Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Measurements and modeling of gas fluxes in unsaturated mine waste materials 2007

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

Item Metadata

Download

Media
ubc_2007-317921.pdf
ubc_2007-317921.pdf [ 32.07MB ]
Metadata
JSON: 1.0081099.json
JSON-LD: 1.0081099+ld.json
RDF/XML (Pretty): 1.0081099.xml
RDF/JSON: 1.0081099+rdf.json
Turtle: 1.0081099+rdf-turtle.txt
N-Triples: 1.0081099+rdf-ntriples.txt
Citation
1.0081099.ris

Full Text

M E A S U R E M E N T S A N D M O D E L I N G O F G A S F L U X E S IN U N S A T U R A T E D MINE W A S T E M A T E R I A L S by LOUIS K A T E L E KABWE Dipl., Institut de Recherche Scientif ique, 1977 B . S c , Universite du Quebec a Montreal, 1983 M . S c , M . S c , The University of Saska tchewan, 1994, 2001 A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Mining Engineering) T H E U N I V E R S I T Y O F BRIT ISH C O L U M B I A June 2007 © Louis Katele Kabwe, 2007 Abstract Accurate measurements and predictions of surface CO2 f luxes are needed to quantify b iogeochemical reaction rates in unsaturated geologic media and soi ls. However, no standard appears to exist for establ ishing the accuracy of field measurements of soil respiration rates. A s a result, a technique to measure CO2 f luxes - from the soil surface to the atmosphere was recently deve loped and verified in m e s o c o s m s over the range of C0 2 f luxes reported for field condit ions. The method, termed the dynamic c losed chamber ( D C C ) , was shown to accurately measure CO2 f luxes from ground surface to the atmosphere in m e s o c o s m s . The main advantage of this direct technique is the almost instantaneous estimation of the CO2 flux. Al though the D C C is a promising technique, its ability to accurately quantify surface C0 2 flux under field condit ions remains to be verif ied. The field application of the D C C is investigated in this thesis with a particular focus on quantifying reaction rates in waste-rock piles at the Key Lake uranium mine in northern Saska tchewan , C a n a d a . It should, however, be noted that the dominant geochemica l reactions in the two waste-rock piles at the Key Lake mine were not typical of acid rock drainage (ARD) waste-rock piles. The CO2 f luxes measured in this study occur in the organic material underlying the waste rocks, in contrast to A R D waste-rock piles where O2 consumpt ion and CO2 production are the results of sulphide oxidation and carbonate buffering. This work provided a complete suite of measurements required to character ize spatial distribution of CO2 f luxes on larger-scale studies of waste-rock pi les. There has been no previous f ield-scale study to quantify C0 2 f luxes across a waste-rock pile. ii The ability of the D C C method to accurately quantify field soil respiration was demonstrated by comparing the DCC fluxes to those obtained using two other C 0 2 flux measurement techniques: the static closed chamber (SCC) and eddy covariance (EC) methods. The DCC yielded comparable data but had distinct advantages over the two other methods in terms of speed and repeatability. The DCC was also used to investigate C 0 2 fluxes under the climatic variables (e.g., rainfall and evaporation) that affect soil water content at the Deilmann north (DNWR) and Deilmann south (DSWR) waste-rock piles, at the Key Lake uranium mine. The effects of rainfall events on waste-rock surface-water conditions and C 0 2 fluxes were of short duration. A simple model for predicting the effects of soil water content on C 0 2 diffusion coefficient and concentration profiles was developed. The model was verified with measured C 0 2 fluxes obtained from meso-scale columns of unsaturated sand. Verification of the model showed good agreement between predicted and measured data. The model was subsequently used to predict C 0 2 diffusion and concentration profiles in response to changes in soil water contents in the piles and also to predict surface C 0 2 fluxes from the DNWR and DSWR for a 6-d test period [August 1 (day 3) to August 6 (day 8) 2002] following a 72.9 mm precipitation event over the initial 48-h [July 30 (day 1) to July 31 (day 2) 2002]. The model predicted surface C 0 2 fluxes trends that were very similar to the measured surface C 0 2 fluxes from the DNWR and DSWR piles during the test period Based on the tests conducted in this thesis the D C C method has shown to be suitable for field applications to quantify C 0 2 fluxes and to characterize the spatial and temporal dynamics of C 0 2 fluxes from unsaturated C-horizon soils and waste-rock piles. i i i Table of Contents Page Abstract ii Table of Contents iv List of Tables ix List of Figures x Abbreviations and Symbols xxiii Acknowledgements xxvii Dedication xxviii CHAPTER 1 : Introduction 1 1.1 Introduction 1 1.2 Research Objectives 3 1.3 Organization of the Thesis 5 CHAPTER 2: Literature Review 7 2.1 Introduction 7 2.2 Waste-Rock Piles 8 2.3 Sources of C 0 2 in Subsurface Soils and Waste-Rock Piles 19 2.3.1 C 0 2 Production by Microbial Respiration (Biotic) 20 2.3.2 C 0 2 Production by Pyrite Oxidation-Carbonate Buffering (Abiotic) 24 2.4 Studies of C 0 2 in Subsurface Pore Gas and Associated Surface Gas Fluxes from Waste Rock and non Waste-Rock Systems 27 2.4.1 Studies of Subsurface C 0 2 Gas and Surface C 0 2 Fluxes from Waste Rock Piles 28 2.4.2 Studies of Subsurface C 0 2 from Non-Waste-Rock Material 31 iv 2.5 Climatic Variables Affecting Subsurface and Surface Gas Fluxes: Precipitation and Evaporation 39 2.5.1 Evaporation 40 2.5.2 Methods of Predicting Evaporation 44 2.5.3 SoilCover Program 47 2.5.4 Chapter Summary 50 C H A P T E R 3: Materials and Methods 53 3.1 Introduction 53 3.2 Laboratory Program 53 3.2.1 Sample Collection 53 3.2.2 Grain-Size Analysis 54 3.2.3 Water Retention Curve 56 3.2.4 Saturated Hydraulic Conductivity 58 3.3 Laboratory Mesocosm and Minicosms (sand columns) 62 3.4 Field Program 65 3.4.1 Site Location and Description 66 3.4.2 Field C 0 2 Flux Measurements Methods 72 3.4.2.1 Measuring C 0 2 Fluxes using Dynamic Closed Chamber (DCC) Method 74 3.4.2.2 Measuring C 0 2 Fluxes using Static Closed Chamber (SCC) Method 77 3.4.2.3 Measuring C 0 2 Fluxes using Eddy Covariance (EC) Method 80 3.4.2.4 Gravimetric Water Content Measurement 83 3.4.2.5 Meteorological Weather Station 83 3.4.2.6 Chapter Summary 84 C H A P T E R 4: Results and Data Interpretation 86 4.1 Laboratory Tests Program 86 4.1.1 Grain-Size Distribution 86 4.1.2 Water Retention Curve 91 4.1.3 Hydraulic Conductivity 99 4.2 Field Tests Program 104 4.2.1 Diurnal Variation In C 0 2 Flux 104 4.2.2 Spatial and Temporal Variation in C 0 2 Flux Measured using the Dynamic Closed Chamber at the Deilmann South Waste-Rock Pile 107 4.2.3 Spatial and Temporal Variations in C 0 2 Flux Measured using the Dynamic Closed Chamber at the Deilmann North Waste-Rock Pile 112 4.2.4 Cross-Statistical Comparison Between C 0 2 Fluxes Measured from across the DNWR and DSWR 115 4.3. Comparison of C 0 2 Fluxes Measured using the DCC to those Measured using Static Closed Chamber (SCC) and Eddy Covariance (EC) Methods on the Deilmann South Waste Rock Pile (DSWR) 117 4.3.1 Introduction 117 4.3.2 D C C Fluxes 118 4.3.3. S C C Fluxes 118 4.3.4 EC Fluxes 123 4.3.4 Summary of the advantages and disadvantages of the dynamic closed chamber (DCC) method 129 C H A P T E R 5: Ana lys is and D iscuss ion 131 5.1. Introduction 131 vi 5.2 Effects of Rainfall Events on Waste-Rock Surface Water Conditions and C 0 2 Fluxes Across the Surfaces of the Deilmann North (DNWR) and Deilmann South (DSWR) Waste Rock Piles 131 5.2.1 Short Term Effects of Rainfall Events on Near Surface-Water Conditions 132 5.2.2 Short Term Effects of Rainfall Events on C 0 2 Fluxes 138 5.3 Predictions of Evaporative Fluxes and Near-Surface Water Contents Profiles 142 5.3.1 Short Term Predictions of Evaporative Fluxes 144 5.3.2 Short Term Predictions of Near Surface Water Contents Profiles 150 5.4 C 0 2 Diffusion Prediction and Model Proposed 153 5.4.1 C 0 2 Diffusion 153 5.4.2 Biotic C 0 2 Production Rate 159 5.4.3 Development of the Partial Differential Equation 163 5.4.4 Finite Difference Formulation 166 5.5 Computer Code Program 169 5.6 Application of the "C02" Model Using Measured Values in Sand Minicosms ..173 5.6.1 Prediction of C 0 2 Concentration Profiles in Response to Changes in Water Contents Profiles 173 5.6.2 Simulations of C 0 2 Concentrations Profiles using Sand Minicosms Measured Data 176 5.7 Prediction of C 0 2 Diffusion and Concentration-Depth Profiles in Response to Changes in Water-Depth Profiles in the DSWR 181 vii 5.8 Predictions of C 0 2 Diffusion and Surface C 0 2 Flux from the DNWR and D S W R Piles Following Rainfall Events 188 5.9 Chapter Summary 192 C H A P T E R 6. Summary and Conc lus ions 193 References 197 Append ices 219 Appendix A: Measuring 0 2 Fluxes Using the Dynamic Closed Chamber (DCC) System 219 Appendix B: Eddy Correlation Method: a Brief Theory 227 Appendix C: Computer Code for C 0 2 Diffusion Model 231 Appendix D: Waste-Rock Samples Analyses Results 244 Appendix E: C 0 2 Flux Measurements Results obtained at the Deilmann south (DSWR) and Deilmann north (DNWR) Waste-rock Piles using the Dynamic Closed Chamber (DCC), Static Closed Chamber (SCC) and Eddy Covariance (EC) Methods 252 Appendix F: Data for Measurements of Water Contents and C 0 2 Fluxes Across the Surfaces of the DSWR and DNWR Piles after Rainfall Events 257 Appendix G : Minicosms Data used for Validation of the C 0 2 Diffusion Model 259 Appendix H: Climatic Parameters used in Simulations with SoilCover and recorded at the weather station installed on the Deilmann south waste-rock (DSWR) pile 262 Appendix I: SoilCover Run Summary Pages for Simulations of Evaporative Fluxes At the DSWR and DNWR Piles during the Field Tests 254 V l l l List of Tables Page Table 2.1 In-situ thermal conductivity measurements in waste-rock dumps material I 5 Table 2.2 In-situ air permeability measurements in waste-rock dump material... 15 Table 2.3 In-situ oxygen diffusion coefficient measurements in waste-rock dumps 15 Table 2.4 Typical physical properties of A R D waste-rock piles (Ritchie, 1994a). 1 6 Table 2.5 Physicochemical properties of the Doyon and Nordhalde waste rock piles (Lefebvre et al., 2001) 1 7 Table 2.6 Typical characteristics of A R D (Ritchie, 1994a) 18 Table 2.7 Summary of CO2 concentrations for waste-rock material 28 Table 2.8 Summary of C 0 2 concentrations for non-waste-rock material 31 Table 2.9 Summary of surface C 0 2 fluxes for non-waste-rock material 33 Table 4.1 Nature, origin, and basic geotechnical properties of various granular materials 87 Table 4.2 Nature and origin of data for the hydraulic conductivity "k" value of various granular materials 103 Table 4.3 Summary of results of C 0 2 flux measurements using the dynamic closed chamber system (DCC) for the test period of 2000-2002 at Deilmann south waste-rock pile (DSWR) 112 Table 4.4 Summary of results of C 0 2 flux measurements using the dynamic closed chamber system (DCC) for the test period of 2000-2002 at Deilmann north waste-rock pile (DNWR) 113 List of Figures Page Figure 2.1 Diagram describing the complex flow system develop within the waste-rock dump in response to precipitation and climatic conditions 9 Figure 2.2 Conceptual model of water flow and vapour transport in a waste- rock dumps 11 Figure 2.3 (A) Depth geologic profile for Deilmann south waste-rock (DSWR) pile (Adapted from Birkham et al., 2003 and (B) Map showing the Deilmann north (DNWR), Deilmann south (DSWR) and outline of former lake bed at the Key Lake mine, Saskatchean, Canada 22 Figure 2.4 0 2 consumption rates vs C 0 2 production rates for forest soils, lake bottom sediments, and gneissic waste rocks units: u.mol/kg/week; x represents gne iss ic waste rocks (DNWR); A represents lake bottom sed iments col lected from beneath the waste-rock pile (DNWR): • represents forest soi ls (natural forest site adjacent to the DNWR) (Lee et al., 2003b) 23 Figure 2.5 (A) Relation of evaporation (flux) to time under different evaporativities. (B) Relation of relative evaporation rate (actual rate as a function of the potential rate) to time, indicating the three stages of the drying process 42 Figure 3.1 Photographs showing the mechanical sieve machine and sedimentation process setups 55 Figure 3.2 Schematic diagram and photograph of Tempe cell setup for measurement of soil water characteristic curve 57 Figure 3.3 Schematic diagram and photograph of permeameter cell setup for measurement of saturated hydraulic conductivity 60 Figure 3.4 Schematic diagram and photographs of the columns (mesocosm and minicosms) used to calibrate and verifify the dynamic closed chamber method 63 Figure 3.5 Map of Saskatchewan showing the location of the Key Lake uranium mine, Saskatchewan, Canada 67 Figure 3.6 Photograph showing the Deilmann pit, the Deilmann north waste- rock (DNWR) and Deilmann south waste-rock (DSWR) piles at the Key Lake uranium mine, Saskatchewan, Canada 69 Figure 3.7 Depth geologic profile for Deilmann south waste-rock (DSWR pile at the Key Lake mine, Saskatchewan, Canada (Adapted from Birkham etal. , 2003) 71 Figure 3.8 Map of the Deilmann north waste-rock (DNWR) and Deilmann south waste-rock (DSWR) piles at the Key Lake mine, Saskatchewan, Canada, showing the chambers and the meteorological weather station locations 73 Figure 3.9 Schematic diagram and photograph of the dynamic closed chamber (DCC) setup for surface C 0 2 flux measurements 75 XI Figure 3.10 (A) Typical slopes of direct measurement of concentration versus time using the dynamic closed chamber (DCC) method (B) schematic diagram and (C) photograph of the D C C method setup for measuring surface C 0 2 gas fluxes 78 Figure 3.11 Photograph showing the meteorological weather station and the eddy covariance (EC) sensors for measuring C 0 2 flux installed on Deilmann south waste-rock (DSWR) 81 Figure 3.12 (A) Schemat i c d iagram and (B) photograph of meteorological weather station instal led on Deilmann south waste-rock (DSWR) pile at the Key Lake mine, Saskatchewan, Canada 85 Figure 4.1 Particle size distribution curves (without gravel and boulder-sized) for the samples of waste-rock from Deilmann south waste-rock pile (DSWR) for ground surface sand (curve with symbols) and core sand/sandstone (curves with full lines). Symbols represent the measured data from this thesis. The full lines show the one standard deviation range of grain-size data obtained by Birkhman et al. (2002) 88 Figure 4.2 Particle size (without gravel and boulder-sized) distribution curves for the samples of waste-rock from Deilmann north waste-rock pile (DNWR) for ground surface sand (curve with broken line and symbols) and core basement-rock (curves with solid lines). Symbols represent the measured data from this thesis. The full lines show the one standard deviation range of grain-size data obtained by Birkhman et al. (2002) 89 x i i Figure 4.3 Water retention curve (WRC) of the sample of waste-rock (with fine fraction only) from the Deilmann north waste-rock (DNWR) pile. Symbols represent the measured data and the solid line the best fit curve generated with SoilCover (SoilCover, 1997) 92 Figure 4.4 Water retention curve (WRC) of the sample of waste-rock (with fine fraction only) from the Deilmann south waste-rock (DSWR) pile. Symbols represent the measured data and the solid line the best fit curve generated with SoilCover (SoilCover, 997) 93 Figure 4.5 Water Retention Curve (WRC) of the sample of waste-rock from the Deilmann north (DNWR) pile: Symbols represent the measured data and the solid line represents the best fit curve generated with SoilCover(SoilCover, 1997) 95 Figure 4.6 Water Retention Curve (WRC) of the sample of waste-rock from the Deilmann south (DSWR) pile: Symbols represent the measured data for and the solid line represens the best fit curve generated with SoilCover (SoilCover, 1997) 96 Figure 4.7 Characteristic of the sample of the waste-rock from the Deilmann north waste-rock pile (DNWR): hydraulic conductivity curve (K). The value of saturated hydraulic conductivity (K s a t) was measured in the laboratory but the unsaturated hydraulic conductivity (K) was derived from the Brooks and Corey mode (Brooks and Corey, 1964) 101 x i i i Figure 4.8 Characteristic of the sample of the waste-rock from the Deilmann north waste-rock pile (DNWR): hydraulic conductivity curve (K). The value of saturated hydraulic conductivity (K s a t) was measured in the laboratory but the unsaturated hydraulic conductivity (K) was derived from the Brooks and Corey mode (Brooks and Corey, 1964) 102 Figure 4.9 Short-term (hourly) variations in the C 0 2 flux measured at DSF1 on August 6, 2000. Fluxes were determined using the dynamic closed chamber (DCC) method and averaging a series of four to eight measurement cycles, with each cycle lasting from 2- to 8-min (depending on the magnitude of the flux). The shaded box represents the 95% confidence interval (±17 mg C 0 2 rrf2 h"1) around the calculated daily mean (235 mg C 0 2 nrf2 h"1) 106 Figure 4.10 (A) C 0 2 fluxes measured using the dynamic closed chamber (DCC) at twenty selected sampling stations (DSF1 - DSF20) at the Deilmann south waste-rock pile (DSWR) (Figure 3.8) during the summers of 2000 and 2002 (B) average flux values (mg C 0 2 m"2 h" 1) measured from sampling locations (•) on the DSWR 108 xiv Figure 4.11 Daily variations in the C 0 2 flux measured at the Deilmann south waste rock (DSWR) pile in (A) July, (B) August, and (C) September 2000. Flux measurements were obtained at three to four locations on each sampling date. At each location, the flux was determined using the dynamic closed chamber (DCC) method and averaging a series of four to eight measurement cycles, with each cycle lasting from 2- to 8-min (depending on the magnitude of the flux). The overall mean for each monthly sampling period is represented by the dashed lines ( ). Within months, symbols labeled with the same letter are not significantly different at the P < 0.05 level of probability 109 Figure 4.12 Box & Whisker plot characterizing the spatial and long-term temporal variability in the C 0 2 flux measured using the dynamic closed chamber (DCC) method at the Deilmann south waste-rock (DSWR) pile in 2000 and 2002. The estimated, time-averaged flux = 170 (±51) mg C 0 2 rrf2 h"1. The minimum and maximum flux values are marked by asterisks (*). Note: values occurring beyond the "whiskers" were identified as outliers and were not included in the analysis of variance 1 1 1 Figure 4.13 C 0 2 fluxes measured using the dynamic closed chamber (DDC) at nine sampling stations (DNF1 - DNF9) at the Deilmann north waste-rock pile (Figure 3.6) during the summers of 2000 and 2002: (A) Data points presented on a XY (scatter) and (B) average flux values (mg C 0 2 rrf2 h"1) from samplings locations on the DNWR. ... 1 1 4 X V Figure 4.14 Spatial and temporal variations in C 0 2 fluxes measured during the summer of 2000 and summer 2002: box-and-wisker plots showing the mean, standard deviation, and extreme values for Deilmann north waste-rock (DNWR) pile data 116 Figure 4.15 Box-and-wisker plot for flux measurements obtained using the DCC method at the Deilmann south waste-rock (DSWR) pile during the period from August 24 t h to August 25 t h , 2002 (set of data for comparison with the other two methods: S C C and EC). The minimum and maximum flux values are marked by asterisks (*). Note: values occurring beyond the "whiskers" were identified as outliers and were not included in the analysis of variance 119 Figure 4.16 (A) C 0 2 flux values (mg nrT2 h"1) obtained using the static closed chamber (SCC) at eleven selected sampling stations (•) at the Deilmann south waste-rock (DSWR) pile in the morning (AM) (between 10:00 and 11:00) and afternoon (PM) between 16:30 and 17:30) on August 24, 2002 (B) averages fluxes (mg C 0 2 m 2 h"1) from the sampling locations 120 Figure 4.17 Box-and-wisker plots for C 0 2 flux measurements obtained from the Deilmann south waste rock pile on August/24/2002. Measurements were obtained using the static closed chamber method between the hours of 10:00 and 11:00 AM and 16:30 and 17:30 P M . The minimum and maximum flux values are marked by asterisks *. Note: values occurring beyond the "whiskers" were identified as outliers and were not included in the analysis of variance 121 XVI Figure 4.18 Comparison of the dynamic closed chamber (DCC) and static closed chamber (SCC) methods for measuring C 0 2 fluxes. Flux measurements were obtained at the Deilmann south waste-rock (DSWR) pile site during the period from August 24 t h to August 25 t h , 2002. The minimum and maximum flux values are marked by asterisks (*). Note: values occurring beyond the "whiskers" were identified as outliers and were not included in the analysis of variance 122 Figure 4.19 Diurnal variations in the C 0 2 flux measured from 10:00 to 17:00 on August 25, 2002 using the EC method at the Deilmann south waste-rock (DSWR) pile. The shaded box represents the 95% confidence interval (± 24 mg C 0 2 nrf2 h"1) around the calculated daily mean (150 mg C 0 2 m"2 h"1) 124 Figure 4.20 Measured C 0 2 fluxes using Eddy covariance (EC) at the Deilmann south waste-rock (DSWR) pile. Measurements were obtained on a continuous basis during the period from June 25 t h to August 25 t h 2002. Each data point represents the daily mean value averaged over the period from 10:00 to 17:00 hours: The shaded box (B) represents the 95% confidence interval (±10 mg C 0 2 m~2 h"1) around the overall mean (150 mg C 0 2 m~2 h"1). Note: gaps in the data set represent precipitation events during which no useful data were collected by the EC system 1 2 5 XVII Figure 4.21 Comparison of the eddy covariance (EC) and chamber-based methods for measuring the C 0 2 flux from the Deilmann south waste-rock (DSWR) pile 127 Figure 5.1 Rainfall and volumetric water contents measured over an 8-day test period [July 30 (day 1) to August 6 (day 8), 2002] at station DNF1 with time at the Deilmann north waste-rock (DNWR) pile 133 Figure 5.2 Rainfall and volumetric water contents measured over an 8-day test period [July 30 (day 1) to August 6 (day 8), 2002] at sampling station DSF1 with time at the Deilmann south waste-rock pile (DSWR) 134 Figure 5.3 Volumetric water content profiles measured over an 8-d test period (30 July (day 1) to 6 August (day 8) 2002 at station: (A) DSF1 at the Deilmann south waste-rock (DSWR) pile and (B) DNF1 at the Deilmann north waste-rock (DNWR) pile with time 137 Figure 5.4 Rainfall, water contents, and C 0 2 fluxes measured at sampling station DNF1 over an 8-d test period (30 July (day 1) to 6 August (day 8) 2002) at the Deilmann north waste-rock (DNWR) pile with time 139 Figure 5.5 Rainfall, water contents, and C 0 2 fluxes measured at sampling station DSF1 over an 8-d test period (30 July (day 1) to 6 August (day 8) 2002) at the Deilmann south waste-rock (DSWR) pile with time 140 xvi i i Figure 5.6 Variations in C 0 2 flux measurements with surface-water saturation (S=8/n) measured over an 8-d test period [July 30 (day 1) to August 6 (day 8), 2002] at DSF1 and DNF1 at the Deilmann south waste-rock pile (DSWR) and Deilmann north waste-rock pile (DNWR) respectively 143 Figure 5.7 (A) Rainfall, water contents measured, and SoilCover predicted evaporative fluxes at the Deilmann north waste-rock (DNWR) pile (B) ratio of actual (AE) and potential (PE) evaporation as a function of time over an 8-d test period [30 July (day 1) to 6 August (day 8) 2002] 145 Figure 5.8 (A) Rainfall, water contents measured, and SoilCover predicted evaporative fluxes at the Deilmann north waste-rock (DSWR) pile (B) ratio of actual (AE) and potential (PE) evaporation as a function of time over an 8-d test period [30 July (day 1) to 6 August (day 8) 2002] 146 Figure 5.9 SoilCover predicted evaporative fluxes (actual A E and potential (PE) and (B) ratio of A E / P E at the Deilmann south waste-rock pile (DSWR) over a 27-d test period (29 July (day 1) to 24 August (day 27) 2002) with time 148 Figure 5.10 SoilCover predicted evaporative fluxes (actual A E and potential (PE) and (B) ratio of A E / P E at the Deilmann north waste-rock pile (DNWR) over a 27-d test period (29 July (day 1) to 24 August (day 27) 2002) with time 149 xix Figure 5.11 Comparison of (A) measured and (B) SoilCover simulated water content profiles and (B) SoilCover predicted water contents for a 6- d test period [30 July (day 3) to 4 August (day 8), 2002] at the Deilmann north waste-rock (DNWR) pile 151 Figure 5.12 Comparison of (A) measured and (B) SoilCover simulated water content profiles and (B) SoilCover predicted water contents for a 6- d test period [30 July (day 3) to 4 August (day 8), 2002] at the Deilmann south waste-rock (DSWR) pile ; 152 Figure 5.13 (A) Hypothetical water content profiles obtained by reducing the initial water profile (d1) by a factor of 0.8 consecutively (B) Simulated effective diffusion coefficient (De) of C 0 2 as a function of water content using artificial data presented above 158 Figure 5.14 Simulated microbial respiration rates as a function of temperature and water content using Equation 5.21 163 Figure 5.15 Representative elementary volume, REV, for derivation of partial differential equation 164 Figure 5.16 Three nodes and the mass fluxes entering and exiting node 1 for development of the finite difference formulation 167 Figure 5.17 Flowchart for Visual Basic program 170 Figure 5.18 Stability curves generated by the model for different iterations using time steps of 0.05 day 172 X X Figure 5.19 (A) Hypothetical water contents profiles in a sand material described in Figure 5.13A (B) Model predicted C 0 2 concentrations profiles in a HT sand column obtained with hypothetical simulated water contents profiles (Figure 5.19A) and an initial measured C 0 2 concentrations profiles (dl) in HT column (Kabwe et al., 2002) 175 igure 5.20 Measured volumetric water content profiles in the low (A) temperature (LT) (thermostat set at 5 °C) and (B) high (21 temperature (HT) (room tem[perature) minicosms. V represent the water table 177 Figure 5.21 Measured C 0 2 concentration profiles in the (A) high temperature (HT) (21 - 23 °C) and, low temperature (LT) (5 °C) minicosms (Richards, 1998; Kabwe, 2001) 179 Figure 5.22 Model predicted C 0 2 concentration profiles in the (A) high temperature (HT) (21 - 23 °C) and (B) low temperature (LT) (5°C) minicosms 180 Figure 5.23 Relationship between measured and simulated C 0 2 concentrations from (A) low temperature (LT) and (B) high temperature (HT) minicosms plotted on a 1:1 scale 182 Figure 5.24 Depth profiles for Deilmann south waste-rock (DSWR) pile (A) Geologic profile (B) mean C 0 2 concentration (Vol.) and (C) mean volumetric water contents values (Adapted from Birkham et al, 2003) 184 XXI Figure 5.25 (A) Hypothetical water-depth profiles in DSWR pile and (B) model predicted effective diffusion coefficients (De) in response to changes in water contents (Figure 5.32A). Curve d1 represents the actual measured mean water- depth profile (Birkham et al., 2003). The subsequent profiles were generated by reducing the initial measured water-depth profile by a factor of 0.1 consecutively 186 Figure 5.26 Model predicted changes in: (A) effective diffusion coefficient (De) and (B) C 0 2 concentrations profiles in response to changes in water contents profiles described in Figure 5.32A 187 Figure 5.27 Rainfall, measured surface water content and C 0 2 flux and predicted effective diffusion coefficient (De) and surface C 0 2 flux at the Deil;mann North was;te-rock (DNWR) pile over an 8-day test period [30 July (day 1) to 6 August (day 8) 2002] with time 189 Figure 5.28 Rainfall, measured surface water content and C 0 2 flux and predicted effective diffusion coefficient (De) and surface C 0 2 flux at the Deilmann South waste-rock (DSWR) pile over an 8-day test period [30 July (day 1) to 6 August (day 8) 2002] with time 190 XXll Abbreviation and Symbols 1 . Abbreviations Meaning A Area Arh Inverse of relative humidity of air AE Actual evaporation AEV Air entry value B R N Inverse of relative humidity of soil surface C Concentration C p Specific heat capacity C u Uniformity coefficient C v Specific heat of the soil C h Volumetric specific heat of the soil C w Coefficient of consolidation with respect to water phase CV Coefficient of variation D* Bulk diffusion coefficient D e Effective diffusion coefficient D a Diffusion coefficient through air phase D H Equivalent particle diameter D w Diffusion coefficient through water phase D v Diffusion coefficient of water vapor Dvap Molecular diffusion of water vapor DAS Data acquisition system DCC Dynamic closed chamber DNF# Collar location # at Deilmann north DSF# Collar location # at Deilmann south DNWR Deilmann north waste-rock DSWR Deilmann south waste-rock E Vertical Evaporatve flux E a Actual evaporation E p Potential evaporation Eh Sensible heat flux XXlll EC Eddy covariance e Void ratio e s Vapour pressure at the soil surface e a Vapour pressure of the air above the evaporating surface f(u) Wind mixing function Fco2 Flux of carbon dioxide g Acceleration due to gravity G Production rate Go Reference production rate G r Ground heat flux G s Groung heat flux h Height h r Relative humidity of the soil surface "w Total head Hh Pressure head H Henry's Law coefficient HT High temperature k Hydraulic conductivity kr Constant in Arrhenius equation ks Saturated hydraulic conductivity L Pore-size distribution index U Latent heat of evaporation of water m Total mass of gas P Total Pressure PE Potential evaporation P s Saturation vapour pressure of the soil PSv Saturation vapour pressure of soil P v Vapor pressure within the soil q Humidity Qn Net radiant energy available at the surface R Universal gas constant REV Representative elementary volume Rc Concentration ratio of two isotopes xxiv S Degree of saturation S r Residual saturation S e Effective saturation SCC Static closed chamber SWCC Soil water characteristic curve t time T Temperature T a Air temperature T s Surface temperature T r a Tortuosity coefficient for air Trw Tortuosity coefficient for water phase U a wind speed U* Friction velocity. V Volume V s Volume of voids V v Volume of solids W v Molecular weight of water w Vertical wind XXV 2. Symbols Meaning a Tortuosity factor of the soil (3 Cross-sectional area of the soil available for vapor transfer A Slope of the saturation vapor pressure versus temperature curve y Psych rometric constant p w Mass density of water p s Mass density of soil p a Density of dry air X Thermal conductivity of the soil v|/ Total suction \|/r Residual suction 6 Total porosity 0 eq Equivalent porosity 6 a Air porosity 8 w Water porosity 8 Factor representing the ratios of 1 3 C and 1 2 C xxvi Acknowledgements I would like to express my sincere gratitude to my supervisor Dr. Ward Wilson for the time and energy he has expended, along with his willingness to share his expertise, experience, guidance, encouragement and support throughout the course of my MSc and this PhD thesis works. His constant enthusiasm, positive attitude and Christian faith were a source of inspiration for me. My thanks are also extended to all members of my Advisory Committee: Dr Scott Dunbar, Dr. Bern Klein and Dr Marek Pawlik for their valuable suggestions and constructive feedbacks. I gratefully acknowledge the assistance and guidance of my former MSc thesis co- supervisor and co-author, Dr. Jim Hendry from the University of Saskatchewan. Cogema Resources Ltd, Cameco Corporation, and Natural Sciences and Engineering Research Council of Canada are acknowledged for providing financial support through an Industrial Research Council (IRC). The assistance by the mine staff of Cameco at the Key Lake uranium mine, and Ray Kirkland and Tyler Birkham from the University of Saskatchewan, is gratefully acknowledged. My sincere thanks are also extended to Dr Henrique Rubio and Dr Bruno Bussiere who reviewed my thesis draft. May those who have contributed either directly or indirectly to this work, especially, my sister Bernadette Kapembe and Dr Musangu Ngeleka, and whom I have failed to mention individually, find in this work the fruit of their conjugated efforts (none mentioned, none forgotten). Finally I would like to thank my wife Florence Chabu Kabwe for her patience, understanding and for the support throughout the course my studies. My son Fiston, my daughter Berthe, and my grand-daughter Joy Makayla were a source of joy and inspiration for me. xxvii Dedication To my late mother Saya Kasuba and my late father Jonas Mumba Katele: their dreams have been fulfilled. (Romans 5: 1-5) ...knowing that suffering produces endurance, and endurance produces character, and character produces hope, and hope does not disappoint us, because God's love has been poured into our hearts through the Holy Spirit that has been given to us. To my Creator and Saviour, Jesus Christ. xxviii Chapter I: Introduction Page 1 CHAPTER I INTRODUCTION 1.1 Introduction Accurate measurements and predictions of surface gas fluxes such as CO2 and O2 are of great importance to the mining engineers and/or researchers in the development of a long-term management plan for reactive mine waste dumps. Measurements of C 0 2 fluxes are needed to quantify biogeochemical reaction rates in unsaturated geologic media and soils (Hendry et al., 1993, 1999, 2001; Wood et al., 1993; Wood and Petraitis, 1984; Affek et al., 1998; Keller and Bacon, 1998; Lee et al., 2003, Birkham et al., 2003). These flux measurements can also provide needed input for global warming models (Hanson et al., 1993; Sundquist, 1993; Holland et al., 1995; Sellers et al., 1995; Thierron et al., 1996; Wickland and Striegl, 1997; Buchmann et al., 1999). Quantification of gas (e.g., O2) diffusion rates can be used to determine the extent of sulfide oxidation in unsaturated waste-rock piles (e.g., Harries and Ritchie, 1985; Davis and Ritchie, 1987; Hockley et al., 2000; Timms and Bennett, 2000; Bennett et al., 2003; Molson et al., 2005) and mine tailings impoundments (Elberling and Nicholson, 1996; Wunderly et al., 1996; Elberling et al., 2000; Elberling and Damgaard, 2001). Oxygen gas diffusion rates can also be used to establish how effective soil cover systems would be at reducing gas diffusion into the waste rock and tailing profile (Harries and Ritchie, 1985; Yanful et al., 1993a; O'Kane et al., 1995; Smolensky and Hockley, 1999; Aubertin et al., 2000; Timms and Bennett, 2000; Mbonimpa et al., 2002, Chapter I: Introduction Page 2 2003; Bussiere et al., 2002, 2003; Cook et al., 2004; Martin et al., 2006; Aubertin et al., 2006). Soil C 0 2 flux is a complex process controlled by biotic and abiotic factors (Buchmann, 2000). The presence of C 0 2 also exerts an important control on the pH of the pore water in unsaturated zones (Lowson et al., 1982; Neal and Whitehead, 1988). The C 0 2 dissolved in water has a major influence on water chemistry (Neal and Whitehead, 1988) and soil acidification (Elberling and Jakobsen, 2000) and it drives carbonic acid weathering of silicate and carbonate minerals (Reardon et al., 1979). Over the past two decades considerable attention has focused on radiatively important biogenic trace gas such C 0 2 because of the concern of global warming (Blake and Rowland, 1988; Matson and Harris, 1995; Trumbore et al., 1996, Brooks et al., 1997; Fahnestock et al., 1998; Hobbie et al., 2000; Burkins et al., 2001). Similarly, because of the concern over acid rock drainage (ARD), most studies involving pore gases in mine wastes (including waste-rock piles) have focused on 0 2 consumption rates as an indication of the rate of sulfide mineral oxidation. To date, there has been comparatively little attention paid to the C 0 2 side of biotic and abiotic gas production and fluxes from the subsurface C-horizon soils and mine waste dumps. Monitoring of 0 2 - C 0 2 fluxes may provide a practical tool for identifying and understand the different important mechanisms in the waste dumps such as the zones of microbial respiration and pyrite oxidation-carbonate buffering in mine waste dumps (Lee et al., 2003) as well as providing an indication as to the extent of microbial activity in the waste dumps. Fluxes of C 0 2 from waste-rock piles, though important in determining geochemical reactions rates, are poorly characterized and standards establishing the accuracy of field measurements of C 0 2 surface fluxes are lacking Chapter I: Introduction Page 3 (Nakayama, 1990; Norman et al., 1992; Rayment and Jarvis, 1997; Janssens et al., 2000; Scott-Denton et al., 2003). In previous work, the author (Kabwe, 2001, and Kabwe et al., 2002) developed and verified in mesocosms a technique to measure C 0 2 fluxes from the soil surface to the atmosphere. The technique termed the dynamic closed chamber (DCC) method is based on direct measurement of the change in C 0 2 concentration with time in the headspace of a chamber installed on ground surface over a relatively short period of time. The D C C method was shown to accurately measure C 0 2 surface fluxes from ground surface to the atmosphere in mesocosms. This is a direct technique of measurements and it provides an almost instantaneous indication of the reaction rate under field conditions, regardless of climatic or moisture conditions in the waste dumps. This laboratory-verified technique provided the opportunity to quantify temporal and spatial C 0 2 fluxes under field conditions and at the same time, compare these fluxes measurements to those obtained from two other methods: the static closed chamber (SCC) and eddy covariance (EC) methods. This thesis work presents the results of the field applications of the D C C to quantify reaction rates and other processes at work in mine waste-rock piles at the Key Lake uranium mine in northern Saskatchewan. The work provided a complete suite of measurements required to characterize spatial distribution of C 0 2 fluxes on waste rock. The author is not aware of any larger-scale studies that quantify C 0 2 fluxes across a waste-rock pile. Chapter I: Introduction Page 4 1.2 Research Objectives The main objective of this thesis was to extend the application of a novel and laboratory-verified device (the dynamic closed chamber D C C system) designed and developed by the author (Kabwe, 2001; Kabwe et al., 2002) for CO2 flux measurements under field conditions on the DSWR and DNWR mine dumps at the Key Lake mine in northern Saskatchewan. It should be noted that this thesis represents the Phase II work of a 'Collaborative Research Program in the Mining Industry for Waste-Rock Hydrology', between the University of Saskatchewan and the University of British Columbia and funded by Cameco Mining and Cogema Resources. Phase I of the research work (Kabwe, 2001, and Kabwe et al., 2002) involved the design and testing of a dynamic closed chamber method for measurements of COaflux in mesocosms. The specific objectives of this thesis were: (1) To compare the D C C field CO2 fluxes data with those obtained using two other field soil respiration techniques: static closed chamber (SCC) and eddy covariance (EC) methods. (2) To measure the drying rate on surfaces of waste-rock piles after rainfall events (3) To predict evaporation on the surfaces of waste-rock piles using the SoilCover (Unsaturated Soils Group, 1997) computer model. (4) To predict changes in water content profiles on waste-rock piles after rainfall events using SoilCover computer model. Chapter I: Introduction Page 5 (5) To design and develop a numerical model for CO2 gas production and diffusion in unsaturated materials. (6) To validate the CO2 model using measured data. (7) To use the CO2 model to predict C 0 2 diffusion and concentration- depth profiles in the waste-rock piles in response to changes in water- depth profiles at the Key Lake mine. (8) To use the CO2 model to predict the effects of rainfall events on the surface effective diffusion coefficient and surface CO2 flux on the waste-rock piles at the Key Lake mine. The work presented in this thesis applies the D C C method that was previously developed and verified in mesocosms to measure C 0 2 flux from ground surface to the atmosphere (Kabwe et al., 2002) and was shown to accurately measure CO2 surface fluxes. In this study the D C C will be tested under field conditions to quantify and determine biogeochemical reaction rates in waste-rock piles. The method can be of great value to the mine engineer in the development of closure designs for mine waste- rock piles at the Key Lake mine. The method can also be extended to other mine waste dumps to quantify biogeochemical reaction rates in unsaturated geologic media and soils at other mine sites in Canada and world wide. 1.3 Organization of the Thesis Chapter 2 of the thesis presents a literature review for studies of subsurface C 0 2 and O2 production and consumption rates and the associated fluxes in waste-rock Chapter I: Introduction Page 6 systems and other soil ecosystems. The chapter ends with a brief review of climatic variables affecting the gas fluxes: precipitation and evaporation. Chapter 3 presents material characterizations. The objective was to determine the soil properties and characteristics of near-surface waste rock which influence the CO2 gas flux. The tests conducted include: grain size analysis, water retention curve (WRC) (or soil water characteristic curve), and saturated hydraulic conductivity. Chapter 4 discusses results of field application of the dynamic closed chamber (DCC) method for measurements of CO2 fluxes. The chapter ends with a comparison of the DCC fluxes to those obtained from two other methods: static closed chamber (SCC) and eddy covariance (EC) methods. Chapter 5 presents results of the investigations for the climatic variables affecting CO2 fluxes (e.g., effects of rainfall and evaporation on soil moisture and CO2 fluxes) precipitation and evaporation. The " C 0 2 " diffusion model developed in this work is discussed at the end of Chapter 5. The theoretical background, development of the partial differential equation including verification of the model developed are also presented. Final conclusions are summarized in Chapter 6. Chapter II: Literature Review Page 7 CHAPTER II Literature Review 2.1 Introduction This chapter presents a literature review for studies of subsurface CO2 and O2 production, consumption rates and the associated surface fluxes in waste-rock systems and establishes the need for the research. Section 2.2 presents a physical description of a waste-rock pile to help conceptualize the complex flow developed within the waste rock in response to climatic conditions. Section 2.3 discusses the sources of CO2 in waste-rock piles, more specifically on biotic and abiotic reactions in waste-rock piles. Section 2.4 presents results of studies for C 0 2 production and O2 consumption rates and surface fluxes in waste-rock and non-waste-rock systems reported in literature. Section 2.5 provides a discussion on the two important climatic variables that affect subsurface and surface gas fluxes namely, precipitation and evaporation. The chapter summary is presented in section 2.6. Chapter II: Literature Review Page 8 2.2 W a s t e - R o c k P i l e s Waste-rock piles from mining operations are constructed from the excavation and surface deposit of overburden rock, which commonly contain sulfide minerals. These above-ground, coarse-grained deposits tend to be heterogeneous in structure as a result of placement methods (e.g., end dumping, lift placement and compaction). The volume of a waste rock dump can be as large as hundreds of millions of cubic meters, making its size several kilometers in width and hundreds of meters in depth. These unsaturated and exposed to atmospheric precipitation, energy (e.g., solar energy) and gases (oxygen, carbon dioxide) and have the potential for generating sulfuric acid ( H 2 S 0 4 ) in the presence of sulfide minerals (e.g., F e S 2 , Fe-|.xS) (Nordstrom and Alpers, 1999; Keit and Vaughan, 2000; Rimstidt and Vaughan, 2003; Elberling, 2005) (see Figure 2.1). This acid can dissolve heavy metals in the mine waste and produce acid rock drainage (ARD), which is potentially toxic to plants, animals and humans. The oxidation rate of sulphide minerals (e.g., pyrite, FeS 2 ) depends on a number of factors which define the environment within the waste-rock dumps, including temperature, pH, oxygen concentration, chemical composition of the pore water, and microbial population (Ritchie, 1994). Oxidation is an exothermic process that produces a large amount of heat (Elberling, 2005). Field measurements show that the temperature can be as much as 63 K warmer than the atmosphere in waste-rock piles with heights of 20-30 m (Harries and Ritchie, 1981; Gelinas and Choquette, 1992). Another factor that complicates the oxidation process is the presence of bacteria. Certain species of bacteria (e.g., Thiobacillus ferrooxidans) were found to increase the rate of oxidation by two orders of magnitude (Lorenz and Tarpley, 1963; Brierley, 1978). Chapter II: Literature Review Page 9 FLUXES PROCESSES Energy Water Gas 0 2 & C 0 2 l 1 1 Hydrology Reaction Rates Geochemistry C H 2 0 + 0 2 = H 2 0 + C02( B ) F e S 2 ( s ) + 7/20 2 + H 2 0 = F e 2 + + 2SO4"4 + 2 H + FeS2(S) + C a C 0 3 ( s ) + 7 / 2 0 2 ( g ) = F e 2 + + C a / + + 2SO4"4 + C O ^ , 2+ ISSUES Measurement Scale » Environmental Loadings Time Evolution Figure 2.1 conceptual models of the complex flow system and pollutant generation and transport processes developed within a pyritic waste-rock dump in response to precipitation and climatic conditions. Chapter II: Literature Review Page 10 In short, the physico-chemical-microbial environment within a waste-rock dump determines the sulfide mineral oxidation rate, which in turn determines the physico- chemical-microbiological environment (Ritchie, 1994). Figure 2.2 shows a conceptual model of water and gas flow, and vapour transport in a waste-rock dump and internal structure and material segregation of a waste rock pile (Herasymuik, 1995; Fala et al., 2005). MEND (2001) proposed four hydrostratigraphic types to characterize waste-rock piles. The types differ depending upon the materials and construction methods, and the characteristics of flow. The four types are: i. Non-segregated coarse-grained rock piles that transmit water rapidly to the base of the pile; ii. Non-segregated fine-grained rock piles that are likely to contain a basal saturated zone; iii. The segregated rock piles that contain a fine-grained crest zone that may not permit the passage of significant quantities of water; and iv. Layered, segregated dumps that contain a finer-grained crest and sandy gravel layers to the face of the rock pile. Segregated waste-rock dumps exhibit a graded stratigraphy caused by the segregation that occurs as materials roll down the pile at the angle-of-repose (Figure 2.2A). Finer sandy gravels are present at the crest, while coarser materials accumulate further down-slope. Dawson and Morgenstern (1995) have shown that when materials consisting mostly of finer sandy gravel are end-dumped, little segregation occurs and a finer grained layer is formed in the dump. A layer of finer material is typically found on the surface of the rock and a layer of coarser material is found at the base (Fala et al., Chap te r II: Literature Rev iew P a g e 1 1 natural soil or rock Figure 2.2. Conceptua l model of internal structure and material segregat ion of a waste rock pile. Chapter II: Literature Review Page 12 2005). The bulk grain size distribution then includes alternating fine and coarse-grained material layers (see Figure 2.2) (Morin et al., 1991). When the grain size distribution is more variable, the vertical pile profile can be irregular and the structure of alternating fine and coarse layers less distinct. Layering inside the pile can be locally enhanced by construction traffic (heavy equipment), which tends to crush and compact the surface material, creating layers that can be up to 1 m thick (Aubertin et al., 2002; Martin, 2004; Fala et al.,2005). The grain and pore size distribution within a waste rock piles affects its hydraulic properties, which in turn control internal flow. Preferential flow can be caused by continuous macropores or by vertical, horizontal, or inclined layers of relatively high hydraulic conductivity that often control the movement of water within a pile. When the layering occurs as a fine-grained unit above a coarse-grained zone, a capillary barrier is formed in which water is preferentially retained in the fine grained material due to capillary forces (Nicholoson et al., 1989; Buissiere et al., 2003; Fala et al., 2005). Capilarry barriers have ben proposed for use in waste rock piles to control air and water flow (Poulin et al., 1996; Lefebvre et al., 2001b). The unsaturated condition and heterogeneous, coarse-grained nature of the waste rock deposits often make geochemical and geotechnical parameters difficult to measure (Pantelis et al., 2002). The problem is exacerbated when the dump is > 10 m high, as this is as high as a column of water can be supported by atmospheric pressure. For these reasons, few data on the chemical composition of pore water within waste-rock dumps are in the literature. Monitoring the chemistry of the gas1 phase in unsaturated environments is a relatively easy process. The gas can be easily drawn from sampling ports within the pile and the technology to measure chemically important components (i.e. 0 2 and CO2) is Chapter II: Literature Review Page 13 readily accessible (Russell and Appleyard, 1915; Enock and Dasberg, 1971; de Jong and Schappert, 1972; Rightmire and Hanshaw, 1973; Rightmire, 1978; Hass et al., 1983; Jaynes et al., 1983a; Wallick, 1983; Wood and Petraitis, 1984; Harries and Ritchie, 1985; Ceding et al., 1991; Gelinas et al., 1992; Elberling et al., 1993; Hendry et al., 1993, 2001; Ritchie, 1994a, 1994b; Elberling and Nicholson,1996; Lee, 1997; Keller and Bacon, 1998; Russell and Voroney, 1998; Helgen et al., 2000; Hockley et al., 2000; Kabwe etal . , 2002). The O2 and CO2 concentrations in the pore gas can be expected to vary because there are sinks and sources for these within the waste-rock dumps. The source of 0 2 in a waste-rock dump is at the outer surface of the dump. 0 2 concentrations vary with increasing distance into a dump in a manner that depends on the prevailing O2 transport mechanisms and on the oxidation rates. It should be noted that 0 2 concentrations less than 0.2% mole fraction have been measured (Bennett et al., 1999), and values as low as 0.01% mole fraction have been reported (Goodman et al., 1992). C 0 2 concentrations in the pore space of a waste-rock dump can range up to about 20% mole fraction and are frequently in the range 1-10% (Harries and Ritchie, 1983; Schuman et al., 1992). This is much higher than atmospheric levels of 0.03%. Elevated concentrations of C 0 2 increase the oxidation rate of pyrite by moderate thermophiles (Norris, 1989), some workers have reported increased growth rate of thiobacillus ferrooxidans with increasing levels of CO2 (Holuigue et al., 1987; Beyer et al., 1990), but others have reported little change up to 7% (Kelly and Jones, 1978; Norris, 1989). Haddadin et al. (1993) observed that increased C 0 2 concentrations increased the pyrite oxidation rates, but that concentrations of 4% were inhibitory to all three of the microbial populations involved in pyrite oxidation in the system studied. Chapter II: Literature Review Page 14 The monitoring of O 2 and C 0 2 concentrations in pore gas is commonly used as an indication of the occurrence, location, rate and type of chemical reactions occurring in subsurface environments (Ritchie, 1994). 