, (ASTM D5731, 2008), decreases with increasing porosity, estimated according to International Society for Rock Mechanics (ISRM) 1985.   Franklin (1972) indicated slake durability index could be used as an indicator of resistance against abrasion. Slake durability of rocks directly depends on their degree of porosity and defects. Slake durability index shows the slaking resistance of a rock to weakening and disintegration. The slake durability index depends on the amount of porosity, ore mineralogy, chemical bonding, grain size, pH and ore fluid characteristic (Franklin & Chandra, 1972).   The slake durability apparatus contains a drum that is made of 2 mm woven-wire square-mesh sieve. The cylindrical drum has a diameter of 140 mm and is 100 mm long. The drum is placed horizontally in a 41   trough and rotates about its axis. The trough can be filled to 20 mm in the drum below the slaking solution. The device is equipped with a motor that rotates the drum at 20 rpm speed for 10 minutes.   The slake durability test procedure includes two cycles of wetting and drying. The slake durability index is a term that indicates dry mass percentage of pieces that are larger than 2 mm and remaining after two wetting and drying cycles. The samples are dried in an oven, and the wetting cycle is conducted by 10 minutes of tumbling and abrasion while samples are rotated in water. The slake durability index is calculated by Equation 15 (ASTM D4644, 2008).   Id (2) = [(Wf-C)/ (B-C)]*100                                         (15) where Id (2) is slake durability index, B, drum and dried specimens mass before the first cycle (g), Wf, drum and dried specimens mass after the second cycle (g) and C, drum mass (g). Figure 3-4 shows a photo of the test apparatus.  Figure 3-4 Slake durability apparatus    42   The slake durability test procedure conducted was a modified form of ASTM D4644. This modified ASTM D4644 test procedure was used for smaller particles in comparison with the standard procedure. In the standard test procedure the sample contains approximately 10 fragments with the mass range of each ranging from 40 to 60 grams, for a total weight of 450 to 550 grams (ASTM D4644, 2008). Subsamples of cone crusher and HPGR between the ranges of 2 to 16 mm were screened to four size fractions for these tests. For the feed, two size fractions between 2 to 32 mm were used. The mass of approximately 500 g was used for the whole size sample tests.   3.2.3 He pycnometry This test was used to measure the porosity for both HPGR and cone crusher materials fine samples. The apparatus used for the tests was the Micromeritics AccuPyc II 1340 Analysis System. The porous, non-porous or powered samples volume is measured by admitting the gas under pressure to the chambers. There are two chambers in the equipment, one for placing the sample and the other empty one as a reference volume.   The materials from HPGR and cone crusher within the size ranges of 600 to 300, 300 to 150 µm, and under 150 µm were used for the analysis. The run time for each test was approximately 60 minutes. The chamber was filled with He and by knowing the chamber volume and the volume of the He filled the chamber and their differences, the volume of sample was calculated.   In addition, the sample mass before the test was used to calculate the bulk powdered density. To calculate the connected plus isolated porosity, measurement of dimensional density is also needed. For dimensional density, the volume of samples was divided by the mass of each sample. The porosity is calculated as the difference between bulk and dimensional density divided by dimensional density. The parameter 43   measured by this test is the bulk-powdered density (ρB). Using the dimensional density, ρD, connected plus isolated porosity, φT, is calculated by Equation 16.  φT = (ρB −  ρD)/ ρB                            (16)  3.2.4 Nitrogen adsorption test Specific surface area (m2/g) of samples prepared by HPGR, cone and feed were investigated by the gas adsorption test based on Brunauer–Emmett–Teller (BET) model (Brunauer, Emmett & Teller, 1938). BET model is the most precise method to measure the specific surface area, if the the BET plot (1/ [VSTP (Po/P)-1] as a function of P/Po) (VSTP is ideal gas volume at standard temperature and pressure, Po/P is relative pressure of nitrogen) is a straight line between 0.05 and 0.3 relative pressure. Figure 3-5 shows the BET plot of Cu/Au sample under 150 μm size crushed by cone crusher.   Figure 3-5 BET plot of Cu/Au sample (under 150 μm size) crushed by cone crusher    44   All BET plots of the samples used in this study were straight line between the relative pressures of 0.05 and 0.3. Therefore, the physisorption analysis was completed for each sample using multi-points BET with 120 selected points for adsorption and desorption.  BET theory is based on the multi-layer physical gas adsorption (usually nitrogen) on solid surface to measure the specific surface area. In the BET method, the nitrogen gas adsorbs on the entire particle surface, including irregularities and internal pores, through the reversible physisorption process. In physisorption, nitrogen molecules form weak bonds with the particle surface, and no chemical reaction occurs (Lowell & Shields, 1998). Nitrogen adsorption on a cold particle is a function of pressure. The stages of nitrogen adsorption on the particles are as below: 1- Micropores filling in very low pressures; 2- Monolayer adsorption in low pressures; 3- Multilayer adsorption in medium pressures; 4- Capillary condensation in high pressures.  The final stage of nitrogen adsoption is in microcracks at the highest relative pressure. The total microcrack volume (cm3/g) is calculated based on the volume of gas when the relative pressure shows the beginning of the capillary condensation in isotherm plots in BET modeling.  The calculation of the average microcrack diameter is based on the cylindrical tube geometry shape pores assumption. The end plateau of the isotherm, shows the highest amount of adsorbed gas. Using the surface area from BET, the microcracks diameter is calculated.   D= 4V/A                                                                          (17) 45   Where D is average microcracks diameter (Angstrom), V is total gas absorbed (cc/g) and A is surface area from BET calculations (m2/g).  After outgassing, the standard mesopore analysis was performed using liquid nitrogen. Depending on the mineralogy, size and porosity in the samples, the duration of the physisorption tests varied from 6 to 40 hours. An Autosorb IQ Automated Gas Sorption Analyzer was used for the test works. To begin the test, outgassing was done for 24 hours for each sample. By outgassing, contaminants, mainly moisture, are completely removed from the surface and pores. Therefore, during the test nitrogen can be adsorbed on all exposed particle surfaces. Figure 3-6 shows the equipment used in this study.  Figure 3-6 A photograph of Autosorb IQ Automated Gas Sorption Analyzer   46   3.2.5 Water absorption test To understand the effectiveness of microcracks on the water absorption capacity, a water absorption test was conducted. This test was used to measure the amount of water that is absorbed into the ore particle’s microcracks in a particle assemblage.   The standard method is for coarse aggregates that are not lightweight, e.g. silica vs. pumas. The coarse aggregate term indicates the granular materials such as crushed stones that are coarser than 4.75 mm (No. 4 sieve). The water absorption quantities for aggregates indicate their durability and the amount of absorbed asphalt binder.   The water absorption test was carried out for HPGR and cone crusher materials according to modified ASTM procedure (ASTM C127, 2012). The mass of the samples prepared was less than the standard test procedure due the limitation of available sample mass. The absorption percentage indicates the amount of water that penetrated into the particle microcracks, because the sample surface is dried first.   One kilogram of samples from cone crusher and HPGR materials were used in this test. The samples were placed in a wire basket of 3.35 mm size and the samples and basket were submerged in the water for 24 hours. The samples were weighed and oven dried for 24 hours at 100 to 110°C and reweighed. The water absorption percentage was calculated by dividing the sample weight loss by the dry weight (ASTM C127, 2012) shown in the Equation 18.   Absorption % = [(B-A) / A] * 100                                    (18) where B is mass of saturated sample (surface dry) in air (g) and A is mass of oven dry sample in air (g).  47   3.2.6 Soil/water characteristic curve analysis To evaluate the differences between cone and HPGR samples unsaturated flow behavior, column tests were conducted. Column tests provided the data required to develop the soil/water characteristic curve (SWCC). The curve obtained by plotting the volumetric water content as a function of matric suction is referred to as SWCC (Buckingham, 1907). SWCC provided the parameters that then were used in the HYDRUS program to estimate the unsaturated hydraulic conductivities of the crushed ores.   Column test In this study, column tests have been conducted to obtain desorption (drying) curve of SWCC. The simple column test usually provides satisfactory results for coarse soils (Fredlund & Rahardjo, 1993). The distance from the water level at the bottom of the column can be converted to matric suction values. To be able to convert the height to matric suction the assumption is that hydrostatic pressure conditions exist for the water within the column.   Figure 3-7 (a) and (b) show the schematic diagram and a photograph of the column used for the soil/water characteristic tests, respectively. The column height was 60 cm and the water outlet tube was at 51 cm height. The column diameter was 9.8 cm. The column consisted of three, 20 cm-long sections, assembled together, for easy loading and unloading of samples.   48    Figure 3-7 (a) Schematic diagram and (b) a photograph of the column used for SWCC      (a) (b) Water outlet Drainage valve 51 cm Drainage valve Water inlet 9.8 cm 49   The procedure used for the column test was as follows: (Chapuis, R. P. , Masse, I., Madinier,. B., Aubertin, 2014): 1. Weighing the empty column; 2. Filling the column with water and weighing it again;  3. Filling the empty column with the crushed ore with whole particle size range and recording the ore mass placed in each 2.5 cm of the column height;  4. Calculating the whole column mass after filling the column, by adding the column weight and added ore mass; 5. Calculating the porosity (φ) of each section before water addition using the ore volume (calculated using the mass and density) and calculating the volume of each column section (using the diameter and height);  6. Filling the column with water from the bottom and measuring the amount of added water and the time needed for filling the pore spaces with water until the water flows from the water outlet; 7. Calculating the mass of filled column with water; 8. Waiting 24 hours for the complete penetration of water into the pores;  9. Starting the drainage and measuring the water discharge with time (for drainage volume and time plot);  10. Waiting 72 hours after the water discharge stops;  11. Removing the material from each column section (every 2.5 cm) by scoop and weighing the removed materials; 12. Placing the samples in oven with 100 degrees centigrade for 24 hours and then weighing the dried materials to calculate water contents; 13. Measuring the water content of the samples by measuring the wet and dried ore weights for each specific elevation in column (every 2.5 cm); 50   14. Calculating volumetric water content (θ) for each section; 15. Plotting the matric suction (the elevation of sample in the column) with corresponding water content (SWCC).   Table 3-4 lists the parameters obtained from the column test and other parameters calculated using equations. These parameters were used to construct the SWCC plots.   51   Table 3-4 Measured parameters of the column test Parameter  Description M1  Mass of empty column (g) Mwo  Mass of water in empty column (g) Me  Mass of column and fittings when filled with water (g) Ms  Mass of dry ore in column with moisture (g) M2  Mass of column and ore before adding water (g) Vw  Water volume added for filling column (cm3) Mtot  Total mass of column, ore and water after filling column (g) ρw  Density of water (g/cm3) ρ  Dry density (ore density with considering porosity) (g/cm3) ρs   Solid density (ore) (g/cm3) Vo  Volume for each section (cm3) V  Total volume of column (cm3) Mso Weight of added solid with moisture (g) Vh Volume of water from moisture in ore before adding water (cm3) Ms1 Weight of dry ore before adding water (g) (Ms1/ρs) Ore volume (cm3) before adding water in each section N Porosity (void volume/section volume) before adding water, (1-(ore volume/ section volume))/ 100 Viw  Initial water content (moisture + added water) (cm3) Ts Filling column time (min) Td Drainage time (day)    52   Table 3-4 Measured parameters of the column test (Cont.) Parameter  Description Vdw Drained water (cm3) M01 Weight of ore after drainage (g) M02 Weight of ore after drying in the oven after drainage (g) Vo= M02/ ρs Volume of solids (cm3)  Mw= M02 -M01 Weight of water (g) Wc= Mw/ M02 Water content (weight of water/weight of dried solid) Sr= Vw/(Vs-Vo) Degree of saturation (volume of water /volume of void (section volume minus ore volume after drying)) Suction (kPa)  0.24 to 5 kPa   After plotting the SWCC with the results obtained from the column tests, some parameters including residual water content, air entry value and saturated water content can be estimated from the plots. The air-entry value, Ψb (AEV), is the magnitude of the matric suction required for air to enter the largest pore in the specimen. The residual water content, θr, is the amount of water retained in the soil, the further removal of water requires high increases in suction (Fredlund & Xing, 1994). Figure 3-8 shows these parameters (Zhai & Rahardjo, 2012).    53   Figure 3-8 SWCC variables (Zhai & Rahardjo, 2012)   With the data obtained from SWCC desorption curve and the method explained later in this section, the unsaturated hydraulic conductivity (Krw) for six samples prepared by HPGR and cone crusher was estimated with HYDRUS modeling.  SWCC data modeling  The relationship between the unsaturated hydraulic conductivity, degree of saturation and water content can be obtained from the SWCC with different types of modeling, as suggested by Burdine (Burdine, 1952), Brooks and Corey (Brooks & Corey, 1964) and others (van Genuchten, 1982). In this study, the Brooks and Corey model was used. The reason why Brooks and Corey was applied related to the validity of the model for both sand-size ores and larger sizes.  First, the Brooks and Corey model is explained in this section. For estimation of unsaturated hydraulic conductivities, the HYDRUS program evaluations were based on the Brooks and Corey model.  Tangent line 54   In the Brooks and Corey modeling, first, the estimation of the effective degree of saturation (Se) (Brooks and Corey, 1954) is required. From the saturation and matric suction curve obtained from the column test, the residual degree of saturation (Sr) can be estimated. Then, Se for each matric suction is calculated by Equation 19.   Sୣ =S − S୰1 − S୰                                                                                    (19)                           The air entry value ψb or (ua – uw)b and Se are used to calculate the pore size distribution index (λ) according to Equation 20.  Sୣ = ቊ(uୟ – u୵)b(uୟ – u୵)ቋ஛                                                                    (20)    where (ua - uw)b is ψb and (ua - uw) refers to ψ or matric suction.  The unsaturated hydraulic conductivity is calculated by Equation 21 (Brooks and Corey, 1964). For (ua – uw) ≤ (ua – uw)b, the kw= ks and ks is the hydraulic conductivity at saturation (S=100%).  