UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Comprehensive review of plunge pool performance at four of BC Hydro dam sites and assessment of scour… Monfette, Maryse 2004

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

Item Metadata

Download

Media
831-ubc_2004-0713.pdf [ 26.29MB ]
Metadata
JSON: 831-1.0063412.json
JSON-LD: 831-1.0063412-ld.json
RDF/XML (Pretty): 831-1.0063412-rdf.xml
RDF/JSON: 831-1.0063412-rdf.json
Turtle: 831-1.0063412-turtle.txt
N-Triples: 831-1.0063412-rdf-ntriples.txt
Original Record: 831-1.0063412-source.json
Full Text
831-1.0063412-fulltext.txt
Citation
831-1.0063412.ris

Full Text

COMPREHENSIVE REVIEW OF PLUNGE POOL PERFORMANCE AT FOUR OF BC HYDRO DAM SITES AND ASSESSMENT OF SCOUR EXTENT by MARYSE MONFETTE B.A.Sc. (Geological Engineering), Universite Laval, 1999 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES CIVIL ENGINEERING THE UNIVERSITY OF BRITISH COLUMBIA June 2 0 04 ® Maryse Monfette, 2004 ABSTRACT The plunge pool performance at four of BC Hydro dam sites is reviewed. The sites of Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam are described in terms spillway design, plunge pool geology, and historical spillway outflows. The main factors susceptible to affect plunge pool scour are examined and the importance of each is established through a comparative analysis between sites. The spillway layout affects the plunge pool scour development mainly through the design unit discharge and total head drop for a given outflow and reservoir level. The plunge pool scour configuration is affected by the prevailing geological conditions. The magnitude of spillway flows appears predominant over the frequency and duration of spills. Observations support the concept of equilibrium conditions. A total of 14 sets of scour data were assembled from the review of plunge pool performance at the four sites of study. This information was used to evaluate the conventional methods of scour assessment such as hydraulic model studies and empirical equations in comparison with a new approach called the Erodibility Index Method. Ten well-known empirical formulas were tested. The Erodibility Index Method was applied methodically to the four sites of study and a sensitivity analysis was performed. The results of downstream scour tests from small-scale model studies were comparable to prototype observations in one of the four study cases. The empirical equations were seen to provide on average conservative values for design, but the great variability in results for a single data set and the inconsistency of a given formula from one site to another are problematic. The expression by Damle (1966) was seen to give the best combination of precision and accuracy with a standard error of estimate of 16 ft. The performance of the Erodibility Index Method in the assessment of maximum scour depth was disappointing and did not outclass the conventional methods. The Erodibility Index Method was characterized by a tendency for underestimation and a standard error of estimate of 53 ft. The main weakness of the approach is that the vertical distribution of power available for scour in the plunge pool is essentially related to the submerged jet velocity profile which do not reflect the changing magnitude of spillway discharges. TABLE OF CONTENTS Abstract ii Table of Contents iii List of Tables v List of Figures vi Acknowledgements viii CHAPTER I INTRODUCTION 1 CHAPTER II LITERATURE REVIEW 5 2.1 PHYSICAL PROCESSES 5 2.1.1 Free-Trajectory  Jets  6 2.1.2 Jet  Diffusion  and Hydraulic  Action Within  the Plunge Pool. . 8 2.1.3 Scour Mechanisms  10 2.2 PLUNGE POOL SCOUR ASSESSMENT 12 2.2.1 Hydraulic  Model  Studies 13 2.2.2 Empirical Formulas  14 2.2.3 The Dam Foundation  Erosion Study 16 CHAPTER III DAM SITES DESCRIPTION 22 3.1 PEACE CANYON DAM 22 3.1.1 General Arrangement  22 3.1.2 Spillway  Characteristics  23 3.1.3 Plunge Pool Geology  24 3.1.4 Historical  Spills 25 3.1.5 Scour Hole Development 2 6 3.2 SEVEN MILE DAM 2 8 3.2.1 General Arrangement  2 8 3.2.2 Spillway  Characteristics  2 9 3.2.3 Plunge  Pool Geology  3 0 3.2.4 Historical  Spills 32 3.2.5 Scour Hole Development 3 3 3.3 PORTAGE MOUNTAIN PROJECT 3 5 3.3.1 General Arrangement  3 5 3.3.2 Spillway  Characteristics  36 3.3.3 Plunge  Pool Geology  . 