0 2 consumption by sulphide mineral oxidation is of particular importance in assessing the impact of waste-rock piles on the environment because of the associated acid generation (Molson et al., 2005). A decrease in 0 2 concentrations in waste-rock piles does not necessarily indicate the occurrence of sulphide mineral oxidation. Organic oxidation is another common, but less environmentally harmful, process during which 0 2 is consumed. C 0 2 in pore gas is an indicator of the types of oxidation processes occurring. For example, for a given amount of 0 2 consumed, carbonate buffering of acid generated from sulphide mineral oxidation will typically produce less C 0 2 than organic oxidation (discussed further in this Chapter). In addition, carbonate minerals and organic molecules have different ratios of stable carbon isotopes (discussed briefly in section 2.3): consequently, the stable carbon isotope signature of pore gas C 0 2 may be used to trace the C 0 2 source (Hendry et al., 2002). Ritchie (1994a) and Lefebvre et al. (2001) provided a summary of typical physical properties and typical characteristics of A R D waste-rock with data on bulk properties of waste-rock dumps (Tables 2.1, 2.2, 2.3, 2.4 and 2.5). The gas concentration and temperature profiles in a waste-rock dump depend on the magnitude of a number of physical properties of the dump material, including air permeability, the gas diffusion coefficient, and the thermal conductivity. Ritchie (1994a) pointed out that the set of data in these Tables indicate that, at least for the parameters measured Chapter. 11: Literature Review Page 15 Table 2.1. In-situ thermal conductivity measurements in waste-rock dumps material Mine site location # of measurement | Range points in waste dump | (Wm'1 K"1) P Average (Wm-1 K"1) Aitik mine, Swedenx 8 0.71-1.63 1.2 +_0.4 Heath Steele, Canada x 3 1.04-1.22 1.2 + 0.1 Kelian, Kalimantanx 7 1.57-3.31 2.1 +0.6 Rum Jungle, Australia" 6 1.77-3.12 2.2 + 0.5 Doyen mine, Canada y 6 2.5 Nordhalde, Germanyy 8 1.0 Table 2.2. In-situ air permeability measurements in waste-rock dumps material I Mine site location j # of measurement B points in waste dump Range (m2); Aitik mine, Sweden" 27 (0.6 +0.2) x 1fJ 1 1 - (1.4 +0.1) x 10"a Heath Steele, Canada" 24 (1.6+ 0.15) x 10"11 - (4.7+ 0.5) x 10"9 Kelian, Kalimantan" 18 (3.9 + 0.1)6) x 10"9 j Rum Jungle, Australiay 144 (8.89 + 0.19) x 10"13 - (1.49 +_0.21) x 10"9 Doyen mine, Canada y 8.1 x10" 1 0 J Nordhalde, Germanyy I 2.5 x 10"12 Table 2.3. In-situ oxygen diffusion coefficient measurements in waste-rock dumps Mine site location # of measurement points in waste dump Range nrrV^xlG- 6) , Aitik mine, Sweden" 2 (2.25 + 1.04)-(6.85-1.02) Heath Steele, Canada" 3 (2.65+ 0.55)-(3.35-0.25) Woodlawn, Australia" 2 (3.49+ 1.64)-(5.07-0.39) Doyen mine, Canada y j 6 2.85 Nordhalde, Germanyy j 8 5.70 X = Ritchie, 1994; y = Lefebvre et al., 2001 Chapter II: Literature Review Page 16 Table 2.4. Typical physical properties of ARD waste-rock piles (Ritchie, 1994a) Property Unit Typical value Approximate Range of values Height m 20 2 to 150 j Area ha 30 0.1 to 150 Density Kg/m a 1500 1300 to 1900 Sulfur content as pyrite Wt. % 2 0.5 to 30 Climate type Tropical to polar Rainfall m/yr 0.1 to 5 Water content within dump Vol. % 10 (at infiltration of 0.5 m yr"1) 5 to 25 I Porosity % 40 Carbonate density 0.6 kg m"a 0.04 % O 2 diffusion coefficient m 2/s 5 x 10' 6 2 x 10 - 6 to 6x10" 6 Air permeability m 2 | 1x10" 1 2 to 1x10"9 Temperature within J dump °C -7 to 65 Chapter II: Literature Review Page 17 Table 2.5. Physicochemical properties of the Doyon and Nordhalde waste rock piles (Lefebvre etal . , 2001) Propert ies Unit Doyen Nordhalde Volume of waste rocks m* 11.5x10 b 27.0x10 b Maximum thickness m 35 80 Main Rock Type Sericite schists Slates Solid density [ Kg/m a 2740 2751 Porosity Dim. | 0.00 0.30 Average water saturation % 42 63 Effective vertical air permeability m 2 8.1x10- 1 u 2.5x10" 1 2 Water infiltration rate m/year 0.350 0.166 J Average thermal j conductivity W/m °C | 2.5 1.0 Effective oxygen diffusivity IT^/S 2.13x10^ 2.13x10"b J Range of Temperature within dumps °C 1 - 6 5 3-16 Sulfur content as pyrite % j 1 - 2% Chapter II: Literature Review Page 18 Table 2.6. Typical characteristics of A R D (Ritchie, 1994a). I Property Typical associated chemical species Range Impact I Acidity (pH) Sulfuric acid 2 to 4 Mobilization of metal ions Iron j Ferrous and ferric ions; ferric oxides, hydroxides; jarosites Concentration 100 to 3000 mg L"1 Discoloration and turbidity in receiving waters as pH increases and ferric salts precipitate Heavy metals Cu, Mn, Zn, Cd , Hg, Pb, As , Ra 1 to 200 mg L"1 Reduction in aquatic flora and fauna; bioaccumulation; reduction in quality of potable groundwater supplies Total dissolved salts C a , Mg, Al , S 0 4 2 " | Reduction in quality of | potable groundwater 100 to 30000 | supplies; reduction in mg L"1 J quality of water supplies | for livestock Chapter II: Literature Review Page 19 (though small data-set) the heterogeneity of the dump material and layering consequent on the method of dump formation do not carry through in any marked way. These bulk properties are required both to predict the environmental conditions within a dump, and to quantify oxidation rates. The extents to which these properties vary from place to place in a dump and from dump to dump provide some insight on the impact that dump-construction methods and details of dump composition have on these bulk properties (Fala et al., 2005). The gas concentration and temperature profiles in a waste-rock dump depend on the magnitude of a number of physical properties of the dump material, including air permeability, the gas diffusion coefficient, and thermal conductivity. It was noted that the variation from dump to dump (from this data set) was about the same as that within a dump and that gas transport in a dump was dominated by diffusion when the air permeability was 10" 1 0 m 2 or less (Bennett et al., 1989; Pantelis and Ritchie, 1991a). 2.3 Sources of C 0 2 in Subsurface Soils and Waste-Rock Piles The generation of soil C 0 2 flux is a complex process controlled by biotic and abiotic factors (Buchmann, 2000; Shi et al., 2006). Gas-filled pores in soil typically contain 10-100 times higher concentrations of C 0 2 than the atmosphere (Welles et al., 2001), primarily due to soil C 0 2 production from respiration in living roots and heterotrophic soil microorganisms (Elberling, 2003). C 0 2 in pore gas may be used to identify its source. The ratio of stable carbon isotopes ( 1 3 C / 1 2 C ) in C 0 2 from pore gas indicates if the C 0 2 source is organic, inorganic, or a combination of both. It should be noted that the carbon isotopes technique was used by Birkham et al. (2003) to determine the source of C 0 2 in the waste-rock piles (DNWR & DSWR) and beneath at Chapter II: Literature Review Page 20 the Key Lake mine. Birkham et al. (2003) presented measured values of carbon isotopes ratio for the waste-rock piles at the Key Lake mine and concluded that pore gas CO2 in the DNWR pile (see Figure 2.3) likely originated from a combination of organic (biotic) and inorganic (abiotic) sources (Birkham et al., 2003). Carbon isotopes ratio values for the DSWR indicated the majority of pore gas CO2 from DSWR originated from an organic source underlying the waste rock (e.g., Figure 2.3, dewatered lake) (Birkham et al., 2003). 2.3.1 C 0 2 production by microbial respiration (biotic) Microbial aerobic respiration and oxidation of organic matter are generally considered to be the primary sinks for O2 and the main sources of elevated biogenic C 0 2 concentrations in the subsurface. Rates of aerobic microbial degradation of organic matter and contaminants in the subsurface are greater than anaerobic degradation (Hendry et al., 2002). An understanding of the physical transport mechanisms and the biochemical processes that control C 0 2 and O2 concentrations and fluxes to and within the subsurface are needed (Bennet and Ritchie, 1990; Hendry et al., 1999; Hendry et al., 2002; Pantelis et al., 2002; Lee et al., 2003b). O2 consumption and CO2 production by microbial respiration in unsaturated media can be represented by the general biotic reaction (Stumm and Morgan, 1981; Lee et al., 2003b): C H 2 0 + 0 2 -> C 0 2 ( g ) + H 2 0 [2.2] where C H 2 0 represents a simple carbohydrate. In this simple case of organic oxidation Chapter II: Literature Review Page 21 one mole of 0 2 consumed results in the production of one mole of C O 2 . More complex organic molecules (e.g., C106H263O-110N16P) may have molar ratios of 0 2 consumpt ion to C O 2 production of c loser to 1:0.77 (Drever, 1997): Cio6H263011oN16P + 1 3 8 0 2 ( g ) - > 2 + [2-3] 1 0 6 C O 2 ( g ) + I 6 N O 3 + H P O J " + 1 2 2 H 2 0 + 1 8 H + B a s e d on the Equat ions 2.2 and 2.3, respiratory consumpt ion of 1 mol of 0 2 should produce 0.8 or 1 mol of C 0 2 . It was noted in the literature review that mine waste-rock piles are, in some c a s e s , constructed upon organic carbon-r ich dewatered lake bottoms (Birkham et a l . , 2003 ; L e e et a l . , 2003b) (see Figure 2.3). Microbial respiration in these buried deposi ts can also consume 0 2 and produce C 0 2 . Lee and co-workers (Lee et a l . , 2003) found that these st iochiometric ratios are very similar to those observed for microbial respiration in forest soi ls ( 1 O 2 : 0 . 7 C O 2 ) (see Figure 2.3B) and in buried lake sediments beneath mine waste-rock piles ( 1 O 2 : 0 . 5 C O 2 ) . They found a positive correlation between the rates of 0 2 consumpt ion and C 0 2 production and organic carbon content (i.e., higher organic carbon contents in forest soil than lake bottom sediments) and suggested that the difference in 0 2 / C 0 2 ratios were due to dif ferences in the stoichiometry of the organic carbon. Other researchers (Amundson et a l . , 1988; W a n g et a l . , 1999) reported positive correlation between respiration rates and organic carbon content in unsaturated zones . Measurements of 0 2 - C 0 2 f luxes, therefore, may provide an indication of the zones of respiration and the extent of microbial activity in the waste- rock pile. Chapter II: Literature Review Page 22 Figure 2.3. (A) Depth geologic profile for Deilmann south waste-rock (DSWR) pile. (Adapted from Birkham et al., 2003) (B) Map showing of the Deilmann north (DNWR), Deilmann south (DSWR) and outline of former lake bed at the Key Lake mine, Saskatchewan, Canada. Chapter II: Literature Review Page 2 3 Figure 2.4. 0 2 consumption rates vs C 0 2 production rates for forest soils, lake bottom sediments, and gneissic waste rocks (units: umol/kg/week; x represents gneissic waste rocks (DNWR); A represents lake bottom sediments collected from beneath the waste-rock pile (DNWR): • represents forest soils (natural forest site adjacent to the DNWR) (Lee et al., 2003b). Chapter II: Literature Review Page 24 2.3.2 C 0 2 production by pyrite oxidation-carbonate buffering (abiotic) Soil C 0 2 derived from unsaturated mine waste-rock piles can also be produced in abiotic (e.g. sulfide minerals) reactions in situ (Elberling and Nicholson, 1996; Timms and Bennett, 2000; Birkham et al., 2003; Lee et al., 2003b). If gaseous 0 2 is present in the unsaturated waste-rock piles, the oxygen can be consumed by microorganisms in the chemical oxidation of minerals (e.g. pyrite) and can lead to formation of acid and sulfate (Stumm and Morgan, 1981; Ritchie, 1994b; Lee et al., 2003b): FeS 2 ( s ) + 3.50 2( g) + H 2 0 Fe(ll)+ 2 S O ^ + 2 H + [2.4] The resulting Fe(ll) in Equation 2.4 can be oxidized to Fe(lll) by : 2Fe(ll) + 0.5O 2( g) + 2 H + 2Fe(lll) + H 2 0 [2.5] Combining Equations 2.4 and 2.5 we obtain: 2FeS 2 ( s ) + 7 .50 2 ( g ) + H 2 0 2Fe(lll) + 4 S O j ~ + 2 H + [2.6] In a solution with pH > 3, F e 3 + can precipitate from solution to produce additional acid (Dubrovsky et al., 1984; Janzen et al., 2000) by: Fe(lll) + 3 H 2 0 Fe(OH) 3 + 3 H + [2.7] Precipitation of other Fe(lll)-bearing phases, such as goethite (a-FeOOH) or schwertmannite (Fe808(OH)6S04), may occur in acid mine waters (Bigham et al., 1990). Alternatively, Fe(lll) can be consumed through further oxidation of sulphide minerals in acidic water (Wiersma and Rimstidt, 1984; Blowes et al., 1995) by: Chapter II: Literature Review Page 25 F e S 2 + 1 4 F e 3 + + 8 H 2 0 -> 1 5 F e 3 + + 2 S 0 2 . - +16H + [2.8] Carbonate minerals are often present in natural subsurface environments and have a buffering effect on the pH of subsurface pore water (Freeze and Cherry, 1979). The acid generated by pyrite oxidation can dissolve available carbonates to produce C 0 2 gas by: C a C 0 3 ( s ) + 2 H + -> C a 2 + + H 2 0 + C 0 2 ( g ) [2.9] Combining Equations 2.4 and 2.9 yields Equation 2.10: F e S 2 + C a C 0 3 +3.50 2( g) ->Fe( l l )+Ca 2 + + 2 S 0 2 . - + C 0 2 ( g ) [2.10] In addition, combining Equations 2.6 and 2.9 yields Equation 2.11 2 F e S 2 + C a C 0 3 +7 .50 2 ( g ) ->2Fe( l l l )+Ca 2 + + 4 S 0 2 T + C 0 2 ( g ) [2.11] Furthermore, combining Equations 2.6, 2.7, and 2.9 yields Equation 2.13 for near neutral pH solution: FeS 2 ( s ) + 2 C a C Q 3 + 3 .750 2 ( g ) +1.5H 2 0 -> Fe(OH) 3 + 2 C a 2 + + 2 S 0 2 . - + 2 C 0 2 ( g ) [2.12] Based on Equations 2 . 1 0 - 2 . 1 2 , consumption of 1 mol of 0 2 by pyrite oxidation with carbonate buffering may produce 0.1, 0.3, or 0.5 mol of C 0 2 and between 0.5 and 0.6 mol of sulphate. The C 0 2 produced is thus an indirect measure for the carbonate buffering and an indicator of the types of oxidation processes occurring. It is therefore suggested that both sulfide oxidation-carbonate buffering and microbial respiration may control 0 2 and C 0 2 gas concentrations in the unsaturated waste-rock piles (Lee et al., 2003b). C 0 2 produced in unsaturated soils and waste-rock piles undergoes redistribution via gas transport and geochemical reactions with water and various mineral phases Chapter II: Literature Review Page 26 (Hendry et al., 1993) and will diffuse upward to the atmosphere (soil respiration) and downward to the water table under concentration gradients. Microbially produced CO2 can dissolve in the recharging water and react to produce bicarbonate (HCO3). These species are then transported to the water table in the dissolved state. In addition to sinks attributed to CC>2-carbonate mineral reactions, CO2 also dissolves in water to produce carbonic acid. Although this latter flux is small compared to the soil C 0 2 efflux, it has a major influence on water chemistry (Neal and Whitehead, 1988) and soil acidification (Elberling and Jakobsen, 2000) and it drives carbonic acid weathering of silicate and carbonate minerals (Reardon et al., 1979; Elberling, 2003). Many studies have shown that the dominant sink for CO2 from unsaturated zones is the atmosphere (Solomon and Cerling, 1987; Hendry et al., 2001). For example, Hendry et al. (1993) showed that 2% of the C 0 2 produced in a 3.2 m thick sandy unsaturated zone under high recharge conditions was removed by the recharging ground water. Solomon and Cerling (1987) determined that about 4% of the CO2 produced in an unsaturated zone was removed by the recharging ground water. The dominant mechanism for gas transport in soil pores is generally accepted to be concentration controlled molecular diffusion through air-filled pores (Keen, 1931; Grable, 1966; Weeks et al., 1982; Elberling, 2003). Variations in pore gas composition due to thermal convection and atmospheric pressure variations were observed by Harries and Ritchie (1985), Bell et al. (1991), Hockley et al. (2000), Lefebre et al. (2001), and Molson et al. (2005). While diffusion is typically limited to a near-surface zone of a few meters depth, advection (due to a thermal gradient and/or wind pressure gradients or barometric pumping) and barometric pumping have the potential to move air (and oxygen) to much greater depths into the pile. In general, the more permeable Chapter II: Literature Review Page 27 general, the more permeable the waste-rock material, and the greater the height-to- width ratio of the waste-rock pile, the greater is the potential for advective air movement. The reactivity of the waste-rock material as well as the coarsenes (hence air permeability), and the spatial variability of these properties within a pile, have a strong influence on the magnitude of thermally induced advection (Wels and Robertson, 2003). In contrast, air movement due to barometric pumping is controlled by the waste rock porosity, changes in ambient air pressure and the heterogeneity of air permeability of the waste-rock dump. 2.4 Studies of C 0 2 in Subsurface Pore Gas and Associated Surface Gas Fluxes from Waste-Rock and non Waste-Rock Systems The following sections present literature review of studies of CO2 in subsurface pore gas and associated surface gas fluxes from waste-rock and non waste-rock systems. It should be noted that this literature review serves to establish the need for surface CO2 fluxes measurements on both mine waste-rock and non-waste-rock systems. This is because the dominant oxidation reactions and associated C 0 2 fluxes measured at the DSWR occur in the organic material underlying the waste-rock pile (Birkham et al., 2003). Birkham et al. (2003) also suggested that pyrite oxidation- carbonate buffering and the resulting O2 consumption and C 0 2 production are more likely to be observed in the gneissic waste rocks at the DNWR. Many studies have investigated C 0 2 in pore gas in subsurface and surface gas fluxes for non waste-rock material and only few attempts have been made to quantify C 0 2 production or surface flux (Birkham et al., 2003) in waste-rock piles. Chapter II: Literature Review Page 28 2.4.1 Studies of subsurface C0 2 gas and surface gas fluxes from waste- rock piles The following section along with Tables provide the literature review results for typical studies of C 0 2 production and consumption rates and surface fluxes for waste- rock piles. Table 2.7. Summary of C 0 2 concentrations for waste-rock material. Sources Locations Waste Rock: size and geologic material C 0 2 maximum concentration (%) Harries and Ritchie (1985) Rum Jungle Australia 15 to 25 m high waste-rock piles silty sand to rocks, 1 to 3% pyrite > 20 Gelinas et al. (1992) La Mine Doyen, Quebec 30 to 35 m high waste-rock pile, 3.5 to 4.5% pyrite 7 Hockley et al. (2000) Germany 1 to 3% sulphides, high cone, of carbonates 60 Birham et al. (2003) Key Lake mine Saskatchewan Canada 20 to 28 m high, sand/sandstone 8 Harries and Ritchie (1985) measured pore gas C 0 2 and 0 2 concentrations in the within a pyritic waste-rock pile (average height 25 m) at the Rum Jungle uranium mine in Australia to identify oxidation zones and measure rates of oxidation. The waste-rock pile consisted mainly of pyritic, graphitic shale. C 0 2 concentrations varied from near Chapter II: Literature Review Page 29 atmospheric levels to greater than 20 %. 0 2 concentrations varied from near atmospheric levels to 0 %. Important conclusions of this study were that 0 2 supply was a rate-limiting factor for oxidation and that both diffusion and advection (due to thermal and atmospheric effects) resulted in gas migration through the pile. Advection due to thermal effects (buoyancy forces) was significant in regions where temperatures were elevated (>50 °C) relative to monthly mean temperatures (25 to 30 °C). Advection due to changes in atmospheric pressure was observed at depths of up to 7.5 m where 0 2 concentration fluctuations matched the semidiurnal changes in atmospheric pressure. Gelinas et al. (1992) studied the physico-chemical conditions for La Mine Doyon in Quebec. A R D from the waste-rock piles had pH values around 2 and total dissolved solids (mainly sulfates, Fe, Al and Mg) values up to 200 000 mg/L; pyrite concentrations in the waste-rock piles were approximately 3.5 to 4.5 % (by mass). Porosity of the piles and the dry bulk density were estimated at 35 % and 1850 kg/m 3, respectively. Maximum temperatures typically ranged from 40 to 50°C, C 0 2 pore gas concentrations typically increased to about 7 % at a depth of 30 m and 0 2 pore gas concentrations typically decreased to about 2.5 % at a depth of 30 m. Air convection (due to thermal effects) was identified as a key gas transport process as air venting from the waste piles was observed during cold weather. Hockley et al. (2000) measured temperature, C 0 2 and 0 2 pore-gas concentrations, and air pressure in a waste-rock pile at a uranium mine in Germany. Typical seepage from the waste-rock pile had a pH of 2.7 and sulfate concentrations above 10 000 mg/L. Acid generating material had sulfide mineral concentrations ranging from 1 to 3 % (by mass). Thermal convection of pore gas was observed at some sites during winter months as stable temperatures deeper in the pile were Chapter II: Literature Review Page 30 between the summer and winter ambient temperatures. Gas transport due to barometric pressure fluctuations was also observed. C 0 2 concentrations typically increased (up to 60%) with increasing depth. C 0 2 production was attributed to high concentrations of carbonate material. 0 2 concentrations typically decreased (down to 0%) with increasing depth; Birkham et al. (2003) measured CO2 concentrations profiles, C 0 2 consumption and production rates, and C 0 2 fluxes from two waste-rock (the Deilmann south waste- rock (DSWR) and Gaertner (GWR) piles at the Key Lake Uranium Mine in northern Saskatchewan. The concentrations exhibited a linear increase for C 0 2 in concentrations with depth through the piles and suggested that the dominant sites of reactions occurred below the piles. Mean C 0 2 concentrations at the DSWR changed little with depth (change in C 0 2 concentrations less than 1 % from atmospheric concentrations). C 0 2 concentration increased from 10 to 20 m, decrease from 20 to 30 m, and increased from 30 to 40 m. They found that oxidation of organic matter beneath the waste-rock pile dominated the pore-gas chemistry and that significant changes in 0 2 and C 0 2 concentrations within mine waste piles may not be the result of sulfide mineral oxidation/carbonate buffering. The gas consumption and production values ranged between 0.04 and 0.15 ug C 0 2 / g soil/day C 0 2 . C 0 2 concentration depth profiles at the Key Lake mine were similar to those presented in other waste-rock studies (Harries and Ritchie, 1985; Hockley et al., 2000) in that C 0 2 concentrations were negatively correlated to 0 2 concentrations. Although C 0 2 concentrations increased to a maximum of 20 % at the Rum Jungle mine in Australia (Harries and Ritchie, 1985), 0 2 + C 0 2 values were usually less than 15 %. CO2 concentration depth profiles presented by Hockley et al. (2000) had CO2 concentrations Chapter II: Literature Review Page 31 that usually increased to much greater than 20 % (up to 60 %) at depths below 20 m; 02+C0 2 values ranged from approximately 5 % to approximately 60 %. In summary, waste-rock studies in the literature indicated that only one attempts have been made to quantify C 0 2 production and C 0 2 concentration depth profiles (Birkham et al., 2003) and surface flux. A need, therefore, exists for measurements of surface C 0 2 fluxes for large waste-rock piles. 2.4.2 Studies of subsurface C0 2 from non waste-rock material The following section along with Tables provide the literature review results for typical studies of C 0 2 production and consumption rates and surface fluxes for non waste-rock piles. It should be noted that the range of non-waste-rock systems is wide, and only selected literature review results will be presented. Table 2.8. Summary of C 0 2 concentrations for non-waste-rock material. Source Location Waste rock size and geologic material Max. C0 2 cone. (%) De Jong and Schappert (1972) Canadian prairie (1.5 m of unsaturated heavy clay) 2.26 Rightmire and Hanshaw (1973) Florida (sand, forest and grassland) Atkinson (1977) England (limestone soils, depths to 130 m) 1.8 Chap te r II: Literature Rev iew P a g e 32 Table 2.8 Continued. Source Location Waste rock size and geologic material Max. C0 2 cone. (%) Reardon et a l . (1979) Ontar io (up to 11 m of unsaturated, ca lca reous s a n d , forest region) 0.8 J a y n e s et a l . (1983b) Eas te rn United States (recla imed coa l strip mine) 18.7 H a a s et a l . (1983) North Dakota) Great Pla ins(greater than 13 m of ca lca reous c laystone and si l tstone, lignite present, vegetated. 19 to 20 Wa l l i c k (1983 ) A lber ta ( less than 13 m thick unsaturated z o n e , rec la imed coa l mine a rea , high carbonate content in s o m e areas) 24 W o o d and Petrait is (1984) Southern High P la ins , T e x a s (51 to 77 m of ca lca reous geo log ic material) 3.02 S o l o m o n and Cer l ing (1987) Utah (approximately 2 m of unsaturated montane soi l , vegetated) 1.24 W o o d et a l . (1993) Southern S a s k a t c h e w a n (7 m of unsaturated silt loam/ti l l , vegetated) 3 T rumbore et a l . (1995) Eas te rn A m a z o n i a (45 m of unsaturated clay, forest and pastureland) 7 L e e (1997) Massachuse t t s (0.5 to 12 m of unsaturated sand) 5 Chapter II: Literature Review Page T a b l e 2 .9 . Summary of surface C 0 2 fluxes for non waste-rock material. Source Location/geologic material C 0 2 surface flux (mmol C/m2/day) De Jong and Schappert(1972) Canadian prairie (at least 1.5 m of unsaturated heavy clay), (d=1710 kg/m3) Up to 241 Wood and Petraitis (1984) Southern High Plains, Texas (51 to 77 m of calcareous geologic material), (d=1710 kg/m3) 2.5x10^ to 1.2x10"2 Solomon and Cerling (1987) Utah (approximately 2 m of unsaturated montane soil, vegetated) (d=2070 kg/m3) 7.48x10"2 to 0.64 Wood et al. (1993) Washington state (loess, vegetated) d=1869 kg/m3) 9.63x10"4 to 8.18x10"2 Wood et al. (1993) Southern Saskatchewan (7 m of unsaturated silt loam/till, vegetated)(d=2056 kg/m3) 0 to 2.58x10"3 Trumbore etal. (1995) Eastern Amazonia (45 m unsatuareted clay, forest and pastureland) 220 to 580 Lee (1997) Massachusetts (0.5 to 12 m of unsaturated sand) 19.6 (low veg.) 372 (golf course) 50 (woodland) 123 (grassy area) Russell and Voroney (1998) central Saskatchewan (calcareous till, forest) 53 to 807 Hendry etal. (1999) Southern Saskatchewan (5.75 m of unsaturated sand 30.8 (average) Chapter II: Literature Review Page 34 de Jong and Schappert (1972) calculated CO2 production rates and CO2 surface flux in the Canadian prairies using a Fickian approach (using measured CO2 concentration depth profiles and an estimated diffusion coefficient). The maximum CO2 concentration measurement within the top 1.5 m was 2.26 %. CO2 production rates ranged up to 241 /jg C/g/day and the surface flux of CO2 ranged up to 1 473 mmol C/m 2/day. The geologic material was heavy clay. Russell and Voroney (1973) measured CO2 surface fluxes ranging from 53 to 807 mmol C/m 2/day in central Saskatchewan (forest region). Root respiration was estimated to contribute 60 % of the C 0 2 production and a strong correlation was found between CO2 surface flux and temperature, pore-gas C 0 2 concentration in the humus layer (surficial organics) and moisture content. The geologic material was medium to fine-textured, medium to strongly calcareous, glacial till. Volumetric soil moisture content ranged from about 10 to 35 %. Atkinson (1977) measured pore gas CO2 concentrations in a limestone environment in England. CO2 concentrations increased up to 1.8 % with measurements taken at depths as great as 130 m. Oxidation of down-washed organic matter was noted as a possible source of C 0 2 at depth. Maximum total carbon content was 11 % in the limestone soils. Reardon et al. (1979) measured pore gas CO2 concentrations values in a forest area in Ontario. CO2 % generally increased with depth (up to between 0.3 and 0.8 %). The water table was 6 to 11 m deep and the total porosity was approximately 0.38. At one site a C 0 2 concentration gradient at the water table was observed indicating that the groundwater was a source of CO2 (degassing). A C 0 2 concentration gradient at the water table was not observed at the other site. Chapter II: Literature Review Page 35 Jaynes et al. (1983a) measured pore-gas CO2 concentrations at a reclaimed coal strip mine in eastern United States. CO2 concentrations ranged from near atmospheric to greater than 15 %. O2+CO2 values were less than 20.9 %, and O2 and CO2 concentrations were negatively correlated. 0 2 and CO2 concentrations were weakly correlated to temperature. The average bulk dry density was 1560 kg/m 3. The average pyritic S concentration was 0.18 % (by mass, ranging from 0.02 to 2.0 %); the average C concentration was 4.4 % (by mass, ranging from 0.4 to 22.8 %). Haas et al. (1983) measured pore-gas CO2 concentrations, at 8 sites in North Dakota (Great Plains region). C 0 2 concentrations fluctuated seasonally indicating a relationship to root respiration. Organic and lignite (coal) oxidation were dominant sources of C 0 2 in the pore gas. Wallick (1983) measured pore-gas C 0 2 concentrations in the Battle River Mine area in Alberta to indicate when a reclaimed mined area had reached geochemical equilibrium with unmined landscapes. C 0 2 concentrations ranging from atmospheric to 24 % and originated from both carbonate dissolution ( C 0 2 degassing) and organic oxidation. Wood and Petraitis (1984) calculated O2 consumption rates, CO2 production rates and C 0 2 surface fluxes at two sites on the Southern High Plains of Texas. O2 consumption rates varied from approximately 2.3 x 10"3 to 2.5 x 10~2 \xg 0 2 /g/day (assuming bulk dry density of 1735 kg/m3); CO2 production rates varied from approximately 2.5 x 10"4 to 1.2 x 10~2 p,g C/g/day; and surface fluxes were 2.2 mmol C/m 2/day and 10.5 mmol C/m 2/day. CO2 production was calculated as a function of depth and gas migration was attributed to diffusion. C 0 2 concentrations increased with depth (up to approximately 1 and 2.5 %); O2 concentrations decreased with depth Chapter II: Literature Review Page 36 (down to approximately 19 and 16 %). Although the sites contained significant amounts of calcium carbonate (up to 80 % in many zones), C 0 2 production was attributed to oxidation of organic particles deep in the unsaturated zone. An important conclusion of this study was that a very small amount of C 0 2 production at depth had a more profound effect on the geochemistry of the system than a similar production rate near the surface. The unsaturated zones studied were 51 m (gas probes to 36 m) and 77 m (gas probes to 21 m) thick. Solomon and Cerling (1987) calculated C 0 2 production rates ranging from 7.48 x 10"2 to 0.64 JJQ C/g/day in a montane soil in Utah using the concentration-gradient approach. Pore gas C 0 2 increased to 1.24 % during the growing season and also during the winter due to the capping effect of the snowpack. C 0 2 in the pore gas was noted as an important factor in the weathering of certain minerals (i.e. albite). Also, C 0 2 removal by dissolution in infiltrating water was determined to be minimal (less than 4 % of annual budget of C 0 2 in pore gas); as much as 15 % of the C 0 2 in pore gas could be removed during periods of low C 0 2 production. Wood et al. (1993) calculated C 0 2 production rates in Washington state (9.63 x 10"4 to 8.18 x 10"2 fjg C/g/day) and south central Saskatchewan (0 to 2.58 x 10"3 /jg C/g/day) considering diffusive fluxes and partitioning of C 0 2 into infiltrating water; C 0 2 sorption onto solid phase was considered negligible. C 0 2 production rates were reasonably correlated with temperature and microbial abundance at the Washington site; no correlation was observed at the Saskatchewan site. Production at the Saskatchewan site was found to be isolated within 1 m above the water table (approximately 6.5 m below ground surface) and to within the top 2 m during the growing season (root respiration). The water table at the Washington site was 5 to 6 m Chapter II: Literature Review Page 37 below the ground surface. CO2 and O2 pore gas concentrations were negatively correlated at both sites; C 0 2 % generally increased with depth (up to approximately 2 % in Washington and 3 % in Saskatchewan). CO2 production was attributed to root respiration and the oxidation of soil organic carbon. Total organic carbon (TOC) of the loess at the Washington site ranged from 1.5 % near the surface to 0.03 % deeper in the unsaturated zone; TOC of the till at the Saskatchewan site ranged from 1.3 % near the surface to 0.2 % deeper in the unsaturated zone. Volumetric moisture contents at the Saskatchewan site ranged from 11.78 to 21.88 %. Trumbore et al. (1995) measured CO2 surface fluxes (220 to 580 mmol C/m 2/day) from clay soils in Eastern Amazonia by collecting C 0 2 from the subsurface as it was released. The study sites ranged from pastureland to forest and CO2 production was attributed to both root respiration and oxidation of soil organic matter. Total soil C concentrations ranged from 0.10 to 0.20 % below 3 m to 2.52 to 3.18 % near the surface. Dry bulk density ranged from 960 to 1 220 kg/m 3. CO2 concentrations in the pore gas increased (up to about 8 %) with depth (down to 8 m). Lee (1997) calculated the surface flux of CO2 at four different sites (gravel-pit area, woodland, golf course, and grassy area using a Fickian approach. CO2 surface fluxes and maximum CO2 concentrations were: 19.6 mmol C/m 2/day and 0.7 % (gravel- pit area), 372 mmol C/m 2/day and 5 % (golf course), 50 mmol C7m2/day and 1.1 % (woodland), and 123 mmol C/m 2/day and 3.2 % (grassy area). C 0 2 concentrations were measured to a depth of 3.5 m and generally increased with depth. The geologic material in the study area was sandy (quartz and Na feldspar) with no carbonate minerals present and low organic C content (< 0.1 %). Chapter II: Literature Review Page 38 Hendry et al. (2001) measured pore gas C 0 2 concentrations, and calculated C 0 2 production rates and surface fluxes, and measured field C 0 2 surface fluxes for a 5.7 m thick sandy, unsaturated zone (in central Saskatchewan). Volumetric moisture content ranged from 3 % to over 6 %; mean density and porosity values were 1510 kg/m 3 and 0.43, respectively. C 0 2 concentrations fluctuated seasonally with maximum concentrations occurring in the summer (0.85 to 1.22 %) and minimum concentrations occurring in the winter (0.04 to 0.24 %). A numerical model constrained by measured CO2 concentrations and fluxes, temperature and moisture contents and assuming gas migration due to diffusion was used to calculate C 0 2 production rates between 5 jjg C/g dry soil/day (summer respiration in the soil horizon) and less than 10"4 mg C/g dry soil/day in unsaturated sections of the C horizon. It was also noted that microbial activity ( C 0 2 production) might be very low despite the presence of microorganisms in the unsaturated zone. In summary, studies indicated that C 0 2 pore gas concentrations typically increased (absolute maximum of 24%) with increasing depth. Reardon et al., 1979, Wood and Petraitis, 1984, Solomon and Cerling, 1987, Hendry et al., 1993, Wood et al., 1993, Trumbore et al., 1995, Lee 1997, Hendry et al., 1999, and Birkham et al.2003, all presented C 0 2 concentration depth profiles in which C 0 2 concentrations generally increased with depth. Readon et al., 1979, Wood et al., 1993, Lee 1977, and Hendry et al., 1999 presented concentration depth profiles in which C 0 2 concentrations at shallow depths were elevated due to seasonal root respiration. C 0 2 concentrations were elevated at shallow depths in Hendry et al., 1993 due to increased concentrations of organic matter in the top 0.3 m. Chapter II: Literature Review Page 39 2.5 Climatic Variables Affecting Subsurface and Surface Gas Fluxes: Precipitation and Evaporation Climate has the potential to enhance or reduce soil CO2 fluxes. Precipitation can create changes in soil water content and gases (e.g., C 0 2 ) profiles within unsaturated zones; the extent of the effect depends on the intensity and duration of rainfall (Freeze 1969; Capehart and Carlson 1997). Heavy rainfall events, which close the air pathways to the atmosphere in the upper layers of the soils, may results in an inverted C 0 2 profile for a short period and in lower surface C 0 2 flux (Osozawa and Hasegawa, 1995). Soil C 0 2 flux decreases as the soil moisture decreases (Davidson et al.,1998). The influence of the soil water content on gas flux measurements and diffusion is important only when the soil is at a high water content (Davidson and Trumbore, 1995; Moncrieff and Fan, 1999; Conen and Smith, 2000; Hutchinson et al., 2000). Moreover, Moncrieff and Fan (1999) pointed out that no available theory completely describes the influence of high water content on the C 0 2 flux from each soil layer. Evaporation from mine wastes (tailings and waste rocks) is a crucial component of the water balance, partitioning incoming precipitation into water losses back to the atmosphere and controlling water available for soil moisture storage and deep drainage (Carey et al., 2005). Soil covers are widely used in mine waste (tailings and waste rocks) to prevent the generation of acid. To assess the long-term performance of a cover, it is necessary to study the total water balance, including evaporation (Wilson et al., 1994; Wilson et al., 1997; Yanful et al., 2003a; Carey et al., 2005). For an effective soil cover, the soil must maintain a high degree of saturation (Yanful, et al., 2003a; Aubertin et al., 2006). Soil water evaporation significantly affects water content, and as Chapter II: Literature Review Page 40 a result the degree of saturation of the soil. Therefore, knowledge of the rate of evaporation at the soil-atmosphere interface is required to estimate the water content of candidate cover soils (Wilson et al., 1994; Yanful et al., 2003a; Carey et al., 2005). The following sections briefly present the theory and methods of estimating evaporation, including SoilCover (Unsaturated Soils Group, 1997) computer model. SoilCover was used to estimate evaporative fluxes for comparison with direct measurements of evaporation using eddy covariance (EC) method (Carey et al., 2005). SoilCover model is well-established in the literature, and has been shown by several research to give reasonable accuracy solutions to real-world problems (Rykaart et al., 2001; Scanlon et al., 2002; Noel and Rykaart, 2003; Yanful and Mousavi, 2003a; Yanful et al., 2003b; Vermaak and Beznuidenhout, 2003);. 2.5.1 Evaporation Evaporation involves the change in state of water from a liquid to a vapour. The process occurs when water molecules, which are in constant motion, possess sufficient energy to overcome the surface tension at the liquid surface and escape into the atmosphere (Gray, 1995). The evaporation demand is governed by environmental conditions, such as air temperature, relative humidity, net radiation and wind speed (Wilson, 1990; Unsaturated Soils Group, 1997). Evaporation from soil surfaces is strongly controlled by the water content and water transmission properties of the soil. The rate of movement of water from soil to air depends on the energy gradient and the resistance offered by each pathway through which water moves. Under the same climatological conditions, the evaporation rate can be expected to differ from the rate of evaporation from a free water surface because of the influence of the soil on the mass Chapter II: Literature Review Page 41 and energy exchange processes (Hillel, 1980). For example: 1. The surfaces of soil in a natural environment usually are unsaturated, therefore the vapor pressure is less than the saturation vapor pressure at the surface temperature. 2. The capillary conductivity, which controls the rate of capillary flow of water in an unsaturated soil under a specific energy gradient, is largely a function of the soil moisture content, the size, shape and distribution of the soil pores, and fluid properties. There is a distinct difference between potential (PE) and actual evaporation (AE). The actual rate of evaporation from a soil surface depends on the availability of water (Thornthwaite, 1948; Penman, 1948; Holmes, 1961; Bouchet, 1963; Priestley and Tayor, 1972; Brutsaert, 1982; Morton, 1983; Wilson et al., 1994 and 1997). The maximum potential rate occurs only when the soil surface is fully saturated and water is present on the ground surface. The actual rate of evaporation begins to decline once the soil surface becomes unsaturated. The rate of evaporation continues to decline as the soil surface continues to desiccate. Hillel (1980) showed typical curves for evaporation rates versus drying time for soil (Figure 2.4). The soil drying process has been observed to occur in three recognizable stages (Fisher, 1923; Pearce et al., 1949; Hillel, 1980): (1) An initial constant-rate stage, which occurs early in the process, while the soil is wet and conductive enough to supply water to the site of evaporation at a rate commensurate with the evaporative demand. During this stage, the evaporation rate is limited by, and hence also controlled by, external meteorological conditions (i.e., radiation, wind, air humidity, etc.) Chapter II: Literature Review Page 42 (A) Time Evaporation at Potential (B) / rim** Figure 2.5 . (A) Relation of evaporation (flux) to time under different evaporativities (curves 1 - 4 are in order of decreasing initial evaporation rate). (B) Relation of relative evaporation rate (actual rate as a function of the potential rate) to time, indicating the three stages of the drying process. Chapter II: Literature Review Page 43 rather than by the properties of the soil profile. As such, this stage, being weather controlled, is analogous to the flux- controlled stage of infiltration in contrast with the profile-controlled stage. The evaporation rate during this stage might also be influenced by soil surface conditions. In a dry climate, this stage of evaporation is generally brief and may last only a few hours to a few days. (2) An intermediate falling-rate stage, during which the evaporation rate falls progressively below the potential rate (the evaporativity). At this stage, the evaporation rate is limited or dictated by the rate at which the gradually drying soil profile can deliver moisture toward the evaporation zone. Hence, it can also be called the soil profile-controlled stage. (3) A residual slow-rate stage, which is established eventually and which may persist at a nearly steady rate for many days, weeks, or even months. This stage apparently comes about after the surface-zone has become so desiccated that further liquid-water conduction through it effectively ceases. Water transmission through the desiccated layer thereafter occurs primarily by the slow process of vapor diffusion, and it is affected by the vapor diffusivity of the dried surface zone and by the adsorptive forces acting over molecular distances at the particle surfaces (Hillel, 1980). This stage is often called the vapor diffusion stage and can be important where the surface layer is such that it becomes quickly desiccated. The transition from the first to second stage is generally a sharp one, but the second stage generally blends into the third stage so gradually that the last two cannot be Chapter II: Literature Review Page 44 separated so easily. This can be explained by the fact that during the initial stage, the soil surface gradually dries out and soil moisture is drawn upward in response to steepening evaporation-induces gradients (Hillel, 1980). The rate of evaporation can remain nearly constant as long as the moisture gradients toward the surface compensate for the decreasing hydraulic conductivity (resulting from the decrease in water content). Hillel (1980) noted that since, as the evaporation process continues, both the gradients and the conductivities at each depth near the surface are decreasing at the same time, it follows that the flux toward the surface and the evaporation rate inevitably decreases as well. As shown in Figure 2.4, the end of the first, i.e., the beginning of the second stage of drying can occur rather abruptly. 