Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The use of clinoptilolite as permeable reactive barrier substrate for acid rock drainage Lai, Ranee Wan Man 2005-12-22

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-ubc_2005-105259.pdf [ 30.74MB ]
Metadata
JSON: 831-1.0063368.json
JSON-LD: 831-1.0063368-ld.json
RDF/XML (Pretty): 831-1.0063368-rdf.xml
RDF/JSON: 831-1.0063368-rdf.json
Turtle: 831-1.0063368-turtle.txt
N-Triples: 831-1.0063368-rdf-ntriples.txt
Original Record: 831-1.0063368-source.json
Full Text
831-1.0063368-fulltext.txt
Citation
831-1.0063368.ris

Full Text

THE USE OF CLINOPTILOLITE AS PERMEABLE REACTIVE BARRIER SUBSTRATE FOR ACID ROCK DRAINAGE by RANEE WAN MAN LAI B.A.Sc, The University of British Columbia, 1996 M.A.Sc, The University of British Columbia, 1998 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (^Civil Engineering^ THE UNIVERSITY OF BRITISH COLUMBIA August 2005 © Ranee Lai, 2005 ABSTRACT This study investigated the use of clinoptilolite as a permeable reactive barrier (PRB) substrate for retaining heavy metals from Acid Rock Drainage (ARD). PRBs are one of the options for retaining metals from ARD, allowing a cleaned groundwater plume to the receiving water bodies. The mineral clinoptilolite, a molecular sieve which has high cation exchange capacity, can retain heavy metals. Clinoptilolite is available in many locations and is inexpensive (~CDN$100-200/tonne). The suitability of clinoptilolite for the use as a PRB substrate was evaluated based on its chemical stability and metal retention in acidic environments. Results showed that clinoptilolite is chemically stable in ARD environment, the clinoptilolite structure is stable at pH >1.5. Clinoptilolite was found to retain 130.6mg Cu/kg soil (63.8% of Cu), 22.65mg Fe/kg soil (82.1% of Fe), 158mg Zn/kg soil (39.5% of Zn) and 215.4mg Al/kg soil (89.7% of Al) from the Britannia Mine natural ARD (pH 3.28) in batch equilibrium adsorption tests. Pretreatment of clinoptilolite with NaCl solution helped improve the retention of metals and reduced the leaching of Mn from the clinoptilolite. The performance of the clinoptilolite was evaluated using column leaching tests to simulate on-site conditions. Breakthrough curves were obtained at various flowrates and influent metal concentrations. Copper and zinc were the major contaminants of concern in the ARD. The breakthrough of copper occurs at 40 pore volumes (pv), manganese at 13 pv, zinc at 45 pv and aluminum at 38 pv, whereas iron precipitated once introduced to the leaching cell. Metal retention was found to be dependant on the flow rate. Selective extractions of metals on the clinoptilolite was conducted. Results ii indicated that partitioning was dependent on the flow conditions and the chemical characteristics of the leachate (pH and chemical composition). An algorithm was developed in the geochemical model PHREEQC for the design of the clinoptilolite barrier. The model helps ^to predict the performance and the transport of contaminants based on the amount of exchange sites, influent composition and concentration within the clinoptilolite PRB system, which are useful for estimating the service life and thickness required in the design of clinoptilolite PRB systems. iii TABLE OF CONTENTS ABSTRACT II TABLE OF CONTENTS IV LIST OF FIGURES VIILIST OF TABLES XACKNOWLEDGEMENTS XVII CHAPTER 1 INTRODUCTION 1 1.1 STATEMENT OF PROBLEMS1.2 RESEARCH OBJECTIVES 6 1.3 RESEARCH PLAN 7 1.4 RESEARCH CONTRIBUTIONS 10 1.5 ORGANIZATION OF THESIS 1 CHAPTER 2 LITERATURE REVIEW 3 2.1 ACID ROCK DRAINAGE 12.1.1 Generation of Acid Rock Drainage 16 2.1.2 Characteristics of ARD 25 2.2 BACKGROUND OF BRITANNIA MINE2.2.1 Geological characteristics at the Britannia Mine 27 2.3 PERMEABLE REACTIVE BARRIER SYSTEMS 30 2.3.1 Design requirements 32 2.3.2 Uses of permeable reactive barrier systems 35 2.4 PROPERTIES OF ZEOLITE 7 2.4.1 Minerological properties of Clinoptilolite 38 2.4.2 Stability in low pH 39 2.4.3 Adsorption capability 41 2.4.4 Surface Sorption Phenomena 2 2.4.5 Availability of the substrate material 49 2.5 GEOCHEMICAL MODELS 50 CHAPTER 3 MATERIALS AND METHODS 55 3.1 ARD SAMPLING3.2 LABORATORY INVESTIGATION OF ZEOLITE CLINOPTILOLITE 57 iv 3.2.1 Physical-chemical characteristics of zeolite clinoptilolite 57 3.2.2 Minerological characteristics of zeolite clinoptilolite 60 3.2.3 Adsorption capacity and chemical stability 62 3.2.4 Column leaching tests 67 3.2.5 Chemical analyses for clinoptilolite and water sample after column leaching tests ....74 Water sample analyses 73.3 APPLICATION OF THE GEOCHEMICAL MODEL 78 CHAPTER 4 CHARACTERISTICS OF CLINOPTILOLITE AND ACID ROCK DRAINAGE AND THEIR COMP AT ABILITY 80 4.1 CHARACTERISTICS OF BRITANNIA MINE ACID ROCK DRAINAGE 80 4.1.1 Chemical characteristics of ARD used in batch equilibrium adsorption tests 81 4.1.2 Chemical characteristics of ARD used in column-leaching tests 82 4.2 PHYSIO-CHEMICAL, PHYSICAL AND MINERALOGICAL PROPERTIES OF CLINOPTILOLITE 83 4.2.1 Physical-chemical properties 84.2.2 Mineralogical properties 4 4.3 PH STABILITY TESTS 90 4.3.1 Chemical stability4.3.2 Physical stability - hydraulic conductivity 93 4.3.3 Mineralogical changes 94 4.4 NEUTRALIZATION ABILITY OF CLINOPTILOLITE 96 CHAPTER 5 RETENTION OF METALS ON CLINOPTILOLITE 99 5.1 ADSORPTION OF METALS 95.1.1 Adsorption isotherms of metals on clinoptilolite 99 5.1.2 Adsorption of metals affected by pH .• 104 5.2 EFFECT OF RETENTION OF COPPER WITH CONTACT TIME 107 5.2.1 Effects of mixing in batch adsorption tests on the retention of copper 108 5.2.2 Effects of flow rates in mini-column leaching cell tests on the retention of copper ..109 5.2.3 Effects of copper concentration in column leaching tests on the adsorption onto clinoptilolite Ill 5.2.4 Summary of comparison of the retention of copper onto clinoptilolite in different testing methods 112 CHAPTER 6 COLUMN LEACHING CELL TESTS 115 6.1 COLUMN LEACHING TESTS (I) - BULK CLINOPTILOLITE 11v 6.2 COLUMN LEACHING TESTS (II) - PRETREATED CLINOPTILOLITE 119 6.2.1 Leachate analyses 116.2.2 Analysis of clinoptilolite - post-leaching 139 6.3 COLUMN LEACHING TESTS (III) - CHANGES IN CLINOPTILOLITE WITH TIME 155 6.3.1 Leachate analyses 156.3.2 Analyses of the clinoptilolite 166 6.4 DESIGN OF CLINOPTILOLITE PERMEABLE REACTIVE BARRIER SYSTEM 179 6.4.1 Conceptual design of the clinoptilolite permeable reactive barrier 179 6.4.2 Recognition of time frame 180 6.4.3 Design parameters 1 6.4.4 Application of clinoptilolite PRB in a hypothetical ARD scenario 184 6.4.5 Design recommendations 190 CHAPTER 7 GEOCHEMICAL MODEL 2 7.1 DATA COLLECTION 197.1.1 Characterization of solid media 192 7.1.2 Characterization of groundwater 3 7.1.3 Assumptions in the model 197.2 CONSTRUCTION OF THE MODEL ALGORITHM 194 7.2.1 Components included in the model7.2.2 Types of geochemical reactions included in the model 195 7.2.3 Selection of model thermodynamic and kinetic data 197 7.3 SCRIPT DEVELOPMENT OF THE CONCEPTUAL MODEL 197.4 INTERPRETATION OF MODELING RESULTS 200 7.5 MODEL PREDICTIONS 202 7.6 LIMITATIONS AND POSSIBLE SOURCES OF ERRORS 203 CHAPTER 8 CONCLUSIONS AND RECOMMENDATIONS 205 8.1 CONCLUSIONS 208.2 RECOMMENDATIONS FOR FURTHER RESEARCH 209 REFERENCES 212 APPENDICES 8 Appendix A Methods of chemical analyses A-l Selective sequential extractions on tested clinoptilolite 220 A-2 Water sample analyses 222 vi Appendix B Chemical compositions of ARD and residue B-l Chemical composition of ARD 224 B-2 Chemical composition of residues from ARD filtration 225 B-3 Comparison of chemical compositions between filtered and non-filtered mine drainage from the 4150 Portal 226 Appendix C Mineralogical characteristics of clinoptilolite C-l Investigations of the X-Ray Diffractorgrams for semi-quantitative analysis of the mineral composition of clinoptilolite 227 C-2 Major elements in the clinoptilolite sample by X-Ray Fluorescence (XRF) and Inductive Coupled Plasma (ICP) spectroscopy 239 C-3 Compaction curve of clinoptilolite 242 C-4 Sample calculation for the compaction of material and its compaction properties . 243 Appendix D Langmuir Isotherm and Langmuir constant determination 245 D-l Adsorption isotherm of metals (Cu, Fe, Mn, Zn, Al) on clinoptilolite in single species solutions 248 D-2 Adsorption isotherm of metals (Cu, Fe, Mn, Zn, Al) on clinoptilolite in multiple species solutions - Distilled water with nitric acid as background solution 253 D-3 Adsorption isotherm of metals (Cu, Fe, Mn, Zn, Al) on clinoptilolite in multiple species solutions - Natural acid rock drainage as background solution 261 Appendix E Leaching of metals from clinoptilolite 266 Appendix F Determination of pretreatment methods 8 Appendix G Column leaching cell tests water sample analyses results 270 Appendix H Column leaching cell tests soil analyses results H-l Results of metal fractionations on soil samples extracted from column leaching cells 332 H-2 Numerical and graphical presentation of metal fractionations on soil samples after deducting background metal concentrations 350 Appendix I Results of water sample analyses from Critical Path Analysis column leaching cells 373 Appendix J Results of soil analyses on tested clinoptilolite from Critical Path Analysis leaching column cells J-l Results of metal partitioning on clinoptilolite extracted from leaching cell after 15, 26, 38, 45, 57, 73, 85, and 96 pore volumes 406 J-2 Numerical and graphical presentation of metal partitioning on clinoptilolite after subtracting background concentration from IM treated clinoptilolite 415 J-3 Changes of the amount and percentage partitioning of metals (Cu, Fe, Mn, Zn, Al) 432 Appendix K Geochemical model development K-l Description of mode codes 438 K-2 Model output 44vii LIST OF FIGURES Figure 2.1 Bacterial activities of Thiobacillus ferrooxidans 17 Figure 2.2 Effect of temperature on biological and chemical oxidation rate 19 Figure 2.3 Effect of pH on biological and chemical sulphide-oxidation rate 20 Figure 2.4 Stages in the formation of acid rock drainage (Adapted from Guidelines for ARD prediction in the North, Indian and Northern Affairs Canada, 1993) 23 Figure 2.5 Regional location map of the Britannia Mine site 26 Figure 2.6 Regional map showing sampling sites at the Britannia Mine (adapted from Price etal, 1995) 28 Figure 2.7 Interior of the adit at the Britannia Mine 29 Figure 2.8 Water containing high iron precipitates ponding at downstream of 4100 Portal, Britannia Mine site 2Figure 2.11 The diffuse-ion layer (Adapted from Mohamed and Antia,1998) 43 Figure 2.12 An illustration of metal ion sorption on (hydr)oxide (Adapted from Sparks, 1995) 47 Figure 3.1 Sampling location at the 4150 Portal, Britannia Mine 55 Figure 3.2 Sampling location at the 4100 Portal, Britannia Mine 6 Figure 3.3 Setup of the variable-head permeameter 60 Figure 3.4 The work plan of the retention kinetics study of copper onto clinoptilolite 64 Figure 3.5 Schematic diagram of the procedures for non-mixing batch adsorption tests for retention kinetics 66 Figure 3.6 Schematic diagram of the experimental setup of mini-column-leaching cell 67 Figure 3.7 Setup of mini-column leaching cell 6Figure 3.8 Configuration of column-leaching cells 9 Figure 3.9 Assemblage at the end-capsFigure 3.10 Setup of column leaching cells 70 Figure 3.11 Schematic diagram of the column leaching cell 7Figure 4.1 Diagrams illustrating layer-weathering and edge-weathering of micas by exchanging interlayer potassium with hydrated exchangeable cations. Mica particles would normally be much wider in relation to thickness than the zone represented in the diagram. (Adapted from Dixon and Weed, 1977) 86 Figure 4.2 Scanning electron microscopic image of bulk clinoptilolite (particle size 2-50 Hm) 9Figure 4.3 Leaching of ions from clinoptilolite acid treatment at various pHs with time: (a) Amount of Al ions leached; (b) Amount of K ions leached 91 Figure 4.4 The leaching of Si and Na relative to Al in multiple species solutions with background matrix of: (a) nitric acid, pH 2; (b) natural ARD, pH 3.28 92 Figure 4.5 Scanning electron microscopic (SEM) diagrams of clinoptilolite with particle size ranges 2 to 50 um: (a) after water treatment; (b) after ARD treatment 96 Figure 4.6 The changes of pH in the ARD after 24-hour exposure of clinoptilolite to ARD of various pH levels 9Figure 5.1 Adsorption of metal ions in multiple Cu, Al, Fe and Zn species on clinoptilolite with natural ARD (pH 3.28) as background solution 102 Figure 5.2 Multiple Cu, Al, Fe and Zn species adsorption on clinoptilolite with background solution of nitric acid/distilled water at pH 2 10Figure 5.3 The leaching of metals from clinoptilolite in 24-hr batch equilibrium tests with natural ARD at various pH level: (a) overview of the adsorption curve, (b) zoomed-in view of Figure 5.3 a at the metal leaching to adsorption transition 104 Figure 5.4 Comparison of the retention of Cu onto clinoptilolite between mixing and non-mixing batch tests 108 Figure 5.5 Comparison of the retention of Cu onto clinoptilolite at different flow rates in mini-column leaching tests 110 Figure 5.6 Comparison of the retention of Cu onto clinoptilolite in mini-column-leaching tests with different solute concentrations 111 Figure 5.7 Adsorption of Cu with time on clinoptilolite in various testing conditions 113 Figure 6.1 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell with bulk clinoptilolite, non-filtered natural ARD, and flow rate of 1 m/day. (Column ID: NTNFC1) 118 Figure 6.2 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; and (h) concentration of Zn, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, non-filtered natural ARD, and flow rate of 1.5 m/day. (Column ID: PTB1C1) 120 Figure 6.3 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, filtered ARD of ix maximum design concentration, and flow rate of 1.5 m/day. (Column ID: PTB3) 121 Figure 6.4 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, filtered ARD of maximum design concentration, and