Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanisms and spatial variability of rainfall infiltration on the Claude waste rock pile Bellehumeur, Tracy M. 2001

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

Item Metadata

Download

Media
831-ubc_2001-0152.pdf [ 18.22MB ]
Metadata
JSON: 831-1.0089913.json
JSON-LD: 831-1.0089913-ld.json
RDF/XML (Pretty): 831-1.0089913-rdf.xml
RDF/JSON: 831-1.0089913-rdf.json
Turtle: 831-1.0089913-turtle.txt
N-Triples: 831-1.0089913-rdf-ntriples.txt
Original Record: 831-1.0089913-source.json
Full Text
831-1.0089913-fulltext.txt
Citation
831-1.0089913.ris

Full Text

MECHANISMS AND SPATIAL VARIABILITY OF R A I N F A L L INFILTRATION ON THE CLAUDE WASTE R O C K PILE  by  TRACY M. BELLEHUMEUR  B.A.Sc, University of Waterloo, 1996  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Earth and Ocean Sciences  We accept the thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA March, 2001 © Tracy M. Bellehumeur, 2001  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  Enril  ,W  QLe&«SutJiu.}  The University of British Columbia Vancouver, Canada  Date Mor-di H , Uoi  Abstract The transport of acid rock drainage from a highly heterogeneous and unsaturated waste rock pile is based on an understanding of fluid flow within a pile. This study provides a quantitative evaluation of spatial variability in rainfall infiltration and identifies relations between hydrologic pathways, contact areas and structural features in the Claude waste rock pile. Thirty-five fixed ring infiltrometer tests were conducted to measure infiltration capacity over the surface of the pile. Hydrologic pathways into the Claude pile were investigated by releasing dye staining tracers at the surface, followed by excavation and mapping of the stained waste rock. The west side of the pile is characterized by a coarse, 'untrafficked' surface, in which infiltration at the surface is immediate. Recharge on the east side of the pile is controlled by a fine traffic compacted material and surface topography. Intense rainstorms cause ponding and overland flow to concentrate in coarse drains at the bottom of small catchment areas. In an intense rainstorm, the majority of recharge is through the drains on Claude east. In a typical low intensity rainstorm, the majority of recharge is through the fine matrix material on Claude east. Following the infiltrometer test, a volume of dye tracer, rhodamine WT, was ponded in the ring infiltrometers. Excavation provided a visual record of Darcy flow and flow through macropores. Dye distribution decreased rapidly and converged in macropores near the surface. Three large-scale dye tracer releases (800L to 2400L) were applied at the catchment drains on Claude east and distributed over a small area on Claude west. Excavation of the tracer revealed that infiltrating water migrates over the surface of coarse particles, avoiding the fine matrix material (boulder hopping). Significant lateral spreading was caused by ponding on old traffic surfaces encountered below the pile surface. Two mechanisms of flow were observed during excavation of the large-scale tracer tests, preferential flow in macropores, and boulder hopping. Darcy flow was observed in the fine, traffic compacted layers. This thesis documents the strong influence of macropore flow and preferential flow on the distribution of rainfall recharge into the Claude pile.  Table of Contents Abstract  ii  List of Tables  v  List of Figures  vi  Acknowledgements 1. Introduction  viii 1  1.1 Objectives of the study  2  1.2 Site description  4  1.3 Overview of Thesis  5  2. Literature R e v i e w - Understanding infiltration in coarse waste rock  5  2.1 Infiltration Theory  6  2.2 Factors Affecting Infiltration - Crust-Topped Soils, Compaction, Macropores  7  2.3 Spatial variability of infiltration  8  2.4 Field studies on flow through waste rock 3. F i e l d and Laboratory Methods  10 10  3.1 Runoff Experiment  10  3.2 Fixed Ring Infiltrometer Testing  10  3.3 Small volume dye tracers in ring infiltrometers  12  3.4 Physical Properties  13  3.4.1 Density and Porosity Measurements  13  3.4.2 Constant head permeameter - saturated hydraulic conductivity  13  3.4.3 Moisture content  14  3.4.4 Surface texture  15  4. Rainfall Infiltration  16  4.1 Density, porosity and saturated hydraulic conductivity  16  4.2 Moisture content  16  4.3 Surface texture  17 iii  4.4 Infiltration measurements  18  4.4.1 Catchment 2  18  4.4.2 Catchment 8  20  4.4.3 Claude West  20  4.4.4 Summary o f infiltrometer measurements  21  4.5 Small volume dye tracers in ring infiltrometers  22  4.6 Summary of small volume dye tracers in ring infiltrometers  23  4.7 Limitations of infiltrometer testing  24  4.8 Rainfall partition between infiltration and runoff on Claude east  25  4.8.1 Direct measure o f runoff in Catchment 7  25  4.8.2 Predictions o f rainfall partition in Catchment 2 and 8 o f Claude east  27  4.9 Summary of rainfall distribution  5. Large-Scale D y e Tracer Tests  29  31  5.1 Field methods  31  5.2 Waste rock texture and moisture below the pile surface  32  5.3 Results of Large-Scale Tracer Tests  33  5.3.1 Tracer test in Catchment 1 on Claude east  33  5.3.2 Tracer test in Catchment 2 on Claude east  34  5.3.3 Distributed tracer test on Claude west  35  5.4 Summary of large-scale dye tracer tests  36  6. Conclusions  38  References  40  A p p e n d i x 1: Rainfall record at C l u f f Lake for M a y to August, 1998  43  A p p e n d i x 2: Rainfall statistics for extreme events at C l u f f Lake  44  A p p e n d i x 3: Evaluation o f visual dye tracers  45  A p p e n d i x 4: G r a i n size analysis o f Claude pile a) surface samples; b) below-surface samples  46  A p p e n d i x 5: U S D A soil textures to wetting front suction  48  iv  List of Tables  Table 4 - 1 : Infiltration parameters estimated from infiltrometer testing in Catchment 2. Table 4-2: Infiltration parameters estimated from infiltrometer testing in Catchment 8. 5°  Table 4-3: Statistical properties o f steady state infiltration and sorptivity. Table 4-4: Summary o f measured runoff to the drain o f Catchment 7.  Table 4-5: Summary o f predicted distribution o f recharge to catchment surface and drains in a rainfall event. 51  Table 5-1: Summary o f field methods for large-scale dye tracer tests.  52  Table 5-2: Summary o f observations o f large-scale dye tracer tests.  $Z  v  List of Figures 5^  Figure 1-1: A e r i a l photo o f the Claude waste rock pile. Figure 1-2: Claude waste pile after contouring. Claude east, west and six experimental sites.  &H  Figure 1-3: Difference in surface texture on the Claude waste rock pile, a) Traffic-compacted Claude east and b) coarse Claude west.  ** **  Figure 1-4: Catchment and drain on Claude east during two-year return storm in August o f 1997.  & (p  Figure 1-5: R a i n water in a ponded state on Claude east after a rainfall. Figure 2-1: Conceptual diagram o f steady state infiltration and sorptivity.  5?  Figure 3-1: R u n o f f experiment in Catchment 7.  ^  Figure 3-2: F i x e d ring infiltrometer experiment set-up.  fe  0  Figure 3-3: Infiltrometer installed in driving path on Claude west.  '  Figure 3-4: Constant head permeameter experiment set-up.  ^ ^  Figure 4 - 1 : Cumulative infiltration in Catchment 2 o f Claude east, a) L o n g time steady state behavior and b) early-time, capillarity-dominated flow.  fe  3  Figure 4-2: Frequency histogram for infiltration parameters measured in Catchment 2. a) Steady state infiltration and b) field sorptivity.  k ^ (o5  Figure 4 - 3 : Infiltrometer test 1 - Catchment 2 o f Claude east. Figure 4-4: Cumulative infiltration in Catchment 8 o f Claude east, a) L o n g time steady state behavior and b) early-time, capillarity-dominated flow.  (0(0  Figure 4-5: Frequency histogram for infiltration parameters measured in Catchment 8. a) Steady state infiltration and b) field sorptivity.  t>^"  Figure 4-6: Infiltration on Claude west, a) Cumulative infiltration and b) infiltration capacity.  feS  Figure 4-7: Location o f infiltrometer test 1 on Claude west.  (pi  Figure 4-8: V e r t i c a l dye distribution in infiltrometer tests in Catchment 2.  ~f0  Figure 4-9: D y e tracer distribution 0.2cm below surface o f pile, a) Infiltrometer test 2 and b) infiltrometer test 3.  T" I  Figure 4-10: D y e tracer distribution from infiltrometer test 2. a) Excavation depth o f 0.2cm and b) stained macropore network at excavation depth o f 27cm. Figure 4-11: Macropore flow through fractures and over surface o f boulder below infiltrometer test 3.  2 ? "b  Figure 4-12: D y e tracer distribution below infiltrometer test 4. a) U n i f o r m dye front to 0.8cm and b) preferential flow in coarse zones at 1cm depth.  ^ ^  Figure 4-13: Vertical dye distribution in infiltrometer tests on Claude west.  ^ 5  Figure 4-14: Cross-section o f dye distribution below infiltrometer test 1 on Claude west.  7 b  Figure 4-15: Schematic o f flow mechanisms in shallow surface o f Claude East.  ~J ~\  Figure 4-16: a) L o n g duration rainfall events in which runoff was measured and b) short duration rainfall  8H  vi  events in which runoff was measured. Figure 4-17: R u n o f f predictions for rainfall events with different peak times.  ?5  Figure 5-1: Tracer release 1 in drain o f Catchment 1 on Claude east.  t IP  Figure 5-2: Textural differences in the Claude Waste R o c k Pile.  8 8 88  Figure 5-3: Compacted and crushed waste rock at old traffic surface. Figure 5-4: Schematic o f Tracer test 1.  9 S  Figure 5-5: F l o w over the surface o f large boulder excavated from 0.5m below surface - Tracer Test 1.  ID  Figure 5-6: Large boulder excavated from 0.5m below surface - Tracer Test 1.  <\o  Figure 5-7: D y e tracer moving from cobble to cobble or boulder hopping - Tracer Test 1.  ^ I  Figure 5-8: Boulder hopping - Tracer Test 1.  1 I  Figure 5-9: F i n a l distribution of Tracer test 2.  1 %  Figure 5-10: Cross-section o f previous traffic surface - Tracer test 2.  T3  Figure 5-11: Extent o f ponding on Claude west.  °\ H  Figure 5-12: F i n a l distribution o f dye tracer on Claude west.  ^ ^  Figure 5-13: Macropore f l o w in fractures o f a boulder below old traffic surface.  1 lo  vii  Acknowledgements I would like to thank Leslie Smith, my supervisor, for his guidance and patience throughout my research. Many thanks to Craig Nichol who helped me through my field work at Cluff Lake. This work has been funded by Cogema Resources, Cameco Incorporated and the Natural Sciences and Engineering Research Council of Canada.  Chapter 1.  Introduction  The generation of acidic water and the release of trace metals from waste rock piles continues to be a serious environmental challenge facing the global mining industry. The rate of release of metals, total contaminant loading and the time scale of impact On the environment have important implications when planning or decommissioning a rock pile. Literature on waste rock has focused on the mineral weathering of waste rock and on rate limiting factors, such as oxygen transport and microbial catalysis [Eriksson et al, 1997]. The rate of release of metals and dissolved metal concentrations are also determined by hydrological transport mechanisms, in addition to geochemical weathering. Few field investigations into the hydrogeologic structure of waste rock piles have been conducted to assess mechanisms of flow within a pile. Even fewer studies have been conducted on the spatial distribution of rainfall recharge into a rock pile. Uncertainty regarding water recharge and flow within waste rock piles can be a serious problem in predicting the behavior of new and decommissioned waste rock piles. Waste rock piles are typically unsaturated and highly heterogeneous, thus the occurrence of preferential flow, also termed macropore and bypass flow, is common in waste rock [Eriksson et al, 1997; Newman, et al, 1997; Smith et al, 1995; Diodato andParizek, 1994]. For the prediction of acid rock drainage (ARD), the hydraulics of a waste rock pile are more often treated as a homogeneous porous medium, in which the occurrence of preferential flow is not adequately described. A better understanding of how rainfall recharges a waste rock pile and what mechanisms of flow control infiltration is needed to improve our predictions of acidic drainage, metal transport and the long-term contaminant loading to the environment. This thesis presents the spatial distribution of rainfall recharge and mechanisms for rainfall infiltration into the Claude waste rock pile, located at Cogema Resources' Cluff Lake Mine. This project is one component of a collaborative study with Cogema Resources, Cameco Incorporated and NSERC for understanding flow and solute transport in waste rock piles located in northern Saskatchewan. 1.1  Objectives of the Study  There are two principal objectives in this study. The first objective is to evaluate the distribution of rainfall recharge through the surface of an unsaturated mine waste rock pile, the Claude waste rock pile. This was achieved through observations during rainfall events, fixed ring infiltrometer testing and measurements of rainfall intensity and runoff. The results will provide insight into the proportion of infiltration at the surface that is characterized as fast flow or gravity driven flow and the amount that is porous-media flow, described by Darcy's law. The second objective is to identify fluid pathways and mechanisms of infiltration and relate these to contact areas and material and structural properties within the Claude waste rock pile. This was achieved by releasing rhodamine WT, a dye tracer, on the surface of the rock pile to stain the travel paths of infiltrating water. Excavation of the stained waste rock followed the release and provided a visual record of the dye distribution with depth.  1  Recharge and infiltration mechanisms are important in modeling and predicting acid rock drainage (ARD) and the impact on the environment. The significance of this research is to improve our understanding of the mechanisms of infiltration and determine what proportion of waste rock is contacted by rainwater at the surface and throughout the pile, and thus able to transport ARD products to the environment. The key questions addressed by this research are the following: What is the spatial variability of rainfall infiltration into the waste pile? To what extent does rain water contact waste rock below the surface of the pile, to a depth of four meters? What are the mechanisms of infiltration into the pile? Based on observations of the Claude waste rock pile, a hypothesis can be made to answer the following question: what proportion of rainwater recharges deep into the pile and eventually to groundwater? 1.2  Site Description  The Claude waste rock pile is located at Cogema Resources' Cluff Lake Mine in the Athabasca basin of northern Saskatchewan. The pile was built between 1984 and 1989 from overburden and waste material from the open pit mining of uraniumfromthe Claude pit (Figure 1-1). Final grading of the top and sides of the pile was completed in 1993. The 7.2 million ton Claude waste rock pile stretches over a length of 900m, with a width of 600m at the south end. The average height of the pile is 30m, with sides sloped at 27%. An outline of the final footprint and contours of the pile is shown in Figure 1-2. The water table is believed to be located beneath the rock pile, at a depth greater than 30m below the surface of the pile. The pile is unsaturated, with the exception of perched zones, discovered during excavation of the Claude waste rock pile. Old traffic compacted layers were excavated at depths between 0.5m and 3m below the surface of the pile, therefore, the number of lifts or benches to give a total height of 30m may be between 10 and 60 benches. The Claude waste rock pile has a coarse texture with waste rock ranging in size from boulders to clay. Pile construction was a combination of dumping over the end of 0.5m to 3m lift sizes and free dumping over a previous lift, the result of which creates a number of compacted traffic surfaces within the pile. Due to different construction methods at closure, the surface of the east side of the pile (Claude east) is characterized by a compacted traffic surface, while the west side (Claude west) has a coarse and 'untrafficked' surface. The final lift on Claude west was built byfreedumping small piles over the previous lift and grading only once. The east and west sides of the pile are located as shown in Figure 1-2. The following six experimental site locations are also shown in this schematic of the Claude pile: Site 1: Catchment 1 - Tracer test 1; Site 2: Catchment 2 - Infiltrometer testing and Tracer test 2; Site 3: Catchment 7 - Runoff experiment; Site 4: Catchment 8 - Infiltrometer testing; Site 5: Ponded zone; Site 6: Claude west - Distributed tracer test.  2  It is believed that there exist perched water tables above old traffic surfaces within the rock pile. Evidence of these perched water tables was found during excavation of waste rock on the Claude pile. A tightly compacted layer or old traffic surface was excavated and waste rock on top of the compacted layer was saturated, however, waste rock was dry below the traffic surface. The structure of the coarse-grained rock pile is very loose on the surface of Claude west permitting free movement of water and air into the waste rock. The implications of this loose structure are important in ARD oxidation reactions and transport of reactants. An infinite supply of oxygen is made available for the oxidation of acid producing minerals. The products of oxidation can then be transported by infiltrating rain water to produce acidic leachate. The compacted and crushed surface of Claude east has a massive structure, in which the waste rock is tightly packed and cohesive. The location of Claude east and Claude west is shown in Figure 1-2 and the textural difference is illustrated in Figure 1-3. During a two-year return period rainfall event on August 13 , 1997, it was observed that rainfall on the east side of the Claude rock pile was ponding at the surface. Excess rainfall flowed overland to discharge into small zones (typically less than 100cm ) in topographic lows where an open structure in the rock mass permitted immediate infiltration into the pile. A number of catchment outlines were delineated during this 20minute storm by observing the patterns of rainfall runoff. A total of eight catchment areas on Claude east were delineated during heavy storms in August and September of 1997. The catchment areas range in sizefrom20m to 390m . Surface runoff in these catchments appears to collect around a 'drain', in which free water available at the surface during runoff is conducted downward very quickly through an open void, as small as 3 cm, or through a group of open voids at the surface. Recharge at a drain is characterized as fast flow, while infiltration through the surface crust is typical of Darcy flow. During the intense rainfall event on August 13th, rainfall runoff to a drain was photographed in a large catchment area of size 387m , hereafter referred to as Catchment 2 (Figure 1-4). The location of the stake shown in Figure 1-4 marks the drain location, where water was observed to swirl