37 3.3.4 Historical  Spills 38 3.3.5 Scour Hole Development 3 9 3.4 REVELSTOKE DAM 4 0 3.4.1 General Arrangement  40 3.4.2 Spillway  Characteristics  4 0 3.4.3 Plunge  Pool Geology  41 3.4.4 Historical  Spills 43 3.4.5 Scour Hole Development 4 3 CHAPTER IV COMPARATIVE  ANALYSIS OF FACTORS AFFECTING PLUNGE POOL SCOUR . 88 4.1 SPILLWAY CHARACTERISTICS 88 4.2 PLUNGE POOL GEOLOGY 91 4.3 SPILLWAY DISCHARGES 93 4.4 SCOUR RATE 96 4.5 TAILRACE IMPROVEMENTS 99 CHAPTER V CONVENTIONAL ASSESSMENT OF PLUNGE POOL SCOUR 105 5.1 HYDRAULIC MODEL STUDIES 105 5.1.1 Peace Canyon Dam 105 5.1.2 Seven Mile  Dam 106 5.1.3 Portage  Mountain  Project 107 5.1.4 Revel stoke Dam 108 5.2 SCOUR DEPTH EMPIRICAL FORMULAS 110 5.2.1 Descriptive  List of Equations  110 5.2.2 Performance  of Empirical Formulas  113 CHAPTER VI THE ERODIBILITY INDEX METHOD 13 6 6.1 CONCEPTUAL APPROACH 136 6.2 PRACTICAL APPLICATIONS 13 8 6.2.1 Erodibility  Index Characterization  138 6.2.2 Jet  Hydraulics  - Computational Methods  139 6.2.3 Maximum  Scour Depth 141 6.3 SENSITIVITY ANALYSIS 143 6.3.1 Erodibility  Index 144 6.3.2 Jet  Velocity  Profile  145 6.3.3 Jet  Impact Velocity  146 6.3.4 Jet  Air Concentration at Impact 148 6.4 EVALUATION 14 9 CHAPTER VII CONCLUSION 177 7.1 LIMITATIONS 177 7.2 CONCLUSIONS 180 7.3 RECOMMENDATIONS AND EXPECTATIONS 182 Selected Bibliography 185 Appendix I Erodibility Index Charts 193 Appendix II Plunge Pool Surveys 197 Table 3.1: - Peace Canyon Dam/Spillway Characteristics 46 Table 3.2: Peace Canyon Dam/Historical Spills 47 Table 3.3: Peace Canyon Dam/Scour Hole Development 48 Table 3.4: Seven Mile Dam/Spillway Characteristics 49 Table 3.5: Seven Mile Dam/Historical Spills 50 Table 3.6: Seven Mile Dam/Scour Hole Development 51 Table 3.7: Portage Mountain Project/Spillway Characteristics 52 Table 3.8: Portage Mountain Project/Historical Spills 53 Table 3.9: Portage Mountain Project/Scour Hole Development 54 Table 3.10: Revelstoke Dam/Spillway Characteristics 55 Table 3.11: Revelstoke Dam/Historical Spills 56 Table 3.12: Revelstoke Dam/Scour Hole Development 57 Table 4.1: Comparison of Spillway Characteristics 101 Table 5.1: List of Empirical Formulas for Ultimate Scour Depth Prediction . 116 Table 5.2: Scour Data Sets 118 Table 5.3: Statistical Analysis of Scour Depth Empirical Formulas Performance 119 Table 6.1: Peace Canyon/Erodibility Index Characterization 151 Table 6.2: Seven Mile/Erodibility Index Characterization 152 Table 6.3: Portage Mountain/Erodibility Index Characterization 153 Table 6.4: Revelstoke/Erodibility Index Characterization 1 . . . . 154 Table 6.5: Jet Hydraulics - Computational Steps 155 Table 6.6: Jet Hydraulics - Basic Parameters 157 Table 6.7: Sensitivity Analysis of Erodibility Index 158 List of Figures Figure 1.1: Free-Trajectory Jet Issued from a Typical Ski-Jump Spillway at Seven Mile Dam (Trail, B.C.) 5 Figure 2.1: Main Parameters and Physical Processes Involved in Plunge Pool Scour 19 Figure 2.2: Jet Behaviour in the Atmosphere 20 Figure 2.3: Jet Behaviour in the Plunge Pool 2 0 Figure 2.4: Erodibility Threshold for Rock and Other Complex Earth Materials . 21 Figure 3.1: Geographic Location of Dam Sites 58 Figure 3.2: Peace Canyon Dam/General Arrangement 5 9 Figure 3.3: Peace Canyon Spillway 59 Figure 3.4: Peace Canyon Dam/Spillway Foundation Geology 60 Figure 3.5: Peace Canyon Dam/Spillway Outflow Hydrograph - 1979 to 2001. . . . 61 Figure 3.6: Peace Canyon Dam/Plunge Pool Topography Before Spillway Operation. 