For (ua – uw) > (ua – uw)b  k୵ = kୱ ൜(uୟ– u୵)b(uୟ – u୵)ൠଶାଷ஛                                                         (21)  Finally, the relative unsaturated hydraulic conductivity (krw) is calculated for each matric suction based on the Equation 22:  Krw= kw (100)/ ks                                                                           (22) 55   Also, this form of the Brooks and Corey (1964) model explained below can be used to analyze the unsaturated hydraulic conductivity as it is evaluated in HYDRUS. Equation 23 and 24 show the Se and Krw.  Se=|αh|(-n)              h < -1/α                                                    (23) Se= 1                     h ≥ -1/α Krw= Ks* Se(2/n+I+2)                                                                 (24) where α is the inverse of the air-entry value, h is suction, n is pore size distribution index, I is pore connectivity parameter, Ks is saturated hydraulic conductivity and Krw is unsaturated hydraulic conductivity.   To estimate the saturated hydraulic conductivity (Ks), permeability of each ore can be evaluated with the Kozeny (Kozeny & Wien, 1927), later modified by Carman (Carman, 1956) model. In this permeability-porosity relation model, the permeability of the porous medium is calculated by knowing the surface areas in materials with the assumptions of having cylindrical tube pores shape with uniform radii. This equation can be used in laminar flow. The version of simplified Kozeny is Equation 25 (Xu and Yu, 2008).   k= ε3/ (5× (1-ε)2 × S2)                                                            (25) where k is permeability (cm2), ε is porosity (cm3/cm3) and S is surface area (cm2/cm3).  To estimate the saturated hydraulic conductivity from the permeability values, Equation 26 is used.   Ks= k ρwg/μw                                                                              (26) where Ks is saturated hydraulic conductivity (m/day), k is permeability (m2), ρw is water density (kg/m3), g is gravitational acceleration (m/day2) and μw is water viscosity (kg/m day). 56   HYDRUS Program HYDRUS program was used to measure the unsaturated hydraulic conductivities with the Brooks and Corey model. HYDRUS 1D is a modeling program for analyzing the water flow and transport in unsaturated porous media. The software is based on a one-dimensional finite element model for simulating the transportation of water and multiple liquids in unsaturated media. PC-PROGRESS Engineering developed the HYDRUS program and the model is free for users (Simunek, 2009).  In inverse modeling with HYDRUS, the parameters obtained from SWCC plots, the quantities of the drained water versus time data and the parameter from drainage plots were used to evaluate the unsaturated hydraulic conductivity. In the modeling, for the boundary conditions, the upper boundary condition was selected as constant flux. The lower boundary condition was selected constant pressure head. Ks was calculated based on the Equation 26. In this model, by using the water content percentage and related time obtained from the draining plot, soil hydraulic properties of the ore determined. After running the inverse model in HYDRUS, the plots of water content, hydraulic conductivity and degree of saturation versus height for large scale were obtained.    57   Chapter 4: Results In this Chapter, results from the laboratory tests and modeling are presented for three different types of samples. The samples included gold oxide (Au), copper sulphide (Cu) and gold/copper sulphide (Cu/Au) ores. The terms ‘Au’ is used for gold oxide, ‘Cu’ for copper sulphide and ‘Cu/Au’ for copper/gold sulphide ores. All the samples were crushed by jaw crusher after receiving them from the mines. A jaw crusher was first used to prepare the samples with proper sizes that are suitable for limitation on feed sizes in HPGR and cone crushers. The next step was to crush them further with HPGR and cone crusher.   The laboratory test work included PSD analysis, SEM imaging, slake durability, He pycnometry, nitrogen adsorption, water absorption and column tests. Modeling of the column test results was performed using the HYDRUS program.   The results showed finer PSD, lower slake durability index, higher porosity, higher water absorption, higher surface area and lower hydraulic conductivity for materials crushed by the HPGR.   4.1 PSD of ores In this section, the PSDs for the samples used in this study and those from other studies are presented. The relationship between the ratio of particle passing size percentages for cone crusher and HPGR, at different pressure ranges, are presented. Figure 4-1 (a), (b) and (c) show the PSDs for feed, HPGR, and cone crusher for Au, Cu and Cu/Au ores, respectively. For these lab tests, HPGR pressures were 4.5 N/mm2 for Au ore, 2.75 N/mm2 for Cu ore and 2.5, 3.5, 4.5 and 6 N/mm2 for Cu/Au ore. Data for Cu cone PSD calculation is in Appendix A.    58    Figure 4-1 PSD for Au (gold oxide) (a), Cu (copper sulphide) (b) and Cu/Au (copper gold sulphide) (c) ores      010203040506070809010010 100 1000 10000 100000Cumulative passing %Particle size (micron) Au HPGR(4.5 N/mm2)Au ConeAu Feed(a) 59   Figure 4-1 PSD for Au (gold oxide) (a), Cu (copper sulphide) (b) and Cu/Au (copper gold sulphide) (c) ores (Cont.)        010203040506070809010010 100 1000 10000 100000Cumulative passing %Particle size (micron)Cu HPGR(2.75 N/mm2)Cu ConeCu Feed010203040506070809010010 100 1000 10000 100000Cumulative passing %Particle size (micron)Cu/Au HPGR(3.5 N/mm2)Cu/Au HPGR(6 N/mm2)Cu/Au HPGR(2.5 N/mm2)Cu/Au HPGR(4.5 N/mm2)Cu/Au FeedCu/Au Cone(b) (c) 60   Table 4-1 shows the 50 and 80 percent passing sizes (the size of the screen for which 50% and 80% of sample mass passes through) for the feed, cone and HPGR. In addition, the reduction ratio (F50 feed/ F50 crushed sample of cone and HPGR) for each crusher was calculated based on the sizes of the feed over crushed ore. For the Cu/Au ore the results show that the reduction ratio increases with increase in HPGR pressure.   Table 4-1 PSD for the ores used in this study Ore Condition F50 (µm) F80 (µm) Reduction ratio Au  Feed 7000 17000  Cone 3200 5800 2.2 HPGR (4.5 N/m2) 2000 4000 3.5 Cu  Feed 12000 20300  Cone 4500 7000 2.7 HPGR (2.75 N/mm2) 2600 7000 4.6 Cu/Au  Feed 20000 28000  Cone 5500 7500 3.6 HPGR (2.5 N/mm2)   3700 9000 5.4 HPGR (3.5 N/mm2) 2700 7000 7.4 HPGR (4.5 N/mm2) 2000 6000 10 HPGR (6 N/mm2) 1800 5000 11.1  Figure 4-2 shows the PSD of feed and cone for all three ores. As the plots show, the finest to coarsest feed samples are Au, Cu and Cu/Au. The finest to coarsest cone samples are the same as the feed.     61   Figure 4-2 Feed and cone PSDs for Au, Cu and Cu/Au ores  Figure 4-3 presents the PSD data for only HPGR crushing. As the plots show, the finest to coarsest HPGR samples are Au 4.5 N/mm2, Cu/Au 6 N/mm2, Cu/Au 4.5 N/mm2, Cu/Au 3.5 N/mm2, Cu 2.75 N/mm2 and Cu/Au 2.5 N/mm2.         010203040506070809010010 100 1000 10000 100000Cumulative passing %Particle size (micron)Cu/Au FeedAu FeedCu FeedCu/Au ConeAu ConeCu Cone62   Figure 4-3 HPGR PSDs for Au, Cu and Cu/Au ores   Comparing the PSD of feed and HPGR shows despite the finer material of Cu feed in comparison with Cu/Au, the HPGR Cu/Au is finer than Cu except at the lowest pressure. Therefore, the PSD of HPGR greatly depends on the pressure not only the feed size.   Table 4-2 lists the F(80) for all samples of HPGR and cone materials of this study. Calculation of F(80) is based on the mass retained on each screen and the related screen sizes. By calculating the cumulative mass for the specific sizes, the screen through which eighty percent mass passed was determined.   The studies indicated finer PSDs (in comparison with cone crusher) for HPGR, regardless of ore type, HPGR pressure and the cone crushed size. Also, Table 4-2 shows the equal passing sizes of cone crusher 010203040506070809010010 100 1000 10000 100000Cumulative passing %Particle size (micron) Au HPGR(4.5 N/mm2)Cu HPGR(2.75 N/mm2)Cu/Au HPGR(2.5 N/mm2)Cu/Au HPGR(3.5 N/mm2)Cu/Au HPGR(4.5 N/mm2)Cu/Au HPGR(6 N/mm2)63   calculated for the related HPGR F(80), F(40) and F(10). These particle passing sizes were selected to compare the large, medium and fine sizes for each crusher. The percentage of the passing sizes of cone samples mass versus HPGR mass is lower in all cases. It indicates for all ore types, HPGR samples are finer than cone crusher.   Table 4-2 PSD for different ore types crushed with cone crusher and HPGR of this study Type F (40) HPGR (µm) F (80) HPGR versus cone F (40) HPGR versus cone F (10) HPGR versus cone Figure Au ore  (4.5 N/mm2) 1400 F (57) F (27) F (7) Figure 4-1 (a) Cu ore  (2.75 N/mm2) 1800 F (80) F (18) F (5) Figure 4-1 (b) Cu/Au ore  (2.5 N/mm2) 2400 F (90) F (22) F (3) Figure 4-1 (c) Cu/Au ore  (3.5 N/mm2) 1700 F (74) F (17) F (2.5) Figure 4-1 (c) Cu/Au ore (4.5 N/mm2) 1300 F (60) F (13) F (2.5) Figure 4-1 (c) Cu/Au ore  (6 N/mm2) 1100 F (52) F (12) F (2.5) Figure 4-1 (c)  HPGR pilot scale test can be used to predict PSDs of HPGR industrial size in hard rock mining. The issue with the pilot scale test is that a relatively large amount of ore mass (250-300 kg) is required for the test. However, lab-size cone crusher test is easy to conduct and available in most commercial laboratories. Therefore, if PSD of cone crusher can be used to predict the PSD of HPGR, the method will be cost-effective and time saving.  64    To determine the relationship between HPGR and cone crusher particle sizes, first the F(80), F(40) and F(10) of the six ores used in this study were plotted. Figure 4-4 shows the size relation from this study for cone crusher and HPGR samples.   Figure 4-4 F(80), F(40) and F(10) of HPGR versus cone crusher (this study)   Based on the results obtained from this study, the relationship between HPGR and cone crusher PSDs can be approximately predicted. Considering the HPGR pressures and the relationship between PSD of cone crusher and HPGR is shown in Figure 4-4, the results were divided in two categories. The first category is for HPGR pressures under the 4.5 N/mm2 and the second is HPGR pressures between 4.5 to 9.5 N/mm2. The average quantities were calculated based on the simple mean average method because only six values were available.  01020304050607080901000 20 40 60 80 100Cone cumulative passing %HPGR cummulative passing %Cu ore (2.75 N/mm2)Au ore (4.5 N/mm2)Cu/Au ore (2.5 N/mm2)Cu/Au ore (3.5 N/mm2)Cu/Au ore (4.5 N/mm2)Cu/Au ore (6 N/mm2)65   Figure 4-5 shows the relationship between cone and HPGR cumulative passing size of this study. The plot shows the ore size categories based on the average sizes of these six samples.   Figure 4-5 Cumulative passing sizes of cone and HPGR for two ore categories with different pressure ranges (this study)   As Figure 4-5 shows, the HPGR finer PSDs, the F(40) and F(10), are half the percentage of cone crusher (F(20) and F(5) of cone crusher). On the other hand, for the larger size, F(80) of HPGR, the ratio is different for the two ore categories. For the first category, pressure under 4.5 N/mm2, the F(80) of both are almost the same. For the second ore category, the pressure between 4.5 and 9.5 N/mm2, the F(80) of HPGR equals to F(57) of the cone.   01020304050607080900 20 40 60 80 100Cone cumulative passing %HPGR cumulative passing %Pressure under 4.5 N/mm2(This study)Pressure between 4.5 and 9.5N/mm2 (This study)66   To confirm the relationship obtained for the PSDs of ores crushed by HPGR and conventional method, i.e., cone crusher, PSD data from several studies were collected, replotted and evaluated. Figure 4-6 shows the PSD for different type of ores crushed by HPGR and cone crusher.  Figure 4-6 PSD of HPGR and cone crusher for different ores (other studies)        010203040506070809010010 100 1000 10000Cumulative passing %Particle size (micron)Cone HPGR0102030405060708090100100 1000 10000 100000Cumulative passing %Particle size (micron)Cone HPGR010203040506070809010010 100 1000 10000Cumulative passing %Particle size (micron)Cone product HPGR010203040506070809010010 100 1000 10000Cumulative passing %Particle size (micron)Cone HPGR(c) (d) (a) (b)  67   Figure 4-6 PSD of HPGR and cone crusher for different ores (other studies) (Cont.)   Figure 4-6 shows PSD for (a) base metal sulphide ore (Palm et al., 2010), (b) unknown ore (van der Meer & Gruendken, 2010), (c) Sphalerite ore (Chapman et al., 2013), (d) platinum group metals (Chapman et al., 2011), (e) zinc ore (Ghorbani et al., 2012) and (f) PGM (Humphries et al., 2006). Table 4-3 shows the equal passing sizes of cone crusher calculated for the related HPGR F(80), F(40) and F(10).    010203040506070809010010 100 1000 10000 100000Cumulative passing %Particle size (micron)Cone HPGR01020304050607080901001000 10000 100000Cumulative passing %Particle size (micron)Cone HPGR(e) (f) 68   Table 4-3 PSD for different ore types processed with cone crusher and HPGR of other studies Type F(80) HPGR (µm) F (80) Cone (µm) F (40) HPGR (µm) F (80) HPGR versus cone F (40) HPGR versus cone F (10) HPGR versus cone Reference Figure Base metal sulphide ore  (9 N/mm2) 3000 4000 500 F (62) F (20) F (5) (Palm et al., 2010) Figure 4-6 (a) NA (pressure is not mentioned in the study) 8000 8000 700 F (78) F (10) NA (van der et al., 2010) Figure 4-6 (b) Sphalerite ore  (9 N/mm2) 2200 2800 400 F (60) F (24) F (4) (Chapman et al., 2011) Figure 4-6 (c) Platinum group metals (PGM)  (9 N/mm2) 2800 4100 420 F (66) F (18) F (2) (Chapman et al., 2011) Figure 4-6 (d) Zinc ore  (9.5 N/mm2) 10000 10800 1900 F (35) F (13) F (2) (Ghorbani, et al., 2012) Figure 4-6 (e) PGM  (3.8 to 5.3 N/mm2) 5000 9000 1000 F (55) F (20) NA (Humphries et al., 2006) Figure 4-6 (f)  To check if the relationship from this study shown in Figure 4-5 could be applied for all other ore types, data for five samples from other studies in Table 4-3 was used. Figure 4-7 shows the size relation from other studies for cone crusher and HPGR materials.    69   Figure 4-7 F(80), F(40) and F(10) of HPGR versus cone crusher (other studies)   After calculating the mean average, the results from this study and other studies are shown in Figure 4-8.   Figure 4-8 Cumulative passing sizes of cone and HPGR for five ores of other studies and six ores of this study  0102030405060700 20 40 60 80 100Cone cumulative passing %HPGR cummulative passing %Base metal sulfide oreSphalerite orePlatinum group metals(PGM)Zinc orePGM01020304050607080900 20 40 60 80 100Cone cumulative passing %HPGR cumulative passing %Pressure under 4.5N/mm2 (This study)Pressure between 4.5 and9.5 N/mm2 (This study)Pressure between 4.5 and9.5 N/mm2 (Otherstudies)70   As it shows in Figure 4-8, the trend is virtually the same for six ores of current studies and five ores from previous studies by other researchers (1 to 3 percent differences). Therefore, it can be concluded the same relationship between cone and HPGR crushed ores for these two range of pressure may be used to predict the sizes.   4.2 Scanning electron microscopy (SEM) After analysis of PSD, SEM imaging was conducted to visually investigate the existence and shape differences of microcracks. Figure 4-9, Figure 4-10, Figure 4-11, Figure 4-12, Figure 4-13 and Figure 4-14 show SEM images of feed, cone and HPGR for Cu and Cu/Au (3.5 N/mm2) ores. These two ores were selected because Au ore was very brittle and probably would break more during the preparation procedure.   