2.5.2 Methods of predicting, evaporation Evaporation can be calculated with a formulation of Dalton Equation (Wilson et al., 1994): e s = vapour pressure at the soil surface e a = vapour pressure of the air above the evaporating surface. The actual evaporation rate is governed by the vapor pressure difference (e s - e a) and the potential evaporation rate by the vapor pressure difference ( e a - e a ) (for a specific [2.13] Where: f(u) = a wind mixing function Chapter II: Literature Review Page 45 set of conditions of net available energy, Q, drying power, E A , surface temperature, T s , and surface vapor pressure, e s). Penman (1948) formulated an equation for evaporation from a well-watered, short-grass surface by incorporating net radiation and energy balance into Dalton equation: • E = A Q n + Y E a [2.14] A + y where; E = Vertical evaporative flux (mm day"1), A = Slope of the saturation vapor pressure versus, temperature curve at the mean temperature of the air (mmHg/°C), Q n = Net radiant energy available at the surface (mm day"1), y = Psychrometric constant, E a = f ( u ) ( e a - e a ) Penman equation is well-known, and many variations on it have been developed over the years (Burman and Pochop, 1994). There are several other commonly used methods for the calculation of potential evaporation, including the Thornthwaite (Thornthwaite and Mather, 1955) for montly calculations, and Priestley-Taylor method (Priestley and Taylor, 1972; Wilson, 1990). Wilson (1990) and Burnman and Pochop (1994) provide a more detailed review of potential evaporation calculation methods. Another analytical approach to the prediction of actual evaporation was presented by Granger (1989), who suggested that the actual rate of evaporation from the soil could be determined through the Dalton equation, and the actual vapour pressure at the soil surface. Granger did not present a method for calculation of vapour Chapter II: Literature Review Page 46 pressure at the soil surface. Wilson et al. (1994, 1997) expanded on the work of Penman (1948) and developed a coupled thermal, vapor, and liquid water flow model for predicting actual evaporation from a bare soil. Dalton's Law was utilized to calculate evaporation rate based on the suction at the soil surface: E = A Q " + y E a [2.15] A + y A r h where; E = Vertical evaporative flux (mm day"1), E a = f(u)e a(B r h-A rh) where, f(u) = Function dependent on wind speed, surface roughness, and eddy diffusion, 0.35(1+0.1 U a), U a =Wind speed (km hr"1), e a =Vapor pressure in the air above the evaporating surface, A r h =lnverse of the relative humidity of the air, B rh =lnverse of the relative humidity at the soil surface. In this equation, the parameter A r h (inverse of relative humidity) at the soil surface becomes unit in the case of saturated vapour pressure in the soil surface, and the equation simplifies to the original Penman equation. In order to predict the actual evaporation with this equation, it is necessary to solve for the vapour pressure at the soil surface. The solution for evaporation events is equally complex because the rate of Chapter II: Literature Review Page 47 potential evaporation is determined by both the rate of potential evaporation established by climatic conditions and the suction at the soil surface. 2.5.3 SoilCover program SoilCover is a one-dimensional finite-element package that models transient liquid and water vapor flow, based on a theoretical model for predicting th rate of evaporation from soil surfaces presented by Wilson et al. (1994). The model is based on a system of equations for couple heat and mass transfer in soil (Yanful et al., 2003). The flow of water vapor and liquid water are described on the basis of Fick's Law and Darcy's as follows: 5t w 5y x 5 h w + c f 'y J D . 5 y ; [2.16] where: h w = Total head (m) t = Time (s) C j y = Coefficient of consolidation with respect to the liquid water phase C 1 = J - Pw9 p w = Mass density of water (kg m"3) g = Acceleration due to gravity (m s"2) y = Position (m) K w = Hydraulic conductivity (m s"1) Chapter II: Literature Review Page 48 = Coefficient of consolidation with respect to the water vapour phase P(Pw) 2gm 2 w = Slope of the moisture retention curve (1/kPa) P = Total pressure in the air P h Dv=aP t V 3 P R T J = diffusion coefficient of water vapor through the soil (kg m kl\T1 a = p 2 7 3 is the tortuosity factor of the soil; and p is the cross- sectional area of the soil available for vapor transfer D v a p = 0.229 x10"4 f 1+ T ,1 .75 V 27315j = is the molecular diffusivity of water vapor in air (m 2 s"1) T = temperature (K) W v = the molecular weight of water (0.18 kg kmol"1) R = the universal gas constant (8.314 J mol"1 K"1). Temperature is evaluated on the basis of conductive and latent heat transfer as follows: 5T 5 V 8vJ _ | r(p+pv)i 8 f S P v l 8t " = 5y I p J 5 y 8 y J [2.17] Chapter II: Literature Review Page 49 where: T = temperature (°C) Ch = C v p s = the volumetric specific heat of the soil as a function of water content (J rrf3 °C-1) C v = the specific heat of the soil (J kg"1 °C) p s = the mass density of the soil (kg m"3) X = the thermal conductivity of the soil (W m"1 °C"1) L v = the latent heat of vaporization of water (J kg"1). SoilCover calculates the vapor pressure in the soil using the relationship provided by Edlefsen and Anderson (1943), in which vapor pressure is calculated on the basis of the total suction in the liquid phase: Pv=Psvhr [2.18] Where: P v = Actual vapour pressure within the soil P s v = Saturation vapour pressure of the soil at its temperature, T h r = relative humidity of the soil surface as a function of temperature fygWv h r = e^ R T v|/ = Total suction in the total suction in the unsaturated soil (m). Atmospheric coupling is achieved by calculating the soil evaporative flux. Soil evaporative flux is a function of the vapor pressure gradient between the cover surface Chapter II: Literature Review Page 50 and the atmosphere. A modified Penman formulation proposed by Wilson (1990) in Equation 2.14 is used. The surface temperature may be estimated using the following relationship (Wilson, 1990): T s = T a + ^ ( Q " E _ G s ) [ Z 1 9 ] where: T s = the temperature at the soil surface (°C) T a = the temperature of the air above the soil surface (°C) G s = the ground heat flux (mm day"1 of equivalent latent heat). y = Psychrometric constant, 2.5.4 Chapter Summary In summary, from the literature review, it was noted that many studies have investigated 0 2 and C 0 2 in subsurface pore gas and surface 0 2 and C 0 2 fluxes for natural ground profiles. Very few studies have focused on quantifying surface C 0 2 fluxes and C 0 2 production rates for waste-rock systems. The literature review for waste- rock studies also indicated that quantification of 0 2 consumption rates has been completed almost exclusively by Australian researchers using using temperature profiles. A need, therefore, exists for measurements of surface C 0 2 fluxes and C 0 2 production rates in waste-rock piles. Chapter II: Literature Review Page 51 O2 is geochemically important and active because it is a strong oxidizing agent (has strong affinity for electrons). Examples will be outlined later in the thesis to illustrate the role of O2 as an oxidant in a waste-rock environment (e.g., oxidation of sulphide minerals and oxidation of organic matter). The production of CO2 gas is important because it dissolves in the pore water and produces an increase in the activity of H + (increase acidity). It is also important to note that carbonate minerals are often present in natural subsurface environments and have a buffering effect on the pH of subsurface pore water (Freeze and Cherry, 1979). This effect will be discussed later in the thesis. Pore gas migration in this study will be attributed to diffusion. This assumption is consistent with previous sub-surface gas studies (de Jong and Schappert, 1972; Elberling and Nicholson, 1996; Harries and Ritchie, 1985; Solomon and Cerling, 1987). Different diffusion models describe the interaction between gas molecules and the porous media through which the gas is diffusing. The Knudsen model depends upon the molecular weight and temperature of the gas as well as the pore size through which it is diffusing, but it is not influence by the presence of other gas molecules. The molecular diffusion process assumes that gas molecules collide only with other gas molecules. Molecular diffusion depends upon the molecular weights and temperatures of all the gases in a particular system and does not consider the physical nature of the porous media. A third diffusion model assumes that gas molecules collide with each other and with the porous media. Diffusion is dependent upon pore size, molecular weights and temperatures of the gases, and the physical nature of the porous media. This gas diffusion model will be used in the present work and the development of related equations will be described later in the thesis. Chapter II: Literature Review Page 52 The influence of soil water content on gas fluxes measurements and diffusion is important when the soil is at high water content (Davidson and Trumbore, 1995; Moncrieff and Fan, 1999). The water content of soil depends on several factors: soil texture, temperature, soil respiration rate, environmental conditions of adjacent layers. These factors, which control C 0 2 fluxes, vary in different ecosystems and under different climatic conditions. As pointed above, climate has the potential to enhance or reduce soil C 0 2 fluxes. The total water balance, including evaporation is necessary, example to assess the long-term performance of a cover (Wilson et., 1994, and 1997; Aubertin et al., 2006). The dependency of the effective diffusion coefficient on soil water content for different textured soils is well documented (Klute and Letey, 1958; Rowell et al., 1967; Mbonimpa et al., 2003). The diffusion coefficient of C 0 2 in water is about four orders of magnitude slower than that in the air-filled voids. The knowledge of the rate of evaporation at the soil-atmosphere interface is therefore required to estimate the water content of candidate cover soils (Wilson et al., 1994; Yanful et al., 2003a) In conclusion, the work described in the subsequent chapters is primarily directed at the measurement of C 0 2 fluxes from a waste rock surface. A review of literature shows there is need for further study in this important area of mine waste management. A new instrument is developed and tested using other methods. In addition, the new instrument is used to measure C 0 2 from a waste rock surface under natural field conditions. The influence of surface water conditions with respect to the diffusion coefficient of C 0 2 and associated fluxes is also investigated. Chapter III: Materials and Methods Page 53 CHAPTER III Materials and Methods 3.1 Introduction The methods used in this thesis consist of laboratory tests and field measurements. The objective of the laboratory program was to determine the hydraulic properties and characteristics of the soil that influence the C 0 2 gas surface fluxes. The tests were conducted in the geotechnical laboratory of the department of mining engineering at the University of British Columbia. The tests conducted include: grain size analysis, water retention curve (WRC) measurements, and saturated hydraulic conductivity tests. The field program consisted of (i) measuring the C 0 2 surface fluxes at the Deilmann north (DNWR) and Deilmann south (DSWR) waste-rock piles using the dynamic closed chamber (DCC) method during the summers of 2000 and 2002 and compare the results with those obtained using the static closed chamber (SCC) and eddy covariance (EC) methods and (ii) investigating the effects of climatic variables (e.g., rainfall and evaporation) which affect the gas fluxes. 3.2 Laboratory Program The laboratory program consisted of sample collection and testing for hydraulic properties. 3.2.1 Sample collection A 5-kg sample of waste-rock material from the ground and near ground surface (0-0.15 m) was collected at three different locations around DNF1 and DSF1 (Figure Chapter III: Materials and Methods Page 54 3.8) at the DNWR and DSWR using a sampling scoop. The triplicates samples from each of the waste-rock piles were combined and placed in zippered airtight plastic bags. The samples were shipped to the department of mining engineering of the University of British Columbia for laboratory tests. All samples were stored in the laboratory at room temperature. It should be noted that the waste-rock samples were not representative of the entire DNWR and DSWR piles because physical weathering of the waste rock would have occurred at the surface and near surface of each pile over the years. The grain size of the samples collected was < than 5 cm. It should be noted that the DNWR and DSWR piles consisted of sand and sandstone, and basement gneiss rock, respectively, and that, after physical weathering, had broken down to soil with texture of a medium sand. Therefore, the samples of waste-rock material collected were not representative of the entire waste-rock piles. 3.2.2 Grain-size analysis The particle-size analysis of the soil samples was determined by sieve analysis according to A S T M Designation: D 422-63. Two tests were performed on each sample: (i) the distribution of particle sizes larger than 75 pm (retained on the No. 200 sieve) was determined by sieving (ii) the distribution of particle smaller than 75 pm was determined by a sedimentation process, using a hydrometer to secure the necessary data (Figure 3.1). Chapter III: Materials and Methods Page 55 Chapter III: Materials and Methods Page 56 (i) Particles larger than 75 um: Approximately 200 g of each waste-rock sample larger than 75 pm (retained on the No. 200 mesh) was dried at 110 °C for 24 h. The oven dried sample was sieved through sieves with mesh sizes of 4, 10, 20, 40, 60, 80, 100,140, 200 and 270 on a shaker for 10 minutes. The mass and percent of waste-rock retained on each sieve were determined by weighing and plotted against the size of the sieves openings. (ii) Sedimentation process: Approximately 70 g of each sample passing through 200 mesh opening sieve was oven dried for 24 h and the mass was subsequently recorded. The sedimentation process was done in a 1-L glass cylinder using Sodium hexametaphosphate as a dispersing agent dissolved in distilled water (Figure 3.1). The solution was adjusted to a pH of 9.5 using sodium carbonate. After agitation of the slurry hydrometer readings were taken at specified time intervals up to 24 h. A sieve analysis was then performed on the material after the suspension was washed with tap after and oven dry at 110 °C. The percentage of soil remaining in suspension at the level at which the hydrometer was measuring the density of the suspension was calculate using a formula (ASTM Standard D 422-63, 1998). 3.2.3 Water retention curve Water retention curves (WRCs) (or soil water characteristic curves) for the two waste-rock samples were determined in a Plexiglas Tempe cell apparatus (0.1 m dia. x 0.14 m height) using standard methods (Fredlund and Rahardjo, 1993) (Figure 3.2). In this test approximately 75 percent of the cell volume was filled with the waste-rock sample. The samples were tested using a 1 bar ceramic stone conducted at room temperature of approximately 20 °C. Atmospheric pressure was maintained at the Chapter III: Materials and Methods Page 57 ( A ) J5L II I Confining Ring 0-Rn[ - Air Flush Port W«t«r FiUed Grooves SAMPLE Top Ceramic Stole B u t Water Dbckirp -*• Collection (Ur-») Figure 3.2. (A) Schematic diagram of Tempe cell and (b) water retention curve (SWCC) measurement setup. Chapter i l l : Materials and Methods Page 58 discharge face of the porous stone. Air did not flow through the cell unless the air pressure exceeded the air entry value of the ceramic disk. Small amounts of the air diffused through the water in the pores of the high air entry disk and were subsequently flushed from the base of the cell. However, the test was not affected as the air pressure in the cell was maintained by the inlet pressure. The high air entry disk at the base of the apparatus must be saturated prior to the start of the test. The sample was slowly saturated from the base upwards with distilled water until the sample surface was flooded. The sample was left saturated over night prior to the measurements. After saturation of the waste-rock specimen, increasing pressures of 0.2, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 50, 80 and 100 kPa were applied to the air phase within the cell. The total mass of the waste-rock filled Tempe cell was continually monitored during the drainage phase of each pressure increment. Equilibrium was achieved when zero discharge (measured as change in mass) was observed over a 24 to 72 hour period. Upon reaching equilibrium at 100 kPa of applied suction, the sample was removed from the Tempe cell. The water content corresponding to the highest matric suction (100 kPa) was measured by oven-drying the waste-rock sample. This water content together with the previous changes in weight were used to back-calculate the water contents corresponding to the other suction values. The matric suctions were then plotted against their corresponding water contents to yield the S W C C . Fredlund and Rahardjo (1993) provide a discussion on the measurement of the matric suction. 3.2.4. Saturated hydraulic conductivity The saturated hydraulic conductivity (Ks at) of the samples was determined by performing a falling-head hydraulic conductivity test in a stainless steel permeameter Chapter III: Materials and Methods Page 59 cell (0.101 m dia. x 0.116 m height) using an A S T M Standard Test Method, D 5856, 1995 (Figure 3.3). It should be noted that the falling-head test is usually recommended for soil having a K s a t > 10"4 - 10"5 m s"1 and the constant head test is recommended for coarse-grained soils. The base and top plates of the permeameter were sealed using rubber O-rings. The top plate was connected to a 100 ml standing pipe burette (0.015 m dia. x 0.70 m height). The base plate was connected to a constant head reservoir. Oven-dried waste-rock samples were uniformly and loosely poured into the cell to about 95 percent of the cell volume^ The weight of the dry sample was determined by the difference between the weight of the waste-rock-filled cell and the empty cell. The sample was saturated downward with distilled water flowing from the burette through All air bubbles were removed from the apparatus system by downward flushing of the system with distilled water. Water from the standing pipe burette was allowed to flow through the waste-rock sample using a regulated valve The time for water to fall between two defined elevations on the standing pipe burette was recorded for each test. The test was repeated until a constant time for water to fall a given height was achieved. The final sample height was then measured before removing the sample from the permeameter cell. The K s a t was calculated using the following equation (ASTM Standard Test Method, D 5856, 1995): aL * L o g - 9 - [3.1] where: a = cross-sectional area of the burette, L= Length of the waste-rock sample in the permeameter, Chapter 111: Materials and Methods Page 6 0 (A) Graduated cylinder. LA Figure 3.3. (A) Schematic diagram and (B) experimental setup of the saturated hydraulic conductivity measurement. Chapter III: Materials and Methods Page 61 A= cross-sectional area of the permeameter, T 0= time when water in the standing pipe isat H 0 , T-i= time when water in the standing pipe is at H i , H 0 and Hi= are the heads from the stand pipe to the bottom constant head. The hydraulic conductivity (K) of an unsaturated soil is a function of the degree of saturation (or the volumetric water content) or soil matric suction (y) (Huang et al. 1998). A number of empirical relationships have been proposed to determine K as a function of volumetric water content, or matric suction or \\i (Richards 1931; Wind 1955; Gardner 1956; Davidson et al. 1969; Philip 1986; Ahuja et al. 1988, and Frudlund and Rahardjo, 1993). However, the models proposed by Brooks and Corey (1964) and Mualem (1978), appear to be of wider applicability than other models. We used the Brooks and Corey (1964) relation to calculate the unsaturated hydraulic conductivity in which the measured K s a t is defined as above at \\i < V | / A E V , M / A E V is the suction corresponding to air-entry value (AEV) and, (K): K ( V ) = K sat > V > MMEV [3.2] n = 2 + 3L [3.3] L = the pore-size distribution index, Chapter III: Materials and Methods Page 62 L = - Aiog(se) Alog(^) [3.4] S e = effective saturation, S - S r 1 - S r [3.5] S = degree of saturation at and S r = residual saturation. 3.3 Laboratory Mesocosm and Minicosms (sand columns) Previously, Kabwe et al. (2002) tested the dynamic closed chamber (DCC) method in the laboratory in well-constrained mesocosm (2.4 m dia. x 3.2 m thick) and two minicosms (0.58 m dia. x 1.2 m thick) that was shown to accurately measure C 0 2 fluxes from ground surface to the atmosphere (Figure 3.4). One minicosm was maintained at 18-23 °C (HT) and the other at 5 °C (LT). Data measured from these columns were used to validate the numerical model developed in this thesis [see also Appendix E). The data include: the water contents, C 0 2 and 0 2 concentrations and temperature profiles measured in the columns. These large physical models, referred to as mesocosms, are considered better surrogates for the natural ecosystem because they have biogeochemical cycles and gradients representative of the natural system. They are of sufficient size that relevant physical, chemical, and biological processes are active and thus permit natural behavior under controlled conditions and can provide an important link in the validation and extrapolation of results to ecosystems (Hendry et al., 2001; Lawrence and Hendry, 1995). The description, construction and filling of the mesocosm are presented in Lawrence et al. (1993), Hendry and Lawrence et al. (1993), Chapter III: Materials and Methods Page 63 1 F i g u r e 3 . 4 . Figures showing (A) schematic diagram and (B) photograph of the mesocosm column (Lawrence et al., 1993), and (C) minicosms columns, C 0 2 gas analyzer and small chamber setup used in the calibration and verification of the D C C method (Kabwe, 2001; Kabwe et al., 2002; Richards, 1998).. Chapter III: Materials and Methods Page 64 Lawrence and Hendry (1995), and Hendry et al. (2001). The following section briefly presents the physical descriptions of the mesocosm and minicosms. Mesocosm: (Figure 3.4A) Fine-grained, poorly graded sand was excavated from the C-horizon of an unsaturated zone at a field site located 10 km south of Saskatoon, Saskatchewan (Hendry et al., 2001). The sand was excavated to a depth of 6 m using a 2 m diameter solid-stem auger in November 1992. As it was excavated, the 65 tons of sand were placed on plastic sheets and transported to the laboratory. The bottom 0.5 m of the cylinder mesocosm (2.4 diameter x 4.6 m high) was filled with 6 to 12 diameter gravel to facilitate control of the water table. The sand excavated from the field site was placed on top of the gravel in the order opposite to which it was removed from the field site. The volume of the excavated material yielded a final sand thickness of 3.6 m. and minicosms shown in Figure 3.4. Biologically produced CO2, was considered to be the primary source for the CO2. Studies (Hendry et al., 2001) on microbial aspects of the mesocosm indicated the presence of microbial activity throughout the unsaturated zone. These results indicated that biological activity within the mesocosm was likely sufficient to account for the generation of C 0 2 throughout the profile. Minicosms: (Figure 3.4B) Two minicosms were constructed from a 0.58 m ID polyvinylchoride (PVC) tube, 1.3 m in height, fitted with removable airtight lids. The minicosms were filled with about 634 kg of sand excavated from an unsaturated C- horizon (no A or B horizons) at a field site located 10 km south of Saskatoon, Saskatchewan. The texture and chemistry of the C-horizon sand are described in Hendry et al. (1999, 2000). The methods of filling the minicosms and installation of the instrumentation are described in Richards (1998). On day 1 of the study (August 06, Chapter III: Materials and Methods Page 65 1995), the water tables in the minicosms were lowered from ground surface to a depth of 0.95 m. One minicosm was maintained at room temperature (21-23°C) (HT -High Temperature) and the other minicosm at 5 + 2 °C (LT - Low Temperature). Results of studies of microbial populations also indicated that biological activity within the minicosms was likely sufficient to account for the generation of C 0 2 throughout the profile (Hendry et al., 1999; Richards, 1998). 3.4 F ie ld P r o g r a m The field investigations were carried out at the Key Lake uranium mine, northern Saskatchewan, Canada (57° 12' latitude, 105° 35' longitude) during the summers of 2000 (June to September) and 2002 (August to September). Two waste-rock piles were selected for study: the Deilmann north waste-rock (DNWR) and the Deilmann south waste-rock (DSWR) piles. The DSRW was selected because of its overall simplicity. Specifically, (1) it is texturally uniform with a grain size similar to that of the sand used in mesocosms to verify the D C C method (Kabwe et al., 2002); (2) the surface of the pile was devoid of plant cover and there was no soil development, hence any spatial or temporal variability associated with surficial respiration was minimal. A weather station was installed at the DSWR in April 2000 to characterize basic climatic variables. C 0 2 flux collars for the dynamic close chambers (DCC) method were installed on the DNWR and DSWR in April 2000 by the author of this thesis. C 0 2 flux collars for the static closed chamber (SCC) method were installed on the DNWR and DSWR in Summer 2002 by the author of this thesis. Sensors for measuring C 0 2 flux using the eddy covariance (EC) technique were installed on the tripod of weather station on DSWR in Summer 2002. Chapter III: Materials and Methods Page 66 The following sections describe materials and methods of the field tests along with the results and interpretation. 3.4.1 S i te l oca t ion a n d d e s c r i p t i o n The Key Lake uranium mine is located at the southern rim of the Athabasca Basin in north-central Saskatchewan, approximately 750 km north of Saskatoon, Canada (57° 12' latitude, 105° 35' longitude) (Figure 3.5). The average mean annual temperature at the mine site from 1977 to 1998 was -1.33 °C. From 1977 to 1998, average winter precipitation (October to April inclusive; predominantly snow) was 163.6 mm, average summer precipitation (May to September inclusive; predominantly rain) was 294.8 mm and average total precipitation was 457.4 mm (Birkham et al., 2003). The average annual evaporation (potential) for this time period was 652.9 mm (data obtained at the Key Lake mine site). Basement gneiss rock is unconformably overlain by Athabasca Group sandstone (Key Lake Mining Corporation, 1979). The Deilmann ore body was mined from 1984 to 1997. Predominant uranium-bearing minerals were coffinite (USi0 4 ) and pitchblende (U0 2 ) (Key Lake Mining Corporation, 1979). Arsenide, nickel and sulfide minerals associated with the Key Lake deposits included niccolite (NiAs), gersdorffite (NiAsS) and millerite (NiS) (Key Lake Mining Corporation, 1979). Dissolution of these minerals would potentially increase the concentrations of Ni and As in infiltrating waters. Minor amounts of pyrite (FeS 2 ) and cobaltite ((Co, Fe)AsS) were also present (Key Lake Mining Corporation, 1979) F i g u r e 3 . 5 . Map of Saskatchewan showing the location of the Key Lake uranium mine., Saskatchewan, Canada. Chapter III: Materials and Methods Page 68 Gangue materials consisted of sandstone and basement gneiss. The minerals comprising the sandstone included quartz (Si0 2 ) , chlorite ((Mg, Fe, AI)6(AI, Si) 4Oio(OH) 8), kaolinite (AI 2Si 205(OH)4), calcite (CaC0 3 ) and siderite (FeC0 3 ) (Key Lake mining Corporation, 1979. The basement gneiss was typically composed of quartz, muscovite (KAI3(AISi30io)(OH)2), chlorite and feldspars (Key Lake Mining Corporation, 1979). Graphitic gneiss was also present. Acid base accounting results for sand/outwash till, sandstone and basement rock indicated that both the sulfur and carbonate contents were very low and that the waste- rock piles were not clearly acid generating or consuming (Steffen Robertson and Kirsten (Canada) Inc., 1993). The ratio of neutralization potential to acid generation potential (NP/AP) for sand/outwash till was 1.6 with a standard deviation (s.d.) of 1.30 and a sample size (n) of 29. The mean total sulfur content was 0.03 % (s.d. = 0.02, n = 29). The NP /AP for sandstone was 0.8 (s.d. = 1.93, n = 68) and the total sulfur content was 0.04 % (s.d. = 0.02, n = 68). The NP/AP for basement rock was 1.7 (s.d. = 1.5, n = 27) and the total sulfur content was 0.11 % (s.d. = 0.05, n = 27). Temperatures within the piles ranged from 0 to 2°C. The excavation of the Deilmann pit resulted in the concurrent construction of two main waste-rock piles from 1984 to 1997, the Deilmann north waste-rock (DNWR) and the Deilmann south waste-rock (DSWR) piles (Figure 3.6). The waste-rock piles were constructed in lifts approximately 8 m in height. Haul ramps were used to transport material to each new lift pad where the waste rock was then dumped and pushed off the edge of the pad to maintain a flat top. Compaction and physical weathering of the waste rock would have occurred at the surface of each lift as a result of machinery traffic. Chapter III: Materials and Methods Page 69 F i g u r e 3 . 6 . Photograph showing the Deilmann pit, the Deilmann north waste-rock (DNWR) and Deilmann south waste-rock (DSWR) piles at the Key Lake uranium mine, Saskatchewan, Canada. Chapter III: Materials and Methods Page 70 The degree of compaction and weathering could be expected to be the greatest nearest the haul ramp and to decrease further from the haul ramp. The DSWR was constructed between 1984 and 1995 and consists exclusively of sand and sandstone (Key Lake Mining Corporation, 1979) (Figure 3.7). The maximum height of the DSWR pile is 28 to 31 m above the original ground surface. The bottom of the pile has elevated concentrations of organic matter derived from both lake bottom sediments and forest soils (e.g., Figure 3.7). The original ground surface was not scraped of organic material before construction creating a layer of organic-rich sand at the bottom of a large area of the waste-rock pile (Figure 3.7) The DNWR pile was constructed from 1984 to 1997 and consists of a mixture of sand, sandstone and basement rock. The maximum height of the DNWR pile is approximately 42 m above the original ground surface. Several lakes near the pits were drained, exposing lake bottom sediments. The bottom of the pile has also elevated concentrations of organic matter derived from both lake bottom sediments and forest soils. Similarly, the original ground surface of the DNWR was also not scraped of organic material before construction creating a layer of organic-rich sand at the bottom of a large area of the waste-rock pile (Birkham et al., 2003). The geochemical conditions of the waste-rock piles at the Key Lake mine were unique because sulfur contents < 0.11 % S (Steffen Robertson and Kirsten (Canada) Inc., 1993) were low compared to most other geochemical studies (Jaynes et al., 1983a; Gelinas et al., 1992; Elberling et al., 1993; Ritchie, 1994a; Elberling and Nicholson, 1996; Keller and Bacon, 1998; Hockley et al., 2000). Geochemical conditions are also unique because of the cold climate and, as previously mentioned, the piles are constructed of overburden sand, sandstone and gneissic basement rock Chapter III: Materials and Methods Page 71 Figure 3.7. Depth geologic profile for Deilmann south waste-rock (DSWR) pile at the Key Lake uranium mine, Saskatchewan, Canada (Adapted from Birkham et al., 2003). Chapter III: Materials and Methods Page 72 that were dumped on the original ground surface. Approximately 44 million cubic meters of waste rock produced at this site is more than most waste-rock volumes reported in the literature (Gelinas et al., 1992; Ritchie, 1994a; Hockley et al., 2000). Geochemical conditions are also unique because of the cold climate. 3.4.2 Field C 0 2 flux measurement methods This section presents the three methods for C 0 2 flux measurements: the dynamic closed chamber (DCC), the static closed chamber (SCC), and eddy covariance (EC) methods. It should be noted again that the D C C was designed, verified and applied on field by the author of this thesis. The S C C was designed by the Department of Soil Science of the University of Saskatchewan (U of S), however, the field measurements were carried out by the author of this thesis, but the gas analysis for C 0 2 were done at the Department of Soil Science of the U of S (Farrell et al., 2002). Sensors for measuring C 0 2 fluxes using EC method were installed by the Department of Geography of the U of S on the tripod of the weather station installed on the DSWR by the author of this thesis on April 28, 2000 (see later in this section). Data analysis was also done at the same Department of the U of S. Twenty collars were installed on the DNWR and nine on the DSWR, between April 27 and 29, 2000 (Figure 3.8). The collars were manually driven into the waste-rock piles, leaving the top 0.01 rrrof the collar above the waste-rock surface. The ground surface was leveled by rotating a straight edge template (0.01 m thick) on top of the collar. The collars were allowed to stabilize in the sand for about 60 days prior to the start of the C 0 2 flux measurements. Chap te r III: Mater ia ls and Methods P a g e 73 Deilmann South Waste Rock Pile Figure 3.8. Map of the De i lmann north waste- rock ( D N W R ) and De i lmann south waste- rock ( D S W R ) piles at the Key Lake mine, S a s k a t c h e w a n , C a n a d a , showing the chambers and the meteorological weather station locat ions. Chapter III: Materials and Methods Page 74 3.4.2.1 Measuring C0 2 fluxes using dynamic closed chamber (DCC) method A technique to measure C 0 2 fluxes from the soil surface to the atmosphere was recently developed and verified in mesocosms over the range of C 0 2 fluxes reported for field conditions by the author of this thesis (Kabwe, 2001 and Kabwe et al., 2002). The technique termed the dynamic closed chamber (DCC) method, is based on direct measurement of the change in C 0 2 concentration with time in the headspace of a chamber installed on ground surface. Carbon dioxide concentrations were directly measured using a portable C 0 2 gas analyzer (ADC 2250, BioScientic Ltd). The work of this thesis focused on the field application of the DCC method to quantify reaction rates in waste-rock piles. Full details of the design, construction, and operation of the DCC are presented in Kabwe (2001) and Kabwe et al. (2002). The following section briefly described the DCC method. Chamber collars for the DCC were fabricated from fiberglass rims (0.76m dia.x 0.15m height); the chamber lid (0.76m dia. x 0.05m thick) was fabricated from Plexiglas (Figure 3.9). A rubber O-ring was installed into a groove on the underside of the lid to provide an air-tight seal between the lid and the collar. Inlet and outlet brass-fittings were installed in the lid. A perforated tube (1.20 m long) with one end connected to the inlet fitting was installed into a groove on the underside of the lid to provide air dispersion in the chamber headspace. The lid was attached to the collars with nuts and bolts. Carbon dioxide analyses were performed using an ADC 2250 differential infrared C 0 2 gas analyzer (ADC BioScientific Ltd). The analyzer provided simultaneous absolute and differential gas measurements. All CO2 measurements were corrected for pressure Chapter III: Materials and Methods Page 75 Figure 3.9. (A) Typical slopes of direct measurement of concentration versus time using the dynamic closed chamber (DCC) method (B) schematic diagram and (C) photograph of the D C C method setup for measuring surface C 0 2 gas f luxes. Chapter III: Materials and Methods Page 76 broadening and dilution effects caused by water vapor; single bench (CO2) peak-to- peak noise was typically <0.2 ppmV at 350 ppmV CO2. Measurements were made by sealing the lid onto the collar and continuously circulating air from the chamber (top, center) through the A D C 2250 C 0 2 analyzer and back into the chamber through the perforated air-dispersion ring on the underside of the lid (see Figure 3.9). Prior to measuring a flux, the ambient C 0 2 concentration was measured at the collar. The C 0 2 was then scrubbed from the air in the sealed chamber (using soda lime in an on-line trap) to lower the C 0 2 concentration to below ambient (to yield improved accuracy at low flux levels). In the measurement mode, the analyzer measured the C 0 2 concentrations in the chambers as they increased from sub-ambient to ambient and higher concentrations (Figure 3.9). During this period, the CO2 concentration in the chamber was measured at 1 s intervals, with mean concentrations recorded every 10 s. The flow rate through the chambers was maintained at approximately 39 L h"1. The flux of C 0 2 from the soil surface was calculated from the rate of change in C 0 2 concentrations in the chambers as follows: F c o 2 = where F c o 2 ' s the C 0 2 flux from the soil surface, C is the concentration (mg m"3) in the chamber at ambient temperature and pressure, t is time, h is chamber height (m), and dC/dt is the slope of the best fit of the time series as time approaches zero. Fluxes were determined by averaging a series of four to eight measurement cycles. The final flux reported here equals the flux observed at the ambient air CO2 concentration, which was determined prior to measurements. The time required to determine one series of flux dC dt x h [3.6] Chapter III: Materials and Methods Page 77 measurements ranged from 2 to 8 min, depending on the magnitude of the flux. To minimize temperature variation within the chamber, it was shielded from the sun during the measurement period. Note: in all cases, actual C O 2 concentrations were measured as mixing ratios (i.e., volume per unit volume of air) and were converted to a mass basis as described by Hutchinson and Livingston (2000). 3.4.2.2 Measuring C0 2 fluxes using static closed chamber (SCC) method Ambient fluxes of C 0 2 also were measured using a static closed chamber (SCC) consisting of a P V C cap fitted with a vent tube and Swagelok™ sampling port (see Figure 3.10). Collars (15 cm x 20.3 cm i.d.) for the chambers were manually driven into the collar. To minimize the effects of soil disturbance on the C 0 2 flux, the collars were inserted into the waste rock about one week prior to the start of the measurement period. Each chamber had a volume-to-surface area ratio of about 9:1, with an internal headspace volume (including the above-ground portion of the collar) of 1750 cm 3 and a surface area of 201 cm 2 . Once the chamber was sealed to the collar, gas samples were collected at 20-min intervals. Gas samples were collected from the enclosed headspace using a disposable, 20-cc syringe equipped with a 25-gauge, 5 / 8- inch needle. Gas samples were withdrawn through the sampling port (sealed with a gray butyl rubber septum) in the top of each chamber; injected into pre-evacuated (ca. 5 x 10"3 atm), 12- cc Exotainers™, and analyzed using gas chromatography (Farrell et al., 2002). The gas samples were stored under a positive pressure of approximately 2-atm (i.e., 20-cc of headspace gas was injected into each 12-cc collection tube) to minimize any gaseous exchange with atmospheric air. The gas samples were then shipped to the Department Chapter III: Materials and Methods Page 78 Figure 3.10. (A) Schematic diagram and (B) photograph of the static closed chamber (SCC) setup (collar and cap) installed on the Deilmann south waste-rock ( D S W R ) pile. Chapter III: Materials and Methods Page 79 of Soil Science of U of S for C 0 2 gas analysis. Carbon dioxide concentrations were determined using a Varian Model CP2003 Micro-GC equipped with a micro-TCD and Poraplot U column (injector temperature = 110°C, column and detector temperature = 50°C). Ultra-high purity (UHP) helium was used as the carrier gas. The vertical flux density for C 0 2 above the soil surface (mg C 0 2 m"2 h"1) was determined by measuring the change in gas concentration beneath the sealed chamber at set (equally spaced) time intervals. The vertical flux was then calculated using the diffusion-based estimation model proposed by Hutchinson and Mosier (1981) (see also Appendix I): V ^ - C p ) 2 , ^ - C Q ) 0 0 2 A t 1 ( 2 C 1 - C 2 - C 0 ) ( C j - C O and t 2 = 2t 1 and ( C l ~ C q ) > 1 [3.8] where: V is the volume (m3) of enclosed chamber air, A is the area (m2) of soil that is covered by the chamber, C 0 is the initial C 0 2 concentration (mg m"3), and C i and C 2 are the C 0 2 concentrations (mg m"3) at times U (0.33 h) and t2 (0.67 h). The 20 min time interval between samples was long enough for the C 0 2 concentration in the chamber headspace to increase to a measurable level, yet short enough that the C 0 2 concentration in the chamber neither leveled off nor declined during the interval from ti to t2. Carbon dioxide concentrations were converted to a mass basis after correcting for variations in temperature (i.e., 15°C), vapor-pressure (to correct for wet gas), and atmospheric pressure. Chapter III: Materials and Methods Page 80 3.4.2.3 Measuring C 0 2 fluxes using eddy covariance (EC) method Sensors for measuring C 0 2 flux using eddy covariance (EC) technique were installed on the same tripod of weather station in 2002 between 2 July and 25 August (Figures 3.11 and 3.12). A three-dimensional sonic anemometer (CSAT3, Campbell Scientific Inc., Logan, UT) was mounted on a 1.5 m boom with the mid-point of the sonic head approximately 1.7 m above the ground surface within the constant flux layer. The instrument height is justified considering the surface is vegetation-free without wake elements and above the height of the roughness sub-layer. An open-path gas analyzer (LI-7500, LI-COR Inc., Lincoln, NB) was placed on the boom adjacent to the sonic anemometer at the same height with approximately 0.2 m separating the mid-point of each sensor. Various wind components were recorded every half hour using a CR-23X data-logger (Campbell Scientific Inc., Logan, UT). Although the D S W R site was flat, it had a fetch of only 150 to 300 m. The 'flux footprint' of a tower (which is a function of wind speed and direction and the height of the tower) was calculated as described by Schuepp et al. (1990). The peak for the flux footprint ranged from 35 to 50 m upwind, with approximately 90% of the cumulative flux footprint within 150 m upwind of the Both tower. The E C method was used to measure the C 0 2 flux on a continuous basis (Baldocchi et al., 1988). Changes in C 0 2 concentration were measured using the LI- 7500 open-path C 0 2 / H 2 0 gas analyzer. Latent and sensible heat fluxes were measured concurrently. Wind speed and gas concentration measurements were obtained at a frequency of 10 Hz. The C 0 2 flux (Fco 2) was calculated as the product of the mean covariance of the vertical wind speed fluctuations (w' j and the scalar fluctuations in C 0 2 Chapter III: Materials and Methods Page 81 Figure 3 .11. Photograph showing the meteorological weatherstation and the eddy covariance (EC) sensors for measuring C 0 2 flux installed on Deilmann south waste-rock pile (DSWR). Chapter III: Materials and Methods Page 82 F C 0 2 = P a w ' C 0 2 [3.9] where p a is the density of the dry air and the prime (') denotes the deviation from the mean (see also Appendix A). The running mean was based on a 300-s time constant; the resultant mean fluxes and various wind components were recorded every 30 minutes. Corrections and adjustments to the Fco 2 a r © summarized as follows. First, fluxes were corrected for changes in air density (Webb et al., 1980) and were removed when u* < 0.1 m s"1 (Note: u* is the friction velocity as measured by EC) due to poor energy balance closure at low wind speeds (Twine et al., 2000; Barr et al., 2002). Second, F C o 2 was corrected for underestimation by eddy covariance by adjusting for energy-balance closure, assuming that eddy covariance underestimated Fco 2 by the same fraction that it underestimated sensible and latent heat fluxes (Black et al., 2000, Twine et al. 2000; Barr et al., 2002), i.e., by 24% for all measurement periods (r2 = 0.86, n = 3035). Additionally, as a check on the energy balance closure method, a power spectral density function was computed using high-frequency (20 Hz) data to determine if sampling interval was sufficient to capture low and high frequency eddies (i.e., Moore, 1986). This analysis indicates that the tower captured 79% of the energy being transferred (sampling at 10 Hz and integrating over 30 minutes). However, this high- frequency data was collected during one 6-hour period only, and as such the energy- balance method of correction was used as it was considered more representative over all stability conditions. Chapter III: Materials and Methods Page 83 3.4.2.4 Gravimetric water content measurement Waste rock samples were retrieved in triplicates from selected locations around DNF and DSF (Figure 3.8) at four different depths (0, 0.05, 0.10 and 0.15 m) at the DNWR and DSWR, from the period of July 29 to August 06, 2002. The samples weighing between 200 to 250 g were immediately placed in zippered plastic bags to preserve the in situ moisture in the samples. The samples were transported to the on site Key Lake metallurgical laboratory and kept in the refrigerator. The samples were tested within 24 h. The gravimetric water content for waste rock samples was measured according to the ASTM Standard Test Methods for Laboratory Determination of Water (Moisture) Content of Soil and Rock by Mass (D2216-05). The method consisted of taking the waste rock samples weighing between 100 and 200g. The samples were put in a container of known weight, weighed (subtracting the weight of the container gives the weight of the wet soil). The samples were then put in an oven, and dried at 105 °C for about 24 h until all the water had evaporated. After drying the samples, they were weighed again (subtracting the weight of the container gives the weight of the dry soil). The moisture content on a weight basis was the difference between the wet and dry weights divided by the dry eight. The gravimetric water contents were converted to volumetric water contents using data from the W R C and specific gravity. 