62 Figure 3.7: Peace Canyon Dam/Plunge Pool Topography As Surveyed on 15 April 1980 63 Figure 3.8: Peace Canyon Dam/Plunge Pool Topography As Surveyed on 5 September 1981 64 Figure 3.9: Peace Canyon Dam/Plunge Pool Topography As Surveyed on 4 August 1996 65 Figure 3.10: Seven Mile Dam/General Arrangement 66 Figure 3.11: Seven Mile Spillway 66 Figure 3.12: Seven Mile Dam/Plunge Pool Bedrock - Geological Mapping 67 Figure 3.13: Seven Mile Dam/Photographs of Plunge Pool Bedrock 68 Figure 3.14: Seven Mile Dam / Spillway Outflow Hydrograph - 1979 to 2001. . . . 69 Figure 3.15: Seven Mile Dam/Plunge Pool Topography Before Spillway Operation. . 70 Figure 3.16: Seven Mile Dam/Plunge Pool Topography As Surveyed on 14 December 1979 71 Figure 3.17: Seven Mile Dam/Plunge Pool Topography As Surveyed on 11-12 August 1982 72 Figure 3.18: Seven Mile Dam/Plunge Pool Topography As Surveyed on 20 September 1984 73 Figure 3.19: Seven Mile Dam/Plunge Pool Topography As Surveyed on 21-26 October 1997 74 Figure 3.20: Portage Mountain Project/General Arrangement 75 Figure 3.21: Portage Mountain Spillway 75 Figure 3.22: Portage Mountain Project/Right Cliff Rock Strata 76 Figure 3.23: Portage Mountain Project/Plunge Pool Geology 77 Figure 3.24: Portage Mountain Project/Spillway Outflow Hydrograph -1972 to 2001 78 Figure 3.25: Portage Mountain Project/Plunge Pool Topography Before Spillway Operation 79 Figure 3.26: Portage Mountain Project/Plunge Pool Topography As Surveyed on 15-16 May 1973 80 Figure 3.27: Portage Mountain Project/Plunge Pool Topography As Surveyed on 4 August 1996 81 Figure 3.28: Revelstoke Dam/General Arrangement 82 Figure 3.29: Revelstoke Spillway 82 Figure 3.30: Revelstoke Dam/Plunge Pool Geological Information 83 Figure 3.31: Revelstoke Dam/Spillway Outflow Hydrograph - 1983 to 2001 84 Figure 3.32: Revelstoke Dam/Plunge Pool Topography Before Spillway Operation. . 85 Figure 3.33: Revelstoke Dam/Plunge Pool Topography As Surveyed on 15 May 1984 . 86 Figure 3.34: Revelstoke Dam/Plunge Pool Topography As Surveyed Following Spillway Tests on August 1986 87 Figure 4.1: The Latest Plunge Pool Scour Configuration at Peace Canyon Dam and Seven Mile Dam 102 Figure 4.2: The Latest Plunge Pool Scour Configuration at Portage Mountain Project and Revelstoke Dam 103 Figure 4.3: Spillway Flow Duration Curve of Each Site of Study 104 Figure 5.1: Peace Canyon Dam/Comparison of Prototype Scour and Model Scour . 120 Figure 5.2: Seven Mile Dam/Model Scour Patterns in Non-Cohesive Bed 121 Figure 5.3: Seven Mile Dam/Model Scour Patterns in Cohesive Bed 122 Figure 5.4: Portage Mountain Project/Plunge Pool Scour From Model Tests. . . 123 Figure 5.5: Revelstoke Dam/Scour Configuration From Model Studies 124 Figure 5.6: The Veronese Equation/Predicted Plunge Pool Floor Elevation. . . 125 Figure 5.7: The Jaeger Equation/Predicted Plunge Pool Floor Elevation. . . . 126 Figure 5.8: The Damle Equation/Predicted Plunge Pool Floor Elevation . . . . 127 Figure 5.9: The Chian Min Wu Equation/Predicted Plunge Pool Floor Elevation. 128 Figure 5.10: The Martins Equation/Predicted Plunge Pool Floor Elevation . . . 129 Figure 5.11: The Taraimovich Equation/Predicted Plunge Pool Floor Elevation ! 130 Figure 5.12: The Mason A Equation/Predicted Plunge Pool Floor Elevation . . . 131 Figure 5.13: The Mason B Equation/Predicted Plunge Pool Floor Elevation . . . 132 Figure 5.14: The Wang Shixia Equation/Predicted Plunge Pool Floor Elevation . 133 Figure 5.15: The Yildiz & Uzucek Equation/Predicted Plunge Pool Floor Elevation 134 Figure 5.16: Comparison Between Predicted Versus Observed Plunge Pool Scour Depth 13 5 Figure 6.1: Conceptual Approach of the Erodibility Index Method in the Assessment of Plunge Pool Scour 159 Figure 6.2: The Erodibility Index Method Applied to the Peace Canyon Plunge Pool 160 Figure 6.3: The Erodibility Index Method Applied to the Seven Mile Plunge Pool 161 Figure 6.4: The Erodibility Index Method Applied to the Portage Mountain Plunge Pool 162 Figure 6.5: The Erodibility Index Method Applied to the Revelstoke Plunge Pool 163 Figure 6.6: The Erodibility Index Method/Predicted Plunge Pool Floor Elevation 164 Figure 6.7: Peace Canyon/Sensitivity Analysis of Jet Velocity Profile. . . . 