The samples were mounted and polished with diamond suspension down to one µm for SEM analysis. There were significant removal of ore fragments for cone and HPGR samples during polishing. This indicates reduced strength and high density of pre-existing microcracks for these particles (cone and HPGR). For the feed materials, on the other hand, the polished ore surface was significantly less damaged during polishing. This shows the lower density of pre-existing microcracks for the feed samples. In images, the darker parts show the microcracks in the ores.   71   Figure 4-9 Cu feed SEM in 50 and 10 µm scale  Figure 4-10 Cu cone SEM in 50 and 10 µm scale  Figure 4-11 Cu HPGR SEM in 50 and 10 µm scale    Cu HPGR  Cu HPGR  Cu Feed Cu Feed Cu Cone Cu Cone Cracks Cracks Cracks Cracks 72   Figure 4-12 Cu/Au feed SEM in 50 and 10 µm scale  Figure 4-13 Cu/Au cone SEM in 50 and 10 µm scale  Figure 4-14 Cu/Au HPGR 3.5 N/mm2 SEM in 50 and 10 µm scale     Cu/Au HPGR 3.5 Cu/Au HPGR 3.5 Cu/Au Feed Cu/Au Feed Cu/Au Cone Cu/Au Cone Crack Cracks Cracks 73   SEM images visual comparison is difficult. To visualize the microcracks more clearly, the images below of the observed cracks were drawn by hand with the use of transparent papers. Figure 4-15 and Figure 4-16 show the microcracks in samples.  Figure 4-15 Microcracks in feed, cone and HPGR of Cu samples (50 µm scale)        Cu Cone Cu Feed Cu HPGR  50 μm 50 μm 50 μm 74   Figure 4-16 Microcracks in feed, cone and HPGR of Cu/Au samples (50 µm scale)      As these show, the presence of microcracks were higher in the HPGR materials. SEM provided the images to observe the microcracks and it is not a quantitative method. Although, due to the particle damage during the samples preparation, this method is not conclusive and is not suggested for further investigation.  4.3 Slake durability  Table 4-4 lists the results for slake durability tests along with the sample weight for the feed, cone, and HPGR of the Au, Cu and Cu/Au materials for different size ranges. For Au ore, three size ranges were used instead of the four used for the other two ores. Au samples were very fine and only samples could be Cu/Au Cone Cu/Au HPGR 3.5 Cu/Au Feed 50 μm 50 μm 50 μm 75   prepared were under 12.7 mm. Detailed method of calculation for Cu samples is presented in Appendix B. The repeatability of the test is discussed in Appendix C.  Table 4-4 Slake durability indices (Id(2) (%)) and related sample weights (W (g)) Sample Size (mm) Au 4.5 N/mm2 Cu 2.75 N/mm2 Cu/Au 2.5 N/mm2 Cu/Au 3.5 N/mm2 Cu/Au 4.5 N/mm2 Cu/Au 6 N/mm2 Id(2) W Id(2) W Id(2) W Id(2) W Id(2) W Id(2) W  Feed 32–12.7 95 159 99 645.5 99 517.4 99 517.4 99 517.4 99 517.4 12.7– 2 82 290.8 89 481 93 177.9 93 177.9 93 177.9 93 177.9 Cone crusher 16–12.7 NA 0 98 16.2 100 21.3 100 21.3 100 21.3 100 21.3 12.7– 9.51 98 10.8 98 62.9 99 81.1 99 81.1 99 81.1 99 81.1 9.51–6.73 97 209.4 98 183.5 99 207.3 99 207.3 99 207.3 99 207.3 6.73–2 85 328.3 92 253.3 92 285.1 92 285.1 92 285.1 92 285.1 HPGR 16–12.7 NA 0 98 44.6 100 69.9 99 63.4 99 72.7 99 37.5 12.7–9.51 96 9.7 93 61.6 98 166.8 97 263.9 97 184.5 98 53.3 9.51–6.73 57 82 90 89.9 98 140.5 96 257.1 98 281.7 95 119.3 6.73–2 47 288.7 70 136.3 74 235.7 84 168.6 91 708.9 82 321.8 Cone  Whole sample 59 397 87 450.6 99 362.1 99 362.1 99 362.1 99 362.1 HPGR Whole sample 31 556.2 42 490.3 54 437.5 51 365.6 55 492.6 48 582.2  When the sample mass is low, the interaction between particles may not happen during tumbling. Therefore, the impact and compression forces on the particles are not enough to break them. Therefore, data for sample weight less than 100 g was not used. In the samples with higher mass, the data are valid 76   due to higher sample weight. Highlighted results in Table 4-4 are the quantities that are not considered valid.  Figure 4-7 shows the ratio of slake durability indices for HPGR divided by that for cone for Au, Cu and Cu/Au ores as a function of size ranges (for samples weights larger than 100g). The numbers smaller than one indicate lower slake durability indices for HPGR materials in comparison with cone crusher. This allows easier comparison.   The difference between cone and HPGR samples becomes clear as the sample size decreases. There is no clear relationship between the HPGR pressure and the slake durability index. The slake durability indices of the whole sample are almost two times higher for the cone crusher material in comparison with that of HPGR product for all ores.   Comparing the ore mineralogy and slake durability results showed the slake index for Cu/Au is higher than that for Cu and Cu is higher than that for Au ore. This means the mechanical stability of Cu/Au is higher than Cu and Cu is higher than Au ore. Even at the highest HPGR pressure, Cu/Au ore is more resistant to slaking than the others are.   77   Figure 4-17 HPGR to cone slake durability index ratios for Au, Cu and Cu/Au ores    Changes in ore mechanical properties are related to the different breakage mechanisms of HPGR and cone crusher. The compression breakage mechanism that occurs in the bed of particles in HPGR influences the ore mechanical properties more than the impact (faster rate of loading) breakage forces in cone crusher. The existence of more microcracks because of the interparticle breakage mechanism with HPGR is the main reason that decreases the particle resistance to the tumbling action in slake durability test.     0.4 0.5 0.6 0.7 0.8 0.9 1 1.19.51–6.73 mm6.73–2 mmWhole sampleHPGR/cone slake durability indices ratioAuCuCu/Au HPGR 2.5 N/mm2Cu/Au HPGR 3.5 N/mm2Cu/Au HPGR 4.5 N/mm2Cu/Au HPGR 6 N/mm278   4.4 He pycnometry To understand the porosity differences between HPGR and cone crusher samples pycnometry tests were conducted. This test could only be used for fine particle sizes due to equipment limitation. Three different fine size ranges were used in the tests.   Table 4-5 shows the results for porosity (%) from He pycnometry tests for HPGR and cone crusher for size ranges under 600 µm. The average quantities are calculated based on the PSD percentage for each particular size. Figure 4-18 shows the plots for HPGR to cone porosity ratios for all three types of ore. The numbers smaller than one indicate higher porosity for cone materials. This allows easier comparison. Detailed calculations are presented in Appendix D for Cu samples. The repeatability of the test is discussed in Appendix E.  Table 4-5 Porosity (%) of the samples from the He pycnometry test Sample Size (µm) Au  Cu  Cu/Au  (2.5 N/mm2) Cu/Au  (3.5 N/mm2)  Cu/Au  (4.5 N/mm2)  Cu/Au  (6 N/mm2)  Cone  600–300 52 30 48 300–150 55 42 50 <150 64 51 61 Average  54 38 51 HPGR  600–300 54 47 53 55 44 36 300–150 62 52 68 52 55 53 <150 62 63 76 65 70 64 Average  59 51 62 55 52 46    79    Figure 4-18 HPGR to cone porosity ratios for Au, Cu and Cu/Au ores    The porosity for HPGR samples is higher than that for cone crusher materials for all samples, except for Cu/Au ore at 4.5 to 6.0 N/mm2 HPGR pressures for 600-300 µm and at 6.0 N/mm2 HPGR pressures for average size ranges. For Cu ore, the difference in porosity of HPGR and cone materials is significant (25%). This can be due to more microcracks for HPGR samples as opposed to cone materials. The porosity is the highest for Au and the lowest for Cu ores regardless of the crushing method. The porosity for Cu/Au ore is between those for Au and Cu ores.   0.6 0.8 1.0 1.2 1.4 1.6 1.8600 µm–300 µm300 µm–150 µm<150 µmAverageHPGR/cone porosity ratioAuCuCu/Au (2.5 N/mm2)Cu/Au (3.5 N/mm2)Cu/Au (4.5 N/mm2)Cu/Au (6 N/mm2)80   To verify these observations, the surface area for these fine samples as well as single, large sample (7 mm) of each ore have been measured with gas adsorption testing.   4.5 Nitrogen adsorption test Table 4-6 lists the specific surface area (m2/g) for particle assemblage and single particles of feed, cone and HPGR materials for Au, Cu and Cu/Au ores. The bold text indicates the surface area for HPGR material divided by that for cone material. Therefore, numbers larger than one indicate higher surface area for HPGR materials. This makes the comparison easier. To calculate the average values of the surface area for particles assemblages, data for three different sizes and the percentage of each sizes from the PSD plots were used.   Figure 4-19 shows the ratios of HPGR and cone specific surface area for all ore samples at different sizes of assemblage and 7 mm single particles. Detailed results from BET test for Cu/Au HPGR 6 N/mm2 (under 150 µm) is explained in Appendix F. The repeatability of the test is discussed in Appendix G.   81   Table 4-6 BET specific surface area (m2/g) for particle assemblage and single particle Ore  Sample Size Assemblage Single 600-300 µm 300-150 µm <150 µm Average 7mm Au Cone 11.0 11.0 11.0 10.9 2.0 HPGR (4.5 N/mm2) 5.0 6.0 7.0 5.9 6.0 Feed - - - - 2.5 Cu Cone 5.0 6.0 10.0 6.1 1.0 HPGR (2.75 N/mm2) 7.0 7.0 6.0 6.7 2.5 Feed - - - - 4.3 Cu/Au Cone 2.0 2.0 3.0 2.3 1.0 HPGR (2.5 N/mm2) 4.0 4.0 4.0 4.0 1.0 HPGR  (3.5 N/mm2) 2.0 3.0 3.0 2.3 1.1 HPGR  (4.5 N/mm2) 2.0 3.0 4.0 2.8 1.0 HPGR  (6 N/mm2) 2.0 3.0 4.0 2.8 1.3 Feed - - - - 4.3  The results show that:  For Au assemblage samples, the surface area for cone crushed materials is higher (almost double) than that for HPGR sample for all particle size ranges;  For Cu assemblage samples, the surface area for HPGR materials is higher than that for cone sample for all particle sizes, except for particles smaller than 150 µm;  For Cu/Au assemblage samples, the surface area for HPGR materials is either equal or higher than that for cone crusher. For the lowest HPGR pressure (2.5 N/mm2), the surface area for HPGR materials is clearly higher than that for cone crusher materials;  82    For single particles of 7 mm, the surface areas for HPGR samples are significantly higher than that for cone crusher materials for Au and Cu ores. For Cu/Au, however, there is no significant difference in surface area for HPGR (regardless of pressure) and cone crusher materials.   The higher surface area of single particles in HPGR samples are related to both the existence of higher density of microcracks and non-uniform shape of samples surfaces. The particles crushed with cone crusher had a smoother surface compared to those crushed with HPGR.  83   Figure 4-19 HPGR to cone specific surface area ratios for Au, Cu and Cu/Au ores as a function of particle size   For 7 mm single particles of Cu and Au ores, the specific surface area of HPGR samples are 2.5 and 3.0 times higher than that of cone samples. Particle morphology and internal flaws obtained with different crushing techniques may be responsible for differences in surface area of single particles. The results show HPGR is effective to increase ore microcracks for particle size of 7 mm. This can indicate that there is a certain size range for HPGR effectiveness in terms of microcracks generation.   Nitrogen adsorption test results for both assemblage and single particles showed Au ore has higher specific area than Cu, and Cu ore has higher specific area than Cu/Au. This order reverses when ores are 0 0.5 1 1.5 2 2.5 3 3.5600-300 µm300-150 µm<150 µmAverage7mm (single)HPGR/cone specific surface area ratioCuAuCu/Au HPGR (2.5 N/mm2)Cu/Au HPGR  (3.5 N/mm2)Cu/Au HPGR  (4.5 N/mm2)Cu/Au HPGR  (6.0 N/mm2)84   compared in terms of PSDs. This result for the assemblage samples shows higher surface area in HPGR samples is related to the finer PSD and microcracks. However, in the tests on single particles higher surface area is related to the existence of more microcracks and the shape of the particles.  Table 4-7 shows the average microcracks volume and microcracks width for each ore. The bold numbers in Table 4-7 show the ratios of microcracks volume for HPGR samples divided by that for cone materials. Figure 4-20 shows a plot of these ratios for each ore.  Table 4-7 Average microcracks volume and microcracks width from BET test for assemblage Sample Average microcracks volume (cc/g) Average microcracks width (Å) Cu Cone  0.0027 74 Cu HPGR 0.0057 62 Au Cone 0.0094 53 Au HPGR 0.0084 56 Cu/Au Cone 0.0024 94 Cu/Au HPGR (2.5 N/mm2) 0.0033 77 Cu/Au HPGR (3.5 N/mm2) 0.0032 85 Cu/Au HPGR (4.5 N/mm2) 0.0030 87 Cu/Au HPGR (6 N/mm2) 0.0038 85    85   Figure 4-20 HPGR to cone microcracks volume ratios for Cu, Au and Cu/Au ores   Except for Au ore, all other results show equal or slightly higher microcracks volumes for HPGR in comparison to cone materials. Au ore is very soft and, hence, it develops fines instead of surface cracking when crushed by HPGR. The highest ratio of microcracks volume for Cu/Au samples is for the 2.5 N/mm2 pressure.  The next step was to understand the effects of microcracks on the leach solution absorption. To study this, water absorption and column tests were used. These two tests helped understand the ore/water behaviour that influences the degree of recovery in heap leaching.  0 0.5 1 1.5 2 2.5HPGR/cone microcracks volume ratio86   4.6 Water absorption The water absorption test was carried out for the cone and HPGR samples of Au, Cu and Cu/Au ores. For this purpose, one kilogram of HPGR and cone crusher samples with 3.35 to 16 mm size was used based on the modified standard ASTM test procedure (ASTM C127, 2012). Table 4-8 shows the results for Au, Cu and Cu/Au ores. Figure 4-21 plots the water absorption for HPGR material divided by that for cone material. Numbers larger than one indicate higher water absorption for HPGR materials.   Table 4-8 Water absorption results Sample Crusher Water absorption% Au Cone 2.7 HPGR 2.7 Cu  Cone 7.6 HPGR 9.4 Cu/Au   Cone 1.3 HPGR (2.5 N/mm2) 1.4 HPGR (3.5 N/mm2) 1.6 HPGR (4.5 N/mm2) 2 HPGR (6 N/mm2) 2.8    87   Figure 4-21 HPGR to cone water absorption ratios for Au, Cu and Cu/Au ores   The water absorption percentages include the water absorbed inside the microcracks in the particles. For all materials, the water absorption values were higher for HPGR in comparison with cone crusher, except for Au ore. This ore showed the same water absorption for HPGR and cone crusher samples. The higher water absorption for HPGR samples is due to the effect of microcracks.   Water absorption for Cu ore is higher than Au, and Au is higher than Cu/Au ore for both cone and HPGR samples. The reason for higher absorption for Cu ore in both cone and HPGR materials is due to better absorption of water inside the microcracks in this ore. By increasing the HPGR pressure in Cu/Au ore, the amount of absorbed water becomes slightly higher. This can be related to both finer PSD and higher microcracks.  0 0.5 1 1.5 2 2.5CuAuCu/Au (HPGR 2.5 N/mm2) Cu/Au (HPGR 3.5 N/mm2)Cu/Au (HPGR  4.5 N/mm2)Cu/Au (HPGR 6.0 N/mm2)HPGR/cone water absorption ratio88   The higher water absorption for HPGR samples shows the effectiveness of microcracks in increasing the amount of water absorbed. Higher water absorption by microcracks indicates the better leach solution penetration inside ore particles.   The next question was if this amount of absorbed water could be drained from the heap. In addition, understanding how this phenomenon affects the unsaturated hydraulic behavior in the heap was necessary. Therefore, column tests have been conducted to investigate this behaviour of the ores.  4.7 Column test In this part of the laboratory testing program, column tests for all nine samples were conducted. The column tests provided the information necessary for obtaining the SWCC curves. In addition, drainage as a function of time was obtained from these tests. Using the information from the SWCC plots and free drainage plots, parameters for modeling were estimated.   For modeling, the HYDRUS program was used to analyze the unsaturated hydraulic behaviour of the samples. The results of the modeling focused on unsaturated hydraulic conductivity data. The outcome from these analyses were used to evaluate the effect of comminution method on the hydraulic conductivity of soil.   The data obtained from column tests were degree of saturation, average porosity in the column and the amount of water needed to saturate the column before drainage. In the column test, the degree of saturation before drainage, the porosity and initial water content were calculated by measuring the mass of the input ore and volume of the added water. The moisture content of each ore was measured prior to the tests. As Table 4-9 shows, the moisture content of all ores was less than 1% and these values were considered in all water calculations.   89   Table 4-9 Moisture content of ores Ores Moisture content% Cu Cone 0.23 Cu HPGR 0.73 Au Cone 0.23 Au HPGR 0.58 Cu/Au Cone 0.75 Cu/Au HPGR (2.5 N/mm2) 0.28 Cu/Au HPGR (3.5 N/mm2) 0.08 Cu/Au HPGR (4.5 N/mm2) 0.21 Cu/Au HPGR (6 N/mm2) 0.17  After filling the column with ores and water, parameters calculated before drainage are shown in Figure 4-22 for Cu cone ore. Detailed results from column test for Cu cone ore are given in Appendix H.  