around an open gap in the surface and disappear into the pile. During the same intense rainfall event on August 13 , rainfall was observed to infiltrate immediately into the coarse surface on Claude west. A twenty-four hour rainfall event with a 100-year return period occurred in July, 1998, in which still no ponding was observed on the surface of Claude west. Clear evidence of an extensive traffic surface throughout Claude west was observed one day after this rare storm. A number of seeps were observed flowing out of the west side of Claude west at approximately 10m depth below the pile surface. It is believed that rain water was infiltrating vertically, ponding and flowing laterally over the surface of a compacted traffic layer. A measurement of the flow rate discharging from a prominent seep was 20L/min. One large spring was observed flowing from the toe of the pile in the northwest corner of Claude west. A number of springs were observed on the east side of Claude east, of which three or four seeps emerged at lm depth and two seeps emerged at 10m depth, below the pile surface. There were also significant seeps at the base of Claude east, adjacent to the ramp and th  2  th  3  flowing into the Claude pit. The majority of seeps stopped flowing within three days of the storm event. These observations were made by L. Smith, R. Beckie and C. Nichol in the period between July 12 and July 15 , 1998. Runoff from one catchment was observed to collect in a low permeability zone on Claude east, representing less than 0.1% of the pile surface, and remain in a ponded state for several days after a rainfall event. The location of the ponded zone is approximately 80m northeast of the top of the access ramp. This ponding area was approximately 140m on August 31 , 1997, as shown in Figure 1-5. The difference between this catchment and others observed on the pile is the low permeability at the drain. During later heavy rainstorms, in which the rainfall intensity was great enough to cause ponding on the surface of Claude east, rainfall excess was also observed to flow down the access ramp towards Claude pit. The sloped sides around the entire waste rock pile appear to behave similar to the surface of Claude west, in which infiltration was immediate and negligible amounts of overland flow occurred even in the most intense rainfall event during 1997 and 1998. Evidence of a surface crust can be seen over the surface of Claude east, especially in lower permeability zones. In many cases, active formation of a surface crust can be seen where overland flow occurs and the redistribution of surface fines point in the direction of runoff to a drain. In a runoff event fines are picked up and transported over the surface of the pile, only to be redeposited as runoff flows decrease. This process helps to reinforce the low permeability layer at the surface. Other evidence to support this is that gravel-sized and larger waste rock appear to be cemented with the fine surface materials and some force is required to remove large particles of waste rock from the surface. The imprint left behind often had a thin crust of fines curved up around the edge of the imprint in a shape fitted to seal around the removed waste rock particle. th  2  1.3  th  st  Overview of Thesis  Infiltration theory, factors affecting infiltration and the effect of spatial variability of infiltration capacity on runoff events are discussed in Chapter 2. A review of field investigations and dye tracer studies on waste rock and waste rock piles are also discussed in Chapter 2. Chapter 3 describes field methods for characterizing the distribution of rainfall recharge into the Claude pile, including fixed ring infiltrometer testing and measurements of runoff to a drain. In addition, techniques used for visualizing flow paths for rainfall infiltration into the matrix material on the surface of the pile are described. The results of fixed ring infiltrometer testing and quantitative estimates of the partition of rainfall between direct infiltration and runoff to a drain are discussed in Chapter 4. This chapter also provides physical properties related to the infiltration capacity of the waste rock. Chapter 5 describes the methods and results of three large-scale dye tracer tests on Claude east and west. The conclusions are summarized in Chapter 6.  4  Chapter 2.  Literature Review - Understanding Infiltration in Coarse Waste Rock  The following review provides an understanding of rainfall infiltration and factors affecting infiltration. The theory behind the fixed ring infiltrometer method of characterizing infiltration is discussed. The effects of macropores, surface crusting and traffic compaction will be discussed in relation to infiltration pathways and rates. The effect of spatially variable infiltration capacities on the hydrologic response in a catchment system is also reviewed. Finally, a review of field investigations and dye tracer studies for characterizing flow through waste rock is also provided. Understanding flow through a coarse waste rock pile is complicated by heterogeneous permeability and layered structures due to pile construction. Typical to a coarse waste rock pile are large voids or non-capillary pores in the waste rock, some of which are open at the waste rock surface. Referred to as macropores, these capture runoff or ponded water at the surface or within the pile and conduct it rapidly downward, bypassing the waste rock matrix material. The matrix material comprises finer grain sizes, in which capillarity is a dominant mechanism for flow. Characterizing infiltration is also complicated by the conditions on the surface of the pile. Traffic during construction on the pile surface creates a layered waste rock profile due to crushing waste rock and compacting voids, in which the permeability of the subsurface layer is greater than that of the layer above. The surface of Claude east is characterized by such afine,compacted traffic surface. Due to pile construction in 0.5m to 3m lifts, traffic-compacted surfaces exist at various depths through both sides of the Claude pile. Extensive traffic surfaces throughout Claude east and west were evident during the 100-year return period storm. 2.1  Infiltration Theory  Infiltration of water into waste rock is controlled by two driving forces: capillary potential (capillarity) and gravitational potential. Capillary force is the hydrogen bonding of water and waste rock or surface tension between water and waste rock. The capillary potential gradient can be in any direction and the strength of the gradient depends on pore size and moisture gradient. The dominant driving force for infiltration in the waste rock matrix material changes through time from capillary potential during early-time infiltration to gravitational potential, at which time the flow reaches a constant rate. Infiltration capacity tends to be high in the early stages of rainfall and then slows to the steady state infiltration capacity when the soil becomes saturated. This decrease in infiltration capacity results mainly from the decrease in the capillary potential gradient, as the wetting front moves farther from the surface. The decrease in infiltration capacity is also due to a decrease in the rate of uptake of water into storage. The relationship between water content, matric suction and conductivity is basically that when water content increases, matric suction decreases and the unsaturated hydraulic conductivity increases until saturation. When a soil becomes saturated, the hydraulic conductivity is at a steady state and matric suction is zero.  5  Early-time infiltration capacities will vary depending on the initial wetness and capillarity of the waste rock. An initially 'wet' waste rock has a small matric-potential gradient and therefore a lower infiltration rate, and reaches a steady infiltration sooner. Water ponded on dry waste rock, in which the matric potential gradient is large, will have a high initial infiltration rate and longer time to reach steady state. The initial infiltration capacity and time to approach steady-state flow is, of course, also influenced by the hydraulic conductivity of the soil. Infiltration capacity was measured on the Claude waste rock pile using a fixed ring infiltrometer. The theory of flow for ring infiltrometers assumes that the cumulative infiltration (i[L]) in the early stages of flow is linear with the square root of time and that at long times i should be linear with time [White et al., 1992]. The initial slope of the cumulative infiltration versus the square root of time graph is defined as the sorptivity (S[LT" ]) and the final slope is the steady state infiltration capacity (i [L/T]). Refer to Figure 2-1 for a diagram illustrating this concept. The following equations are a summary of the early-time and steady state behavior: limt=>o di/dt => So limt=> di/dt => i The sorptivity is a measure of the wetting front suction or matric-potential gradient and is a function of the initial water content and porosity, or storage available. A measure of sorptivity provides the transient response of infiltration at the start of a rainfall event. Predictions of runoff on the Claude pile can be in gross error if the early-time infiltration capacity is not accounted for. A soil profile with a high sorptivity will have a rapid initial infiltration and reach steady state sooner, as compared to a soil profile with a low sorptivity. 1/2  ss  00  2.2  ss  Factors Affecting Infiltration - Crust-Topped Soils, Compaction and Macropores  An important factor affecting infiltration on the surface of Claude east is the formation of a thin crust or seal at the surface. These conditions often develop when a soil is exposed to the beating action of raindrops, which break down soil structure and redistribute fines into the pore spaces at the surface. The formation of a surface crust is enhanced because there is no vegetation cover on the Claude pile. The restricting surface crustfromraindrop compaction is typically 1mm to 1cm thick [Klute, 1986] and was observed on the Claude pile to range from 1mm to 5mm. The effect of traffic compaction causes a similar low permeability barrier, however the scale of aggregate sizes that are broken down and redistributed are much larger thanfromraindrop compaction. Traffic compaction creates a layered permeability effect on the surface of the waste rock pile, in which permeability increases with depth below the surface. Waste rock piles may have many such compacted traffic layers within the pile. These layers tend to create a barrier to flow, as suggested by Diodato and Parizek (1994) who conducted geophysical logging of the hydrologic properties of an unsaturated waste rock pile. They found that sharp increases in volumetric moisture content, from 8% up to 46%, were coincident with high density layers within the pile, suggesting perched water-tables. For low-intensity storms, estimating infiltration is difficult from ponded ring infiltrometer measurements as these are heavily influenced by macropores and these do  6  not participate in rainfall infiltration until surface ponding occurs (Clothier and White, 1981; 1982). In a natural rainfall, ponding depths are near zero, much less than the pressure required to overcome the air entry value of some macropores. Therefore, macropores will often stay dry and will not participate in transmitting water during a lowintensity rainfall. In addition, T. Talsma has proven that the sorptivity of a soil depends on the hydraulic head applied at the surface, especially where infiltration is dominated by structural cracks and roots (Talsma, 1969). In a heavy-rainfall event in which ponding occurs, infiltration capacity depends to a large extent on the size and volume of noncapillary pores and cracks, hereafter referred to as macropores (Berndtsson and Larson, 1987). Bypass flow through macropores at the surface can be rapidly transmitted through a network of macropores or can be taken up by capillarity into the surrounding and drier matrix zones, thus speeding up the rate of inflow (HUM, 1980; Beven and Germann, 1982). The effects of macropores on water flow and solute transport can be significant (Beven and Germann, 1982). Flow bypass causes an initial rapid transport of contaminants followed by a "tailing", in which the concentration of solutes is controlled by the rate of diffusion from the matrix to the macropores (Wildenschild et al, 1994). 2.3  Spatial Variability of Infiltration  Variability of infiltration at the surface is due to the heterogeneity of waste rock physical properties. Heterogeneity may be due to non-uniform spatial compaction and crushing of waste rock, big boulders within the deposit, structuring imposed by pile construction, and different geologic materials from around the ore deposit. The spatial variability of infiltration capacity has primarily been studied in natural catchment areas that include agricultural, urban or forested areas. Infiltration capacity varies greatly even within small watersheds (Berndtsson and Larson, 1987; Sullivan et al, 1996; Merz and Plate, 1997). In a small semi-arid catchment, Berndtsson and Larson (1987) conducted fifty-two ring infiltrometer tests and found initial infiltration capacities to vary between near zero and 33mm/min and steady state infiltration capacities to vary between near zero and 6mm/min. The coefficient of variation, which expresses the standard deviation over the mean, was over 110% for the steady state infiltration capacity. Warrick and Nielsen (1980) gathered the results of different field infiltration studies and found coefficients of variation of 170-400% for steady state infiltration capacity (Merz and Plate, 1997). Another example of variability of infiltration in a catchment was shown by Burgy and Luthin (1956), who measured the average infiltration capacity of a clay soil by flooding a 335m basin and measuring a decline in the ponded surface of 1.8mm/min. This was followed by ring infiltrometer measurements conducted in 119 sites throughout the basin. Infiltrometer values ranged from near zero to 1 lOmm/min with an average value of 2.8mm/min, slightly higher than the basin infiltration capacity (Dingman, 1994). Infiltration parameters for the Claude pile are expected to be highly variable in space, due to the complexity of spatial patterns of texture, degree of compaction and moisture within the surface of the pile. Merz and Plate (1997) studied the effect of spatial variability in soil properties on runoff. They showed that spatial variability results in runoff that has a complex and 2  7  event-dependent behavior, i.e. runoff is different with changing storm size. They found that spatial variability in runoff behavior plays a negligible role in very small and large events. Runofffrommedium-sized rainfall events is significantly affected by the spatial variability in soil properties. 2.4  Field Studies on Flow through Waste Rock  Diodato and Parizek (1994) conducted a study on the hydrogeologic properties of a coarse waste rock pilefroma coal strip mine. They used geophysical logging to quantify volumetric moisture content and bulk density and determine unsaturated hydraulic conductivity during rainfall events. They found that changes in these parameters were very short lived, rapid pulses and concluded that groundwater recharge from the waste rock pile occurs in short lived pulses, corresponding to infiltration and recharge events. In addition, Diodato and Parizek (1994) applied a concentrated solution of bromide on the surface of the waste rock and allowed the tracer to be transported by naturally occurring rainfall. They sampled for bromide from pressure-suction lysimeters installed at 2m depth below the surface. Two concentration peaks, of similar magnitude, were observed, separated by a period of six months. Their hypothesis was that a dualpermeability flow may be occurring in which water is conducted rapidly through macropores and through the finer grained matrix much later. The late concentration peak was held in storage for a longer period of time. Another important observation of the bromide tracer was that breakthrough curves had a long tail, suggesting that water is retained during dry periods in the finer-grained waste rock, and flows when the moisture content rises in response to another rainfall event. The preferential flow of water infiltrating through coarse gravel and broken rock in unsaturated conditions was documented by Elboushi (1975). A dilute solution of white paint was ponded and infiltrated into a rectangular pile of unsaturated coarse rock to observe the pathways for infiltration. The height of the pile was approximately lm with a surface area of 1.7m . The distribution of paint coverage with depth showed that water infiltrating the coarse rock tends to concentrate in small rivulets, some of which end abruptly and only a few continue to exit the bottom of the rock pile. The paint coverage decreased from 100% at the surface to 20% of the cross-sectional area coated at 90cm depth. Elboushi also recorded that the average paint coverage on particle surfaces decreases with depth, such that only 10% to 20% of the surface area of individual rock particles was coated at deeper levels in the pile of coarse rock. A similar paint-flush study was conducted to examine the small-scale flow paths through the Eskay Creek waste rock pile (Morin et al, 1997). The study followed the approach of Elboushi (1975) and used white latex paint as a water tracer. However the latex paint was only diluted 1:1 with water, whereas Elboushi diluted the paint 1:5 with water. Twelve paint flush sites were established, in which 5 to 20 gallons of paint were applied in cylinder rings embedded into the rock. Due to the relative impermeability of the crushed and compacted surface of waste rock that resembled a well-compacted silt, virtually none of the paint infiltrated in most of the site locations, over a 24-hour period. A few paint flush sites were chosen where coarse rock was visible at the surface of the Eskay Creek pile, which was not the common condition for this pile. Near vertical 2  movement through relatively homogenized waste rock was described, however the paint encountered a similar fine compacted surface of waste rock and thus was impeded from migrating further than one meter. In one paint flush site, paint was described to pond on a fine compacted layer, located approximately 10cm below the surface of the pile, and then migrate further by flowing around the surface of a pebble embedded in the fine layer. Morin et. al (1997) concluded for the Eskay Creek waste rock pile that fine layers of crushed and compacted waste rock can significantly impede vertical infiltration, however even a pebble can breach its integrity. Erriksson, Gupta and Destouni (1997) investigated preferential flow and transport in mining waste rock through tracer tests in field lysimeters. They found indications of water flowing preferentially in approximately 55% to 70% of the total water content while 30% to 45% of the total water content in the field is immobile. Their theory for the dominant mechanism of preferential flow is big boulders channeling water flow at the surface and within the pile. However, they did not credit the