165 Figure 6.8: Seven Mile/Sensitivity Analysis of Jet Velocity Profile 166 Figure 6.9: Portage Mountain/Sensitivity Analysis of Jet Velocity Profile. . 167 Figure 6.10: Revelstoke/Sensitivity Analysis of Jet Velocity Profile 168 Figure 6.11: Peace Canyon/Sensitivity Analysis of Jet Impact Velocity . . . . 169 Figure 6.12: Seven Mile/Sensitivity Analysis of Jet Impact Velocity 170 Figure 6.13: Portage Mountain/Sensitivity Analysis of Jet Impact Velocity . . 171 Figure 6.14: Revelstoke/Sensitivity Analysis of Jet Impact Velocity 172 Figure 6.15: Peace Canyon/Sensitivity Analysis of Jet Air Concentration . . . 173 Figure 6.16: Seven Mile/Sensitivity Analysis of Jet Air Concentration . . . . 174 Figure 6.17: Portage Mountain/Sensitivity Analysis of Jet Air Concentration . 175 Figure 6.18: Revelstoke/Sensitivity Analysis of Jet Air Concentration . . . . 176 ACKNOWLEDGEMENTS I would like to acknowledge BC Hydro for providing access to all information needed to perform this study and for their permission to use the gathered sets of data. The financial contribution of BC Hydro through the Professional Partnership Program is also gratefully acknowledged. CHAPTER I INTRODUCTION Free-trajectory jets are commonly used in large dam operation as a means of releasing excess water from dam reservoirs. Included are classical overfalls of overtopping dams, pressure outflows through the dam itself, and deflected discharges at the toe of overflow structures. The ski-jump or flip bucket device is an economical way of throwing the jet away from the structure, thus enhancing the energy dissipation before it reaches the downstream riverbed. Such spillway arrangement is shown in Figure 1.1 at Seven Mile Dam in Trail, British Columbia. The remaining energy is dissipated through the excavation of a scour hole, or plunge pool. Plunge pools are effective if the tailwater level is uncertain, the resultant scour hole does not endanger surrounding structures, and space limitations preclude use of a conventional stilling basin (Hager, 1998). Breusers and Raudkivi (1991), in their book on scouring, note the following adverse effects of scour due to plunging jets: 1) The endangering of the stability of the structure itself by structural failure or increased seepage; 2) The endangering of the stability of the downstream riverbed and side slopes; and 3) The formation of a mound of eroded material, which can raise the tailwater level at the dam. Experience has shown that free-trajectory jets can create unforeseen and dangerous erosion of scour holes, leading to costly remedial measures. Documented cases of such situations include the Grand Rapids Generating Station (Canada) (Manitoba Hydro, 1970), Tarbela Dam (Pakistan) (Lowe III et al., 1979), Kariba Dam (Zimbabwe) (Whittaker and Schleiss, 1984), Cahora-Bassa Dam (Mozambique) (Whittaker and Schleiss, 1984; Quintela et al., 1979), and Picote Dam (Portugal) (Da Cunha and Lencastre, 1966). The potential for excessive downstream scour should be addressed early in the feasibility studies in order to establish the viability of the layout, and even the project itself. The engineer must be able to assess the most probable scenario of plunge pool development resulting from the expected use of the spillway. Throughout the life of a project, the plunge pool response to changes, in spillway operation must also be foreseen. Over the past decade, the reevaluation of the Probable Maximum Flood (PMF) at some sites has led to the recognition of insufficient spillway capacity. The problem of scour in dam foundations is considered so critical in the US that the Federal Energy Regulatory Commission (FERC) requires owners of hydro-electric plants to demonstrate that scour resulting from overtopping under PMF conditions will not endanger their dams (Annandale, 1994). The difficulty in estimating plunge pool