3.4.2.5 Meteorological weather station A meteorological weather station (Campbell Scientific, Inc.) was installed on DSWR (Figure 3.12) approximately 5 m southwest of the collar located at DSF1 (Figure 3.8) on April 28, 2000 by the author of this thesis. The meteorological station sensors Chapter III: Materials and Methods Page 84 and data acquisition systems (DAS) were installed on a tripod. The wind monitor and net radiometer were mounted on a steel cross-arm at a height of approximately 3 m above the ground surface. The air temperature and relative humidity probe was housed in a radiation shield (approximately 1.8 m above the ground surface) to minimize the effects of solar radiation. The tipping bucket rain gauge was installed on a wooden plank near the tripod (approximately 1 m above the ground surface). The DAS consisted of a CR10 data- logger (Campbell Scientific Inc.), a storage module and a solar panel with 12 volt battery system. The station collected hourly average air temperatures. 3.4.2.6 Chapter Summary In summary, in a previous study using large-scale, laboratory mesocosms filled with sand [Kabwe et al., 2002], the DCC method was shown to accurately measure CO2 fluxes from ground surface to the atmosphere. This laboratory-verified technique, therefore, provided the opportunity to quantify CO2 fluxes under field conditions. The following chapter 4 presents results of the field application of the D C C method. The DCC method was used to determine the magnitude of spatial and, to a lesser degree, temporal variations in the C 0 2 efflux on the DNWR and DSWR piles. In addition, fluxes measured using the D C C method were compared to those obtained from two other methods: static closed chamber (SCC) and eddy covariance (EC) methods. Chapter III: Materials and Methods Page 85 Figure 3.12. (A) Schematic diagram and (B) photograph of meteorological weather station installed on Deilmann south waste-rock (DSWR) pile at the Key Lake mine, Saskatchewan, Canada. Chapter IV: Results and Data Interpretation Page 86 CHAPTER IV Results and Data Interpretation 4.1 Laboratory Tests Program The laboratory program consisted of testing for hydraulic properties for samples from the Deilmann north (DNWR) and Deilmann south (DSWR) waste-rock piles . 4.1.1 Grain-size distribution The laboratory tests results for the near-ground surface ( 0 - 0 . 1 5 m) samples from DSWR and DNWR for grain-size distribution are plotted respectively in Figures 4.1 and 4.2. The detailed tests results are presented in Appendix C. The grain-size distribution curves from DSWR sample (Figure 4.1, curve with symbols) indicated that 90% of the material was sand size with 10% silt- and clay-size particles. The sand sizes ranged from coarse (6%), medium (32%), and fine (52%). The uniformity coefficient (C u) of the sample (C u = D6o/Dio) was found to be about 3.6 (e.g., D i 0 = 0.015 cm is the size such that 10% of the particles are smaller than that size). For comparison, a washed beach sand would have a C u of about 2 to 6 whereas a sample with a C u< 4 is considered well sorted while a sample with a C u > 6 is considered poorly sorted. The void ratio (e) of the sample (e = V V A/ S ) was found to be 0.560 (e.g., V v is the volume of voids and V s is the volume of solids). The C u , e and D-io values of the sample are typical of the values for granular non-consolidated sand materials reported in Table 4.1. Chapter IV: Results and Data Interpretation Page 87 Table 4.1. Nature, origin, and basic geotechnical properties of various granular materials. Source Material Dio(cm) C u e . Coarse sand 0.05800 1.3 0.750 Sydor (1992) . Borden sand 0.00910 1.7 0.590 . Modifield Borden sand.. 0.00800 1.8 0.640 Kissiova (1996) . Secrete sand 0.01450 3.5 0.570 MacKay (1997) . Ottawa sand 0.00937 1.7 0.634 . Beaver Creek sand consolidated at 5 kPa. . . . 0.00930 2.6 0.269 Bruch (1993) . Beaver Creek sand consolidated at 10 kP.. . . 0.00930 2.6 0.267 . Beaver Creek sand Lim eta l . (1998) consolidated at 5 kPa. . . . 0.00930 2.6 0.618 The mean ± one standard deviation grain-size distributions (curves with solid lines) obtained from 106 core samples from DSWR (Birkham, 2002) are also presented in Figure 4.1 for comparison. The partial grain-size (without gravel and boulder-sized) distribution for the material evaluated in this thesis (curve with symbols) was within the envelope of the core samples up to the grain size > 0.3 mm).. Chapter IV: Results and Data Interpretation Page 88 U.S. Sieve openings in inches 1/2 U.S Standard Sieve numbers 10 20 50 100 140 270 ' I I I at c "w CO a. c <u o 03 100 10 0.1 0.01 Gravel Sand Silt or Coarse Fine Coarse Medium Fine Clay Figure 4 . 1 . Particle size distribution curves (without gravel and boulder-sized) for the samples of waste-rock from Deilmann south waste-rock pile (DSWR) for ground surface sand (curve with symbols) and core sand/sandstone (curves with full lines). Symbols represent the measured data from this thesis. The full lines show the one standard deviation range of grain-size data obtained by Birkhman et al. (2002). Chapter IV: Results and Data Interpretation Page 89 Figure 4.2. Particle size (without gravel and boulder-sized) distribution curves for the samples of waste-rock from Deilmann north waste-rock pile (DNWR) for ground surface sand (curve with broken line and symbols) and core basement-rock (curves with solid lines). Symbols represent the measured data from this thesis. The full lines show the one standard deviation range of grain-size data obtained by Birkhman et al. (2002). Chapter IV: Results and Data Interpretation Page 90 The C u of the mean grain-distribution (not presented) for the core samples was about 3.3 and was typical of the value obtained for this study. Birkham (2002) noted, however, that the grain-size distributions of waste-rock samples were not representative of the entire DSWR pile as gravel and boulder-sized particles were excluded from the analysis The grain-size distribution for the near-surface sample collected from DNWR of (Figure 4.2, curve with symbols) indicated that 83% of the material was sand size with 17% silt- and clay-size particles. The sand sizes ranged from coarse (16%), medium (42%), and fine (25%). The C u of the sample was found to be about 6.3 (e.g., D-m = 0.018 cm). The DNWR sample is considered to be poorly sorted than that from the DSWR. The e of the sample was found to be 0.591 and was within the range of the granular sand materials reported in Table 4.1. The mean ± one standard deviation of 26 grain-size distributions (curves with solid lines) obtained from core samples from DNWR (Birkham, 2002) are also presented in Figure 4.2. The grain-size distribution curve measured in this study was outside the mean ± one standard deviation envelope for the basement-rock core samples. The C u of the mean grain-size distributions (not shown) for the basement-rock was determined to be about 30. Birkham (2002) found that 40% of the DNWR basement-rock bulk sample was coble-sized. This was consistent with the visual observation that the basement-rock in the DNWR generally had larger particles compared with the sand-sandstone material. As was the case for the grain-size distributions from the DSWR, the grain-size distributions of samples from the DNWR could not be considered representative of the entire DNWR pile because boulder-sized particles were excluded from the analysis (Birkham, 2002). Chapter IV: Results and Data Interpretation Page 91 4.1.2 Water retention curve For a given porous media, the relationship between the soil water content (9) and the soil suction (vj/) or matric potential is known as either the water retention curve (WRC) (Marshall et al., 1996; Aubertin et al., 1998; Delleur, 1999), the soil water characteristic curve (SWCC) (Fredlund and Xing, 1994; Barbour, 1998), the soil suction curve (Yong, 2001), or the soil moisture-retention curve (Kovacs, 1993; Hillel, 1980; Looney and Falta, 2000). In this thesis, the W R C will be used to represent the relationship between 6 and Figures 4.3 and 4.4 show the W R C s for the DNWR and DSWR samples respectively measured in the laboratory using Tempe cells. The solid symbols are measured data (from 0.2 to 100 kPa suction) and the solid lines (0 to 1 million kPa suction) represent the best fit curves generated with SoilCover model using an equation developed by Fredlund and Xing (1994). The water content (6) can also be expressed in terms of saturation (S ) (S = 0/n), where n is the soil porosity (Figures 4.5 and 4.6). It should be noted that 9 at zero suction is equivalent n. The n for the samples for DNWR and DSWR were found to be 0.36 and 0.38, respectively. The W R C describes the soil's ability to store and release water (Fredlund and Rahardjo, 1993; Barbour, 1998). It also represents the drying curve for the soil material and provides useful information on the water retention and water transmission behavior of a waste-rock pile and helps to describe the effects of waste-rock texture and void ratio (e) on the distribution of the water phase in the waste-rock pile, and thus, the gas diffusion in this pile (Barbour, 1998). The W R C can be seen as a representation of the pore-size distribution function with assumption based on the capillary model (Mualem, 1986). Aubertin et al. (2003) also developed a model to predict the S W C C from basic Chapter IV: Results and Data Interpretation Page 92 F i g u r e 4 . 3 . Water retention curve (WRC) of the sample of waste-rock (with fine fraction only) from the Deilmann north waste-rock (DNWR) pile. Symbols represent the measured data and the solid line the best fit curve generated with SoilCover (SoilCover, 1997). Chapter IV: Results and Data Interpretation Page 93 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 Matric suct ion (kPa) Figure 4.4 Water retention curve (WRC) of the sample of waste-rock (with fine fraction only) from the Deilmann south waste-rock (DSWR) pile. Symbols represent the measured data and the solid line-the best fit curve generated with SoilCover (SoilCover, 1997). Chapter IV: Results and Data Interpretation Page 94 geotechnical properties. In general, the W R C is described as having three parts: (i), the upper horizontal line of the curve represents approximately 100% saturation of the sample (ii) the fast decreasing slope, and (ii) the slow decreasing slope represents the residual water content. The W R C s show that the soil samples remain saturated when suctions are lower than the air-entry values (AEVs). The A E V corresponds to the suction at which the soil sample will begin to ciesaturate and, depending on the soil type, may or not be well defined. SoilCover model calculations yielded values of A E V s of 2.4 and 1.3 kPa for the DNWR and DSWR, respectively. The capillarity of the soil allows it to remain saturated at suction less than the A E V (Aubertin et al., 2003). The slope of the curve defines the volume of water taken on or released by a change in pore-water pressure. The W R C s show that the A E V of the sample from the DSWR is better defined (with steep slope) than that from the DNWR (with smooth slope) (Figures 4.3 and 4.4). The A E V can be defined graphically as the intersection of the best-fit lines of the two linear segments of the W R C (as shown in Figures 4.3 and 4.4) (Fredlund and Xing, 1994 and Barbour, 1998). The tangent (graphical) method yielded values of approximately 1.5 and 1 kPa for the DNWR and DSWR piles respectively. SoilCover model simulations yielded values of A E V s of 2.4 and 1.3 kPa for the DNWR and DSWR, respectively. The slight difference in the A E V s values is due to slight variations in the waste rock textures and porosity. Fine grained soils tend to have flat (or smooth) functions with high A E V s , whereas coarse grained soils tend to have steep functions with low AEVs . For example, the DSWR sample contained less fine-grained (e.g., 10% silt- and clay-size particles), than the DNWR (e.g., 17% silt- and clay-size particles) and the rest of the material was sand-size. Chapter IV: Results and Data Interpretation Page 95 Figure 4.5. Water retention curve (WRC) of the sample of waste-rock from the Deilmann north (DNWR) pile. Symbols represent the measured data and the solid line the best fit curve generated with SoilCover (SoilCover, 1997). Chapter IV: Results and Data Interpretation Page 96 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 Matric Suct ion (kPa) Figure 4.6. Water retention curve (WRC) of the sample of waste-rock from the Deilmann south (DSWR) pile. Symbols represent the measured data and the solid line the best fit curve generated with SoilCover (SoilCover, 1997). Chapter IV: Results and Data Interpretation Page 97 Yanful et al. (2003a) measured the S W C C for fine sand and found an A E V value 3 kPa. This value is close to the value of the DSWR sand sample measured in this thesis (i.e., 2.4 kPa). Wilson et al. (1994) and Newman (1999) also measured the WRCs for fine-grained materials (Beaver Creek sand) and both found an A E V of approximately 3 kPa. Since the soils are close to saturation up to 2.4 and 1.3 kPa suctions for the DNWR and DSWR, respectively, almost all the pore spaces are filled with water and thus the CO2 flux is expected to be significantly reduced. It should be noted that the free diffusion coefficient of CO2 is about four orders of magnitude larger in air than in water, diffusive transport in the water-filled pores is much slower than that in the air- filled voids. The results show that above the AEVs , the water contents (or saturation) decrease rapidly with matric suction. The W R C s show the two samples drain rapidly between values of matric suctions of 2.4 and 10 kPa and 1 and 10 kPa for the DNWR and DSWR, respectively. At 10 kPa suction, the samples retained about 20% and 10% water for the DNWR and DSWR samples, respectively. This behavior is characteristic of uniform sand and sand/silt materials and has been also described by others (Wilson et al., 1994; Barbour, 1998). As the matric suction increased the samples reach slow residual values. The residual water content is controlled primarily by the fine fraction and the surface area of the sample. The residual suctions (¥,-) (suction at residual water content) were determined using the tangent method applied to the W R C s as described by Fredlund and Xing (1994) and were found to be 11 and 6 kPa for the DNWR and DSWR, respectively (Figures 4.5 and 4.6). Chapter IV: Results and Data Interpretation Page 98 Aubertin et al. (2003) provides also the following expression to evaluate 0-42 r 4 1 l "•-<^r [ ] and also T A E V (suction at AEV): ( e D H ) x where e is the void ratio, and D H is an equivalent particle diameter for a heterogeneous mixture and b and x are fitting parameters. For practical geotechnical applications, the value of D H can also be approximated using the following function (Aubertin et al., 1998; Mbonimpa et al., 2000, and 2002): D H = [ l + 1.17log(C u)]D 1 0 [4.3] where D 1 0 is the diameter corresponding to 10% passing on the cumulative grain-size distribution curve, and C u is the coefficient of uniformity (C u = D 6 0 / D 1 0 ) . For the equivalent capillary rise in granular soils b can be approximated using the following function (Aubertin etal . , 1998): b = ? — x — [4.4] 1.17log(C u)+1 Using the values of C u and D 1 0 for the DNWR and DSWR samples (see Section 4.1.1) of this thesis), the D H for the DSWR and DNWR were found to be 0.02478 cm and 0.03488 cm, and b for the DSWR and DNWR were found to be 0.143 and 0.388, respectively. Equation 4.2 yielded values of of 22.07 cm (9.21 kPa) and 55.90 cm (5.59 kPa) for the DNWR and DSWR, respectively. These values are close to those determined graphically using the tangent method (11 and 6 kPa for the DNWR and DSWR, respectively). Similarly, Equation 4.2 yielded values of ^ P A E V of 18.8 cm (1.88 Chapter IV: Results and Data Interpretation Page 99 kPa) and 10.36 cm (1.04 kPa) for the DNWR and DSWR, respectively. These values are very close to those determined graphically using the tangent method (1.5 and 1 kPa) for the DNWR and DSWR, respectively. It should be noted that Equation 4.1 is frequently quite practical for fine-grained soils because D-io and C u are often unknown. 4.1.3 Hyd rau l i c conduc t i v i t y The hydraulic conductivity is a measure of the ability of the soil to transmit water and depends upon both the properties of the soil and the fluid (Klute and Dirksen, 1986). Total porosity, pore-size distribution, and pore continuity are the important soil characteristics affecting hydraulic conductivity and S W C C . The hydraulic conductivity at or above the saturation point (e.g., AEV) is referred to as saturated hydraulic conductivity (K s a t ) , and for water contents below saturation, it is called the unsaturated hydraulic conductivity 'K' (Figures 4.7 and 4.8). Laboratory tests described in the previous section were conducted to determine the saturated hydraulic conductivity 'k s at ' . using the falling-head permeability tests. The tests yielded values of 'K s a t ' of 1.20 x 10"6 m s"1 and 1.49 x 10"5 m s"1 and for the DNWR and DSWR near-ground surface (0-0 .15 m) samples, respectively. These values are characteristic of sand and sand/silt materials. The nature and origin of various data of K are given in Table 4.1. Wilson et al. (1994) measured a value of 3.9 x 10"6 m s"1 for Beaver Creek sand using the falling- head permeability tests. Newman (1999) also measured value of saturated hydraulic conductivity of 6.2 x 10"5 m s"1 for Beaver Creek sand. Yanful et al. (2003) obtained values of K s a t of 1.9 x 10"6 m s"1 and 7.3 x 10"6 m s"1 for fine sand and coarse sand, respectively. Hatanaka et al. (1997) measured values of 1.5 x 10"5 - 4.3 x 10"4 m s"1 for undisturbed sands (12 results). Mbonimpa (1998) determined values of 8.2 x 10"5 - 1.1 Chapter IV: Results and Data Interpretation Page 100 x 10" 3 m s"1 for uniform sand (30 results). These values are very similar to those obtained in this work for DNWR and DSWR. The DNWR contained more fine sand (52%) than the DSWR (25%) and had a comparatively lower Ksat. The hydraulic conductivity 'K' of an unsaturated soil is a function of matric suction y . Laboratory testing was not conducted to measure 'K' at different values of matric suction. Various methods of calculating the hydraulic conductivity 'K' were described above. The relation between the 'K' and y derived from the Brooks and Corey (1964) model for the samples is shown in Figures 4.7 and 4.8. The 'K' of the samples from DNWR and DSWR decreased rapidly with increasing \\> past the A E V s at 1.3 and 2.4 kPa suctions, respectively. As suction was increased by two orders of magnitude, the 'Ks' are predicted to decrease by more than 10 orders of magnitude. At \\i = 100 kPa, both K values decreased to <10"15 m s" 1. In summary, the W R C s and associated 'Ksat' of the samples from DSWR and DNWR showed that the near-ground surface (0 - 0.15 m) sample on DNWR retained more water at saturation associated with increasing matric suction than that on DSWR. This behavior is due to slight variations in the waste-rock textures that control soil water. Chapter IV: Results and Data Interpretation Page 101 1 . 0 E - 0 4 Figure 4.7. Characteristic of the sample of the waste-rock from the Deilmann north waste-rock pile (DNWR): hydraulic conductivity curve (K). The value of saturated hydraulic conductivity (K s at) was measured in the laboratory but the unsaturated hydraulic conductivity (K) was derived from the Brooks and Corey mode (Brooks and Corey, 1964). Chapter IV: Results and Data Interpretation Page 102 1.0E-04 1.0E-06 + 1.0E-16 4 0.1 10 100 Matr ic suc t i on (kPa) Figure 4.8. Characteristic of the sample of the waste-rock from the Deilmann south waste- rock pile (DSWR): hydraulic conductivity curve (K). The value of saturated hydraulic conductivity (K s a t) was measured in the laboratory but the unsaturated hydraulic conductivity (K) was derived from the Brooks and Corey mode (Brooks and Corey, 1964). Chapter IV: Results and Data Interpretation Page 103 Table 4.2. Nature and origin of data for the K value of various granular materials Source of results Type of material (number of results) Range of K values measurd (m s'1) Wilson et al. (1993) Beaver Creek sand 3.9 x 10"b Hatanaka eta l . (1997) Undisturbed sands (12 results) 1.5 x 1 0 " 5 - 4 . 3 x 10"4 Mbonimpa (1998) Uniform sand (30 results) 8.2 x 10~ 5 - 1.1 x 10"3 Newman (1999) Beaver Creek sand 6.2 x 10"b Yahful et al. (2003) Fine sand . Coarse sand 1.9 x 10"b 7.3 x 10-6 Chapter IV: Results and Data Interpretation Page 104 4.2 Field Tests Program This section presents the results of the field tests described in the previous sections. The tests were conducted at the Deilmann north waste-rock (DNWR) and Deilmann south waste-rock (DSWR) piles at the Key Lake uranium mine, northern Saskatchewan, over a period of two years (summers of 2000 and 2002). The CO2 flux results were obtained using the D C C and compared to those obtained using two other methods: dynamic closed chamber (DCC), static closed chamber (SCC), and eddy covariance (EC) methods. The data presented include the results of: 1. Diurnal variation in CO2 flux measured with the D C C at the Deilmann south waste- rock (DSWR) pile. 2. Quantification of spatial and temporal variations in CO2 flux using the DCC at the Deilmann north waste-rock (DNWR) and DSWR piles. 3. Measurements of C 0 2 flux using S C C at the DSWR. 4. Measurements of C 0 2 flux using E C at the DSWR. 5. Measurements of near- and surface-water contents and associated CO2 fluxes after heavy rainfall events at the DNWR and DSWR piles. 4.2.1 Diurnal variation in C 0 2 flux Temporal variability was addressed on a diurnal and long-term basis. The short- term (hourly) variations in the C 0 2 flux was measured using the DCC at a single sampling station (DSF1) (Figure 3.8) over a 9-h period (09:00 to 17:00) on August 6, 2000 at the DSWR. The corresponding average hourly air temperature was recorded from the weather station installed on DSWR. The measurements were repeated two to Chapter IV: Results and Data Interpretation Page 105 three times during the test period to reflect the diurnal variation in C 0 2 flux due to perturbations in daily weather conditions such as cloudy and rainy days. Representative results of both the C 0 2 flux measurements and air temperature are presented in Figure 4.9. The C 0 2 flux ranged from 219 to 250 mg C 0 2 m"2 h"1 (Figure 4.9), with a mean value of 235 (± 14) mg C 0 2 m"2 h"1. Coefficients of variation (CV) for the individual sampling periods ranged from 4.6 to 6.5% and were comparable to those reported under more controlled conditions in laboratory mesocosms (Kabwe et al., 2002). Short- term (hourly) variations in the flux were not significant (P < 0.05) (e.g., P is test statistic on which a decision rule is based for a test of hypotheses). At DSF1 , both the magnitude of the C 0 2 flux and the daily variation in the flux were smaller than the values generally reported for agricultural or forest soils (Brumme & Beese, 1992; Loftfield et al., 1992; Rochette et al., 1992; Ambus & Robertson, 1998; Frank et al., 2002). There was only a weak diurnal pattern to the flux and no correlation between the C 0 2 flux and air temperature (r = 0.548) (e.g., r is a coefficient of correlation). Parkin and Kaspar (2003) reported that diurnal changes in the soil-to-atmosphere C 0 2 flux were strongly correlated with air temperature (more so than with soil temperature) when C 0 2 production at the surface was a major component of the total measured C 0 2 flux (Wohlfahrt et al., 2005; Shi et al., 2006). This, together with the results of Birkham et al. (2003) and Lee et al. (2003a, 2003b), suggests that the C 0 2 flux from the surface waste-rock pile may be a result of the upward migration of gas produced during organic matter oxidation at depth and its subsequent transport to, and diffusion across the waste rock/air interface (see Figure 3.7). Chapter IV: Results and Data Interpretation Page 106 Figure 4.9. Short-term (hourly) variations in the C 0 2 flux measured at DSF1 on August 6, 2000. Fluxes were determined using the dynamic closed chamber (DCC) method and averaging a series of four to eight measurement cycles, with each cycle lasting from 2- to 8-min (depending on the magnitude of the flux). The shaded box represents the 95% confidence interval (+17 mg C 0 2 m"2 h"1) around the calculated daily mean (235 mg C 0 2 m"2 h"1). Chapter IV: Results and Data Interpretation Page 107 4.2.2 Spatial and temporal variation in C 0 2 flux measured using the DCC at the Deilmann south waste rock (DSWR) pile The D C C was used to quantify spatial and temporal variations in the C 0 2 flux at 20 sampling stations (DSF1 - DSF20) (Figure 3.8) at the DSWR pile. The measurements were over three periods during summer 2000 (July 1-11; August 1-11, and September 8-16) and twice during summer 2002 (July 13-22 and August 21-26). Results of the C 0 2 flux measurements are presented in Figures 4.1 OA and Figure 4.10B. Table 4.1 and Figures 4.11 A, 4.11B, and 4.11C present results of statistical analysis of the C 0 2 fluxes measured in July, August, and September 2000. During each 4 to 6 day sampling period, the C 0 2 flux was measured at a minimum of 12 sampling stations, with three to four stations sampled each day. Differences between sampling stations were generally small (average CV = 24%), indicating that the degree of spatial variability was relatively low. Moreover, the analysis of variance (ANOVA) indicated that within each sampling period differences between the daily C 0 2 fluxes were not significant (F j u i = 2.87; F A u g= 1.17; F S e p = 0.60). The F distribution is the ratio of the variances of two independent samples from normal populations and is given by: where S 2 is the variance associated with samples of size n from a normal distribution with variance a 2 and x 2 is the chi-square distribution with u = n - 1 degrees of freedom. F = X2l»2 , and x2 = [4.5] Chapter IV: Results and Data Interpretation Page 108 ( A ) v ' 4 0 0 _ 3 5 0 "•c 3 0 0 CN H 2 5 0 • Ui 2 0 0 • X ^ 1 5 0 O 1 0 0 5 0 A A • • 0 • 0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 C h a m b e r l o c a t i o n ( D S F # ) • July-00 BAugust-00 • Sept- 00 OJuly-02 • 02-Aug ( B ) Figure 4.10. (A) C 0 2 fluxes measured using the dynamic closed chamber (DCC) at twenty selected sampling stations (DSF1 - DSF20) at the Deilmann south waste-rock pile (DSWR) (Figure 3.8) during the summers of 2000 and 2002 (B) average flux values (mg C 0 2 m"2 h"1) measured from sampling locations (•) on the DSWR. Chap te r IV: Resu l t s and Data Interpretation P a g e 109 400 CM 300 - E - Ui E 200 - X 3 - - CM 100 - o o - a • ab ab July 2000 ab ( A ) 180 182 184 186 188 190 Julian Day 400 CM 300 E Ui £ 200 X 3 H - CM 100 o o (B) ; a - i i r a i , i i i August 2000 212 213 214 215 216 Julian Day 217 400 .c CM 300 E - Ui E 200 - X 3 - 100 o o - i 1 1 r (C) September 2000 O 1> o 0 243 244 245 246 247 Julian Day 248 Figure 4.11. Dai ly variat ions in the C 0 2 flux measu red at the De i lmann south was te rock ( D S W R ) pile in (A) Ju ly , (B) August , and (C) Sep tember 2000 . F lux measu remen ts were obtained at three to four locat ions on each sampl ing date. At e a c h locat ion, the flux w a s determined using the dynamic c losed chamber ( D C C ) method and averaging a ser ies of four to eight measurement cyc les , with e a c h cyc le lasting from 2- to 8-min (depending on the magni tude of the flux). The overal l m e a n for each monthly sampl ing period is represented by the d a s h e d l ines ( ). Within months, symbo ls labeled with the s a m e letter are not signif icantly different at the P < 0.05 level of probability. Chapter IV: Results and Data Interpretation Page 110 Working under the assumption that both samples are from the same normal population (as was in this case) then Equation becomes: F = | | [4.6] The F ratio was then compared to the expected value of F(ui = n-i - 1, u 2 = n 2 - 1) using the nearest table entry. With the exception of the July 2000 data set, differences between sampling periods were not significant (average—yielding a long-term average flux of 194 (± 75) mg C 0 2 m"2 h"1. Again, these results suggest a the laterally extensive source of C 0 2 at the base of the pile (see Figure 3.7) and indicate that the production and upward migration of C 0 2 through the waste-rock pile is relatively uniform both spatially and temporally. As noted above, the July 2000 data set yielded a mean C 0 2 flux (238 mg C 0 2 m"2 h"1) that was significantly greater (P < 0.05) than the mean flux calculated for any of the other sampling periods. The flux data collected in July 2000 also exhibited a much wider range in values as indicated by the size of the 'box' and length of the 'whiskers' (Figure 4.12) suggesting a greater degree of spatial and temporal variability during this measurement period. Flux measurements from across the site also were obtained during the summer of 2002. Differences among sampling stations obtained during the summer 2002 at the DSWR were also relatively small (overall C V = 31 %) and were not significant—yielding an overall average flux 174 (± 31) mg C 0 2 m"2 h"1. Both calculated mean C 0 2 fluxes for the summer 2002 periods were not significantly different from those obtained during the summer 2000 periods at the DSWR, with the exception of the July 2000 data set. Chapter IV: Results and Data Interpretation Page 111 500 f 400 CM i E, 300 * 200 CM 8 100 L J I * Jul'00 Aug'OO Sep'00 Jul'02 Aug'02 Sampling Date Figure 4.12. Box & Whisker plot characterizing the spatial and long-term temporal variability in the C 0 2 flux measured using the dynamic closed chamber (DCC) method at the Deilmann south waste-rock (DSWR) pile in 2000 and 2002. The estimated, time-averaged flux = 170 (± 51) mg C 0 2 m"2 to"1, The minimum and maximum flux values are marked by asterisks (*). Note: values occurring beyond the "whiskers" were identified as outliers and were not included in the analysis of variance. Chapter IV: Results and Data Interpretation Page 112 Table 4.3. Summary of results of C 0 2 flux measurements using the dynamic closed chamber system (DCC) for the test period of 2000-2002 at Deilmann south waste-rock pile (DSWR). Summer 2000 Summer 2002 Mean Std a CV b Mean Std CV mg m"2 h"1 mg m"2 h"1 % mg m"2 h"1 mg m"2 h"1 % July 238(n=19) 86 36 185(n=15) 51 28 August 175(n=18) 51 29 162(n=10) 58 36 Sept. 150(n=10) 44 29 Overall 194 75 39 174 53 31 Overall average (summer 2000 and summer 2002): (188 ± 68 mg m"2 h"1) a Std : standard deviation b C V : coefficient of variation Consequently, the data from each sampling period were pooled and replotted as Box- and Whisker plots (Figure 4.12) to better illustrate the relative consistency of the spatial and short-term temporal variability associated with C 0 2 flux measurements at the DSWR. 4.2.3 Spatial and temporal variations in C 0 2 flux measured using the DCC at the Deilmann north waste-rock (DNWR) pile The spatial and temporal variations in the C 0 2 flux were measured at the Deilmann north waste-rock (DNWR) pile using the D C C at 9 sampling stations (DNF1 - DNF9) (Figure 3.8). The measurements were assessed three times during summer 2000 (July 1-11; August 1-11, and September 8-16) and twice during summer 2002 Chap te r IV: Resu l ts and Data Interpretation P a g e 113 Table 4.4. Summary of results of C 0 2 flux measurements using the dynamic closed chamber system (DCC) for the test period of 2000-2002 at Deilmann north waste-rock pile (DNWR). S u m m e r 2000 S u m m e r 2002 M e a n a S t d b C V M e a n Std C V mg m" 2 h"1 mg m" 2 h"1 % mg m" 2 h"1 mg m- 2 h"1 % July 159(n=9) 41 25 302(n=9) 83 27 Augus t 203(n=9) 50 18 249(n=9) 91 37 Sept . 169(n=9) 52 31 Overal l 177 50 28 276 89 32 Overal l ave rage (summer 2000 and s u m m e r 2002): (217 + 83 mg m" 2 h"1) a Std: standard deviation b C V : coefficient of variation Data of the C 0 2 f l u x measurements are presented in Figures 4.13A and 4.13B. Table 4.4 and Figure 4.14 summarize the results of statistical analysis of the C 0 2 fluxes measured during the summer of 2000 (July to September) and during the summer of 2002 (July to August 2002) at the DNWR. Differences among sampling stations obtained during the summer of 2000 were small (overall C V = 28%) and were not significant, yielding an overall average flux of 177 (+ 50) mg m"2 h"1. Similarly, differences among sampling stations obtained during the summer of 2002 were also relatively small (overall C V = 32%) and were not significant, yielding an overall average flux of 276 (±89) mg m"2 h"1. However, both calculated mean C 0 2 fluxes for the summer of 2002 were significantly different from the mean flux calculated for other sampling periods in the summer of 2000 at the DNWR. To better illustrate the relative consistency of the spatial and short-term temporal variability associated with C 0 2 flux measurements at the DNWR, the data from each Chapter IV: Results and Data Interpretation Page 114 ( A ) 400 i 350 300 250 E 200 g 150 £ 100 ° 50 • • 8 • A • • O • O • • * 9 o • • A • O  **  O • t 0 1 2 3 4 5 6 7 8 9 10 Chamber location (DNF#) • July-02 • Aug-02 ASept-02 OJul-02 • Aug-02 (B) DNWR 2000 — — DNWR 2002 \ 146 * t238 \ \ 184 186 \ \ 2 3 8 * # 1 2 5 \ | \ 228 \ \ 261 \ \ 302 $™ \ \ 313»V5 \ 1 \ | I 0 m ' \ I SOO m \ ^"™\ 0 m 500 m \ 230, ! j 5 \ Figure 4.13. C 0 2 fluxes measured using the dynamic closed chamber (DDC) at nine sampling stations (DNF1 - DNF9) at the Deilmann north waste-rock (DNWR) pile (Figure 3.6) during the summers of 2000 and 2002: (A) Data points presented on a X Y (scatter) and (B) average flux values (mg C 0 2 m"2 h"1) from samplings locations on the DNWR. Chapter IV: Results and Data Interpretation Page 115 sampling period were also pooled and re-plotted as Box-and-Whisker plots (Figure 4.14). As indicated by the size of the "boxes" and length of the "whiskers" the fluxes data obtained in July and August 2002 also exhibited much wider ranges in va lues- suggesting a greater degree of spatial and temporal variability during these measurement periods. 4.2.4 Cross-statistical comparison between C 0 2 fluxes measured from across the DNWR and DSWR piles Results of cross-statistical comparison tests between the overall averages C 0 2 fluxes calculated for the summers of 2000 and 2002 at the DSWR and the DNWR (Tables 4.1 and 4.2) showed that only the summer 2002 data set for the DNWR yielded an overall average C 0 2 flux that was significantly different from other summer sampling periods. Differences among the remaining sampling periods were not significant. The degree of spatial variability at both sites was generally small (average CV is 28%-39%). These minor differences were attributed to slight variations in the waste-rock textures that control soil water. The overall averages of C 0 2 fluxes at the DNWR and the DSWR over the 2-year test period (summer 2000 and summer 2002) were 217 ± 83 mg rrf2 h"1 and 188 + 68, respectively. In summary, the above results appear to reflect the laterally extensive source of C 0 2 determined to originate in the dewatered organic-rich lake-bottom sediments at the base of the piles (Figures 3.7 and 3.8) (Birkham et al.,2003; Lee et al., 2003) and suggest that the production and upward migration of C 0 2 through the waste-rock piles is relatively uniform both spatially and temporally. At the DSWR the dominant oxidation reactions occur in the organic material underlying the waste rock, the more minor gas reactions in the waste rock are masked (Birkham et al., 2003). However, it should be noted that the acid-base accounting and humidity test results (Cameco, unpublished) Chapter IV: Results and Data Interpretation Page 116 _ 5 0 0 4 0 0 h I <V 3 0 0 E ut E 1< 2 0 0 _3 **- CM O O 100 I T 1 1 Ju l 'OO A u g ' O O Sep 'OO J u l ' 0 2 A u g ' 0 2 Sampling Date Figure 4.14. Spatial and temporal variations in C 0 2 fluxes measured during the summer of 2000 and summer 2002: box-and-wisker plots showing the mean, standard deviation, and extreme values for Deilmann north waste-rock (DNWR) pile data. Chapter IV: Results and Data Interpretation Page 117 suggested that pyrite oxidation-carbonate buffering and the resulting O2 consumption and C 0 2 production are more likely to be observed in the gneissic waste rocks at the DNWR that yielded much greater acid generating potential (AP) (3-11.2 kg C a C 0 3 eq./tonne) and acid neutralizing potential (NP) (3.1-4.2 kg C a C 0 3 eq./tonne) than sandy waste rocks (AP: 0.9-1.2 kg C a C 0 3 eq./tonne; NP: 0.9-1.1 kg C a C 0 3 eq./tonne). It may be concluded that the difference between the overall averages C 0 2 fluxes calculated for the summer of 2000 and 2002 at the DNWR could be the result of sulphide oxidation and carbonate buffering. Birkham et al. (2003) also concluded that the sulphide oxidation and carbonate buffering may have been dominant reactions within the DNWR. 4.3 Comparison of C 0 2 Fluxes Measured using the DCC to those Measured using Static Closed Chamber (SCC) and Eddy Covariance (EC) Methods on the Deilmann South Waste-Rock Pile (DSWR) 4.3.1 Introduction In a previous study using large-scale laboratory mesocosms filled with sand the DCC method was shown to accurately measure C 0 2 fluxes from ground surface to the atmosphere over the range of CO2 fluxes reported for field conditions (Kabwe et al., 2002). However, to ascertain whether the DCC best approximated field CO2 fluxes, the DCC measurements were compared to those obtained from across the DSWR using static closed chamber (SCC) and Eddy covariance (EC) methods. The comparison was based on flux data measured on the same period (August 24 to 25, 2002). Chapter IV: Results and Data Interpretation Page 118 4.3.2 DCC fluxes Subsets of the D C C data obtained during the period from August 24 to 25, 2002, were pooled and replotted as Box-and-Whisker plot (Figure 4.15) for comparison with those obtained using S C C and E C methods. The A N O V A indicated that differences between the daily C 0 2 fluxes measure using the DCC from August 24 and 25 t h 2002 were not significant (P < 0.05)—yielded an average flux of 162 + 58 mg CO2 m"2 h"1 (n=12)forthe 2-d period. 4.3.3 SCC fluxes Flux measurements using S C C were conducted on August 24, 2002. Headspace gas samples were collected in both the morning (between 10:00 and 11:00) and afternoon (between 16:30 and 17:30). Flux measurements obtained in both the morning (between 10:00 and 11:00) and afternoon (between 16:30 and 17:30) on August 24, 2002 are presented in Figures 4.16 and 4.17. Results of statistical analysis are presented as box-and-whisker plots in Figure 4.18. Temporal variations were generally small and there was no significant difference between the average C 0 2 f l u x measured in the morning (181 ± 60 mg C 0 2 m"2 h~1) and that measured in the afternoon (173 ± 62 mg C 0 2 m"2 h"1). Presumably, this reflects the fact that the morning and afternoon measurement periods occurred at the beginning and end of the diurnal cycle. Nevertheless, these results support those obtained earlier using the D C C method, which indicated that the spatial and short-term temporal variability associated with the C 0 2 flux was relatively small. Flux measurements of both Chapter IV: Results and Data Interpretation Page 119 Figure 4.15. Box-and-wisker plot for flux measurements obtained using the D C C method at the Deilmann south waste-rock (DSWR) pile during the period from August 24 t h to August 25 t h , 2002 (set of data for comparison with the other two methods: S C C and EC). The minimum and maximum flux values are marked by asterisks (*). Note: values occurring beyond the "whiskers" were identified as outiiers and were not included in the analysis of variance. Chapter IV: Results and Data Interpretation Page 120 (A) 350 300 s 250 H 200 H x 150 r5 100 o 50 • A M 1 O • o 2 A M 2 O O 4 6 8 10 12 14 16 Sampling location (DSF#) • PM1 O P M 2 18 20 22 Figure 4.16. (A) C 0 2 flux values (mg m"2 h"1) obtained using the static closed chamber (SCC) at eleven selected sampling stations (•) at the Deilmann south waste-rock (DSWR) pile in the morning (AM) (between 10:00 and 11:00) and afternoon (PM) between 16:30 and 17:30) on August 24, 2002 (B) averages fluxes (mg C 0 2 m 2 h"1) from the sampling locations. Chapter IV: Results and Data Interpretation Page 121 Figure 4.17. Box-and-wisker plots for C 0 2 flux measurements obtained from the Deilmann south waste rock (DSWR) pile on August 24, 2002. Measurements were obtained using the static closed chamber (SCC) method between the hours of 10:00 and 11:00 (AM) and 16:30 and 17:30 (PM). The minimum and maximum flux values are marked by asterisks (*). Note: values occurring beyond the "whiskers" were identified as outliers and were not included in the analysis of variance. Chapter IV: Results and Data Interpretation Page 122 (A) g X 3 O o Chamber position rJ DCCS • Static chambers (B) Figure 4.18. Comparison of the dynamic closed chamber (DCC) and static closed chamber (SCC) methods for measuring C 0 2 fluxes. Flux measurements were obtained using the two methods at the Deilmann south waste-rock (DSWR) pile site during the period from August 24 t h to August 25 t h, 2002. The minimum and maximum flux values are marked by asterisks (*). Note: values occurring beyond the "whiskers" were identified as outliers and were not included in the analysis of variance. Chapter IV: Results and Data Interpretation Page 123 the S C C and DCC methods yielded comparable results (Figure 4.19) during the August 2002 test period, with no significant difference (P< 0.05) between the mean CO2 fluxes obtained using the two methods. These results demonstrate that both the S C C and DCC methods are equally applicable to the measurement of C 0 2 fluxes from the surface of the waste-rock pile. 4.3.4 E C f l u x e s Near-continuous measurements of the C 0 2 flux were obtained using the EC method during the period from June 25 t h to August 25 t h 2002. Measurements were recorded on a data logger installed on the weather station on DSWR. At the end of the test period the data logger was shipped to the Department of Geography of the University of Saskatchewan for data analysis. It should be noted that data collected during precipitation events by the EC system are no useful (Carey et al., 2005). Subsets of the data were then used to assess the magnitude of the diurnal cycle and compare the EC and chamber-based methods (DCC and SCC) . Figure 4.20 shows the diurnal variation in C 0 2 flux measured from 10:00 to 17:00 on August 25, 2002 using the EC at the DSWR The greater temporal resolution provided by the E C system revealed that the CO2 flux exhibited a distinct diurnal pattern in Figure 4.20. As was the case in August 2000, there was no correlation between the magnitude of the C 0 2 flux and air temperature—again suggesting that temperature—induced changes in C 0 2 production at the surface of the waste-rock pile were not a major contributor to the total measured C 0 2 flux. Compared to the diurnal variations in the C 0 2 flux (average s.d. = ±48 mg C 0 2 rrf2 h"1), day-to-day variations in the average flux were generally small (average Chapter IV: Results and Data Interpretation Page 124 3 0 0 - ^ 250 E o E 15f> X 3 CM o o ioa 2 3 i . i i i_ 09:0010:0011:0012:0013:0014:0015:0016:0017:00 Time Figure 4.19. Diurnal variations in the C 0 2 flux measured from 10:00 to 17:00 on August 25, 2002 using the E C method at the Deilmann south waste-rock (DSWR) pile. The shaded box represents the 95% confidence interval (± 24 mg C 0 2 m*2 h"1) around the calculated daily mean (150 mg C 0 2 m"2 h"1). Chapte r IV: Resu l ts and Data Interpretation P a g e 125 Figure 4.20. Measu red C 0 2 f luxes using E d d y covar iance ( E C ) at the De i lmann south was te - rock ( D S W R ) pile. Measu remen ts were obtained on a cont inuous bas is during the period from J u n e 2 5 t h to Augus t 2 5 t h 2002 . E a c h data point represents the daily m e a n va lue ave raged over the period from 10:00 to 17:00 hours: T h e shaded box (B) represents the 9 5 % conf idence interval (±10 mg C 0 2 m" 2 h"1) around the overal l m e a n (150 mg C 0 2 rrf 2 h"1). Note: gaps in the data set represent precipitation events during which no useful data were col lected by the E C sys tem. Chapter IV: Results and Data Interpretation Page 126 s.d. = ±35 mg CO2 m"2 h"1). The near-continuous measurements of the CO2 flux obtained during the period from June 25 t h to August 25 t h 2002 are shown in Figures 4.21 and 4.22. Despite a slight downward trend in the daily CO2 flux with time, the A N O V A revealed that there was no significant difference (P < 0.05) between the averages for July (160 ± 36 mg C 0 2 m"2 h"1) and August (136 ± 32 mg C 0 2 m"2 h"1). A monthly similar trend was observed in 2000 (using the DCC method), which suggests that there may be a small, but distinct seasonal fluctuation in the C 0 2 flux. Flux measurements obtained using both the E C and chamber-based methods occurred on six occasions during July and August 2002 are presented in Figure 4.23 Differences in the C 0 2 flux measured on individual sampling dates, though sometimes large, were not significant (P < 0.05). There was no consistent trend; i.e., the chamber-based methods yielded daily flux values that were less than those obtained using the E C method on the first three sampling dates and greater than those obtained using the E C method on the last three sampling dates. As a result, the time-averaged C 0 2 flux calculated from the E C data (171 ± 39 mg CO2 m"2 h"1) was comparable to that calculated from the corresponding chamber data (178 ± 31 mg C 0 2 m"2 h"1). In summary, the D C C results showed that the flux of CO2 from the surface of the waste-rock pile to the atmosphere as relatively uniform, both spatially and temporally. Presumably, this reflects the combined influence of a relatively constant rate of C 0 2 production in the organic-rich zone at the base of the waste-rock pile (Birkham et al., 2003) and the textural uniformity of the overburden material (sand) used to construct the pile (Birkham, 2002). That is, these factors combine to exert a controlling influence on the composition and upward migration of pore gases and, in turn, the efflux of gases Chapter IV: Results and Data Interpretation Page 127 3 0 0 r 2 5 0 * _ 2001 D) £ , 1 5 0 [ x 3 CM o o 1 0 0 5 0 • 0 • Eddy covariance • Chan.ber-based 3 1 9 4 1 9 6 1 9 7 1 9 9 2 3 6 2 3 7 J u l i a n D a t e (2002) Figure 4.21. Comparison of the eddy covariance (EC) and chamber-based methods for measuring the C 0 2 flux from the Deilmann south waste-rock (DSWR) pile. Chapter IV: Results and Data Interpretation Page 128 from the surface to the atmosphere. Whereas the chamber-based (DCC and S C C ) methods yielded comparable data, with an overall time-averaged C 0 2 f l u x of 171 ± 54 mg CO2ITI"2 h"1; the E C method than that calculated from the chamber data. Underestimation of the F c o 2 associated with soil respiration by EC-based methods relative to chamber-based methods has been reported widely in the literature (e.g., Goulden et al., 1996; Norman et al., 1997; Law et al., 1999; Janssens et al., 2000; Davidson et al., 2002). Though not excessively large, these differences presumably reflect the different processes measured by the two methods. The chamber data exhibited slightly greater standard deviations than the E C data (i.e., D C C = ±58 mg C 0 2 m"2 h"1; S C C = ±59 mg C 0 2 m"2 h"1; E C = ±32 mg C 0 2 m"2 h"1). This most likely reflects the fact that the variability associated with the chamber-based measurements includes both a spatial and temporal component, whereas the variability associated with the EC method is primarily temporal in nature. Thus, it was concluded that both chamber types were suited to the quantification and spatial resolution of C 0 2 fluxes associated with waste-rock piles at the Key Lake mine and that the E C method provided the best estimate of the temporal variability in the C O 2 flux. It is important to note that no single method of measuring soil-atmosphere gas exchanges can meet all objectives. Thus, the choice of method will depend on the type of information required and the characteristics of the site being investigated. Chamber- based methods are especially useful for characterizing spatial variability as well as providing more detailed information regarding local-scale processes. Though more expensive and technically more complex, the eddy covariance method (and other micrometeorological techniques) provides a powerful tool Chapter IV: Results and Data Interpretation Page 129 that allows for spatial integration and near-continuous, long-term monitoring of the so i l - atmosphere flux. Finally, the presence of oxygen in the waste-rock atmosphere is critical for the determination of reaction rates within waste-rock piles. Measurements of surface O2 fluxes is required to complement the field C 0 2 flux data. 4.3.4 Summary of the advantages and disadvantages of the dynamic closed chamber method (DCC) Some of the advantages of the DCC method can be summarized as follow: 1. The D C C method presents a relatively fast method of measuring field C 0 2 fluxes (2 to 10 min. depending on the magnitude of the fluxes). 