Figure 4-22 Phase diagram of Cu cone (Vt-total volume, Va-air volume, Vw-water volume, Vs-solid volume, Mw-water mass, Ms-solid mass, Mt-total mass)       3845 cm3 1428.9 cm3 134.5 cm3 1428.9 g 2281.5 cm3 7475 g 6046.1 g 90   Table 4-10 shows these values (mass of the solids in the column, void volume and initial water added to fill the column) for all samples.   Table 4-10 Initial values in column tests for samples Ores Mass of solids (g) Volume of void (cm3) Initial water volume (cm3) Cu Cone 6046 134.5 1429 Cu HPGR 6209 206.2 1296 Au Cone 5817 311.6 1338 Au HPGR 6319 233.5 1227 Cu/Au Cone 5077 32.6 1406 Cu/Au HPGR (2.5 N/mm2) 6858 167.6 1089 Cu/Au HPGR (3.5 N/mm2) 7014 162.5 1036 Cu/Au HPGR (4.5 N/mm2) 6633 190.0 1152 Cu/Au HPGR (6 N/mm2) 6705 139.5 1175  From these values, the degree of saturation and porosity in the column before and after adding water were calculated from Equations 27 and 28.  Sr=Vw/Va                                                                      (27) Porosity= Va/Vt                                                          (28) where Vw is water quantity as moisture and added water to fill the column and Va is total volume of column minus volume of the solid.  91   Table 4-11 shows the initial results from the column tests for nine samples of cone crusher and HPGR with the whole sample size range. The degree of saturation before drainage, porosity before and after adding the water.  Table 4-11 Initial results from column tests  Ores Sr before drainage (%) Porosity before adding water (%) Volume of air remaining (%) Cu Cone 91 40 3.5 Cu HPGR 86 39 5.4 Au Cone 81 42 8.1 Au HPGR 84 38 6.0 Cu/Au Cone 98 43 1 Cu/Au HPGR (2.5 N/mm2) 87 33 4.4 Cu/Au HPGR (3.5 N/mm2) 86 31 4.2 Cu/Au HPGR (4.5 N/mm2) 86 35 4.9 Cu/Au HPGR (6 N/mm2) 89 34 3.6  Figure 4-23 shows the Sr (%), porosity before adding water (%) and initial water volume (cm3) for HPGR divided by those for cone samples. Numbers higher than one indicate larger parameters for HPGR materials. This allows easier comparison.    92   Figure 4-23 HPGR/cone ratios for Sr (%), porosity before adding water (%) and initial added water volume (cm3) in column test    The observations for Au, Cu and Cu/Au ores of HPGR and cone crusher materials indicated:  Higher degree of saturation before drainage in column for: o Cu cone crusher materials in comparison with Cu HPGR; o Au HPGR materials in comparison with Au cone; o Cu/Au cone crusher materials in comparison with Cu/Au HPGR;  Higher column porosity for cone crusher materials because of the coarser PSD for cone materials;  More water needed for filling the column for cone crusher materials because of the higher porosity. 0.6 0.7 0.8 0.9 1 1.1 1.2Sr%  before drainagePorosity before addingwater %Initial water volume(cm3)HPGR/cone ratioCu/Au HPGR (6 N/mm2)Cu/Au HPGR (4.5 N/mm2)Cu/Au HPGR (3.5 N/mm2)Cu/Au HPGR (2.5 N/mm2)AuCu93    These results are related to the finer PSD in HPGR samples.   The times taken to fill the column with water were between one to five hours for all ores except Cu ore, which took 15 minutes for filling. For Au HPGR and Cu/Au HPGR (6 N/mm2), filling times were as long as about five hours. The Cu/Au cone crusher materials showed the highest degree of saturation among other ores. This ore had the largest PSD and lowest amount of fines in comparison with Au and Cu/Au ores.  After filling the column with water, samples were left for 24 hours to allow as much water as possible to go into the microcracks. After drainage was completed for each ore, samples were dried separately for every 2.5 cm of the column height. Using the weight of the materials before and after drying, the water content was calculated, based on the formula explained in Table 3-4.  Data from column tests were used to obtain the SWCC plots. Figure 4-24 (a), Figure 4-24 (b) and Figure 4-24 (c) show SWCC plots for Cu, Au and Cu/Au samples, respectively. Figure 4-25 shows SWCC data for Cu/Au ore prepared at different HPGR pressures.   94   Figure 4-24 SWCC plots for Cu (a), Au (b) and Cu/Au (c)     05101520253035400 1 2 3 4 5 6 7 8 9 10Volumetric water content (%)Matric suction (kPa)Cu coneCu HPGR2 per. Mov. Avg.(Cu cone )2 per. Mov. Avg.(Cu HPGR)(a)05101520253035400 1 2 3 4 5 6 7 8 9 10Volumetric water content (%)Matric suction (kPa)Au coneAu HPGR2 per. Mov.Avg. (Aucone)2 per. Mov.Avg. (AuHPGR)(b)95   Figure 4-24 SWCC plots for Cu (a), Au (b) and Cu/Au (c) (Cont.)  Figure 4-25 Volumetric water content data for Cu/Au ore at different HPGR pressures  The methods explained in section 3.2.6 could be used to estimate the residual water content (θr) and saturated water content (θs) from the SWCC plots for modeling. However, the SWCC plots showed none of the tests reached the residual matric suction and moisture content as that is definitely at a matric suction much higher than 6 kPa. 05101520253035400 1 2 3 4 5 6 7 8 9 10Volumetric water content (%)Matric suction (kPa)Cu/Au coneCu/Au HPGR 2.5 N/mm22 per. Mov. Avg. (Cu/Aucone )2 per. Mov. Avg. (Cu/AuHPGR 2.5 N/mm2 )(c)05101520253035400 1 2 3 4 5 6 7 8 9 10Volumetric water content (%)Matric suction (kPa)Cu/Au HPGR 3.5 N/mm2Cu/Au HPGR 4.5 N/mm2Cu/Au HPGR 6.0 N/mm2Linear (Cu/Au HPGR 3.5 N/mm2)Linear (Cu/Au HPGR 4.5 N/mm2 )Linear (Cu/Au HPGR 6.0 N/mm2 )96    The SWCC plots indicate:   The cone samples had lower volumetric water contents than the HPGR due to the extra fines in the HPGR crushed ores;  The plots for the Cu/Au at different pressures were less logical as the lines did not plot in the order of low to high pressures. The highest pressure plot (6 N/mm2) was a little lower than the second highest (4.5 N/mm2). This shows that the water contents of these two samples were about the same and the differences were most probably due to laboratory sampling effects. Therefore, further analysis was not conducted for these two samples;  The Cu plots are the closest to a typical SWCC curve with the air entry value at about 2 kPa for both cone and HPGR samples. However, it was expected that the air entry value for the cone sample would be a little lower due to less fines and larger PSD;  For all the other tests there were definitely influences of sampling effects as the volumetric water contents varied quite a bit;  The results showed that the air entry value for all these other tests except Cu cone, Cu HPGR and Au HPGR samples is in the order of 1 kPa as there is not a clear straight line in the beginning for these samples.   97   The parameter θs obtained from SWCC curves are listed in Table 4-12.   Table 4-12 Saturated water content from SWCC curve for each ore Ore type Saturated water content (θs) %  Cu cone 16 Cu HPGR 21 Au cone 24 Au HPGR 20 Cu/Au cone 7 Cu/Au HPGR 2.5 N/mm2 16 Cu/Au HPGR 3.5 N/mm2 16  Data was used in inverse modeling with HYDRUS to evaluate the soil unsaturated behavior. The unsaturated behavior of soil required to understand the effect of microcracks and PSD of the two different types of crushing. To run HYDRUS with inverse modeling, the amount of drained water versus time was used as the input data. The amount of drained water versus time was measured during the free drainage in the column tests. Figure 4-26, Figure 4-27, Figure 4-28, Figure 4-29, Figure 4-30, Figure 4-31 and Figure 4-32 show the drained water volume versus time after the free drainage for all ore types.         98   Figure 4-26 Cu cone drainage plot    Figure 4-27 Cu HPGR drainage plot     02004006008001000120014000.1 1 10 100 1000 10000Drained water volume (cm3 )Time (min)Time and drainage010020030040050060070080090010000.1 1 10 100 1000 10000Drained water volume (cm3 )Time (min)Time and drainage99   Figure 4-28 Au cone drainage plot  Figure 4-29 Au HPGR drainage plot     010020030040050060070080090010000.1 1 10 100 1000 10000Drained water volume (cm3 )Time (min)Time versus drainage01002003004005006007000.1 1 10 100 1000 10000Drained water volume (cm3 )Time (min)Time versus drainage100    Figure 4-30 Cu/Au cone drainage plot  Figure 4-31 Cu/Au HPGR 2.5 N/mm2 drainage plot    020040060080010001200140016000.01 0.1 1 10 100 1000 10000Drained water volume (cm3 )Time (min)Time versus drainage010020030040050060070080090010000.1 1 10 100 1000 10000Drained water volume (cm3 )Time (min)Time versus drainage101   Figure 4-32 Cu/Au HPGR 3.5 N/mm2 drainage plot  To estimate the Krw by modeling using HYDRUS, the estimation of the residual water content was required. Because the column tests could not provide the residual water content values due to the limited height, θr was estimated from the drainage plots. The volume of water when the drained water volume reached a constant value was assumed to be the residual water content. Knowing the volume of drained water, the volume of water in the column was calculated by subtracting it from the total water added to fill the column. Dividing the volume of the water in the column and column volume, the estimated residual water volumetric content percentage was obtained.   The shapes of the drainage curves also indicate that most of the columns represent a dual porosity behavior. This is observed by the rapid release of water in the beginning, typically of water draining from large sized pores. The drainage that follows has a lower rate of release indicating the presence of smaller pores representing the finer PSD. Table 4-13 shows the quantities.   01002003004005006007008009000.1 1 10 100 1000 10000Drained water volume (cm3 )Time (min)Time versus drainage102   Table 4-13 Residual water content from drainage versus time curves for samples Ore type Residual water content (θr) %  Cu cone 6 Cu HPGR 11 Au cone 12 Au HPGR 13 Cu/Au cone 2 Cu/Au HPGR 2.5 N/mm2 6 Cu/Au HPGR 3.5 N/mm2 5  For all samples, the residual water content and saturated water content are higher for HPGR in comparison with cone ores. The reason is the finer PSD and microcracks in HPGR samples that held the water in the column.   After estimating data from column tests, inverse modeling was completed while considering the effects of microcracks and PSD on the unsaturated hydraulic conductivities. In addition, saturated hydraulic conductivity data was required in the inverse modeling. The next section explains the model used for this estimation.   4.7.1 Analyzing saturated hydraulic conductivity with Kozeny model In this model, specific surface area and porosity values are required to calculate the saturated hydraulic conductivities as explained in section 3.2.6. Surface area was measured from gas adsorption test and the average quantities were calculated by knowing the surface area for four samples and their related PSD percentage in the ores. Porosity data was estimated from the column tests. Ks calculated for ores are shown in Table 4-14. The data shows higher values for cone crusher materials except for Cu/Au samples.  103   Table 4-14 Ks estimation from Kozeny model Ore type Porosity (cm3/ cm3) Estimated surface area (m2/g) k (cm2) Ks (cm/min) Cu cone 40 1.59 8.58E-12 4.36E-03 Cu HPGR 39 3.70 1.46E-12 7.39E-04 Au cone 42 3.95 1.64E-12 8.34E-04 Au HPGR 38 6.00 5.07E-13 2.58E-04 Cu/Au cone 43 1.06 2.47E-11 1.26E-02 Cu/Au HPGR 2.5 N/mm2 33 1.90 3.18E-12 1.62E-03 Cu/Au HPGR 3.5 N/mm2 31 1.52 4.06E-12 2.06E-03  4.7.2 Analyzing hydraulic conductivity with Brooks and Corey model In the Brooks and Corey model previously explained in section 4.7.2, estimation of α and n for each sample was required. The air entry values were estimated from SWCC plots and α is the inverse of the air-entry value. The other parameter was n (pore size distribution index) that was estimated by the HYDRUS program. Table 4-15 shows α and n data.    104   Table 4-15 α and n parameters for Brooks and Coreys model Ore type α n Cu cone 0.5 1.1 Cu HPGR 0.5 1.3 Au cone 1 1.1 Au HPGR 0.5 1.1 Cu/Au cone 1 1.7 Cu/Au HPGR 2.5 N/mm2 1 1.7 Cu/Au HPGR 3.5 N/mm2 1 1.7  4.7.3 Analyzing hydraulic conductivity with HYDRUS program Using the θr estimated from free drainage curve, θs from SWCC plots, Ks from Kozeny model, initial column data on the amount of drained water versus time and α and n from Table 4-15, the unsaturated hydraulic conductivities were calculated with inverse modeling with HYDRUS as shown in Table 4-16. Detailed HYDRUS data and results are listed in Appendix H.   Table 4-16 Hydraulic conductivity for ores from HYDRUS inverse modeling Suction (kPa) Krw Cu (cm/min) Krw Au (cm/min) Krw Cu/Au (cm/min) Cone HPGR Cone HPGR Cone HPGR 2.5 N/mm2 HPGR 3.5 N/mm2 0.98 0.2575E-04 0.2712E-05 0.4725E-05 0.1181E-05 0.2575E-04 0.2712E-05 0.4725E-05 1.96 0.7374E-04 0.8467E-05 0.1353E-04 0.3382E-05 0.7374E-04 0.8467E-05 0.1353E-04 2.94 0.2957E-03 0.3805E-04 0.5425E-04 0.1356E-04 0.2957E-03 0.3805E-04 0.5425E-04 3.92 0.3394E-02 0.5339E-03 0.6228E-03 0.1557E-03 0.3394E-02 0.5339E-03 0.6228E-03 5 0.4360E-02 0.7000E-03 0.8000E-03 0.2000E-03 0.4360E-02 0.7000E-03 0.8000E-03  105   Figure 4-33, Figure 4-34 and Figure 4-35 show unsaturated hydraulic conductivities for all ore types. Note that best fit lines are shown using the type of function indicated in the legends to each figure.  Figure 4-33 Krw from inverse modeling for Cu cone and HPGR   Figure 4-34 Krw from inverse modeling for Au cone and HPGR    0.00E+001.00E-032.00E-033.00E-034.00E-035.00E-036.00E-037.00E-038.00E-030 1 2 3 4 5 6Krw(cm/min)Suction (KPa) Cu oreCu coneCu HPGRExpon. (Cu cone)Expon. (Cu HPGR)0.00E+001.00E-032.00E-033.00E-034.00E-035.00E-036.00E-037.00E-038.00E-030 1 2 3 4 5 6Krw(cm/min)Suction (KPa) Au oreAu coneAu HPGRPower (Au cone)Power (Au HPGR)106   Figure 4-35 Krw from inverse modeling for Cu/Au cone, Cu/Au HPGR 2.5 N/mm2 and Cu/Au HPGR 3.5 N/mm2  The differences between unsaturated hydraulic conductivities shown in the plots indicate the effects of microcracks and PSD from the two different methods of crushing. The amount of water that remains between and inside the particles causes the difference in speed of water flow. The effects of these differences are further discussed in Chapter 5.   0.00E+001.00E-032.00E-033.00E-034.00E-035.00E-036.00E-037.00E-038.00E-030 1 2 3 4 5 6Krw(cm/min)Suction (KPa) Cu/Au oreCu/Au coneCu/Au HPGR 2.5 N/mm2Cu/Au HPGR 3.5 N/mm2Power (Cu/Au cone)Power (Cu/Au HPGR 2.5N/mm2)Power (Cu/Au HPGR 3.5N/mm2)107   Chapter 5: Discussions and analysis In this section, the results of laboratory tests presented in chapter 4 are discussed. This Chapter discusses the following:   Relationship between HPGR PSD and slake durability indices;  Estimation of HPGR microcracks generation and suggested method to calculate the quantity;  Estimation of water penetration into microcracks;  Analysis of the effect of microcracks on unsaturated flow through heap materials.   5.1 HPGR finer PSDs effect on slake durability indices The first laboratory test was PSD analysis to obtain the differences between cone crusher and HPGR samples. Results showed that at high HPGR pressures, the PSD is finer for HPGR samples in all particle size ranges either large or fine. At low HPGR pressures, the sizes of smaller than the F(80) sizes of HPGR are finer in comparison with cone crusher products. However, for larger sizes (between F(80) and F(100)) PSD are almost the same for both crushers at low HPGR pressures. The finer PSD obtained from HPGR is advantageous for heap leaching due to the higher surface area in grains. In addition, better liberation of target grains with lower energy consumption would be feasible. However, the percolation of the leach solution may be impacted if there are large percentages of fine materials present.  The finer F(80) of HPGR for all pressures impacted the slake durability indices. In the slake durability tests, two types of size ranges were used. The first size range was four specific ranges between 2 and 16 mm that was selected for tests. The specific size ranges were selected to decrease the effect of PSD on slake durability indices and to obtain the better understanding of ores mechanical stability. The sizes were 16–12.7, 12.7-9.51, 9.51-6.73 and 6.73-2 mm. For all sizes, the slake durability indices were lower for 108   HPGR that are related to the lower resistance of HPGR samples to the slaking and tumbling actions because of the microcracks.   