scour remains in the numerous interrelated physical processes involved. In the analysis, one must consider the plunging jet diffusion and inner core decay coupled with air entrainment in both the atmosphere and plunge pool, the mean and fluctuating pressures distribution on the plunge pool floor, the fracturing and progressive dislodgment of the plunge pool material, the abrasion process, the downstream sediment transport capacity, and the induced shear on the scour hole boundaries caused by secondary currents. This is further complicated by the changing scour hole geometry and the accumulation of material forming a mound at the downstream margin of the hole. Although the physical processes are generally recognised and addressed by many authors, most research to date has focused on developing empirical formulas for ultimate scour depth, based on experiment and prototype observations. Physical models that respect the Froude law scaling are used in laboratory to simulate scour downstream of overflow structures. Both the theoretical approach and the laboratory test approach to scour depth forecasting have met with considerable difficulties (Yildiz and Uziicek, 1994). As an attempt to improve technology for the prediction of scour caused by overtopping flows, the Dam Foundation Erosion Study Team was formed in 1993 under the management of the U.S. Bureau of Reclamation. The new technology is based on a threshold relationship between the rate of energy dissipation of flowing water and a geomechanical classification of earthen materials (Erodibility Index) using 150 field observations (Annandale, 1995). The main objective of this study is to evaluate new technology for estimating plunge pool scour and determine whether there is an improvement in accuracy when compared to conventional methods such as empirical formulas and small-scale model studies. The second objective is to establish the influence of factors such as spillway design, geological conditions, and spillway discharges upon plunge pool development. Four dam sites were selected which are all owned and operated by B.C. Hydro, public corporation of British Columbia: Peace Canyon Dam, Seven Mile Dam, Portage Mountain . Project, and Revelstoke Dam. Plunge pool performance data were gathered for each site along with spillway characteristics, plunge pool geology, and spilling history. The sites chosen feature both gated overflow and long chute spillways. Plunge pool geology varies from nearly horizontal sedimentary rock strata to metamorphic argillite, gneiss, and quartzite. The history of spillway discharges is unique to each site. Chapter II includes a review of the physical processes involved in scour formation as reported in the literature and documents the main areas of research in the evaluation of scour extent. Chapter III is a descriptive summary of each of the four selected dam sites (Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam) given the context of plunge pool scour investigation. A discussion of the main factors affecting plunge pool performance at each site is presented in Chapter IV. The conventional methods of plunge pool scour assessment are evaluated in Chapter V through a comparison between scour configurations from hydraulic model studies, computed scour depths using empirical formulas, and actual scour development on sites. The Erodibility Index Method is presented in Chapter VI and the new approach to the assessment of plunge pool scour is applied to the four sites of study. Chapter VII summarizes the limitations and conclusions of the study with qualitative observations on the most probable scour progression to expect at each site in the event of the Inflow Design Flood. To facilitate describing the sites and the plunge pool configuration, a standard convention of reference is used: the right and left sides of the river channel are defined from the perspective of someone looking downstream of the dam. The use of imperial