2. It is a direct method and provides an almost instantaneous indication of the flux measurements regardless of climatic or moisture conditions in the waste dumps. 3. The D C C S uses a portable C 0 2 analyzer and can be used to measure the CO2 fluxes in situ at the same locations using the same chambers with minimal disturbance of the soil. The disadvantages of the D C C method can be summarized as follow: 1. The method requires a very sensitive CO2 gas analyzer, thus high initial capital investment ($19,000 CDN). 2. The C 0 2 flux measurements can be influenced by solar radiation and strong wind, and to changes in chamber pressure and temperature during longer measurement cycles (> 1 h). 3. Although the actual measurement time of change in concentrations in the chamber headspace was short in most cases, the set up of the experiment that include Chapter IV: Results and Data Interpretation Page 130 the attachment of the lid to the chamber prior to measurement was labour intensive, taking several minutes to attach the lid. This long time period was due to the size of the chamber and large number of bolts (n=24) required to ensure a gas-tight seal between the collar and the lid. Chapter V: Analysis and Discussion Page 131 CHAPTER V Analysis and Discussion 5.1 Introduction The previous chapter was directed at the primary objective of the present study with respect to the development of a reliable apparatus (i.e., the dynamic closed chamber (DCC) method) for measuring C 0 2 fluxes from waste rock. This chapter extends the application of the D C C method and presents the results of the investigation for the influence of a short-term, multi-day (29 July to 5 August 2002) heavy rainfall event on waste-rock water conditions and associated C 0 2 fluxes from Deilmann north (DNWR) and Deilmann south (DSWR) waste-rock piles at the Key Lake uranium mine in northern Saskatchewan. The partial differential equation used for the C 0 2 model used in this thesis to quantify C 0 2 production and diffusion through unsaturated soils is also described. Results of the model validation and its application on mine waste-rock piles are presented. The main objectives of this chapter were to predict the influence of soil water on C 0 2 fluxes from mine waste-rock piles and to validate and apply the " C 0 2 " model to predict concentration-depth profiles and surface C 0 2 fluxes in the waste-rock piles. 5.2 Effects of Rainfall Events on Waste-Rock Surface Water Conditions and C 0 2 Fluxes Across the Surfaces of the Deilmann North (DNWR) and Deilmann South (DSWR) Waste-Rock Piles Chapter V: Analysis and Discussion Page 132 C 0 2 fluxes from both the DNWR and the DSWR were measured three times during the summer of 2000 (July to September 2000) and twice during the summer of 2002 (July to August 2002) as described in the previous sections. The total precipitation recorded for each year from 2000, 2001 and 2002 were 483.4, 524.6, and 548.9 mm, respectively. During this period, the year 2002 recorded the greatest precipitation (approximately 1.13 times higher) than that of 2000. From 1977 to 1998, average winter precipitation (October to April inclusive; predominantly as snow) was 163.6 mm, average summer precipitation (May to September inclusive; predominantly rain) was 294.8 mm (data obtained at Key Lake mine site). In general, rainfall accounts for approximately 64% of the average total precipitation at the Key Lake mine site. The effects of the rainfall events on the DNWR and the DSWR surface-water conditions and surface C 0 2 fluxes are discussed in the following sections. 5.2.1 Short-term effects of rainfall events on near surface-water conditions Figures 5.1 and 5.2 show the changes of measured volumetric (6) water contents at near ground surfaces (0 - 0.15 m) with time, following the cessation of 75.9 mm rainfall over an initial 48-h period [July 30 (day 1) to July 31 (day 2)] with a gradual decrease in rainfall from August 1 (day 3) to August 3 (day 5), at the DNWR and the DSWR. Results show that the ground surfaces of the piles (0 m, open circles) dried rapidly, whereas the drying rates at greater depths (0.05 m and below) decreased slowly with time (Figures 5.1 and 5.2) (see data in Appendix E). The ground surface of the DSWR (Figure 5.2) drains more rapidly than that at the DNWR. For example, on 31 July 2002 (day 3) the ground surface water content on the D S W R was about 0.06 compared with 0.23 on the DNWR. Both ground surfaces continued to dry rapidly with Chap te r V : Ana lys i s and D iscuss ion P a g e 1 3 3 40 35 + 30 + 25 + E E 75 20 + DC 15 + 10 + 5 + 3 Rainfall r i Stage II n = 0.36 S = 69.5% Stage III + 0.25 + 0.20 ^ CD \ \ • "•EL \ \ o J Z 1 +• 0.30 + 0.15 c CD C o a> CO + 0.10 + 0.05 0.00 4 5 Day # 6 8 o Water cont.: 0 m • 0.05 m A 0.10 m O 0.15 m Figure 5.1. Rainfal l and (30 Ju ly (day 1) to 6 Augus t waste- rock ( D N W R ) pile. volumetr ic water contents (0) measured over an 8-d test per iod! (day 8) 2002) at station DNF1 with t ime at. the De i lmann north; Chap te r V : Ana lys i s and D iscuss ion P a g e 134 40 35 + 30 + 25 + E = 20 + CO c •5 rx 15 + 10 + 5 + Slage Stage II Stage III + 0.25 + 0.20 n = 0.38 S = 31.5% *•* a" —• . H \ • • o A o J U L o • o • Rainfall Water cont.: 0 m 0.05 m 0.10 m 0.15 m +• 4 5 Day# 0.30 0 ; f c 0.15 ~ o o + 0.10 a> co + 0.05 I- 0.00 8 Figure 5.2. Rainfal l and water contents measu red over an 8-d test per iod (30 July (day 1 1) to 6 Augus t (day 8) 2002) at sampl ing station D S F 1 with t ime at the De i lmann south w a s t e - . rock ( D S W R ) pile. T h e porosity n=0.38. .; I Chapter V: Analysis and Discussion Page 135 time to water content values of about 0.10 and 0.004 (day 8), respectively, on the DNWR and the DSWR. The drying rates eventually diminished with time, although at greater depths (0.05 m and below), water contents remained elevated at the end of the test period (day 8). This behavior is caused by the reduction of the unsaturated hydraulic conductivity due to the decrease in surface-water content as evaporation continues (Shuttleworth, 1993; Wilson et al., 1994; Capehart and Carlson 1994; Ek and Cuenca 1994; Capehrt and Carlson 1997). The soil surface curves can be described as having three stages of drying as described in Hillel (1980) and Wilson et al. (1994). Stage I drying occurred during the wet period from day 1 to day 3 when the soil surfaces were nearly saturated. Stage II drying starts from day 3 after the cessation of heavy rainfall events. The beginning of the second stage of drying occurs rather abruptly (see Figures 5.1 and 5.2) as the soil surfaces rapidly dried out. The length of time for the second stage of drying lasts depends upon the intensity of the meteorological factors that determine atmospheric evaporativity, as well as upon the conductive properties of the waste-rock itself. The soil surface at the DNWR (Figure 5.1) dried out more slowly than that at the DSWR (Figure 5.2). For example, on day 3 (August 1, 2002) the surface-water content on the DNWR was about 0.23 as compared to 0.06 on the DSWR. The empirical rate of the decrease of ground soil surface-water (0 m) content (d8w/dt) can be described by (Gray, 1995): [5.1] Chapter V: Analysis and Discussion Page 136 where 6W is the volumetr ic water content (cubic metre of water per cubic meter of air), t is the t ime, and a and b are parameters related to the boundary condit ions and conductance properties of the soi l , respectively. The exponent b, which is related to soil diffusivity, is obviously most important, and the greater its value, the greater the dec rease in water content. The use of this model to develop descript ive equat ions for the rate of drying of the ground surface at the DSWR ( - = 28.67 * r 5 - 0 8 , R 2 = dfi 0.948) and the DNWR ( - - ^ - = 7.19 * t~ 3 - 3 0 , R 2 = 0.826) y ielded high correlation coefficients (using Microsoft Excel ) . The drying equat ions indicate that the drying rate at the DSWR is greater than that at the DNWR (e.g., the exponent b for DSWR is greater than that for the DNWR). Gray (1995) pointed out that, if the drying rates were limited only by a diffusion- limited process (i.e. vapor diffusion across the drying zone) , the exponents in the drying rate functions would be 0.5. Equat ion 5.1 is purely empir ical and does not attempt to account for flow mechan isms. For example , during drying, the water is s imul taneously redistributing away from the waste-rock ground sur faces (e.g., F igures 5.3) because of both upward flow due to evaporat ion and downward drainage due to gravity; thereby speeding decay of the surface drying rates. The redistribution tends to persist longer in the waste rock at the DNWR than that at the DSWR. The t ime-variable rate of redistribution depends not only on the hydraulic properties of the waste rocks, but a lso on the initial wetting depth, as well as on the relative dryness of the bottom layers (Hillel, 1980). Chap te r V : Ana lys i s and D iscuss ion P a g e 137 Figure 5.3. Vo lumetr ic water content (6) profi les measu red over an 8-d test period [30 Ju ly (day 1) to 6 Augus t (day 8) 2002] at (A) station D N F 1 at the De i lmann north waste- rock ( D N W R ) pile and (B) station D S F 1 at the De i lmann south waste- rock ( D S W R ) pile with t ime. Chapter V: Analysis and Discussion Page 138 5.2.2 Short-term effects of rainfall events on C 0 2 fluxes The changes in measured C 0 2 fluxes from ground surface following the cessation of 75.9 mm rainfall over the initial 48-h period [July 30 (day 1) to July 31 (day 2)] are presented in Figures 5.4 and 5.5 (solid circles). On 31 July 2002 (day 3), C 0 2 fluxes measured from the DNWR and the DSWR were 3% and 36 % of their initial average values of 217 and 188 mg m"2 h"1, respectively. The figures showed that the changes of surface C 0 2 fluxes with time were negatively correlated with measured surface water contents for the-waste rock sand. As the water contents at ground surfaces decreased exponentially, the surface C 0 2 fluxes increased exponentially from day 3 to day 8. These inverse linear relationships yielded correlation coefficients of R 2 = -0.997 and R 2 = -0.820 (using Microsoft Excel) for the DNWR and the DSWR, respectively. By the end of the 8-d test period, the surface C 0 2 fluxes had increased by factors of 4 and 45 while the ground surface water contents had decreased from 6.7% to 0.04% and from 25.0% to 1.5% at the DNWR and the DSWR, respectively, and the measured C 0 2 gas fluxes approximated their initial mean flux values. This observation suggested that it takes about 5 to 6 d after a heavy rainfall event for the gas fluxes to approach pre-rainfall values. In addition it is further suggested that the impact of rainfall events on C 0 2 fluxes from the waste-rock piles is of relative short duration. Chapter V: Analysis and Discussion Page 139 Figure 5.4. Rainfall, water contents, and C 0 2 fluxes measured at station DNF1 over an 8-d test period [30 July (day 1) to 6 August (day 8) 2002] at the Deilmann north waste-rock (DNWR) pile with time. Chapter V: Analysis and Discussion Page 140 Day# mmmA Rainfall o Surf, water cont. (0 m) • CQ2 F i g u r e 5.5. Rainfall, water contents, and C 0 2 fluxes measured at the DSF1 over an 8-d test period [30 July (day 1) to 6 August (day 8) 2002] at the Deilmann south waste-rock (DSWR) pile with time. Chapter V: Analysis and Discussion Page 141 The functional relationship between the measured surface CO2 flux and the surface-water content is also shown in Figure 5.6. Results showed that the surface CO2 flux is sensitive to changes of waste-rock surface-water content after heavy rainfall event, exhibiting a power decrease with surface-water content of the form: F C 0 2 = a * e - b [5.2] where Fco2 is the surface C 0 2 flux (milligrams per square meter per hour), 0W is the volumetric water content (cubic metre of water per cubic metre of air), and a and b are parameters related to the boundary conditions and conductance properties of the porous media, respectively. The use of this model to develop descriptive equations showed that a good relationship between the surface C 0 2 flux and the ground surface water content of the waste rock at the DNWR (FC 02(N) = 4 .71*9~ 1 1 5 , R 2 = 0.790) and DSWR (F C 0 2(S) = 53.50 *0~ a 2 1 , R 2 = 0.846) (Figure 5.6). The difference in the coefficients a and b between the two piles is attributed to textural variability that affects the water content and the diffusivity of CO2, which is also a function of water content. In summary, measurements showed that the gas-flow conditions at the ground surfaces of the DNWR and DSWR were transient after a heavy rainfall. The transient effects were attributed to rapid drainage and evaporation. The effect of heavy rainfall on water-content profiles and C 0 2 fluxes was of a relatively short duration. Chapter V: Analysis and Discussion Page 142 5.3 Predictions of Evaporative Fluxes and Near-Surface Water Contents Profiles It was noted from the previous sections that rainfall events can create changes in soil water content and CO2 gas profiles within unsaturated zones and that the extent of the effect depends on the intensity and duration of the rainfall. Evaporation from mine wastes is a crucial component of the water balance (Carey et al., 2005). Similarly, soil water evaporation significantly affects water content, and as a results, the degree of saturation of the soil and the gas diffusion. Knowledge of the rate of evaporation at the soil-atmosphere interface is required to estimate the water content of candidate cover soils. The one-dimensional SoilCover computer model (Unsaturated Soils Group, 1997) was used to estimate evaporative fluxes at the DNWR and the DSWR over the 8-d test period [(30 July (day 1) to 6 August, 2002 (day 8)]. The model calculates daily evaporation on a site-specific basis, using weather data collected at the site as a boundary condition for the calculation of actual evaporation. The weather data (radiation, air temperature, humidity, and wind speed, etc.) is used in combination with soil characteristics and the calculated changes in soil moisture (details data input, output and results are also presented in Appendix H). The SoilCover predicted evaporative fluxes at the DSWR were compared to published measured values obtained by Carey et al. (2005) from the DSWR during the same test period. Chap te r V : Ana lys i s and D iscuss ion P a g e 143 400 350 300 -- . f 250 -- E E. 200 f x 3 « 150 O O 100 -- 50 F C 0 2 = 5 3 . 5 0 X 0 2 1 0 1 R 2 = 0.846 + 0.0 DNWR DSWR - - 1 Power (DNWR) Row er (DSWR) F c c e = 4 . 7 1 x - " ^ R 2 = 0.790 0.2 0.4 Saturation (S) 0.6 0.8 Figure 5.6. Var ia t ions in C 0 2 flux measurements with sur face-water saturat ion (S=8/n) measu red over an 8-d test period [30 Ju ly (day 1) to 6 Augus t (day 8) 2002] at stat ions D N F 1 and D S F 1 of the D N W R (n=0.36) and D S W R (0.38) pi les, respect ively. Chapter V: Analysis and Discussion Page 144 5.3.1 Short-term predictions of evaporative fluxes Figures 5.7 and 5.8 show the SoilCover model predictions of cumulative evaporative fluxes and the ratio of actual (AE) and potential (PE) evaporation (AE/PE) as a function of time for the 8-d test period at the DNWR and the DSWR, respectively. Simulations results indicated that during the period of heavy rainfall events from day 1 to day 3 the evaporation rate was relatively low. During this period the cumulative P E and A E were equal at both the DNWR and the DSWR. This stage is being referred as stage I drying (Wilson et al., 1994). Wilson et al. (1997) noted that the A E is approximately equal the P E rate of evaporation until the value of matric suction reaches approximately 3000 kPa. During this period the evaporation is controlled by external meteorological conditions (Hilled, 1980; Wilson et al., 1994). As the ground surfaces continued to dry from day 3 to day 5 the rate of evaporation started increasing rapidly. The cumulative evaporation was slightly higher at the DSWR (PE = 5.3 mm) than at the DNWR (PE = 4.5 mm) on day 5. After day 5, the values of Actual rate of evaporation and Potential rate of evaporation started to progressively diverge with A E less than P E , but slight faster at the DSWR than at the DNWR (Figures 5.7B and 5.8B). Moreover, results showed that during the separation of the A E and P E , the water contents had dropped dramatically from 0.25 on day 3 to about 0.003 on day 5 at the DNWR and from 0.07 on day 3 to about 0.001 on day 5 at the DSWR, respectively. At the end of the 8-d test period the model simulation results indicated 9.5 and 10.9 mm cumulative PEs for the DNWR and the DSWR. These values represent averages daily evaporation rate of 1.2 and 1.4 mm d"1 for the 8-d test period for the DNWR and DSWR respectively. Carey and co-workers (Carey et al., 2005) directly measured summer evaporation (6 June to 25 August, 2002) using eddy (EC) Chap te r V : Ana lys i s and D iscuss ion P a g e 145 (A) E x -12 -10 + o CL ro > a> > -8 + -6 + -4 + 0 4 - I Rainfall • P E A E - H -Vo l .wa te r Figure 5 .7 . (A) Rainfa l l , water contents measu red , and So i lCove r predicted evaporat ive f luxes at the De i lmann north waste- rock ( D N W R ) pile (B) ratio of actual (AE) and potential (PE) evaporat ion ( A E / P E ) as a funct ion of t ime over an 8-d test per iod [30 Ju ly (day 1) to 6 Augus t (day 8) 2002]. Chap te r V : Ana lys i s and D iscuss ion P a g e 146 Day* Rain PE AE - H - Vol. water ( B ) 0.4 -- 0.2 -- 0.0 -I 1 1 1 1 r 1 2 3 4 5 6 7 8 Day* Figure 5.8. (A) Rainfa l l , water contents measu red , and S o i l C o v e r predicted evaporat ive f luxes at the De i lmann south waste- rock ( D S W R ) pile and (B) ratio of actual (AE) and potential ( P E ) ( A E / P E ) evaporat ion as a function of t ime over an 8-d test per iod [30 Ju ly (day 1) to 6 Augus t (day 8) 2002]. Chapter V: Analysis and Discussion Page 147 covariance method at the DSWR: They measured the cumulative A E s (data not shown here) for the 8-day test period and found the cumulative A E of 8.0 mm with an average evaporation of 1 mm d"1 at the DSWR (Carey et al., 2005). The results showed good agreement between SoilCover model predicted and E C measured A E s data for the 8-d test period at the DSWR. Carey et al. (2005) also noted that the measured A E was significantly less than the PE at the DSWR due to high surface albedo that reduce available energy for evaporation. Figures 5.9 and 5.10, respectively, show SoilCover model simulation results of cumulative P E and A E for a 27-d period (July 29 to August 24, 2002) for the DNWR and the DSWR. During this period subsequent rainfall events occurred between day 16 and day 19 where a total of 26.9 mm fell at the sites. These rainfall events are depicted in Figures 5.9B and 5.10B where the ratios A E / P E equal to unit (AE/PE=1). At the end of the 27-d simulation period the model results yielded cumulative A E s of 32 and 35.6 mm with ratio of PE /AE of 1.44 and 1.37 for the DNWR and the DSWR, respectively. These results represent averages A E evaporation of 1.2 and 1.3 mm d"1 at the DNWR and DSWR respectively. Carey et al. (2005) field-measured data indicated cumulative A E of 37 mm with an average of 1.4 mm d"1 for the 27-d test period at the DSWR. These results show good agreement between model (SoilCover) simulations and measured A E s values for the 27-d test period at the DSWR. In summary, the comparison between the SoilCover predicted A E evaporation and the E C measured A E (Carey et al., 2005) indicated the ability of the SoilCover model to predict, with sufficient accuracy the A E at the surfaces of the waste-rock materials. Chap te r V: Ana lys i s and D iscuss ion P a g e 148 F i g u r e 5 . 9 . So i lCove r predicted evaporat ive f luxes (A) actual A E and potential P E and (B) the ratio of A E / P E at the De i lmann north waste- rock ( D N W R ) pile over a 27-d test period [29 Ju ly (day 1) to 24 Augus t (day 27) 2002] with t ime. Chap te r V: Ana lys i s and D iscuss ion P a g e 149 -50 A 1 1 1 1 1 1 1 1 1 1 1 r 1 3 5 7 9 11 13 15 17 19 21 23 25 27 Day# (B) 0.2 -- 0.0 -I 1 1 1 1 1 1 1 1 1 1 1 1 1 3 5 7 9 11 13 15 17 19 21 23 25 27 D a y # F i g u r e 5 . 1 0 . So i lCove r predicted evaporat ive f luxes (A) actual A E and potential P E and (B) the ratio of A E / P E at the De i lmann south waste- rock ( D S W R ) pile over a 27-d test period [29 Ju ly (day 1) to 24 Augus t (day 27) 2002] with time. Chapter V: Analysis and Discussion Page 150 5.3.2 Short-term predictions of near-surface water contents profiles SoilCover was also used to predict the changes in water content profiles after the rainfall events over the 8-d test period [(30 July (day 1) to 6 August, 2002 (day 8)] at the DNWR and DSWR. In the simulations, the initial water content profile was required in order to predict the subsequent water profiles. The initial measured water contents profiles and the soil properties of the DNWR and DSWR were used as inputs data for the model simulations. The upper boundary value of water content of 0.03 (3%) was also specified in the model as initial water conditions during the simulations. Figures 5.11A and 5.12A present the measured water contents profiles and Figures 5.11B and 5.12B present the SoilCover predicted water contents profiles at the DNWR and DSWR piles. Both the measured and SoilCover predicted data show that the water contents conditions at the ground surfaces of the DNWR and the D S W R were transient after the heavy rainfall events. The transient conditions at both the ground surfaces were attributed to the sandy texture of the waste-rock piles and their associated high saturated hydraulic conductivities. The water redistribution appears to persist longer in the waste rock at the DNWR than at the DSWR and this was attributed to slight variations in the waste-rock textures that control soil water. Chap te r V: Ana lys i s and D iscuss ion P a g e 151 Figure 5.11. Compar i son of (A) measured and (B) So i lCove r predicted water content profi les for the 8-day test period [July 30 (day 1) to Augus t 6 (day 8), 2002] at the De i lmann north waste- rock ( D N W R ) pile. Chap te r V : Ana lys i s and D iscuss ion P a g e 152 Figure 5.12. C o m p a r i s o n of (A) measured and (B) S o i l C o v e r predicted water content profi les for the 8-day test period [July 30 (day 1) to Augus t 6 (day 8), 2002] at the De i lmann south waste- rock ( D S W R ) pile. Chapter V: Analysis and Discussion Page 153 5.4 C0 2 Diffusion Prediction and Model Proposed This section presents the important theoretical relationships used to quantify CO2 production and diffusion through unsaturated geologic media. The theoretical of the partial differential equation describing the change in C 0 2 concentration with depth and time as a function of C 0 2 production and diffusion is also presented. The section is concluded with the description of a relatively simple computer code program used to solve the finite difference formulation of the governing equation of the model developed in this work. The processes of diffusion, production of CO2, the equations describing one-dimensional transient C 0 2 diffusion and production are described below. 5.4.1 C 0 2 diffusion Gas dynamics in most soil systems is a three-dimensional problem, but the use of one-dimensional models is generally accepted (de Jong and Schappert, 1972; Collin and Rasmuson, 1988). One-dimensional CO2 diffusion in a gaseous environment is commonly described by Fick's First Law that defines the mass flux of C 0 2 in a given direction as directly proportional to the negative of the concentration gradient in that direction (Fetter, 1993): [5.3] where: FCo2 = mass flux of C 0 2 (kg m"2 s"1), D = the free air diffusion coefficient (m 2 s"1), C = C 0 2 concentration (kg m"3), and Chapter V: Analysis and Discussion Page 154 Z = depth (m). The use of Equation 5.3 assumes that Fick's law adequately describes the diffusive gas flux. For gases such as C 0 2 , which have sources or sinks in the system and constitute a small fraction of the total system pressure, this appears to be true (Thorstenson and Pollock, 1989). The diffusion coefficient in air (similar to atmospheric composition) for C 0 2 is 1.39x10"5 m 2 s" 1 (at 0°C) (Weast and Astle, 1981). The diffusion coefficient increases with increasing temperature and decreasing molecular weight (Fuller etal . , 1966). Aubertin et al. (2000) described the use of an equivalent porosity to represent the effective porosity available for the diffusion of oxygen. Applying this relationship for C 0 2 diffusion transforms the water porosity into an equivalent air porosity by portioning it with Henry's Law coefficient as follows: 0eq=e a+ewH [5.4] where: 9 e q = equivalent porosity (m 3 m"3), 6 a = air porosity (m 3 m"3), 9 W = water porosity (m 3 m"3), and H = Henry's Law coefficient (approximated as 0.03 for C 0 2 in air and water at 25 °C) (Hendry et al., 993). Increasing water saturation decreases the equivalent and effective porosity and reduces C 0 2 diffusion. Using Henry's law to represent phase partitioning of a reactive gas, such as C 0 2 , is an approximation of the true process (Hendry et al., 1993). Chapter V: Analysis and Discussion Page 155 Fick's First Law defining C 0 2 diffusion through porous media as a function of the equivalent porosity is defined by: F c o 2 = - e e q D ^ [5.5] where: F = mass flux of C 0 2 (Kg m"2 s"1), B e q = equivalent porosity (m 3 m"3), D* = bulk diffusion coefficient (m 2 s"1), C = C 0 2 concentration (Kg m"3), and Z = depth (m). The equivalent porosity and the bulk diffusion coefficient (D*) are often combined into a variable D e , the effective diffusion coefficient, to give: D e = 9 e q D * [5.6] Then Equation 5.5 can be written in terms of the effective diffusion coefficient as: Fco2=- D e § E5-7] In soils both diffusion and chemical reactions will determine the C 0 2 gradient as described by Fick's second law. Assuming steady state conditions, this law can be written as (Hendry et al., 1999): Chapter V: Analysis and Discussion Page 156 D d2C edz2 = - G [5.8] where G is a reaction rate (pg C O 2 *g dry soil"1*d"1). Aubertin et al. (2000) and Mbonimpa et al. (2003) also defined the effective diffusion coefficient (D e) from Equation 5.6 as a function of the components of the diffusion in the air and water phase as represented in Equation 5.9. D e = D a + H D w [5.9] Where: D a = diffusion coefficient component through air phase (m 2 s"1), D w = diffusion coefficient component through water phase (m 2 s"1), H = Henry's coefficient as defined above. = 0 a D ° T a and D w = 9 W D ° T w — " w ^ w 1 w [5.10] where: D° = diffusion coefficient of C 0 2 through air (m 2 s"1), D° = diffusion coefficient of C 0 2 through water (m 2 s"1), T a = tortuosity coefficient for air phase, and T w = tortuosity coefficient for water phase. The tortuosity coefficients are related to the properties of the material through the following equations (Collin and Rasmuson, 1988; Mboninpa et al., 2003): Chapter V: Analysis and Discussion Page 157 2x+1 [5.11] 0 2 y + [5.12] e|x +(i-e a) x =1 [5.13] e 2 w y+(i-ew) y=i [5.14] where: 9 = total porosity. Mbonimpa et al. (2003) noted that a reasonable estimation of the value of the variables x and y is 0.75. Using this value for x and y and combining Equations 5.9 - 5.12, the diffusion coefficient equation can be simplified to the Equation 5.15 (Aachib et al., 2002, 2004). Equation 5.15 was used in the model adopted in this thesis for the evaluation of CO2 diffusion. An example of the variation of the effective diffusion coefficient D e as a function of water porosity 6 W is illustrated in Figure 5.13B. The water contents profiles in Figure 5.13A represent hypothetical drying forcing conditions generated in unsaturated sand material. The initial water-depth profile (curve d1, Figure 5.13A), however, represents De=4P°a 0a 5 + H D w0w 5 e [5.15] Chapter V: Analysis and Discussion Page 158 1.0 -I 1 1 1 1 0 10 20 30 40 Vol. water content (%) Figure 5.13A. Hypothetical water content profiles in unsaturated sand Material. Curve d1 is an actual measured water profile in the HT minicosm. d 4.0E-06 -- 3.5E-06 -- 3.0E-06 -- 2.5E-06 -- 2.0E-06 -- 1.5E-06 -- 1.0E-06 -- 5.0E-07 -- A d1 • d2 • d3 • d4 A d5 « d 6 o d6 o d7 o d8 X d9 + + 10 20 30 Volumetric water content (%) •fir 40 Figure 5.13B. Simulated effective diffusion coefficient (De) of C 0 2 as a function of water content using hypothetical data presented above. Chapter V: Analysis and Discussion Page 159 the measured mean water-depth profile in the HT minicosm column described in Section 3.3 of Chapter 3 of this thesis and in Kabwe et al. (2002). The subsequent profiles were generated by reducing the initial water-depth profile (curve d1) by a factor of 0.1 consecutively. The corresponding simulated changes of the effective diffusion coefficient D e of C 0 2 was computed using Equation 5.15 with the parameters: 6 = 0.40 and = 1.39x10"5 m 2 s" 1 and H = 0.03. The general trend shows an increase in the D e with a decrease in water content or vice versa, as shown in Figure 5.13B. The dependency of the D e on soil water content for different textured soils is well documented (Klute and Letey, 1958; Rowell et al., 1967; Collin and Rasmuson, 1988; Mbonimpa et al., 2003). The diffusion coefficient of C 0 2 in water is about four orders of magnitude slower than that in the air-filled voids. 5.4.2 Biotic C0 2 production rate As discussed in the previous chapter C 0 2 can be produced in biotic reaction (e.g., microbial respiration) or in abiotic (e.g., carbonate buffering) reactions. However the model developed in this thesis will only be focused on biotic reaction. It should be noted that the D C C method was tested and verified in mesocosms filled with fine- grained sand excavated from the C-horizon of an unsaturated zone at a local field site located near Saskatoon, Saskatchewan (Hendry et al., 1993; Kabwe et al., 2002). Studies on microbial aspects of the mesocosm indicated that biological activity within the mesocosm was likely sufficient to account for the generation of C 0 2 throughout the profile (Hendry et al., 2001; Lawrence et al., 1993). The " C 0 2 " model described in this thesis will be also tested and validated with mesocosms measured data. Chapter V: Analysis and Discussion Page 160 The CO2 production (microbia) rate (G) can be described by a function similar to that used by Hendry et al. (1999): G = 6^G 0[g(T)g(ew)g(z)] [5.16] Where: G = CO2 production rate (kg C kg"1 dry soil day"1), G 0 = reference production rate ( kg C kg"1 dry soil day"1), g(T) = the production contribution based on temperature, g(9w) = the production contribution based on soil moisture content, g(z) = the production contribution based on depth. The function provides the option of determining the G term as a function of temperature, soil moisture content, and/or depth. The functional dependence of production upon temperature, soil moisture content and depth are (Hendry et al., 1999): g(T)=e k< T- T> [5.17] when: T> T m i n and k is arbitrary The lowest temperature at which CO2 production occurs is T m j n . g(T) = 0 [5.18] When: T < I min Chapter V: Analysis and Discussion Page 161 g(ew)=0w [5.19] where: a = arbitrary g(z)=e -bz [5.20] where: b = arbitrary. These functions represent the influence of the primary independent variables (Hendry et al., 1999): g(T) is the Arrhenius equation, where a Qio (=e 1 0 k ) value determines the degree to which respiration increases with a 10°C increase in temperature; g(z) represents the commonly observed (e.g., Simunek and Suarez, 1993) exponential decrease in productivity with depth; and the combination of Oag(0w) serves to reflect the reduction in activity which typically occurs at high and low water contents (Ekpete and Cornfield, 1965; Rixon, 1968; Grant and Rochette, 1994; Hendry et al., 1999). The biotic production rate used in the model developed in this thesis was represented by a function similar to that used by Hendry et al., 1999 (the Equation 5.21): G = G 0 e ! e w e k r [5.21] where: kr = constant in the Arrhenius equation (°C 1 ) (k r=0.044°C"1), T = measured temperature (°C), and Chap te r V: Ana lys i s and D iscuss ion Page 162 T = reference temperatures (°C). The parameters a and b are fitting parameters. Note that the coefficient Qio(=e10kr) is often used to represent the relative increase in respiration intensity per 10°C increase in temperature. The value of k=0.044°C indicates an average Q10 of 1.6. The variability in Q10 is most likely attributed to differences in microbial community structure. It is acknowledge that sensitivity analysis demonstrated that G is only weakly dependent on 1 < b< 3 (Hendry et al., 1999). Figure 5.14 illustrates the simulated microbial respiration rates as a function of temperature and water content using Equation 5.21. Values of the parameters in Equation 5.14 used in the simulations were specified by Hendry et al. (2001): a = 2; b =1.25, k = 0.044 °C; G 0 = 207 ^g C.g" 1.d" 1 when T = 6.17 °C. The production of C 0 2 was attributed to microbial activity in the C-horizon sand (Hendry et al., 2000). The simulated results show that at low water content (9W), C 0 2 production decreases because of a lack of water; at high 6 W , production also decrease because of excess water filling the pore spaces. The model simulation shows that the maximum microbial CO2 production occurred at a water content of 0.25 that corresponds to a water saturation of 70% (e.g., for 6 = 0.40). This is within values reported in the literature. Linn and Doran (1984) observed that soil incubated with 60% soil pore space filled with water, supported maximum aerobic microbial activities. Chapter V: Analysis and Discussion Page 163 3 . 0 E - 1 1 Soil water content Figure 5.14. Simulated microbial respiration rates as a function of tempe- rature and water content using Equation 5.21. 5.4.3 Development of the partial differential equation The following section presents the development of the partial differential equation describing the change in CO2 concentration with depth and time as a function of CO2 diffusion and production. Let's consider a representative elementary volume (REV) for derivation of the partial differential equation as illustrated in Figure 5.15. Chapter V: Analysis and Discussion Page 164 "out R E V F i g u r e 5.15. Representative elementary volume, REV, for derivation of partial differential equation. By applying the conservation of mass principle to a porous, cubic volume (dimensions: dx, dy, dz) through which a gaseous species is diffusing (the mass flux entering the volume (F i n ) minus the mass flux exiting the volume (F o u t ) must equal the change in storage (Sm/dt): where F = mass of gaseous species per unit of area per unit of time in the z direction. (Fin - Fout )dxdy + G(dxdydz) = ^ [5.22] at where: G = production rate of gaseous species within cubic volume (kg C kg"1 dry soil day"1), m = total mass of gaseous species within cubic volume (kg m"3), and t = time (s). Chapter V: Analysis and Discussion Equation 5.22 can be rewritten as: Page 165 f f Fin" V V SF ^ F i n + — dz , n dz dxdy + G(dxdydz) = am [5.23] The mass of gaseous species found in the pore gas and pore water of the cubic volume can be represented by: m = ( C a 6 a d x d y d z ) + ( H C a e w d x d y d z ) [5.24] where: C a = mass of gaseous species /volume of pore gas (kg m"3), H = Henry's law coefficient, 6 a = air porosity (m 3 rrf3), and 9 W = water porosity (m 3 m"3). Considering Equations 5.8 ( F = - D e — ) and 5.24, Equation 5.23 can be rewritten: dz D e ~ dz dxdydz • + G(dxdydz) = g (C a 9 a dxdydz + H C a 9 w d x d y d z ) [5.25] Assuming that 9 a and D e do not vary over the depth interval dz and that 9 a , 9 W ) H, dx, dy, and dz do not vary over time interval dt, Equation 5.25 can be simplified to Equation 5.26 similar to Hendry et al. (2001).: Chapter V: Analysis and Discussion Page 166 D e ^ - + G = ( e a + H 6 w ) 8Ca [5.26] dz dt Equation 5.26 is the governing equation used in the model. 5.4.4 Finite difference formulation The simple model developed in this thesis used a finite difference numerical method to solve the partial differential equation given by Equation 5.26. The numerical method offers a discrete approximation to problems with complex physical properties and geometry, but requires numerous calculations, which are lessened by the use of digital computers capable of performing numerous calculations quickly. The finite difference formulation of Equation 5.26 is a simple calculation based on approximating the derivatives of the function, resulting in a solution only at the discrete points (Lin et al., 1997). For a number of nodes there will be n linear equations, hence the problem may be solved (Freeze and Cherry, 1979). The finite difference method has many advantages, including: simple problems are easily solved, abundance of literature, successful algorithms are available to solve the system of equations and the accuracy is good. To develop the finite difference formulation defining the change in concentration at a given node, three nodes were defined as shown in Figure 5.16. The nodes represent the center of the finite difference element. The three mass fluxes (Fin, F o ut. and Fprad) are defined in Equations 5.27, 5.28, and 5.29: Chapter V: Analysis and Discussion Page 1 6 7 0 . I prod out Figure 5.16. Three nodes and the mass fluxes entering and exiting nodel for development of the finite difference formulation. F o u t = " D e — = - D dC _ ( C a - C O 8z~ " e ^ \ ( z 2 - z i ] [5.27] F - - D ^ - - D F . n - ° e a z D e o , 1 | ( Z i _ Z o j [5.28] Fprod = G A Z 0 , 1 , 2 [5.29] Substituting these three equations and Equation 5.23 into Equation 3.19 results in Equation 5.30 fo-Cp) ( C 2 - C l ) 9 e q A C i K i . 2 | ' e o , 1 i( Z l- Z or e1-2 i ( z 2 - Z i i At [5.30] Chapter V: Analysis and Discussion Page 168 and solving Equat ion 5.30 for the change in concentrat ion ( A C i ) g ives Equat ion 5.31, which def ines the change in concentrat ion at node 1 over a given t ime-step (At): A C 1 = At [ n ( C j - C p ) ( C z - C Q + G [5.31] where: Co(t), Ci(t) and C 2 ( t ) are the concentrat ions of C 0 2 at t ime (t) at three adjacent nodes of increasing depth numbered 0, 1 and 2; z 0 , Z i and z 2 are the depths below ground surface of the three nodes; D eo,i(t) and D e i , 2 ( t ) are the effective diffusion coefficients between nodes 0 and 1, and 1 and 2 respectively, determined at t ime (t); 0 e q i ( t ) is the equivalent porosity at node 1 at t ime t and; A z 0 , i , 2 is the distance between the midpoint between nodes 0 and 1 and the midpoint between nodes 1 and 2. D e n ,n+i( t ) is calculated from the mean of D e n ( t )and D e n +i( t ) . The max imum length of each t ime step within the diffusion model was determined by: ^ 4 = 0.5 [5.32] A Z 2 The variable defined as A z 0 , i , 2 in Equat ion 5.31 is the average of the two spaces on either s ide of the node 1. A value of At was calculated at every node in the profile, and the smal lest t ime step was used . For all the var iables, the subset numbers separated by c o m m a s indicate the node(s) from which the variable must be calculated. Va lues of T and 0 W were interpolated onto the grid in both t ime and space . Chapter V: Analysis and Discussion Page 169 The boundary condit ion for the finite difference solution is a constant concentrat ion at the top. Atmospher ic concentrat ion is the constant value for the sur face node (e.g., 0 .036% CO2 a tmospher ic concentration). 5.5 C o m p u t e r C o d e P r o g r a m A s imple computer program cal led " C 0 2 " was written using Mac ro V isua l Bas i c of Microsoft Exce l to so lve Equat ion 5.30. The full V isua l Bas i c codes for the model is provided in Append ix B. A flow chart for the program is shown in Figure 5.17. The model uses water content matrix (depth and time) as the input for the diffusion and production calculat ions. The model is therefore, able to use the So i lCove r water content (or saturation) output as input to the " C 0 2 " model to calculate the change in CO2 concentrat ion with depth and t ime as a function of CO2 diffusion and production. The upper boundary condit ion (depth of 0 m) of the model was constrained to volumetric concentrat ion of 0.036 % for C 0 2 . Th is represents the relative concentration of CO2 in the atmosphere. The model required va lues for soil porosity (bulk), volumetric water contents profile and temperature. C 0 2 concentrat ion profi les were a lso required to run the model . Initial concentrat ion depth profiles provided a starting point for the model while concentrat ion profiles at a later t ime provide the model with va lues it could attempt to match. C 0 2 production was determined as a function of G 0 , soil air porosity (6 a), and soil moisture content (0W) with G 0 being the only fitted parameter. T h e arbitrary parameter " a " was set at 2 which maximized the product of of and 0 W at a degree of saturation of 0.70 (where degree of saturation is the ratio of the vo lume of water-fil led voids to the vo lume of total void space) . Max imum reactivity at a degree of saturation of 0.70 was reasonable because ample amounts of both water and CO2 (from pore gas) Chapter V: Analysis and Discussion Page 170 Set-up Major Matrices Write spreadsheet of water contents matrix Write values of constants User Input Input # of nodes Input initial cone. Ifiput porosity Input Go value Variable Assignments Interpolate a saturation values for each node Calculate the time-step for each node and compare the minimum value to max/min time-step. Calculate the diffusion coefficient for each node. Calculate the average value of the diffusion coefficient and nodal spacing Finite Difference Calculation Calculate change in concentration Calculate new concentration profile. 1 Output - Write concentration matrix - Write Diffusion coefficient matrix Figure 5.17. Flowchart for theJVisual Bas i c program Chapter V: Analysis and Discussion Page 171 for reactions would be avai lable. A degree of saturation of 0.70 a lso agreed with literature va lues of max imum respiration rates. The program structure consis ted of two nested loops: the innermost loop and the outer loop (Figure 5.17). The innermost loop occurs for each t ime-step and is where the finite difference calculat ion takes place. The outer loop occurs for each "day" where the program starts by creating all major matr ices. The user uses the input spreadsheet (descr ibed later in this section) to input the initial concentrat ion profile and the total porosity (assume to be the s a m e throughout the profile). The minimum and max imum t ime-step value is specif ied but can be changed if des i red. T h e s e va lues limit how smal l or how large the t ime-step value get. The t ime-step is calculated as a function of the coefficients in the finite difference equat ion. The formula used to calculate the t ime-step is given in Equat ion 5.30 that def ines the t ime-step required for mathematical stability (Zill and Cu l len , 1992). It was determined from trial simulat ions that for most model ing scenar ios , 350 iterations were required to reach stability (See Figure 5.18). T h e model calculates the t ime-step for each node then takes the minimum value and compares it to the max imum and minimum time-step specif ied by the program or the user. A s noted in the Figure 5.18, convergence w a s ach ieved after 350 iterations. T h e stability w a s poor below 200 iterations. Hence , 350 iterations were performed for each simulat ion. The output of the model is a spreadsheet file containing the: day #, iteration #, nodes, new concentrat ion, concentrat ion changes , diffusion, saturation and time difference calculated va lues are presented in Append ix B. Chapter V: Analysis and Discussion Page 172 Ui c o ( 0 g 8.0E-09 + | 6.0E-09 H O 4.0E-09 2.0E-09 0.0E+00 F i g u r e 5.18 Stability curves generated by the model for different iterations using time steps of 0.05 day. Chapter V: Analysis and Discussion Page 173 5.6 Application of the C0 2 Model Using Measured Values in Sand Minicosms The theory and development of a numerical model for C 0 2 diffusion and transport were presented in the previous sect ions. The model was based on the finite difference method to so lve the one-d imensional diffusion equation using the program Mac ro V isua l Bas i c . The exper iments for the dynamic c losed chamber ( D C C ) method were des igned and carr ied out to evaluate the ability and accuracy of the D C C to measure C 0 2 f luxes under actual field condit ions on the sur faces of the D N W R and D S W R . However , no instrumentation was installed to measured C 0 2 concentrat ions and gradients in the shal low profile within the upper meter of the waste rock pi les. Therefore, it is not possib le to rigorously test the full utility of the C 0 2 model following the heavy rainfall event similar to the So i lCove r model ing that was conducted to predict changes in soil water content. The " C 0 2 " model is evaluated in this sect ion for the prediction of C 0 2 concentrat ions measured in the min icosms exper iments previously descr ibed in sect ion 3.3 (Kabwe 2001 and Kabwe et a l . , 2002). The simulation results are interpreted and d i scussed in the following sect ions. 5.6.1 Prediction of C0 2 concentration profiles in response to changes in water contents profiles It was shown in the previous sect ions that changes in microbial respiration can result from changes in temperature and water content. In the following simulat ions hypothetical water content profiles for the sand column were used to illustrate the Chapter V: Analysis and Discussion Page 174 effects of water content on the effective diffusion coefficient and C O 2 gas concentrat ion depth- profiles in the sand co lumn. Figure 5.19A shows the hypothetical water profiles in a sand co lumn. T h e s e profiles were generated by progressively reducing the initial water content profile (curve d1) by a factor of 0.8 consecut ively. It should be noted that the initial water content profile (curve d1) is a real measured water profile of a sand column (HT) descr ibed in Kabwe (2001) and Kabwe et al.