cause of preferential flow to geological heterogeneity, deposition of waste rock and compaction variability (Eriksson et al, 1997). Smith et al. (1995) reviewed the hydrogeological properties of four waste rock piles from different mine sites: Myra Falls, B.C., Island Copper, B.C., Elkview Mine, B.C., and Golden Sunlight Mine, Montana. The emphasis of the study was on pile hydrostratigraphy and the response to a rainfall event, and the large-scale hydrogeologic characterization of a waste rock pile inferred from outflow hydrographs recorded in toe drains. Smith et al. (1995) determinedfromthese data sets that water infiltrating a coarse pile could take only a few hours to travel through the pile, suggesting rapid channel flow. However, one conclusion of the study was that no single comprehensive data set of the parameters required to characterize the hydrologic behavior of a waste rock pile exists for a single mine. The following study is an attempt to provide a comprehensive data set for characterizing rainfall infiltration into a coarse unsaturated waste rock pile.  9  Chapter 3. Field and Laboratory Methods for Characterizing Rainfall Infiltration  3.1  Runoff Experiment  A catchment of 195m on Claude east was instrumented to measure the volume of direct runoff to the catchment drain, and the volume of infiltration into the finer-grained surface of the catchment. This information is important as it provides a ratio of the amount of slow and distributed infiltration, or Darcy flow, to the amount of rapid inflow at the catchment drains. The runoff and drain area, hereafter referred to as Catchment 7, is illustrated in Figure 3-1 and the location is shown as site 3 in Figure 1-2. The approximate size of the catchment was outlined during a typical short and intense rainfall event on June 25th, 1998, which had a one year return period at Cluff Lake. Refer to Appendix 1 and 2 for rainfall records and statistics for extreme rainfall events at Cluff Lake. Markers were dropped on what appeared to be a flow divide around the perimeter of the catchment during ponding and runoff. The appearance of a surface flow divide was usually shown by a waste rock mound and water running in opposite directions from the mound during the intense rainfall event. The creation of catchments appears to be due to final grading and vehicle ruts on the pile. A tipping bucket rain gauge and data logger were installed beside the catchment to monitor rainfall during a stormy period at the end of June and July, 1998. A 200L drum was embedded into the pile at the drain location at the bottom of the catchment. One edge of a small plastic sheet was placed around the perimeter of the barrel and the other edge was embedded and sealed with silly clay in the surface crust of the catchment (Figure 3-1). The volume of water in the barrel was measured daily, following a rainfall event. A continuous record of total rainfall and also rainfall intensity from a nearby climatic station, about 3km away from the rock pile, is provided for May, June and July, 1998 in Appendix 1. These data provide some idea of relative peak rainfall intensities and moisture conditions before and during the experimentation period. No infiltrometer testing was performed in Catchment 7 to maintain the integrity of the surface for runoff measurements. 2  3.2  Fixed Ring Infiltrometer Testing  (...Old beer cans do not make good infiltrometers . H.Bouwer, 1986) The majority of infiltration measurements on Claude east were taken during a relatively dry period in the second week of June 1998. During the seven days of experimentation, the total rainfall observed at a nearby climatic station was only 5.4mm. Only 9.8mm of rainfall occurred during the preceding two-week period (Appendix 1). The first set of infiltration measurements on Claude east and the measurements on Claude west were taken during a rainy period in September, 1997. The flooding-type, single ring infiltrometer was used to measure infiltration. It is recommended for providing the best measure of vertical infiltration (K.Loague, 1990). The ring infiltrometer is an inexpensive cylindrical design, made simplyfromcut sections 1  10  of a forty-five gallon drum. The dimensions and set-up of the infiltrometer used are shown in Figure 3-2. A sledgehammer was used to drive the infiltrometer into the waste rock to a depth of about 5cm. The ring infiltrometers were installed with little disturbance to the existing surface conditions, however loose rocks that were caught under the cylinder edge were removed. The inside and outside wall were sealed with silty clay to prevent flow from bypassing the surface crust through the edge of the ring. The effective diameter for infiltration was 58cm. Water was rapidly ponded in the ring infiltrometer and readings of the drop in the ponded water surface were taken from a ruler taped on the inside of the cylinder. Different depths of ponded water were applied in the ring infiltrometers during the two different experimentation periods. Approximately 12cm to 15cm of water was ponded in the ring infiltrometers during the 1997 experimentation period and only 5 cm to 7cm of water was ponded during the 1998 experimentation period. This change in methodology was meant to reduce ponding pressures at the surface, which tend to force entrapped air out of macropores and provide a flow bypass through the soil matrix, as per the discussion in Section 2.2. In addition, the high ponded pressure may have also caused piping through the surface crust and between macropores in the waste rock, creating an interconnected macropore system (communication with R. Sidle, 1998). Another reason for reducing the ponded depth was to approximate a gradient of one, such that the infiltration capacity and hydraulic conductivity of the soil are identical. The limitations of infiltrometer testing will be discussed further in Section 4.7. Water was rapidly poured onto a wide plastic lid placed in the infiltrometer to avoid direct impact of water on the waste rock surface. Careless application of water can cause erosion, and fine particles to become suspended and settle onto the surface during ponding, resulting in a restricting layer. Measurements of the drop in the ponded surface were taken at intervals of one minute or less during the start of infiltration, and less often as a steady state infiltration was approached. The infiltrometer was constantly refilled to maintain a ponded depth of approximately 5cm on the surface of the waste rock. Infiltrometer testing was conducted in two catchment areas on Claude east, which were outlined during an intense two-year return storm in August of 1997. Many catchments on the surface of Claude east were identified and outlined as discussed in Section 3.1. Catchment 2 is a large catchment of size 387m and is shown during a rainstorm in Figure 1-4. Catchment 8 is a smaller catchment of size 76m . The locations of Catchment 2 (site 2) and 8 (site 4) are shown in Figure 1-2. A total of sixteen infiltrometer tests were performed in random locations within Catchment 2. Ten infiltrometer tests were carried out in Catchment 2 in the dryer period in June, 1998 and six were carried out in the wetter period in September, 1997. Fourteen infiltrometers tests were performed in random locations within Catchment 8 in June, 1998. Only the data collected during the infiltrometer testing in June, 1998 were used to estimate sorptivity, as more care was given to the collection of early-time data. Field methods in June, 1998 involved careful sealing of the infiltrometer edges. If leakage was observed, the test would be restarted at a new unwetted location. Total rainfall in the period before and during infiltrometer testing in June, 1998 is provided in Appendix 1. During the same rainstorm in August, 1997, observations of infiltration on Claude west revealed no ponding. Rainfall recharge appeared to be immediate everywhere on the  11  surface. Four infiltrometer tests were carried out in random locations over the surface of Claude west in September, 1997. A fifth infiltrometer test was completed on a pathway created by our truck driving over the coarse 'untrafficked' surface of Claude west. (Figure 3-3) The diameter of infiltrometer 5 was 25cm to fit into the rut created by the wheels of our truck. To characterize the infiltration rate of the low permeability ponding zone near the access ramp on Claude east, an infiltrometer was placed in the deepest part of standing water and the drop in the ponded surface was monitored through August and September, 1997 (Figure 1-5). The size of the ponded area was measured to spread over 140m on August 31 , however, the ponding area decreased by 50% by September 4 . The ponding area covers less than 0.1% of the pile surface, at its largest size, thus it represents very little of the recharge into the waste pile. The drop in the ponded surface was monitored during the period between September 9 to September 24 , 1997. The drop in the ponded surface was measured twice per day from a ruler on the inside wall of the infiltrometer. Evaporation was also responsible for the drop in the ponded surface area, therefore, evaporation rates were taken into consideration for determining the water balance during this period. Evaporative fluxes were assumed for the top of the waste pile from measurements in an evaporation pan located approximately 3 km away in a tailings management area. The evaporation in the vegetated tailings area is an approximation to the windy and exposed surface of the waste pile. 2  st  th  th  3.3  th  Small-Volume Dye Tracers in Ring Infiltrometers  Rhodamine WT tracer was released in the ring infiltrometer at the end of infiltrometer tests 2, 3, 4, 5 and 6 in Catchment 2 and infiltrometer tests 1 and 2 in the coarse surface of Claude west to trace the pathways of infiltrating water through the matrix material on the surface of the waste rock pile. The concentration of the applied tracer was chosen to be 1.25ml of a 5% rhodamine solution, per 1 liter of water. The concentration was determined by testing the visibility of rhodamine on different samples of Claude waste rock. Approximately 20L (8cm) of rhodamine solution was ponded in the ring infiltrometer and allowed to infiltrate. The volume of dye tracer was chosen so that the expected penetration depth was 20cm, assuming a porosity of 0.40. Excavation of waste rock using hand tools was manageable to this depth below the surface of the pile. The infiltration of rhodamine was followed by careful excavation of the stained waste rock. Tools, such as a garden shovel and brush, were used to remove the waste rock in 1mm to 5cm lifts. A photograph was taken at each excavation depth to record the dye tracer distribution for further analysis. To maintain a scaled three-dimensional perspective of dye distribution with depth, a one-meter square grid with 10cm spacing was placed in the same horizontal location at variable depths and photographed. Using the photographs, dye contact areas were estimated from the ratio of stained waste rock to unstained waste rock, within the footprint of the ring infiltrometer. Dye coverage with depth provides an estimate of the area of waste rock in contact with the infiltrating water. In summary, rhodamine WT was assumed to stain the transmission routes for water infiltrating at a steady rate.  12  Rhodamine WT was chosen as a visual dye tracer because of its mobility and high visibility for photographing. Experiments conducted in the laboratory showed that rhodamine WT was coincident with the wetting front and produced a dark red stain that was clearly visible on the travel paths of water. Afieldstudy by Pang et al. (1998) also determined that rhodamine WT was a nonreactive tracer and used this dye to characterize preferential flowpaths in an alluvial gravel aquifer. An evaluation of four other dyes, typically used for staining the travel paths of water, proved to be heavily sorbed and retarded from the wetting front migration. Refer to Appendix 3 for details of the experiment setup and results of dye testing.  3.4  Physical Properties  3.4.1  Density and porosity measurements  The degree of compaction by traffic on Claude east is an important influence on the permeability of the surface. A simple test was used to determine the density of the compacted traffic surface of Claude east in Catchment 2 and 8. Two waste rock samples, one from each of Catchment 2 and 8, were carefully excavated and collected from a circular pit with a diameter and depth of approximately 30cm. The pit was then refilled with a known volume of coarse silica sand. The waste rock sample was oven dried and weighed. The known dry mass divided by the volume of silica sand yielded a rough estimate of dry bulk density for the surface of Catchment 2 and 8. From this information and an assumed particle density of 2.65g/cm , the porosity (n) of the in-situ waste rock from the surface of Claude east can be determined from the following: n = e/(l + e), in which e is the void ratio, defined as e = Gsp /p -l, in which Gs is the particle density, p = density of water and pb = dry bulk density of waste rock. w  b  w  3.4.2  Constant head permeameter - saturated hydraulic conductivity  Waste rock from the top 0.5m of Claude east was shipped from Cluff lake to the University of British Columbia to determine the saturated hydraulic conductivity of the waste rock. This sample of waste rock was excavated from the surface at Site 1, located as shown in Figure 1-2. Keep in mind that this waste rock has undergone changes in texture, density and weathering state due to traffic compaction and exposure on the surface, before and after completion of the pile in 1989 andfinalgrading in 1993. Excavating and shipping the waste rock, from in-situ conditions, into 200L drums and repacking the waste rock in a laboratory experiment does not produce permeability values representative of in-situ conditions. However, a laboratory measurement of saturated hydraulic conductivity may be useful to compare to field 'saturated' values of steady state infiltration capacity. A constant head permeameter was built to measure the saturated hydraulic conductivity of waste rock. The permeameter was set up to produce an upward flow through the waste rock sample, such that the head below the sample is greater than the  13  head at the top of the sample. The pressure head at the bottom of the sample was supplied by a constant head reservoir situated slightly higher than the sample height. A twenty-liter pail served as a constant head reservoir, connected through a flexible nylon hose through the bottom of the waste rock sample. The column permeameter was constructed from plywood and lined with a 0.5mm plastic geotextile. The dimensions of the column are 60cm x 60cm x 60cm. The experiment set-up is illustrated in Figure 3-4. Waste rock was loosely placed in the column by dumping approximately seven 6cm lift sizes (seven 20L volumes) for a total sample height of 41cm. The waste rock was not compacted, however the surface was leveled off between each lift. The waste rock was placed over a 3 cm layer of coarse and clean gravel that allowed for a uniform pore pressure distribution below the waste rock sample. A piezometer was installed in the gravel layer below the sample and a shorter piezometer was installed in the ponded water over the surface of the sample so that head differences could be measured accurately. To minimize entrapped air during saturation of the waste rock, a slow rate of inflow was applied into the gravel beneath the waste rock, such that pore spaces were saturated from the bottom of the column, allowing air to escape. The sample was saturated in three days. To prevent bypass flow along the walls, dry silty sand was poured into a 1cm space between a cardboard liner and the inside of the column walls. The cardboard liner was raised with each new lift placed in the column. After saturation of the waste rock, a head difference was applied by raising the constant head reservoir. The applied head difference was approximately 7cm, low enough so that piping and erosion did not occur and high enough to measure a flow rate out of the permeameter. Settlement of 0.5cm occurred over the first two days of monitoring, however the density remained unchanged thereafter. Outflow rates and gradients were monitored over four days, however after the second day the system was at a steady state outflow and thus the conductivity remained unchanged. The density of the waste rock in the permeameter was measured after hydraulic testing by removing a measured volume and weighing the oven dried waste rock. The density in relation to the measured saturated conductivity of the column will provide a comparison to the field-measured density and field 'saturated' conductivity from Catchment 2 and 8. 3.4.3  Moisture content  Moisture conditions before ponding were determined on Claude east from waste rock samples collected from the surface 5cm within Catchment 2 and 8. The volume of each sample was roughly 5 liters or 8kg. The gravimetric moisture content was determined from the difference in weight between the wet and oven dried waste rock sample divided by the dry weight of the sample. An approximation to the volumetric moisture content (9 ) was calculatedfromthe measured dry bulk density and gravimetric moisture content for an assumed particle density of 2.65g/cm . The volumetric moisture content is the following: ew = v / v = e (p /p ) where 0 = gravimetric moisture content, p - density of water and W  w  t  g  b  w  g  w  14  Pb = dry bulk density of waste rock. Waste rock grain density, particle distribution, and pore sizes are very heterogeneous on the Claude rock pile, so there is some uncertainty in estimating a representative volumetric moisture content. Initial moisture content for catchment 2 was estimated from only one sample taken on June 15 , 1998. Initial moisture conditions were estimated for Catchment 8 for three samples taken on June 11, 12 and 13, 1998. th  3.4.4  Surface texture  The texture of waste rock at the surface of the rock pile is key to the infiltration capacity, sorptivity, and the geometry of fluid pathways. Grain size sieve analyses were conducted on five samples of waste rock taken from the shallow surface of Catchment 2 on Claude east. Four samples were from the top 1cm at infiltrometer locations 2, 3, 4 and 6, and another sample was takenfroma range between 5cm and 10cm below the surface at infiltrometer location 2. Two waste rock samplesfrominfiltrometer location 2 provide a comparison of textural changes with increasing depth. Waste rock texture on the top 20cm of Claude west was approximated from a sieve analysis of four samples, taken during excavation of the dye tracer release in infiltrometer test 1 and 2 on Claude west. Two samples were collected from below infiltrometer test 1, onefromthe top 10cm and anotherfrom20cm below the first sample. Two samples were collected from the same depth at 18cm below infiltrometer test 2 on Claude west.  15  Chapter 4.  