units throughout this study was motivated by the available data and the existing dam reference grid (in the Imperial system of measurement) which was the basis for the plunge pool scour configuration. Figure 1.1. Free-trajectory jet issued from a typical ski-jump spillway at Seven Mile Dam (Trail, B.C.) CHAPTER II LITERATURE REVIEW The study of plunge pool scour covers several disciplines. Erosion phenomena downstream of spillways involve complex interactions between mechanical and hydraulic processes (Simoes and Vargas JR, 2001) . Although the physical processes are generally recognised and addressed by many authors, these processes have usually been isolated and investigated separately. All these factors intervene, in a more or less marked manner, in the erosive process, but our knowledge of the phenomena that occur in the erosion zone is insufficient, so that there are no formulations that enable theoretical models to be prepared for forecasting the behaviour of failure of the mass under the action of jets from flood spillways (Ramos, 1982). Instead, the main stream of work was centred on developing empirical formulas for ultimate scour depth, based on experiment and prototype observations. In laboratory, physical models that respect the Froude law scaling are used to simulate scour downstream of overflow structures. In the mid 1990s, a new approach to scour was proposed by Annandale (1994, 1995) which formed the working basis of the Dam Foundation Erosion Study started in 1993 under the management of the U.S. Bureau of Reclamation. First, background theory on the physical processes related to scour including jet hydraulics and mechanical interaction with the rock mass is presented, followed by a brief overview at the most documented methods on the assessment of scour downstream of spillways. 2.1 PHYSICAL PROCESSES The power (kW/m2) available to erode material is a function of the jet hydraulics (Annandale, 2002c) . The more energy dissipated along the jet trajectory the less the scouring potential. The energy of the jet is dissipated in three phases: (i) during the aerial phase, when the reach of the jet is decreased by air resistance; (ii) during the submerged phase, where diffusion processes in the water cushion attenuates the erosive power of the jet. This phase is more efficient as the depth of the scour zone increases; (iii) in the rock mass after impact, causing the erosion (Simoes and Vargas JR, 2 001). The main physical processes associated with the energy dissipation of the jet through air, water, and rock mass are summarized in Figure 2.1 and described herein. 2.1.1 Free-Trajectory Jets Water flowing through a ski-jump spillway leaves the terminal structure as a free-trajectory jet and falls into the river channel some distance away from the spillway. The trajectory assumed by the jet depends on the flow energy at the bucket lip and the exit angle in accordance with the Newtonian laws of motion. As the jet travels through the atmosphere and encounters air resistance, its physical characteristics evolve. Important aspects of plunging jet behaviour are: the ability to spread laterally; the ability to become increasingly distorted during the plunge; and the ability to break up if the plunge length is sufficiently long (Ervine and Falvey, 1987). Figure 2.2 illustrates the jet behaviour in the atmosphere. Ervine and Falvey (1987) suggest that the initial turbulence intensity determines the angle of lateral spread of the jet. The turbulence intensity Tu is defined here as the ratio of the root mean square value of the instantaneous axial velocity to the average axial velocity. Turbulence intensity for free overfall jets is less than 3%, for flip bucket jets between 3 and 5% and for orifice jets between 3 and 8% (Bollaert, 2002, as reported by Schleiss, 2002). The turbulence intensity enhances the breakdown of large eddies into smaller ones and the fine-grained turbulent structure is more