On the other hand, for the whole size range samples used in slake durability tests, the indices for HPGR samples were almost half the values for cone crusher samples. The softer crushed ore from HPGR as a result of microcracks has finer PSD and breaks more easily in the slake durability machine.    In the slake durability tests, the results obtained for Cu and Cu/Au whole size samples clearly show the weakness of HPGR samples. For HPGR samples of these two ores, the indices can almost be classified as low (42-55%). In contrast, for cone crushed samples of these two ores the indices are very high (87-99%). The marked differences between indices for HPGR and cone for Cu and Cu/Au ores are due to the lower resistance of HPGR materials, not only the finer PSD. Therefore, it is concluded that the low resistance of HPGR samples is due to the microcracks.   5.2 Microcracks generation from HPGR In this section, by analyzing the result of each test, the lab tests which provided useful information for evaluation of microcracks in crushed ores are indicated. SEM images showed higher densities of microcracks for HPGR materials.. During sample preparation, there was significant damage caused by grinding/polishing, for both HPGR and cone crusher materials. The results from this test are not conclusive in investigating the formation of microcracks, thus it is not suggested for inclusion in the testing protocol.  As explained, slake durability tests showed lower resistance for HPGR materials during the tumbling and slaking actions. The slake durability indices for whole size samples of HPGR of different ore types were 109   lower than those for the cone crusher. The lower indices indicate the lower stability of HPGR samples. This lower resistance is due to the existence of more microcracks in HPGR samples.    A little higher porosity was obtained from He pycnometry for HPGR samples. Although, the problem with this test was a very fine sample sizes could be used for the tests. Sample sizes were under 600 µm and differences between porosity of HPGR and co`ne samples were low. These differences are related to higher microcracks density and finer PSD for HPGR samples that allowed higher values of penetrated He into the HPGR samples.  The higher surface area from BET for HPGR single particles (7 mm) for all ore types also indicates more microcracks in the samples. Especially, for Au and Cu ores, the values were 2.5 to 3 times higher for HPGR ores. For fine sizes (under 600 µm) the surface area of two crushers were almost the same except for Au ore. Au samples had a very low surface area in HPGR crushed ore in comparison with cone crushed samples. Also, the average microcracks volume of Au ore for fine sizes was less for HPGR.   The water absorption test provided the information on the percentage of absorbed water in microcracks. The results showed higher absorption of water into the microcracks for HPGR samples except for Au ore in comparison with cone crusher samples (the values were the same for Au samples). The water absorption values of samples from this test are compared with the results from column test in Figure 5-5.  The results from BET tests showed higher surface area for single particles for Au, Cu and Cu/Au samples respectively. The amounts of absorbed water from the water absorption tests were higher for Cu, Au and Cu/Au ores, respectively. The fewer microcracks in the Cu/Au ore is the reason for smaller values of absorbed water in this sample.   110   From the column tests, lower porosity was observed for HPGR samples. The lower porosity of HPGR samples due to the finer PSD needed less added water for filling the column. On the other hand, the degree of saturation in the column was a little less for HPGR as a result of the existence of more fine particles and trapped air between them.  For Au ore that had the highest surface area for single particles, the porosity was highest in the column. It has the highest surface area and it results in the highest porosity in the column. Also, from the column test, the Au HPGR sample had the highest residual water content, highest saturated water content and the highest residual degree of saturation in comparison with all other ores.   Residual water content and saturated water content values were higher for all HPGR samples in comparison to the same ore from cone crushed materials. Finally, the modeling with HYDRUS program showed the lower unsaturated hydraulic conductivities for HPGR that is related to both finer PSD and microcracks.   Some of the results of lab testing discussed in this section, clearly showed the higher amounts of microcracks for HPGR samples. In the next section, by using the BET test and column test results the estimation of the rough percentage of microcracks over total porosity in the samples is evaluated.  5.2.1 Estimation of microcracks percentage to the total porosity for whole size ores Results from BET and column tests were used to calculate the microcracks percentage of the whole size samples. The volume of microcracks per gram for the single particle was known from BET tests. The porosity values needed for the calculations, were obtained from the porosity measurements in column tests. Table 5-1 shows the calculation for the Cu ore. 111    Table 5-1 Microcracks percentage over total porosity in whole size sample for Cu cone Cu cone  Results Total sample dimensional volume in column (cm3) 3690 Average porosity from column test (%) 40 Empty space based on porosity (cm3) (Porosity× sample dimensional volume) 1476 Filled space based on porosity (cm3) (Dimensional volume minus empty space) 2214 Microcracks volume (cm3/g) (From BET test) 0.0015 Mass of ore (g) 6047 Total microcracks volume in sample (cm3) (Microcracks volume × mass) 8.88 Total microcracks volume/total empty space (%) 0.60  Values for all other samples were calculated using the same method of calculation for Cu cone and are shown in Table 5-2. Figure 5-1 shows the results for all ore types.   112   Table 5-2 Microcracks percentage over total porosity in whole size samples Ore Microcracks percentage over total porosity  of whole size range sample (%) Cu cone 0.6 Cu HPGR 1.7 Au cone 1.2 Au HPGR 3.7 Cu/Au cone 0.6 Cu/Au HPGR 2.5 N/mm2 1.0 Cu/Au HPGR 3.5 N/mm2 1.5 Cu/Au HPGR 4.5 N/mm2 0.8 Cu/Au HPGR 6 N/mm2 1.5   Figure 5-1 Microcracks percentage over total porosity in whole size samples  00.511.522.533.54Cu cone Cu HPGR Au cone Au HPGR Cu/Au cone Cu/AuHPGR 2.5N/mm2Cu/AuHPGR 3.5N/mm2Microcracks percentage %Type of oreMicrocracks percentage of whole size samples113   As Figure 5-1 shows, HPGR generates more microcracks in comparison with cone crusher. This result is in agreement with the surface area values obtained from BET tests for large particles that were higher for HPGR samples.   In conclusion, the BET and column tests are useful methods to roughly estimate the microcracks quantities for samples. Although, for the ore types used in this study, the differences are very small. In the next section, a method to understand the water absorption mechanism into the microcracks of ores is presented.   5.3 Analysis of water inside the microcracks As Table 5-3 shows calculation of the water content was conducted with two different methods. The residual water content estimated from the drainage and time plots showed the amount of water remaining on particles and inside microcracks. Because the calculated values were based on the wet condition (before drying the samples) after free drainage.  The second values are water content estimated from SWCC plots. The water content estimated from SWCC plots showed the quantity of water kept inside microcracks and remaining on the surface of particles. This quantity was based on the difference between water volume of the ore after flowing out of the column and after drying in the oven. The difference between these two quantities is the amount of water that is held inside microcracks and remaining on the surface of particles. Figure 5-2 shows these two methods.   114   Figure 5-2 Schematic diagram to show the water in microcracks and on the surface calculation             To estimate the percentage of water in microcracks and on surface over the total remaining water between particles after drainage a method was suggested. In this method, the results from BET and column tests were used.   First of all, the estimation of total microcracks volume inside the ores was required. From the BET test, the microcracks volume for single particle was known. The total dry mass of the ore in the column was known too. Knowing the dry mass (g) and microcracks volume (cm3/g), by multiplying them, the total microcracks volume in the column was calculated (cm3). This total microcracks volume is the maximum water volume that can penetrate into the microcracks.   The assumption made was that the total microcracks spaces could be filled with water. By subtracting the total microcracks volume (the same amount as water in the microcracks) and the difference of water Sample after free drainage  Sample after drying in the oven  Sample out of the column before drying Contains water between particles and inside microcracks Contains water in microcracks and on particle surface Contains zero water θr estimated from free drainage plots Water content estimated from SWCC plots 115   remaining after drainage between the wet and dry conditions, the amount of water inside microcracks and on the surfaces and also the amount of water between particles were calculated. Table 5-3 shows the calculations and results for all nine samples.   Table 5-3 Percentage of water in microcracks and on particle surface over the total remaining water Parameters Cu cone Cu HPGR Au cone Au HPGR Cu/Au cone Cu/Au HPGR  2.5 N/mm2 Cu/Au HPGR  3.5 N/mm2 θr after free drainage (from drained water versus time plot) (%) 6 11 12 13 2 6 5 θ after drying samples (from SWCC plot) (%) 11 14 15 17 3 9 10 Volumetric water content in microcracks and on surfaces (difference between drainage and SWCC plots) (%) 5 3 3 4 1 3 5 Water volume based on two method differences (cm3)  184.5 110.7 129.1 147.6 36.9 125.5 169.7 Initial water volume (cm3) 1429 1296 1338 1227 1406 1089 1039 Drained down water (cm3) 1244 1185 1209 1079 1369 964 869 Percentage of remaining water to initial added water (%) 12.9 8.5 9.6 12.0 2.6 11.5 16.3 Average microcracks volume from BET (cm3/g)  0.0015 0.0039 0.0033 0.0083 0.0019 0.0018 0.0024 Total ore mass in column (g) 6047 6210 5817 6319 5077 6858 7014 Microcracks volume in column (cm3) 8.88 24.53 19.34 52.37 9.72 12.36 17.14 Water in microcracks (cm3) 8.88 24.53 19.33 52.37 9.72 12.36 17.14 Water between particles (cm3) 176 86 110 95 27 113 153 Percentage of water in microcracks and on surface over water between particles (%) 4.8 22.2 15.0 35.5 26.3 9.9 10.1  Figure 5-3 shows the percentage of water remaining after free drainage over the initial added water. For all HPGR samples, the values of remaining water after free drainage over the added water are higher than cone crusher samples except Cu ore. The remaining water values for HPGR samples are higher for Au, 116   Cu and Cu/Au (except 3.5 N/mm2 HPGR pressure) ores respectively. From Figure 5-1, the microcracks percentage for whole size range sample were higher for Au, Cu and Cu/Au ores. Also, results provided in chapter 4 on PSD analysis showed the same relationship (Au is finer than Cu and Cu is finer than Cu/Au). Therefore, the higher volume of remaining water can be the results of presence of more microcracks and finer PSD of the sample.   Figure 5-3 Percentage of water remaining after free drainage over the initial added water  Figure 5-4 shows the percentage of water remaining inside microcracks and on the particle surfaces over the water between pores. These values are higher for HPGR materials except Cu/Au samples. This sample was the coarsest sample among all samples that retained the lowest water in the pores.     024681012141618Cu cone Cu HPGR Au cone Au HPGR Cu/Au cone Cu/AuHPGR 2.5N/mm2Cu/AuHPGR 3.5N/mm2Water percentage %Type of orePercentage of remaining water to initial added water %Percentage of remaining water to initial added water %117    Figure 5-4 Percentage of water in microcracks and on surface over water in pores    Comparing water inside microcracks and on surface of particles from two different tests During the laboratory tests in this study, the amount of remaining water in microcracks and on surfaces was obtained from two different methods. The first method was the simple water absorption test. The results of this test are shown in Table 4-8. The second method was the one explained in this section in Table 5-3. Figure 5-5 shows the values from these two methods.    0510152025303540Cu cone Cu HPGR Au cone Au HPGR Cu/Au cone Cu/AuHPGR 2.5N/mm2Cu/AuHPGR 3.5N/mm2Water percentage %Type of orePercentage of water in microcracks and on surface over water between particles %Percentage of water in microcracks over water between particles %118    Figure 5-5 Water content  in microcracks and on surface of particles in terms of the mass of dry ores (%)  As it is shown in Figure 5-5, the amount of water inside the microcracks and on surfaces is higher for HPGR samples in comparison with cone crusher except for Cu ores. The average values are calculated for all ores from the water absorption and column tests. In the calculation based on the data from the column test, the assumption was that all microcracks can be filled with water. Results show that there are not significant differences by making this assumption. The reason can be the small percentage of microcracks in these ores. However, the conclusions drawn from a comparison of these two methods (water absorption test and column test) are that both tests can be used to estimate the water absorption in microcracks. The water absorption test is very simple and preparation of the sample is very easy. Therefore, this test can be a good indicator of water absorption quantity of ores in the beginning of a project.  5.3 5.62.5 2.511.6 2012345678910Cu cone Cu HPGR Au cone Au HPGR Cu/Au cone Cu/Au HPGR2.5 N/mm2Cu/Au HPGR3.5 N/mm2Water percentage%Type of oreWater absorption test Water percentage from column Average119   5.4 Effect of microcracks on unsaturated hydraulic conductivity In this section, the results from HYDRUS program were used to compare the unsaturated hydraulic conductivity for two different conditions. Figure 5-6 shows the results at 4 kPa.  Figure 5-6 Hydraulic conductivity from inverse modeling with HYDRUS   Figure 5-6 shows results of modeling for six type of samples for unsaturated hydraulic conductivities from Table 4-16. As it shows unsaturated hydraulic conductivity is the highest for Cu/Au ore. In the Cu/Au sample the calculated microcracks percentage was the lowest, the remaining water after the free drainage was the lowest and the water inside microcracks and on surface were the lowest too. Therefore, this ore shows the highest unsaturated hydraulic conductivities.   Results from PSD analysis showed the amounts of fine particle sizes are higher for HPGR materials. This higher quantity of fines along with microcracks affected the unsaturated hydraulic conductivities that can reduce metal recovery. All Krw quantities obtained from HYDRUS were lower for HPGR in comparison with cone crushed material. To enhance the permeability of water flow through the heap and to prevent water trapping between fine particles agglomeration of ore can be used for the HPGR crushed materials. 0.00E+005.00E-041.00E-031.50E-032.00E-032.50E-033.00E-033.50E-034.00E-03Cu cone Cu HPGR Au cone Au HPGR Cu/Au cone Cu/AuHPGR 2.5N/mm2Cu/AuHPGR 3.5N/mm2Krw (cm/min)Ore typeKrw 120   5.5 Suggested testing protocol Conducting the laboratory tests on samples from cone and HPGR crushers provides criteria and parameters to decide if further investigation on a specific ore is feasible. This testing protocol provides data that illustrates the amenability of the ore to be crushed by HPGR for heap leaching instead of conventionally used cone crusher.   