Rainfall Infiltration  The following results estimate the partitioning of rainfall between direct infiltration, and surface runoff to a drain in three catchments on Claude east. Direct measures of runoff in Catchment 7 are presented for different storm events in June, 1998. The infiltration behavior of Catchment 2 and 8 was characterized from the average of thirty ring infiltrometer tests. Estimates of the mean infiltration behavior in Catchment 2 and 8 of Claude east are applied to actual storm intensities from the summer of 1997 and 1998 to estimate the partition between recharge to the fine matrix waste rock and recharge to the coarse drains. Results of the five infiltrometer tests performed on Claude west are also reported in this chapter. 4.1  Density, Porosity and Saturated Hydraulic Conductivity  The in-situ density of waste rock from the top 30cm of Catchment 2 was estimated from a sample volume of 9L with a dry weight of 14 700g. The dry bulk density measured in Catchment 2 was approximately 1.63g/cm . The dry bulk density measured from the top 30cm of Catchment 8 was 1.69g/cm , in which the sample volume was 7.7L and weight was 13000g. A large sample volume of 33L was dried and weighed to estimate the dry bulk density in the column permeameter. Recall that this waste rock was loosely packed into the column and settled 0.5cm, over a sample length of 41cm. The waste rock packed in the column permeameter was measured to have a similar density to Catchment 2 of 1.63g/cm . The porosity of waste rock taken from the shallow surface of Claude east was estimated using the dry bulk density, as determined in Section 3.4.1. The estimated porosity of waste rock taken from the top 30cm of Catchment 2 and Catchment 8 was 0.39 and 0.36, respectively. The porosity of the waste rock in the column permeameter was also estimated to be 0.39. The saturated hydraulic conductivity of waste rock loosely packed in the column permeameter was 4.6 +/- 0.1 (x 10") m/s, which is representative of a medium grained sand and a Darcian material. The range of error in the measurement of hydraulic conductivity was chosen from the smallest unit on the ruler used to measure head drop over the surface of the permeameter. The hydraulic conductivity corresponds to 2.8mm/min, in terms of infiltration capacity units. This is a high infiltration capacity, likely due to the loose packing of the waste rock matrix material. 3  3  5  4.2  Moisture Content  The preponding moisture conditions of Catchment 8 appeared to remain fairly constant over the three sampling days on June 11th, 12th and 13th, even though a light rainshower occurred on the evening of the 11th. The three samples taken from different locations within Catchment 8 had a gravimetric moisture content of 5%, 5% and 8% on the respective days of June 11th, 12th and 13th. The different moisture content measured on June 13th is likely due to spatial variability in sampling, as no rain occurred since the last moisture measurement of 5%. Gravimetric moisture content in Catchment 2 was measured to be 4% from one sample taken on June 15 , 1998. During the week of th  16  infiltrometer testing in June, 1998, pre-ponding volumetric moisture conditions on Claude east were estimated to be in the range between 7% and 13%, which is approximately 18% to 37% saturation of pore spaces (calculated from 4% to 8% gravimetric moisture content, bulk density and porosity). The average initial infiltration behavior (ie.sorptivity value) for Catchment 2 and 8 is representative of matric suction gradients that develop between saturation of 18% to 37% of the pore space. The pre-ponding moisture content was unimportant during infiltrometer testing in August, 1997, because sorptivity was not measured at that time. Recall that initial moisture content has no effect on steady state infiltration during field-saturated conditions. Moisture content during the wet conditions experienced in 1997 are provided here for insight as to how the moisture content can vary on the surface of Claude east. During field experimentation in August, 1997, the gravimetric moisture content was estimated between 10% and 12% in Catchment 2, which corresponds to an approximate degree saturation of 45% to 54%. These moisture measurements are meant to be firstorder approximations of the moisture contents in the matrix material of the shallow surface of Claude east. 4.3  Surface Texture  The texture of waste rock at the surface of the rock pile is key to the infiltration capacity, sorptivity, and, of course, the geometry of fluid pathways. The results of the grain size analyses conducted on near-surface samples of waste rock taken from Claude east and west showed distinctly different textures, as shown in Appendix 4. The mean grain size of the four samples taken from the top 1cm of Claude east represents a texture that is approximately 33% gravel, 44% sand and 23% fines. Grain size analysis on three samples from the top 18cm of Claude west represent a mean grain size distribution of 75% gravel of larger particles, 21% sand and 4% files. Two samples taken from differing depths below infiltrometer test 2 on Claude east showed an increasing grain size distribution with depth. The sample taken from the top 5mm of the surface crust had a 10% greater fraction of sands, compared to the waste rock material sampled between 5 cm and 10cm below the surface, which had a 10% greater fraction of gravel. The surface of Claude east is characterized by a layered permeability that increases with depth. This is consistent with observations of the compacted traffic surface on Claude east. The texture of waste rock was analyzed at deeper locations within the pile from samples taken during excavation of the large-scale dye tracer tests. Six samples were taken from 35 cm to 300cm below the surface of Claude east and four samples were taken from 50cm to 150cm below the surface of Claude west. Claude east and west have a similar waste rock texture at these depths, as shown in Appendix 4. The mean grain size distribution of waste rock below the surface of the pile is typically 74% gravel or larger particles, 20% sand and 6% fines. As expected, this waste rock texture below the surface of the pile is similar to the surface of Claude west, where no traffic compaction has occurred.  17  4.4  Infiltration Measurements  The results of thirty infiltrometer tests in Catchment 2 and 8 on Claude east and five tests on Claude west are presented in this section. Infiltration data are presented as cumulative infiltration(mm) versus elapsed time(min). The steady state infiltration rate is estimated from a regression analysis of the straight-line behavior at later time. The earlytime infiltration behavior is described by the sorptivity, estimated from a regression analysis of the cumulative infiltration versus square root time. Ponding experiments on Claude west resulted in negligible early-time, capillarity behavior. Therefore, only steady state infiltration values are presented for Claude west. 4.4.1  Catchment 2  Figure 4-1 shows measurements of the cumulative infiltration versus time for all sixteen infiltrometer tests. The infiltrometer measurements showed a large range in infiltration capacities. Steady state infiltration behavior in Catchment 2 was attained after an average twenty-five minutes of capillarity-dominated flow. On average, two hours of infiltration was monitored to estimate a steady state flow rate into each ring infiltrometer. Estimates of the steady state infiltration capacity and sorptivity for each infiltrometer are included in Table 4-1. The distribution of these measured infiltration parameters, with the exception of infiltrometer test 1, is positively skewed as shown in the frequency histogram in Figure 4-2. Most of the steady state infiltration rate values (87% of the tests) were in a wide range from 0.04mm/min to 0.7mm/min, however an infiltration capacity of more than 24.8mm/min was measured in infiltrometer test 1. The high infiltration rate of infiltrometer test 1 is at least two magnitudes greater than the other 15 infiltrometer tests. Features on the surface at infiltrometer test 1 are similar to those of a drain, in which cobbles and large voids are open at the waste rock surface, allowing a bypass through the surface crust (Figure 4-3). Infiltrometer test 1 confirms that the surface of Claude east has indeed a bi-modal permeability. Infiltration in infiltrometer test 1 is rapid and very different from the slow infiltration capacity of the fine matrix waste rock, which covers the majority of the surface on Claude east. Infiltrometer test 1 was on a local topographic high within the catchment, and thus it could not function as a drain, although it has an infiltration capacity representative of a drain. Infiltrometer test 1 was removed from the population sample for estimating the average infiltration behavior of Catchment 2. Infiltrometer test 5 was also removed from the estimate of the average steady state behavior of Catchment 2 as it did not appear to reach steady state during the short monitoring period of 30 minutes. Early-time data from the infiltrometer tests conducted in 1997 do not provide a reliable indication of sorptivity in Catchment 2. Over an hour of non-linear infiltration can be seen in infiltration tests 2, 3 and 5, performed in the wetter period in 1997 (Figure 4-la). In wet soil conditions, infiltration should reach steady state sooner, due to a lower suction gradient. This prolonged period was likely due to poor sealing and leakage around the edge of the infiltrometers and not due to capillary behavior or macropores filling. The length of time for non-linear infiltration from infiltrometer test 2, 3 and 5 are too great to be explained by capillary behavior. Infiltrometer testing of in-situfieldsoils,  18  in which the soil-water potential at the surface was greater than zero, were found to have sorptive time scales of 2.4 minutes (Bundendore fine sand), 18 minutes (Fowlers Gap light clay) and up to 50 minutes for Fowlers clay loam (White and Sully, 1987). Infiltrometer tests 4 and 6 were measured after a period of saturating the surface and repairing leaks, therefore, early-time behavior was not recorded. The quality of earlytime data from the six infiltration tests performed in 1997 are poor, as the goal then was only to measure the steady state behavior. The leaks and clay seals around the infiltrometers were repaired during infiltration, so that an accurate measure of steady state infiltration capacity could be obtained. The greater ponding depth used in 1997 (12cm to 15cm instead of the 5cm to 7cm depth applied in 1998) may have forced more entrapped air out of the soil profile, possibly increasing infiltration in the macropores as well as creating a greater crosssection for flow. In addition, the increased pressure may have perforated the low permeability surface crust and initiated piping between separate macropores, which may cause an overestimate of the infiltration capacity. However, the steady state infiltration capacity measured in infiltrometer tests 2 to 6, during the period in 1997, falls within the range of infiltration capacity measured in infiltration tests 7 to 16, performed in June, 1998. During the second site visit in June, 1998, careful attention was paid to sealing the edges of infiltrometers 7 to 16, in addition to ponding a maximum of 5cm to 7cm of water. The average sorptivity was estimated from infiltrometer tests 7 to 16. Infiltration data from the first 25 minutes of ponding showed an appreciable period of non-linear infiltration and sufficient data was obtained to determine the field sorptivity. The range of slopes in the cumulative infiltration versus square root time graph varied between 1.4mm/min" and 6mm/min" and had a similar positively skewed distribution as compared to the distribution of steady state infiltration capacity (Figure 4-2). The average catchment response was estimatedfromthe arithmetic mean of fourteen steady state infiltration capacities and the arithmetic mean of ten sorptivities. The mean steady state infiltration capacity for Catchment 2 was 0.22mm/min with a standard error on the mean of 0.04mm/min. The mean sorptivity was estimated to be 1/2  1/2  1/9  •  1 /9  3.0mm/min with a standard error on the mean of0.6mm/min'" (Table 4-1). The standard error of the mean for any sample distribution is given as the standard deviation divided by the square root of the sample size. An average catchment response for infiltration under ponded conditions was generatedfromthe mean steady state infiltration capacity and sorptivity, as shown in Figure 4-la. This average catchment response will later be used to estimate runoff volumes in different rainstorms on the Claude pile. Sorptivity measurements on the surface of Claude east are consistent with published sorptivity values for a fine sand in its virgin state and a cultivated sandy loam. The sorptivities of the Bungendore fine sand and the Cowra sandy loam were measured at a slight suction in a ring tensiometer to eliminate the effect of macropores and were found 1/9  1/9  to have an average value of 3.1 mm/min and 1.7mm/min , respectively (Clothier and White, 1981). An estimate of the sorptivity on the surface of Claude east can also be calculated from assumed Green-Ampt parameters. Sorptivity was approximated using the following equation developed by Youngs (Maidment, 1993): 19  S = V(2(n-6i)KS ) where Sf is the Green-Ampt wetting front suction, mm; n is the porosity; 9j is the initial water content and K is the effective hydraulic conductivity, mm/min. An effective hydraulic conductivity of 0.2mm/min was assumed from the average of the steady state infiltration capacity of Catchment 2. A moisture deficit (n-0j) of 0.24 was determined from measurements of porosity (0.36) and maximum saturation (33% or 0.12) in Catchment 2. The Green-Ampt wetting front suction was taken from a chart relating USDA soil textures to wetting front suction (Appendix 6). The average grain size of the top 3cm of the surface of Claude east was measured to have 33% gravel, 44% sand and 23%) silt and clay, which correlates to a Green-Ampt wetting front suction of 18cm. The approximate sorptivity was then calculated to be 4.1mm/min , which fits very nicely within the standard error for the mean of sorptivity measurements in Catchment 2. f  1/2  4.4.2  Catchment 8  The cumulative infiltration for fourteen infiltrometer tests conducted in Catchment 8 is shown in Figure 4-4. Estimates of the steady state infiltration capacity and sorptivity for each infiltrometer are included in Table 4-2. Cumulative infiltration in Catchment 8 reached a steady state condition after a short period of ten to fifteen minutes, almost half the time for infiltration in Catchment 2 to reach steady state. The shorter period of capillarity-dominated flow is in agreement with higher infiltration rates observed in Catchment 8 and visual evidence of a slightly coarser grain size at the surface, as compared to Catchment 2. Over 60% of the steady state infiltration capacities were in a narrow range between 0.13mm/min and 0.28mm/min, as shown in the frequency histogram in Figure 45. The remainder of the steady state infiltration capacities were widely spread over increasing capacities up to 0.84mm/min. Sorptivity values in Catchment 8 ranged between 1.5mm/min and 6.4mm/min , which is the same range of sorptivities that were measured in Catchment 2. The distribution of both the sorptivity and steady state infiltration capacity is positively skewed, also similar to Catchment 2 (Figure 4-5). Waste rock around the outside of one ring infiltrometer became increasingly wet during ponding in test location 11, indicating non-vertical infiltration. Rapid early-time infiltration and nonsteady flow for over 20 minutes after ponding began are likely indicators that a leak had occurred beneath the edge of the ring infiltrometer (Figure 44a). For this reason, datafrominfiltrometer test 11 were discarded. The average catchment response was estimatedfromthe arithmetic mean of thirteen steady state infiltration rates and the average of thirteen field sorptivities. The average steady state infiltration capacity for Catchment 8 is estimated to be 0.34 mm/min with a standard error of 0.05mm/min (Table 4-2). The average early-time infiltration behavior (sorptivity) in Catchment 8 is estimated to be 3.2mm/min with a standard error of 0.4mrn/min . A plot of the average early-time and steady state behavior of Catchment 8 is shown in Figure 4-4. 1/2  1/2  1/2  20  4.4.3  Claude West  The first four infiltrometer tests performed on the west side of the Claude pile revealed a wide distribution of high infiltration capacities from 4mm/min to an incredible 200rnm/min, as shown in Figure 4-6. In just forty-five seconds, two 20L pails of water infiltrated through ring infiltrometer test 1. The very coarse and clean (negligible amount of fines) waste rock at the location of infiltrometer test 1 on Claude west is illustrated in Figure 4-7. Refer also to Figure l-3b to see the coarse surface of Claude west and the location of infiltrometer test 4. The implication of such high infiltration capacities on the surface of Claude west is such that all rainfall will immediately infiltrate. The cumulative infiltration from the start of ponding appears to be at a steady state, which suggests that gravity-driven flow is dominant from the start of ponding. An estimate of sorptivity on Claude west is unnecessary to characterize the early-time infiltration behavior. The average infiltration capacity measured on Claude west is 45mm/min and the standard error is greater than the mean, due to the wide variability of high infiltration capacities on Claude west and the small size of the sample population (Table 4-3). The effects of traffic compaction are demonstrated by infiltrometer test 5, installed in the wheel ruts from our truck (Figure 3-3). Infiltrometer test 5 reached a steady state infiltration capacity of O.lmm/min after 10 minutes of capillarity-driven flow. The sorptivity was estimated to be 1.25mm/min . This infiltration capacity is similar to the infiltration capacity measured on the compacted surface of Claude east. Since infiltrometer test 5 represents less than 0.1% of the surface area of Claude west, it was not included in the random sample population of infiltration capacity on Claude west. This example is used to demonstrate how the permeability of the waste rock on the Claude pile can be significantly altered, due to different pile construction methods at closure. Waste rock from the Claude pit is malleable, easily crushed and compacted due to traffic. 