flow rate of 1.5 m/day. (Column ID: PTB4) 122 Figure 6.5 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, filtered natural ARD, and flow rate of 1.5 m/day. (Column ID: PTB5) 123 Figure 6.6 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, filtered natural ARD, and flow rate of 1.5 m/day. (Column ID: PTB6) 124 Figure 6.7 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, filtered natural ARD, and flow rate of 0.5 m/day. (Column ID: PTB8) : 125 Figure 6.8 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, filtered natural ARD, and flow rate of 0.5 m/day. (Column ID: PTB9) 126 Figure 6.9 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) x concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, filtered ARD with maximum design concentration, and flow rate of 0.5 m/day. (Column ID: PTB11) 127 Figure 6.10 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, filtered ARD with maximum design concentration, and flow rate of 0.5 m/day. (Column ID: PTB12) 128 Figure 6.11 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, non-filtered ARD with maximum design concentration, and flow rate of 0.5 m/day. (Column ID: PTB13) 129 Figure 6.12 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, non-filtered ARD with maximum design concentration, and flow rate of 0.5 m/day. (Column ID: PTB14) 130 Figure 6.13 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 10cm; (k) 20 cm; (1) 30 cm; (m) 40 cm; and (n) 50 cm for leaching cell IM NaCl treated clinoptilolite, non-filtered natural ARD, and flow rate of 0.5 m/day. (Column ID: PTB15) 131 Figure 6.14 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 30 cm; (k) 40 cm; and (1) 50 cm for xi leaching cell IM NaCl treated clinoptilolite, filtered natural ARD, and flow rate of 0.5 m/day in a packed bed system. (Column ID: PTB17) 132 Figure 6.15 The change in chemical characteristics in water sample from sampling ports along the leaching cell with pore volumes for: (a) pH; (b) redox potential, mV; (c) specific electrical conductivity, mS/cm; (d) pH and pE at outlet; (e) concentration of Cu, mg/L; (f) concentration of Fe, mg/L; (g) concentration of Mn, mg/L; (h) concentration of Zn, mg/L; and (i) concentration of Al, mg/L. Breakthrough curves for metals at: (j) 30 cm; (k) 40 cm; and (1) 50 cm for leaching cell IM NaCl treated clinoptilolite, non-filtered natural ARD, and flow rate of 0.5 m/day in a packed bed system. (Column ID: PTB18) 133 Figure 6.16 Concentration of Mn in the water sample in (a) bulk clinoptilolite - without pretreatment, (b) clinoptilolite after IM NaCl pretreatment 134 Figure 6.17 pH of the water sample in (a) bulk clinoptilolite - without pretreatment, (b) clinoptilolite after IM NaCl pretreatment 135 Figure 6.18 Comparison of breakthrough volume with various ARD concentrations in columns of flow rates: (a) 0.5 m/day; (b) 1.5 m/day 138 Figure 6.20 Partitioning of Cu in the clinoptilolite at equilibrium from column leaching cell tests with conditions: (a) non-filtered natural ARD, 1 m/day, bulk clinoptilolite; treated clinoptilolite with (b) non-filtered natural ARD, 1.5 m/day, (c) filtered maximum design ARD at 1.5 m/day, (d) filtered natural ARD at 1.5 m/day, (e) filtered natural ARD at 0.5 m/day, (f) filtered maximum design ARD at 0.5 m/day, (g) non-filtered maximum design ARD at 0.5 m/day, (h) non-filtered natural ARD at 0.5 m/day 143 Figure 6.21 Partitioning of Fe in the clinoptilolite at equilibrium from column leaching cell tests with conditions: (a) non-filtered natural ARD, 1 m/day, bulk clinoptilolite; treated clinoptilolite with (b) non-filtered natural ARD, 1.5 m/day, (c) filtered maximum design ARD at 1.5 m/day, (d) filtered natural ARD at 1.5 m/day, (e) filtered natural ARD at 0.5 m/day, (f) filtered maximum design ARD at 0.5 m/day, (g) non-filtered maximum design ARD at 0.5 m/day, (h) non-filtered natural ARD at 0.5 m/day 144 Figure 6.22 Partitioning of Mn in the clinoptilolite at equilibrium from column leaching cell tests with conditions: (a) non-filtered natural ARD, 1 m/day, bulk clinoptilolite; treated clinoptilolite with (b) non-filtered natural ARD, 1.5 m/day, (c) filtered maximum design ARD at 1.5 m/day, (d) filtered natural ARD at 1.5 m/day, (e) filtered natural ARD at 0.5 m/day, (f) filtered maximum design ARD at 0.5 m/day, (g) non-filtered maximum design ARD at 0.5 m/day, (h) non-filtered natural ARD at 0.5 m/day 145 Figure 6.23 Partitioning of Zn in the clinoptilolite at equilibrium from column leaching cell tests with conditions: (a) non-filtered natural ARD, 1 m/day, bulk clinoptilolite; treated clinoptilolite with (b) non-filtered natural ARD, 1.5 m/day, (c) filtered maximum design ARD at 1.5 m/day, (d) filtered natural ARD at 1.5 m/day, (e) filtered natural ARD at 0.5 m/day, (f) filtered maximum design ARD at 0.5 m/day, (g) non-filtered maximum design ARD at 0.5 m/day, (h) non-filtered natural ARD at 0.5 m/day 146 Figure 6.24 Partitioning of Al in the clinoptilolite at equilibrium from column leaching cell tests with conditions: (a) non-filtered natural ARD, 1 m/day, bulk clinoptilolite; treated clinoptilolite with (b) non-filtered natural ARD, 1.5 m/day, (c) filtered maximum design ARD at 1.5 m/day, (d) filtered natural ARD at 1.5 m/day, (e) filtered natural ARD at 0.5 m/day, (f) filtered maximum design ARD at 0.5 m/day, (g) non-filtered maximum design ARD at 0.5 m/day, (h) non-filtered natural ARD at 0.5 m/day 147 Figure 6.25 Comparison of retention of Cu in column leaching cells in adsorption isotherms 151 Figure 6.26 pH of water samples from sampling ports along the column leaching cells up to pore volumes (pv) of: (a) 15 pv, (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 158 Figure 6.27 Redox potential (expressed in millivolts, mV) of water samples from sampling ports along the column leaching cells up to pore volumes (pv) of: (a) 15 pv, (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 159 Figure 6.28 Electrical conductivity of water samples from sampling ports along the column leaching cells up to pore volumes (pv) of: (a) 15 pv, (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 160 Figure 6.29 Concentration of Cu in water samples from sampling ports along the column leaching cells up to pore volumes (pv) of: (a) 15 pv, (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 161 Figure 6.30 Concentration of Fe in water samples from sampling ports along the column leaching cells up to pore volumes (pv) of: (a) 15 pv, (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 162 Figure 6.31 Concentration of Mn in water samples from sampling ports along the column leaching cells up to pore volumes (pv) of: (a) 15 pv, (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 163 Figure 6.32 Concentration of Zn in water samples from sampling ports along the column leaching cells up to pore volumes (pv) of: (a) 15 pv, (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 164 Figure 6.33 Concentration of Al in water samples from sampling ports along the column leaching cells up to pore volumes (pv) of: (a) 15 pv, (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 165 Figure 6.34 Partitioning of Cu on the clinoptilolite at different depths along the column leaching cells with natural non-filtered ARD and flowrate of 0.5 m/day with IM NaCl pretreated clinoptilolite after: (a) 15 pore volumes (pv), (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 167 Figure 6.35 Partitioning of Fe on the clinoptilolite at different depths along the column leaching cells with natural non-filtered ARD and flowrate of 0.5 m/day with IM NaCl pretreated clinoptilolite after: (a) 15 pore volumes (pv), (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 168 Figure 6.36 Partitioning of Mn on the clinoptilolite at different depths along the column leaching cells with natural non-filtered ARD and flowrate of 0.5 m/day with IM NaCl pretreated clinoptilolite after: (a) 15 pore volumes (pv), (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 169 Figure 6.37 Partitioning of Zn on the clinoptilolite at different depths along the column leaching cells with natural non-filtered ARD and flowrate of 0.5 m/day with IM NaCl pretreated clinoptilolite after: (a) 15 pore volumes (pv), (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 170 Figure 6.38 Partitioning of Al on the clinoptilolite at different depths along the column leaching cells with natural non-filtered ARD and flowrate of 0.5 m/day with IM NaCl pretreated clinoptilolite after: (a) 15 pore volumes (pv), (b) 26 pv, (c) 38 pv, (d) 45 pv, (e) 57 pv, (f) 73 pv, (g) 85 pv, and (h) 96 pv 171 Figure 6.39 Correlation between the total amount of Cu retained on clinoptilolite with pore volumes 172 Figure 6.40 Percentages of partitioning of Cu in the form of exchangeable, ionically bound to amorphous materials, bound to Fe- & Mn- oxides, oxidizable compounds and residual fraction along the length of the column leaching cell 173 Figure 6.41 Percentage of partitioning of Fe in the form of exchangeable, ionically bound to amorphous materials, bound to Fe- & Mn- oxides, oxidizable compounds and residual fraction along the length of the column leaching cell 174 Figure 6.42 Percentage of partitioning of Mn in the form of exchangeable, ionically bound to amorphous materials, bound to Fe- & Mn- oxides, oxidizing compounds and residual fraction along the length of the column leaching cell 175 Figure 6.43 Correlations between the sum of amount of Zn from SSE with pore volumes 176 Figure 6.44 Percentage of partitioning of Zn in the form of exchangeable, ionically bound to amorphous materials, bound to Fe- & Mn- oxdies, oxidizable compounds and residual fraction along the length of the column leaching cell 177 Figure 6.45 Percentage of partitioning of Al in the form of exchangeable, ionically bound to amorphous materials, bound to Fe- & Mn- oxides, oxidizable compounds and residual fraction along the length of the column leaching cell 178 Figure 6.46 Correlations between mass loading and the amount of metals on soils 181 Figure 6.47 Relationship of the amount of metals retained on clinoptilolite to the metal concentrations in the effluent in column leaching cells 182 Figure 6.48 Illustration of the hypothetical problematic ARD scenario 185 Figure 6.49 Configuration of a clinoptilolite PRB system in a hypothetical problematic ARD scenario 186 Figure 7.1 Model predictions for 0.5 m/day, non-filtered ARD leaching column with treated clinoptilolite 200 xiv LIST OF TABLES Table 2.1 Composition of mine drainage water from various types of mines (Ripley et al., 1978) 25 Table 2.2 The range in water quality, flow and discharge data for 2200 and 4100 discharges, and for Jane and Britannia Creeks (adapted from Price et al., 1995) 30 Table 2.3 Considerations in designing permeable reactive barrier systems 34 Table 2.4 Previous uses of permeable reactive barriers in metal contaminants removal 35 Table 2.5 Major elemental compositions from ICP analysis on clinoptilolite (Beck, 1996) 39 Table 2.6 Factors that could affect retention of metals and the mechanisms 42 Table 2.7 Functions of popular geochemical modeling programs (Source: Zhu and Anderson, 2002) 5Table 3.1 Column leaching tests conducted and the design parameters 72 Table 3.2 Concentrations of metals in ARD of maximum design concentrations 73 Table 3.3 Difference in chemical composition between filtered and non-filtered ARD 73 Table 3.4 Reaction time of leaching cells for critical path analysis 74 Table 3.5 Reagents in selective sequential extractions and their purposes 77 Table 4.1 Chemical characteristics of ARD from the 4150 and 4100 Portal at the Britannia Mine,B.C 80 Table 4.2 Major chemical composition and differences between natural (non-filtered) and filtered acid -rock drainage obtained from the 4150 Portal 81 Table 4.3 Concentrations of metal ions (mg/L) in natural ARD (pH 3.28) 82 Table 4.4 Chemical characteristics of the 4150 and the "adjusted" 4100 drainages 82 Table 4.5 Physio-chemical properties of clinoptilolite 83 Table 4.6 Trace minerals associated with the clinoptilolite sample 84 Table 4.7 Semi-quantitative analysis of the composition of major minerals in clinoptilolite from X-Ray diffraction 85 Table 4.8 Major elements in the clinoptilolite sample by X-Ray Fluorescence (XRF) and Inductive Coupled Plasma (ICP) Spectroscopy 88 Table 4.9 Hydraulic conductivity of clinoptilolite with various permeates 93 Table 5.1 Constant related to binding strength, k [L/mg], from Langmuir equation obtained in batch equilibirum adsorption tests in single and multiple-species solutions 100 Table 5.2 Percentage removal of metals in the natural Britannia copper mine drainage (pH 3.28) by the clinoptilolite (24-hour batch test) 105 Table 5.3 Ionization energies and ionic radii of Al , Fe , Cu and Zn ions 106 Table 5.4 Comparison of retention of Cu onto clinoptilolite using different methods of treatment 113 xv Table 6.1 Summary of the design parameters of the column leaching cell tests and the chemical characteristics of the leachates at equilibrium 134 Table 6.2 Comparison of breakthrough volume at different flow rates 136 Table 6.3 Comparison of breakthrough volume between the filtered and non-filtered ARD at 0.5 m/day 139 Table 6.4 Summary of the partitioning of Cu expressed in % of the total from SSE in each fraction 148 Table 6.5 Summary of the partitioning of Zn expressed in % of the total amount obtained from SSE in each fraction 153 Table 6.6 Empirical relationships between the sum of the amount of Cu on clinoptilolite with the amount of leachate reacted 172 Table 6.7 Empirical relationships between the sum of the amount of Zn on clinoptilolite with the amount of leachate reacted 176 Table 6.8 Empirical relationships between the total amount of Cu and Zn retained on clinoptilolite with mass loading in the column leaching cell 182 Table 6.9 Costs associated with a clinoptilolite PRB in the hypothetical ARD scenario 187 Table 6.10 Costs associated with a water