( 2002). In this example , the initial water contents at 0, 0.45 and 0.9 m depths (curve d1) were 12, 20 and 34 % respectively. The final hypothetical water contents at 0, 0.45 and 0.9 m depths (curve d10) were 3, 4 and 8 % respectively. The corresponding starting CO2 concentration profile (Figure 5.19B, curve d1) represented the actual measured CO2 concentrat ions for the sand column descr ibed in Kabwe et a l . (2002). In this example , the CO2 concentrat ions at 0.15, 030 and 0.60 m depths were 0.082, 0.14 and 0 .15% respectively. The subsequent simulated changes in C 0 2 profiles in the column in response to changes in the water contents profiles presented in Figure 5.19A are shown in Figure 5 .19B (curves from d1 to d lO) . A s the soil water content changes from wet to dry condit ions (Figure 5.19, curves from right to the left) the CO2 concentration profiles a lso increase proportionally (Figure 5.19B, curves from left to the right). At the end of the simulation the CO2 concentrat ion at 0.15, 0.30 and 0.60 m depths were 0.13, 0.17 and 0.20 %, respectively. S ince a constant C 0 2 flux was appl ied to the base of the HT column during the simulat ion, the change in the C 0 2 concentrat ions profiles was due to the change in the effective diffusion coefficient (D e ) (Equat ion 5.15) for C 0 2 , which is a function of water content. The general trend showed a dec rease in the D e with an increase in water content. Chapter V: Analysis and Discussion Page 175 1.0 -I 1 1 < 1 0 10 2 0 30 4 0 Vol. water content (%) F i g u r e 5 .19A. Hypothetical water contents profiles in a sand material described in Figure 5.13A. 1.0 -I 1 1 1 1 1 0 0.05 0.1 0.15 0.2 0.25 C 0 2 c o n c e n t r a t i o n (% b y vol.) F i g u r e 5 .19B. Model predicted C 0 2 concentrations profiles in a HT sand column obtained with hypothetical simulated water contents profiles (Figure 5.19A) and an initial measured C 0 2 concentrations profile (d1) in HT column (Kabwe et al., 2002). Chapter V: Analysis and Discussion Page 176 5.6.2 Simulations of C 0 2 Concentration Profiles using Sand Minicosm- Measured Data In order to test the ability of the " C 0 2 " model to predict the C 0 2 diffusion in sand material, s imulat ions were performed using the min icosm-measured data descr ibed in sect ion 3.3 of Chapter 3 of this thesis and in Kabwe (2001), Kabwe et a l . (2002) and Richards (1998). O n e min icosm was kept at room temperature (18 - 23 °C) (HT) and another at 5 °C (LT) (see Appendix F). The min icosm experiments started after the min icosms were filled with about 634 kg of sand excavated from an unsaturated C - horizon at a field descr ibed in Kabwe et a l . (2002) and Richards (1998). A constant application rate of water (2 L l/week) was appl ied to the min icosms from the beginning of the exper iments. However , each min icosm demonstrated relatively high water re lease rates during the first 70 days of exper iments (Richards, 1998). Effluent rates stabi l ized after 60 days from the beginning of the experiment. The water contents profiles shown in Figures 5.20A and 5.20B represent mean va lues of measured water profiles in the min icosms for the period of 100 days from the start of exper iments. A s expected, the water content increases with increasing depth to the water table. T h e s e water "profiles were used to predict the changes of C 0 2 concentrat ions profiles in high (HT) (21 - 23 °C) and low (LT) (5 °C) temperature min icosms (Kabwe et a l . , 2002; R ichards, 1998). The measured temperature profiles in the HT (18 - 23 °C) and LT (5 °C) min icosms over the first 100 days after filling are presented in Append ix X . Temperatures remained near constant and the standard deviation was < 1.0 °C (Richards, 1998). Chapter V: Analysis and Discussion Page 177 F i g u r e 5.20. Measured volumetric water content profiles in the (A) low temperature (LT) (thermostat set at 5 °C and (B) high temperature (room temperature) minicosms. V represent the water table. Chapter V: Analysis and Discussion Page 178 R e s p o n s e s to short-term fluctuations in room temperature for both LT and HT min icosms were observed only at the 0.02 m depth. For simulation purposes, the average temperatures profiles were used for each min icosm. Figures 5.21 A and 5.21 B show the C 0 2 concentrat ion profiles for the HT and LT min icosms respectively, measured during the first 100-d period from the beginning of the exper iments (Kabwe (2001) and Richards (1998). The C 0 2 concentrat ions increased with depth, reaching the greatest concentrat ions at the capil lary fr inges. During approximately the first 60 days of the experiment, C 0 2 concentrat ions were not yet stable. The general ly higher concentrat ions during the first 60 days were attributed to the disturbance of the soil at the time of excavat ion and min icosms fillings (Lawrence et a l . , 1993; Chappe l le , 1996). Stable concentrat ion profiles were reached after 80 days at 0.75 m depth (Richards, 1998) but after the initial period of stabi l ization, C 0 2 concentrat ions at all posit ions tended to dec rease at a low constant rate. The " C 0 2 " model was used to predict the min icosms concentrat ions profiles due to changes in the water content profiles descr ibed above (Figure 5.20). T o simulate the C 0 2 concentrat ions profiles, the starting C 0 2 concentrat ions profiles and the reference production rate (G 0 ) were required for the min icosms. The initial concentrat ions profiles on day 12 (Figures 5.21 A and 5.21 B) for the LT and HT were used as inputs to predict subsequent concentrat ions profiles due to changes in water contents profiles (Figure 5.20). Figures 5.22A and 5.22B show model simulated C 0 2 profiles within the LT and HT min icosms respectively, for the case where G * was character ized by constraining a = 2, b = 1.25, k = 0.04 °C in Equat ion 5.21, and G 0 = 270 ugC g d" 1 . (Hendry et a l . , 1999). Chapter V: Analysis and Discussion Page 179 0.6 + 0.8 + 0.9 H 1 1 1 h 0 0.2 0.4 0.6 0.8 1 1.2 C 0 2 c o n e . (% vol.) 0.0 0.1 + 0.2 4- 0.3 + 0.4 + 0.5 + 0.6 4- 0.7 0.00 0.05 0.10 0.15 0.20 C 0 2 c o n e . (% vo l ) F i g u r e 5.21. Measured C 0 2 concentration profiles in the (A) high temperature (HT) (21 - 23 °C) and (B) low temperature (LT) (5 °C) minicosms (Richards, 1998; Kabwe, 2001). Chapter V: Analysis and Discussion Page 180 F i g u r e 5.22. Model predicted C 0 2 concentration profiles in the (A) high temperature (HT) (21 - 23 °C) and (B) low temperature (LT) (5 °C) minicosms. Chapter V: Analysis and Discussion Page 181 Compar i son between the measured (Figures 5.21A and 5.22A) and predicted C 0 2 concentrat ion profiles (Figures 5.21 B and 5.22B) shows that the model c losely approximates the measured CO2 concentrat ion profiles in both the LT and HT min icosms, except in the region between 0.2 and 0.4 m depth. The relationship between measured and model prediction is shown in Figure 5.23 for the LT and HT min icosms, respectively. Data for the LT min icosm yield a good correlation ( R 2 =0.98) between measured and model prediction as compare to R 2 = 0.74 for the HT min icosm. In summary, a s imple one-d imensional numerical model for the prediction of changes in the effective diffusion coefficient of C 0 2 and its redistribution in subsur face sand material due to changing water contents was deve loped and val idated using min icosm-measured data for unsaturated sand co lumns. The match between the simulated and the measured concentrat ion profiles for the two min icosms was good. The LT min icosm yielded the best fit (R 2=0.98) between the measured and simulated profiles as compared to R 2 =0.79 for the HT min icosm. It should a lso be noted that the change in CO2 concentrat ion profile in the LT min icosm was smal ler than that in the HT min icosm over the 100-day test period. 5.7 Prediction of C 0 2 Diffusion and Concentration-Depth Profiles in Response to Changes in Water-Depth Profiles in the DSWR The " C 0 2 " model was a lso used to predict CO2 diffusion and concentrat ion-depth profiles in the D S W R pile. The D S W R was selected, for simulat ions because it has a grain s ize similar to that of the sand used in min icosms and m e s o c o s m to verify the D C C method (Kabwe et a l . , 2002). Moreover, the model was a lso constrained to biotic production rate and that the D S W R pile was controlled by the oxidation of organic Chapter V: Analysis and Discussion Page 182 c o c5 o 3 (A 0 . 2 0 0 . 1 5 + 0 . 0 5 + 0 . 0 0 j • ^/• R 2 = 0 . 9 8 i 1 0 . 0 0 0 . 0 5 0 . 1 0 0 . 1 5 S i m u l a t e d C 0 2 c o n c e n t r a t i o n (%) • d l2 • d19 Ad26 + d34 Xd47 • d75 Ad96 0 . 2 0 c o c o c o o 6 o •a 3 (/> CO 0 . 2 0 . 4 0 . 6 S i m u l a t e d C 0 2 c o n c e n t r a t i o n (%) 0 . 8 F i g u r e 5.23. (A) Relationship between measured and simulated C 0 2 concentrations in the low temperature (LT) (5 °C) minicosm plotted on a 1:1 scale. Chapter V: Analysis and Discussion Page 183 carbon present in the lake-bottom sediments (Birkham et a l . , 2003; Lee et a l . , 2003) which are at a constant temperature (0 - 1 °C) and moisture content (25 V o l . %) (Birkham et a l . , 2003) (see also Figure 5.24). Figure 5.24 shows the geologic profile, and the mean CO2 concentrat ion- and water content-depth profi les for D S W R (Birkham et a l . , 2003) . B i rkham et a l . (2003) reported that the trends of C 0 2 - d e p t h profile was stable over t ime, suggest ing near steady-state condit ions with respect to gas concentrat ions, and thus, reaction rates. The CO2 concentrat ion increases with increasing depth up to the organic layer of the pile (Figure 5.24B) and suggested that the dominant sites of reaction occurred below the pile. Be low the organic layer the more vertical C 0 2 concentrat ion-depth profile is observed and that supported the interpretation that the dominant site of production was from the organ ic- r ich material at the base of the pile (Birkham et a l . , 2003). The volumetric water content (Figure 5.24C) va lues general ly ranged from 2 to 3 0 % , with standard deviat ions at each depth general ly less than 2 % , suggest ing near steady-state water condit ions (Birkham et a l . , 2003). Zone of increased water contents (>20%) were measured at the natural ground surface between 18 and 20 m, and near the original ground surface. The deepest z o n e s of elevated water content corresponded to the underlying organic layer (Figure 5.24C). S ince the C 0 2 - and water-depth profiles were stable over t ime (Birkham et a l . , 2003) with smal l variations in measured va lues, it w a s not possib le to simulate or predict subsequent changes in C C V d e p t h profiles assoc ia ted with changes in water content. For illustration purposes, hypothetical drying forcing condit ions were generated in the pile by reducing the initial measured mean water-depth profile (Figure 5.24B and Figure 5.25(A), curve dO) by a factor of 0.1 consecut ively as illustrated in Figure 5.25A. Chapter V: Analysis and Discussion Page 184 (A) Depth 3 0 4 Sand et layer with ice crystal Sand Original ground Organic/sand (B) C02conc. (%) 0 2 4 6 8 1 0 • (C) Water content (%) 0 0 10 20 30 40 5 • 10 • 15 • £. 20 • £ 25H 30 • 35 • 40 - 45 * • t t • • • • • I 0 5 10 15 20 25 H 30 35 40 45 A F i g u r e 5.24. Depth profiles for Deilmann south waste-rock (DSWR) pile (A) Geologic profile (B) mean C 0 2 concentration (Vol.) and (C) mean volumetric Water contents values (Adapted from Birkham et al, 2003). Chapter V: Analysis and Discussion Page 185 B e c a u s e of the unsaturated condit ion in the waste-rock pile, the variat ions in water contents were relatively smal l , except near the original ground and between 18 and 20 m depth. Figure 5.25B shows the model predicted effective diffusion coefficients De-profiles in response to changes in water-depth profiles (hypothetical) (Figure 5.25A) within the D S W R pile. The model ing approach incorporated oxidation reactions limited to the organic-r ich material at the base of the pile (up to 30 m depth) (Birkham et a l . , 2003). The water content-depth profiles va lues were also limited to the organic layer. S ince the production rate of the waste-rock material was not known or determined, an arbitrary number was used instead. Th is may cause some errors in the va lues. The plots in Figure 5.25B show a dec rease in D e -depth profiles with decreas ing water- depth profiles through the pile. T h e s e profiles trends are a lso illustrated in Figure 5.26A for the plots of the D e as a function of water content. A s expected the D e dec reases with increasing water content. The model predicted changes in CO2 concentrat ion-depth profiles are shown in Figure 5.26B. The changes in CO2 concentrat ion-depth profiles in response to changes in hypothetical water-depth profiles were not significant due to the low initial starting water-depth profile va lues in the waste-rock pile. Th is interpretation is supported by the measured standard deviat ions of less than 2 % (Birkham et a l . , 2003), suggest ing near steady-state water condit ions in the pile over t ime. However, the trends showed that as the soil water-depth profiles changes from wet to dry condit ions (Figure 5.25A, curves from right to the left) the CO2 concentrat ion-depth profiles a lso increase proportionally (Figure 5.26B, curves from left to the right). In summary, the model was used to est imate C 0 2 diffusion and concentrat ion- depth profi les in D S W R in response to changes in water-depth profi les. B e c a u s e of the unsaturated condit ion of the waste-rock pile and the near steady-state condit ions with Chapter V: Analysis and Discussion Page 186 F i g u r e 5.25. (A) Hypothetical water-depth profiles in D S W R pile and (B) model predicted effective diffusion coefficients (D e) in response to changes in water contents in Figure 5.32A Curve dO in Figure 5.32A represents the actualmeasured mean water-depth profile in Figure 5.31 C in the D S W R (Birkham et al.,2003). The subsequent profiles (hypothetical water profile) were generated byreducing the initial measured water-depth profile by a factor of 0.1 consecutively. Chapter V: Analysis and Discussion Page 187 |̂  Water content (%, Vol.) | | _ F i g u r e 5.26. Model predicted changes in: (A) effective diffusion coefficient (De) as a function of water content and (B) C 0 2 concentrations depth-profiles in response to changes in water contents profiles described in Figure 5.32A. Chapter V: Analysis and Discussion Page 188 respect to gas concentrations and water content profiles in the DSWR simulations results showed relatively small variations in predicted values. 5.8 Predictions of C 0 2 Diffusion and Surface C 0 2 Flux from the DNWR and DSWR Piles Following Rainfall Events A simplified form of the " C 0 2 " model was also used to predict the effects of rainfall events on the surface effective diffusion coefficient (D e) and surface C 0 2 flux at the DNWR and DSWR during the 6-day test period [August 1 (day 3) to August 6 (day 8) 2002] following 75.9 mm rainfall event over the initial 48-h period [July 30 (day 1) to July 31 (day 2) 2002]. It should be noted that the effects of rainfall events on surface water conditions and C 0 2 fluxes at the DNWR and DSWR were discussed in details in Section 5.2 of Chapter 5 of this thesis. Figures 5.27 and 5.28 show the measured surface water contents, C 0 2 fluxes and rainfall events during the 8-d test period as discussed in Chapter 5, along with the model predicted surface D e and C 0 2 fluxes from the DNWR and DSWR piles, respectively. The model was simulated using the measured surface water contents (curves with broken lines and open marks) for each day of the 6-d (day 3 to day 8) test period. Figure 5.27 shows that as the DNWR ground surface continued to dry gradually from day 3 to day 8, the model predicted surface D e (curve with solid line with triangle marks) also continued to gradually increase with time. It should be noted that the predicted surface D e and measured surface C 0 2 flux exhibit very similar trends. They both increased initially at a fast rate from day 3 to day 5, then at a slow rate and eventually reached a plateau on day 7. At the end of the test period on day 8 the surface D e at the DNWR was found to be 4.25x10"6 m 2 s"1.The similar trends were also Chapter V: Analysis and Discussion Page 1 8 9 Figure 5.27. Rainfall, measured surface water content and C 0 2 flux and predicted effective diffusion coefficient (De) and surface C 0 2 flux at the Deilmann North waste-rock (DSWR) pile over an 8-day test period [30 July (day 1) to 6 August (day 8) 2002] with time. Chapter V: Analysis and Discussion Page 190 4 5 Rainfall Predicted surfacevDe 0 . 5 0 + 0 . 4 5 + 0 . 4 0 + 0 . 3 5 0 . 3 0 j » < P f ~£ 0 . 2 5 | ~ o o 0 . 2 0 | Q ° 5 + 0 . 1 5 + 0 . 1 0 0 . 0 5 0 . 0 0 I Rain •Water" C02flux De — - C02 flux (Pred) - - -Po ly . (De) F i g u r e 5.28. Rainfall, measured surface water content and C 0 2 flux and predicted effective diffusion coefficient (De) and surface C 0 2 flux at the Deil;mann North waste-rock (DNWR) pile over an 8-day test period [30 July (day 1) to 6 August (day 8) 2002] with time. Chapter V: Analysis and Discussion Page 191 observed at the D S W R in Figure 5.28. Similarly, both the measured surface C 0 2 flux and predicted surface D e initially increased at a fast rate from day 3 to day 5 and eventually reached a plateau on day 7. At the end of the test period the surface D e at the D S W R was found to be 4.40 x 10" 6 m 2 s" 1 . To predict the surface C 0 2 flux using F ick 's 1st and 2 n d law (e.g., Equat ion 5.3 used in the model) the concentrat ion gradient (e.g., dC /dz ) must be known. The concentrat ion gradients, however, were not measured during the test period. But based on the measured average surface C 0 2 flux and the corresponding model predicted sur face D e from the D N W R and D S W R piles, the concentrat ion gradients can be est imated using F ick 's first law (e.g., Equat ion 5.3). The concentrat ion gradients were found to be: DC - f w" A ^ = 2 . 0 3 x 1 0 - 2 U ^ - and ^ = 1 .24x10- 2 dz l m 3 * m j 3z kg V m 3 * r r v for the D N W R and D S W R pi lesrespectively. It w a s a s s u m e d in the model simulat ions that the C 0 2 w a s produced at a s teady rate below the piles (e.g., the dominant sites of reactions) (Birkham et a l . , 2003) and that the shal low C 0 2 gradient near the ground surface would remain constant during the relatively short-term wetting event. It should be pointed out that this assumpt ion is not completely val id; however actual C 0 2 measurements in the sand profiles immediately below the ground sur faces of the waste-rock piles were not obtained and thus it is not possib le to accurately constrain the lower boundary of the transient model . B a s e d on this simplifying assumpt ion, the calculated surface C 0 2 f luxes during the 6-d test period at the D N W R and D S W R are a lso shown respectively in F igures 5.27 and 5.28 (curves with solid l ines with circle marks) along with the predicted surface D e s . Resul ts showed that the calculated (predicted) and measured Chapter V: Analysis and Discussion Page 192 C 0 2 f luxes exhibited very similar trends. T h e s e trends supported the interpretation that the flux is proportional to the D e t imes the concentrat ion gradient. 5.9 Chapter Summary In summary, results showed that the water content at ground surface is transient after a heavy rainfall and is an important factor in controlling CO2 f luxes. Both the E C measured A E and P E (Carey et a l . , 2005) and So i lCover predicted A E and P E va lues showed good agreement. The " C 0 2 " model predicted, as expected, a dec rease in D e -depth profiles and increase in the C 0 2 concentrat ion-depth profiles with decreas ing water-depth profiles through the D S W R pile. The model a lso predicted surface CO2 f luxes trends that were very similar to the measured surface CO2 f luxes during the 6-d test period from the D N W R and D S W R following heavy rainfall events. Chapter VI: Summary and Conclusions Page 193 CHAPTER VI Summary and Conclusions A recently deve loped and laboratory-verif ied dynamic c losed chamber ( D C C ) method has been tested under field condit ions on waste-rock piles at the Key Lake uranium mine. The method has been used to quantify the magnitude of spatial and temporal variations in the C 0 2 flux on the Dei lmann north ( D N W R ) and Dei lmann south ( D S W R ) waste-rock piles over a period of two years (summer 2000 - summer 2002). The ability of the D C C to accurately quantify field respiration was demonstrated by compar ing the D C C f luxes to those obtained using two other field C 0 2 flux measurement techniques: the static c losed chamber ( S C C ) and eddy covar iance (EC) methods. The main advantage of this direct technique is that it provides an almost instantaneous indication of the reaction rate under field condit ions, regardless of cl imatic or moisture condit ions in the waste dumps. The D C C was a lso used to investigate the effects of cl imatic var iables (e.g., rainfall and evaporation) on near-surface waste-rock-water condit ions which also affect surface C 0 2 gas f luxes. A relatively s imple " C 0 2 " model was developed to predict the changes in the effective diffusion coefficient of C 0 2 , sur face C 0 2 flux and its redistribution in subsur face material in response to changes in soil water contents. At the D S W R site the D C C was used to demonstrate that the C 0 2 flux was relatively uniform across the pile (with a C V of only about 30%). Th is C V reflects the combined influence of a relatively constant rate of C 0 2 production in the organic-r ich Chapter VI: Summary and Conclusions Page 194 zone at the base of the waste-rock pile and the textural uniformity of the overburden material (sand) used to construct the D S W R pile (Birkham, 2002). That is, these factors combine to exert a controlling influence on the composi t ion and upward migration of pore g a s e s and , in turn, the flux of gases from the surface to the atmosphere. Compar i son between the D C C and the static c losed chamber ( S C C ) showed that there was no significant difference (p < 0.05) between the mean C 0 2 f luxes obtained using the two methods at the D S W R . W h e r e a s the chamber -based ( D C C and S C C ) methods yielded comparab le data from the D S W R , with an overal l t ime-averaged C 0 2 flux of 171 + 54 mg C 0 2 m" 2 h"1; the eddy covar iance (EC) method yielded a t ime- averaged C 0 2 flux (150 + 35 mg C 0 2 m" 2 h"1) that was about 1 2 % lower than that calculated from the chamber data. Underest imat ion of the C 0 2 flux assoc ia ted with soil respiration by E C - b a s e d methods relative to chamber -based methods has been reported widely in the literature [e.g., Gou lden et a l . , 1996; Norman et a l . , 1997; Law et a l . , 1999; J a n s s e n s et a l . , 2000; Dav idson et a l . , 2002]. Though not excess ive ly large, these dif ferences presumably reflect the different p rocesses measured by the two methods. The chamber data exhibited slightly greater standard deviat ions than the E C data (i.e., D C C = + 58 mg C 0 2 m" 2 h"1; S C C = + 59 mg C 0 2 m" 2 h"1; E C = + 32 mg C 0 2 m" 2 h~1). It is bel ieved that this likely reflects the fact that the variability assoc ia ted with the chamber -based measurements includes both a spatial and temporal component , whereas the variability assoc ia ted with the E C method is primarily temporal in nature. The overall averages of C 0 2 f luxes at the D N W R and D S W R measured with the D C C over the 2-year test period (summer 2000 and summer 2002) were 188 + 68 and 217 + 83 m" 2 h" 1, respectively. Chapter VI: Summary and Conclusions Page 195 B a s e d on these results, it was conc luded that the D C C is well-suited to the quantification and spatial resolution of C 0 2 f luxes assoc ia ted with waste-rock pi les. Th is work showed that the effects of heavy rainfall events on the C 0 2 flux and near-surface water condit ions were of short duration. The short-term effects of rainfall events were reflected in the lack of long-term spatial and temporal variations in C 0 2 f luxes (average C V is 28%-39%) at both sites over a 2-year test period (summer 2000 and summer 2002). B e c a u s e of lack of temporal and spatial variation in C 0 2 f luxes, it is conc luded that rainfall events had little long-term effects on C 0 2 flux from waste-rock piles. During the test period, So i lCove r was used to predict the rate of evaporat ion on the D S W R and results were compared to publ ished f ie ld-measured evaporat ion using eddy covar iance (EC) method on the D S W R (Carey et a l . , 2005). Resul ts showed very good agreement between the model predicted and E C measured va lues. Both the field- measured and predicted data indicated an average evaporat ion rate of approximately 1.1 mm per day at the D S W R for the 8-day test period. Verif ication of the " C 0 2 " model developed showed good agreement between predicted and sand co lumn-measured data. Simulat ions results for the deep profile in D S W R showed relatively smal l variations in predicted C 0 2 concentrat ion-depth profiles assoc ia ted with change in water content during a simulated drying event. A simplif ied model was a lso used to predict surface C 0 2 f luxes on the D N W R and D S W R at the Key Lake mine following rainfall events. Resul ts showed the model predicted surface D e and C 0 2 f luxes that exhibited very similar trends with measured data. In summary, the C 0 2 model , a long with others capab le of predicting changes in water content profiles with time such as So i lCover , can be of value in the prediction and monitoring of Chapter VI: Summary and Conclusions Page 196 b iogeochemical p rocesses occurr ing in the unsaturated geologic material and natural soi ls. Finally, the field results showed that the D C C method is especia l ly useful for character iz ing spatial variability as well as identifying zones of sulphide oxidation and carbonate buffering in the waste-rock pi les. The method has distinct advantages over the traditional methods in terms of accuracy, speed , and repeatability and it can be used to measure the CO2 f luxes in situ at the s a m e locations using the s a m e chambers with minimal disturbance of the soi l . The method can be extended to any other mine waste dumps to quantify b iogeochemical reaction rates in unsaturated geologic media and soi ls e lsewhere in C a n a d a . In conc lus ion, it is bel ieved the objectives of this thesis work were met as stated in the introduction. References Page 197 References Aachib, M., Aubertin, M., Mbonimpa, M. 2002. Laboratory measurements and predictive equations for gas diffusion coefficient of unsaturated soils. Proceedings of the 55 t h Canadian Geotechnical and Joint IAH-CNC and C G S Groundwater Speciality. Niagara Falls Ont.. Canadian Geotechnical Society, Ottawa: 156-171. Aachib, M. Mbonimpa., Aubertin, M. 2004. Measurement and Prediction of Oxygen Coefficient in Unsaturated Media, with Application to Soil Covers. Water, Air and Soil Pollution 156: 163-193. Affek.-H., Ronen, D., and Yakir, D. 1998. Production of C 0 2 in the capillary fringe of a deep phreatic aquifer. Water Resour. Res. 34 (5), 989-996. Ahuja, L.R., Ross, J.D., Bruce, R.R., and Cassel , D.K. 1988. Determining unsaturated Hydraulic conductivity from tensiometric data alone. Soil Sci . Soc. Am. J . 52: 27- 34. Ambus, P. and Robertson, G.P. 1998. Automated near-continuous measurement of carbon dioxide and nitrous oxide fluxes from soil. Soil Sci . Soc. Am. J . 62: 394- 400. Amundson, R.G., Chadwick, O.A., Sowers, J .M. , and Doner, H.E. 1988. Relationship between climate and vegetation and the satble carbon isotope chemistry of soils in the eastern Mojave desert. Nevada. Quat. Res. 29: 245-254. Atkinson, T.C. 1977. Carbon dioxide in the atmosphere of the unsaturated zone: an important control of groundwater hardness in limestones. J . Hydrol. 35, 111-123. Aubertin, M., Ricard, J.-F. , and Chapuis, R.P. 1998. A predictive model for the water retention curve: application to tailings from hard-rock mines. Can. Geotech. J . 35: 55-69. References Page 198 Aubert in, M. , Aach ib , M. , and Authier, K. 2000. Evaluat ion of diffusive flux through covers with a G C L . Geotext i les and G e o m e m b r a n e s 18:215-233. Aubert in, M. , Mbon impa, M., and Chapu is , R . P . 2003. A model to predict the water retention curve from basic geotechnical properties. C a n . Geo tech . J . 40: 1104- 1122. Aubert in, M. , Mo lson , J . , Buss ie re , B., and Dagena is , A . M . 2006. Investigations of layered cover acting as oxygen barriers to limit acid mine drainage. Proc . 5 t h I C E G Environmental Geotechn ics : Cha l lenges and Responsibi l i t ies for Environmental Geotechn ics , T h o m a s H.R. (ed.) 26-30 June 2006 Thomas Tel ford, vol . 2 , pp 827-835. Ba ldocch i , D.D., H icks, B .B. , and Meyers , T . P . 1988. Measur ing biosphere atmosphere exchanges of biologically related gases with micrometeorological methods. Eco l . 69: 1331-1340. Barbour, S .L . 1998. Nineteenth Canad ian Geotechn ica l Co l loqu ium: The soil-water characterist ic curve: a historical perspect ive. C a n . Geo tech . J . 35(5): 873-894. Barr, A . G . , Griffis, T . J . , B lack, T.A. , Lee , X . , Staebler , R . M . , Fuentes, J .D . , C h e n , Z . , and Morgenstern, K. 2002. Compar ing the carbon budgets of boreal and temperate dec iduous forest stands. C a n . J . For. R e s . 32: 813-822. Bel l , A . V . , Mil ls, A .L . , and Herman, J . S . 1991. A n update of the acid waste rock field trials at heath Stee le Mine, New Brunswick. S e c o n d International Conference on the Abatement of Ac id ic drainage: 321-340. Bennett, J . W . and Ritchie, A . I .M. 1990. "Measurements of the Transport of Oxygen into Two Rehabil i tated W a s y e - R o c k dumps". In: Proceed ings of the S e c o n d International Conference on the Abatement of Ac id ic Drainage. M E N D 3 pp 289- 298 September 16-18, 1990, Montreal , C a n a d a . Refe rences P a g e 199 Bennett, J.W. and Ritchie, A.I.M. 1992. The installation of monitoring sites in waste rock at the Aitik Mine site August-September (1991). Austral. Nuclear Sci. Technology Org. ReportANSTO/C263. Bennett, J . W., Comarmond, M. J. , and Jeffery, J. J . 2000. Comparison of sulfidic oxidation rates measured in the laboratory and the field. International Conference on Acid Rock Drainage (ICARD) 2000 Proceedings, Denver, Colorado. The Society for Mining, Metallurgy, and Exploration, Inc., 171-180. Black, T.A., Chen, W.J., Barr, A.G., Arain, M.A., Chen, Z., Nesic, Z., Hogg, E.H, Neumann, H.H., and Yang, P.C. 2000. Increased carbon sequestration by a boreal deciduous forest in years with warm spring. Geophys. Res. Lett. 27: 1271- 1274. Blowes, D.W., Al, T., Lortie, L., Gould., W.D., and Jambor, J.L. 1995. Microbial, chemical, and mineralogical characterization of the Kidd Creek mine tailings impoundment, Timmins area, Ontario. Geomicrobiol. J . 13: 13-31. Bigham, J.M., Schwertmann, U., Carson, L., and Murad, E. 1990. A poorly crystallized oxyhydroxysulfate of iron framed by bacterial oxidation of Fe(ll) in acid mine waters. Geochim. Cosmochim. Acta 54: 2743-2758. Birkham, T.K., M.Sc. Dissertation, University of Saskatchewan 2002. Birkham, T.K., Hendry, M.J., Wassenaar, L.I., Mendoza, C.A., and Lee, E.S. 2003. Characterizing Geochemical Reactions in Unsaturated Mine Waste-Rock Piles Using Gaseous 0 2 , C 0 2 , 1 2 C 0 2 , and 1 3 C 0 2 . Environ. Sci. Technol. 37: 496-501. Brooks, R.H., and Corey, A.T. 1964. Hydraulic properties of porous media. Hydrology Paper, No. 3, Colorado State University, Fort Collins, Colorado. Brierley, C.L. 1978. Bacterial leaching. CRC Critical Rev. Microbiol. 6: 207-262. Brumme, R., and Beese, F. 1992. Effects of liming and nitrogen-fertilization on References Page 200 emiss ions of C 0 2 and N 2 0 from a temporate forest. J . Geophys . R e s . 97(D12): 12851-12858. Buchmann and Ernst-Detlef Schu lze , 1999. Net C 0 2 and H 2 0 f luxes of terrestrial ecosys tems. Globa l B iogeochemica l Cyc les , 3, 743-750. Buchmann , N. 2000. Biotic and Abiot ic factors controll ing soil respiration rates in P i cea ab ies stands. Soi l Bio l . B iochem. 32: 1625-1635. Burkins, M.B. , Virgin ia, R.A. and Wa l l , D.H. 2001 . Organic carbon cycl ing in Taylor Val ley , Antarct ic: quantifying soil reservoirs and soil respiration. G loba l C h a n g e Biology 7: 113-125. Burman, R., and Pochop , L .O. 1994. Evaporat ion, evapotranspirat ion, and cl imatic data (Developments in Atmospher ic Sc i ence 22). E lsev ier Sc i ence , Amste rdam, The Nether lands. Buss ie re , B., Dagena is , A . , Mbon impa, M. , and Aubert in, M. 2002. Modif ication of oxygen-consumpt ion testing for the evaluation of oxygen barrier performance. Proceed ings of the 5 5 t h Canad ian Geochemica l and 3 r d Joint C G S Groundwater Spec ia l Conference, Niagara Fal ls , Ontario, pp. 139-146. Buss ie re , B ., Aubert in, M. , and Chapu is , R . P . 2003 . The behavior of inclined covers used as oxygen barrier. C a n . Geo tech . J . 40 : 512-535. Buyanovsky , G . A . , Wagner , G . H., and Gantzer , C . J . 1986. Soi l respiration in a winter wheat ecosys tem. Soi l S c i . S o c i . A m . J . 50: 338-344. Capehar t , W . J . , and Car l son , T . N . 1994. Estimating near-surface soi l-moisture availability using a meteorological ly driven soil water profile model . J . Hydrol. 160(1-4): 1-20. Capehar t , W . J . , and Car l son , T . N . 1997. Decoupl ing of surface and near-surface soil water content: A remote sens ing perspect ive. Wate r Resour . R e s . 33(6): 1383- Refe rences P a g e 201 1395. Carey , S .K . , Barbour, S .L . , and Hendry, M . J . 2005. Evaporat ion from a waste-rock surface, Key Lake , Saska tchewan . C a n . Geo tech . J . 42 : 1189-1199. Ced ing , T . E . , So lomon , D.K., Q u a d e , J . , and Bowman , J . R . 1991. O n the isotopic composi t ion of carbon in soil carbon dioxide. G e o c h i m . et C o s m o c h i m . Ac ta 55: 3404-3405. Clark, I. and Fritz, P. 1997. Environmantal Isotopes in Hydrology. Lewis Publ ishers . Col l in , M. and R a s m u s o n , A . 1988. G a s Diffusivity Mode ls for Unsaturated Porous Med ia . In Soi l Sc ience Amer i ca Journal . 52: 311-333. C o n e n , F., and Smith, K.A. 2000. A n explanat ion of l inear increases in gas concentrat ion under c losed chambers used to measure gas exchange between soil and atmosphere. Eur. J . Soi l S c i . 51 : 111-117. Cook , F . J . , Dobos , S .K . , Car l in , G .D . , and Millar, G . E 2004. Oxidat ion of pyrite in acid sulfate soi ls: in situ measurements and model ing. Austral ian Journal of Soi l Resea rch 42 : 499-507. Cra ig , H. 1953. The geochemistry of the stable carbon isotopes. Cosmoch im ica et a l . 3, 53-92. Dav idson, E.A. , and Trumbore, S . E . 1995. G a s diffusivity and production of C 0 2 in deep soi ls of the eastern A m a z o n . Tel lus Ser . B 47 : 550-565. Dav idson, E.A. , Belk, E. , Boone , R .D. 1998. Soi l water content and temperature as independent or confounded factors controlling soil respiration in a temperate mixed hardwood forest. G loba l C h a n g e Biol . 4, 217-228. Dav idson, E.A. , S a v a g e , K., Bo ls tad, P. , Clark, D.A., Curt is, P . S . , El lsworth, D .S . .Hanson , P . J . , Law, B .E . , Luo, Y . , Pregitzer, K .S . , Rando lph , J . C . , and Zak, D. 2002. Belowground carbon al location in forests est imated from litterfall and References Page 202 I R G A - b a s e d soil respiration measurements . A g . For. Met. 113: 39-51. Dav idson, J . M . , S tone, L.R., N ie lsen, D.R., and Larue, M .E . 1969. Field measurement and use of soil-water properties. Water Resour . R e s . 5: 1312-1321. Davis , G . B . , and Ritchie, A . I .M. 1987. A model of oxidation in pyretic mine wastes: Part 3. Importance of particle s ize distribution. App l . Math. Mode l . 11 (December) : 417-422. Dawson , R., and Morgenstern 1995. Liquifaction F lowsl ides in Rocky Mountain and Mine W a s t e Dumps , P h a s e 3, Final Report. Natural Resou rces C a n a d a , Ottawa, Ontario. Contract Report 23440-9135/01. de J o n g , E. , and Schapper t , H. J . V . 1972. Calculat ion of soil respiration and activity from C 0 2 profiles in the soi l . Soi l S c i . 113(5): 328-333. Delleur, J . W . 1999. The Handbook of Groundwater Engineer ing. C R C P r e s s , New York. Drever, J.I. 1997. The Geochemis t ry of Natural Waters , 3 r d ed . Prent ice Hal l . Upper Sadd le River, N J . 436 pp. Ek, M. , and C u e n c a , R .H . 1994. Variat ion in soil parameters: Implications for model ing sur face f luxes and atmospher ic boundary layer development. Boundary Layer Meteorol . 70(4): 369-383. Elber l ing, B., N icho lson, R.V. and David, D .J . 1993. Field evaluation of sulphide oxidation rates. Nordic Hydrology 24: 323-338. Elberl ing, B., N icho lson, R.V. , Reardon , E . J . , and Tibbie, P. 1994. Evaluat ion of sulfide oxidation rates: laboratory study compar ing oxygen f luxes and rates of oxidation product re lease. C a n . Geo tech . J . 31 : 375-383. Elberl ing, B., and Nicho lson, R.V. 1996. Field determination of sulfide oxidation rates in mine tail ings. Water Resour . R e s . 32: 1773-1784. Elber l ing, B., and J a k o b s e n , B .H . 2000. Soi l solution pH measurements of using in-line References Page 203 chambers with tension lysimeters. Canad ian J . Soi l S c i . 80: 283-288. Elber l ing, B. 2003. Seasona l trends of soil CO2 dynamics in a soil subject to freezing. J . Hydrol . 276: 159-175. Elber l ing, B. 2005. Temperature and oxygen control on pyrite oxidation in f rozen mine tail ings. Co ld Reg ions Sc ience and Techno logy 4 1 : 121-133. Enoch , C . W . and Dasberg 1971. The occurrence of high C 0 2 concentrat ions in soil air. G e o d e r m a 6: 17-21. Fahnestock, J .T. , Jones , M.H. , Brooks, P D . , Walker , D.A., Welker , J . M . 1998. Winter and early spring C 0 2 efflux from temdra communit ies of northern A l a s k a . J . Geophy . . R e s . 103: 29023-29027. Fa la , O. , Mo lson , J .W. , Aubert in, M. , Buss ie re , B. 2005.Numer ica l Model ing of F low and Capi l lary Barrier Effects in Unsaturated W a s t e Rock P i les . Mine Water and Environment 24(4): 172-185. Farrel l , R . E . , de J o n g , E. , and Elliott, J .A . 2002. G a s Sampl ing and Ana lys is . P p . 1075- 1111. In: J . H . Dane and G . C . Topp (eds.) Methods of Soi l Ana lys is . Part 4. Phys ica l Methods. S S S A Book Ser ies , no. 5. Soi l S c i . S o c i . of A m . J . , Mad ison , W l . Frank, A . B . , L iebig, M.A., and Hanson , J .D . 2002. Soi l carbon dioxide f luxes in northern semiar id grass lands. Soi l B io l . B iochem. 34: 1235-1241. F reeze , R.A. 1969. The mechan ism of natural ground-water recharge and d ischarge 1. One-d imens iona l , vertical, unsteady, unsaturated flow above a recharging ground-water flow system. Water Resour . R e s . 153-171. F reeze , R.A, . and Cherry, J .A . 1979. Groundwater. Prent ice-Hal l , Inc. Fredlund, D .G . , and Rahardjo, H. 1993. Soi l mechan ics for unsaturated soi ls. John Wi ley & S o n s , Inc., New York. References Page 204 Fredlund, D .G . , and X ing , A . 1994. Equat ions for the soi l-water characterist ic curve. C a n . Geo tech . J . , vol . 31 , pp. 521-532. Fuller, E. N., Schett ler, P. D., and Gidd ings, C . J . 1966. A new method for prediction of binary gas -phase diffusion coefficients. Industrial and engineering chemistry 58(5), 19-27. Gardner , W . R . 1956. Calculat ion of capil lary conductivity from pressure plate outflow data. Soi l S c i . S o c . A m . Proc . 20: 317-320. Ge l inas , P. , Lefebvre R., and Choquet te M. 1992. Monitoring of acid mine drainage in a waste rock dump. In Environmental Issues and W a s t e Management in Energy and Minerals Product ion (ed. S . e. al.), pp. 747-756. Ba l kema. Gou lden , M.L., Munger, J .W. , F a n , S . M . , Daube, B .C . , and Wofsy , S . C . 1996. Measurements of carbon sequestrat ion by long-term eddy covar iance: methods and a critical evaluation of accuracy. G lob . C h a n g e Biol . 2: 169-182. Grab le , A . R . 1966. Soi l aeration and plant growth, Adv . Agron . 18: 57-106. Granger , R . J . 1989. A complementary relationship approach for evaporat ion from nonsaturated sur faces. J . Hydrol , 111: 3 1 - 3 8 . Gray , D .M. 1995. Handbook of Hydrology, University of Saska tchewan , Saska toon , S K , C a n a d a . Haddadin , J . , Mor in, D., Ollivier, P. & Fick, M. 1993. Effect of different carbon dioxide concentrat ion on ferrous iron and pyrite oxidation by a mixed culture of iron and/or sulfur-oxidizing bacteria. E n z y m e Microbiol . Techno logy 15, 832-841. H a s s , H., F isher, D.W., Thors tenson, D.C. , and W e e k s , E . P . 1983. 1 3 C 0 2 and 1 4 C 0 2 measurements on soil a tmosphere sampled in the subsoi l unsaturated zone in the westerm Gread Pla ins of the U.S . Rad iocarbon 25: 301-314. References Page 205 Harr ies, J . R. and Ritchie, A . I .M. 1981. "The use of temperature profiles to est imate the pyretic oxidation rate in a waste-rock dump from an opencut mine". Water , Ai r and Soi l Pollution 15: 405-423. Harr ies, J . R. and Ritchie, A . I .M. 1983. The microenvironmnt within waste rock dumps undergoing pyritic oxidation. In recent Progress in Biohydrometal lurgy (G. Ross i & A . E . To rma, eds.) A s s o c . Minerer ia Sa rda , Iglesias, Italy, 377-392 Harr ies, J . R. and Ritchie A. I .M. 1985. Pore gas composi t ion in waste rock dumps undergoing pyritic oxidation. Soi l S c i . 140(2), 143-152. Hanson , P . J . , Wul lschleger , S . D . , Boh lman, S .A . , and Todd , D.E. , 1993. Seasona l and topographic patterns of forest floor CO2 efflux from an upland oak forest. Tree Physio logy, 13, 1-15. Hatanaka, M. , Uch ida , A . , and Takehara , N., 1997. Permebil i ty characterist ics of high- quality undisturbed sands measured in a triaxial cel l . So i ls and Foundat ion, J a p a n e s e Society of Soi ls Mechan ics and Foundat ion Engineer ing 37(3): 129- 135. He lgen, S . , Davis , A . , and Byrns, C . 2000. Measurement of oxygen , temperature, and geochemica l profiles in sulf ide and oxide waste rock dumps of different ages . International Conference on Ac id Rock Drainage ( ICARD) 2000 Proceed ings , 269-275. Herasymuik, G . M , Wi l son , G .W. , Barbour, S .L . , and Smi th , T., 1995. The character izat ion of hydrologic properties and moisture migration pathways of a waste rock dump. Proceed ings of the Nineteenth Annua l British Co lumb ia Mine Rec lamat ion Sympos ium. Hendry, M .J . , Lawrence, J . R . , Zanyk, B .N . , and Kirk land, R. 1993. Mircobial production of CO2 in unsaturated geologic media in a mesosca le model . Water Resour . R e s . References Page 206 29: 973-984. Hendry, M .J . , M e n d o z a , C .A . , Kirk land, R.A. , and Lawrence, J . R . 1999. Quantif ication of transient C 0 2 production in a sandy unsaturated zone . Wate r Resour . R e s . 35 (7): 2189-2198. Hendry, M .J . , M e n d o z a , C .A . , Kirk land, R., Lawrence, J . R . 2001 . A n assessmen t of a m e s o c o s m approach to the study of microbial respiration in a sandy unsaturated zone . Groundwater 39 (3): 391-400. Hendry, M .J . , W a s s e n a a r , L.I., and Bi rkham, T .K. 2002. Microbial respiration and diffusive transport of 0 2 , 1 6 0 2 , and 1 8 0 2 , in unsaturated soi ls: A mesocosm experiment. G e o c h i m . et C o s m o c h i m . Ac ta , V o l . 66, No. 19, pp. 3367-3374. Hil lel, D. 1980. Appl icat ions of Soi l Phys ics , A c a d e m i c P r e s s , Inc., New York. Hobbie, S . E . , Sch ime l , J . P . , Trumbore, S . E . , Randerson , J . R . 2000. Controls over carbon storage and turnover in high-latitude soi ls. G loba l C h a n g e Biology 6: 196- 210. Hockley, D., Smolensky , J . , J a h n , S . , and Pau l , M. 2000. Geochemica l investigations and gas monitoring of an acid generating waste rock pile. International Confe rence on Ac id Rock Drainage ( ICARD) 2000 Proceed ings , 181-189. Hol land, E.A. , Townsend , A . R . , Vi tousek, P . M . , 1995. Variabil i ty in temperature regulation of C 0 2 f luxes and N mineralization from five Hawai ian soi ls: Implications for a changing cl imate. Globa l C h a n g e Bio l . , 1, 115-123. Holl ings, P. , Hendry, M .J . , and Kirk land, R.A. 2000. Quantif ication of oxygen consumpt ion rates for gneiss ic waste rock pi les, K e y Lake uranium mine, northern Saska tchewan , C a n a d a : application of the kinetic cell technique. International Conference on Ac id Rock Drainage ( ICARD) 2000 Proceed ings , 163-169. References Page 207 Hutchinson, G.L . , Livingston, G . P . , Healy, R.W. , and Str iegl, R . G . 2000. C h a m b e r measurement of sur face-atmosphere trace gas exchange: numerical evaluation of dependence on soi l , interfacial layer, and source/s ink properties. J . Geophys . R e s . 105: 8865-8875. J a n s s e n s , I.A., Kowalsk i , A . S . , Longdoz, B., and Ceu lemans , R. 2000. A s s e s s i n g forest soil C 0 2 efflux: an in situ compar ison of four techniques. Tree Phys . 20: 23-32. Jaynes , D.B. , and Rogowsk i , A . S . 1983a. Applicabil i ty of F ick 's law to gas diffusion. Soi l S c i . S o c . A m . J . 