The first most conclusive set of procedures including PSD analysis, slake durability test, gas adsorption (BET) and water absorption tests is suggested. The main parameter of the PSD analysis is the percentage  of fines in the ores. In the case of excessive amount of fines in the sample that might decrease the percolation of the leach solution, agglomeration must be considered.  The slake durability index that indicates the mechanical stability of the ore provides information of the amenability of the ore to be crushed with HPGR. High slake durability indices can result when there are fewer microcracks that can reduce the particle integrity of the crushed ores. On the other hand, low indices show the low hardness of the ore that may indicate the ore is not amenable to be stacked and effectively leached in the heap. Higher values for surface area results from gas adsorption test is beneficial for heap leaching regarding the higher total surface of the particle that is in contact and accessible to the leach solution. In the water absorption test, the values of absorbed water indicate the enhanced leach solution penetration inside the microcracks. Deeper and higher absorption of solution in microcracks improves the metal extraction from the inside grains in the particle.   The second set of  procedures includes column tests. Conducting the column test provides the parameters required for constructing the SWCC plots. The SWCC plot is used to estimate the parameters that are required to understand the hydraulic behavior of the ores during heap leaching. Analyzing the unsaturated hydraulic behavior of the ores with HYDRUS program, can result in a better understanding of the 121   amenability of the ore to be crushed by HPGR for heap leaching and indicates the amount of fines that are very significant. Low unsaturated hydraulic conductivity results in low flows through the heap that may interfere with metal recovery. Also, the column test results were useful to evaluate the existence of microcracks and penetrated water inside them. The information obtained showed the differences between cone and HPGR materials on water absorption and drainage. The design and construction of the column test was simple, however sampling the material in thin layers following leaching presents the introduction of uncertainties. The time needed for each column test was almost one week. The suggested test procedure for the study of ores for the scoping level evaluation is as shown in Figure 5-7.   122    Figure 5-7 Suggested test protocol for determination of ore amenability to be crushed by HPGR for heap leaching in scoping evaluation   5.6 Amenability of samples to be crushed by HPGR for heap leaching After conducting the test procedures on all the samples, the most informative parameters based on the test results and those indicated the characterization of the ores were selected and are shown in Table 5-4. In PSD analysis, the percentage of fines (particles under 200 micron) and F80 are the parameters that were used to indicate the amenability of the ore. Slake durability index for the whole size samples was also selected as a criterion that shows the mechanical stability of the ores. From the gas adsorption test, the •HPGR•Cone crusherComminution•PSD analysis•Slake durability testMechanical characteristics•Gas adsorption test (BET)Particle characteristics•Gas adsorption test (BET)•Water absorption•Column test•HYDRUS modelingAssemblage characteristics123   most important measured value was surface area of the particles that affects the leach solution accessibility to the grains and metal recovery. The amounts of absorbed water in the ores was also chosen as another creterion. Unsaturated hydraulic conductivity data was calculated for the specific matric suction levels and due to its dependency on the saturated hydraulic conductivity, saturated hydraulic conductivity was selected as a parameter. Finally, the residual water content from the column tests was chosen to indicate the drainage behavior of the crushed ores. Hydrualic conductivity through the heap and residual water content in heap leaching affect the metal recovery of the ores.    124   Table 5-4 Parameters selected as creteria to indicate the ore (as an assemblage) amenability to be crushed by HPGR for heap leaching Ore type Fine% (under 200 μm) F(80) (mm) Slake durability index%  Surface area (m2/g)  Water absorption% Ks (cm/min) θr% Cu cone 5 7 87 1.59 7.6 4E-03 6 Cu HPGR 10 7 42 3.70 9.4 7E-04 11 Au cone 10 5.8 59 3.95 2.7 8E-04 12 Au HPGR 15 4 31 6.00 2.7 3E-04 13 Cu/Au cone 4 7.5 99 1.06 1.3 1E-02 2 Cu/Au HPGR (2.5 N/mm2) 10 9 54 1.90 1.4 2E-03 6 Cu/Au HPGR (3.5 N/mm2) 12 7 51 1.52 1.6 2E-03 5 Cu/Au HPGR (4.5 N/mm2) 14 6 55 1.43 2 - - Cu/Au HPGR (6 N/mm2) 14 5 48 1.82 2.8 - -  The following conclusions were drawn by comparing the parameters from the laboratory tests:  Au ore generated the highest surface area in comparison with other samples. However, due to the highest amount of fines and lower F80, this ore is not selected as an amenable ore to be crushed by HPGR for heap leaching. The finer PSD and higher amount of fines resulted in higher volume of water retained between particles and lower hydraulic conductivity that will affect the metal recovery in heap leaching.  Cu ore surface area showed the high values and the fine quantity and F80 was also suitable for heap leach operation. Amount of absorbed water inside microcracks were the highest comparing 125   to other samples and hydraulic behavior of this ore showed the better results in comparison with Au ore. Therefore, Cu ore is suggested for further investigation.  In Cu/Au ores, although the F80 is higher than Cu, the generation of fines had higher values in this ore. The higher slake durability indices, lower surface area and lower absorbed water indicated the lower amount of microcracks. This ore is suggested only to be crushed by HPGR in lower pressures to reduce the amount of fines.   Based on the results from laboratory tests, the following specific range of parameters are suggested for each test. These ranges are suggested based on the nine samples were used in this study and must clearly be refined with further testing.  Ores with fines quantity under 10 percent with the F80 sizes between 7 to 9 mm are considered as amenable ores for HPGR crushing for heap leaching.   Slake durability indices between 40 and 55 percent for the whole size samples.  Surface area of 1.8 m2/g and higher.   Values of absorbed water for 1.5 to 10 percent.   Ks higher than 7E-04 cm/min.   Residual water content under 11 percent.   5.7 Limitation of the research Limitations of this research are categorized in two groups:  1. Samples  In this research, three types of ores with different mineralogies were used for analysis. The mineralogy of the samples affect the test results in several directions. The hardness of the ore 126   affects the extent of microcracks generation. Also, by different mineralogy finer or larger PSD is obtained after crushing. PSD affects the results of the tests used for whole size samples in this study. However, to eliminate the PSD effects, specific size ranges were chosen for all laboratory tests except column tests.  The pressures of the HPGR used for each sample were different as samples had been previously crushed for other laboratory tests. For Au and Cu ores one HPGR pressure was used. However, for Cu/Au ores, four different applied pressures could show the effects of pressure differences.   The samples received from each mine had different feed PSDs. By selecting the specific size ranges of samples, the effects of PSD on test results reduced.  2. Laboratory tests  The slake durability test conducted was a modified version of the standard ASTM D4644 test procedure because of the limitation of sample sizes. Therefore, the indices measured can only be used for comparison and they should not be considered as slake durability indices for the samples.  Porosity values from He pycnometry tests could only be measured for fine samples (under 600 µm). The size of samples could not be larger because of the size limitation in the machine.   BET tests could not be conducted for samples larger than 7 mm. Because of the larger sizes of materials in heap leaching, it would be valuable if the sample sizes could be larger.  In the column test, the materials placement and movement could not be conducted exactly based on every 2.5 cm. Moving the wet materials after drainage introduced some errors. Therefore, in the SWCC plots the ranges for some of the samples contained errors.   In the column tests, because of the height of the column that was 60 cm, only SWCC obtained up to 5 kPa matric suction was possible. In some cases taller columns may provide the residual water 127   content values from SWCC. However, because of the moderate accuracy of removing materials from each 2.5 cm from the shorter column, taller columns were not attempted for this evaluation.   The sample masses received from the mines were limited for this research and water absorption and column tests could not be repeated to evaluate their repeatability.  The repeatability of slake durability, He pycnometry and BET tests were analyzed. Estimated standard deviation and coefficient of variation measured for some samples. The results showed the realistic outcome based on the samples tested using the procedures.    128   Chapter 6: Conclusions, contributions and future work This chapter discusses the conclusions and contributions of this research work and suggests future work to further investigate this topic. This research has been completed to address the main question and objectives listed in the introductory Chapter. The main question of this study was:   Can a simple test protocol be used or developed to evaluate the suitability of using HPGR instead of conventional comminution technique (i.e. cone crusher) in a heap leach circuit for a specific ore in the early stage of the studies?  To address the main question, various laboratory tests and computer modeling were conducted. Individual test results and the combination results from selected tests were analyzed to address four objectives of the research. The general outcome after completing objectives was used to address the main question.  The first objective of this research was to identify the major differences between materials with different mineralogy crushed by the two different techniques in terms of both PSD and microcracks.   To address this objective, results from several laboratory tests were evaluated. PSD analysis showed that all HPGR materials had finer PSDs in comparison with cone crushed samples. In addition, using the results from this study and PSD results from other literature, an empirical relationship between the PSD from cone crusher and that of the HPGR was obtained. This relationship can be used to predict the HPGR F(80), F(40) and F(10) from the lab cone crusher in the early stages of the studies.  SEM imaging used to observe the morphology of microcracks. The microcracks density was higher for HPGR samples in comparison with cone crusher samples. However, because of the particle damage imposed during the sample preparation this test is not suggested in the testing protocol.  129    Slake durability test results indicated the lower mechanical stabilities for HPGR samples on the whole range of particle sizes. This can be related to both finer PSD and lower resistance of HPGR samples resulting from microcracks.   The results of porosity from He pycnometry were higher of the fine samples for HPGR in comparison with cone crusher samples. The data from BET test for single particles of 7 mm showed higher surface area for HPGR materials. Higher amount of absorbed water measured with water absorption tests for HPGR materials. These results are related to the existence of microcracks.   Finally, the column test results showed a lower degree of saturation in the column for HPGR materials that are related to the finer PSD. In addition, the residual water content showed higher values for HPGR that indicates the existence of microcracks and finer PSD that retained more water in the column.  In conclusion, the main benefits of HPGR application for heap leaching are both finer PSD and microcracks. The laboratory test results from this research indicated both these benefits play important roles on factors that impact the amenability of the ores to the heap leach operations. Finer particles and more microcracks impacted the water flow in the column tests. As the results of the HYDRUS modeling indicated, the permeability for solution within particles will be less in the heap for HPGR crushed materials. Generally, finer PSD and microcracks from HPGR increase the surface area of particles and affect the metal recovery in the heap leach operation.     130   The second objective of this study was to establish a method to determine the quantity of microcracks.   This objective was addressed by analyzing data from two laboratory tests. To quantify microcracks for whole size samples, porosity data from column test and microcracks volume data from BET test were used. For whole size range samples, the percentage of microcracks over the total porosity were higher for HPGR materials. Higher density of microcracks improves the accessibility of the leach solution to the target grains inside the particles. Also, surface area of cracked particles is higher than particles with lower amounts of microcracks. Higher accessibility to target minerals and higher surface area inside the particles will improve the metal recovery.  The third objective of this study was to do an investigation the effects of these microcracks on water penetration/absorption.   The values of absorbed water from water absorption test were higher or the same for HPGR materials. BET and column test results were used to further investigate the water penetration inside particles. The differences between residual water contents from drainage plots and water remaining in particles after removing materials from the column indicated the volumetric water percentage remaining inside microcracks and on the particle surfaces. These values were higher for HPGR materials for all ore types except for Cu samples.   In addition, total volume of microcracks in the column was estimated from BET test data. The amounts of water inside the microcracks and on the surfaces of particles remaining at the end of the column tests could be calculated. The ratio of these values were higher for all samples of HPGR except Cu/Au that had the coarsest PSD. The larger quantities of penetrated water inside microcracks and on the particles surface 131   due to the existence of more microcracks and higher surface area in HPGR materials will improve the metal extraction in the heap leach operations.   The fourth objective of this research was to analyse the effect of microcracks and water penetration/absorption on the hydraulic behavior of heap leach ore in heap leaching.  After completing the column tests, plots of drainage as a function of time were used to estimate the residual water contents for each sample. Further investigation of the data of column test provided the information to plot SWCC for six ores. SWCC plots were used to predict the parameters such as saturated water content and air entry values of samples. With the parameters obtained from column test and further investigation by modeling with HYDRUS the unsaturated hydraulic behaviors of samples crushed by cone crusher and HPGR were evaluated. The differences between outputs of the program for cone crusher and HPGR samples showed the effects of microcracks and PSD on the unsaturated hydraulic conductivities and on retaining water in samples.   