4.4.4  Summary of infiltrometer measurements  A summary of the statistical properties of measured infiltration capacities on Claude east and west is given in Table 4-3. The average steady state infiltration capacity in Catchment 2 and 8 is 0.22 +/- 0.04mm/min and 0.34 +/- 0.05mm/min, respectively. The rapid infiltration measured in infiltrometer test 1 (24.8mm/min) suggests that the surface of Claude east has indeed a bi-modal permeability. The average steady state infiltration capacity in Catchment 8 is 50% greater than Catchment 2. By comparison, Catchment 8 appears to be coarser on the surface. During a heavy rainfall event on June 25th, 1998, very little overland flow was observed in Catchment 8, as compared to Catchment 2 (Refer to Appendix 1 for rainstorm details). A high spatial variability in infiltration capacity was measured in the matrix material on the surface of Catchment 2 and 8. This is expected to be a general trend over the entire surface of Claude east. Infiltration in a drain is rapid and very differentfromthe slow infiltration capacity of the fine matrix waste rock, which covers the majority of the surface on Claude east. Mean infiltration capacity on Claude west was measured to be one-hundred times greater than the east side of the pile, consistent with observed textural differences (Refer to Appendix 4). Loose gravel and cobble sized waste rock with fewfinesallows water to infiltrate rapidly. A rainstorm with a 100-year return period would not cause surface  21  ponding even in the location of the lowest measurement of infiltration on Claude west (neglecting infiltrometer test 5, conducted in the wheel rut). Rainfall infiltrates as fast as it arrives on Claude west, thus it will always be a supply-controlled infiltration rate.  4.5  Small-Volume Dye Tracers in Ring Infiltrometers  The small-scale flow paths in the matrix material on the surface of Catchment 2 were investigated by releasing rhodamine WT at the end of infiltrometer tests 2, 3, 4 and 6, during the experimentation period in 1997. Dye tracer was released at the end of a period where steady state infiltration had been established. One important observation was that flow patterns did not diverge below the ring infiltrometer, confirming an approximate one-dimensional infiltration. Therefore, a larger diameter or double ring infiltrometer was not necessary to reduce the effects of diverging flow. Edge effects were only noticeable in the top 1cm below infiltrometer test 3, however these were due to a poor seal around the ring infiltrometer. Stained waste rock was excavated below the edge of infiltrometer test 3 to a depth of 1cm, after which edge effects were not visible. Excavation of waste rock below infiltrometer test 5 was not completed, due to limited time before a rainfall flushed and faded the stained patterns. All five dye tracer tests, in the matrix material on the surface of Claude east, exhibited decreasing dye coverage with depth, as shown in Figure 4-8. A decreasing cross-sectional area for infiltrating dye indicates significant preferential flow. Significant macropore flow was observed in infiltrometer tests 2 and 3. This may be the reason for the long non-linear period of rapid infiltration at the start of ponding, as discussed in Section 4.4.1. Dye coverage at the surface of infiltrometer test 2 and 3 was 100%, however removal of the top 0.2cm, of the dense compacted surface crust, revealed the uniform dye front did not migrate beyond the first millimeter (Figure 4-8). Excavation of infiltrometer test 2 and 3 showed that dye coverage decreased to approximately 14% over the area of the ring infiltrometer at an excavation depth of 0.2cm and was concentrated around seven to ten cobble and gravel-sized particles that were embedded in the dense surface. Refer to the photo of infiltrometer test 2 shown in Figure 4-9a and infiltrometer test 3 shown in Figure 4-9b. Further excavation revealed that dye flowed around the sides of these coarse particles streaking approximately 20% to 50%) of the coarse particles (Figure 4-9b). Below the restricting surface crust, the dye distribution increased to approximately 20% over the area of the ring infiltrometer at 5cm below infiltrometer test 2 and 1cm below infiltrometer test 3 (Figure 4-8). The dye bypassed the surface crust through the higher permeability pathways around the coarsefragmentsof waste rock and then spread slightly outward into the higher permeability matrix material. A deeper macropore network was observed below infiltrometer test 2 in the location shown by the dark dye stain in Figure 4-9a. The stained macropore network consisted of coarse clean waste rock from 3cm to 27cm. The stained pathway was excavated to a depth of 27cm, the bottom of which is shown in Figure 4-10. The excavation of waste rock from a depth below 10cm showed that all the dye tracer had converged into this one macropore network. Deeper excavation below 27cm depth was not feasible. Large particles close to the macropore network were observed to be heavily stained on one side of the waste rock (Figure 4-10).  22  Further excavation of infiltrometer 3 exposed similar trends of preferred flow through macropores. At 6cm below the surface of infiltrometer 3, dye migration was localized in two areas, representing 16% of the source area. Dye tracer was first observed to concentrate in a clean gravel zone at a depth of 6cm, below which, cobbles and gravel provided a macropore network for dye to migrate to 16cm below the surface of the pile. The second area where dye migration occurred was in the macropores along the surface and through fractures of a 20cm boulder (Figure 4-11). The boulder transmitted dye over a depth from 1cm to 19cm below the surface of infiltrometer test 3. Waste rock below the boulder was also heavily stained, however no further excavation below 19cm was completed. In summary, dye tracer migrated to an approximate depth of 19cm through two macropore networks, representing 6% of the original source area. Dye tracer in infiltrometer test 4 showed no signs of major macropore flow, however, a uniform dyefront(with fully coated fine particles of waste rock) was observed to migrate to a depth of 0.8cm by Darcy or matrix flow (Figure 4-12a). Recall that the infiltration capacity of infiltrometer 4 was approximately one-third of the infiltration capacity measured in infiltrometers 2 and 3, likely due to the lack of macropores below infiltrometer 4. Further excavation showed that dye coverage decreased from 85% at 0.5cm below the surface to 18% at a depth of 1cm (Figure 4-12b). Dye distribution rapidly decreased to 2% at a depth of 3.5cm (Figure 4-8). This shallow infiltration of dye suggests that water is held in capillary tension close to the surface, where it is expected to evaporate before significant flow and transport can occur. The excavation of dye in infiltrometer 6 was unfinished due to an untimely rainfall that washed away the dye tracer. Infiltrometer 6 revealed a similar distribution of dye as compared to infiltrometer 4 with an even distribution of dye in the top 1cm. Dye coverage was 80% at 0.5cm and decreased to 40% at 3cm depth below the surface (Figure 4-8). The penetration depth of the dye or the presence of macropores below a depth of 3 cm is unknown. Dye migration below infiltrometer 1 and 2 on Claude west was observed to migrate to depths of at least 60cm below the surface of the pile (Figure 4-13). Excavation below the infiltrometers on Claude west revealed large macropores in a clast supported structure of coarse gravel and cobbles, with very few fines. Excavation of infiltrometer 1 showed that the 8cm of ponded dye penetrated to a depth of at least 60cm, where dye coverage was significant, approximately 85% of the source area (Figure 4-14). Coarse particles were randomly streaked with dye. Further excavation was not feasible. The high infiltration capacity measured on Claude west is consistent with deeper dye penetration. The excavation of infiltrometer 2 revealed dye penetration to 65cm, in which 18%> of the source area was stained with dye. The small amount of fines below infiltrometer 1 and 2 were not stained, indicating that the infiltrating dye flowed only in the coarse fraction of waste rock and avoided the fines. The infiltrating dye avoided the small amount of fine waste rock material (unstained). No surface crust was evident during excavation of the waste rock. Infiltrometer 1 and 2 showed significant dye coverage over a depth up to 50cm in infiltrometer 1 and 18cm in infiltrometer 2. Recall that infiltrometer test 1 had a great infiltration capacity of 200mm/min,  23  4.6  Summary of small-volume dye tracers in ring infiltrometers  Significant macropore flow was recorded in two of the four small-scale dye tracer tests performed on the surface of Catchment 2 on Claude east. Each dye tracer test on Claude east showed that dye coverage decreased with depth, concentrating in the macropores. Dye coverage decreased from 100% at the pile surface to a mean dye coverage of 18% at 3 cm below the surface crust. The 8cm of ponded dye tracer penetrated to variable depths, ranging from approximately 3.5cm (infiltometer 4) to 27cm (infiltrometer 2 and 3). This depth is within the evaporative zone, therefore a portion of this infiltrated water is expected to evaporate. Considering that the pile height is 30m and a number of traffic surfaces impede vertical migration through the matrix material, it is expected that pore water near the surface of the catchments on Claude east is held in storage for a long period of time before recharging the water table. Pore water may eventually be released when the moisture content rises in the matrix material, in response to another rainfall. Two mechanisms of flow were observed in the matrix material on the surface of Claude east. Matrix or Darcy flow was observed in some portion of all the small volume dye tracer tests on Claude east, particularly in the top 1cm of the pile (Figure 4-12a). Preferential flow in the coarser sand-sized particles was most often observed in the shallow depth of Claude east (Figure 4-12b). However, this is considered a type of matrix flow due to heterogeneity. The second mechanism was macropore flow observed in rock fractures, between the matrix material and coarse particles and in the larger voids between coarse waste rock particles. A schematic of these mechanisms observed in the shallow surface of Claude east is shown in Figure 4-15. All four small-volume dye tracer tests performed on Claude east revealed some macropore flow below the top 1cm of the pile surface. The small volume dye tracer tests on Claude west displayed more uniform dye distribution with depth, however, dye distribution was observed in streaks, which indicates the macropores were not filled during infiltration. Dye coverage was recorded to decrease from 100% at the surface to approximately 90% at 10cm below the surface. The infiltrating dye appeared to migrate vertically, rather than converging in the coarser areas, as observed on Claude east. The same volume of dye, as compared to the tests on Claude east, migrated beyond 60cm below the surface of the pile on Claude west. Infiltrating water appears to flow over the surface of coarse particles and from contact to contact between the large particles, here named 'boulder hopping'. The mechanisms of flow in the shallow surface of Claude west appear to be macropore flow, and boulder hopping, in which capillarity has no effect and pore spaces never fill. 4.7  Limitations of Infiltrometer Testing  One of the limitations in infiltrometer testing is that the fixed ring infiltrometer tends to force entrapped air out of macropores in the soil profile, providing a rapid transmission of flow and a bypass through the soil matrix, as per the discussion in Chapter 2.2. In addition, a high ponded pressure can cause piping and erosion through the surface crust and between gaps in the waste rock, creating an interconnected  24  macropore system (R. Sidle, 1998). The ponded ring infiltrometer tends to overestimate infiltration capacity, due to the higher pressure head applied at the surface. Driving ring infiltrometers into the waste rock involves a degree of disturbance to the surface integrity, often at the level of enthusiasm of the driver, which may cause an overestimation of the true infiltration capacity. Systematic overestimation of rainfall infiltration can also result from infiltrometer testing, due to the lateral divergence of flow below the ring infiltrometer. However, vertical infiltration was indeed documented in the single ring infiltrometer tests, validating the use of a single ring infiltrometer of sixty centimeters diameter. The lower permeability surface crust on Claude east helps to ensure vertical flow. An interesting effect of infiltration into layered profiles with a lower permeability unit at the surface is that true vertical flow can be expected, regardless of the diameter of a ring infiltrometer (H.Bouwer, 1986). The effect of flow divergence in the underlying unsaturated material has a negligible effect on the infiltration rate inside the infiltrometer, for infiltrometers greater than 30cm in diameter (H.Bouwer, 1986). 4.8  Rainfall Partition Between Infiltration and Runoff on Claude East  This section provides measurements and predictions of the rainfall partition between surface infiltration and runoff to drains on Claude east. The results of the runoff experiment in Catchment 7 are discussed in Section 4.8.1, in which the partition between infiltration and runoff was measured from four different rainfall events on June 25th, 26th, and July 18th and 24th, 1998. Refer to section 3.1 for a summary of the experiment set-up. These events are typical Cluff Lake rainfall events representing both long and short duration events during the summer thunderstorm season. Compared to the extreme rainfall statistics for Cluff Lake (Appendix 2), these rainfall events have a return period of less than two years. The only way to estimate how partitioning will vary with rainfall intensity was to compare the infiltration capacity measured on the waste rock pile to different rainfall intensities. In the following section (4.8.2), infiltrometer measurements in Catchment 2 and 8 were used to predict the partition between infiltration and runoff for different types of rainfall events in which the peak intensity occurs at different times during the storm. The rainfall event on July 7 , 1997 has a five-year return period and a peak rainfall intensity that occurs late in the storm. Alternatively, the rainfall event on August 13 , 1997 has a peak intensity that occurs within the first five minutes of the storm; this is also an intense storm with a return period of two-years. No direct runoff data are available for these storm events on July 7 and August 13 , therefore predictions were used to demonstrate how the partition between infiltration and runoff varies for different storm types. Estimates of runoff in Catchment 2 and 8 were also made for the June 25 and 26 events for comparison to measured values of runoff in Catchment 7. The rainfall events in 1997 were measured by a rain gauge 3km away, while the 1998 rainfall intensities were measured on the Claude pile. th  th  th  th  th  th  4.8.1  Direct measure of runoff in Catchment 7  The type of rainfall event that occurred on June 25 , 1998 is common during the summer thunderstorm season. The total rainfall was 10.9mm over 121 minutes, however th  25  three peaks of high intensity rainfall occurred, during which time 8.6mm fell over 22 minutes (Figure 4-16a). The surface of the pile was fairly dry before this rainfall event, as the last occurrence of rainfall was a light sprinkle the day before (Appendix 1). A summary of the measured runoff and estimated infiltration in Catchment 7 is shown in Table 4-4. Runoff to the barrel at the bottom of Catchment 7 was measured to be 158L, which is approximately 0.8mm distributed over the area of the catchment (195m ). The estimated partition between infiltration and runoff to the drain of Catchment 7 was 91% and 9%, respectively. The assumption made in this statement is that runoff was generated uniformly across Catchment 7. Observations of the surface of the catchment suggest there are coarse zones with higher infiltration capacity, which likely did not contribute to runoff during this storm event. The fine versus coarse zones are visible in Catchment 7, as shown in Figure 3-1. For low intensity storms, the source areas for runoff are the fine zones in which rainfall intensity is greater than the infiltration capacity. In any event, the important implication of this is that the majority of the rainfall event on June 25 recharged the matrix material on the surface of the pile and 158L (9% of the total rainfall) recharged the drain as rapid gravity driven inflow. If 50% of the water that infiltrated the matrix material is assumed to evaporate, then only 45% of the rainfall on June 25 is estimated to recharge deep into the pile through the matrix material. This assumption for 50% evaporation is supported by the early findings of Craig Nichol, who has measured the flow balance in a small-scale pile constructed of waste rock from the same mine site as the Claude pile (C. Nichol, 2001). All runoff entering the drain is assumed to escape below the evaporative zone and recharge deep into the pile. The average infiltration capacity of Catchment 7 can be estimated if we assume that ponding occurred during the intense period, where 8.6mm fell over 22 minutes. Assuming that the runoff amount of 0.8mm occurred at this time, the infiltration into the matrix material on the surface was 7.9mm over 22minutes. Thus, the average infiltration capacity for Catchment 7 is approximately 0.45mm/min. The June 25 rainfall intensity rarely exceeded 0.45 mm/min (Figure 4-16a), therefore, it is apparent that zones of higher infiltration capacities exist within the catchment, where ponding did not occur. The source of ponding and runoff during this event depended on the lower permeability material. The surface texture in Catchment 7 was observed to have very fine-grained waste rock on the east side of the catchment and sediment deposited along the flow path to the drain (Figure 3-1). A similar storm duration occurred on the following day, however the intensity of the June 26th rainfall event was relatively low. Before the start of this rainfall event the surface of the pile was initially wetted, due to light rainshowers in the morning. A maximum rainfall intensity of 0.2mm/min was recorded, which is about 20% of the intensity on June 25th, as shown in Figure 4-16a. A total of 7.1mm of rainfall was recorded over 100 minutes by the rain gauge set up next to Catchment 7 (Table 4-4). The total volume of rain that fell within the 195m area was 1400L and the volume of runoff measured in the barrel at the bottom of the catchment was 72L. The partition of rainfall to runoff was approximately 5% or 0.4mm. In this lower intensity rainfall, 95% of the water was taken up in the matrix material on the surface of the Claude pile. Less infiltration was expected because of the antecedent moisture conditions, however the low th  th  th  26  intensity of the rainfall did not overcome the infiltration capacity of the majority of the material within the catchment. Regions of very fine-grained waste rock on the east side of the catchment and along flow paths to the drain are the likely contributors to runoff during this low intensity rainfall event. Measurements of runoff were even lower for two very short rainfall events on July 18th and 24th, 1998 (Figure 4-16b). These rainfall events are characterized by a high intensity, (greater than 0.2mm/min) during the first five minutes of rainfall. A total rainfall of 2mm fell over 17 minutes on July 18th, however, 1.4 mm occurred in the first 5 minutes. The volume of rainwater that fell within Catchment 7 was approximately 400L and a negligible amount of 4L was measured to runoff into the drain. In summary, 99% of the rainfall infiltrated the matrix material on the surface