treatment facility for the hypothetical ARD scenario 188 Table 7.1 PHREEQC script to calculate the concentration of metals from outlet of a clinoptilolite leaching cell 19Table 7.2 Comparison of the breakthrough volumes from geochemical model and experimental results 202 xvi ACKNOWLEDGEMENTS I would like to express my sincerest gratitude to individuals and organizations in completing this dissertation. First, I gratefully thank Dr. Loretta Li, for all her patience and support throughout the duration of this research and for her motivation and encouragement for the author in achieving tasks in the last few years. I would like to extend my thanks to Dr. Li for her guidance and critical review of the thesis. My research committee deserves a special note insofar as they have always prompted me to see my research problems in ways that were not immediately obvious to me. I wish to thank Df. Lavkulich, who has provided me tremendous support and advice on technical and theoretical issues, provided me encouragement and confidence in completing this research. I wish to express my sincere appreciation and thanks to Professor Atwater, who has never given up on providing me the most valuable feedback and considerations in practical issues which I would never have thought about. I would like to thank Dr. Ken Hall, for his patience, teaching and inspirational commentary on water chemistry topics; nevertheless, apart from the research, little knowledge about birds. I again would like to thank my research committee for giving me the chance to make everything possible! I will remain grateful for all their support, guidance and encouragement. I would like to express my gratitude to Ms. Susan Harper and Ms. Paula Parkinson, in turning an ugly untidy little duck in the environmental lab into a clean, organized and responsible individual in these few years. With their support, advice and guidance, my laboratory work could not be any better. Thanks to Mr. Harald Schempp and Mr. Bill Leung in the machine shop for helping me getting every piece of equipment built, repaired, and working in no time. I would like to send my best friend, Ms. Linlin Hui, my most gracious thanks for her encouragement, and for the many discussions and long hours working together in the laboratory. With her guidance in the laboratory work in the early stage of my program, I have gained a lot of priceless hints in making things work better. I would like to thank Ms. Joan Liu, Mr. Ken Pun, Mr. Andrew Wu, Mr. Juan Garfias, and (Dr.-to-be) Humberto Preciado for their assistance in the laboratory work, sampling outdoor in cold, hot, rain and sun, peer support and valuable discussions from the view of an outsider. I would like to thank the Anaconda Britannia Mine, B.C., and Canmark Resources Inc. for generous provision of the necessary materials required for completing the experiments. I would also like to thank the United States Geological Survey in providing the geochemical software for free of charge and Dr. Tony Appelo for his valuable feedbacks in my geochemical model. I would like to thank my mother, for her endless support, emotionally and financially, throughout these years, I would not be able to make it through without her help. Thank you mom for your remarkable patience, understanding and encouragement. I wish I could give you a long awaiting birthday gift this year - the purple hat - ©. I wish to thank my sister and brother, Winky and Jonny, for their continuous encouragements in the last few years. I thank all my friends, for their continuous perpetual encouragements, for their patience listening, valuable advices on everything, keeping the faith in me and keeping my spirits high. Thank you for making my dream comes true! Ranee July 2003 xvii CHAPTER 1 INTRODUCTION 1.1 Statement of Problems Acid rock drainage (ARD) is considered to be one of the most serious environmental problems faced by the mining industry. Acid rock drainage occurs as a result of the natural oxidation of sulphide minerals contained in rock that is exposed to air and water. Subsequently, sulphuric acid is produced, lowering the pH of the contaminated drainage. In addition, at a certain pH level (~ pH 3.5), the process is accelerated by acid-generating bacteria, mainly Thiobacillus ferrooxidans and Leptospirillum ferrooxidans, which further lower the pH of the leachate from the exposed sulphide minerals and further mobilize and increase the metals in the drainage. Depending on geological factors and the mine site, the metals in the mining waste may include aluminum, cobalt, copper, cadmium, chromium, gold, iron, lead, manganese, silver and zinc. The pH of ARD may be as low as 2 in some geological environments, where there is no buffering capacity from other secondary minerals. At least 26 of the 1809 abandoned mining sites in British Columbia, Canada, have been identified as acid-generating and 20 additional sites have been identified as potentially acid-generating (EMCBC 1997). With high metal concentrations and low pH, ARD impacts the quality of the environment and the health of organisms in receiving water bodies. Remediation options for the ARD include active treatment such as neutralizing the mine drainage with lime in a treatment plant, eliminating mobile cations in the effluent by precipitation, and passive treatment systems such as wetland and permeable reactive barriers. However, the cost for active treatment can be enormous and might not be feasible for ARD that contains metal concentrations too low for metal recovery but high enough to be toxic to aquatic life and to exceed the criteria for effluent disposal. Reactive barrier systems may be a suitable remediation option for areas where low metal concentrations and flow of ARD occur. Since ARD is characterized by high 1 acidity and metal concentrations, it is necessary that the reactive barrier material be able to attenuate the metal cations from the drainage and be able to neutralize the ARD plume as it passes through the barrier. In addition, the costs of using reactive barrier systems are lower than active treatment facilities due to the minimal maintenance (labor and consumables) required. Zeolite has been found to have potential as an effective metal exchange medium in wastewater treatment (Blowes and Ptacek 1992; Zamzow and Murphy 1992; Robertson and Cherry 1995; Singer and Berkgaut 1995; Baker, Blowes et al. 1997; Vaca Mier, Lopez Callejas et al. 2001), acid rock drainage treatment (Vos 1992; Waybrant, Blowes et al. 1995; Beck 1996; Benner, Blowes et al. 1997; Gilbert, O'Meara et al. 1999) and as landfill liners (Singer and Berkgaut 1995). As a result, and since it is relatively inexpensive, it has the potential to be a good candidate as a reactive barrier material. However, no studies have been done to address the chemical integrity and multi-component attenuation of metal ions on clinoptilolite in low pH environments. There is also a lack of information on the hydraulic conductivity, partitioning of metals, and factors affecting the attenuation process of clinoptilolite in permeable reactive barrier systems. Zeolites are tectosilicates consisting of three-dimensional frameworks of SiO/" tetrahedral units, wherein all the oxygen atoms (O) of each tetrahedron are shared with adjacent tetrahedra. The most important characteristic of zeolites, from a water treatment perspective, is its net negatively charged structure, which results from isomorphic substitution of the Al3+ ion for Si4+ ion of the tetrahedra (Sherry 1971; Boles, Flanigen et al. 1977; Dixon and Weed 1989). Zeolites are also characterized by large void volumes (20% to 50%) and large internal surface areas of several hundred thousand square meters per kilogram, (Boles, Flanigen et al. 1977; Gottardi and Galli 1985; Dyer 1988; Dixon and Weed 1989) making it a good adsorbent. The other unique feature of zeolite is its ability to hydrate and dehydrate reversibly, and to exchange constituent cations without disrupting bonds (Dyer 1988). Each zeolite has its own arrangement and number 2 of channels, providing unique molecular sieve properties. Thus zeolites adsorb cations selectively according to the size and speciation of the cations. Studies have been conducted using zeolite to remove and/or exchange metals with metals-contaminated water (Zamzow and Murphy 1992; Jacobs and Forstner 1999; Vaca Mier, Lopez Callejas et al. 2001). A study was done using zeolite amended with flyash in the supernatant for industrial wastewater treatment containing lead and cadmium (Singer and Berkgaut 1995). It was shown that zeolited flyash displayed high selectivity for Pb2+>Sr2+>Cr2+>Cd2+>Zn2+>Cs+ in competition with Na+, especially at low concentration of these ions. Furthermore, the zeolite was effective in removing Pb2+ and Zn2+ from industrial wastewater of pH 7 (Singer and Berkgaut 1995). Zamzow & Murphy (1992) found that clinoptilolite is efficient in taking up metal cations in copper-mine wastewater such as Al3+, Cd2+, Cu2+, Fe3+, Ni2+, Pb2+ and Zn2+, with relatively more Al3+ and Fe3+ removed than the other cations. This result is consistent with ion-exchange theory (Semmens and Seyfarth 1978; Yuan 1999), which states that the negative sites on crystal lattices have the highest affinity for trivalent ions, followed by divalent and monovalent ions respectively. In case of a series of metal cations of a given valence, the preference is correlated with the molecular size and the ionic radius of the cations. Zamzow & Murphy (1992) reported that Ca2+ and NFLt4" in wastewater interfered with the uptake of heavy metals, especially Ni2+. Since sodium (Na+) is the main exchangeable cation in the clinoptilolite, Zamzow & Murphy (1992) found that when the concentrations of metals are very high in the leachate, the amount of Na+ exchanged from the zeolite would start to build up to such a high level that no more ion exchange can take place. In that case, metal contaminants were not completely removed using the ion exchange sites on the zeolite alone. Previous studies have shown that zeolite clinoptilolite has high retention capacity for heavy metals in ARD (Vos 1992; Gilbert, O'Meara et al. 1999). Vos (1992) conducted a study of 3 zeolite for treatment of ARD from the Britannia mine site, and recent research by Gilbert et al. (1999) using clinoptilolite in ARD remediation found that clinoptilolite is effective in removing metal cations from ARD. In a study by the United States Geological Survey (USGS) (Gilbert, O'Meara et al. 1999) on the adsorption capabilities of clinoptilolite in respect to mine drainage (pH 2-3), slight (120 u.g of Al per gram of soil) dissociation of Al from the clinoptilolite-rich rock was observed. The leaching of Al ions from the clinoptilolite was not investigated further. An investigation has been conducted on the zeolite used in this study for the treatment of ARD from the Equity Silver Mine, B.C. (Beck 1996). It was found that the zeolite was effective in reducing the metals in the ARD (copper by 68%, zinc by 49%, arsenic by 97% and iron by 93%). However, test results on the adsorption were limited to the first 18 hours of reaction through a batch test in a static column, which may not represent the equilibrium condition on the zeolite. The stability of the zeolite, which may be affected by the acidity of the environment, was not investigated. Zeolite clinoptilolite was used in this study. Clinoptilolite is a sub-category of the mineral zeolite. It is also the most abundantly found zeolite. Clinoptilolite is characterized by its high cation-exchange capacity, which is anticipated to have high-metal attenuation. Clinoptilolite also has the highest Si/Al ratio among all the other groups of zeolites, with Si/Al ratios ranging from 4.3 to 5.3 (Breck 1974; Meier and Olson 1978; Dixon and Weed 1989). Zeolites, which possess higher Si/Al ratios, have an advantage that when placed in low pH environments, its structure is less easily destroyed by the loss of Al from the framework to the environment (Boles, Flanigen et al. 1977; Gottardi and Galli 1985; Dyer 1988; Dixon and Weed 1989). The cage-like structure of zeolite clinoptilolite also allows it to have a high hydraulic conductivity compared to other minerals. Clinoptilolite is found abundantly in sedimentary deposits (saline, alkaline lakes, deep-sea sediments, and low-temperature tephra systems). It can often be obtained at a reasonable cost 4 (Hay and Sheppard 2001; Dobelin and Armbruster 2003). An open-system zeolite clinoptilolite is found in Asp Creek and Sunday Creek in B.C.; Nye County in Nevada; John Day Formation in Oregon; Death Valley Junction in California, USA; and Yellow tuffs near Naples in Italy (Sheppard and Simandl 1999). The cost of the zeolite clinoptilolite, depending on the particle size and the quality of the clinoptilolite, ranges from CDN$ 100-200 per tonne. The formation of clinoptilolite in alkaline environments and its high pH may contribute in neutralizing the ARD through the course of treatment. The abundant availability of zeolite clinoptilonite in British Columbia and many parts of the world enable it to be easily accessible for use in permeable reactive barrier applications. ARD is a serious global problem and a local issue in B.C., and the extent of the problem requires solutions to be economically feasible. In search of a suitable substrate for a permeable reactive barrier system, other than the chemical compatibility, attenuation ability for contaminants and its service life, it is also important for economic reasons that the barrier substrate be locally available. This is necessary because a large amount of substrate material would be needed in constructing a reactive barrier system. In view of the above criteria, the research focused on evaluating the performance, in terms of metal-attenuation ability and chemical stability of clinoptilolite as a permeable reactive barrier substract for ARD treatment. The results obtained from experiments are used to calibrate the parameters used in the geochemical model PHREEQC (Parkhurst and Appelo 1999; Parkhurst and Appelo 2003). The geochemical model will be used to predict the retention of metal ions on clinoptilolite, the breakthrough time, and hence the service life and thickness of the permeable reactive barrier. Acid rock drainage from the Anaconda Britannia Mine was used in this study for the treatability of metals on the clinoptilolite. Mine drainage from the Britannia Mine was chosen because it is the most serious acid-generating site in B.C.