47: 425-430. Jaynes , D.B. , Rogowsk i , A . S . , P ionke H.B., and Jacoby , Jr. E.L. 1983b. Atmosphere and temperature changes within a reclaimed coal strip mine. Soi l S c i . 136(3), 164-177. Kabwe , K.L. M . S c , Dissertat ion, University of Saska tchewan , 2001 . Kabwe , K.L. , Hendry, M .J . , W i l son , G . W . , and Lawrence, J . R . 2002. Quantifying C 0 2 f luxes from soil sur faces to the atmosphere. J . Hydrol . 260: 1-14. K e e n , B.A. 1931. The Phys ica l Propert ies of So i ls , Longman , London. Kei th, C . N . , and V a u g h a n , D.J . 2000. Mechan i sms and rates of sulphide oxidation in relation to the problems of acid rock (mine) dra inage. In Environmental Mineralogy: Microbial Interactions, Anthropogenic solution and the inf luences, Contaminated Land and W a s t e Management . Mineralogical Soc iety Spec ia l Publ icat ion. Kel ler, K. C . and Bacon D. H. 1998. Soi l respiration and georespirat ion dist inguished by transport ana lyses of vadose C 0 2 , 1 3 C 0 2 , and 1 4 C 0 2 . G lob . B iogeochem. Cyc les 12(2), 361-372. Key Lake Mining Corporat ion 1979. Environmental Impact Statement for the Key Lake Project. References Page 208 Klute, A . , and Letey, J . 1958. The dependence of ionic diffusion on the moisture content of non- adsorbing porous media. Soi l S c i . S o c i . A m . Proceed ings , 22: 213-215. Klute, A . , and Di rksen, C , 1986. "Hydraul ic Conductivi ty and Diffusivity-Laboratory Methods," in A . Klute, ed . , Methods of Soi l Ana lys is , Part l=Physical and Mineralogical Methods, Amer ican Society of Agronomy Monograph 9, 2d ed . , pp. 687-734. Kovacs , G . 1981. S e e p a g e Hydraul ics. E lsev ier Sc ience Publ ishers , Ams te rdam. Law, B . E . , Ba ldocch i , D.D., and Anthoni , P . M . 1999. Be lowcanopy and soil C 0 2 f luxes in a ponderosa pine forest. Agr ic . For. Meteorol . 94: 171-188. Lawrence, J . R . , Zanyk, B .N . , Wolfaardt, G . M . , Hendry, M .J . , Robarts , R.D. , and Caldwel l , D .E . 1993. Design and evaluation of a meso sca le model vadoze zone and ground water sys tem. Ground Water 31 , 446-455 . Lawrence, J . R . , and Hendry, M .J . 1995. M e s o c o s m s for subsur face research. Water Qual i ty Resea rch J . C a n a d a 30, no. 3:493-512. Lee , E . S . , B i rkham, T.K. , W a s s e n a a r , L.I., and Hendry, M . J . 2003a . Microbial respiration and diffusive transport of O 2 , 1 6 0 2 , and 1 8 0 2 in unsaturated soi ls and geologic sediments. Environ. S c i . and Techno l . 37: 2913-2919. Lee , E . S . , Hol l ings, P. , and Hendry, M . J . 2003b. U s e of O2 consumpt ion and C O 2 production in kinetic cel ls to del ineate pyrite oxidat ion-carbonate buffering and microbial respiration in unsaturated media. J . Cont . Hydro. 65: 203-217. Lee , R. W . 1997. Effects of carbon dioxide variations in the unsaturated zone on water chemistry in a glacial -outwash aquifer. Appl ied Geochemis t ry 12, 347-366. Lefebvre, R., Hockley, D., Smo lensky , J . , and Lamontagne, A . 2001 . Mult iphase Transfer p rocesses in waste rock piles producing acid mine drainage 2: appl icat ions of numerical simulat ion. J . Cont. Hydro. 52: 165 -186 . References Page 209 Linn, D .M. , and Doran, J . W . 1984. Effect of water-fil led pore space on carbon dioxide and nitrous oxide production in tilled and non tilled soi ls. Soi l S c i . S o c . A m . J . 48 , 1267-1272. Lorenz, W . C . and Tarpley, E . C . 1963. Oxidat ion of coal mine pyrite. U S Bureau of M ines Report, Inv. 6247. Pit tsburgh, P A . Looney, B .B. , and Fal ta, R .W. 2000. V a d o z e Zone . Sc ience and Techno logy Solut ions, Vo l s . I and II, Battelle P ress , Co lumbus , Ohio. Lowson , R.T. 1982. A q u e o u s oxidation of pyrite by molecular oxygen. Chemica l Rev iews , Amer ican Chemica l Society, 82(5): 462-496. Loftfield, N .S. , Brumme, R., and B e e s e , F. 1992. Automated monitoring of nitrous oxide and carbon dioxide flux from forest soi ls. Soi l S c i . S o c . A m . J . 56: 1147-1150. Mart in, V . , Aubert in, M. , Buiss iere, B., Mbon impa, M. , Dagena is , A . M . , and Gosse l i n , M. 2006. Proc . Of the 7 t h International Conference on Ac id Rock Drainage ( ICARD) , March 26-30, 2006, St Louis M O . R.I. Barnhisel (ed.) Pub l ished by the Amer ican Society of Mining and Reclamat ion ( A S M R ) , 3134 Montavesta R o a d , Lexington, K Y 40502. Marshal l , T . J . , Ho lmes, J .W. , and R o s e , C . W . 1996. Soi l phys ics. 3 r d ed . Cambr idge University P ress , Cambr idge , U.K. Matson , P.A. , and Harris, R . C . 1995. B iogenic T race G a s e s : Measur ing Emiss ions from Soi l and Water . Blackwel l Scienti f ic. Mbon impa, M. , 1998. Injizierfahigkeit von Feinstb indemit te lsuspensionen zurAbdichung von Lockergeste inen, Mittei lungen des Instituts von Grundbau , Bodenmechan ik and Energ ieasserbau ( IGBE), Universitat. Mbon impa , M, Aubert in, M. , Aach ib , M. , and Buss ie re , B. 2002. Interpretation of field tests to determine oxygen diffusion and reaction rate coefficients of tail ings and References Page 210 soil covers . Proceed ings of the 5 5 t h Canad ian Geotechn ica l Conference - 3 r d Joint International Assoc ia t ion of Hydrogeologists - Canad ian National C h a p t e r - Canad ian Geotechn ica l Soc iety ( I A H - C N C / C G S ) Conference: Ground and Water : Theory to Pract ice, N iagara Fal ls , pp. 147-154. C D - R O M . Mbon impa, M, Aubert in, M. , Aach ib , M. , and Buss ie re , B. 2003. Diffusion and consumpt ion of oxygen in unsaturated cover materials. C a n . Geo tech . J . 40 : 916-932. M E N D Secretar iat C A N M E T , 2001 , Manua l Report 5.4.2, vol . 3 - Predict ion, vol . 4 - Prevent ion and control, vol . 5 - Treatment, vol . 6 Monitoring. Mo loson J . M . , Fa la O., Aubert in M., Buiss iere B. 2005. Numer ical s imulat ions of pyrite oxidation and acid mine drainage in unsaturated waste rock piles. J . Cont. Hydrol. 78 (4): 343-371. Moncrieff, J . B . , and F a n , C . 1999. A model for soil C 0 2 production and transport 2: Appl icat ion to a Florida P inus elliotte plantation. Agr ic. For. Met. 95 : 237-256. Moore , C . J . 1986. Frequency response corrections for eddy correlation sys tems. Boundary-Layer Met. 37: 1 7 - 3 5 . Mua lem, Y . 1978. Hydraul ic conductivity of unsaturated porous media: general ized macroscop ic approach. Water Resour . R e s . 14: 325-334. Mua lem, Y . 1986. Hydraul ic conductivity of unsaturated soi l : Predict ion and formulas. In methods of soil analys is , Part I: Phys ica l and mineralogical methods. Edited by A Klute. Amer ican Society of Agronomy (ASA) and Soi l Sc i ence Society of Amer i ca ( S S S A ) , Mad ison , W is . Agronomy Monograph No. 9. pp. 799-823. N a k a y a m a , F .S . 1990. Soi l respiration. Remote Sens ing Rev . 5 :311-321. Nea l , C , Wh i tehead, P . G . , 1988. the role of C 0 2 in long term stream acidif ication p rocesses : a model ing viewpoint. Hydrological Sc i ence J . 33 : 103-108. References Page 211 N e w m a n , L , M .Sc . Dissertat ion, University of Saska tchewan 1999. N icho lson, R.V. , Gi l lham, R.W., Cherry, J .A . , and Rearddon , E . J . 1989. Reduct ion of acid generat ion in mine tail ings through the use of moisture-retaining cover layers as oxygen barriers. C a n . Geo tech . J . 26: 1-8. N icho lson, R.V. 1994. Iron-sulfide oxidation mechan isms: laboratory studies.. In Short Cou rse Handbook on Environmental Geochemis t ry of Sulf ide M ine-Wastes . Vo l . 22 (ed. J .L . Jambor and D.W. Blowes) , pp. 163-183. Mineralogical Assoc ia t ion of C a n a d a . Noe l , M. , M. , Rykaart , E . ,M 2003. Comparat ive Study of Sur face Flux Boundary Mode ls to Des ign Soi l Cove rs for Mine W a s t e Faci l i t ies. 6 t h International Conference on Ac id Rock Drainage ( ICARD) , Ca i rns , Q L D , 12-18 July 2003 . Nordstrom, D.K., and A lpers , C . N . 1999. Geochemis t ry of acid mine waters. In Environmental Geochemis t ry of Mineral deposit (eds. G . S . P lumlee and M.J . Logsdon) , pp 133-160. Society of Economic Geolog is ts , Little, C O . Norman, J . N . , Ga rc ia , R., V e r m a , S . B . 1992. Soi l CO2 f luxes and the carbon budget of a grass land. J . of Geophys . R e s . 97(17): 18,845-18,853. Norman, J . M . , Kucharik, C . J . , Gower , S T . , Ba ldocch i , D.D., Cri l l , P . M . , Rayment , M. , S a v a g e , K., and Striegl, R . G . 1997. A compar ison of six methods for measur ing soi l -surface carbon dioxide f luxes. J . Geophys . R e s . 102: 28771-28777. Norris, P .R . 1989. Factors affecting bacterial mineral oxidation: the example of carbon dioxide in the context of bacterial diversity. In Proceed ings Internat. Symp . On Biohydrometal lurgy (J. Sal ley, R.G.. I . M c R e a d y & P.L. Wich l lacz , eds.) . C A N M E T Report S P 89-10, 3-14. Dept. Energy Mines R e s o u r c e s C a n a d a . O ' K a n e , M., W i l son , G .D . , Barbour, S .L . , and S w a n s o n , D.A. 1995. Aspec t s on the References Page 212 performance of the till cover sys tem at Equity Si lver M ines Ltd. Proceed ings of the Conference on Mining and the Environment, Sudbury, O N , C a n a d a . O s o z a w a , S . , and Hasegawa , S . 1995. Dial and seasona l changes in carbon dioxide concentrat ion and flux in an And iso l . Soi l S c i . 160: 117-124. Pantel is , D., and Ritchie, A. I .M. , 1991. Macroscop ic transport mechan isms as a rate- limiting factor in dump leaching of pyrite ores. Appl ied Mathemat ics Model l ing, 15:136-143. Pante l is , G . , Ritchie, A . I .M. and Stepanyants , Y . A . 2002. A conceptual model for the descript ion of oxidation and transport p rocesses in sulfidic waste rock dumps. App l . Math Mode l , 26(7): 751-770. P e n m a n , H.L. 1948. Natural evaporat ion from open water, bare soil and grass. Proc . R. S o c . London Ser . A 193, 120-146. Phi l ip, J . R . 1986. L inear ized unsteady multidimentional infiltration. Water Resour . R e s . 22: 1717-1727. P iho l , J . , A lcah iz , J . M . , R o d a , F. 1995. Carbon dioxide effluxes and p C 0 2 in soi ls of three Quercus ilex montane forest. Biochemistry 30: 191-215. Priest ley, C . H . B . , and Taylor, R . J . 1972. "on assessmen t of the surface heat flux and evaporat ion using large-scale parameters", M o n . Weather Rev . , 100: pp. 81-92. R a n k a m a , K. 1954. Isotope Geology. McGraw-Hi l l Book C o . , Inc. Rayment , M.B. , Jarv is , P .D . 1997. A n improved open chamber sys tem for measur ing soil CO2 eff luxes of a Boreal black spruce forest. J . G e o p h y s . R e s . 102: 28779- 287884. Reardon , E . J . , Al l ison G . B . , and Fritz P. 1979. S e a s o n a l chemica l and isotopic variations of soil C 0 2 at Trout Creek , Ontario. J . hydrol. 43 , 355-371. R ichards , L.A. 1931. Capi l lary conduct ion of liquid through porous medium. Journal of References Page 213 Phys ics 1: 318-333. R ichards , G . J . 1998. Min icosm sca le investigations into natural b iogeochemical p rocesses in a sandy vadose zone . Master of Sc i ence Thes i s , Department of Geolog ica l Sc i ences , University of Saska tchewan , Saska toon , Saska tchewan , C a n a d a . Rightmire, C . T., and Hanshaw B. B. 1973. Relat ionship between the carbon isotop composi t ion of soil CO2 and d issolved carbonate spec ies in groundwater. Water Resour . R e s . 9(4): 958-967. Rightmire, C T . 1978. S e a s o n a l variation in P C 0 2 and 13C content of soil a tmosphere. Water Resour . R e s . 14(4): 691-692. Rimstidt, J .D . , Vaughan , D .J . 2003. Pyrite oxidation: a state-of-the-art assessmen t of the reaction mechan ism. G e o c h i m . C o s m o c h i m . Ac ta 67, 873-880. Ritchie, A . I .M. 1994. Short Cou rse Handbook on Environmental Geochemis t ry of sulfide M ine-Was tes ; Jambor , L , B lowes, D. W. , Eds . ; Mineralogical Assoc ia t ion of C a n a d a : Water loo, Ontario, 1994. Ritchie, A . I .M. 1994a. The waste-rock environment. In Short Cou rse Handbook on Environmental Geochemis t ry of Sulf ide M ine-Wastes , Vo l . 22 (ed. J .L . Jambor and D.W. Blowes) , pp. 201-245. Mineralogical Assoc ia t ion of C a n a d a . Ritchie, A . I .M. 1994b. Sulf ide oxidation mechan isms: controls and rates of oxygen transport. In Short Cou rse Handbook on Environmental Geochemis t ry of Sulf ide M ine-Was tes , Vo l . 22 (ed. J .L . Jambor and D.W. Blowes) , pp. 201-245. Mineralogical Assoc ia t ion of C a n a d a . Rochet te, P. , Gregor ich, E . G . , and Desjardins, R.L. 1992. Compar i son of static and dynamic c losed chambers for measurement of soil respiration under field condit ions. C a n . J . Soi l S c i . 72(4): 605-609. References Page 214 Rowel l , D.L., Mart in, M.W., and Nye, P . H . 1967. The measurement and mechan ism of ion diffusion in soi ls. III. The effect of moisture content and soil solution concentrat ion on the self diffusion of ions in soi ls. J . Soi l S c i . 18: 204-222. Russe l l , C .A . , and Voroney, R . P . 1998. Carbon dioxide from the floor of a boreal aspen forest, 1, relationship to environmental var iables and est imates of C respired, C a n . J . Soi l S c i . 78: 301-310. Russe l l , E . J . and App leyard , A . 1915. The atmosphere of,the soi l : its composi t ion and the causes of variation. J . Agr ic. S c i . 7, 1-48. Rykaart , E . M . , W i l son , G .W. , Fredlund, D .G . , and Currey, N.A. 2001 . A spatial flux boundary model for tailings dams . Procedings of the 5 4 t h Annua l Canad ian Geotechn ica l Conferences , 1102-1109. Scan lon , B.R. , Chr is tman, M. , Reedy , R . C . , Porro, I., S imunek, J . , and Flerchinger, G . N . 2002. Intercode compar isons for simulating water ba lance of surficial sediments in semiar id regions. Water Resour . R e s . 38(12): 59-16. Scot t -Denton, L .E . , Sparks , K.L., and Monson , R.K. 2003 . Spat ia l and temporal controls of soil respiration rate in a high-elevation, subalp ine fores. Soi l B io l . B iochem. 35: 525-534. Sel lers , P., Hal l , F., Margol is, H., Kel ly, B., Ba ldocch i , D., den Hartog, G . , Cihlar, J . , R y a n , M .J . , Good i son , B., Gril l, P. , Jon R a n s o n , K., Lettenmair, D., and Wick land , D.E. , 1995. The Boreal Ecosys tem-Atmosphere Study ( B O R E A S ) : an overview and early results from the 1994 field year. Bul l . A m . Meteorol . S o c , 76, 1549-1577. Sh i , Pe i -L i , Zhang , X . Z . , Zhong , Z . M . , and Ouyang , H. 2006. Diurnal and seasona l variability of soil C 0 2 efflux in acropland ecosys tem on the Tibetan P la teau. Agr ic. For. Meteorol . 137: 220-233. References Page 215 Schuepp , P . H . , Leclerc, M.Y. , Macphe rson , J .L , and Desjardins, R.L. 1990. Footprint prediction of sca lar f luxes from analytical solut ions of the diffusion equat ion. Boundary Layer Meteorology, 50: 355-376. S c h u m a n , G . , Carucc ia , F T . & B radham, W . S . 1992. A s s e s s m e n t of factors affecting pyrite reactivity. U .S . Government Report P B 92-140227. Shuttleworth, W . J . 1993. Evaporat ion, in Handbook of Hydrology, McGraw-H i l l , D.R. Maidment , New York. Smolensky , J . , and Hockley, D. 1999. Oxygen Transport P r o c e s s e s in the Nordhalde of the Ronnenburgh Mining District, Ge rmany Mining and the Environment II. Sudbury ' 99 Conference Proceed ings , p271. So lomon , D.K., Ced ing , T . E . 1987. The annual carbon dioxide cyc le in a montaine soi l : observat ions, model ing, and implications for weather ing. Wate r Resour . R e s . 23 (12): 2257-2265. Steffen, Rober tson, and Kirsten (Canada) Inc. 1993. K e y Lake W a s t e Rock Character izat ion prepared for C a m e c o Corp . S tumm, W . , and Morgan, J . J . 1981. Aquat ic Chemist ry . Wi ley- ln tersc ience. alpine- subalp ine sites. Globa l B iogeochem. Cyc les 10(3), 473-482. Sunquist , E.T., 1993. The global carbon dioxide budget, Sc ience (Washington, D. C ) , 259 , 934-941. Sydor , R . C . 1992. Engineered mine tailings cover: verification of drainage behaviour and investigations of des ign. M .Sc . thesis, University of Water loo, Water loo, Ont. Thierron, V . , Laudelout, H., 1996. Contribution of root respiration to total CO2 efflux from the soil of a dec iduous forest. C a n . J . For. R e s . , 26, 1142-1148. Thors tenson, D.C. , and Pol lock, D .P. 1989. G a s transport in unsaturated zones : mult icomponent sys tems and the adequacy of F ick 's law. Wate r Resour . R e s . References Page 2 1 6 25: 477-507. T imms, G . P . , and Bennet, J . W . 2000. In Proceed ing of the International Confe rence on Ac id Rock Drainge ( ICARD) , Denver, Co lo rado 2000: 841-849. Trumbore, S . E. , Dav idson E. A . , de Camargo , P. B., Nepstad D. C , and Martinelli L. A . 1995. Belowground cycl ing of carbon in forests and pastures of Eastern A m a z o n i a . Globa l B iogeochemica l Cyc les 9(4), 515-528. Twine, T . E . , Kus tas , W . P . , Norman, J . M . , Cook , D.R., Houser , P .R. , Meyers , T . P . , Prueger, J . H . , Starks, P . J . , and Wese ly , M.L. 2000. Correct ing eddy-covar iance flux underest imates over a grass land. Agr i . For. Meteorol . 103: 279-300. Unsaturated Soi ls Group 1997, So i lCove r User ' s Manua l , Department of Civi l Engineer ing, University of Saska tchewan , Saska toon , Saska tchewan , C a n a d a . Ve rmaak , R., and Bezuidenhout 2003. The performance of trial cover used for the rehabilitation of acid generat ing coal reject dumps in Northern Kwazu lu-Nata l , South Afr ica. Proceed ings Sixth International Conference on Ac id Rock Drainage ( ICARD) , 399-410. Wal l ick, E.I. 1983. G a s composit ion in the unsaturated zone as an index of geochemica l equil ibrium in reclaimed landscapes . In A s s e s s m e n t of Rec lamat ion Potential and Hydrologic Impart of La rge -Sca le Sur face Mining of Coa l in P la ins A reas of Alberta: Summary Report of Activit ies 1982-1983, V o l . Append ix H. Alberta Chapter C . L . R . A . , Rec lamat ion Conference, 24-25 February, 1983. W a n g , Y . , A m u n d s o n , R., and Trumbore, S . 1999. The impact of land use change on C turnover in soi ls. G loba l B iogeochem. Cyc les 13(1): 47-57. Weas t , R. and Ast le , M . J . 1981. C R C Handbook of Chemist ry and Phys ics . C R C P r e s s , Inc. W e b b , E.K. , Pea rman , G.I., and Leuning, R.L. 1980. Correct ion of flux measurements References Page 217 for density effects due to heat and water vapour transfer. Quarter. Journal of the Roya l Meteorological Soc iety 106: 85-100. W e e k s , E . P . , Earp , D.E. , and Thompson , G . M . 1982. U s e of atmospher ic f luorocarbons F-11 and F-12 to determine the diffusion parameters of the unsaturated zone in the southern High Pla ins of Texas , Water Resour . R e s . 18: 1365-1378. We l les , J . M . , Demet r iades-Shah, T . H . , and McDermitt , D.K., 2001 . Considerat ions for measur ing ground C 0 2 effluxes with chambers . Chemica l Geo logy 177: 3-13. Wick land , K. P. , Str iegl, R. G . , 1997. Measurements of soil carbon dioxide and methane concentrat ions and f luxes, and soil properties at four ages of jack pine forest in the Southern Study, Saska tchewan , C a n a d a , 1993-1995. U .S . G e o l . Surv. O p e n - Fi le R e p . , 97-49. W ie rsma , C .L . , Rimstidt, J .D . , 1984. Rates of reaction of pyrite and marcasi te with ferric iron at pH 2. G e o c h i m . C o s m o c h i m . Ac ta 48 : 85-92.Wi lson, M. R. 1987. P e d o g e n e s i s of fluids assoc ia ted with unconformity-type uranium deposi ts in Saska t chewan , and with the Zor tman-Landusky A u - A g deposi ts in Montana. P h . D . , University of Saska tchewan . Wi l son , G . W . 1990. Soi l Evaporat ive f luxes for Geotechn ica l Engineer ing problems. P h . D . thesis, University of Saska tchewan , Saska toon , Saska tchewan , C a n a d a . W i l son , G . W . , Fredlund, D .G. , and Barbour, S . L . 1994. Coup led soi l -atmosphere model ing for soil evaporat ion. C a n . Geo tech . J . 31 : 151-161. Wi l son , G .W. , Fredlund, D .G. , and Barbour, S .L . 1997. The effect of soil suct ion on evaporat ive f luxes from soil sur faces. C . Geo tech . J . , 34: 145 -155 . Wind , G . P . 1955. Field experiment concerning capil lary rise of moisture in heavy clay soi l . Nether lands Journal of Agricultural Sc ience 3: 60-69. Wohlfahrt, G . , An fang , C , B a h n , M., Haslanter, A . , Neese ly , C , Schmitt , M. , Drosler, References Page 2 1 8 M., Pfdenhauer , and J . , C e r n u s c a , A . 2005. Quanify ing nighttime ecosys tem respiration of a meadow using eddy covar iance, chambers and model ing. Agr i . For. Meteorol . 128: 141-162. W o o d , B.D. , and Petrait is, M . J . 1984. Origin and distribution of carbon dioxide in the unsaturated zone of the southern High Pla ins of Texas . Water Resour . R e s . 20 (9): 1193-1208. W o o d , B. D., Kel ler K. C , and Johnstone, D. L. 1993. In situ measurement of microbial activity and controls on microbial C 0 2 production in the unsaturated zone . Wate r Resour . R e s . 29(3), 647-659. Yanfu l , E .K. , and A u b e , B . C . 1993a. Model l ing moisture retaining soi l covers . In Proceed ings of the 1993 Joint C S C E - A S C E N a - © 2005 N R C C a n a d a 1198 C a n . Geo tech . J . V o l . 42 , 2005 © 2005 N R C C a n a d a . Yanfu l , E.K. , Mousav i , S . M . , and Y a n g , M. 2003a . Model ing and measurement of evaporat ion in moisture-retaining soil covers. A d v a n c e s in Environmental R e s e a r c h , 7: 7 8 3 - 8 0 1 . Yanfu l , E .K . and Mousav i , S . M . 2003b . Est imat ing fall ing rate of evaporat ion from finite soil co lumns. The Sc ience of the Total Environment 313: 151-152 Y o n g , R . N . 2001 . Geoenv i ronmenta l engineer ing - Contaminated soi ls, pollutant fate, and mitigation. C R C P r e s s , B o c a Raton, F la . Zi l l , D . G . and Cu l len , M.R. 1992. A d v a n c e d Enginer ing Mathemat ics. P W S Publ ishing Company , Bos ton , M A . Appendices Page 219 APPENDIX A Measuring 0 2 Fluxes Using the Dynamic Closed Chamber (DCC) System NOTE: Th is sect ion is out of scope of this thesis. This is an on-going research work being carried out by the author of this thesis in the Department of Mining Engineer ing at U B C , under the supervis ion of Prof. Ward Wi lson . A1. Design of the dynamic closed chamber (DCC) Full details of the des ign, fabrication, and descript ion of the D C C chamber are presented in sect ion 3.4.2.1 of this thesis, and in Kabwe (2001) and Kabwe et a l . (2002). The dynamic c losed chamber ( D C C ) used in this work was initially des igned for measur ing C 0 2 f luxes (Kabwe, 2001 and Kabwe et a l . , 2002). It cons is ts of an open- ended rim (collar) with a lid. Full detai ls of the des ign , construct ion, and operation are presented in Kabwe et a l . (2002). C h a m b e r collars were fabricated from f iberglass rims (0.76m dia. x 0.15m height); the chamber lid (0.76m dia. x 0.05m thick) was fabricated from Plexig las. The lid was attached to the col lars with nuts and bolts. The changes of mass of 0 2 within the chamber was measured using an Oxymax E R - 1 0 oxygen gas analyzer. A2. Principle of operation The rate of changes of the mass of 0 2 within the chamber p laced on the surface of the waste-rock dump can be descr ibed by: (A1) Appendices Page 220 where F is the flux of 0 2 through the surface of the waste-rock and A is the area of the base of the chamber. The rate of change of 0 2 concentrat ion within the chamber, dC/dt , is given by (Timms and Bennett, 2001): d C = J _ d r n dt V dt v ; where V is the vo lume of the chamber. The 0 2 flux can then be calculated from the rate of change of 0 2 concentrat ion within the chamber. Combin ing Equat ions 1 and 2 gives: ^ = F - (A3) dt h v ' where h = V / A , the height of the chamber. The critical aspect of the flux chamber design is the need to be able to measure the low f luxes typical of covered sys tems. A3. Description and measurement principle of the Oxymax ER-10 oxygen gas analyzer 0 2 was ana lyzed using an 0 2 - G a s Respi rometer (Micro-Oxymax E R - 1 0 , Co lumbus Instruments, Ohio U S A ) . The E R - 1 0 is a computer ized apparatus for measur ing very low levels of gaseous oxygen uptake. A n I B M - P C compat ib le computer maintains and displays the operation of the Mic ro -Oxymax instrument. Before starting the measurement , the system needs only the time interval between samp les to be speci f ied and the chamber vo lume which is computed automatical ly during sys tem calibration. W h e n the experiment is started, the software a s s u m e s control of the acquisit ion of information and storage of results and/or presentation to the printer. Oxymax E R - 1 0 can measure liquid or solid samp les from 50 mL to 10 L in vo lume. The principle of measurement , involves air sampl ing from the head space of the chamber, circulating it through the gas analyzer and returning back to the sample chamber without any contact with the sample . S a m p l e s are cont inuously aerated with adjustable airflow (100 mL/min. to 1,500 mL/min.), except for the short t ime interval when a particular sample is being measured by the gas analyzer. Cal ibrat ion of 0 2 gas analyzer is performed automatical ly at speci f ic t ime interval with ambient air, thus, Appendices Page 221 removing the need for the mixed gas bottle. Resul ts of measurements are presented in I^ICVmin or as an accumulated (total in u.1) value of 0 2 consumed from the beginning of the experiment. The Oxymax E R 1 0 operates on the principle of using gas sensor to measure the change in the oxygen in the head space of a measur ing cell and using this information to calculate how much oxygen the sample is consuming (oxygen uptake). To compute the oxygen consumpt ion requires two measurements of the head space separated by a span of t ime. The oxygen senso r operates as an oxygen battery (fuel cell), and measures oxygen percentage directly. The sensitivity of the sys tem to oxygen consumpt ion (uptake) is dependent on two factors: the volume of the headspace gas in the measur ing cel l , and the span of t ime between measurements . In genera l , the smal ler the head s p a c e volume, the higher the sensitivity. A l so , the longer the time between measurements , the higher the sensitivity. The volume of the headspace in the cell is automatical ly measured by the apparatus. The apparatus uses a direct method to detect and correct errors in the sensor outputs (resulting from environmental temperature changes , barometric pressure changes , or changes in the sensor) , and thereby ra ises the system's measur ing accuracy. The Mic ro-Oxymax instrument a lso contains a feature cal led 'automatic refresh" which al lows the gas in the headspace of the measur ing cell to be replaced periodically with fresh air or other gas mixtures. Th is feature is important if the level of oxygen consumpt ion by the sample is high enough that the oxygen becomes depleted in the headspace gas . A4. Preliminary Results and Discussion A4.1 Site Location The Syncrude C a n a d a Ltd ( S C L ) mine is located 30 km north of Fort McMurray , Alber ta, C a n a d a . The regional cl imate is continental. The mean annual precipitation is approximately 440 mm of which 310 mm is rain (Meiers et a l . , 2006). The mean annual potential evaporat ion (Penman) is in the range of 600 to 700 mm/year (Boese , 2003; Babour et a l . , 2001 ; E lshorbagy et a l . , 2005). Appendices Page 222 Mildre Lake Mine surface dump and other referred to as S W 3 0 dumps were constructed with marine sal ine-sodic shale overburden removed during mining of soil sands . The S C L mine produces over 200,000 barrels of oil per day (Meiers et a l . , 2006). Up to 14 tonnes of overburden is excavated for each cubic meter of oil p roduced. T h e s e overburden deposi ts are salt rich (saline) and sodic . The glacial soil consis ts of approximately 2 % gravel , 3 8 % sand , and 6 0 % silt and clay s ized particles while the shale consists of approximately 0 .5% gravel, 14 .5% sand , and 8 5 % silt and clay s ized particles (Meiers et a l . , 2006). A4.2 Selection of the height of the DCC chamber The height of the chamber was selected based on the test results conducted on S W D 3 0 dump at Syncrude on August 12, 2001. The rate of change in oxygen concentrat ion within the chamber descr ibed by Equat ion 3 A shows that the smal ler the height, the greater the rate of change of oxygen concentrat ion. However, a smal ler height a lso results in a smal ler gas vo lume and a greater relative uncertainty on the vo lume, due to the irregular profile of the cover surface (Timms and Bennett, 2001). F ig. 1 shows the changes in oxygen concentrat ion measured using three different chamber vo lumes of 2.5, 4 .5 , and 6.3 L with the corresponding chamber heights of 0.01, 0.015 and 0.02 m, respectively. Resul ts clearly indicated that the smal ler vo lume of 2.5 L (e.g., h = 0.01 m) yielded the smal ler rate of change of concentrat ion (e.g., a nearly flat s lope). However, the chamber vo lume of 4.5 L (h = 0.015 m) yielded the greater rate of change of oxygen concentrat ion (e.g., s teeper slope) than the chamber vo lume of 6.3 L (h = 0.02 m). The height (h = 0.015 m) for the headspace of the chamber presented here, was therefore selected as a compromise between maximizing the rate of change of oxygen concentrat ion and minimizing the uncertainty on the gas vo lume of the chamber. Appendices Page 223 21.0 10 20 30 O 4.5 L • 2.5 L • 6.3 L 40 50 Time (min) 60 70 (2) -G 1— 80 •e 90 Fig. 1. C h a n g e s in oxygen concentrat ions in chambers with vo lume of (1) 2.5 L and (2) 4.5 L and (3) 6.3 L as a function of t ime. Measurements were done at the S W D 3 0 dump at the Syncrude mine on 19 July, 2001 . A4.3 Effect of Relative Humidity Fig. 2 showed the oxygen concentrat ion and oxygen consumpt ion rate measured on 12 August , 2001 at the D S W R as a function of t ime. The measurements results yielded a l inear dec rease in oxygen concentrat ion of the form: y = -0.0012x + 20.905 ( R 2 = 0.9934). The plot revealed a slight large initial drop in oxygen concentrat ion within the chamber during the first 10 min time-interval fol lowed by a more gradual dec rease in O 2 concentrat ion. The initial larger drop in O 2 concentrat ion is likely due to the effect of relative humidity. The effect of relative humidity on oxygen concentrat ion within the chamber is well documented in literature. T imms and Bennett (2001) indicated that early measurements with the surface chamber dev ice revealed a large initial drop in oxygen concentrat ion within the chamber when it w a s placed on the ground, fol lowed by a more gradual dec rease . E R - 1 0 employs a drying agent A p p e n d i c e s P a g e 2 2 4 (Anhydrous Magnes ium Perchlorate) in port in which gas is drawn to remove water vapor. 20.92 20.80 0 10 20 30 40 50 Time (min) 60 70 • Cone. • Rate 1000 80 F i g . 2. C h a n g e in oxygen concentrat ion and oxygen consumpt ion rate in a chamber installed at the Dei lmann south waste-rock ( D S W R ) as a function of time. Measurements were done on August 9, 2001. Wate r removal capaci ty depends on the type and s ize of the drying agent employed. F ig . 2 showed the oxygen consumpt ion rate (open symbols) measured within the chamber . The degree of variation w a s smal l (covar iance, C V = 17.6) with a mean rate of 133 + 23.4 \x min" 1 . During the measurements the chamber temperature dec reased from 17.1 to 15.5 °C. Resul ts a lso showed that the sensor 's pressure remained constant (797.9) throughout the duration of the measurements . Moreover the E R - 1 0 detects and corrects errors in the sensors outputs resulting from environmental temperature changes and barometric pressure changes . The effect of high CO2 concentrat ion was el iminated by including soda lime in the drying co lumn. Appendices Page 225 A.5 Effect of soil cover system on 0 2 diffusion Fig. 3 shows the measured oxygen concentrat ions in the chambers installed on a cover (curve with broken line) and non-cover (curve with solid line) sect ions of the D30 Dumps using the E R - 1 0 Respirometer . The measurements were done on July 18, 2001 . Resul ts showed that the oxygen concentrat ion in the headspace of the chamber installed on the cover sect ion did not change significantly throughout the duration of 86 min test-period. The plot, however, revealed a slight drop in oxygen concentrat ion from 20.91 to 20 .87% during the first 10 min time-interval. 20.95 20.90 o > 20.85 + c 20.80 + o 2 20.75 + § 20.70 + c ° 20.65 + o 6 20.60 + 20.55 10 -a- - With cover • Without cover 20 30 40 50 time (min) 60 70 80 Fig. 3. C h a n g e s in oxygen concentrat ions in the chambers installed on (1) a cover and (2) uncover portions of the D30 dump at Syncrude on July 8, 2001 . The slight dec rease in concentrat ion is likely related to the effect of relative humidity on oxygen concentrat ion within the chamber. The oxygen concentrat ion in the chamber installed on uncover sect ion of the Dumps dec reased initially at a faster rate fol lowed by a gradual dec reased with time. The dec rease in oxygen concentrat ion was represented by the function y = - 1 0 " 6 * X 3 + 0 .0003X 2 - 0 .0244X + 20.948 ( R 2 = 0.9948). At the end of the 86-min test period the concentrat ion decreased from 20.91 to 20 .58%. Appendices Page 226 Finally, the above results indicated that the D C C chamber with the E R - 1 0 Respi rometer can be suitable for assess ing the performance of the cover p laced on mine waste dumps. Appendices Page 227 APPENDIX B Eddy Correlation (EC) Method B.1 Introduction This sect ion presents a brief theory and derivation of bas ic equat ions descr ib ing the Eddy correlation (EC) method. B.2 Theory and basic equations The eddy correlation flux, is expressed as where: p is the density of the air (kg m"3); w is the vertical wind (m s"1); c is the mass concentration of substance (kg kg"1) i.e., molecular weights for C 0 2 and air (mjvr\a). Eddy correlation when standard micrometeorogical criteria are met [Hicks et al. 1989] will provide absolute evaluations of vertical fluxes in natural environments without making assumptions associated with diffusivities or the nature of the surface cover. In addition, the exchange rate measured represents a spatially integrated flux and the technique is unobtrusive, therefore not disturbing the environment under study. If we write each term of the right- hand side of the above equation [A1] as the sum of a mean value and an instantaneous departure from that mean, i.e. Flux (kg m 2 s 1) = p.w.c [A1] p-p + p'\ w=w + w'; c-c+c [A2] then the equation becomes: flux = pwc + pwc' + pw'c + pwc' + [A3] p'wc + p'wc' + p'w'c + p'w'c' since Appendices Page 228 [A4] the average value of the flux reduces to: flux = pwc+ pwc'+ w p' c' + c p' w' + p'w'c' [A5] If one ignores density fluctuations (small near the surface) and puts p = p: flux= pwc+ pw'c' [A6] where the first term is the flux due to the mean vertical flow and the second is that due to the eddies. Over uniform surfaces this further reduces to: Application of this method requires measurement of the vertical wind and the substance concentration (i.e. temperature for the heat flux, vapor pressure for the water vapor flux, C 0 2 concentration for the carbon dioxide flux) with sensors of time response short enough to respond to all eddies (fluctuations) (typically a fraction of a second or less). The two instantaneous measurements must be multiplied and their products summed to give flux totals over a period. This is most efficiently done by a measurement system incorporating a small computer. The eddy correlation method derived the latent and sensible heat fluxes using the following relationships: where w, T, q, and C p are the vertical velocity, temperature, humidity, and specific heat capacity of air respectively. F = pw'c' [A7] H = pCpw'T' [AS] E = Lvw'q' [A9] B .3. Wind profile and the transfer of momentum The wind profiles above a stand can be represented by the simple logarithmic equation: Appendices Page 229 "* i i u7 = — l n | k (z^ \ Z o J [A10] where u z is the velocity at height z, u. is the friction velocity, ZQ is the roughness parameter and k is Von Karman's constant (k = 0.4, average size of the eddy). Since momentum equals to mass times velocity, a decrease in wind speed represents a decrease of momentum. This decrease or loss of momentum may be thought of as a downward flux of momentum from the air towards the surface. The momentum flux (x) is expressed as: du T = pK M dz [A11] where, K M is eddy transfer for momentum and x is also called dynamic viscosity. The kinematic viscosity is expressed as: du = K P M dz [A12] Assume: KM = kut(z-Z))and — = u2 P where D is zero-plane displacement, Equation [A12] becomes: du u. [A13] dz k(z-D) This expression represents the wind shear at height z over an aerodynamically rough surface Rearranging and integrating the above equation U-U 2=2 ^du =k J «=0 z=j z-D [A14] gives u, . u= — In z-D \ o J = ^ l n ( z - 2 > ) - l n ( z 0 ) K [A15] Appendices Page 230 If one measure the wind speed u, at several heights, z, and plotting ln(z - D) as a function of u, one can calculate u* and ZQ Equation [14] could also be reported as: k Au Au = — In rz2-D^ v z , - D j u. = In fz2-D^ Kzx-D [A16] The sensible heat and latent heat fluxes can also be expressed as dT H = -pCpKh dz LE = -pAKv de dz [A17] [A18] where, p is air density, C p is the specific heat capacity of air, K h is the eddy diffusivity for heat, and A 6 / A z is the potential temperature gradient, K v is the eddy diffusivity for water vapour, and Ae/Az is the vapor pressure gradient. Similar expressions can be written for gradients of temperature and vapour pressure as follows dT H dz C u,k{z - d) [A19] and de yXE dz pCputk(z-d) where y is the psychrometric constant. [A20] Appendices Page 231 APPENDIX C Computer code for C0 2 diffusion model This section presents the computer codes for the "C02" diffusion model and spreadsheets describing different function of the program including: 1. Input spreadsheet 2. Water content spreadsheet 3. Temperature spreadsheet 4. Output spreadsheet 5. Results spreadsheet 1. Input spreadsheet The user uses this spreadsheet to enter the # of nodes and # of simulation days. All constants values are also entered on this spreadsheet. The elevation (in m), porosity, and initial concentrations profile (in Kg/m3) is required to run the program and should be entered on this spradsheet. The program run button [C02] is also located on this spreadsheet. 2. Water content spreadsheet This spreadsheet contains a table of water content (as %). The Y axis represent the # of node (Elevation) and X axis represents # number of days of simulations. Note that the number of nodes and days are determined in the input spreadsheet. 3. Temperature spreadsheet This spreadsheet is used to enter the temperature profiles for each day if known. However, the average temperature for the profile can be entered throughout the table. 4. Output spreadsheet Appendices Page 232 The output spreadsheet presents the complete results of calculations of new concentration and diffusion coefficient for each iteration. The spreadsheet displays the: day #, iteration*, Nodes, new concentration, concentration changes, diffusion coefficient, saturation and time difference calculated values. 5. Results spreadsheet The new concentrations (in %) and diffusion coefficients (in m/s) for each profile and for each day are printed on this spreadsheet. Appendices Page 233 C.1. Computer code for " C Q 2 " model Publ ic elev(), por(), oldconc(), daysv() A s Double Publ ic watcont(), temp(), G(), Temperature(), allsatu(), GO, k, a , b A s Double Publ ic Nodes , days , maxdelt, mindelt, starow A s Integer S u b read_input() Nodes =Worksheets(" lnput") .Cel ls(5, 3).Value days = Worksheets(" lnput") .Cel ls(6, 3).Value maxdelt = Worksheets(" lnput n ) .Cel ls(7, 3).Value mindelt = Worksheets(" lnput") .Cel ls(8, 3) .Value starow = Worksheets(" lnput") .Cel ls(9, 3).Value GO = Worksheets(" lnput , , ) .Cel ls(10, 3).Value k = Worksheets(" lnput , , ) .Cel ls(11, 3) .Value a = Worksheets(" lnput") .Cel ls(12, 3).Value b = Worksheets(" lnput") .Cel ls(13, 3).Value Tref = Worksheets( , , lnput") .Cel ls(14, 3).Value R e D i m e lev(Nodes) , por(Nodes), o ldconc(Nodes) , daysv(days) R e D i m watcont(Nodes, days) , a l lsatu(Nodes, days) , G ( N o d e s , days) , Temperature(Nodes, days), temp(Nodes, days) Worksheets(" resul ts" ) .Cel ls .ClearContents Worksheets("output" ) .Cel ls .ClearContents Worksheets( "TempDi f f ' ) .Ce l ls .C learContents For i = 1 To Nodes elev(i) = Worksheets(" lnput") .Cel ls(starow + i - 1 , 2) .Value por(i) = Worksheets(" lnput") .Cel ls(starow + i - 1 , 3) .Value oldconc(i) = Worksheets(" lnput") .Cel ls(starow + i - 1 , 4) .Value Next i 'populates the vector containning saturation, G and daysvector matrixes Appendices Page 234 For i = 1 To Nodes For d = 1 To days daysv(d) = Worksheets("watercont") .Cel ls(4, d + 1).Value watcont(i, d) = Worksheets("watercont") .Cel ls( i + 4, d + 1).Value allsatu(i, d) = watcont(i, d) / 1 0 0 Temperature( i , d) = Worksheets("Temperature") .Cel ls( i + 3, d + 1).Value Worksheets("TempDif f ' ) .Cel ls( i , d) .Value = Temperature( i , d) - Tref temp(i, d) = Worksheets( 'TempDi f f ' ) .Cel ls ( i , d) .Value N e x t d Worksheets("resul ts") .Cel ls( i + 2, 2) .Value = oldconc(i) Next i End S u b S u b C o C O N ( ) Dim deltax(), sumdelt() A s Double Dim diffusion(), t imesteps(), waterpor(), airpor() A s Double Dim eqpor(), avgdiffusion(), concchange() , sctime() A s Double Dim difffluxin(), satu(), difffluxout(), coflux(), newconc() A s Double diffcoefair = 0.000018 'diffusion coeficients (De)in m2/s diffcoefwater = 0 .0000000025 'diffusion coeficients (De)in m2/s henry = 0.03 r e a d j n p u t ' average of the s ize of the spaces on either s ide of a node Delev R e D i m del tax(Nodes), di f fusion(Nodes), t imesteps(Nodes) R e D i m waterpor(Nodes), airpor(Nodes), eqpor(Nodes) , avgdif fusion(Nodes) R e D i m concchange(Nodes) , diff luxin(Nodes) R e D i m difff luxout(Nodes), cof lux(Nodes), newconc(Nodes) , satu(Nodes) , sct ime(days) For i = 1 To Nodes - 2 deltax(i + 1) = Abs(((elev(i + 2) - elev(i + 1)) + (elev(i + 1) - elev(i))) / 2) Next i Appendices Page 235 deltax(1) = deltax(2) del tax(Nodes) = del tax(Nodes - 1 ) 'calculate t ime interval in seconds for t imesteps For d = 1 To days -1 sctime(d) = (daysv(d + 1) - daysv(d)) * 86400 N e x t d 'main loop to compute C o 2 concentrat ion per day d = 1 q = 2 Whi le d <= days For i = 1 To Nodes satu(i) = allsatu(i, d) Next i countb = 1 difference = maxdelt + 1 sumdeltat = 0 Whi le sumdeltat < sctime(d) 'statement to avoid extreme va lues For i = 1 To Nodes If satu(i) <= 0 Then satu(i) = 0.00001 End If If satu(i) >= 1 Then satu(i) = 0.9999 End If Next i 'compute Dw, D a , Eqpor, De For i = 1 To Nodes waterpor(i) = satu(i) * por(i) Appendices Page 236 airpor(i) = por(i) - waterpor(i) eqpor(i) = airpor(i) + henry * waterpor(i) diffusion(i) = (1 / por(i) A 2) * (diffcoefair * airpor(i) A 3.5 + henry * diffcoefwater * waterpor(i) A 3.5) t imesteps(i) = eqpor(i) * 0.5 * (deltax(i) A 2) / diffusion(i) G( i , d) = GO * (airpor(i) A a) * (waterpor(i) A b) * Exp(k * (temp(i, d))) Next i deltat = Appl icat ion.WorksheetFunct ion.Min(t imesteps) If deltat > maxdelt Then deltat = maxdelt End If If deltat < mindelt Then deltat = mindelt End If If deltat > difference Then deltat = difference End If sumdeltat = sumdeltat + deltat ' so lve finite difference equation avgdi f fusion( l ) = 0 For i = 1 To Nodes -1 dlev = (elev(i) - elev(i + 1)) avgdiffusion(i + 1) = dlev / ((dlev / (2* diffusion(i))) + (dlev / (2 * di f fusion^ + 1)))) Next i ' compute concchange vector to solve the differential equation concchange ( l ) = 0 For i = 2 To Nodes -1 d1 = (oldconc(i) - oldconc(i -1 ) ) / (Abs(elev(i) - elev(i -1))) d2 = (oldconc(i + 1) - oldconc(i)) / (Abs(elev(i + 1) - elev(i))) concchange( i ) = (deltat / (deltax(i) * eqpor(i)) * (-avgdiffusion(i) * d1 + avgdiffusion(i + 1) * d2 + G( i , d))) Next i Appendices Page 237 concchange(Nodes) = concchange(Nodes - 1 ) For i = 1 To Nodes newconc(i) = oldconc(i) + concchange( i ) If newconc(i) < 0 Then newconc(i) = 0 End If oldconc(i) = (newconc(i)) Next i For i = 1 To Nodes -1 ' Interpolate new saturation va lues for the next iteration satu(i) = allsatu(i + 1, d) + (allsatu(i + 1, d + 1) - allsatu(i + 1, d)) * deltat / sctime(d) Next i countb = countb + 1 difference = sctime(d) - sumdeltat For i = 1 To Nodes Worksheets("output") .Cel ls(q, 1).Value = d Worksheets("output") .Cel ls(q, 2).Value = countb Worksheets("output") .Cel ls(q, 3).Value = i Worksheets("output") .Cel ls(q, 4) .Value = newconc(i) Worksheets("output") .Cel ls(q, 5).Value = concchange( i ) Worksheets("output") .Cel ls(q, 6) .Value = diffusion(i) Worksheets("output") .Cel ls(q, 7).Value = satu(i) Worksheets("output") .Cel ls(q, 8).Value = difference q = q + 1 Next i q = q + 1 W e n d Worksheets("resul ts") .Cel ls(1, 1). Va lue = "New Concentrat ion" Worksheets("resul ts") .Cel ls(4 + Nodes , 1).Value = "Diffusion" For i = 1 To Nodes Worksheets("resul ts") .Cel ls(2, d + 1).Value = daysv(d) Appendices Page 238 Worksheets("resul ts") .Cel ls( i + 2, d + 1).Value = newconc(i) * 100000000 Worksheets("resul ts") .Cel ls( i + 2, 1). Va lue = elev(i) Worksheets("resul ts") .Cel ls(5 + Nodes , d + 1).Value = daysv(d) Worksheets("resul ts") .Cel ls( i + Nodes + 5, d + 1).Value = diffusion(i) Worksheets("resul ts") .Cel ls( i + Nodes + 5, 1). Va lue = elev(i) Next i d = d + 1 W e n d Worksheets("output M ) .Cel ls(1, 1).Value = "Day" Worksheets("output") .Cel ls(1, 2) .Value = "Iteration*" Worksheets("output") .Cel ls(1, 3).Value = "Node" Worksheets("output").Cel ls(1 4) .Value = "Newconc" Worksheets("output").Cel ls(1 5).Value = " C o n c c h a n g e " Worksheets("output").Cel ls(1 6).Value = "Diffusion" Worksheets("output").Cel ls(1 7).Value = "Saturat ion" Worksheets("output").Cel ls(1 8).Value = "Time Difference" End S u b S u b UpdateEmbeddedChar t ( ) '\ attach macro to chart object Act iveSheet .DrawingObjects(Appl icat ion.Cal ler) .Select U s e r F o r m l .Show End S u b S u b UpdateChar tSheetO '\ attach macro to rectangle drawn over chart U s e r F o r m l . S h o w End S u b Appendices Page 239 C.2. T y p i c a l s p r e a d s h e e t s T a b l e C 1 . T y p i c a l i n p u t s p r e a d s h e e t L' Microsoft Excel - Co2b_v2 |1)UpdatedlK2b 1] FJe Edit View Insert Format lools Data Window Help Acrobat A B C D E F G H I J K LT 1 2 3 Constants 4 Description Constant Value 5 Nodes 0 0 6 Days 11 7 Maxdelt at 8600 CQ2 8 Mindelt at 50 9 Data starts 18 10 GO 1.00E-15 11 k 0,044 12 a 2 _ 13 b 1,25 14 Tref 5,17 15 n an 16 U.3U n on 17 i-nodes Elevation(i) Porosity (i) OldCone (i) U.oU 0,70 060 18 1 0,00 4.00E-01 3.60E-09 19 2 0,02 4.00E-01 3.80E-09 050 20 3 0.15 4.00E-01 5.70E-09 040 — 1 i — • — / -«— 21 4 0,30 4.00E-01 7.40E-09 030 22 5 0.45 4.00E-01 9.20E-09 0.20 23 6 0,60 4.00E-01 1.10E-08 010 24 7 0,75 4.00E-01 1.30E-08 0.00 — * ~ T * ^ i 1 1 1 r~ 25 8 0.9C 4.00E-01 1.60E-08 1 2 3 4 5 6 1 N > Hjjnput/ water •nt / Temperature / TempDiff / out ]ut / Bheetl, results / l«1 1 Draw* k Auto5hapes' \ ^ • 0 i 4 Q I Q • I . J T A . mm In Ready 1 Start Q PhDThesisApp... Q C02Programl... 