of the pile during this short duration and moderate intensity rainfall event. Measurements of the runoff rate were obtained during this rainfall event thanks to Craig Nichol, who was on the Claude pile at this time. The runoff rate to the drain (barrel) in Catchment 7 was measured to be 1.2L/min and 0.2L/min at 13 minutes and 23 minutes, respectively, after the start of rainfall. Runoff flowing into the drain was still occuring 6 minutes after the end of the rainfall event on July 18 , which suggests a lag period of at least 6 minutes for runoff from the top of the catchment to reach the drain. The rainfall intensity recorded on July 24th was extremely high, however the storm duration was less than 6 minutes and only 4.4mm of rainfall was recorded (Figure 4-16b). The antecedent moisture conditions were relatively dry, as a week of dry weather conditions preceded this rainfall event. The total volume of rain that fell within Catchment 7 was approximately 860L, of which 5.5L flowed overland into the drain (barrel). A lag period for runoff to reach the drain and for ponding to occur was observed, as runoff began flowing into the drain 9 minutes after the onset of rain, even though rainfall ended after 6 minutes. The flow rate into the drain was measured to be 1 L/min, 10 minutes after the onset of rain. Similar to the July 18th rainfall, the partition between infiltration and runoff was 99% and 1% (Table 4-4). A high capillary potential exists in the dry matrix material at the start of wetting, thus the average infiltration capacity was greater than rainfall intensity in these short duration events. These results suggest that the majority of the rain water is taken up by the surface matrix waste rock during short duration rainfall events. However, somewhere along a connected flow path to the drain, the rainfall intensity exceeded infiltration capacity to generate some runoff to the barrel. th  4.8.2  Predictions of rainfall partition in Catchment 2 and 8 of Claude east  Predictions of the rainfall response in Catchment 2 and 8 are approximated from the mean infiltration capacity, determined from fixed ring infiltrometer testing. The maximum and minimum estimate of the average infiltration capacity of Catchment 2 and 8 are plotted against rainfall events from June 25 and 26 , 1998 (Figure 4-16a) and events from July 7th and August 13th, 1997 (Figure 4-17). Recall that the 1998 rainfall events were recorded on the top of the pile, whereas the 1997 events were recorded 3 km from the pile. The amount of excess rainfall is estimated from the area between the infiltration capacity curve and the rainfall intensity. For simplicity, it is assumed that all th  th  27  excess rainfall flowed overland to the drains in each catchment. Comparing the direct measure of runoff from the June 25th and 26th events to the predicted runoff evaluate this simplification. The transient part of the infiltration capacity curve (early time infiltration) is calculated from the slope of the cumulative infiltration curve, where cumulative infiltration is SoVt. Therefore, the early time infiltration rate (mm/min) can be represented by V2 Sc/Vt (the derivative of the cumulative infiltration). The maximum estimate of infiltration capacity is determined from the maximum sorptivity and steady state infiltration capacity. Refer to Section 4.4.4 for an explanation of the estimated maximum and minimum infiltration parameters. The average infiltration capacities for Catchment 2 and 8, determined from ring infiltrometer testing, are applied to the June 25 storm, in which direct measurements of runoff in Catchment 7 were made. The mean infiltration parameters for Catchment 2 (i = 0.22 +/- 0.04mm/min and So = 3.0 +/- 0.6mm/min ) yield rainfall runoff in the range between 12% and 20% of the total rainfall. Mean infiltration parameters in Catchment 8 are higher (i = 0.34 +/- 0.05mm/min and So = 3.6 +/- 0.4mm/min ) and yield a slightly lower runoff volume in the range between 9% and 12% of the total rainfall. Predictions of the partition of rainfall to runoff are similar to measured runoff volumes (8%) in Catchment 7, for this high intensity storm on June 25 . Given the uncertainty in estimating a mean infiltration capacity through infiltrometer testing and the natural variability of waste rock between different catchments, the average infiltration capacity measured in Catchment 2 and 8 appears to provide a reasonable approximation to runoff from the June 25 storm. The average infiltration capacity curve from Catchment 2 and 8 compared to the low intensity rainfall on June 26 would suggest no runoff was likely (see Table 4-5). The predicted runoff in Catchment 2 and 8 is zero, however, 5% of the rainfall in Catchment 7 was measured as runoff into the drain. More runoff is expected in Catchment 2, since the surface texture of Catchment 2 was observed to have a greater proportion of low permeability areas, as compared to Catchment 7. It is expected that the spatial average of infiltrometer tests performed in each catchment does not represent the low permeability areas of the catchment that generate overland flow in low intensity storms. Runoff on June 26 was likely generated from partial areas of lower permeability, i.e. below average permeability areas that generate ponding during low intensity rainfall events. As a result, it is expected that using the mean infiltration capacity from infiltrometers may not provide the total overland flow and runoff to the drain during low intensity rainfall events. Since there are no direct measurements of rainfall runoff for different types of extreme rainfall events at Cluff Lake, we must rely on predictions. The rainfall events on July 7th and August 13th, 1997, represent very high intensity events that are likely contributors to major recharge deep into the Claude pile. No direct runoff data are available, however these events are important to provide insight into how significantly different the runoff partition is for high intensity rainfall events, and events with peak intensities that occur early versus late in a rainfall event. The rainfall event on July 7th began with a moderate intensity, followed by a peak intensity after 34 minutes, where 10mm of water fell over 10 minutes, as shown in Figure 4-17. Over the duration of 50 th  ss  ss  th  th  th  th  28  minutes, a total of 18mm of rainfall occurred. Extreme rainfall statistics for Cluff Lake show a one-hour storm with 18.3mm of rainfall to have a return period of 5 years (Appendix 2). Assuming that a steady state infiltration in the waste rock was reached before the peak of the storm, mean infiltration parameters for Catchment 2 and 8 yield a prediction of runoff to the drain of 46% and 42%, respectively (Table 4-5). Based on the size of each catchment, the volume that recharged the drain in the July 7th storm is estimated to be over 3000L for Catchment 2 and 500L for Catchment 8. This is a significant amount of recharge to the drains. During the August 13 , 1997 rainfall event, ponding and overland flow was observed to begin in a number of catchment areas, within the first 5 minutes of rainfall (although no direct measures of runoff were made). The peak intensity on August 13 occurred early in the storm, as shown in Figure 4-17. Within the first 21 minutes 10.2mm of water fell. The total rainfall was 11.6mm over a period of 41 minutes, which is approximately a storm with a 2-year return period (Appendix 2). The estimated volume of runoff to recharge the drains was predicted to be in the range between 6% and 14%) for Catchment 2 and between 4% and 6% for Catchment 8 (Table 4-5). Based on the size of each catchment, the volume of runoff to the drains is predicted to be 1000L for Catchment 2 and 100L for the smaller Catchment 8. The majority of rainfall from the August 13 rainfall event is predicted to infiltrate the matrix material on the surface of the Claude pile and less than 15% is predicted to recharge the drains. However, approximately 50% of the rainfall on July 7th is predicted to runoff into the drains. Due to the high initial infiltration capacity of the waste rock when it is dry, less runoff is expected at the start of a rainfall event. Recall that the peak intensity of the July 7th storm occurred after a longer initial period of wetting the surface of the pile (Figure 4-17). The difference in the predicted rainfall partition is due to the occurrence of a peak intensity at an early versus late period of wetting the matrix material on the surface of the waste rock pile. th  th  th  4.9  Summary of Rainfall  Distribution  The partition between rainfall and runoff depends on both variations in the rainfall intensity and the spatial pattern of infiltration capacity and the initial saturation of the waste rock. Fixed ring infiltrometer testing can provide a means of predicting infiltration in the matrix material of the pile surface and runoff to a drain for rainfall events. The partition of a rainfall event to runoff may be underestimated for low intensity rainfall events, since the average of infiltrometer testing in each catchment may not represent the partial areas of low permeability that generate runoff in low intensity events. In a typical rainfall event during the thunderstorm season, the majority of rainfall (over 90%>) is taken up by the waste rock matrix material on the surface of the Claude pile. The excess rainfall that flows overland and concentrates in the small drain area can still be significant. For example, 160L (approximately 10%) of the rainfall) was measured to flow into the drain of Catchment 7 in just a couple hours during a common rainfall event on June 25 , 1998. High intensity and long duration rainfall events cause significant recharge to a drain. In a high intensity rainfall event, recharge to a drain was estimated to be as high as th  29  50% (predicted from the 5-year return storm on July 7 , 1997). During the July 7 storm, 3000L was estimated to recharge the drain of Catchment 2. It is expected that a significant portion of the water infiltrating the matrix material is held in residual capillary tension close to the surface of the pile and is lost to evaporation. The portion of water that does not evaporate is expected to migrate further downward or be released to drains when moisture content rises in the medium. The net recharge into the waste rock pile from the drains versus the matrix material can be roughly estimated by assuming 50% of the water infiltrating the matrix material will be lost to evaporation after the rainfall stops, and the surface dries. All the water that infiltrates the drains is assumed to recharge deeper into the pile, below the evaporative zone. For example, in the July 7 storm, 54% of the rainfall was estimated to infiltrate the matrix material, therefore the net recharge into the pile through the matrix material was estimated to be 27%, versus 46% that recharged through the drain (based on the infiltration capacity of Catchment 2). The more typical rainfall event that occurred on June 25 , 1998 had a different distribution of recharge, in which 43% of the rainfall was estimated to recharge the pile through the matrix material, and 15% was estimated to recharge through the drain (42% was assumed to evaporate from the surface of the pile). The net recharge, estimated from the less intense rainfall event on June 25 , was 58% of the total rainfall, whereas 73% of the total rainfall was estimated to recharge the pile from the more intense event measured on July 7 . In summary, greater recharge into the pile is expected to occur through the fast pathways of the pile, or drains, during high intensity rainfall events, and through the matrix material during low intensity events. The major transport of fluids deep into the pile is expected to occur during intense rainfall events. th  th  th  th  30  Chapter 5. Large-Scale Dye Tracer Tests 5.1  Field Methods  Three large-scale dye tracer experiments were conducted to investigate the pathways for infiltration and flow within the coarse drains of Claude east and the coarse and clean waste rock on Claude west. The same concentration of rhodamine WT that was used for the small-scale dye tracer tests was applied in the large-scale tests (see Section 3.3 for a discussion of dye testing). Tracer release 1 and 2 were applied as point sources into two catchment drains on Claude east, identified during the August 13 , 1997 rainfall event. The third dye tracer release was distributed over a small 4m area chosen randomly on Claude west. The dye tracer applied to the two catchment drains on Claude east was released from the point source outlet of a hose situated upslope of the entry to the drain (Figure 51). A valve on the hose, which was connected to an elevated 55 gallon drum containing rhodamine WT solution, controlled the rate of release into the two drains on Claude east. A continuous flow of dye tracer was released by refilling the reservoir with dye tracer. Surface fines were not disturbed along the flow path to the drain, due to the slow release rate and careful placement of the outlet of the hose. Tracer release 1 was applied in the drain of a small catchment of size 88m and is located as shown in Figure 1-2 (Site 1). Tracer release 1 consisted of releasing 800L over a period of 4.5 hours, which is equivalent to 9.1mm of runoff from an area the size of Catchment 1. This rate of recharge is realistic with respect to the natural events occurring at Cluff Lake, because if an average runoff partition of 20% is assumed (for a storm with an intensity greater than 0.2mm/min) this runoff would be equivalent to a storm with a minimum of a 25 year return period. Tracer release in the drain of Catchment 1 entered the pile in two large macropores at the surface, as pointed to in Figure 5-1. Dye tracer did not pond during the release, but swirled around each macropore before disappearing, much like a sink drain. Tracer release 2 was applied in the drain of Catchment 2, which was the large catchment used for infiltrometer testing (location shown in Figure 1-2). The drain of Catchment 2 was photographed during a rainstorm, as shown in Figure 1-4. Tracer release 2 consisted of 2400L of rhodamine WT solution released at a slower rate over a period of 9 hours. This release volume is equivalent to 6mm of runoff from an area the size of Catchment 2 (387m ). The equivalent storm to generate this volume of runoff to the drain (assuming a runoff partition of 20%) would be one with a 2 year return period. The rate of release of the distributed tracer on the surface of Claude west was limited by the water source. A volume of 1300 L of rhodamine WT was mixed in a hydroseeder (large reservoir connected to a spray pump) and applied by spray over an approximate evenly distributed area of 2m x 2m. Tarps were stretched around a fence surrounding the source area to contain the dye tracer during spraying. The volume of dye tracer is equivalent to 32cm and the duration of the distributed release was only 13 minutes. Unfortunately, this tracer was released at an unrealistic rate for Cluff Lake, representing a rainfall event with a return period that is off the scale of extreme rainfall statistics for Cluff Lake (see Appendix 2). th  2  2  31  Backhoe excavation followed each tracer release, and occurred over a period between two and four days to find and map all the dye tracer. Typically, horizontal lifts were excavated in intervals between 0.3m and 0.5m, however each lift size was influenced by the size of boulders present in the waste rock. The opportunity for excavation along the side of a stained channel or coarse section of waste rock was available during excavation of a deep channel in Claude west and coarse and clean waste rock in tracer release 1. 5.2  Waste Rock Texture and Moisture Below the Pile Surface  Frequent samples were taken during excavation of the three tracer tests. A total of 10 samples were analyzed for texture as shown in Appendix 5b. Waste rock sampling procedures were biased, because the largest particles were often eliminated from the sample for ease of handling and sieving. In addition, the largest sieve had an opening of 25mm. The waste rock texture inside the pile is highly heterogeneous, as shown in the photograph in Figure 5-2. Recall that the Claude pile was constructed by free dumping and grading in separate lifts. This photo, taken during excavation of the distributed dye tracer on Claude west, shows coarse waste rock (up to 0.5m) piled beside a finer grained waste rock overlain by a 20cm layer of compacted and crushed waste rock that was the previous traffic surface. A total of three waste rock samples were taken from the compacted old traffic surface excavated during each tracer experiment. The old traffic surfaces were encountered at an approximate depth of 3m, 0.4m and 0.8m during excavation of tracer test 1, 2 and Claude west, respectively. These samples taken at the old traffic surface showed a greater fraction of fines (between 5% and 15% greater), compared to all other samples below the pile surface (Appendix 5b). The thickness of the compacted and crushed layer of waste rock varied between 20cm and 40cm as shown in Figure 5-2 and 5-3. During the sieve analysis, these samples were difficult to break apart by water or mechanically, due to the extreme compaction and cohesiveness of the clay fraction. Moisture content was measured in the ten samples by weight and varied from dry to nearly saturated. Since the excavation of the dye tracers took place over many days with an open pit, the moisture contents are not representative of in-situ conditions. The moisture content of the samples was generally between 4.3% and 7.3%, with the exception of the samples taken from old traffic surfaces within the pile. The samples from the traffic surfaces were very moist, or saturated. Fully saturated waste rock was found on an old traffic surface at 0.45m depth below the surface of the pile, below tracer release 2. Water flowed into a small pit dug into the surface of the traffic compacted layer. The sample from this compacted layer had a moisture content of 14%. Whereas, 0.5m below the traffic surface, the waste rock was completely dry.  32  5.3  Results of Large-Scale  Tracer Tests  5.3.1  Tracer Test in Catchment 1 on Claude East  Excavation and mapping of the dye tracer followed the day after the tracer release. The top 0.5m of the surface of the pile was excavated. The newly exposed surface appeared to be uncompacted sandy gravel with boulders, different from the compacted and fine surface of Claude east. The dye at 0.5m depth was distributed in two locations within a 0.4m area. Dark staining was observed in a coarse clean waste rock, where all sides of the gravel-sized waste rock was coated. A slight speckling of dye was observed in the fine matrix material on the surface of a large diameter boulder. One centimeter below the clean and fine matrix material revealed dye stains on the surface of the large 1.35m diameter boulder, as shown in Figure 5-5. It appears that the dye tracer flowed over the surface of the boulder, avoiding the matrix material almost entirely. The underside of the large boulder was darkly stained with dye coating 60% of the boulder, as shown in Figure 5-6. At an excavation