(EMCBC 1997); it is also representative of typical chemical characteristics in acidic copper mine drainages. However, no study has been 5 conducted using clinoptilolite as a permeable reactive barrier substrate for removing metal ions from the ARD of the Britannia Mine. In this study, clinoptilolite was first evaluated for its chemical stability in acidic environment at pHs of 0.5-3.0 and its metal attenuation ability using batch equilibrium adsorption tests. Once the suitability of the clinoptilolite was confirmed, clinoptilolite was tested for its performance in column leaching tests, which simulate the reactive wall situations using ARD from the Britannia Mine. Column leaching tests were conducted at various flow rates and concentrations of metals. Results from the column leaching tests provide information for the concentration profiles and breakthrough curves on the clinoptilolite. Partitioning of metals on clinoptilolite was determined by selective sequential extractions. A geochemical model was developed for predicting the breakthrough volumes of metals on the clinoptilolite. 1.2 Research objectives The overall aim of this research is to explore the performance of clinoptilolite as a permeable reactive barrier substrate for acid-rock drainage. The specific objectives of this research are as follow: • To evaluate the suitability of clinoptilolite as a reactive barrier substrate for acid rock drainage by examining its chemical stability (in terms of leaching of ions from the framework of the clinoptilolite structure) in acidic conditions commonly found in ARD environments; • To examine the interactions and partitioning of heavy metals from ARD on the zeolite-clinoptilonite by conducting batch equilibrium adsorption tests and determination of metal adsorptivities; • To determine the performance of the clinoptilolite as reactive barrier substrate in column leaching tests to simulate the on-site flow conditions. Column leaching tests 6 were conducted with the Britannia Mine ARD at natural and maximum design concentrations, and at various flow rates; • To determine if pre-treatment of the clinoptilolite would improve effectiveness; • To predict the transport, retention ability, and partitioning of heavy metals on the clinoptilolite in the barrier system using an existing geochemical model. Such a model is compared against experimental data and the model will be modified to meet the need for the stated application. Design requirements such as flow rates and influent concentrations will be defined in the algorithm to predict the performance of the clinoptilolite permeable reactive barrier. • To provide design recommendations for engineers in using clinoptilolite as a permeable reactive barrier substrate. 1.3 Research Plan To illustrate the objectives in this research, a research plan is presented in Figure 1.1, which consists of four major phases. The four phrases of the research plan are: Phase 1: The selection and characterization of clinoptilolite and acid rock drainage. Phase 2: Evaluation of the suitability and compatibility of the clinoptilolite as a reactive barrier substrate by examining the chemical compatibility (including chemical stability, permeability and neutralizing capacity) between the barrier substrate and the acid rock drainage and its retention ability for metal ions using batch equilibrium adsorption tests; Phase 3: Investigate the performance of the clinoptilolite with the on-site flow conditions. The research also investigates the "worst" case scenarios that could be found on the site with maximum design concentration of metals in the ARD and at a high flow rate that could be encountered at the Britannia Mine site. Phase 4: Apply results from the batch equilibrium adsorption tests and the column leaching tests for the validation of the geochemical model in predicting transport of metals in the clinoptilolite barrier system. 8 c 0) E ro 9> Q or < ro Q) ro E ll (U > o ro , °> IOC (/) ro _ Characteristics of Acid Rock Drainage Characteristics of permeable reactive barrier substrate -Zeolite Clinoptilolite Review of geochemical model for the retention and transport of contaminants Adsorption ability Stability of substrate - chemical - physical Algorithm compilation Figure 1.1 Research Plan Flowchart Batch equilibrium adsorption tests • various concentrations, pHs and solution matrix (1) Column leaching tests - various flowrate, influent concentrations, presence of precipitants until contaminant breakthrough (2) Critical Path Analysis - natural ARD, flowrate, various reaction period Filter bed tests - natural ARD - filtered and non-filtered Batch equilibrium tests - pH (0.5 - 3.0 & 3.86) and time (0 min - 24 hrs) Chemical analysis of leachate: 1. pH 2. reduction-oxidation potential 3. electrical conductivity 4. sulphates 5. dissolved oxygen 6. concentrations of Cu, Fe, Mn, Zn and Al Soil analysis: 1. total digestion 2. selective sequential extraction for metals in fractions of: a. exchangeable b. carbonates c. Fe-Mn oxides d. organics e. residual 3. soil pH Langmuir isotherm, k Activity coefficient, Ka Validation of data into geochemical model Model for predicting the adsorption/retention and transport of metals in the clinoptilolite as substrate for removing metal ions from ARD in PRB systems 1.4 Research Contributions The contributions of this research include providing information on clinoptilolite as used for metal attenuation in ARD treatment and helping predict the performance of a clinoptilolite permeable reactive barrier (PRB) system with a contamination plume of similar source and chemical compositions. As a result of the study, the new knowledge on clinoptilolite to be revealed includes: • Long-term study on the metal retention ability and physio-chemical stability of clinoptilolite in laboratory-scale column leaching tests for the simulation use of clinoptilolite as a permeable reactive barrier system; • The effect of flow rates and ionic strength on the metal retention on clinoptilolite; • The partitioning of metals on clinoptilolite after treatment of acid rock drainage. The use of a clinoptilolite PRB system could benefit the mining and environmental industries by reducing costs for post-mine closure treatment and prevention operations for acidic drainages from exposed sulphide-bearing, acid-generating rocks. The use of locally available clinoptilolite should help promote the local mineral industry. In addition, determination of the pretreatment method for the substrate material helps in prolonging the service life of the permeable reactive barrier system. The empirical and geochemical model can be used to predict the transport, breakthrough time and partitioning of metals in the reactive barrier system. Accuracy in the predictions of the transport of metals within the barrier system will help determine the most suitable scale of the reactive barrier system, thereby minimizing monitoring efforts and post-mine closure operating costs. 10 1.5 Organization of Thesis In Chapter 1, statement of the problem of acid rock drainage has been addressed. The objectives of the research in the use of clinoptilolite in providing a possible treatment method for the heavy metals in ARD are listed. The research plan and the research contributions are outlined. In Chapter 2, a literature review of the problem - Acid Rock Drainage, the proposed solution to the problem - Clinoptilolite permeable reactive barrier, and the possible interactions between the clinoptilolite and the ARD are illustrated. The materials and methods used in the experiments are described in Chapter 3. The results from the characterization of the clinoptilolite and the acid rock drainage used in this study are provided in Chapter 4. Results of the batch adsorption and clinoptilolite stability tests, and discussions of the suitability of clinoptilolite for metal retention from acidic drainage based on its chemical stability and metals adsorptivities are presented in Chapter 5. Provided in Chapter 6 are the results and discussion from the column leaching cell tests, the breakthrough of metals from various flow conditions, and the metal retentions on soil after ARD treatment. Also proposed in this chapter is a design for a permeable reactive barrier using the experimental results obtained in this study. Outlined in Chapter 7 is the algorithm developed in a geochemical model to predict the transport of metals in the clinoptilolite leaching column. In the last chapter, Chapter 8, of this thesis, conclusions made from the findings are reported; recommendations for further researches and possible improvements for the use of 11 clinoptilolite permeable barrier study are provided. Finally, research contributions are summarized. 12 CHAPTER 2 LITERATURE REVIEW 2.1 Acid Rock Drainage Acid rock drainage (ARD) is a natural process whereby sulphuric acid is produced when sulphides in rocks are reduced through exposure to air and water. When large quantities of rock containing sulphide minerals are excavated from an open pit or underground mine, the exposed sulphide minerals react with water and oxygen to create sulphuric acid. When the water reaches a certain level of acidity (about pH 3.5), a naturally occurring bacteria, Thiobacillus ferroxidans, may flourish, accelerating the oxidation and acidification processes, leaching even more trace metals from the waste rocks. Leaching of metals is accelerated in the low pH conditions. Heavy metal pollution results when metals such as arsenic, cobalt, copper, cadmium, lead, silver and zinc contained in the excavated rock are leached out and carried downstream as water washes over the rock surface. Acidic drainage from the mine site can enter nearby streams, rivers, lakes and groundwater, severely degrading water quality, disturbing aquatic and wildlife habitat and making the water virtually unusable. This process could last for hundreds or thousands of years and has a great impact on the environment. Acid rock drainage is a serious global problem and it exists at almost every location where mining activities occur. ARD problems are commonly found in Africa, China, and the Americas. In North America, there are many mining locations where environmental impacts have raised public concerns, such as the Sudbury Mine in Ontario, Canada, and the Iron Mountain Mine in California, U.S.A (Waybrant, Blowes et al. 1995; Dixit, Dixit et al. 1996; Keith, Runnells et al. 2001). 13 According to Environment Canada, acid drainage is considered to be the most serious environmental threat posed by the mining industry. The Canadian mineral industry generates 650 million tonnes of waste every year (one million tonnes of waste rock and 950,000 tonnes of tailings per day), which contains acid-generating sulphides, heavy metals and other contaminants, and is usually stored above ground in large free-draining piles (Government of Canada 1991). The principal minerals and metals produced in British Columbia are coal, copper, crude oil and gold. The waste rock and exposed bedrock walls from which the mineral is excavated are the source of most of the metal pollution caused by mining in British Columbia (EMCBC 1997): Four main types of impacts from mining are acid rock drainage, heavy metal contamination and leaching, processing chemical pollution, erosion and sedimentation, which are interrelated. Among these impacts, acid rock drainage is of most concern because it causes the other impacts on the environment such as heavy metal contamination and leaching. There are 1809 abandoned mine sites across British Columbia, a number of sites have been identified as acid generating and potentially acid generating. The known acid generating mines in B.C. as of 1997 included (EMCBC 1997): • 12 historical mines: Anyox, Big Bull, Coast Copper, Giant Nickel, Mount Washington, Silver Standard, Baker, Britannia, Duthie, Mount Sicker, Silver Butte and Saint Eugene; • 11 closed sites: Bell, Granisle, Equity, Goldstream, Premier, Samatosum, Johnny Mountain, Island copper, QR Gold, Sullivan and Kitsault; • Three operating mines: Eskay Creek, Myra Falls, and Gibraltar. The identified potential acid generating sites include: • Three closed mines: Sable, Scottie Gold, Boss and Snip; 14 • Four operating mines: Elk, Quinsam, Huckleberry and South Kemess; • 12 deposits, in advanced stages of development preparation: Harmony Gold, Prosperity (Fish Lake), Lexington, Mount Milligan, Red Mountain, Tulsequah Chief, Cirque, Kutcho Creek, Lumby Muscovite, Red Chris and Telkwa. Due to both unacceptably high financial (high cost of remediation) and environmental liabilities, historical abandoned sites are frequently left untreated or without a proper closure program. This results in generation of untreated acidic and metal-containing effluents. There are many post-mining sites, as listed above, that have been left unattended once mining is completed. The problem of acid rock drainage has become most serious and costly to deal with and has already posed threats and hazards to the aquatic and wildlife in the receiving water bodies and surrounding areas. There are approximately 200 million tonnes of acid-generating tailings and 420 million tonnes of acid-generating waste rock in B.C. This is also increasing by 25 million tonnes per year (Feasby, Trembla et al. 1997). As copper is the most abundantly mined metal in B.C., acid generation from copper mines is often a concern to the environment. The characteristics of copper mine ARD is that it contains high concentrations of copper, iron, zinc and aluminum. The pH of the rock drainages from copper mines ranges from 2.0 to 7.9. Adequate environmental protection is needed at every hardrock mine site to prevent the runoff and leaching of the contaminants. The approaches for the protection at hardrock mines include rainfall management, surface water control, leak monitoring, reclamation and landscaping, and long-term monitoring. In practice, this protection was not conducted or mines were closed before any environmental criteria for the treatment of ARD were enforced. Methods for treating ARD include active and passive treatments. Active treatment involves removing the metals and neutralizing the acidic drainage in a treatment plant. However, this 15 option is expensive in terms of construction and maintenance, and is only cost-effective for drainages with high flow rates and ones that contain high amounts of metals. Another option to prevent ARD generation is to stop rain water from infiltrating the waste rock pile (Blowes and Ptacek 1994). This option will prevent the oxidation of sulphides, and the release and transport of subsequent reaction products such as heavy metals. Other mitigation measures include conventional treatment of ARD which involves precipitation of Fe-oxyhydroxides by addition of lime. This approach can be effective but often involves high operating costs and produces large volumes of metal-rich sludge, from which metals may subsequently leach (Benner, Blowes et al. 1997). For locations where acid rock drainage is detected, with metal concentrations insufficiently high for metal recovery but with enough toxicity to negatively impact aquatic life, permeable reactive barriers may present an effective and inexpensive option. 