3 PhDthesisApp 3 w * 1:02 PM Friday rosoftExc... j Microsoft Visu... • 6V. i; /Cl/2007 A p p e n d i c e s P a g e 240 Table C2. Typical Water content spreadsheet 2 Microsoft Excel - Co2b_v2 |1)UpdatedlK2b I. ||d ffi I ] Rle Edit View Insert Fymat Tools Data Window Help Acrobat o o t i d s - •• t z - w i i Format Painter IT F23 A B C D E F G H 1 J K L M * 1 2 3 Water concentration 4 5 Nodes\days 1 2 3 4 5 6 7 8 9 10 11 l : 1 12.8000 10,88 9.248 7.8608 6,68168 5,679428 4.827514 4,103387 3,487879 2.964697 2.519992 3.2759' 6 . 2 15.2000 12.92 10.982 9.3347 7,934495 6,744321 5.732673 4272772 4.141856 3 520578 2.992 491 3.89023! 7 3 171500 14,5775 12,391 10.5322 8,952407 7,609546 6,468114 5.497897 4.673213 3.972231 3.376396 4.38931! 8 4 18.5000 15.725 13.366 11.3613 9.657116 8,208548 6,977266 5.930676 5,041075 4.284914 3,542175 4,73482! 9 5 19,0200 16,167 13,742 11,6807 9,928559 8,439275 7,173384 6.097376 5,18277 4.405354 3,744551 4.86791' 10 6 27,7200 23.562 20.028 17,0235 14.47001 12,29951 10,45458 8,886397 7,553437 6,420422 5,457358 7.09456! 11 7 33,4800 28.458 24,189 20,5609 17.47677 14,85525 12,62697 10,73292 9,122983 7,754535 6,591355 8,56876! • 12 8 34,0000 28.9 24.565 20,8803 17.74821 15,08598 12,82308 10,89962 9.264678 7S74976 6,69373 8.70184' 13 14 15 16 ! 17 18 19 20 21 22 23 24 25 26 27 no i i< < • w / Input \watercont/ Temperature / TempDiff / output / Sheetl / results / |< D r a w Ready A y t o S h a p e s 'X \ • 0 | 4 O 1 H * " / ' i * B S g i § . '» Sf3rt 3j PhDTheslsApp.,. Q c O Z P r o g r a r n l . , . .^MiDthesisApp. . . 3 Documents-. . . M MicrosoftExc... j Microsoft Visu,., { [O] % j W ftr Appendices Page 241 Table C3 . Typical temperature spreadsheet L' Microsoft Excel • Co2b_v2 (1)UpdatedlK2b m Rle Edit View Insert Format Tools Data Window Help Acrobat IBB. 1 x A B c D E F G | H 1 J L M N o - 1 2 Temperature 3 Node/ Days 1 2 3 4 5 6 7 8 9 10 11 12 13 4 1 14 14 14 14 14 14 14 14 14 14 14 14 14 5 2 14 14 14 14 14 14 14 14 14 14 14 14 14 6 3 14 14 14 14 14 14 14 14 14 14 14 14 14 7 4 14 14 14 14 14 14 14 14 14 14 14 14 14 8 5 14 14 14 14 14 14 14 14 14 14 14 14 14 9 6 14 14 14 14 14 14 14 14 14 14 14 14 14 10 7 14 14 14 14 14 14 14 14 14 14 14 14 14 11 8 14 14 14 14 14 14 14 14 14 14 14 14 14 12 13 14 15 16 17 18 , 19 X 1 21 22 23 24 25 26 27 28 K 4 • M / Input / watercont ^Temperature / TempDiff / output / Sheetl / results / D r a w AutoShapes' \ \ • 0 H 4 0 Ready , SfS/t jjjPhDThesisApp... ^ C O Z P r o c r a m l . . . QPhDthesisApp, . . 3 Document3-, , , £ Microsoft E x c . j Microsoft Visu.,. { H ff. " * A p p e n d i c e s P a g e 2 4 2 Table C4. Typical model output spreadsheet 2 Microsoft Excel - Co2b_v211 )Updated!K2b D file Edit View Insert F p a t lools Data Window Help Acrobat i t . H22 » '$ 85300.5769261254 19, . i x A B c D E F Diffusion G Saturation H 1 J K 1 Day Iteration* Node Newconc Concchange Time Difference 2 2 1 . 0 0 E € ) 3 .60E-09 O.DE+00 2 .82E-06 2E-01 8 . 6 E 4 Q 4 3 2 2.DDE+0U 2.92E-09 1.2E-10 2 .56E-06 2E-01 8 .6E+04 4 2 3 .00E4O0 5 .63E-09 -7.4E-11 2 .36E-06 2E-01 8 . 6 E ^ 4 5 2 4 . 0 0 E O 0 7.40E-08 3.3E-12I 2 .23E-06 2E-01 8 .8E+04 6 ~T 5.0DE-K30 9 .16E-09 -3 .9E-11 ! 2 .18E-06 3E-01 8.6EHH34 7 2 6 . 0 0 E ^ 1.10E-08 -3.4E-11 1.46E-06 3E-01 B.6E+04 8 2 7.00E+00 1.30E-08 4.2E-11 " 1.09E-D6 3E-C1 8 .6E+04 9 2 8.QQE-H30 1.60E-08 4.2E-111 1.06E-06 3E-01 8 . 6 E ^ 4 10 11 3 1.0DE-H30 3 .60E-09 fl.OEC 2 .5EE-06 2E-01 8 . B E - K M 12 3 2.0QE-KD0 3 .78E-09 •1 .5E-10: 2 .36E-06 2E-01 8 .6E+04 13 3 3 .00E+00 5.59E-Q9 -3.2E-11J 2 .23E-06 2E-01 8.6E-HD4 14 3 4 . 0 0 E 4 O 7 .36E-09 •4.3E-11 2 .18E-06 2E-01 8 .BE+04 15 3 5.0QE+00 9 .11E-09 •47E-11 1.46E-06 3E-01 8 .6E+04 16 3 6.00EHH30 1.10E-08 -1 .0E-12 1.09E-06 3E-01 8 . 6 E ^ 4 17 3 7 .00E+00 1.31E-08 5.9E-11! 1.06E-06 3E-01 8.6E-KD4 18 3 8 .0QE4O0 1.61E-0E 5.9E-11 1 . 0 6 E I B 3E-01 8.6E414 19 20 4 1 . 0 D E - K B 3 .60E-09 0 .OE4O0 2 .56E-06 2E-01 8 .5E- tC4 21 4 2 .00E+00 3.93E-09 1.5E-10 2 .36E-06 2E-01 8.5E-H34 2 2 ] 1 4 3 .00E4OQ 5.54E-09 •4.9E-11 2 .23E-06 2E-01 8.5E+04I 23 1 4 4 0 C E ^ 0 7 .32E-09 -4.2E-11 2 .18E-06 2 E - C " 8 . 5 E 4 0 4 24 4 5.0DE-K10 9.07E-OS 1.10E-0B -4.3E-11 1.46E-06 3E-01 8.5E+C4 25 4 6.DOE+00 -5 .5E-13 1.09E-06 3E-01 8.5E-HD4 2E 4 7.0DE+O0 1 3 2 E - 0 E 5.5E-11 1.06E-06 3E-01 8 .5E+04 27 1 4 8.00E-KD0 1.62E-08 5.5E-11 1.0BE-06 3E-01 8.5E+04 28 29 1 5 1 .ODE+00 3.60E-09 O.OE-rOO 2 .56E-06 2E-01 3 .5E+04 in 1 m n c u m 17cc nc 1 o c i n i s c e n e iz m Draw k Ayto5hapes' \ ^ • O H Q l H * * i Ready , Sf3/t QPhDThesisApp... Q C02Programl,., ^ P r DthesisApp... • 11 $ m % n \ ' P Friday 3 Document3 -... M Microsoft Exc, „ j M rrosoftVisu,.. • l»V. 12/01/2007 Appendices Page 243 Table C5 . Typical model simulat ion results spreadsheet • Microsoft Excel C o 2 b _ v 2 (1 )UpdatedlK2b -J File Edit View Insert Format Tools Data Window Help Acrobat 0£Sy§ §\k7 I I M - •--Infill B|ioo% ^ , J43 A B C D E F G H 1 J K L U N 0 1 New Concentration 2 1 2 3 4 5 6 7 B 9 10 11 3 0 0.36 0.36 0.368416 0.36 0.36 0,36 0.104148 0.36 0.36 0.3B 0.36 0.36 0.36 4 0.02 0.392039 0.497119 0 0.729202 2.00E-01 0.198531 0.219773 0.431767 0.431767 5 0.15 0.526496 0.547786 0.5B4218 0.624813 0.626809 0.593987 0.654541 0.66548 0.672806 0.655853 0.655853 B 0.3 0.B69966 0.713908 0.75623B 0.790302 0.B22791 0.852761 0.872405 0.891883 0.908654 0.924551 0.924551 7 0.45 0.852024 0.912509 0.97024 1.020683 1.063397 1.099365 1.130432 1.156605 1.178888 1.197704 1.197704 5 0.6 1.106526 1.173726 1,238455 1.294453 1.341775 1.381568 1.41537 1.443958 1.468257 1.488863 1.488863 9 0.75 1.40201 1.469232 1.534936 1.591763 1.639664 1.679847 1.713982 1.742813 1.76731 1.788067 1.788067 10 ' 0.9 1.70201 1.7B9232 1.834936 1.891763 1.939664 1.979847 2.013982 2.042813 2.06731 2.088067 2.088067 11 12 Diffusion 13 1 2 3 4 5 6 7 6 9 10 11 14 0 2.56E-06 2.81E-06 3.03E-06 3.23E-06 3.41 E-06 3.57E-06 3.7E-06 3.82E-06 3.93E-06 4.02E-06 4.02E-06 15 0.02 2.3BE-06 2.62E-06 2.87E-06 3.09E-06 3.28E-06 3.45E-06 3.6E-06 3.74E-06 3.85E-06 3.95E-06 3.95E-06 16 0.15 2.23E-06 2.5E-06 2.76E-06 2.99E-06 3.19E-06 3.37E-06 3.54E-06 3.68E-06 3.8E-06 3.91 E-06 3.91 E-06 17 0.3 2.18E-06 2.4EE-06 2.72E-06 2.95E-06 3.16E-06 3.35E-06 3.51 E-06 3.65E-06 3.78E-06 3.89E-06 3.89E-06 18 0.45 1.46E-06 1.78E-06 2.08E-06 2.:7E-0E 2.64E-06 2.88E-06 3.09E-06 3.29E-0B 3.46E-06 3.61 E-06 3.61E-06 19 0.6 1.09E-06 1.41E-06 1.73E-06 2.04E-06 2.33E-06 2.59E-06 2.84E-06 3.06E-0B 3.26E-06 3.43E-06 3.43E-06 20 0.75 1.06E-06 1.38E-06 1.7E-06 2.01 E-06 2.3E-06 2.57E-06 2.82E-06 3.04E-0B 3.24E-06 3.42E-06 3.42E-06 21 0.9 1.06E-06 1.38E-06 1.7E-05 1.01 E-06 2.3E-06 2.57E-06 2.82E-06 3.04E-0B 3.24E-06 3.42E-06 3.42E-06 22 Xi 0.0000045 0.000004 0.0000035 4 0.0000025 n o n / input I watercont / Temperature / TempDiff / output / Sheetl \results -Seriesl - Series2 Series3 - Series4 2.5 15 Draw Ready AutoShapes- \ \ D O | 4 'i Sfflft 3 PhDThesisApp... 3 C02Programl... ^ PhDtheasApp... 3 Document3-... j j Microsoft be. . . j Microsoft Visu... Appendices Page 244 APPENDIX D Waste-rock Sample Analyses Results D.1 Introduction This chapter presents the laboratory results of the waste-rock samples descr ibed in sect ion 2.2.6 of chapter 2. The data presented includes the results of: 1. Gra in s ize distribution curves for the waste-rock samp les from the D S W R and D N W R . 2. Soi l water characterist ic curves ( S W C C s ) for the samples from D S W R and D N W R . 3. Saturated hydraul ic conductivit ies for the samp les from the D S W R and D N W R . D.2. Grain-size distribution Table D1. Grain-size test with dispersing agent The particle-size analysis of the waste-rock samples was determined according to modified A S T M Designation: D 422-63. Results of the laboratory analysis are presented below: University of British Columbia Department of Mining Engineering CO-MIX Laboratory GRAIN SIZE TEST WITH DISPERSING AGENT Sample: TP01 Place: Depth: 0.3 - 0.4 m Date: 10-Dec- 03 Hygroscopic moisture Specific gravity # 2 mm Soil (g) = Appendices Page 245 Tare N° Pycnometer N° Tare (g) Temperature (°C) Tare + W S (fl) Pyc. + water (g) Tare + DS (g) Pyc.+water+soil(g) Moisture (%) G(g/cm 3 ) w(%): 1.09 G s . 2.74 Mass of air dried soil Mt (g) = 90.0 Total mass of dried sample Mdt (g) 89.0 Coarse screening Sieve Opening (mm) Retained (g) Total retained (g) % passing 2" 50.8 0.00 0.00 100.0 1 1/2" 38.1 0.00 0.00 100.0 1" 25.4 0.00 0.00 100.0 3/4" 19.1 0.00 0.00 100.0 3/8" 9.5 0.00 0.00 100.0 4 4.76 0.00 0.00 100.0 10 2.00 0.00 0.00 100.0 Fine Screening Sieve Opening (mm) Retained (%) Total retained (g) % passing 16 1.19 0.00 0.00 100.0 30 0.59 0.00 0.00 100.0 40 0.42 0.30 0.30 99.7 60 0.25 7.80 8.10 90.9 100 0.149 36.50 44.60 49.9 200 0.074 30.40 75.00 15.8 HYDROMETER Mass of wet soil submited to sedimentation Msw (g) 90.01 Hydrometer N ° : 863 Time time (s) temp. (°C) R (g/cm A3) Rh (g/cm A3) a (cm) Q S (%) d (mm) 30 s 30 18.0 1.014 1.0053 13.8 15.4 0.0719 1 min. 60 18.0 1.013 1.0053 14.1 13.6 0.0513 2 min. 120 18.0 1.013 1.0053 14.1 13.6 0.0363 4 min. 240 18.0 1.013 1.0053 12.8 13.6 0.0245 8 min. 480 18.0 1.012 1.0053 13.1 11.9 0.0175 15 min. 900 18.0 1.012 1.0053 13.1 11.9 0.0128 30 min. 1800 18.0 1.012 1.0053 13.1 11.9 0.0090 1 h 3600 18.8 1.012 1.0051 13.2 11.3 0.0063 2 h 7200 18.8 1.011 1.0051 13.3 10.4 0.0045 Appendices Page 246 4 h 14400 18.8 1.011 1.0051 13.3 10.4 0.0032 8 h 28800 18.8 1.011 1.0051 13.3 10.4 0.0023 24 h 86400 18.8 1.010 1.0051 13.6 8.6 0.0013 D (mm) % mat. Pass % mat.ret. MATERIAL* % of material 50.80 100.0 0.0 20<Coarse gravel<60 0.0 38.10 100.0 0.0 25.40 100.0 0.0 19.10 100.0 0.0 6,0<Median gravel<20,0 0.0 9.52 100.0 0.0 4.76 100.0 0.0 2,0<Fine gravel<6,0 0.0 2.00 100.0 0.0 1.190 100.0 0.0 0,60<Coarse sand<2,0 0.0 0.590 100.0 0.0 0,20<Median sand<0,6 29.4 0.420 99.7 0.3 0.250 90.9 9.1 0.149 49.9 50.1 0,06< Fine sand <0,20 57.3 0.074 15.8 84.2 0.0719 15.4 84.6 0,002 < Silt < 0,06 4.0 0.0513 13.6 86.4 0.0363 13.6 86.4 0.0245 13.6 86.4 0.0175 11.9 88.1 0.0128 1.1.9 88.1 0.0090 11.9 88.1 0.0063 11.3 88.7 0.0045 10.4 89.6 0.0032 10.4 89.6 0.0023 10.4 89.6 Clay < 0,002 9.3 0.0013 8.6 91.4 Cu = 300 Appendices Page 247 Grain Size (with dispersing agent) I U U . U - O A n yu.u - o n n OU.U - (U.U - .s in g n a 3 C  %  p a;  3 C  3 C  O A A OU.U - O A A / zu.u - 10.0 - 0.0 - 0.0 -4 >- • — —•- — i 4 010 0.0 100 0.1 Particles diameter (mm) 000 1.0000 Table D2. Grain-size test without dispersing agent 1 • Sample from DNWR: mesh Mm 9 % %retained Cum. Pas % 12.5 100 4 4.75 20 8.00 8 92.04 10 2 25 10.00 18.00 82.04 20 0.85 53.1 21.25 39.25 60.79 40 0.417 53.1 21.25 60.50 39.54 60 0.25 54.3 21.73 82.23 17.81 80 0.177 14.9 5.96 88.19 11.85 100 0.15 6.9 2.76 90.95 9.09 140 0.105 6.4 2.56 93.51 6.53 200 0.075 0.8 0.32 93.83 6.21 270 0.053 5.6 2.24 95.76 4.28 -270 -0.053 8.9 3.56 99.32 Total 250 100.04 Appendices Page 248 2. Sample from DSWR: mesh Mm g % %retained Cum. Pas % 12.5 100 4 4.75 3.1 1.24 1.24 98.80 10 2 14.6 5.84 7.08 92.96 20 0.85 34.6 13.85 20.93 79.11 40 0.417 56.2 22.49 43.42 56.62 60 0.25 60.1 24.05 67.47 32.57 80 0.177 33.8 13.53 80.99 19.05 100 0.15 15.9 6.36 87.35 12.69 140 0.105 13.8 5.52 92.88 7.16 200 0.075 5 2.00 94.88 5.16 -200 -0.075 12.9 5.16 100.04 Total 250 100.04 D.4 Soil Water Characteristic Curve (SWCC) Test Results 1 • Sample from DSWR: Sample D S W R cell 1660.9 Tare 4.1 diameter 6.9 final ht 4.460 cell+sample 1977.6 Tare+wet 53.6 ht 4.64 Dia 6.900 Tare+dry 51.8 ini. vol 173.5026 volume 166.7719 sample 316.7 water 9.4 Water 1.8 Tare 7.9000 soil 307.3 Soil 47.7 tare+wet 328.4800 w.c. 0.030589 w.c. 0.0377 tare+dry 314.2000 suction Weight w.c. final w.c. vol. w.c. %vol. W.c. water 14.2800 0.2 2027.1 0.1917 0.2105 0.3716 37.1637 Soil 306.3000 2026.4 0.5 2026.8 0.1907 0.2095 0.3699 36.9908 w.c. 0.0466 2007.4 1 2026.4 0.1894 0.2082 0.3676 36.7603 1997.3 2 2007.4 0.1276 0.1462 0.2581 25.8094 vol. Water 64.4800 1992.6 3 1997.3 0.0947 0.1132 0.1999 19.9882 vol. Soil 109.0226 1989.7 4 1992.6 0.0794 0.0979 0.1728 17.2793 1987.7 5 1989.7 0.0700 0.0884 0.1561 15.6078 porosity 0.3866 1986.2 6 1987.7 0.0635 0.0819 0.1446 14.4551 void ratio 0.5914 Appendices Page 249 1985.2 7 1986.2 0.0586 0.0770 0.1359 13.5906 1983.1 8 1985.2 0.0553 0.0737 0.1301 13.0142 Gs 2.8095 1980 10 1983.1 0.0485 0.0669 0.1180 11.8039 1978.8 30 1980 0.0384 0.0567 0.1002 10.0171 1977.6 50 1978.8 0.0345 0.0528 0.0933 9.3255 1976.9 80 1977.6 0.0306 0.0489 0.0863 8.6339 100 1976.9 0.0283 0.0466 0.0823 8.2304 D . Sample from DNWR: Sample D N W R Cell 1669.2 Tare 4.1 diameter 5 cell+sample 1960.8 Tare+wet 53.6 Ht 6.9 Tare+dry 51.8 ini. vol 186.964 Sample 291.6 Water 21.9 Water 1.8 6.9 final ht Tare 6.9000 Soil 269.7 Soil 47.7 4.64 dia tare+wet 303.5000 w.c. 0.081201 w.c. 0.0378 173.5026 volume tare+dry 276.5000 suction weight w.c. final w.c. vol. w.c. % vol. W.c. Water 27.0000 0.2 1990.4 0.1910 0.2311 0.3591 35.9072 Soil 269.6000 0.5 1989.5 0.1876 0.2277 0.3539 35.3885 w.c. 0.1001 1 1987.6 0.1806 0.2207 0.3429 34.2934 2 1982.2 0.1605 0.2007 0.3118 31.1811 vol. Water 62.3000 3 1975.4 0.1353 0.1754 0.2726 27.2618 vol. Soil 111.2026 4 1972.3 0.1238 0.1639 0.2548 25.4751 5 1970.5 0.1172 0.1573 0.2444 24.4377 porosity 0.3332 6 1968.5 0.1098 0.1499 0.2328 23.2850 void ratio 0.5602 7 1967.4 0.1057 0.1458 0.2265 22.6510 8 1966.4 0.1020 0.1421 0.2207 22.0746 Gs 2.4244 10 1964.9 0.0964 0.1365 0.2121 21.2101 2.2700 30 1960 0.0782 0.1183 0.1839 18.3859 50 1957.8 0.0701 0.1102 0.1712 17.1179 80 1956.1 0.0638 0.1039 0.1614 16.1381 100 1955.1 0.0601 0.1001 0.1556 15.5617 Appendices Page 250 D.5 Saturated Hydraulic Conductivity Test Results 1 • Sample from DSWR: U B C University of British Columbia Department of Mining Engineering CO-MIX Laboratory Falling Head Permeability Test FLUX U P W A R D Sample: PermDSWRI Golden Sunlight - Tailings Place: Area Depth: 31" - 4 3 " State: Loose - Hig Moist. % Date: 28-Apr-03 Mold No. Height (cm) Diameter (cm) Volumme (cm3) Weight (g) Cylinder 3 11.60 10.10 929.38 3850.7 Small Mold + Sample (g) = 5122.00 Gsample = 1.62 g/cm.3 Across (Tl2) - 0.00801 Lsamole ( m ) = 0.098 Temp, C : 23.0 Vsample (cm3) 785.0 Telaosed (min) h (cm) At (min) A; (%) K (m/s) kava (m/s) 0.000 96.0 - 0.257 51.0 0.257 1.01 E-04 0.000 96.0 - 0.8 0.259 51.0 0.259 1.00 E-04 0.000 96.0 - 1.01 E-04 0.257 51.0 0.257 1.01 E-04 0.000 96.0 - 0.0 0.257 51.0 0.257 1.01 E-04 Appendices Page 251 2, Sample from DNWR: U B C University of British Columbia Department of Mining Engineering CO-MIX Laboratory Falling Head Permeability Test F L U X U P W A R D Sample: PermDNWRI Golden Sunlight - Tailings Place: Area Depth: 3 1 " - 4 3 " State: Loose - Hig Moist. % Date: 28-Apr-03 Mold No. Height (cm) Diameter (cm) Volumme (cm3) Weight (g) Cylinder 3 11.60 10.10 929.38 3850.7 Small Mold + Sample (fl) = 5083.50 Gsample = 1.73 g/cm3 Across (m2) - 0.00801 Lsamole (m) = 0.089 Temp, C: 23.0 Vsample (cm3) 712.9 Teiaosed (min) h (cm) At (min) A t ( % ) K (m/s) kavc (mis) 0.000 96.0 - 0.233 51.0 0.233 1.01 E-04 0.000 96.0 - 9.9 0.256 51.0 0.256 9.19E-05 0.000 96.0 - 9.41 E-05 0.256 51.0 0.256 9.19E-05 0.000 96.0 - 0.0 0.256 51.0 0.256 9.19E-05 Appendices Page 252 APPENDIX E C 0 2 flux measurement results obtained at the Deilmann south (DSWR) and Deilmann north (DNWR) waste-rock piles using the dynamic closed chamber (DCC), static closed chamber (SCC) and eddy covariance (EC) methods Table E1. C 0 2 f luxes measurements obtained using the dynamic c losed chamber ( D C C ) at the Dei lmann south waste rock ( D S W R ) pile. Year 2000 Year 2002 Loc . # July August Sept. July August 1 215 225 204 162 132 2 290 202 182 3 284 143 102 218 187 4 292 291 178 191 5 300 154 137 142 190 6 350 246 200 123 7 356 173 104 136 8 274 180 121 204 132 9 182 202 179 131 115 10 192 182 190 145 11 250 169 203 129 12 247 13 234 116 134 209 14 121 189 213 15 368 164 297 16 224 203 254 288 17 58 96 18 91 113 19 185 20 106 89 144 89 Appendices Page 253 Table E2. C 0 2 f l u xes m e a s u r e m e n t s o b t a i n e d u s i n g the d y n a m i c c l o s e d c h a m b e r ( D C C ) at the D e i l m a n n north w a s t e rock ( D N W R ) pi le. Year 2000 Year 2002 Loc . # July A u g . Sept July A u g . 1 164 231 158 450 254 2 191 274 248 298 317 3 111 205 122 228 294 4 103 136 104 245 211 5 197 178 183 373 384 6 219 266 228 381 246 7 135 132 107 410 305 8 183 204 204 141 89 9 136 198 164 318 142 Table E3 C 0 2 flux measurements obtained using the static closed chamber (SCC) in the morning (between 10:00 and 11:00 on August 24, 2002 at nine selected sampling stations (DSF1 - DSF9).at the Deilmann south waste-rock (DSWR) pile (Figure 4.8A). Loc. AM1 AM2 Avg # mg m"2 h"1 1 226 174 200 3 185 185 4 248 248 5 369 260 314 6 125 125 7 139 139 8 151 151 9 146 146 10 129 139 134 11 164 164 13 20 175 175 Appendices Page 254 Table E4 C 0 2 flux measurements obtained using the static c losed chamber ( S C C ) in the afternoon (between 16:30 and 17:30) on August 24, 2002 at six se lected sampl ing stations ( D S F 1 - D S F 9 ) at the Dei lmann south waste-rock ( D S W R ) pile. Loc . PM1 PM2 A v g . # Mg rn"2 h"1 1 164 261 212.5 3 94.5 269 181.75 4 5 6 7 128 75 101.5 8 9 10 125 125 11 13 145 145 20 270 270 Table E5. Summary of C 0 2 f l u x measurements obtained using the static c losed chamber ( S C C ) in the morning (between 10:00 and 11:00) and afternoon (between 16:30 and 17:30) at nine selected sampl ing stations ( D S F 1 - D S F 9 ) at the Dei lmann south waste-rock ( D S W R ) pile on August 24, 2002. Loc . A M PM # mg m"2 h"1 1 200 213 3 185 182 4 248 5 315 6 125 7 139 102 8 151 9 146 10 134 125 11 164 13 145 20 175 270 Appendices Page 255 Table E6. Tempora l variations in the C 0 2 flux obtained at the Dei lmann south waste-rock ( D S W R ) pile using the Eddy covar iance (EC) method Measurements were obtained during the period from June 2 5 t h to August 2 5 t h 2002. E a c h data point represents the daily mean value averaged over the period from 10:00 to 17:00 hours. Mean C02 Flux C02 Flux Standard Deviation Julian Day Day mg m'z hr'1 mg m"i hr'1 176 25-Jun-02 122 51 177 26-Jun-02 143 67 178 27-Jun-02 134 63 179 28-Jun-02 157 53 180 29-Jun-02 103 50 181 30-Jun-02 182 01-Jul-02 183 02-Jul-02 106 71 184 03-Jul-02 132 73 185 04-Jul-02 186 05-Jul-02 104 37 187 06-Jul-02 100 43 188 07-Jul-02 176 74 189 08-Jul-02 96 49 190 09-Jul-02 118 45 191 IO-Jul-02 133 62 192 11-Jul-02 112 52 193 12-Jul-02 154 23 194 13-Jul-02 160 37 195 14-Jul-02 111 29 196 15-Jul-02 178 56 197 16-Jul-02 156 43 198 17-Jul-02 146 43 199 18-Jul-02 130 84 200 19-Jul-02 87 25 201 20-Jul-02 202 21-Jul-02 181 46 203 22-Jul-02 163 90 204 23-Jul-02 100 48 205 24-Jul-02 146 27 Appendices Page 256 206 25-Jul-02 107 20 207 26-Jul-02 101 48 208 27-Jul-02 96 25 209 28-Jul-02 132 46 210 29-Jul-02 211 30-Jul-02 212 31-Jul-02 213 01-Aug-02 214 02-Aug-02 107 44 215 03-Aug-02 118 36 216 04-Aug-02 119 50 217 05-Aug-02 149 42 218 06-Aug-02 145 17 219 07-Aug-02 102 42 220 08-Aug-02 78 40 221 09-Aug-02 111 75 222 10-Aug-02 105 55 223 11-Aug-02 85 29 224 12-Aug-02 151 48 225 13-Aug-02 226 14-Aug-02 227 15-Aug-02 228 16-Aug-02 79 22 229 17-Aug-02 139 63 230 18-Aug-02 113 55 231 19-Aug-02 152 92 232 20-Aug-02 103 34 233 21-Aug-02 . 93 53 234 22-Aug-02 77 52 235 23-Aug-02 67 23 236 24-Aug-02 104 60 237 25-Aug-02 101 44 Appendices Page 2 5 7 APPENDIX F Data for measurements of near- and surface-water contents and C 0 2 fluxes across the surfaces of the DSWR and DNWR after heavy rainfall events This section presents results of measurements of near- and surface-water contents (0 - 0.15 m) and associated C 0 2 fluxes from the DSWR and DNWR piles over an 8-d test period [30 July (day 1) to 6 August (day 8) 2002] after rainfall events. Waste- rock samples were collected each day during the test period at sampling stations DSF1 and DNF1 (Figure 4.8A) and analyzed for water contents within 24 hours. The gravimetric water contents were measured at 0, 0.05, 0.10, and 0.15 m depths. The gravimetric water contents were converted to volumetric water contents using the waste-rock properties measured in the laboratory (e.g., SWCC, soil specific density and porosity). The climatic parameters for the test site (e.g., rainfall and temperature) were recorded from the weather station installed on DSWR. Results of volumetric water contents, C 0 2 fluxes, rainfall events, and average daily temperatures for the DSWR and DNWR are presented in the Tables below.. Append i ces P a g e '258 Table F1. Water contents and CO2 f luxes measured over an 8-d test period [30 July (day 1) to 6 Augus t (day 8) 2002] at station DNF1 with time at the Dei lmann north waste-rock (DNWR) pile. Date Day # Temp. °C Rainfall (mm) co 2 Flux Depth Jul. 30 1 12.6 39.2 0 m 0.05 m 0.10 m 0.15 m Jul. 31 2 10.0 36.6 mg m"* h' 1 Water content (vol.) Aug. 01 3 11.8 7 7 0.2187 0.1267 0.1432 0.1463 Aug. 02 4 7.5 1 17 0.060 0.1237 0.1191 0.116131 Aug. 03 5 6.5 0.4 264 0.0211 0.1342 0.0950 0.0980 Aug. 04 6 8.4 0 268 0.0256 0.1010 0.0950 0.0950 Aug. 05 7 10.5 0 306 0.0045 0.0980 0.1161 0.1176 Aug. 06 8 13.2 0 316 0.0131 0.1110 0.0794 0.0829 Table F2. Water contents and C 0 2 f luxes measured over an 8-d test period [30 July (day 1) to 6 Augus t (day 8) 2002] at station D S F 1 with time at the Dei lmann south waste-rock ( D S W R ) pile. Date Day # Temp. °C Rainfall (mm) C 0 2 flux Depth Jul. 30 1 12.6 39.2 0 cm 5 cm 10 cm 15 cm Jul. 31 2 10.0 36.6 Mg m"z h"' Water content (vol.) Aug. 01 3 11.8 7 67 0.0571 0.0970 0.0913 0.0870 Aug. 02 4 7.5 1 97 0.0313 0.0785 0.0770 0.0870 Aug. 03 5 6.5 0.4 153 0.0010 0.0728 0.0699 0.0685 Aug. 04 6 8.4 0 138 0.0029 0.0599 0.0542 0.0613 Aug. 05 6 10.5 0 144 0.0017 0.0770 0.0285 0.0514 Aug. 06 8 13.2 0 241 0.00036 0.0728 0.0585 0.0499 Append i ces P a g e 259 APPENDIX G Minicosms data used for simulations with C 0 2 diffusion model This section presents measured column data (Kabwe et al., 2002) used for simulation with C 0 2 diffusion model developed in this thesis. The data were obtained from two minicosms: one kept at low temperature (LT) at about 5 °C and another one at room temperature (HT). The data presented include: water contents profiles, C 0 2 concentrations profiles, and temperatures profiles. G.1. HT Minicosm (column kept at room temperature) Table F1 . Temperature data from HT minicosm measured from Day 12 to Day 96 after filling the column with sand material. Day # Depth (m) D-12 19 26 34 47 75 96 0 9.00 9.02 9.05 9.01 9.02 8.86 8.11 0.02 7.57 7.71 7.71 7.62 7.42 7.21 7.49 0.3 8.34 8.47 8.42 8.35 8.13 8.34 7.74 0.6 8.80 8.91 8.88 8.80 8.65 8.90 8.08 0.9 9.37 9.46 9.44 9.35 9.24 -9.52 8.69 Appendices Page 260 T a b l e G2. Volumetric water contents from HT minicosm measured from Day 12 to Day 96 after filling the column with sand material Depth Day# (m) 11 18 25 32 46 73 95 0 18.29 18.29 18.29 18.29 18.29 18.29 18.29 0.15 22.56 21.03 21.39 21.85 21.32 22.91 24.12 0.30 24.87 23.74 21.43 25.04 21.99 25.63 26.65 0.45 25.89 23.53 24.93 26.61 23.55 29.16 29.69 0.60 27.59 26.38 25.93 28.35 25.12 28.44 29.52 0.7 36.60 39.85 36.65 40.91 25.73 39.98 43.14 0.90 49.80 43.99 44.63 47.12 46.82 48.53 51.31 105 49.52 47.87 46.93 47.58 47.57 48.06 52.46 Table G3. C 0 2 Concentration from HT minicosm measured from Day 1 to Day 96 after filling the column with sand material Depth Day # (m) D12 d19 d26 d33 d47 d75 d96 0 0.036 0.036 0.036 0.036 0.036 0.036 0.036 0.02 0.048 0.04 0.038 0.041 0.041 0.044 0.04 0.15 0.29 0.288 0.279 0.271 0 .263 0 .225 0.203 0.3 0 .323 0 .345 0.343 0.329 0.322 0.269 0.254 0.45 0 .422 0.437 0.51 0.452 0.424 0.371 0.337 0.6 0 .458 0.495 0.5 0.538 0.556 0.494 0.48 0.75 0 .605 0 .573 0.63 0.738 0.86 1.041 1.051 Appendices Page 261 G..2. LT Minicosm (sand column kept at low temperature ~5 °C) Table G4. Volumetric water contents from LT minicosm measured from Day 12 to Day 96 after filling the column with sand material depth Day # (m) 11 18 25 32 46 73 95 0 10.07 10.07 10.07 10.07 10.07 10.07 10.07 0.15 17.79 14.64 15.50 19.81 19.50 18.59 20.19 0.30 15.63 13.66 11.91 17.53 16.49 17.36 16.19 0.45 20.22 17.78 16.70 17.98 18.52 18.99 19.98 0.60 21.46 20.71 17.69 20.88 20.56 21.65 23.37 0.75 18.27 17.14 21.66 19.84 20.18 22.74 20.22 0.90 28.04 28.29 28.84 29.94 30.04 28.76 30.66 1.05 32.93 32.55 30.47 33.98 34.96 35.05 38.42 Table G5. C 0 2 concentrations from LT minicosm measured from Day 12 to Day 96 after filling the column with sand material depth Day# (m) Day 12 Day 19 Day 26 Day 34 Day 47 Day 75 Day 96 0 0.036 0.036 0.036 0.036 0.036 0.036 0.036 0.02 0.051 0.043 0.041 0.047 0.041 0.047 0.043 0.15 0.099 0.098 0.106 0.082 0.099 0.108 0.097 0.30 0.138 0.144 0.147 0.153 • 0.149 0.158 0.143 0.45 0.144 0.153 0.164 0.171 0.173 0.184 0.169 0.60 0.147 0.158 0.173 0.182 0.193 0.196 0.191 0.75 0.000 0.000 0.000 0.000 0.000 0.000 0 Appendices Page 262 APPENDIX H Climatic parameters used in simulations with SoilCover and recorded at the weather station installed on the Deilmann south waste-rock (DSWR) pile Simulat ion of evaporat ive f luxes [potential (PE) and actual (AE)] using So i lCover numerical model required the site weather parameters as inputs. The weather parameters used in simulat ions were recorded at a weather station installed on D S W R . The descript ion of the weather station was presented in chapter 4. The weather parameters used in the model simulat ions are presented in Tab le G1 Appendices Page 2 6 3 Tab le H1. Weather parameters recorded at the weather station installed on Dei lmann south waste-rock ( D S W R ) pile. Key Lake, 2002 T Ai r RH M A X RH MIN Net Rad Net Rad WIND Rain o C Pet. Pet. W/m2 M J m/s Mm 205 Jul-24 21.28583 88.7 39.01 91.9805 7.95 4.289375 0 206 Jul-25 21.22417 91.8 39.93 46.53471 4.02 2.788542 1.8 207 Jul-26 22.51708 96.1 39.06 60.10498 5.19 2.620271 0 208 Jul-27 17.78146 98.4 78.7 24.0446 2.08 3.058292 5.4 209 Jul-28 17.15208 100 58.27 70.62138 6.10 3.287375 7.7 210 Jul-29 13.12063 97.7 70.6 21.79919 1.88 5.67475 39.2 211 Jul-30 11.20646 98.9 .91.5 31.37233 2.71 6.197479 36.6 212 Jul-31 12.05354 100 83 58.67254 5.07 6.021208 7 213 Aug-01 8.072521 97.4 68.36 44.57667 3.85 8.522083 1 214 Aug-02 7.386729 96 66.55 66.0985 5.71 7.299021 0.4 215 Aug-03 10.02648 91.8 51.57 50.56073 4.37 5.679396 0 216 Aug-04 12.50308 88.8 44.88 50.86933 4.40 3.107708 0 217 Aug-05 14.59542 95.5 47.1 61.26404 5.29 5.039354 2.6 218 Aug-06 14.94688 99.8 87.8 20.31873 1.76 3.40325 13.3 219 Aug-07 19.24146 100 41.11 93.0589 8.04 3.4625 0 220 Aug-08 21.89958 98.1 31.3 86.83577 7.50 2.132104 0 221 Aug-09 22.04375 84.8 35.37 70.22992 6.07 3.077438 0 222 Aug-10 19.66188 77 36.82 65.12854 5.63 4.502417 0 223 Aug-11 14.67708 97.3 51.91 30.10281 2.60 3.830146 0 224 Aug-12 16.47563 90.5 38.84 66.93844 5.78 3.648688 0 225 Aug-13 14.16771 98.5 65.73 19.09198 1.65 2.913188 8.8 226 Aug-14 14.02771 99.7 62.58 48.55794 4.20 3.579667 2.9 227 Aug-15 12.06688 98 69.85 44.98344 3.89 2.106208 7.9 228 Aug-16 11.3225 99.1 81.9 22.15146 1.91 2.743583 7.3 229 Aug-17 11.96729 98.9 48.51 73.07313 6.31 3.738729 0 230 Aug-18 13.00967 95.7 28.92 54.86875 4.74 2.488333 0 231 Aug-19 11.39683 95.6 50.55 13.02202 1.13 4.382563 2.2 232 Aug-20 9.112875 87.3 37.3 59.66702 5.16 5.10575 0 233 Aug-21 15.07235 85.4 38.32 62.63533 5.41 3.510646 0 234 Aug-22 20.53417 72.6 38.08 56.01819 4.84 5.302063 0 235 Aug-23 20.70354 94.1 34.59 56.71417 4.90 2.354833 0 236 Aug-24 21.9425 76.8 25.21 60.14081 5.20 5.660375 0 Appendices Page 264 APPENDIX I SoilCover run summary page for simulations of evaporative fluxes at the DSWR and DNWR piles during the field tests The following pages present the daily input and output data and summary pages of So i lCover model simulat ions results of evaporat ive f luxes from the D S W R and D N W R piles obtained during the 8-d and 27-d test periods. The climatic parameters for input data were obtained from the weather station installed at the D S W R . The soil properties of the waste rocks were obtained from laboratory tests. Appendices Page 265 Tab le 1.1 Daily input data for So i lCover simulat ions for evaporat ive f luxes during the 8-d test period at the Dei lmann south waste-rock ( D S W R ) pile. Weather data section Moisture boundaries section Run .. . Max AilTfciT'D V ArTemp Net R=1G Max RH Mm RH '-; '.Windl - * : - : "lop BC Start time (MJ/m2-, dayi (dec) , (krtvriri . '(h'rs) '.hrsj 18.00 10.00 1.8834 0.98 0.71 1.41 3 39.2 0 24 15.50 11.00 2.7110 0.99 0.92 0.5918 3 36.6 0 24 11.00 5.00 5.069 1.00 0.83 0.793 3 7 0 24 4 10.00 3.00 3.8514 0.97 0.68 1.7391 3 1 0 24 13.00 3.50 5.7109 0.96 0.66 2.565 3 0.4 0 24 6 17.00 3.00 4.3685 0.92 0.52 4.4193 3 0 0 24 7 ' - 18.50 5.00 4.3951 0.88 0.45 5.6997 3 0 0 24 8 ' 18.50 5.00 4.3951 0.88 0.45 5.6997 3 0 0 24 Moisture Other daily data section boundaries section Bo; BC Bot • Temp Bot Temp Pan tvfip Write Day Root Roo: .. ', s. hjlii (Type) ( V a i u e ) 1111118 • irrT'day; Out • on. Bo; (CT/, 1 0.11 : I 1 1 0.087 H B 13 I 1 1 ' ~ 1 0.087 HlSPil 1 •SSI 1 0.068 • H 6 5 1 1 BBH \ 0.061 H B 8.25 1 1 ?':":T.Ti 1 1 0.051 i 10 1 1 0.05 11.7 • 1 • • • • M g n M H 1 0.05 1 11.7 1 1 i l i H i j H • Appendices Page 2 6 6 Tab le 12. Daily output data for So i lCover simulat ions for evaporat ive f luxes during the 8-d test period at the Dei lmann south waste rock ( D S W R ) pile. Elapsed Time days Pot Evap (mm) Act Evap (mm)" . Pot ' • Tran (mm) \ Act Tran " (mni) •^«Tot '•"-'lET. (mm) Water Bal (%) Spec Flux (mm) ' Bottom Flux (mm) 0 0 0 0 0 0 0 0 0 1 -0.72 -0.72 0 0 -0.72 -23.775 39.2 0.28 2 -0.745 -0.745 0 0 -0.745 -81.124 36.6 -0.068 3 -1.239 -1.239 0 0 -1.239 192.386 7 -0.13 4 -1.057 -1.057 0 0 -1.057 212.662 1 -0.137 5 -1.615 -1.615 0 0 -1.615 222.733 0.4 -0.266 6 -1.638 -1.468 0 0 -1.468 222.885 0 -0.747 7 -1.922 -1.189 0 0 -1.189 222.831 0 -0.909 8 -1.933 -0.981 0 0 -0.981 223.142 0 -0.92 . Runoff Selected ' Node Fix f '5 Net Infiltration Cum. . PE Cum. A E Cum. ' PT . Cum A T Cum. E T Cum. Precip: (mm) , (mm) r . (mni) (mm) (mm) (mm) (ram' (mm) (mm),1 0 0 0 0 0 0 0 0 0 22.606 0 15.873 -0.72 -0.72 0 0 -0.72 39.2 0 0 35.855 -1.465 -1.465 0 0 -1.465 75.8 0 0 5.761 -2.705 -2.704 0 0 -2.704 82.8 0 0 -0.057 -3.762 -3.762 0 0 -3.762 83.8 0 0 -1.215 -5.377 -5.377 0 0 -5.377 84.2 0 0 -1.468 -7.014 -6.844 0 0 -6.844 84.2 0 0 -1.189 -8.936 -8.034 0 0 -8.034 84.2 0 0 -0.981 -10.869 -9 015 0 0 -9.015 84.2 Cum. Cum. - Cum. Cum. / Runoff Infil. Bott Fl. int fix (mm) (mm) (mm) (mm) 0 0 0 0 22.606 15.873 0.28 0 22.606 51.728 0.212 0 22.606 57.489 0.082 0 22.606 57.432 -0.056 0 22.606 56.217 -0.322 0 22.606 54.749 -1.068 0 22.606 53.56 -1.978 0 22.606 52.579 -2.897 0 Appendices Page .26.7 Tab le 13. So i lCove r simulat ions summary for evaporat ive f luxes during the 8-d test period at the Dei lmann south waste rock ( D S W R ) pile. . SoilCover. V. 4.0] Run Summary Page i. .Project Name: 7. fhl&i Direct tin1: J. Rim hinuM'.ten*: 4 . : M e s l i l i i f n r m a i i m i : 0SWR2b .c;\scv4\ ' no 5. S o i l P r c p t \[\ H i H < •i I I'M1 i f > \ i ' n't i l l s | Oi 1 ,. ii', ' 'c' """"""" ~ ii T. Vi"'ii'i(iiiii v-stii'—'-v i , i ! ' ' , i f i ! 8. R. .r '.H*,'! ' v 32606 Appendices Page 268 Tab le 14. Daily input data for So i lCover simulat ions for evaporat ive f luxes during the 27-d test period at the Dei lmann south waste rock ( D S W R ) pile. Weather data section Moisture boundaries section Ri.i Dry Max A'lTenir V ArTemr Net Max RH RH W-no Speed T O D FJC BC . ,c, v.. -. : :. : • • (dec) (krr'nri ;Va.uej ITS' 1 11.50 9.00 1.8834 0.98 0.71 1.41 3 39.2 0 . 2 " 15.50 11.00 2.711 0.99 0.92 0.5918 3 36.6 0 24 11.00 5.00 5.069 1 0.83 0.793 3 7 0 24 4 \ 10.00 3.00 3.8514 0.97 0.68 1.7391 3 1 0 24 5 13.00 3.50 5.7109 0.96 0.66 2.565 3 0.4 0 24 17.00 3.00 4.3685 0.92 0.52 4.4193 3 0 0 24 18.50 5.00 4.3951 0.88 0.45 5.6997 3 0 0 24 8 18.50 5.00 5.29 0.96 0.47 5 3 2.6 0 24 9 17.00 11.00 1.76 1 0.88 3.4 3 13.3 0 24 24.00 12.00 8.04 1 0.41 3.5 3 0 0 24 11 26.00 13.50 7.5 0.98 0.31 2.1 3 0 0 24 12 28.00 8.00 6.07 0.85 0.35 3.1 3 0 0 24 13 26.50 11.00 5.63 0.77 0.37 4.5 3 0 0 24 21.00 7.00 2.6 0.97 0.52 3.8 3 0 0 24 22.00 3.50 5.78 0.91 0.39 3.6 3 0 0 24 16 --i 20.00 6.00 1.65 0.99 0.66 2.9 3 0 0 24 17 17.50 10.50 4.2 0.1 0.63 3.6 3 8.8 0 24 18 15.00 8.50' 3.89 0.98 0.7 2.1 3 2.9 0 24 12.50 9.00 1.91 0.99 0.82 2.7 3 7.9 0 24 20 15.50 6.00 6.31 0.99 0.49 3.7 3 7.3 0 24 21 17.00 1.00 4.74 0.96 0.3 2.5 3 0 0 24 22 15.00 7.00 1.13 0.96 0.51 4.4 3 0 0 24 23 14.00 0.00 5.16 0.87 0.37 5.1 3 2.2 0 24 ..: 21.00 0.00 5.41 0.85 0.38 3.5 3 0 0 24 25 24.50 16.00 4.84 0.73 0.38 5.3 3 • 0 0 24 26 26.50 10.00 4.9 0.94 0.35 2.3 3 0 0 24 27 26.00 15.00 5.2 0.77 0.25 5.7 3 0 0 24 Appendices Page 269 Moisture Other daily data section boundaries section Bot HC Bot BC fenp Bo- I imp Pan „'•• • Write" Day ' Root; , ' . • . • • .'(mrri/Sayi • i cm) 1 0.11 1 14 ŵ SBIiliillllllf • - 1 0.087 1 10 1 0.087 I H I 13 1 0.068 B I S 8 1 0.061 l l B i 6.5 ( H n R l i l 1 i . -i 0.051 • g i l l 8.25 • H 1 0.05 • 10 1 1 '-'.'";r?'.'i 1 0.05 11.7 1 1 0.05 i 15 • • • H I 1 0.05 . i - 15 H O N 1 " -' 1 ' 1 0.05 1 15 1 0.05 15 • B I B 1 0.05 i i i i i i i 15 B l l l l l l • 1 0.05 Bii i i t 15 1 0.05 15 • • S B ! 1 0.05 ••1 15 l l M B B i '- 1 • -' 1 0.05 111111 15 lliilllllll .1 I-'! 1 0.05 • B i l l 15 • K M 0 1 0.05 IBllIillB 15 1 0.05 15 1 1 0.05 i l l l l i 15 0 i i i i i i i 1 0.05 15 i i i i i i i 1 0.05 mmmm 15 • H I o UB1IB 1 0.05 i 15 ^ f i l l l l l o M M 1 0.05 ilSiili 15 iS&9Ni 1 0.05 15 l l N i 1 0.05 15 o Tab le 15. Daily output data for So i lCover simulat ions for evaporat ive f luxes during the 27-d test period at the Dei lmann south waste rock ( D S W R ) pile. Elapsed Time Pot Evap (mm) Act Evap • (mm) J. Act Trail (min) Tot E T (mm) : .Water Bal • (%) • Spec v Bottom (mm) (mm) 0 1 -0.633 -0.633 0 0 -0.633 -20.61 39.2 0.291 2 -0.745 -0.745 0 0 -0.745 -75.6 36.6 -0.084 3 -1.239 -1.239 0 0 -1.239 -186.932 7 -0.094 4 -1.057 -1.057 0 0 -1.057 -207.207 1 -0.14 5 -1.616 -1.616 0 0 -1.616 -217.28 0.4 -0.255 6 -1.638 -1.464 0 0 -1.464 -217.44 0 -0.686 7 -1.913 -1.174 0 0 -1.174 -217.384 0 -0.836 8 -1.99 -1.987 0 0 -1.987 -221.791 2.6 -0.901 Appendices Page 270 9 -0.571 -0.571 0 0 -0.571 -225.3 13.3 -1.084 10 -2.927 -2.814 0 0 -2.814 -225.433 0 -1.155 11 -2.91 -2.271 0 0 -2.271 -224.448 0 -1.192 12 -2.717 -1.294 0 0 -1.294 -224.843 0 -1.177 13 -2.781 M.197 0 0 -1.197 -224.213 0 -1.17 14 -1.282 -0.42 0 0 -0.42 -223.966 0 -1.107 15 -2.294 -0.731 0 0 -0.731 -224.154 0 -1.127 16 -0.862 -0.291 0 0 -0.291 -224.128 0 -1.085 17 -2.536 -2.522 0 0 -2.522 -235.63 8.8 -1.076 18 -1.307 -1.307 0 0 -1.307 -239.589 2.9 -1.085 19 -0.609 -0.609 0 0 -0.609 -243.004 7.9 -1.058 20 -2.046 -2.046 0 0 -2.046. -246.594 7.3 -1.068 21 -1.788 -1.788 0 0 -1.788 -246.637 0 -1.02 22 -0.827 -0.827 0 0 -0.827 -246.737 0 -1.021 23 -1.968 -1.968 0 0 -1.968 -250.723 2.2 -1.004 24 -2.22 -1.74 0 0 -1.74 -250.604 0 -1.056 25 -2.748 -1.347 0 0 -1.347 -250.551 0 -1.169 26 -2.248 -1.022 0 0 -1.022 -249.817 0 -1.178 27 -3.084 -0.903 0 0 -0.903 -249.701 0 -1.187 Runoff Selected Node -Net ' Cum. Cum. Cum. Cum. Cum. Cum. Fix Infiltration PE A E PT A T E T Precip. (nun) (mm) (mm) (mm) (mm) (mm) (mm) (mm), (mm) HHH3 0 0 0 0 0 0 0 0 23.438 0 15.129 -0.633 -0.633 0 0 -0.633 39.2 0 0 35.855 -1.378 -1.378 0 0 -1.378 75:8 0 0 5.761 -2.617 -2.617 0 0 -2.617 82.8 0 0 -0.057 -3.674 -3.674 0 0 -3.674 83.8 0 0 -1.216 -5.29 -5.29 0 0 -5.29 84.2 0 0 -1.464 -6.928 -6.754 0 0 -6.754 84.2 0 0 -1.174 -8.841 -7.928 0 0 -7.928 84.2 0 0 0.613 -10.831 -9.914 0 0 -9.914 86.8 0 0 12.729 -11.402 -10.485 0 0 -10.485 100.1 0 0 -2.814 -14.329 -13.299 0 0 -13.299 100.1 0 0 -2.271 -17.239 -15.57 0 0 -15.57 100.1 0 0 -1.294 -19.957 -16.864 0 0 -16.864 100.1 0 0 -1.197 -22.738 -18.062 0 0 -18.062 100.1 0 0 -0.42 -24.02. -18.481 0 0 -18.481 100.1 0 0 -0.731 -26.314 -19.213 0 0 -19.213 100.1 0 0 -0.291 -27.177 -19.504 0 0 -19.504 100.1 1.136 0 5.142 -29.713 -22.026 0 0 -22.026 108.9 0 0 1.593 -31.02 -23.333 0 0 -23.333 111.8 0 0 7.291 -31.629 -23.942 0 0 -23.942 119.7 0 0 5.254 -33.675 -25.988 0 0 -25.988 127 0 0 -1.788 -35.463 -27.776 0 0 -27.776 127 0 0 -0.827 -36.29 -28.603 0 0 -28.603 127 0 0 0.232 -38.259 -30.571 0 0 -30.571 129.2 0 0 -1.74 -40.479 -32.312 0 0 -32.312 129.2 Appendices Page 271 0 0 -1.347 -43:227 -33.659 - 0 0 -33.659 129.2 0 0 -1.022 -45.475 -34.681 0 0 -34.681 129.2 0 0 -0.903 -48.559 -35.584 0 0 -35.584 129.2 Cum. ' Cum. Cum. ', - Cum. Runoff (mm) Infil. (mm) Bott FI. (mm) ' int fix (mm) 0 0 0 0 23.438 15.129 0.291 0 23.438 50.985 0.207 0 23.438 56.745 0.113 0 23.438 56.688 -0.027 0 23.438 55.473 -0.282 0 23.438 54.008 -0.968 0 23.438 52.835 -1.805 0 23.438 53.448 -2.705 0 23.438 66.177 -3.789 0 23.438 63.363 -4.944 0 23.438 61.092 -6.136 0 23.438 59.798 -7.313 0 23.438 58.601 -8.483 0 23.438 58.181 -9.59 0 23.438 57.45 -10.716 0 23.438 57.158 -11.802 0 24.573 62.301 -12.877 0 24.573 63.894 -13.962 0 24.573 71.185 -15.02 0 24.573 76.439 -16.088 0 24.573 74.651 -17.109 0 24.573 73.824 -18.129 0 24.573 74.055 -19.133 0 24.573 72.315 -20.189 0 24.573 70.967 -21.357 0 24.573 69.946 -22.536 0 24.573 69.043 -23.722 0 Appendices Page 272 Tab le 16. So i lCove r simulat ions summary for evaporat ive f luxes during the 27-d test period at the Dei lmann south waste rock ( D S W R ) pile. SoilCover V.- 4:0! Hun Summary P a a f ! ' ' ! l | l v l Nil!) 1 I ' m i a l D i m in \ i. Run I ' l t n i m e s m : 4. Mesl i ] i i i ' i i i 'm; i l io»: t >scv4\ 5, Soil I'liilM I i N " M M 6, ROHIMIHI'V Ciiiiflisioiis ' ' 1 • , : / ' C ' ' i ii •> OP •• > - -- - n > < _ Vtffl! Ill*)*' M l l I M M ' :H.- Run O i ' i i i i ' s « i ' i t ' . t i 1 Appendices Page 273 Tab le 17. Daily input data for So i lCover simulat ions for evaporat ive f luxes during the 8-d test period at the Dei lmann north waste rock ( D S W R ) pile. Weather data section Moisture boundaries section Run Day AirTsrrp AiiTemp N&t - Max RH • Speed I 03 BC Top Start • • (C) V GiT/} (dec) • (his. • 1 18.00 1 0 . 0 0 • 1.8834 0.98 0.71 1.41 3 39.2 0 24 18.00 10.00 1.8834 0.98 0.71 1.41 3 36.6 0 24 3 11.50 9.00 2.7106 0.99 0.91 0.5918 3 7 0 24 4 15.50 9.00 5.0693 1 0.83 0.793 3 1 0 24 5 11.00 5.00 3.8514 0.97 0.68 1.7391 3 0.4 0 24 '6'""'' 10.00 3.00 5.7109 0.96 0.67 2.565 3 0 0 24 7 13.00 3.50 4.3685 0.92 0.52 4.4193 3 0 0 24 8 17.00 3.00 4.3951 0.89 0.45 5.6997 3 0 0 24 Moisture Other daily data section boundaries section Got BC Bet BC Toe Temp Bot TtniD rj=n Evap Day Root Root Blllllii!llll •. alue) i>- lC) <C: • • w i l t !;(mm/day.)'.'- Out ccrri •• • ;m)3 1 0.3 • 14 1 1 Wamm flHllll 1 0.3 13 I l l i l B 1 IllISlBI • H i l l 1 0.23 ••1.-' ;. 8 • • • • 1 1118111 • • M i l 1 0.146 6.5 1 iRll l l l l l l l l l l 1 0.116 A 8.25 1 1 ~.J&\ *' 1 0.098 1 ' 10 I B i l l l 1 ' .1 ' i l l f l l i t 1 0.095 1- - 11.7 l l l l l B I 1 1 .1 0.118 ' • " i l l 10 • H 1 Appendices Page 274 Tab le 18. Daily output data for So i lCover simulat ions for evaporat ive f luxes during the 8-d test period at the Dei lmann north waste rock ( D S W R ) pile. Elapsed Time days ; P O t Evap (mm) , Act Evap (mm) Pot, ' Act Tran Tran (mm) (mm). 1 " ' E T -. (mm) Water Bal (%) Spec Flux (mm) Bottom , Flux (mm) 0 0 0 0 0 0 0 0 0 1. -0.72 -0.72 0 0 -0.72 -23.775 39.2 0.28 2 -0.745 -0.745 0 0 -0.745 -81.124 36.6 -0.068 3 -1.239 -1.239 0 0 -1.239 192.386 7 -0.13 4 -1.057 -1.057 0 0 -1.057 212.662 1 -0.137 5 -1.615 -1.615 0 0 -1.615 222.733 0.4 -0.266 6 -1.638 -1.468 0 0 -1.468 222.885 0 -0.747 7 -1.922 -1.189 0 0 -1.189 222.831 0 -0.909 8 -1.933 -0.981 0 0 -0.981 223.142 0 -0.92 Runoff (mm) Selected Node Fix (mm) , Net Infiltration (mm) Cum. PE (mm) Cum. * . A E (mm) Cum. ' . pt- :S • (mm) Cum. • A T (mm) Cum. E T (mm) Cum. Precip. (mm) ' 0 0 0 0 0 0 0 0 0 22.606 0 15.873 -0.72 -0.72 0 0 -0.72 39.2 0 0 35.855 -1.465 -1.465 0 0 -1.465 75.8 0 0 5.761 -2.705 -2.704 0 0 -2.704 82.8 0 0 -0.057 -3.762 -3.762 0 0 -3.762 83.8 0 0 -1.215 -5.377 -5.377 0 0 -5.377 84.2 0. 0 -1.468 -7.014 -6.844 0 0 -6.844 84.2 0 0 -1.189 -8 936 -8.034 0 0 -8.034 84.2 0 0 -0.981 -10.869 -9.015 0 0 -9.015 84.2 Cum. • Cum. v , / Cum. , Cum. Runoff Infil. Bott.FI. int fix (mm) . (mm) (mm) (mm) 0 0 0 0 22.606 15.873 0.28 0 22.606 51.728 0.212 0 22.606 57.489 0.082 0 22.606 57.432 -0.056 0 22.606 56.217 -0.322 0 22.606 54.749 -1.068 0 22.606 53.56 -1.978 0 22.606 52.579 -2.897 0 Appendices Page 275 Tab le 19. So i lCove r simulat ions summary for evaporat ive f luxes during the 8-d test period at the Dei lmann south waste rock ( D N W R ) pile. 1; Hnjjw.t Name: 2, i ' r o j i x ! U i r i ' d o r y : I Rui) I' ,i i H'ti 4, MuMi Immm m « » 5,".Soil Pi i| n% Nii i ' in ii \ ; <> Run Swumaryfage• n,\AKP1c c : » s c v 4 i i B< " i! i n (-it ii>tiitih h.. Hiiii.Out|Hii stimman': >' '.0'.' I .,> ! i l ( I lM ! - i I 1 i f)1 , U ~ _ 1 \ - r 1 T _ ,4 ' i i 1 i ! 1 ! ! I 1 P, • - v ~ user iJt \ •JXfA.m 1 1.SGE-03 OS Append i ces P a g e 276 Table 110. Daily input data for So i lCover simulat ions for evaporat ive f luxes during the 27-d test period at the Dei lmann nortth waste rock ( D N W R ) pile. Weather data section Moisture boundaries section Max A.rlerr'B Aulen-.p Net Max RH RH Speed Top BC Top ••• S:a-1 l ire time • !C> aay; idee) idee"' ikir-'ii) Hype) (Value)' ' (hrs) 1 18.00 10.00 1.8834 0.98 0.71 1.41 3 39.2 0 24 2 18.00 10.00 1.8834 0.98 0.71 1.41 3 36.6 0 24 11.50 9.00 2.7106 0.99 0.91 0.5918 3 7 0 24 15.50 9.00 5.0693 1 0.83 0.793 3 1 0 24 • 11.00 5.00 3.8514 0.97 0.68 1.7391 3 0.4 0 24 ' 6 10.00 3.00 5.7109 0.96 0.67 2.565 3 0 0 24 7 13.00 3.50 4.3685 0.92 0.52 4.4193 3 0 0 24 8 17.00 3.00 4.3951 0.89 0.45 5.6997 3 2.6 0 24 9 ' 17.00 11.00 1.76 1 0.88 3.4 3 13.3 0 24 . 10 18.50 12.00 8.04 1 0.41 3.5 3 0 0 24 • 11 17.00 13.50 7.5 0.98 0.31 2.1 3 0 0 24 12 24.00 8.00 6.07 0.85 0.35 3.1 3 0 0 24 13 26.00 11.00 5.63 0.77 0.37 4.5 3 0 0 24 14 28.00 7.00 2.6 0.97 0.52 3.8 3 0 0 24 15 26.50 3.50 5.78 0.91 0.39 3.6 3 0 0 24 16 21.00 6.00 1.65 0.99 0.66 2.9 3 0 0 24 17 22.00 10.50 4.2 0.1 0.63 3.6 3 8.8 0 24 • 20.00 8.50 3.89 0.98 0.7 2.1 3 2.9 0 24 19 17.50 9.00 1.91 0.99 0.82 2.7 3 7.9 0 24 20' ' ' 15 6.00 6.31 0.99 0.49 3.7 3 7.3 0 24 ;21 12.50 1.00 4.74 0.96 0.3 2.5 3 0 0 24 22 15.5 7.00 1.13 0.96 0.51 4.4 3 0 0 24 23 17.00 0.00 5.16 0.87 0.37 5.1 3 2.2 0 24 24 15.00 0.00 5.41 0.85 0.38 3.5 3 0 0 24 25 14.00 16.00 4.84 0.73 0.38 5.3 3 0 0 23 26. ' . 21.00 10.00 4.9 0.94 0.35 2.3 3 0 0 24 24.50 15.00 5.2 0.77 0.25 5.7 3 0 0 24 Moisture Other daily data section boundaries section Bot BC Bot BC I IsP Bot I enp Par Lvac • n i l Roo: koct Type ;C) ililiiiliillilBl imm'dayi P S I M B M I Bot.(cm) • 1 0.3 '1 14 • 1 IlilB • 1 0.3 1 13 • 1 -!iV"-:. 1 0.23 1 8 1 1 .!..',t. S M I B I M 1 0.146 1 6.5 1 I H l l l l 1 0.116 1 , 8.25 1 • Appendices Page 277 1 0.098 • 10 • 1 H H H H 1 0.095 ' I 1'';' 11.7 1 i l i l f i • 1 0.118 1 • 10 llltllttil 1 1 0.05 '?: 1 15 ' 1 • 1 - 0.05 -. 1>. 15 1 i 1I111S1 1 0.05 t „ 15 i 1 BMm 1 0.05 15 ffijjjfjlllj 1 lllfliliBilliplli 1 0.05 15 1 l l l l l i 1 0.05 1 15 mBmni 1 I B i t y i l i - 1 0.05 ' .1-'' 15 IHHRII 1 1 0.05 • -1'. 15 0 BlBill 1 0.05 :. 1 , 15 0 B11IBI JilillBl 1 0.05 . 1 15 1 0 lllllilill ''-1 . - 1 0.05 1 "• 15 0 1 - iliiiiiij 1 0.05 £?:1 :• 15 1 1 0.05 HH9B 15 IRiiSBi 0 • l i l t . 1 0.05 15 K S H M 0 1 0.05 n m 15 M H I 0 H E f l i 1 0.05 15 • • • • 0 HHB I N H i i 1 0.05 15 • B s l l i 1 lliillli 1 8 H B 1 1 0.05 • 1 15 1 , . . . 1 0.05 1 15 ""v.T-.t' 1 , • 1 i Tab le 111. Daily output data for So i lCover simulat ions for evaporat ive f luxes during the 27-d test period at the Dei lmann north waste rock ( D N W R ) pile. ' Elapsed Pot Act Pot Act Tot , Water Spec Bottom Time days Evap (mm)- ' Evap v (mm) Tran (mm) Tran (mm) E T ' (mm) •-' Bal (%) Flux * (mm) , Flux (mm) 0 0 0 0 0 0 0 0 0 1 -0.735 -0.735 0 0 -0.735 -7.323 39.2 -11.85 2 -0.727 -0.727 0 0 . -0.727 -9.52 36.6 -35.624 3 -0.691 -0.691 0 0 -0.691 -12.549 7 -16.112 4 -1.38 -1.38 0 0 -.1.38 -19.872 1 -8.056 5 -1.096 -1.096 0 0 -1.096 -22.03 0.4 -3.075 6 -1.496 -1.496 0 0 -1.496 -22.851 0 -1.683 7 -1.525 -1.069 0 0 -1.069 -23.114 0 -0.854 8 -1.825 -1.825 0 0 -1.825 -23.299 2.6 0.603 9 -0.579 -0.579 0 0 -0.579 -33.409 13.3 -0.072 10 -2.708 -2.708 0 0 -2.708 -47.271 0 -0.163 11 -2.553 -1.681 0 0 -1.681 -59.167 0 -0.177 12 -2.573 • -1.254 0 0 -1.254 -64.431 0 -0.152 13 -2.774 -0.687 0 0 -0.687 -67.107 0 -0.122 14 -1.419 -0.536 0 0 -0.536 -68.718 0 -0.095 15 -2.447 -0.594 0 0 -0.594 -69.565 0 -0.074 16 -0.877 -0.295 0 0 -0.295 -70.247 0 -0.055 17 -2.676 -2.664 0 0 -2.664 -70.985 8.8 -0.008 18 -1.415 -1.415 0 0 -1.415 -72.585 2.9 -0.005 19 -0.669 -0.668 0 0 -0.668 -78.405 7.9 -0.063 20 -2.027 -2.027 0 0 -2.027 -98.202 7.3 -0.1 Appendices Page 278 21 -1.629 -1.629 0 0 -1.629 110.986 0 -0.165 22 -0.836 -0.835 0 0 -0.835 120.726 0 -0.168 23 -2.087 -2.087 0 0 -2.087 -127.48 2.2 -0.084 24 -1.991 -1.313 0 0 -1.313 129.965 0 -0.107 25 -2.363 -0.971 0 , o -0.971 131.853 0 -0.105 26 -2.084 -0.561 0 0 -0.561 132.993 0 -0.086 27 -3.03 -0.498 0 0 -0.498 133.788 0 -0.067 Runoff Selected Net Cum. Cum. Cum. Cum. . Cum. Cum. Node Fix Infiltration PE A E PT A T E T Precip. (mm) (mm) •„..; (mm) . (mm) (mm) (mm) (mm) (mm) (mm) 0 0 0 0 0 38.465 -0.735 -0.735 0 0 -0.735 39.2 0 0 35.873 -1.461 -1.461 0 0 -1.461 75.8 0 0 6.309 -2.152 -2.152 0 0 -2.152 82.8 0 0 -0.38 -3.532 -3.532 0 0 -3.532 83.8 0 0 -0.696 -4.628 -4.628 0 0 -4.628 84.2 0 0 . -1.496 -6.124 -6.124 0 0 -6.124 84.2 0 0 -1.069 -7.649 -7.193 0 0 -7.193 84.2 0 0 0.775 -9.474 -9.018 0 0 -9.018 86.8 0 0 12.721 -10.053 -9.597 0 0 -9.597 100.1 0 0 -2.708 -12.761 -12.304 0 0 -12.304 '100.1 0 0 -1.681 -15.314 -13.985 0 0 -13.985 100.1 0 0 -1.254 -17.887 -15.239 0 0 -15.239 100.1 0 0 -0.687 -20.662 -15.926 0 0 -15.926 100.1 0 0 -0.536 -22.081 -16.461 0 0 -16.461 100.1 0 0 -0.594 -24.528 -17.056 0 0 -17.056 100.1 0 0 -0.295 -25.405 -17.351 0 0 -17.351 100.1 0 0 6.136 -28.081 -20.015 0 0 -20.015 108.9 0 0 1.485 -29.496 -21.43 0 0 -21.43 111.8 0 0 7.232 -30.165 -22.098 0 0 -22.098 119.7 0 0 5.273 -32.191 -24.125 0 0 -24.125 127 0 0 -1.629 -33.821 -25.754 0 0 -25.754 127 0 0 -0.835 -34.656 -26.59 0 0 -26.59 127 0 0 0.113 -36.743 -28.677 0 0 -28.677 129.2 0 0 -1.313 -38.735 -29.99 0 0 -29.99 129.2 0 0 -0.971 -41.098 -30.961 0 0 -30.961 129.2 0 0 -0.561 -43.181 -31.522 0 0 -31.522 129.2 0 0 -0.498 -46.211 -32.02 0 0 -32.02 129.2 - Cum. Runoff • Cum. . Bott Fl. Cum: int fix Appendices Page 279 (mm) (mm) (mm)''.' r(mm)-,C 0 0 0 .... o.. 0 38.465 -11.85 . o 0 74.339 -47.474 • 0' 0 80.648 -63.586 0 0 80.268 -71.643 0 0 79.572 -74.718 0 0 78.076 -76.4 0 0 77.007 -77.254 0 0 77.782 -76.652 0 0 90.503 -76.724 0 0 87.796 -76.887 0 0 86.115 -77.064 0 0 84.861 -77.216 0 0 84.174 -77.337 0 0 83.639 -77.432 0 0 83.044 -77.506 0 0 82.749 -77.561 0 0 88.885 -77.569 0 0 90.37 -77.575 0 0 97.602 -77.637 0 0 102.875 -77.738 0 0 101.246 -77.903 0 0 100.41 -78.071 0 0 100.523 -78.155 0 0 99.21 -78.262 0 0 98.239 -78.367 0 0 97.678 -78.453 0 0 97.18 -78.52 0 Appendices Page 280 Tab le 112. So i lCove r simulat ions summary for evaporat ive f luxes during the 27-d test period at the Dei lmann north waste rock ( D N W R ) pile. SoilCover V. 101 Run Summary Pace l:Pmj<:d;iV;iiii!-: 2. Project D i rectory : 3. : Run Pitrifmeters: 4. IVteh Infill o t . !» ' DNWRP4x c:\scv4\ 5. .Soil Pn'ihiu suiin . | „ , , , , , 6. Bfluiiilarv (.imdiiimi* M M l l ' 1, Vfiill Hit " Ii " Ii* X. Run Out >i  s mi * ii • | . i , i I i >i 11. ii feUiiimiiiilive E' s i l l I I ..!) Nftf-uniBtavcAFiiiiiiiJ'... 'K i i v in i i i i i i i ) ' II User Kcste ii Oi-Apr-88 Ussr

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 10 0
United Kingdom 6 0
Japan 1 0
Canada 1 0
Germany 1 0
France 1 0
City Views Downloads
Unknown 6 0
Washington 4 0
Sheffield 2 0
Mountain View 2 0
Saskatoon 1 0
Youngstown 1 0
Tokyo 1 0
Charlotte 1 0
Tulsa 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items