depth of 1.12m below the surface (immediately below the large boulder), dye was distributed within a i m area. Dye stains were found on the sides of sixteen cobble to boulder-sized waste rock particles. The dye was most concentrated on the underside of the boulders and covered an average 30% of the surface of each particle. A trace amount of dye tracer was observed in the fine matrix material. The excavated surface exposed at 1.54m below the surface of the pile exhibited a similar pattern, however the footprint of the dye area increased to 2m . No definite single channel was observed. Dye coverage is on the surface of the cobbles and small boulders. During excavation below 1.12m, a 75cm boulder was uncovered to reveal intense dye stains over one side of the boulder. At an excavation depth of 1.69m, dye is localized in one area (1.2m x 0.4m) of cobbles directly beneath the source area of the dye tracer. A cross-section was achieved by excavating deeply along the side of the stained cobbles. At a depth between 1.69m and 2.55m, cobble-sized waste rock with little to no fines was exposed to reveal streaks where dye appeared to migrate vertically over the surface of the cobbles, as shown in Figure 5-7 and 5-8. On average, 20% of the waste rock particle was coated with dye. The small fractions of fine matrix material located next to the cobbles remained unstained (Figure 5-8). It appears that dye had moved from cobble to cobble by way of the contact point between boulders. This characteristic can be described as boulder hopping. A finer texture of waste rock material is encountered at an excavation depth of 2.55m. The matrix material at this depth is a loosely packed and well-graded mix of fine sand to gravel with trace cobbles. The dye tracer is more uniformly spread over an area of 2m directly beneath the source area. Dye coated both the loose matrix material and cobble-sized waste rock. Dye migration ended at 3.0m depth below the surface, where a previous traffic surface was encountered. The dye tracer was observed to be distributed over a minimum area of 3.5m x 2.0m. It is likely that the area of distributed tracer was slightly larger, however not all stained waste rock could be observed at this depth due to the methods of 2  2  2  2  33  excavation. A cross-section was excavated into the compacted surface to reveal a distinct ponding surface for the tracer, as evident by the horizontal extent of tracer (2m in length) shown in Figure 5-3. Waste rock was very moist at the surface of the tightly compacted layer of waste rock. Below the traffic surface, waste rock was observed to be dry and loosely compacted (Figure 5-3). It should also be noted that waste rock within the pile was dry in areas one meter lateral to the stained waste rock. The following calculation answers the question: "Does the extent of dye migration observed account for the volume of dye tracer released into the drain of Catchment 1?". An estimate of the residual volume of water in the stained waste rock was compared to the volume of tracer released (800L). The residual volume of water is estimated by the following equation: Residual Volume = volume of stained waste rock x n x 0 Where, n = porosity and 6 = residual water content. The volume of waste rock stained by infiltrating dye was approximately 1.5mxl.5mx3.0m (Table 5-2). An approximate porosity (0.4) and residual water content (0.2) was assumed for the waste rock. The residual volume is then estimated to be 540L, which is less than the volume released. It is expected that a portion of the dye tracer was taken up in the previous traffic surface during ponding at 3.0m depth. Recall that the final horizontal distribution of dye tracer was measured to be greater than 3.5mx2.0m. W  W  5.3.2  Tracer Test in Catchment 2 on Claude East  The drain of Catchment 2 was excavated the following day after the dye tracer release. Excavation to a depth of 0.35m below ground surface revealed similar behavior to tracer test 1, in which mainly cobbles and some gravel were stained in streaks. Dye covered approximately 30% to 100% of each coarse waste rock particle. The small amount of matrix material at this depth was loosely packed and unstained. The few large 'blocky' boulders located near the source were overturned to reveal heavy staining on the undersides of each boulder. A 60cm boulder that protruded from the top of the pile was 20cm lateral to the drain. Dark dye streaks were found on the underside of the boulder, showing evidence that the dye flowed along the bottom surface of the boulder to a horizontal distance of 60cm. At a depth of approximately 0.5m, a previous traffic surface was encountered. Infiltrating dye appeared to pond and spread laterally on the tightly compacted layer. A large area was excavated to delineate the edge of the dye distribution, as shown in Figure 5-9. Dye tracer stains were darkest on the coarsest waste rock particles, but were also found in the sands and gravel sized particles located on top of the previous traffic surface. Dye tracer was observed to migrate as far as 5.5m south and 4.5m north from the point source location. All sides and bottom of the excavation were cut back further in search of the rhodamine dye tracer. No further dye tracer was uncovered below the previous traffic surface. The vertical extent of dye migration reached a maximum depth of 0.62m, within the compacted traffic layer. Figure 5-10 shows a cross-section of the ponded dye and the tightly compacted waste rock layer, which has the appearance of concrete. The final extent of lateral spreading was measured to be 10.0m x 4.5m (Figure 5-9). It is unknown  34  what the ultimate fate of the tracer water was. The volume of dyed rock accounts for only half of the tracer by residual saturation calculations. Similar to Tracer test 1, it is expected that a portion of the dye tracer was taken up in the previous traffic surface during ponding at 0.5m depth. The volume of dye tracer released in Catchment 2 on Claude East was three times greater than the release in Catchment 1, in hopes that deeper infiltration within the pile could be mapped (Table 5-1). Recall that no ponding in the drain was observed during the intense 2-year return storm in August, 1997. Therefore, it was a surprise to discover the tracer only migrated to a final average depth of 0.5m below the surface. 5.3.3  Distributed tracer test on Claude West  Excavation of the dye tracer on Claude west followed in the afternoon of the distributed release and took about four days to excavate and map all the dye tracer. Recall that the source zone for the infiltrating dye was 2.0m x 2.0m. A cross-section was cut to a depth of 1.0m along one side of the source zone. The dye tracer appeared to migrate vertically below the source zone, decreasing slightly in area (to 1.9m x 1.9m) from the original footprint area. The exposed waste rock was very loose and coarse, with cobbles supporting each other and loose gravelly sand partially filling the voids between boulders and cobbles. There was no evidence of a crushed or compacted layer at the pile surface. Three large 0.5m boulders were removed from the top 1.0m of the pile. All waste rock particles were fully coated within the top 1.0m below the source zone. This may be due to the large volume of dye applied at a fast rate (Table 5-1) over the surface of Claude west. No dye was observed in the silty-sand matrix material, located below 1.0m depth. A ponding layer was encountered at this depth. The extent of ponding was determined by cutting back into the walls of the excavation, while making observations of cross-sections into the compacted traffic layer. Dye spread laterally in all directions from the location of the source zone. The ponded dye spread over an area of 13.8m x 10.0m and ranged in depth between 0.75m and 1.0m below the surface of the pile (Figure 5-11). Further dye migration below the old traffic surface occurred in a number of locations and typically penetrated between 30cm and 60cm. Six locations, in which dye penetrated through the compacted layer, were mapped and photographed. Figure 5-11 shows these locations. A large amount of dye was found to penetrate through the traffic surface in a single channel and migrate to a significant depth of 4.0m. The entrance to this deep channel appeared to be an old "drain" on the previous traffic surface, in which two small boulders created a large gap in the compacted traffic surface. The entrance of the channel was located approximately 2.7m north-west of the source zone and at a depth of 0.9m below the surface (location A). A cross-section was excavated along the side of the channel and is shown in Figure 5-12. As shown, the dye tracer migrated vertically, below the old compacted traffic surface, for 1.0m over the surface of large haphazardly stacked boulders. This can be called boulder hopping, where dye flows over the surface of large boulders and continues at the contact points between boulders. At an approximate depth of 2.0m below the surface of the pile, the dye tracer encountered a finer waste rock material sloped at a 45 degree angle (Figure 5-12). The dye tracer flowed along the  35  interface between the coarse and fine waste rock material to a final depth of 4.0m below the surface of the pile (Figure 5-12). Dye tracer was not observed to penetrate into the fine waste rock material. The fine waste rock material was relatively dry, whereas, the boulder-sized waste rock was moist. The five other locations where dye penetrated the old traffic surface are further described here. Dye was observed to penetrate to 30cm below the old traffic surface (1.3m below the pile surface) by way of macropore flow around coarse cobbles in the tightly compacted traffic layer (location B). Dye migrated to the same depth in location C, however a single large boulder provided a surface for dye to flow around. Dye was not observed in the matrix material compacted around the sides of the boulder. Figure 5-13 shows dye in the macropores of a fractured 50cm boulder, in which dye penetrated to 50cm below the compacted layer (1.5m below the pile surface). The 50cm boulder was removed and heavy dye staining within the footprint was observed. The dye didn't penetrate the matrix material beyond the footprint. Also shown in Figure 5-13 is a small macropore near the surface of the tightly compacted layer (right of the meter stick), which allowed dye to migrate 5cm. This was a typical observation made when cutting sections in the old compacted traffic surface. Near the edge of the ponded extent (location F), a cross-section was revealed to show coarse waste rock in slightly looser matrix material, where dye migrated to lm below the old traffic surface (2m below the pile surface). Dye was only observed on the coarse fractions of waste rock. The average lateral extent of the dye tracer was 2.0m x 2.0m, not including lateral spreading due to ponding on a previous traffic surface. The extent of dye coverage observed (4.0m x 2.0m x 2.0m) would account for the volume of dye tracer released on Claude west, based on residual volume calculations described previously. The residual volume of dye tracer was calculated to be 1280L, assuming a porosity of 0.4 and a residual water content of 0.2. Therefore, the total volume of dye tracer released on Claude west (1300L) was visually mapped and recorded during excavation of the dye tracer. 5.4  Summary of Large-Scale Dye Tracer Tests  In general, dye migrated in a vertical direction over the surface of coarse waste rock particles and avoided the fine matrix material. Dye coverage was observed on the surface of cobbles and boulders. Where the dye tracer did not intercept cobbles or boulders, dye stained the coarser fractions of the matrix material, such as gravel and coarse sand. Dye tracer did not diverge away from the source during migration, except where a traffic surface was encountered. The final depth of migration in the large-scale dye tracer tests was 3.0m, 0.5m and 4.0m for Tracer test 1, 2 and Claude west, respectively. On Claude west, the average lateral extent of dye tracer was similar to the distributed source area (approximately 2m x 2m). The average lateral extent of dye tracer observed in Tracer test 1 was approximately 1.5m x 1.5m (Table 5-2). Significant lateral spreading was caused by ponding on old traffic surfaces encountered between 0.5m and 3.0m below the pile surface. Dye spread 14m laterally on an old traffic surface encountered at 1.0m below the surface of Claude west. A small amount of lateral  36  spreading can be attributed to the surface of large boulders encountered below the source zone, as observed on the 1.35m boulder excavated in Tracer test 1. A sloped interface between coarse and fine waste rock also caused some lateral spreading, as documented in the Claude west tracer test. No definite single channel for transmitting dye was identified in Tracer test 1 and 2. However, a continuous channel was documented by photograph during excavation of the distributed tracer on Claude west, which allowed dye to migrate to 4.0m depth. The dye tracer moved through coarse fractions of waste rock, mostly flowing over the surface of the waste rock. Two mechanisms of flow were observed during excavation of the large-scale tracer tests, including flow in macropores, and boulder hopping. The latter mechanism was best observed in the deep channel excavated on Claude west, in which dye flows over the surface of large boulders and continues at the contact points between boulders. There was no evidence of matrix or Darcy flow in the top lm below the pile surface on Claude west. However, Darcy flow was observed where the dye tracer ponded on a previous traffic surface.  37  Chapter 6. Conclusions  The texture of waste rock at the surface of the rock pile is key to the infiltration capacity, sorptivity, and, of course, the geometry of fluid pathways. The surface of Claude west is characterized by a coarse 'untrafficked' surface of loose gravel, cobblesized waste rock and few fines, in which infiltration at the surface is immediate in even the most extreme rainfall events. A finer compacted traffic surface and the topography of the pile surface control recharge on the east side of the Claude pile. Intense rainstorms cause ponding and overland flow to concentrate in coarse drains at the bottom of small catchment areas, which vary in area from 20m to 370m for those that were mapped. A high degree of spatial variability in infiltration capacity was measured in the matrix material on the surface of Catchment 2 and 8. The average steady state infiltration capacity in Catchment 2 and 8 is 0.22 +/- 0.04mm/min and 0.34 +/- 0.05mm/min, respectively. Rapid infiltration (24.8mm/min) was measured in one infiltrometer in Catchment 2, which suggests that the surface of Claude east indeed has a bi-modal permeability. Infiltration in a drain is rapid and very different from the slow infiltration capacity of thefinematrix waste rock, which covers the majority of the surface on Claude east. Mean infiltration capacity on Claude west was measured to be one-hundred times greater than the east side of the pile, consistent with observed textural differences. All five small-volume dye tracer tests in the matrix material on the surface of Claude east, exhibited decreasing dye coverage with depth, indicating significant preferential flow. Dye coverage decreased from 100% at the pile surface to a mean dye coverage of 18% at 3 cm below the surface crust. The ponded dye tracer (8cm of dye in the infiltrometers) penetrated to variable depths below the infiltrometers at the surface of Claude east, ranging from approximately 3.5cm to 27cm. These shallow penetration depths likely result in water in the matrix material near the surface of Claude east to be held in storage and partially evaporate before the next rainfall event. Two mechanisms of flow were observed in the matrix material on the surface of Claude east from the small-volume tracer tests. Matrix or Darcy flow was observed in the fine waste rock in the shallow surface. Flow typically concentrates in the coarser sand-sized particles immediately below the surface of Claude east. This type of preferential flow is also considered matrix flow due to heterogeneity. The second mechanism was macropore flow observed in rock fractures, in the voids between matrix material and coarse particles, and in the larger voids between coarse waste rock particles. Macropore flow was observed more often than matrix flow. The small volume dye tracer tests on Claude west displayed more uniform dye distribution with depth. Dye coverage was recorded to decrease from 100% at the surface to approximately 90% at 10cm below the surface. Infiltrating water appears to flow over the surface of coarse particles from contact to contact between the large particles, where macropores are only partially saturated. The mechanisms of flow in the shallow surface of Claude west appear to be macropore flow, and flow over the surface of large particles, in which capillarity has no effect and pore spaces never fill. The partition between rainfall and runoff to a drain depends on variations in the rainfall intensity and the spatial pattern of infiltration capacity and the initial saturation of the waste rock. In a typical low intensity rainfall event, the majority of rainfall (over  38  90%) is taken up by the waste rock matrix material on the surface of Claude east. However, only 45% is expected to recharge deep into the pile, because half of the pore water in the matrix material is expected to evaporate. In a high intensity rainfall event, recharge to a drain was estimated to be as high as 50% (estimated from the 5-year return storm on July 7 , 1997). During the July 7 storm, 3000L was estimated to recharge the drain of Catchment 2. The net recharge from the typical storm on June 25 , 1998 was estimated to be 43% to the matrix material on the surface and 15% to the drains, based on the infiltration capacity of Catchment 2 (the remainder was assumed to be lost to evaporation). A more extreme rainfall event, such as the storm on July 7 , 1997, provides a different partition between matrix infiltration and runoff. The net recharge into the pile from the storm on July 7 was estimated to be 27% to the matrix material, versus 46% that recharged to the drain. A greater proportion of rainfall recharged deeper into the pile during the July 7 storm (73%), as compared to the less intense storm on Jun 25 (58%). In summary, the major transport of fluids into the pile is expected to occur through the fast pathways of the pile, or drains, during high intensity rainfall events. The large-scale dye tracer tests showed that water infiltrating through drains migrates in a vertical direction over the surface of coarse waste rock particles and avoids the fine matrix material. Dye coverage was observed on the surface of cobbles and boulders. Where the dye tracer did not intercept cobbles or boulders, dye stained the coarser fractions of the matrix material, such as gravel and coarse sand. Dye tracer did not diverge away from the source during migration, except where a traffic surface was encountered. Significant lateral spreading was caused by ponding on old traffic surfaces encountered between 0.5m and 3.0m below the pile surface. Dye was observed to spread between 4m and 14m laterally on old traffic surfaces. No definite single channel for transmitting dye was identified in Tracer test 1 and 2. However, a continuous channel was documented by photograph during excavation of the distributed tracer on Claude west, which allowed dye to migrate to 4.0m depth. Two mechanisms of flow were observed during excavation of the large-scale tracer tests, macropore flow and boulder hopping, which is infiltrating water flowing over the surface of large boulders and continuing at the contact points between boulders. There was little evidence of matrix or Darcy flow below the surface of the pile, except where the tracer ponded on an older traffic surface. th  th  th  th  th  th  th  39  References  A n d r e i n i , M . S . , T . S . Steenhuis. Preferential Paths o f F l o w under Conventional and Conservation Tillage. Geoderma V . 