2.1.1 Generation of Acid Rock Drainage Acid Rock drainage (ARD) is a natural process whereby sulphuric acid is produced when all three primary components (reactive sulphide minerals, air and water) for acid generation are present to react with each other. When large quantities of rock containing sulphide minerals, such as pyrite (FeS2, the most abundant sulphide containing mineral), are excavated from an open pit or in an underground mine, the rock reacts with water and oxygen to create sulphuric acid. The process is shown in the following chemical reaction: FeS2 + 7/2 02 + H20 -» Fe2+ + 2S042" + 2H+ At pH levels greater than 4 or 5, the ferrous iron (Fe2+) is oxidized (or reacts with oxygen) to form ferric iron (Fe3+) by chemical oxidation reactions: Fe2+ + '/4 02 + H+ -> Fe3+ + lA H20 16 In the early stages of ARD at pH levels above 3.5, ferric iron tends to precipitate out of solution as ferric hydroxide (Fe(OH)3), a highly colored red precipitate. H+ ions released during the reaction, lower the pH. The following equation summarizes the ferric hydroxide precipitate as Fe3+ ion reacts with water. Fe3+ + 3H20 Fe(OH)3 + 3H+ From the above three equations, it can be seen that the oxidation of one mole of pyrite will produce a net, four moles of hydrogen ion. Bacterial activities involved in ARD When the water or acid rock drainage reaches about pH 3.5, commonly occurring bacteria, Thiobacillus ferrooxidans and Leptospirillum ferrooxidans, accelerate the oxidation and acidification processes, leaching even more trace metals from the rocks. The reactions that occur in the presence of the sulphide oxidizing bacteria include: (1) oxidation of ferrous iron (Fe2+) to ferric (Fe3+); and (2) direct oxidation of reduced sulphur (S2~) to form sulphuric acid (H2SO4). An illustration of the two oxidations reactions by Thiobacillus ferrooxidans is shown in Figure 2.1. FeS2 2S" + 302 + 2H20 = 2 H2S04 + 2e" Fe2+ = Fe3++e" & Fe3+ + 3H20 = Fe(OH)3 + 3H+ Figure 2.1 Bacterial activities of Thiobacillus ferrooxidans Thiobacillus ferrooxidans is an chemo-autotrophic bacterium, which derive energy from oxidation of either iron or sulphur of pyrite containing rocks. In the case of ARD, the bacteria utilize the liberated electron (e~) from the oxidation of ferrous to ferric iron for the reduction of 17 carbon dioxide into new cell material. Hence, these bacteria can grow in a minimal nutritional environment since all of the minimal requirements are readily available in the natural acid rock drainage. When Thiobacillus ferrooxidans is oxidizing the reduced sulphur compounds, as well as ferrous iron to the ferric iron state, it increases the rate-limiting step of ferric iron generation. Acid is produced microbiologically as a result of direct oxidation of reduced sulphur compounds. The acid will leach from the rock as long as the source rock is exposed to air and water, and until all sulphides are exhausted. Other bacterial species that participate or are capable of accelerating the oxidation of sulphide minerals, include Thiobacillus thiooxidans and Sulfolobus (Edwards, Bond et al. 2000). Effect of temperature on oxidation of pyrite The effect of temperature on the biological and chemical oxidation of pyrite is illustrated in Figure 2.2. The maximum biological oxidation occurs at 36°C for mesophillic bacteria and at 55°C for thermophillic bacteria. The chemical oxidation rate increases with increasing temperature. In general, both chemical and biological oxidation rates accelerate with an increase in temperature. Since the oxidation process is exothermic, the process generates heat and further increases the temperature in the environment. However, in cold-temperature regions, where temperatures are below 0°C, chemical oxidation is considered insignificant. 18 Temperature (°C) Figure 2.2 Effect of temperature on biological and chemical oxidation rate (adapted from Guidelines for Acid Rock Drainage Prediction in the North, Indian and Northern Affairs Canada, 1993) Effect of pH on oxidation of pyrite The effect of pH on the biological and chemical oxidation of pyrite at acidic pH is illustrated in Figure 2.3. The maximum biological oxidation occurs at a pH around 2.5. Chemical oxidation is insignificant below pH 5.2 compared to the biological oxidation. However, above pH 5.2, biological oxidation ceases and chemical oxidation becomes dominant. Biological oxidation of sulphide below pH 1 becomes insignificant while chemical oxidation below pH 1 starts to increase with decreasing pH. 19 0 1 2 3 4 5 6 Figure 2.3 Effect of pH on biological and chemical sulphide-oxidation rate (adapted from Guidelines for Acid Rock Drainage Prediction in the North, Indian and Northern Affairs Canada, 1993) Geochemistry and Neutralization of ARD Following the oxidation of the sulphide mineral, the resulting acid products may either be immediately flushed away by infiltrating water or removed from solution while in contact with an acid-consuming mineral. The acid-consuming mineral may neutralize the leachate and remove a portion of the acidity and iron from the solution and raise the pH. In evaluating the generation of ARD, it is important to know that not all sulphide minerals are equally reactive nor is acidity equally produced. The ability of a particular rock sample to generate net acidity is a function of the balance between the potentially acid-producing sulphide minerals and the potentially acid-consuming (or neutralizing/acid consuming) material. 20 Geologists determine both the acid-producing and acid-neutralizing mineral contents of samples from a mine site and analyze the results by a method called Acid-Base Accounting (ABA). If the buffering capacity of the minerals in the deposit is exhausted before the acid-generating capacity, acidic drainage will occur. Conversely, when the buffering capacity of the rock is greater than its acid-generating capacity, then acidic drainage will not occur. In that situation, metals will precipitate out of the solution. The approach in estimating the likelihood of the acid generated from waste rock being consumed is through the Neutralization Potential (NP) (Lawrence and Scheske 1997). The interpretation of NP values involves the consideration of mineralogical composition of the waste, and a system of classifying neutralizing minerals according to their relative reactivity. It is based on the contribution of carbonate minerals and the more reactive silicate minerals that can react with acidic drainage under conditions found in mine waste. For example, calcite, dolomite, brucite, and magnesite dissolve at pH 5 and have a relative reactivity of 1; comparing this to the intermediate weathered minerals such as phyllosilicates (chlorite, talc, etc), which have a relative reactivity of 0.02, and clay (vermiculite, montmorillonite, etc) with a relative reactivity of 0.01. Hence, carbonate-based mineral, like calcite, have a much higher neutralizing capacity for acidic drainages. Sulphate ions are stable over a wide range of pH, therefore, the concentration of sulphate in the drainage can be used as an indicator of the extent of acid generation even after neutralization by acid-consuming mineral has occurred. The most common acid-consuming mineral is calcite (CaC03), which consumes acidity by reacting with the H+ ions to produce calcium and bicarbonate ions (HCO3), as summarized in the following equation: CaC03 + H+^ Ca+ + HC03" 21 As neutralization occurs, metallic ions such as Fe , Cu , Zn , Pb , and As will precipitate in the hydroxide form. However, metallic ions in the acid rock drainage also precipitate with sulphates, carbonates, and their hydrated and/or hydroxy-complex forms. Beside calcite, dolomite is also a commonly found acid neutralizing mineral. Calcite and dolomite are capable of raising the pH of the water passing through the soil to around 6.5. Ferrous carbonate (FeCOs) has a capability of buffering drainage water to pH around 5.5, aluminum hydroxide in clay minerals, e.g. gibbsite, could buffer the drainage to pH around 4.0 through the dissolution of aluminum silicates. Ferric hydroxide, Fe(OH)3, which could occur from precipitation of the ferric ions from the naturally occurring acid rock drainage, could buffer the pH of the drainage to around pH 3.0 (Sherlock, Lawrence et al. 1995; Nordstorm and Alpers 1997). Different minerals can neutralize acid drainage at different rates and in different pH ranges. Many ore deposit types in Canada have little or no buffering capacity (AQUAMIN 1996). Therefore, if these deposits are exposed at the surface, or if wastes from the mining of these deposits are exposed, acidic drainage can result. Processes in the development of ARD The generation of ARD could be viewed as 3-stage process defined by the pH of the water that is in contact with the sulphide. Illustrated in Figure 2.4 are the three stages that occur in the process. 22 REACTIONS IN STAGES I AND II FeS2 (s) + 7/2 02 + H20 -» Fe+2 + 2S04"2 + 2H+ Fe+2 + lA 02 + H+ -» Fe+3 + V2 H20 "* *1 FeS2 (s) + 14Fe+3 + 8H20 -> 15Fe+2 + 2S04"2 + 16H+ 0 J— . 1 Time Figure 2.4 Stages in the formation of acid rock drainage (Adapted from Guidelines for ARD prediction in the North, Indian and Northern Affairs Canada, 1993) In the initial stage (Stage 1) of oxidation, the pH is nearly neutral. When oxygen and water first come into contact with the sulphide-containing mineral, the exposed sulphide mineral produces ferric hydroxide as a precipitate and sulphuric acid. This stage is carried out by chemical oxidation. However, in this stage, any calcium-based carbonate acid-neutralizing compound, such as calcite in the rock, would neutralize the amount of acidity produced and maintain the leachate at the neutral pH level (pH~7). Ferric iron is precipitated from solution as a hydroxide, and the rate of pyrite oxidation is controlled at a low rate by the loss of ferric iron from the solution. Therefore, drainage from this stage is affected by the acid-neutralizing capacity of the waste rock and the rate of flushing, regardless if it has enough time for the neutralization to occur. The drainage at this stage is also characterized by having an elevated level of sulphate because the H+ ions are taken up by the buffering agent, such as carbonates. 23 If ARD generation continues and the acid-neutralizing compound is exhausted,, the process moves towards the second stage (Stage 2) of the ARD development. When pH in the environment decreases towards 4.5, both chemical and biological oxidation reactions occur. Once biological oxidation is introduced below about pH 4.5, the oxidation of sulphide increases rapidly. Therefore, the pH of the drainage would also drop progressively. The drainage at this stage is weakly acidic with elevated ferrous iron and sulphate concentrations. The acidity can increase to relatively high levels even though the metal concentration in solution may be low. The acid generation continues when oxygen, water and sulphide-bearing mineral are in contact. When the alkali minerals are consumed or become unavailable, or when acidity is produced at a faster rate than alkalinity, the pH becomes more acidic and the sulphide-oxidizing reaction becomes dominated by biological activities. Ferrous ion is produced from the sulphide oxidation reactions, and biologically oxidized to ferric iron, which, in turn, replaces oxygen as the primary oxidant. The rate of oxidation in this stage is more rapid than in stage 1, whereas in stage 1 the oxidation occurs chemically. The decrease in pH also accelerates the rate of oxidation. The drainage from this stage (Stage 3) is acidic and contains high levels of sulphate and metals in solution (Edwards, Bond et al. 2000). Iron in the solution occurs primarily as ferric iron after biological oxidation. The acid generation continues until the sulphide has been completely oxidized, which depends on the sulphide content in the rock. The pH of the leachate decreases until the rock becomes essentially inert and the ambient pH of the water is not affected. This could take days to hundreds of years and is dependent on the factors required for acid generation, namely the reactivity/solubility of the sulphide-bearing material, the oxygen concentration and the availability of water. For example, in cold-temperature regions, the low temperature would 24 retard the onset of rapid oxidation and thus extend the time scale of the various stages of acid generation. 