Completing the laboratory tests and analyzing the results to address the objectives were used to answer the main objective of the research. The main objective was to evaluate simple test protocol that reliably indicates whether the application of HPGR in heap leach circuit, as a comminution technique, offers advantages over the conventional methods.   The suggested testing protocol includes crushing the ore with cone crusher and HPGR, PSD analysis, slake durability test, BET gas adsorption, water absorption and column tests. After completing the testing protocol scoping level decisions can be made based on the selected ranges of parameters. Those samples selected for further work can then be subjected to bottle roll and column testing, as well as a more extensive testing program. 132    6.1 Contributions The most important contributions of this study are as below.  1. Developing the simple, timely and inexpensive test protocol to evaluate the HPGR crushed ore for heap leaching in the scoping stage of industrial project studies.  The suggested group of laboratory tests in this study has not been conducted on HPGR samples previously. Each of the laboratory test requires short time to be completed and procedures are straightforward. The results of each test were beneficial to address the objectives of this research. Therefore, the suggested protocol is considered as an advance in simplified approaches to be applied in laboratories during the screening stages of heap leach project studies.    2. Designing a new method to measure the microcracks in samples.  In the method used to measure the microcracks in this study, an uncomplicated method of measurement is suggested for whole size range samples. This approach to the analysis is advanced and unique because of its capability to measure values for assemblage samples. Previous methods of microcracks measurements suggested in literature are capable to measure the quantities for a single particle. Therefore, this method is more beneficial and can be used for all type of crushers.  3. Developing a simple method to estimate the volume of water penetrating into microcracks.  The differences between residual water content from drainage plot and water content from SWCC plots indicated the volume of remaining water inside the microcracks and on the 133   surface of particles. It was the first time that this method was used on the HPGR ores. This method is able to estimate the absorbed water inside microcracks using an uncomplicated procedure.   4. Estimating the effects of microcracks on hydraulic behavior of the crushed ores.  Conducting column tests and using HYDRUS to study the hydraulic behavior of the HPGR materials for the heap leaching have not been conducted before. The unsaturated hydraulic properties of the ores indicate how solution penetrates between particles (as assemblage), inside the microcracks and the drainage pattern of the pregnant solution. The unsaturated hydraulic behavior of the solution in the heap leach operations has a direct impact on the metal recovery.   134    6.2 Recommendations for future work The following suggestions are proposed by the author to extend this research:  1. Conducting the test procedure on several oxide gold ores to decrease the effect of different mineralogies.  In this study, three different types of ore mineralogies were used for the investigation. Each of them had a different rock resistance as it was evaluated by slake durability test. Performing the test procedure on three samples of the same mineralogy such as gold oxide ore could indicate the effects of HPGR crushing independent of the ore mineralogy.  2. Conducting the tests on the same size distribution samples to eliminate the effects of PSD.  In this study, the samples received from mines had a different PSD. Also, they have been prepared with different methods prior to the start of the research. However, to eliminate the PSD effects, the specific size ranges were selected for all laboratory tests except the column test. However, investigating samples with the closer PSD will indicate the validity of the suggested test protocol more precisely.   3. Further rock mechanics tests to determine the relationship between ore mechanical properties and generation of microcracks.  Some of the rock mechanical tests required specific sample sizes in their procedures that were not accessible during this research. By performing additional tests and providing data on mechanical stability of ores, a deeper understanding on the amenability of crushing ores by HPGR for heap leaching could be developed.  135   4. Further evaluation of the suggested test procedure.  The test methods of this research can be implemented on a much wider basis in combination with column leaching studies to further evaluate and validate these methods.     136   Bibliography Adamson, A. W. (1967). Capillary transport in a tube. In Physical Chemistry of Surfaces. ASTM C127. (2012). Standard test method for density, relative density (specific gravity), and absorption of coarse aggregate. ASTM. ASTM D4644. (2008). Standard Test Method for Slake Durability of Shales and Similar Weak Rocks. ASTM. ASTM D5731. (2008). Standard Test Method for Determination of the Point Load Strength Index of Rock and Application to Rock Strength Classifications. ASTM. Batchelor, G. K. (1967). An Introduction To Fluid Dynamics. Cambridge University Press. Batterham, R. (2011). Trends in comminution driven by energy. 4th Asian Particle Technology Symposium (APT2009) - II, 22(1), 138–140. Battersby, M. J. ., Kellerwessel, H., & Oberheuser, G. (1992). Important advances in the development of high pressure rolls comminutions for the mineral industry. In Extractive metallurgy of gold and base metal (pp. 159–165). Baum, W., & Ausburn, K. (2011). HPGR comminution for optimization of copper leaching. Minerals & Metallurgical Processing, 28 (2), 77–81. Baum, W., Patzelt, N., & Knecht, J. (1996). The use of high-pressure roll grinding for optimization of copper leaching, ME Annual, 68–96. Bettersby, M. J. G., Kellerwessel, H., Oberheuser, G. (1993). Important advances in the development of high pressure rolls comminution for the minerals industry. In Extractive Mettalurgy of Gold and Base Metals (pp. 159–165). Bioubakhsh, S. (2011). The penetration of chloride in concrete subject to wetting and drying: 137   measurement and modelling. Doctoral thesis. UCL (University College London). Brooks, R. H., & Corey, A. T. (1964). Hydraulic properties of porous medium. Hydrology Paper, 3. Brunauer, S., Emmett, P. H., & Teller, E. (1938). Adsorption of gases in multilolecular layers. Journal of the American Chemical Society, 60(2), 309–319. Burdine, N. T. (1952). Relative permeability calculations from pore-size distribution data. Trans. AIME, 198, 71–77. Carman, P. C. (1956). Flow of Gases Through Porous Media. Butterworths Scientific Publications. Chapman, N. A., Shackleton, N. J., Malysiak, V., & Connor, C. T. (2013). Comparative study of the use of HPGR and conventional wet and dry gringing methods on the flotation of base metal sulphides and PGMs. The Journal of the Southern African Institute of Mining and Metallurgy, 113, 407–413. Chapman, N. A., Shackleton, N. J., Malysiak, V., & O’Connor, C. T. (2011). The effect of using different comminution procedures on the flotation of Platinum-Group Minerals. Minerals Engineering, 24(8), 731–736. Chapuis, R. P. , Masse, I., Madinier,. B., Aubertin, M. (2014). A drainage column test for determining unsaturated properties of coarse materials. Geotechnical Testing Journal, 30(2). County, K. (2015). Soledad Mountain Project Technical Report and Updated Feasibility Study. Daniel, M. J. (2007). Energy efficient mineral liberation using HPGR technology. University of Queensland, PhD Thesis. Dhawan, N., Safarzadeh, M., Miller, J. D., Moats, M. S., & Rajamani, R. K. (2013). Crushed ore agglomeration and its control for heap leach operations. Minerals Engineering, 41, 53–70. http://doi.org/10.1016/j.mineng.2012.08.013 Dhawan, N., Safarzadeh, M. S., Miller, J. D., Moats, M. S., Rajamani, R. K., & Lin, C. (2012). Recent 138   advances in the application of X-ray computed tomography in the analysis of heap leaching systems. Minerals Engineering, 35, 75–86.  Dhawan, N., Safarzadeh, M. S., Miller, J. D., Rajamani, R. K., & Moats, M. S. (2012). Insights into heap leach technology (Vol. SME Annual). Dunn, D. E., LaFountain, L. J., & Jackson, R. E. (1973). Porosity dependence and mechanism of brittle fracture in sandstones , J. Geophys. Dunne, R., Goulsbra, A., & Dunlop, I. (1996). High-pressure grinding rolls and the effect on liberation. In Randol Gold Forum. Randol International (pp. 49–54). Golden, Colorado. Fandrich, R. G. (1999). Mineral liberation through confined particle bed breakage. University of Queensland, PhD Thesis. Franklin, J. A., & Chandra, R. (1972). The slake durability test. International Journal of Rock Mechanics and Mining Science, 9, 325–341. Fredlund, D. G., & Rahardjo, H. (1993). Soil mechanics for unsaturated soils. John Wiley & Sons. Fredlund, D. G., & Xing, A. (1994). Equitions for the soil-water characteristic curve. Canadian Geotechnical Journal, 31(4), 521–532. Ghorbani, Y., Becker, M., Mainza, A., Franzidis, J. P., & Petersen, J. (2011). Large particle effects in chemical/biochemical heap leach processes – A review. Special Issue : Bio and Hydrometallurgy, 24(11). Ghorbani, Y., Mainza, A. N., Petersen, J., Becker, M., Franzidis, J.-P., & Kalala, J. T. (2012). Investigation of particles with high crack density produced by HPGR and its effect on the redistribution of the particle size fraction in heaps. Minerals Engineering, 43, 44–51. Ghorbani, Y., Mainza, A. N., Peterson, J., Kalala, J. T., Becker, M., & Franzidis, J.-P. (2011). Use of x-139   ray computed tomography to quantify the differences in cracks and pores of sphalerite ore particles when comminuted using an HPGR and cone crusher for heap leach feed preparation. In Fifth international conference on Autogenous and Semiautogenous Grinding Technology (Vol. September). Ghorbani, Y., Petersen, J., Becker, M., Mainza, A. N., & Franzidis, J. P. (2013). Investigation and modelling of the progression of zinc leaching from large sphalerite ore particles. Hydrometallurgy, 131–132, 8–23. Ghorbani, Y., Petersen, J., Harrison, S. T. L., Tupikina, O. V, Becker, M., Mainza, A. N., & Franzidis, J. P. (2012). An experimental study of the long-term bioleaching of large sphalerite ore particles in a circulating fluid fixed-bed reactor. Hydrometallurgy, 129–130, 161–171.  Hall, C. (1994). A review of fluid transport theory. Materials and Structures, 27, 291–30. Hatzor, Y. H., & Palchik, V. (1997). The influence of grain size and porosity on crack initiation stress and critical flaw length in dolomites. , Int. J. Ro. Hilden, M. (2005). DEM bed breakage simulations. Report, JKMRC, University of Queensland. Hoshino, K. (1974). Effect of porosity on strength of clastic sedimentary rock, Proc. 3rd. Humphries, G., Rule, C., & Wolmarans, E. (2006). The development of a process flowsheet for the new Anglo Platinum, PPRust north concentrator, incorporating HPGR technology. In International Platinum Conference, Platinum Surges Ahead. The Southern African Institute of Mining and Metallurgy. Klingmann, H. L. (2005). Soledad Mountain Project, Flow sheet development and benefits of the HPGR . Golden Queen Mining Co. Knecht, J., & Baum, W. (1994). Optimising refractory and oxide gold ore operations with high pressure grinding rolls. In SME Annual Meeting. 140   Kodali, P., Dhawan, N., Depci, T., Lin, C. L., & Miller, J. D. (2011). Particle damage and exposure analysis in HPGR crushing of selected copper ores for column leaching. Minerals Engineering, 24(13), 1478–1487. Kozeny, J. (1927). Über Kapillare Leitung des Wassers im Boden. Wien, Akad. Wiss, 271–306. Lowell, S., & Shields, J. E. (1998). Powder surface area and porosity (Third). Chapman & Hall. Martys, N. S., & Ferraris, C. F. (1997). Capillary transport in mortars and concrete. Cement and Concrete Research, 27(5). Mular, A. L., Halbe, D. N., & Barratt, D. J. (2002). Mineral Processing Plant Design, Practice, and Control. SME. Nwaila, G. (2014). Application of HPGR and X-Ray CT to investigate the potential of Witwatersrand gold ore for heap leaching: a process mineralogy approach. University of Cape Town. Palchik, V., & Hatzor, Y. H. (2004). The Influence of Porosity on Tensile and Compressive Strength of Porous Chalks, Rock Mecha. Palm, N. A., Shackleton, N. J., Malysiak, V., & O’Connor, C. T. (2010). The effect of using different comminution procedures on the flotation of sphalerite. Minerals Engineering, 23(11–13), 1053–1057. Patzelt, N., Knecht, H., & Baum, W. (1995). Case made for high-pressure roll-grinding in gold plants, 189 of Pol. Revnivtsev, V. I., Kapralov, Y. P., Finkelstein, G. A., Zarogatsky, L. P., Ivanov, N. A., Blekhman, I. I., & Ivanov, B. G. (1984). Selective liberation of minerals in inertial cone crushers. Powder Technology, 38(2), 195–203. Richards, L. A. (1931). Capillary conduction of liquids through porous mediums. Physics, 1(5), 318–333. 141   Rossi, G. (1990). Biohydrometallurgy,. McGraw-Hill, New York. McGraw-Hill. Schonert, K. (1988). A first survey of grinding with high-compression roller mills. International Journal of Mineral Processing, 22(1–4), 401–412. Schonert, K. (1991). Advances in comminution fundamental, and impacts on technology. In XVII International Mineral Processing Congress, Dresden (pp. 1–21). Simunek, J. (2009). The HYDRUS-1D Software Package for Simulating the One-Dimensional Movement of Water, Heat, and Multiple Solutes in Variably-Saturated Media. SCIENCES, DEPARTMENT OF ENVIRONMENTAL RIVERSIDE, UNIVERSITY OF CALIFORNIA RIVERSIDE, CA. Solomon, N., Mainza, A. M., Becker, M., Ross, V., Petersen, P., & Franzidis, J. P. (2010). Effects of high pressure grinding rolls on platinum-bearing ores and the  flotation response as compared to the conventional ball mill. . International Mineral Processing Congress. AusIMM, Brisbane, 1249–1263. Stephenson, I. (1997). The downstream effects of high pressure grinding rolls processing. JKMRC, University of Queensland, PhD Thesis. Tsakiroglou, C. D., Hajdu, K., Terzi, K., Aggelopoulos, C., & Theodoropoulou, M. (2017). A statistical shrinking core model to estimate the overall dechlorination rate of PCE by an assemblage of zero-valent iron nanoparticles. Chemical Engineering Science, 167, 191–203. van der Meer, F. P., & Gruendken, A. (2010). Flowsheet considerations for optimal use of high pressure grinding rolls. Minerals Engineering, 23(9), 663–669. van Genuchten, M. T. (1982). A comparison of numerical solutions of the one-dimensional unsaturated—saturated flow and mass transport equations. Advances in Water Resources, 5(1), 47–55. Videla, A. R., Lin, C. L., & Miller, J. D. (2007). 3D characterization of individual multiphase particles in packed particle beds by X-ray microtomography (XMT). Special Issue To Honor The Late 142   Professor R. Peter King, 84(1–4), 321–326.  von Michaelis, H. (2005). Real and potential metallurgical benefits of HPGR in hard rock ore processing. Proceedings of the Randol Innovative Metallurgy Forum Held in Perth, WA, Australia, 21–24 August, 31–39. Watling, H. R. (2006). The bioleaching of sulphide minerals with emphasis on copper sulphides—a review. Hydrometallurgy, 84(1), 81–108. Watson, S., & Brooks, M. (1994). KCGM evaluation of high pressure grinding rolls technology. In Fifth Mill Operators Conference (pp. 69–83). Wills, B. A., & Napier-Munn, T. J. (2011). Mineral Processing Technology (Seventh). JKMRC Julius Kruttschnitt Mineral research centre. Xu, Peng; Yu, B. (2008). Developing a new form of permeability and Kozeny–Carman constant for homogeneous porous media by means of fractal geometry. Advances in Water Resources, 31, 74–81. Zhai, Q., & Rahardjo, H. (2012). Determination of soil–water characteristic curve variables. Computers and Geotechnics, 42, 37–43.    