4 6 pp.85-102. Berndtsson, R. and M . Larson. 1987. Spatial variability o f infiltration in a semi-arid environment. Journal o f H y d r o l o g y V . 9 0 pp.117-133. Beven. K . and P. Germann. 1982. Macropores and water f l o w in soils. Water Resources Research V.18(5) pp. 1311-1325. Burgy, R . H . and J . N . Luthin. 1956. A test o f the single and double ring types o f infiltrometers. A m e r i c a n Geophysical U n i o n Transactions V . 3 7 pp. 189-191. Clothier, B . E . and I. White. 1981. Measurement o f sorptivity and soil water diffusivity in the field. S o i l Science Society o f A m e r i c a Journal V . 4 5 pp.241-245. Clothier, B . E . and I. White and G . J . Hamilton. 1981. Constant-rate rainfall infiltration: F i e l d experiments. S o i l Science Society o f A m e r i c a Journal V . 4 5 pp.245-249. Corey, John C . 1968. Evaluation o f dyes for tracing water movement in acid soils. S o i l Science V . 1 0 6 pp. 182-187. D a v i s , S . N . , G . M . Thompson, H . W . Bentley, and G.Stiles. 1980. Ground-water tracers-A short review. Ground Water V . 1 8 pp. 14-23. Diodato, D . M . and R . R . Parizek. 1994. Unsaturated Hydrogeologic Properties o f R e c l a i m e d C o a l Strip M i n e s . G r o u n d Water V . 3 2 ( l ) pp. 108-118. Dingman, S . L . 1994. Physical Hydrology. M a c m i l l a n College Publishing Company. N Y , U S A . Elboushi, Ismail M . 1975. Amount o f water needed to initiate f l o w in rubbly rock particles. Journal o f Hydrology V . 2 7 pp.275-284. Eriksson, N . , A . Gupta and G . Destouni. 1997. Comparative analysis o f laboratory and field tracer tests for investigating preferential f l o w and transport in mining waste rock. Journal o f H y d r o l o g y V . 1 9 4 pp. 143-163. Flury, M . , H . Fluhler, W . A . Jury and J . Leuenberger. 1994. Susceptibility o f soils to preferential f l o w o f water: A field study. Water Resources Research V . 3 0 pp. 1945-1954. Flury, M . and H . Fluhler. 1994. Brilliant B l u e F C F as a dye tracer for solute transport studies - A toxicological overview. Journal o f Environment Quality V . 2 3 pp.1108-1112. Flury, M . and H . Fluhler. 1995. Tracer Characteristics o f Brilliant B l u e F C F . S o i l Sciences V . 5 9 pp.22-27. Germann, P. and K . B e v e n . 1981. Water f l o w in soil macropores 1. A n experimental approach. Journal o f S o i l Science. V . 3 2 pp. 1-13. Ghodrati, M . and W . A . Jury. 1990. A field study using dyes to characterize preferential f l o w o f water. S o i l Sciences V . 5 4 pp. 1558-1563. Hatano, R. and H . W . G . Booltink. 1992. U s i n g fractal dimensions o f stained f l o w patterns in a clay soil to predict bypass flow. Journal o f Hydrology V . 135 pp. 121-131.  40  H i l l e l , D. 1980. Applications o f S o i l Physics. A c a d e m i c Press, Inc. N e w Y o r k . pp.2.5-2.73. H i l l e l , D. and Gardiner, W . R . 1970. Transient infiltration into crust-topped profiles. S o i l Science. V . 109(2) pp.69-76. K l u t e , A . 1986. Methods o f Soils A n a l v s i s . Part 1. Physical and M i n e r a l o g i c a l Methods (Second Edition). A m e r i c a n Society o f A g r o n o m y - S o i l Science Society o f A m e r i c a . W l , U S A . pp. 822-844. Loague, K . 1990. Simple design for simultaneous steady-state infiltration experiments with ring infiltrometers. Water Resources Bulletin V.26(6) pp.935-937. Loague, K. and G . Gander. 1990. Spatial variability o f infiltration on a small rangeland catchment. Water Resources Research V.26(5) pp.957-971. Maidment, D. 1992. Handbook o f Hydrology. Infiltration and S o i l Water Movement pp.5.1-5.49. M e r z , B . and E . J . Plate. 1997. A n analysis o f the effects o f spatial variability o f soil and soil moisture on runoff. Water Resource Research V.33(12) pp.2909-2922. M o r i n , K . A . , N . Hurt, S . G . Hutyt. 1997. History o f Eskay Creek M i n e ' s waste-rock dump from placement to disassembly. Canadian M E N D Report, Government o f Canada, Ottawa. N i c h o l , Craig. 2001. Personal communication Omoti, V . and A . W i l d . 1979a. U s e o f fluorescent dyes to mark the pathways o f solute movement through soils under leaching conditions, 1.Laboratory experiments. S o i l Science V . 1 2 8 pp.28-33. Omoti, V . and A . W i l d . 1979b. U s e o f fluorescent dyes to mark the pathways o f solute movement through soils under leaching conditions, 2. F i e l d measurements. S o i l Science V . 1 2 8 pp.95-104. Pang, L., M . Close and M . N o o n a n . 1998. Rhodamine W T and Bacillus subtilis Transport through an A l l u v i a l G r a v e l Aquifer. Groundwater V . 3 6 ( l ) pp.112-122. Smettem, K . R . J . and N . C o l l i s - G e o r g e . 1985. Statistical characterization o f soil biopores using a soil peel method. Geoderma V . 3 6 pp.27-36. Smettem, K . R . J . and S.T. T r u d g i l l . 1983. A n evaluation o f some fluorescent and non-fluorescent dyes in the identification o f water transmission routes in soils. Journal o f S o i l Science V . 3 4 pp.45-56. Smith, L., D. L o p e z , R. B e c k i e , K. M o r i n , R. Dawson and W . Price. 1995. Hydrogeology o f Waste R o c k Dumps, F i n a l Report. Report by University o f British C o l u m b i a to Department o f Natural Resources Canada, File N o . 23440-4-1317/01-SQ. 130 p. Sullivan, M . , J.J. W a r w i c k and S . W . Tyler. 1996. Quantifying and delineating spatial variations o f surface infiltration in a small watershed. Journal o f Hydrology V.181 pp.149-168. Talsma, T. 1969. In situ measurement o f sorptivity. Australian Journal o f S o i l Research V . 7 pp.269-276. Trudgill, S.T., A . M . Pickles, K . R . J . Smettem, R . W . Crabtree. 1983. Soil-water residence time and solute uptake. l . D y e tracing and rainfall events. Journal o f Hydrology V . 6 0 pp.257-279. T r u d g i l l , S.T., A . M . Pickles, K . R . J . Smettem, R . W . Crabtree. 1983. Soil-water residence time and solute uptake. 2.Dye tracing and preferential flow predictions. Journal o f H y d r o l o g y V . 6 2 pp.279-285.  41  White, I. and M . J . S u l l y . 1987. Macroscopic and microscopic capillary length and time scales from field infiltration. Water Resources Research V . 2 3 pp.1514-1522. White, I. and K . M . Perroux. 1987. Use o f sorptivity to determine field soil hydraulic properties. S o i l Science Society o f A m e r i c a Journal. V.51(5) pp. 1093-1101. Wildenschild, D. and K . Jensen, K . V i l l h o l t h and T. Illangasekare. 1994. A laboratory analysis o f the effect o f macropores on solute transport. G r o u n d Water. V.32(3) pp.381-389.  42  Appendix 1: Rainfall record at Cluff Lake for May to August, 1998  80.000 70.000  June 25 event  60.000 50.000  i n  rainfall  40.000 30.000 20.000 10.000 o.ooo h i R 11 03  2  i  rfl w i i  03  2  Total Rainfall at Cluff Lake for May to August, 1998  c Et E  2.5  June 25  2  rainfall  event  £l.5  0.5 ^4oo  a> 2•  LO  ~) 4  CO CO  oo Cp  OO  3  3  D)  CN  CO  -)  —>  Rainfall intensity at Cluff Lake for May to August, 1998 Note: This rainfall data was collected at Cluff Lake's weather station located 3kmfromthe Claude Waste Rock Pile.  43  Appendix 2: Rainfall statistics for extreme events at Cluff Lake  R E T U R N PERIOD (y) IRaini*a 11 \ Pit rapri 2 35.9  M hr •6'hz  IS  20  25  50  100  i 55.0  59.3'  62.3  64.6  7.1.7  8.7  48',2  52.1  54.9  58.0  63.6  0.1  31.5  36:7  39.7.  43.4  48.3  3.2  22.5  ; ?J:6  31 X  35.6  95  47.3  'iri.s ~4U'~ B.5-  2 hr l'hr  • •5 ' 10  13.0. 10.1  15.0  78  1.1,6  n.9  30.6'  21.8  23.8  25.6  , 2 b, 3 iy.O'  2.8  ' i'B.a  20.O  21.3  ' 22.2  25.3  83  1 6.5 , 17.3  19,6  2.0  1  M.I , IS'.S-  44  Appendix 3: Evaluation of visual dye tracers The following dyes were tested for mobility and visibility in a laboratory experiment. D y e solutions were ponded on the surface o f a column o f unsaturated silty and gravelly sand. T h e column was a clear plastic with a diameter o f 4 c m and length o f 0.5m with a mesh screen fastened to the bottom. T h e concentration o f the applied dyes varied as per experimentation and recommendations from previous literature (See Smetten and T r u d g i l l , 1983; Smetten and Collis-George, 1985; Pang, Close and N o o n a n , 1998; and F l u r y and Fuhler, 1995). A 60ml solution o f dye was ponded on the surface o f the unsaturated soil column and a dye front was observed to move through the column. In the case o f methylene, brilliant and Chicago blue, a saturated wetting front was observed to separate from the dye front, indicating that sorption is retarding the movement o f the dye. F o r example, the migration front o f methylene blue dye had moved 15cm while the saturated wetting front had moved 4 0 c m after 20 minutes o f infiltration. The leading edge o f a m o v i n g water front is missed b y these dyes with the exception o f A c i d red 1 and Rhodamine W T . Rhodamine W T has a very good visibility and was coincident with the wetting front. The f o l l o w i n g is an evaluation o f the mobility and visibilty o f dyes tested.  Concentration  Mobility  Visibility  Methylene blue  0.5g/l  low  high  Brilliant blue G  lg/1  average  high  0.5 to lg/1  average  average  Dye  Chicago blue A c i d red 1 Rhodamine W T  2g/l  high  low  1.25ml/L o f a 5 % solution o f rhodamine  high  high  45  a a E <U U U  s  I  ii Qi  T3  3  C3 O JS •M «tH  O  VI  'an >>  "8 cs « .a .fi  '3 sO  Tt"  X  c  a  o  <  -»—'  co O  CM  E o  E o  -«—'  ro O  O  CN  CO  Infill  E  4—  c  <u E o re  O  c  d>  E  £  LU LU LU 3  a1 3 Q 3  Q  %3 5 3  o  CO  o  o  CN  o  CM  Infill  •3 c  HHHH c  c  c  c  o o  CM  O O  O  00  o  CO  o  sseiu Aq uein jeuji juoojod  o CM  ci  o CD  sseiu Aq ueqj jauy juacuad  Appendix 5: USD A soil textures to wetting front suction  Sand (•%)  (Source: Maidment, D. 1992. Handbook of Hydrology. Infiltration and Soil Water Movement p.5.37)  Table 4-1: Infiltration parameters estimated from infiltrometer testing in Catchment 2. I n f i l t r o m e t e r test  Steady state i n f i l t r a t i o n c a p a c i t y (mm/min)  Sorptivity (mm/minl/2)  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16  n/a 24.8 0.33 0.27 0.12 n/a 0.21 0.32 0.04 0.07 0.08 0.27 0.12 0.69 0.20 0.28 0.14  2.9 1.6 1.4 1.6 1.8 4.8 6.2 1.8 2.6 5.3  0.22 0.17 0.16 0.04  3.0 2.6 1.8 0.56  Mean Geometric mean Standard deviation Standard error  Table 4-2: Infiltration parameters estimated from infiltrometer testing in Catchment 8. I n f i l t r o m e t e r test  Steady state i n f i l t r a t i o n c a p a c i t y (mm/min)  Sorptivity (mm/minl/2)  1 2 3 4  0.13 0.25 0.17 0.84 0.45 0.25 0.28 0.28 0.27 0.49 n/a 0.20 0.34 0.55  2.0 2.5 1.5 6.4 5.2 2.5 3.3 3.2 2.6 3.7 n/a 3.1 4.9 5.4  0.34 0.30 0.19 0.05  3.6 3.3 1.5 0.4  5 6 7 8 9 10 11 12 13 14 Mean Geometric mean Standard deviation Standard error  Table 4-3: Statistical properties of steady state infiltration and sorptivity. Steady state infiltration (mm/min)  Catchment 2  Catchment 8  C l a u d e west  N u m b e r o f measurements Mean Median Standard deviation Standard error o f mean M e a n + standard error (max estimate o f mean) M e a n - standard error (min estimate o f mean)  14 0.22 0.20 0.16 0.04 0.26  13 0.34 0.28 0.19 0.05 0.40  4 45 13 96 48 n/a  0.18  0.29  n/a  Catchment 2  Catchment 8  10 3.0 1.8 0.56 3.6  13 3.6 1.5 0.41 4.0  2.4  3.2  Sorptivity (mm/min ) m  Number o f measurements Mean Standard deviation Standard error o f mean M e a n + standard error (max estimate o f mean) M e a n - standard error (min estimate o f mean)  Note: Outliers and suspect tests that were eliminated are not shown in this table.  Table 4-4: Summary of measured runoff to the drain of Catchment 7. Runoff Experiment  26-June-98  18-July-98  24-July-98  10.9mm  7.1mm  2mm  4.4mm  Length o f Storm:  121min  lOOmin  17min  6min  Duration o f high intensity>0.2mm/min: Rainfall during high intensity  22.5min  0  5min  6min  8.6mm  n/a  1.4mm  4.4mm  158L  72L  4L  5.5L  Total rain:  V o l u m e runoff to barrel: A r e a o f Catchment 7:  25-June-98  195m  2  195m  2  195m  2  195m  2  R u n o f f to barrel:  0.8mm  0.4mm  0.02mm  0.03mm  Infiltration to matrix:  10.1mm  6.7mm  1.98mm  4.37mm  % Recharge to drain:  9% 46%  5% 48%  1% 50%  1% 50%  % Deep recharge through matrix material:*  * A s s u m i n g 5 0 % o f water in the matrix material evaporates  6o  Table 4-5: Summary of predicted distribution of recharge to catchment surface and drains in a rainfall event. Predictions of runoff from infiltrometer data Catchment 2 (3 87rri) Steady state infiltration capacity:  Mean  Min  Max  0.22mm/min  0.18mm/min  0.26mm/min  0.34mm/min  0.29mm/min  0.40mm/min  *  *  *  2.2mm  1.3mm  Catchment 8 (76 m ) 2  Steady state infiltration capacity:  *  *  *  2 5 - J u n e - 9 8 total rainfall:  10.9mm  Length o f storm:  121min  Catchment 2-Excess rain:  1.6mm  % Recharge to drain:  15%  20%  12%  % Deep recharge through matrix material:*  43%  40%  44%  1.3mm  1.3mm  0.95mm  % Recharge to drain:  Catchment 8-Excess rain:  12%  12%  9%  % Deep recharge through matrix material:*  44%  44%  46%  0.08mm  0  2 6 - J u n e - 9 8 Rainfall event:  7.8mm  Length o f storm:  1 OOmin  Catchment 2-Excess rain:  0  % Recharge to drain:  0%  1%  0%  % Deep recharge through matrix material:*  50%  50%  50%  0  0  0  % Recharge to drain:  0%  0%  0%  % Deep recharge through matrix material:*  50%  50%  50%  Catchment 8-Excess rain:  7 - J u l y - 9 7 Rainfall event:  18mm  Length o f Storm:  50min  Catchment 2-Excess rain:  (5 year return - intensity peaks later)  8.3mm  8.9mm  7.8mm  % Recharge to drain:  46%  % Deep recharge through matrix material:*  27%  49% 26%  43% 28%  7.5mm  7.9mm  6.7mm  % Recharge to drain:  42%  % Deep recharge through matrix material:*  29%  44% 28%  37% 28%  Catchment 8-Excess rain:  (2 year return - intensity peaks early)  1 3 - A u g - 9 7 Rainfall event:  11.6mm  Length o f Storm:  41 min  Catchment 2-Excess rain:  2.5mm  3.8mm  1.8mm  9% 46%  14% 43%  6% 47%  1,3mm  1.8mm  1.1mm  % Recharge to drain:  5%  % Deep recharge through matrix material:*  48%  6% 47%  47%  % Recharge to drain: % Deep recharge through matrix material:* Catchment 8-Excess rain:  * A s s u m i n g 5 0 % o f water in the matrix material evaporates  4%  Table 5-1: Summary of field methods for large-scale dye tracer tests. Tracer Test Area  Volume of tracer  Period of release  Catchment size  Equivalent runoff per area of Catchment  Catchment 1  800L  4.5 hrs  88m  9.1mm  Catchment 2  2400L  9hrs  387m  1300L  13 m i n  -  Claude west - 4 m area 2  2  2  6.2mm 30cm  Table 5-2: Summary of observations for large-scale dye tracer tests. Tracer test AreaF i n a l  t r a c e r depth A v e r a g e l a t e r a l extent o f dye (not i n c l u d i n g p o n d i n g zones)  Catchment 1  3.0m  1.5 x 1.5m  Catchment 2  0.5m  n/a*  Claude West  4.0m  2.0 x 2 . 0 m  2  2  D e p t h to previous traffic surface  Horizontal distribution of dye o n p r e v i o u s t r a f f i c surfaces  3.0m  3.5m x 2.0m  0.5m  10.0m x 4.5m  1.0m  14.0m x 10.0m  * V e r t i c a l dye migration was too short to determine an average  52  Time (T) Figure 2-1: Conceptual diagram of steady state infiltration and sorptivity.  A n g l e iron f o r stabiltiy Ruler  30 c m  60 c m  re 3-2: Fixed ring infiltrometer experiment set-up.  0  20  40  60  80  100  120  140  L a p s e d time (min)  Figure 4-1: Cumulative infiltration in Catchment 2 of Claude east, a) Long time steady-state behaviour and b) early time capillarity dominated flow.  63  5  fi 3 u  0-0.1  0.1-0.2  0.2-0.3  0.3-0.4  0.4-0.5  0.5-0.6 0.6-0.7  Steady State Infiltration Capacity (mm/min)  6 y 5 -4 -  1  I  1  0 -I  h-l cb  L+J CN  -A  L, ro <N  (J •f cS  j  I  L+J in 4  |  1  1_|  L+J CD  in  N-  to  Sorptivity (mm/min- ) 1/2  Figure 4-2: Frequency histogram for infiltration parameters measured in Catchment 2. a) Steady state infiltration capacity and b) field sorptivity.  80  T  Figure 4-4: Cumulative infiltration in Catchment 8 of Claude east, a) Long time steadystate behaviour and b) early time capillarity-dominated flow.  7  j  6 5 0)  o o  3+  O  2 + 1 0  -I  1 1 1 1 1 1 [ 0-0.1 0.1-0.2 0.2-0.3 0.3-0.4 0.4-0.5 0.5-0.6 0.6-0.7 0.7-0.8 0.8-0.9 Steady State Infiltration Capacity (mm/min)  5 j 4 cu  _  o 3 2 3  u 2 O  CO I  m  CM  C£>  Sorptivity (mm/min- ) 1/2  Figure 4-5: Frequency histogram for infiltration parameters measured in Catchment 8. a) Steady state infiltration capacity and b) field sorptivity.  160 W Infilt test 1  Infilt test 5  30  40  50  70  60  80  100  90  T i m e (min)  (A at  E o £  2  240mnVrrin 6  8  10  12  14  Infiltration capacity (mm/min)  Figure 4-6: Infiltration on Claude west, a) Cumulative infiltration and b) infiltration capacity.  (B8  Figure 4-7: Location of infiltrometer test 1 on Claude west.  Dye tracer coverage, %  D  D.  10 -  c  t e r ock  "O  1  3 25 -  in  tracer coverage, %  I  Infiltrometer test 4  40  60  80  10C  o-  " .c  15 20 -  20  100  80  (cm)  5 -  60  iste r ock di  (cm)  o-  40  e  Dye tracer coverage, %  Dye tracer coverage, %  20  v  5 10 15 20 -  Infiltrometer test 6  3 25 X Horizontal cross-section/ Excavation surface  Figure 4-8: Vertical dye distribution in infiltrometer tests in Catchment 2.  Figure 4-10: Stained macropore network at 27cm below infiltrometer test 2.  ft  Dye tracer coverage, %  Dye tracer coverage, %  Figure 4-13: Vertical dye distribution in infiltrometer tests on Claude west.  M a c r o p o r e flow  Figure 4-15: Schematic of flow mechanisms in shallow surface of Claude East.  (IHUI/UIUI) uouejijuui JO Aijsuaiui  (uiui/uiui) u o i i e j i u j u i JO  Ajisu9ju|  81  Figure 5-2: Textural differences in the Claude Waste Rock Pile. (Claude west)  Figure 5-3: Compacted and crushed waste rock at old traffic surface. (Tracer test 1 at 3m below surface)  88  O ^cC%yg, o ^oCS^ 0  a  0  °  o  o  °  Ponding on previous — ~ ~ traffic surface  3.0m b.s.  Figure 5-4: Schematic of Tracer test 1.  C\0  Unstained matrix material  Figure 5-8: Boulder hopping - Tracer Test 1.  Figure 5-9: Final distribution of Tracer test 2. (0.5m below pile surface)  (0.98m)  D  C  B / /  _ ,.  Outline of source zone  -Jj (0.96m)  (0.76m)  2.0 m  (0.77m)  (0.88m)  13.8m  F i g u r e 5-11: E x t e n t o f p o n d i n g o n C l a u d e west. ( N u m b e r s i n brackets indicate e x c a v a t i o n depth b e l o w p i l e surface)  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089913/manifest

Comment

Related Items