2.1.2 Characteristics of ARD Depending on geological factors and the type of mine at the site, the metals found in mining waste may include arsenic, antimony, cobalt, copper, cadmium, chromium, gold, iron, lead, molybdenum, silver and zinc. These metals tend to dissolve and mobilize more easily in the acidic waters associated with acid rock drainage. For many rock types, metal leaching will only be significant if the pH drops below pH 5.5. However, elements like molybdenum, zinc, cadmium, antimony and arsenic remain soluble at neutral or alkaline pH values (Price and Errington 1997). Concentrations of the metals found in the acid rock drainage vary largely at every mine and with the type of mine. Typical metal concentrations found in the drainage in the different kinds of mines are shown in Table 2.1. The drainage from copper and uranium mines appear to be the most acidic and contain the highest concentrations of metals. Table 2.1 Composition of mine drainage water from various types of mines (Ripley, Redmann et al. 1978) Copper Copper-Lead-Zinc Copper-Nickel Copper-Molybdenum Gold Iron Ore Mercury Uranium PH 3.8 2.0-7.9 7.5 7.7 7.6 6.4 11.5 2.3 Suspended solids, m g/L . 10-690 15 68 Dissolved solids, mg/L 700 1300 1460 1680 Aluminum, mg/L 0.6 Arsenic, mg/L 0.002-0.03 0.005 Calcium, mg/L 120 240 52 Cadmium, mg/L 0.002 Cobalt, mg/L 0.1 0.004 0.008 416 Chromium, mg/L 0.1 0.002 Copper, mg/L 0.01-83 0.005-76 0.15 0.02 0.24 0.1 3.6 Iron, mg/L 0.08-48 8.2-3200 1.2 0.21 3.6 1.3 30-3200 Mercury, mg/L 0.001 Magnesium, mg/L 4 106 Manganese, mg/L 0.27 0.4 0.43 5.6 Molybdenum, mg/L 0.04 Sodium, mg/L 16 101 Nickel, mg/L 1.0 0.01 0.22 0.1 Lead, mg/L 0.006 0.02-90 0.01 0.1 .01 0.7 Uranium, mg/L 67 Zinc, mg/L 0.01-91 0.04-1600 0.1 0.13 0.06 0.1 0.01 11 Sulphate, mg/L 320-1660 300 810 320 360-7400 25 2.2 Background of Britannia Mine Among the numerous abandoned mine sites in British Columbia, the Anaconda Britannia Mine is one legacy of B.C.'s mining industry. The Britannia Mine is located at Britannia Beach on the east shore of Howe Sound, approximately 48 km north of Vancouver, British Columbia, Canada. Britannia Copper Mine, one of the largest copper producers in North America, operated from the 1920's. Mining operations were stopped and the concentrator was closed in the 1970's. The site has been generating ARD since the mine first operated. A regional location map of the Britannia Mine site is presented in Figure 2.5. Jf-2 4 Roy 0 Jhurloiw Rock Bay cN»<Jales 0 WyattBay Bralorne °M°ha "Pavilion ,Cach« Creek Seton Portage oLillooet Vcroft Savona ;m ^e" Spences Bridge BRITISH COLUMBIA 1J Q Lytton an Campbell River CANADA Me Powell River Courtenay Britannia Mine Cumberland ,o Comox 19A • Got) (4} m X Vancouver Rurnaby PortAlberni w Nanaimo11 Richmond ©Surr^ Vancouver bland ci ^. _, MetSOWL Ucluelet A I 3 ; WASHINGTON ^"t_ BeBn*iaB%.g; m NITED STATES Mount Vernon •J Glacier Peak /Oak Harbor \ Wilderness \ ^nifaprtes Saanich, \ . —*< 20 Figure 2.5 Regional location map of the Britannia Mine site The problem of the acid rock drainage is significant (extremely low pH and high heavy metals concentration) partly due to the large exposure of sulphide-bearing rocks from open pit mining. This has created large basins, trapping heavy snows and rain, and tunneling the water 26 to the mine. The other reason for the low pH in the ARD is that the rocks in the west coast of the Canadian region have very low acid-neutralizing capacity. The runoff rate from surface excavations is approximately 400 to 600 m3 per hour. The two main pathways of drainage from the Britannia Mine site are the 2200 Portal and the 4100 Portal (Figure 2.6). ARD flows from the 2200 Portal into Jane Creek, a tributary of Britannia Creek, and ultimately into Howe Sound.. Since the water in the creek is warmer and fresher than the water in Howe Sound, the creek water tends to float on the upper layer of the Howe Sound water. This harms and kills the salmon, which live in and are attracted to brackish water and sheltered inlet areas such as found in Howe Sound (Price, Schwab et al. 1995). All mine drainages are directed to the 4100-Portal by a 6-km-long adit (Figure 2.7). The drainages from the 4100-Portal then flows into Howe Sound through a 27-m deep outfall at the mouth of the Britannia Creek. The purpose of the deep outfall is to reduce the impact of ARD on the surface water. Figure 2.8 shows a pond created by a large puddle of water downstream of the 4100 Portal. The picture was taken in October 1997. The dark-reddish-orange color shows a large amount of iron precipitate in the water. 2.2.1 Geological characteristics at the Britannia Mine The geological strata of the ore bodies mined at Britannia consist of sulphides, widely disseminated or concentrated in stringers and along bedding planes. The principal economic mineralization in the area is associated with quartz mineralization, and consists of pyrite and chalcopyrite, with important concentrations of sphalerite in certain areas of the deposits. The main mineralogy of ore bodies is simple and fairly constant. Pyrite is by far the most abundant mineral, with less chalcopyrite and sphalerite and minor erratically distributed galena, tennanite, tetrahedrite and pyrrhotite. The main nonmetallic minerals include quartz and muscovite (chlorite), anhydrite and siderite (MEMPR 1991). 27 to 00 Figure 2.6 Regional map showing sampling sites at the Britannia Mine (adapted from Price et al., 1995) Figure 2.7 Interior of the adit at the Britannia Mine Figure 2.8 Water containing high iron precipitates ponding at downstream of 4100 Portal, Britannia Mine site The water quality of the ARD from the 2200 level and the 4100 level are summarized in Table 2.2. The pH of the drainage from the 2200 level and the 4100 level is as low as 2.7, concentration of copper can reach as high as 115 mg/L, zinc can reach up to 48 mg/L, iron up to 55 mg/L and aluminum up to 74 mg/L. The concentrations of these metals far exceed the BC metal mine effluent limits which are: for copper 0.05-0.3 mg/L; zinc 0.2-1.0 mg/L; iron 0.3-1.0 mg/L; aluminum 0.5-1.0 mg/L; and pH limit of 6.5-8.5. The metals are toxic to aquatic life at those concentrations. However, the metal concentrations in the mine drainages are too low for economical metal recovery. Therefore, there is a need to find a cost-effective method to deal with the metal concentrations in the drainage. Reactive wall for ARD treatment is one of the potential solutions. Table 2.2 The range in water quality, flow and discharge data for 2200 and 4100 discharges, and for Jane and Britannia Creeks (adapted from Price et al., 1995) 2200 Drainage Jane Creek Britannia Creek Before 2200 After 2200 Before the 2200 Area Before Main 4100 Input After Main 4100 Input 4100 Drainage Flow, L/s 0-20 4.2-5.6 4.2-69 58-1030 426-2284 - 42-160 pH 2.7-4.6 5.9-7.4 4.0-7.6 5.3-7.0 4.4-6.0 3.6-4.9 3.2-4.5 Cu, mg/L 0.2-115 0.4-4.0 0.1-28 0.001-0.015 0.17-1.9 0.8-3.3 14-31 Zn, mg/L 1.3-48 0.7-1.6 1.1-13 0.006-0.029 0.20-0.70 1.3-3.4 22-27 Al, mg/L 4-74 0.2-2 0-21 0-0.1 0.1-0.9 0.7-3.0 20-32 Fe, mg/L 0.4-55 0.1-5 0.1-16 0.0 0.1-0.9 0.2-1.2 2-34 S04,mg/L 200-1950 46-93 62-545 3-6 14-5 112-238 1140-1900 2.3 Permeable Reactive Barrier Systems Passive treatment technologies use natural materials to facilitate chemical and biological processes. This cost-effective treatment is obtained by manipulating environmental conditions in the treatment system so that particular contaminant removal processes are optimized by utilizing locally sourced substrates. As a result, neither the materials nor the products of the vast majority of passive treatment are hazardous. Some passive systems, such as permeable 30 reactive barriers and wetland systems, are designed to operate for years with minimal requirement for operator intervention and/or costly maintenance (Younger, Ban wart et al. 2002). A permeable reactive barrier (PRB) is a passive in-situ treatment zone of reactive material that degrades or immobilizes contaminants as groundwater flows through it. PRBs could be installed as permanent, semi-permanent or replaceable units across the flow path of a contaminant plume. The transport of contaminants is carried out by natural gradients, through the strategically placed treatment media. The media, or substrate of the barrier system, retain the metals or other pollutants by degradation, sorption, precipitation and/or other retention mechanisms. The barriers may also contain reactants for degrading volatile organics, chelators for immobilizing metals, or other agents to enhance the performance (U.S.EPA. 1999). A schematic diagram of a permeable reactive barrier is presented in Figure 2.9. Figure 2.9 Schematic diagram of a permeable reactive barrier with contaminant plume (Adapted from Puis et al. (1998)) The goal of the permeable reactive barrier is to minimize the possibility that a contaminant plume can move toward and endanger sensitive receptors, such as through drinking water, or 31 discharge into surface waters. Rather than to constrain plume migration, permeable reactive barriers are designed as conduits for the contaminated groundwater flow (Puis, Powell et al. 1998). As contaminated water passes through the reactive zone of the PRB, the contaminants are either immobilized or chemically transformed to a more desirable state (e.g. less toxic, etc.). Passive treatment systems are now sufficiently numerous in North America (and increasingly in Europe) that is possible to identify some of the key pros and cons of the technology. The advantages of the passive treatment system include low operating and capital costs and their ability to work unattended for long periods. Passive systems can often be directly integrated with surrounding ecosystems and be more pleasant in appearance than active treatment systems. However, reliable expertise, research and field-scale practices of passive treatment technology are still scarce. In addition, since day-to-day intervention in the treatment processes is precluded in the operation, precise control of the treatment effluent quality is not feasible at this point. In cases where the flow rate of the contaminant plume is high, large land-area, which may not be available in all the problem locations, is likely necessary for the treatment system. 2.3.1 Design requirements Substrates used in a permeable reactive barrier should be compatible with the subsurface environment. The substrate should not cause adverse chemical reactions or by-products when reacting with constituents in the contaminant plume, and should not itself act as a possible source of contaminant. In order to minimize the cost of a permeable reactive system, it is required that the substrate be able to persist over a long period of time, maintaining its reactivity and integrity during its service life. The substrate material ideally should also be readily available at a low cost or be locally available. In addition, the substrate material for the permeable reactive barrier should 32 maintain a hydraulic conductivity that does not constrain groundwater flow. This could be achieved by using substrates with a wide range of grain sizes, which would prevent blocking of inter-granular spaces. Most PRBs are installed in either the funnel-and-gate, or the continuous trench configuration (Figure 2.10). The funnel-and-gate system employs impermeable walls to direct the contaminant plume through a gate, or treatment zone containing the reactive media. A continuous trench is installed across the entire path of the plume and is filled with reactive media. Figure 2.10 Schematic plan layouts of permeable reactive barriers: (a) continuous trench, and (b) funnel-and-gate There are no established rules for sizing the length of permeable reactive barriers in the direction of flow through the reactive substrate (Younger, Banwart et al. 2002) . However, a few considerations are suggested in the selection of the permeable reactive barrier substrate (Table 2.3): 33 Table 2.3 Considerations in designing permeable reactive barrier systems Design parameter Continuous wall Funnel-and-gate Method of flow Water passes through the barrier under its natural gradient and at its natural flow velocity Funnels direct groundwater towards permeable treatment zones or "gates". The funnel portion of the design is engineered to completely encompass the path of the contaminant plume the overall design must prevent the contaminant plume from flowing around the barrier in any direction. Cross-sectional area The PRB covers an area comparable to the cross-sectional area of the plume Up- and down-gradient surface areas of the aquifer material contacting the PRB should be approximately the same to minimize disruption in the natural groundwater flow relative to the funnel-and-gate design Groundwater flow velocity Groundwater flow velocity through the PRB is similar to the velocity in the aquifer Groundwater velocity within the gate should be higher than those resulting from the natural gradient Hydraulic conductivity To prevent underflow of the contaminated groundwater, hydraulic conductivity of the aquifer should be less than that of the PRB Funnel should be made of low permeability material to direct groundwater to the gate. The funnel typically consists of sheet piling, slurry walls, or some other material and is "keyed" into an impermeable layer to prevent contaminant underflow. Hydraulic conductivity of the reactive material must be equal to or greater than the aquifer permeability to minimize flow restrictions. Depth The PRB should be built to a depth that over-encompasses the vertical and horizontal