143    Appendices Appendix A: PSD of Cu sample Table A-1 Copper sulfide (Cu ore) size distribution data Size (mm) Cone Cumulative passing % HPGR Cumulative passing % Feed Cumulative passing % 32 100 100 99.4 25 100 100 84.1 19 100 100 65.9 16 99.9 99.7 56.4 12.5 99.71 97.3 47.5 8 90.29 84.5 29.4 5.6 62.89 74.1 22.3 2.8 45 63 17 2 26.43 52.9 13 1.4 19.01 42.8 10.3 1 14.56 34.7 8.2 0.71 11.67 27.6 6.5 0.5 9.71 22.9 5.4 0.25 7.96 18.2 4.4 0.18 6 15 2.95 0.125 5.16 11.4 2.9 0.09 4.33 8.8 2.4 0.063 3.4 6.3 1.8 0.045 2.53 4.6 1.3 144    Appendix B: Slake durability for Cu sample Table B-1 Slake durability measurement method for Cu sample for assemblage Samples Size (mm) Oven dried weight (g) Sample and pan weight (g) Sample and pan weight after first cycle (g) Sample and pan weight after second cycle (g) pan weight (g) Slake durability % Feed 1 32-12.7 643.6 942.3 937.4 935.1 298.7 98.9 Feed 2 12.7-2 478.7 641.9 590.7 587.1 163.2 88.6 Cone 1 16-12.7 16.2 543.2 543 542.9 527 98.1 Cone 2 12.7-9.51 62.7 360.2 359.6 359.1 297.5 98.2 Cone 3 9.51-6.73 183 561.3 558.5 558.1 378.3 98.2 Cone 4 6.73-2 252.8 745.4 730.7 725.2 492.6 92.0 HPGR 1 16-12.7 44.4 343.4 342.6 342.6 299 98.2 HPGR 2 12.7-9.51 61.3 353 349.8 349 291.7 93.5 HPGR 3 9.51-6.73 89.4 653.8 645.6 644.6 564.4 90.0 HPGR 4 6.73-2 134.9 298.3 263.2 258.3 163.4 70.3    145   Appendix C: Repeatability of slake durability tests To check the repeatability of results obtained from the slake durability tests, the test was repeated for five samples of Cu ore. Table C-1 shows the results. This Table also shows the estimated standard deviation and coefficients of variation. Except HPGR sample in size range of 9.61-6.73, the data show only negligible variation. Data is expected to be in these ranges.   Table C-1 Results of repeated slake durability tests for Cu samples  Sample Size (mm) Test one (Cu 2.75 N/mm2) Test two (Cu 2.75 N/mm2) Average σ (estimated standard deviation) Coefficient of variation (%) Id(2) Id(2) Id(2) Cone 12.7– 9.51 98 98 98 0.0 0.0 9.51–6.73 98 98 98 0.0 0.0 6.73–2 94 92 93 1.8 1.9 HPGR 9.51–6.73 74 90 82 14.2 17.3 6.73–2 68 70 69 1.8 2.6     146   Appendix D: He Pycnometry for Cu sample Table D-1 Pycnometry measurements of Cu samples VOLUME (cm3) CONE 150 CONE 300 CONE 600 HPGR 150 HPGR 300 HPGR 600 CELL CELL 2 171.1255 171.42 171.93 170.67 171.10 171.21 169.13 169.12 170.9826 171.36 171.863 170.62 171.10 171.28 169.14 169.16 171.1513 171.37 171.98 170.56 171.09 171.28 169.12 169.17 170.9965 171.50 172.01 170.61 171.04 171.27 169.12 169.18 171.0643 171.43 172.00 170.54 171.046 171.30 169.12 169.17 171.1724 171.40 171.98 170.63 171.01 171.30 169.13 169.15 171.0941 171.45 171.97 170.63 171.07 171.26 169.15 169.15 171.2254 171.43 171.99 170.79 171.07 171.31 169.14 169.15 171.1253 171.53 172.01 170.69 171.15 171.36 169.151 169.17 171.1006 171.53 172.04 170.64 171.08 171.32 169.14 169.172 171.1038 171.44 171.98 170.63 171.08 171.294 169.13 169.16 0.075 0.06 0.05 0.05 0.04 0.04 0.01 0.02 Volume (cm3) 1.96 2.30 2.81 1.48 1.91 2.13   0.07 0.06 0.05 0.05 0.04 0.04   MASS (g)        6.67 7.37 8.65 4.87 6.07 6.61   0.01 0.01 0.01 0.01 0.01 0.01   Volume of cube 4 cm3   147   Table D-1 Pycnometry measurements of Cu samples (Cont.) CONE 150 CONE 300 CONE 600 HPGR 150 HPGR 300 HPGR 600 CELL CELL 2 DIMENSIONAL DENSITY (g/cm3) 1.6675 1.84375 2.1625 1.2175 1.5175 1.6525   0.041762 0.046161 0.05412 0.03054 0.03802 0.041388   BULK DENSITY (g/cm3) 3.393867 3.201733 3.069357 3.308626 3.166588 3.101961   0.131234 0.085467 0.059332 0.12387 0.073448 0.064915   POROSITY (%) 50.86725 42.414 29.5455 63.20225 52.07775 46.72725   2.405107 1.395938 0.729287 2.961872 1.518093 1.236409      148   Appendix E: Repeatability of He pycnometry tests To check the repeatability of He pycnometry tests, the tests were repeated for some samples. Table E-1 shows the results. For Au samples, two tests were conducted. For Cu/Au HPGR 3.5 N/mm2 three tests were conducted. This Table also shows the estimated standard deviations and coefficients of variation.  Table E-1 Results of He pycnometry tests for repeated tests on three samples Sample Size (µm) Au Test 1 Au Test 2 Cu/Au Test 1 (3.5 N/mm2) Cu/Au Test 2 (3.5 N/mm2) Cu/Au Test 3 (3.5 N/mm2) Average σ (estimated standard deviation) Coefficient of variation (%) Cone  600–300 52 69 - - - 60.5 15.1 24.9 HPGR  600–300 - - 55 72 49 58.6 13.6 23.2 <150 62 70 - - - 66 7.1 10.7  The standard deviation and coefficient of variation for this test are rather high. For example, the porosity for Au cone sample in the size range of 600-300 µm is 60.5 ±15.1 (between 45.4 and 75.6). The coefficient of variation for smaller size (<150 µm) is however close to 11. The results show a range of outcomes as each assemblage had a different distribution of fractures. These results are expected due to sub-sample variability.   149   Appendix F: BET for Cu/Au HPGR 6-under 150 µm sample  Sample weight: 0.6376 g           Analysis gas: Nitrogen          X sect. area: 16.2 Ų/molecule      Non-ideality: 6.58e-05 Adsbate (DRP): Nitrogen          Bath Temp.: 77.30 °C            Outgas Temp: 30.0 °C           Outgas Time: 140.0 hrs         Analysis Time: 596.7 min P/Po tolerance: 0                 Equilibrium time: 10                 End of run:  04/22/2015 21:04 Station #: 1                  PC sw. version: 1.52              Temp Comp: On           150                                                          Table F-1 Multipoint BET for Cu/Au HPGR 6 N/mm2 (under 150 µm) P/Po Volume (cc) 1/ (W ((Po/P)-1)) [cc/g]  7.0470e-03 0.5870 9.673E+00 8.1225e-03 0.6016 1.089E+01 9.2121e-03 0.6141 1.211E+01 1.3147e-02 0.6423 1.660E+01 2.1917e-02 0.6892 2.602E+01 3.2023e-02 0.7260 3.646E+01 1.1036e-01 0.9048 1.097E+02 1.5911e-01 0.9856 1.536E+02 2.0911e-01 1.0572 2.001E+02 3.1034e-01 1.2066 2.984E+02   Area = 3.669E+00 m²/g Slope = 9.451E+02 Y -Intercept = 4.147E+00 Correlation Coefficient = 0.999921 C = 2.289E+02                                                                                     AREA-VOLUME-PORE SIZE SUMMARY SURFACE AREA DATA Multipoint BET..............................................  3.669E+00 m²/g MICROCRACKS VOLUME DATA Total Microcracks volume for pores with Diameter Less than 402.2 Å at P/Po = 0.94970.........................  7.730E-03 cc/g 151    PORE SIZE DATA Average Microcracks diameter.......................................  8.428E+01 Å  DATA REDUCTION PARAMETERS Thermal Transpiration: ON Effective Molecule Diameter (D) 4.0000 Å Effective Cell Stem Inner Diameter (d) 1.0000 mm Last Po Acquired 512.00 mm Hg Maxi Dose: ON Initial Fill: OFF Dose Wizard: OFF  BJH/DH Moving Average Size: 1 Thickness method: DeBoer Interaction Constant (K) 2.9600 nm^3 x kJ/mol ADSORBATE MODEL PARAMETERS  Adsorbate Type = Nitrogen Adsorbate Temp. = 77.3500 K Molecular Weight = 28.0134 g/mol Cross-Sect. Area = 16.2000 Ų/molecule Liquid Density = 0.8060 g/cc Critical Temp. = 126.2000 K Critical Pressure = 33.5000 atm Average Diameter = 0.3000 nm Polarizability = 1.4600 (cc/molec) x 1e-24 152   Magnetic Suscept. = 2.0000 (cc/molec) x 1e-29 Molecular Density = 6.7000 (molec/cm²) x 1e14 Surface Tension = 8.8500 erg/cm² Contact Angle = 0.0000 degrees Supercritical Ads. Constant K = 1.0000  ADSORBENT MODEL PARAMETERS Adsorbent Type = Carbon Atom Diameter = 0.3400 nm Polarizability = 1.0200 (cc/molecule) x 1e-24 Magnetic Suscept. = 13.5000 (cc/molecule) x 1e-29 Surface Atom Density = 38.4500 (molec/cm²) x 1e14 Adsorbent Density = 2.2460 g/cc DR exponent (n) = 2.000  Figure F-1 Isotherm plot for Cu/Au HPGR 6 N/mm2 (under 150 µm)    153   Appendix G: Repeatability of gas adsorption test To check the repeatability of gas adsorption test, the test was repeated for some samples. Table G-1 shows the results. This Table also shows the estimated standard deviation and coefficient of variation.  Table G-1 Results of gas adsorption tests for repeated tests on five samples Surface area (m2/g) Cu/Au HPGR 2.5 N/mm2 (300-150 µm) Cu feed (7 mm) Cu/Au cone (600-300 µm) Cu/Au  HPGR 6 N/mm2 (7 mm) Cu/Au HPGR 2.5 N/mm2 (7 mm) Test 1 3.79 4.35 1.84 0.7 0.94 Test 2 3.06 3.84 1.68 1.31 0.88 Test 3 3.67 - - 1.85 - Average 3.54 4.1 1.76 1.28 0.91 σ (estimated standard deviation) 0.4  0.45  0.14  0.67  0.1  Coefficient of variation (%) 12.5  11.0  8.1  52.8  5.8   Except one sample (Cu/Au HPGR 6 N/mm2 (7 mm)), the coefficients of variations are in acceptable ranges. It is related to the different density of microcracks and different shapes for one single particle have been chosen. The results show a range of outcomes as each particle had a different distribution of fractures. These results are expected due to sub-sample variability.  154   Appendix H: Column test and HYDRUS results for Cu sample Table H-1 Initial measurements of column test for Cu cone Parameters Values M1 (mass of the empty column) (g) 6042.8 Mwo (mass of the water in empty column) (g) 3667.6 Me (mass of the column, fittings when filled with water (g) 9710.4 Ms (mass of dry ore in the column with moisture) (g) 6060.2 M2 (mass of the column and ore before adding water) (g) 12103 Vw (water volume added for filling the column) (cm3) 1415 M tot (total mass of the column, ore and water after filling the column) (g) 13518 ρw (density of water) (g/cm3) 1 Dry density (ore density with considering porosity) 1.59383 ρs (solid density of ore) (g/cm3) 2.65 Vo (column for each section) (cm3) 188.479 V (total volume of the column) (cm3) 3844.96    155   Table H-2 Column test parameters to plot SWCC for Cu cone (1) Height (cm) Suction (cm) Suction (kPa) Volume of water from moisture in ore before adding water (cm3) Weight of added solid with moisture (Mso) (g) 0-2.5 2.5 0.24 0.71 309.40 2.5-5 5 0.49 0.44 193.40 5-7.5 7.5 0.73 0.63 274.00 7.5-10 10 0.98 0.77 336.50 10-12.5 12.5 1.23 0.68 297.10 12.5-15 15 1.47 0.68 296.50 15-17.5 17.5 1.72 0.64 280.10 17.5-20 20 1.96 0.68 297.30 20-22.5 22.5 2.21 0.68 297.30 22.5-25 25 2.45 0.71 309.00 25-27.5 27.5 2.7 0.73 317.50 27.5-30 30 2.94 0.68 293.50 30-32.5 32.5 3.19 0.62 269.20 32.5-35 35 3.43 0.83 361.30 35-37.5 37.5 3.68 0.68 294.60 37.5-40 40 3.92 0.80 346.70 40-42.5 42.5 4.17 0.51 222.70 42.5-45 45 4.41 0.68 297.30 45-47.5 47.5 4.66 0.71 309.80 47.5-50 50 4.9 0.65 281.80 50-51 51 5 0.40 175.20   156   Table H-3 Column test parameters to plot SWCC for Cu cone (2) Height (cm) Suction (cm) Ore volume (cm3) (Ms1/ρs) Ore volume/section volume Porosity (n) (void volume/section volume) 0-2.5 2.5 116.49 0.62 38.20 2.5-5 5 72.81 0.39 61.37 5-7.5 7.5 103.16 0.55 45.27 7.5-10 10 126.69 0.67 32.78 10-12.5 12.5 111.86 0.59 40.65 12.5-15 15 111.63 0.59 40.77 15-17.5 17.5 105.46 0.56 44.05 17.5-20 20 111.93 0.59 40.61 20-22.5 22.5 111.93 0.59 40.61 22.5-25 25 116.34 0.62 38.28 25-27.5 27.5 119.54 0.63 36.58 27.5-30 30 110.50 0.59 41.37 30-32.5 32.5 101.35 0.54 46.23 32.5-35 35 136.03 0.72 27.83 35-37.5 37.5 110.91 0.59 41.15 37.5-40 40 130.53 0.69 30.75 40-42.5 42.5 83.84 0.44 55.52 42.5-45 45 111.93 0.59 40.61 45-47.5 47.5 116.64 0.62 38.12 47.5-50 50 106.10 0.56 43.71 50-51 51 65.96 0.87 12.51   157   Table H-4 Column test parameters to plot SWCC for Cu cone (3) Height (cm) Suction (cm) Weight of ore after drainage (g) Weight of dry ore before adding water (Ms1) (g) Weight of ore after drying in the oven after drainage (g) 0-2.5 2.5 306.10 308.69 268.50 2.5-5 5 272.90 192.96 242.50 5-7.5 7.5 292.20 273.37 269.00 7.5-10 10 347.20 335.73 306.40 10-12.5 12.5 307.00 296.42 279.60 12.5-15 15 324.40 295.82 293.30 15-17.5 17.5 329.80 279.46 296.80 17.5-20 20 284.80 296.62 247.70 20-22.5 22.5 277.90 296.62 261.10 22.5-25 25 356.20 308.29 330.50 25-27.5 27.5 348.70 316.77 321.40 27.5-30 30 345.80 292.82 317.40 30-32.5 32.5 381.60 268.58 352.60 32.5-35 35 296.10 360.47 280.20 35-37.5 37.5 275.20 293.92 258.40 37.5-40 40 297.00 345.90 282.90 40-42.5 42.5 302.40 222.19 288.40 42.5-45 45 270.60 296.62 257.60 45-47.5 47.5 331.10 309.09 312.90 47.5-50 50 357.50 281.15 329.00 50-51 51 115.70 174.80 102.48   158   Table H-5 Column test parameters to plot SWCC for Cu cone (4) Height (cm) Suction (cm) Volume of ore (cm3) Weight of water (Mw) (g) Volume of water (cm3) (Mw/ρ) 0-2.5 2.5 101.32 37.60 37.60 2.5-5 5 91.51 30.40 30.40 5-7.5 7.5 101.51 23.20 23.20 7.5-10 10 115.62 40.80 40.80 10-12.5 12.5 105.51 27.40 27.40 12.5-15 15 110.68 31.10 31.10 15-17.5 17.5 112.00 33.00 33.00 17.5-20 20 93.47 37.10 37.10 20-22.5 22.5 98.53 16.80 16.80 22.5-25 25 124.72 25.70 25.70 25-27.5 27.5 121.28 27.30 27.30 27.5-30 30 119.77 28.40 28.40 30-32.5 32.5 133.06 29.00 29.00 32.5-35 35 105.74 15.90 15.90 35-37.5 37.5 97.51 16.80 16.80 37.5-40 40 106.75 14.10 14.10 40-42.5 42.5 108.83 14.00 14.00 42.5-45 45 97.21 13.00 13.00 45-47.5 47.5 118.08 18.20 18.20 47.5-50 50 124.15 28.50 28.50 50-51 51 38.67 13.22 13.22   159   Table H-6 Column test parameters to plot SWCC for Cu cone (5) Height (cm) Suction (cm) Volume of water from moisture in ore (cm3) Volumetric water content%  (volume of water /total volume of the section) Degree of saturation Sr% (volume of water /volume of void section) (ore volume minus ore volume after drying) 0-2.5 2.5 4.03 19.95 43.14 2.5-5 5 3.64 16.13 31.35 5-7.5 7.5 4.04 12.31 26.68 7.5-10 10 4.60 21.65 56.00 10-12.5 12.5 4.19 14.54 33.02 12.5-15 15 4.40 16.50 39.97 15-17.5 17.5 4.45 17.51 43.15 17.5-20 20 3.72 19.68 39.05 20-22.5 22.5 3.92 8.91 18.68 22.5-25 25 4.96 13.64 40.31 25-27.5 27.5 4.82 14.48 40.63 27.5-30 30 4.76 15.07 41.34 30-32.5 32.5 5.29 15.39 52.33 32.5-35 35 4.20 8.44 19.22 35-37.5 37.5 3.88 8.91 18.47 37.5-40 40 4.24 7.48 17.25 40-42.5 42.5 4.33 7.43 17.58 42.5-45 45 3.86 6.90 14.24 45-47.5 47.5 4.69 9.66 25.85 47.5-50 50 4.94 15.12 44.30 50-51 51 1.54 17.54 36.00  160   Table H-7 Column test parameters to plot SWCC for Cu cone (6) Time (min) Drained water (cm3) Q (cm3/min) Volumetric water content (%) 0.5 560 1120 25.33 1 835 835 18.18 2 1025 512.5 13.24 3 1056 352 12.43 4 1076 269 11.91 5 1087 217.4 11.62 6 1095.5 182.6 11.40 7 1103.5 157.6 11.20 15 1134.5 75.6 10.39 25 1156.5 46.3 9.82 45 1180.5 26.2 9.19 59 1193.5 20.2 8.85 124 1217.5 9.8 8.23 179 1233.5 6.9 7.81 234 1242.5 5.3 7.58 314 1250.5 4 7.37 389 1257.5 3.2 7.19 1369 1280.5 0.9 6.59 1829 1285.5 0.7 6.46 3289 1292.5 0.4 6.28 4321 1293.5 0.3 6.25 5449 1298 0.24 6.14 8329 1298.5 0.16 6.12 9349 1298.5 0.14 6.12  161   Figure H-1 HYDRUS main page (Cu cone)   Figure H-2 HYDRUS main process (Cu cone)    162   Figure H-3 HYDRUS inverse modeling page (Cu cone)   Figure H-4 HYDRUS inverse modeling geometry information (Cu cone)     163   Figure H-5 HYDRUS inverse modeling time information (Cu cone)   Figure H-6 HYDRUS inverse modeling iteration criteria (Cu cone)     164    Figure H-7 HYDRUS inverse modeling soil hydraulic model (Cu cone)   Figure H-8 HYDRUS inverse modeling water flow parameters (Cu cone)     165   Figure H-9 HYDRUS inverse modeling water flow boundary conditions (Cu cone)   Figure H-10 HYDRUS inverse modeling data (Cu cone)    166   Figure H-11 HYDRUS inverse modeling profile information (Cu cone)   Figure H-12 HYDRUS inverse modeling pressure head versus water content (Cu cone)  167    Figure H-13 HYDRUS inverse modeling pressure head versus hydraulic conductivity (Cu cone)   Figure H-14 HYDRUS inverse modeling pressure head versus effective water content (Cu cone)  168   Appendix I: Information for mines using HPGR for Heap leaching Table I-1 List of mines with HPGR in heap leach operations Mine Country Process Ore type References  Tarkwa Ghana HPGR and heap leaching Au https://www.goldfields.co.za/pdf/investor_day/west_africa_region.pdf Soledad USA HPGR and heap leaching  Au and Ag http://www.goldenqueen.com/s/Projects_Overview.asp St Ives Australia HPGR and heap leaching (FS) Au  Lindero  Argentina HPGR and heap leaching (FS) Au and Cu http://www.goldrockmines.com/i/pdf/reports/2013-May-31-Lindero-43-101-Feasibility-Study-FINAL%20.pdf Volcan Chile HPGR and heap leaching (FS) Au http://www.marketwired.com/press-release/Andina-Timing-Update-for-Volcan-Gold-Project-Studies-TSX-VENTURE-ADM-1122011.htm Lobo Marte Chile HPGR and heap leaching (FS) Au http://www.infomine.com/companies-properties/reports/propertyreport.aspx?pid=14272&part=2  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0357256/manifest

Comment

Related Items