dimensions of the contaminant plume The bottom of the "gate" should be built into an impermeable zone/strata to mitigate the potential for contaminant underflow Thickness The thickness of barrier should be sufficient to remediate the contaminant of concern to the established concentration goals, i.e. sufficient contact, or residence time between the reactive material and the contaminant. Since the reaction mechanism associated with retention of contaminants in PRB are surface area, cation exchange capacity dependent, the amount of reactive material needed should be proportional to the mass flux of contaminant requiring treatment 34 In the actual field implementation of PRBs, it is possible that additional substrates be required than that estimated from theoretical calculations, particularly if the plume is very broad and requires a relatively long PRB. In this case, there is a need to be certain that the PRB has no un-reactive gaps or flow paths throughout its length to maintain full coverage of the plume. There will also be some limiting thicknesses that have to be maintained to assure the integrity of the PRB throughout its volume. The cost of PRBs would depend on the depth, width and saturated thickness of the plume, which controls the overall dimension of the system. The major components for the capital cost of a PRB are the reactive materials, "funnel" material (if installing a funnel-and-gate system), and installation cost. 2.3.2 Uses of permeable reactive barrier systems Permeable reactive barriers have been investigated and used in contaminants including chlorinated solvents, metals and inorganics, fuel hydrocarbons, nutrients, radionuclides and other organic contaminants. Some examples of previously studied full and pilot-scale PRBs for metals contaminants are listed in Table 2.4. Table 2.4 Previous uses of permeable reactive barriers in metal contaminants removal Location Targeted metal contaminants Reactive media Nickel Rim Mine Site, Sudbury, Ontario, Canada Ni, Fe, sulphate Organic carbon: Reactive mixture of municipal compost, leaf compost and wood chips. Pea gravel was added to increase hydraulic conductivity. PRB installed using cut-and-fill technique. Tonolli superfund site, Nesquehoning, PA, USA Pb, Cd, As, Zn, Cu Limestone groundwater trench. U.S. Coast Guard Support Center, Elizabeth City, NC, USA Cr+b Continuous trench wall of zero-valent iron (Fe0). LEAP Permeable Barrier Demonstration Facility, Portland, OR, USA Cr+b Surfactant-modified zeolite (SMZ) permeable reactive barrier hung in the center of the simulated aquifer. 35 In this study, zeolite clinoptilolite was selected as the substrate material for a permeable reactive barrier system. An evaluation and the suitability of clinoptilolite to be a substrate material are discussed in the following Section 2.4. To be a good permeable reactive barrier material, it has to satisfy the following criteria: • have enough adsorption capacity to hold target contaminants in the contaminated groundwater plume • be chemically stable so that under acidic conditions, the barrier material is able to maintain its structure chemically, that the cations within the framework of the chemical structure not dissociate nor that any of the ions leach into the environment • be physically stable and be able to maintain the required hydraulic conductivity to handle the flow of groundwater in the field condition, avoiding underflow of the contaminated plume to the aquifer In the design of the permeable reactive barrier, field conditions must be defined in order to achieve the treatment goal efficiently. Such parameters include: • flow rates • concentrations and type of contaminants in the contaminated plume • width or horizontal spread of the contaminated plume • depth of the contaminated plume • anticipated service life of the PRB system • required effluent concentration from PRB system Once the field conditions are defined, the thickness of the PRB could be determined using the equation: Thickness = mass / (width * depth) Where mass = mass flux of contaminant in influent / contaminant retention capacity of substrate 36 2.4 Properties of zeolite Zeolites are tectosilicates, that is they consist of three-dimensional frameworks of S1O44" tetrahedra wherein all the oxygen (0) of each tetrahedron is shared with adjacent tetrahedra. The zeolite structure would be electrically neutral if each tetrahedron were to contain Si as its central cation. However, in isomorphic substitution, the Si4+ ion is replaced by Al3+ ion, which leads to a deficiency of a positive charge in the framework. From the structural view of zeolite, each O atom is shared between two tetrahedra with no mobile anions present. The resulting zeolite structure has a net negative charge due to the presence of Al-centered tetrahedra, these negative charges are counter-balanced by the presence of alkali and alkaline earth cations (such as Na , K — monovalent cations; Ca , Mg , Sr , Ba — divalent cations) within the existing pores elsewhere in the structure (Sherry 1971; Boles, Flanigen et al. 1977; Dixon and Weed 1989). The zeolite structures can be visualized as linking the primary building units of SiO"4 and AIO4 tetrahedra into secondary building units (SBUs) in simple geometrical forms. The SBUs range in complexity from simple rings of 4 or 6 tetrahedra (4-rings or 6-rings) to cubo-octahedra (8- or 12- rings). SBUs may be linked together in a variety of ways, giving rise to a crystal structure possessing a unique set of physical and chemical properties. For example, the structure of clinoptilolite is based on a complex linkage of 4- and 5- rings of tetrahedra. These hydrated aluminosilicates consisting of tetrahedral framework of O atoms, surrounding either a Si or an Al atom, extended in a three-dimensional network provide structural channels (Boles, Flanigen et al. 1977; Gottardi and Galli 1985; Dyer 1988; Dixon and Weed 1989). Since each zeolite has its own arrangement and number of channels, making its unique molecular sieve properties, it will adsorb cations selectively according to the size and the speciation of the cations. 37 Zeolites are also characterized by the large void volumes of 20% to 50% and the large internal surface areas of several hundred thousand square meters per kilogram, which are the requirements of a good adsorbent. The other unique feature of zeolite is its ability to hydrate and dehydrate reversibly, and to exchange constituent cations without disrupting the bonds of the framework (Dyer 1988). Clinoptilolite, a sub-category of zeolite, which means "oblique feather stone" in Greek, received its name because it was thought to be the monoclinic (or obliquely inclined) phase of the mineral ptilolite, as in "oblique ptilolite". But ptilolite was later found to be the earlier named mineral mordenite; consequently ptilolite is no longer in use. Clinoptilolite is very closely related to heulandite and is currently being considered for disuse itself as it may just be a variety of heulandite. It differs from heulandite significantly only in its enrichment in potassium and slightly higher silica content, and it is argued that a separate mineral is not needed. The name clinoptilolite is widely recognized and used among zeolite industries, mineral collectors and mineralogists and recognized as a legitimate, distinct mineral (Breck 1974). 2.4.1 Minerological properties of Clinoptilolite The representative unit-cell formula for clinoptilolite is (Na3K3){Al6Si3o072}»24H20. The Si/Al ratio is 4.3-5.3 (Dixon and Weed 1989). In this type of zeolite, the net negative charge created by the replacement of Si4+ by Al3+ cations in isomorphic substitution, is balanced by the exchangeable cations K+ and Na+. A summary of the major elemental composition in the zeolite clinoptilolite is shown in Table 2.5. The Si/Al ratio of clinoptilolite used in this study is 5.4. 38 Table 2.5 Major elemental compositions from ICP analysis on clinoptilolite (Beck, 1996) Element Concentration (//g/g or ppm) Aluminum 110,000 Potassium 40,000 Iron 36,000 Sodium 18,000 Zinc 210 Copper 71 2.4.2 Stability in low pH One of the concerns of the use of zeolite as a reactive material with ARD, is its resistance to low pH. It is often assumed that zeolites have a low resistance to mineral acids. Since zeolites are mainly made up of silica and aluminum (in the tetrahedra and octahedra layers), they are susceptible to partial dissolution by either acidic or basic environments. This ease of dissolution can be linked to the readily removed aluminum from tetrahedral site frameworks where Si:Al is 1:2. The leached aluminum hydrolyses to a variety of species in which the metal is hexa-coordinated. However, at higher Si:Al ratio, even though the framework aluminum was leached by the mineral acids, zeolitic structures can still be retained due to the high ratio of silica in the framework (Dyer 1988). If the zeolite is infdtrated with a chemical that would destroy the structure of the zeolite itself, or cause dissociation of the zeolite, the purpose of retaining the contaminant or the metal cations would not be served. For example, leachates of very low pH can dissolve aluminum, iron, alkali metals and alkaline earths; bases dissolve silica, depending on the resistance of the mineral (Gottardi and Galli 1985). The solubility of the zeolite depends on the nature of the 39 acid, the acid concentration, the acid to clay ratio, and the temperature and duration of the treatment (Grim 1968). The loss of Al from the zeolite framework will cause a consequent loss of capacity and ultimately collapse of the framework. The limit of acid resistance of zeolite is usually about pH 3-4 (Dixon and Weed 1989). When the pH drops even lower, zeolite with Si/Al ratio in the range of 1-2 will readily lose Al from their framework. The leached Al is readily hydrolyzed to a variety of species in which the metal is hexa-coordinated (Dyer 1988). For higher Si:Al ratio zeolites, such as clinoptilolite (Si/Al ratio ranges 4.3-5.3) (Boles, Flanigen et al. 1977; Gottardi and Galli 1985; Dyer 1988; Dixon and Weed 1989), the effect of mineral acid is that Al leaches from the framework, and decomposes to an aqueous aluminosilicate gel. In some instances, a zeolitic structure can be retained when there is more Si than Al ions in the framework, because the structure is less easily destructed by the loss of Al. Clinoptilolite has been shown to retain crystal habit and integrity even after six months of exposure to 8 M nitric acid. Leaching can be promoted by agents other than mineral acids and the same effect can be attained by treatment, for example, with ethylenediamine tetra-acetic acid (H4EDTA), silicon tetrachloride, fluorosilicates, organic acids and even acetyl-acetone, which form stable Al complexes. It has been claimed that clinoptilolite can be stripped of all aluminum to leave a silica pseudomorph of each structure. Studies on the reaction of zeolites with moderate acid molarities demonstrate that the first stage of the leaching occurs by a cation exchange, whereby hydronium ions (H30+) replace the indigenous cations (Dyer 1988). De-cationated forms of clinoptilolite were obtained by Barrer and Coughlan (1968) by treatment with 0.25, 0.5, 1 and 2 N acid solutions with the contemporaneous removal of 42%, 67%, 93%, and 100% of the original Al, whose charges in the framework are substituted by 3 H+, whereas the charges of the leached extra framework cations are substituted by H30+. The 40 acid treated material, even with 2N HC1, retains the original x-ray powder pattern of clinoptilolite, with only minor variations (Gottardi and Galli 1985). In a recent study by the USGS (Gilbert, O'Meara et al. 1999) on the adsorption capabilities of clinoptilolite to mine drainage (pH 2-3), a slight (120ug per gram of soil) desorption of Al in the clinoptilolite-rich rock had been observed. This could be explained by the displacement of Al in the zeolite framework by the H+ ion in the acidic solution. 2.4.3 Adsorption capability An earlier study on the adsorption of heavy metals from ARD was conducted on the clinoptilolite sample used in this research (Beck 1996). It was found that zeolite is effective in reducing Equity Silver's ARD's copper by 68%, zinc by 49%, arsenic by 97% and iron by 93%. However, the test results only revealed the adsorption in the first 18 hours of reaction, and the stability of the zeolite, which may be affected by the acidic environment, had not been investigated. A study conducted by Vos (1992) in using zeolite for treatment of ARD from Britannia mine, and recent research done by Gilbert et al. (1999) in using clinoptilolite for ARD remediation has found that clinoptilolite is effective in removing metal cations such as aluminum, iron, lead, cadmium and copper in the ARD. The US Geological Survey tested three Clinoptilolite-Rich Rocks (CRR): sodium-rich, calcium-rich and potassium-rich, for their ability to exchange copper, lead and zinc in the presence of low concentrations of calcium and potassium, which compete with metal ions in the zeolite exchange process. Samples were exposed to acidified solutions of pH 2.1 containing 1.2 mg/L copper, 1.3 mg/L lead, 15 mg/L zinc, and 2 mg/L calcium repeatedly. The sodium-rich CRR tested was able to removed approximately 65% of the zinc, 55% of the copper, and 95% of 41 the lead from solution and was more effective at removing copper and zinc from solution than the calcium- and potassium-rich CRR. Another experiment was conducted using a solution with a pH of 2.6 and containing 100 mg/L copper, 0.31 mg/L lead, 19 mg/L zinc, 3.1 mg/L calcium, 3.3 mg/L potassium, and 15 mg/L sodium. Again, sodium-rich CRR removed more copper and zinc from solution than the calcium- and potassium-rich CRR (Desborough 1997; Virta 1997) However, these results were based only on batch adsorption tests and short term monitoring. The chemical compatibility to clinoptilolite and its stability in the ARD environment was not investigated. 2.4.4 Surface Sorption Phenomena Retention mechanism at a surface is termed as sorption, which includes adsorption, surface precipitation and polymerization. Factors that could affect retention of metals include ion association, gas-water reaction, ion exchange reactions, sorption and mineral-solution reactions. The mechanisms of the factors are listed in Table 2.6. Table 2.6 Factors that could affect retention of metals and the mechanisms (Adapted from Sparks, 1995) Factor Mechanism Ion association reactions Ion pairing, complexation (inner and outer sphere), and chelation type reactions in solution Ion exchange reactions Electrostatic ion replacement reactions on charged solid surfaces Sorption Simple physical adsorption, surface complexation (inner- and outer- sphere), and surface precipitation reactions Mineral-solution reactions Precipitation/dissolution reactions involving discrete mineral phases and coprecipitation reactions by which trace constituents can become incorporated into the structure of discrete mineral phrase Electrostatic ion reactions 42 Clay crystals carry a permanent net negative charge as a result of isomorphous substitution. The net negative charge is compensated by cations, which are located on the layer surfaces. In the presence of water, these compensating cations have a tendency to diffuse away from the layer surface since their concentration is smaller in the bulk solution. On the other hand, they are attracted electrostatically to the charged layers. The result of these opposing trends is the creation of a distribution of the compensating cations in diffuse electrical double layer on the exterior layer surfaces of a clay particle. Theses cations between the layers are confined to the narrow space between opposite layer surfaces. The compensating cations act as the counter-ions of the double layers, like all counter-ions, they are exchangeable for other cations (Mohamed and Antia 1998). Distance Figure 2.11 The diffuse-ion layer (Adapted from Mohamed and Antia, 1998) The first layer of ions (Stern layer) is not immediately at the surface, but at a distance away from it. The counter-ion charge is separated from the surface charge by a layer of thickness (S) in which no charge exists. A simple model relating the density of charge on the surface to the electrical potential across the double layer was developed by Gouy (1910) and Chapman (1913). 43 The model was based on the assumptions that: (1) the adsorbent surface is a uniform plane of charge density; (2) the adsorptive ions are point species that interact mutually and with the adsorbent through the Coulomb force. Their only mechanism of adsorption is the diffiise-ion layer; and (3) the aqueous solution phase is a uniform continuum of dielectric constant in which the point-ion adsorptive is immersed. The relationship between the thickness of the double layer (—) and the ion concentration in the equilibrium solution as described by the K Gouy-Chapman theory is presented as follows: 2 %nne1z1 _2 K = cm ekT where — = thickness of the double layer K n = counter ion concentration in the equilibrium solution (ions/cm ) = molarity x 10"3 x Avogardo's number K = Boltzmann constant (kT = 0.4x 10" ergs at room temperature) T = absolute temperature Z = valences of ions e = elementary charge (4.80 x 10"10 esu) e = dielectric constant (80) The Guoy-Chapman theory predicts that double-layer thickness ( — ) is inversely K proportional to the product of ion concentration and valency of the electrolyte in the external solution (Sparks 1995). It is also noted that the actual thickness of the electrical double layer cannot be measure, but it is defined mathematically as the distance of a point from the surface where the change of potential (y/) with distance (x) is 0 (dy/1 dx - 0). Adsorption 44 Adsorption is the net accumulation of masses at the interface between a solid phase and an aqueous solution phase (Sposito 1984). Adsorption can include the removal of solute molecules from the solution, solvent from the solid surface, and attachment of the solute molecule to the surface (Stumm 1992). Both physical and chemical forces are involved in adsorption of solutes from solution. Physical forces include: van der Waals forces (e.g. partitioning) and electrostatic outer-sphere complexes (e.g. ion exchange). Chemical forces result from short-range interactions that include inner-sphere complexation that involves a ligand exchange mechanism, covalent bonding, and hydrogen bonding (Stumm and Morgan 1981; Sparks 1986). Surface functional groups is "a chemically reactive molecular unit bound into the structure of a solid at its periphery such that the reactive components of the unit can be bathed by a fluid" (Sposito 1984). Surface functional groups may be organic (e.g. carboxyl, carbonyl, phenolic) or inorganic molecular units. The major inorganic surface functional groups in soils are the siloxane surface associated with the plane of oxygen atoms bound to the silica tetrahedral layer of a phyllosilicate and hydroxyl groups that are associated with the edges of inorganic minerals such as kaolinite, amorphous materials, and metal oxides, oxyhydroxides, and hydroxides. When the interaction of a surface functional group with an ion or molecule present in the soil solution creates a stable molecular entity, it is called a surface complex and the reaction is referred to as surface complexation. There are two types of surface complexes that can form: (1) outer-sphere and (2) inner-sphere (Sparks 1995). An outer-sphere surface complex is defined as the complex that is obtained when at least one water molecule is interposed between the surface functional group and the ion or molecule it binds. Outer-sphere surface complexes involve electrostatic coulombic interactions and bonding mechanisms, and thus are less stable than inner-sphere complexes in which the binding is covalent or ionic. Outer-sphere 45 complexation is usually a rapid process that is reversible, and the adsorption via this mechanism is affected by ionic strength of the aqueous phase. Adsorption by outer-sphere complexation occurs only on surfaces that are of opposite charge to the adsorbate. This type of adsorption is called non-specific adsorption (Sposito 1984; Mohamed and Antia 1998). An inner-sphere surface complex is defined as the complex that is obtained when no water molecule is interposed between the surface functional group and the ion or molecule it binds. The inner-sphere surface complexes involve either ionic or covalent bonding, or some combination of the two. The interaction is usually slower and often not reversible and adsorption by this mechanism is weakly affected by the ionic strength of the aqueous phase. Inner-sphere complexation can increase, reduce, neutralize, or reverse the charge on the sorptive regardless of the original charge. Adsorption of ions via inner-sphere complexation can occur on a surface regardless of the surface charge. Since covalent bonding depends significantly on the particular electron configurations of both the surface group and the complex ion, inner-sphere surface complexation is termed specific adsorption (Sposito 1984; Mohamed and Antia 1998). It is also noted that outer- and inner-sphere complexation can, and often do, occur simultaneously (Sparks 1995). If a solvated ion does not form a complex with a charged surface functional group, but instead neutralizes surface charge, it is said to be adsorbed in the diffuse-ion layer. This adsorption mechanism involves ions that are fully dissociated from surface functional groups and are free to move in the soil solution. The diffuse-ion layer involves electrostatic bonding. There is only a weak dependence on the electron configuration of the surface group and the adsorbed ion. This type of adsorption is also called non-specific adsorption (Mohamed and Antia 1998). 46 Surface precipitation As the amount of metal cation or anion sorbed on a surface increases to a high surface coverage, a precipitate of the cation or anion can form with the ions of the mineral and is know as surface precipitation. It is the process whereby a three-dimensional growth mechanism occurs on the solid. There is a continuum between surface complexation (adsorption) and surface precipitation as shown in Figure 2.12 (Sparks 1995). At low surface coverage, surface complexation dominates and as surface coverage increases, nucleation occurs or distinct entities or aggregates occur on the surface. As surface loading further increases, surface precipitation becomes the dominant mechanism. When the precipitate covers the entire surface, it is referred to as a "surface precipitate". When the precipitate grows away from the surface before covering it, it is referred to as a "surface cluster" (Stumm 1992; Sparks 1995). Figure 2.12 An illustration of metal ion sorption on (hydr)oxide (Adapted from Sparks, 1995) Low surface coverage Dominated by isolated Site hinrlinri „.,, 'covering it 47 Relative strengths amount sorption mechanisms The relative affinity that a given metal cation has for a soil adsorbent depends in a complex manner way on the soil solution composition. But as a first approximation, the selectivity of a soil for an adsorption metal cation can be rationalized in terms of inner-sphere and outer-sphere surface complexation, and diffuse-ion layer concepts. The relative order of increasing interaction strength among these three adsorption mechanisms is: diffuse-ion layer > outer-sphere complex > inner-sphere complex. For the diffuse-ion layer only the metal cation valence and surface charge are critical in determining adsorption affinity. The outer-sphere complex is intermediate, in that valence is probably the most important factor (Mohamed and Antia 1998). Hence, the relative affinity of a soil adsorbent for a free metal cation will increase with the tendency of the cation to form inner-sphere surface complexes. Effect of ionic radii In case of a series of metal cations of a given valence, the tendency is correlated positively with the ionic radius(Mohamed and Antia 1998). The first reason is that metal cations with larger ionic radii will create a smaller electric field (since ionic potential = valence/ionic radius) and be less likely to remain solvated for complexation by a surface functional group. The second reason is that metal cation, which has a larger ionic radius implies a larger spread of the electron configuration in space and a greater tendency for a meal cation to polarize in response to the electric field of a charged surface functional group, which polarization is a necessary prerequisite for the distortion of the electron configuration leading to covalent bonding (Mohamed and Antia 1998). Given the above considerations, the relative adsorption affinity series, or so-called the selectivity sequence, of the metal cations found in this study based on ionic radius is: 48 K+>Na+>Zn2+ With respect to transition metal cations, ionic radius is not adequate as a single predictor of adsorption affinity because electron configuration plays a very important role in the complexes of these cations (e.g. Mn, Fe, Ni). The relative affinities of the transition metal cations found in this study are listed as follow, and tend to follow the Irving Williams order (Mohamed and Antia 1998): Cu2+>Fe2+>Mn2+. Effect of pH As pH increases, surface charge decreases toward negative values, and the electrostatic attraction of a soil adsorbent for a metal cation is enhanced. If a soil is reacted with a series of aqueous solutions containing a metal cation at the same initial concentration but having an increasing valence, the amount of metal cation adsorbed will increase with pH (Mohamed and Antia 1998). 2.4.5 Availability of the substrate material In search of a suitable substrate for a permeable reactive barrier system (other than the chemical compatibility, attenuation ability for the contaminants and its service life), it is also important that the barrier substrate be locally available for economical purposes, as a large amount of substrate material could be needed in constructing and the placement of the reactive-barrier system. Clinoptilolite may be found in large quantities from sedimentary deposits (saline, alkaline lakes, deep-sea sediments, and low-temperature tephra systems). Therefore, it may be obtained at a competitive price (Hay and Sheppard 2001; Dobelin and Armbruster 2003). Open-system zeolite clinoptilolite is found at Asp Creek, Sunday Creek in B.C.; Nye County in Nevada, John Day Formation in Oregon, Death Valley Junction in California, USA; and Yellow tuffs near Naples in Italy (Sheppard and Simandl 1999). Depending on the quality 49 and specification, the price of clinoptilolite ranges between 50 and 300 US$ per ton (Armbruster 2001). 2.5 Geochemical models Geochemical models have been used widely in field-based environmental problems of contamination and water-resources studies. Common practices in using geochemical model for contamination problems include determining the concentrations in groundwater for risk assessment, to evaluate feasibility of remedial alternatives, and to demonstrate potential migration of regulated chemical species or other adverse environmental impacts in applying for permits for mining or waste disposal/storage facilities. In maximizing the performance of a permeable reactive barrier, it would be an advantage to have the knowledge and predictions of the solute concentrations in space and time within the barrier system. Knowing the fate of the contaminants in the barrier system, one could predict the extent of contamination and to predict the service life of a barrier system. Predictions of how the contaminants react with the substrate can be used to make estimates of the amount of substrate needed for the barrier system. This in turn, minimizes the cost and maximizes the use of the subst