efficient to entrain air. As air becomes entrained in the jet, conservation of mass requires that the jet expand (Davies and Jackson, 1982). Hence, the jet forms an expanding aerated outer shell with a decaying inner core of water (Figure 2.2) . The angle of jet core decay may be as small as 15-20% of the angle of lateral spread, ignoring core contraction due to gravity (Ervine and Falvey, 1987) . In highly turbulent jets, initial disturbances present on the surface of the jet and associated with fine scale turbulent eddies are propagated as transverse or sinuous wave-like disturbances (Davies and Jackson, n.d.). Surface tension acts to resist the growth of the surface disturbances but the protuberances are amplified by the action of aerodynamic drag forces. As the jet falls, the undulations at the surface of the jet increase in amplitude to reach a point where the turbulent surface fluctuations are large enough to penetrate the core of the jet. Instead of a core of solid water with an aerated outer shell, the jet becomes a conglomeration of individual water drops, each having separate acceleration vectors (Lewis et al. , 1996). The jet break-up length is the free-fall distance before the complete dissipation of the jet inner core. The break-up point marks the limit between two types of jets: undeveloped and developed (Figure 2.2) . Undeveloped jets are characterized by a solid, non-aerated water core as opposed to developed jets which are fully aerated and composed of discrete water segments. As the jet free falls, the internal turbulence intensity, drag due to air resistance, and air entrainment processes cause the jet to transition from an undeveloped to a developed condition (Bohrer et al. , 1998) . The deceleration caused by air drag becomes a significant factor when the jet loses its coherence in free-fall (Lewis et al. , 1996) . Surface tension and turbulence effects determine the distance to the break-up point (Ervine et al. , 1997) . Although a clear understanding of the air entrainment process has not yet been developed, the direction is more towards a recognition of turbulence as the most significant air-entraining mechanism (Ervine and Falvey, 1987). In turbulent jets a large proportion of the entrained air is entrained from the surface undulations on the jet surface, and the remainder from the boundary layer surrounding the jet (Ervine et al. , 1980) . One criterion for the onset of aeration is that the radius of the eddies in the jet have the same order of magnitude as the surface disturbance amplitude (ibid.). The air entrained by the free jet during flight tends to cushion the impact with the tailwater surface. Most authors agree that aerated water will produce a lesser scour than unaerated, "solid" water (Mason and Arumugam, 1985) . The result of air entering the jet is that the effective time average density is considerably reduced and therefore the time average jet dynamic pressure is also reduced in spite of the fact that negligible jet deceleration may have occurred (Davies and Jackson, 1982). Ervine et al. (1997) show that the combined effects of jet spreading and air entrainment decrease the mean and fluctuating dynamic pressures on the pool floor. If the mixing of the air is incomplete and the jet is not fully established [developed] upon impact with the plunge pool, there will be no cushioning effect in the core region (Spurr, 1985) . 2.1.2 Jet Diffusion and Hydraulic Action within the Plunge Pool An approximate description of the hydraulic phenomena involved in the diffusion of a free falling jet in the tailwater pool is possible when the theory of the free jet turbulence is applied (Hartung and Hausler, 1973). Upon impact with the tailwater surface, the trajectory of the free jet follows a straight line sloping at the angle of penetration down to the plunge pool invert. Gravitational effects on the j et would be minimal once the jet entered the tailwater, and the jet would tend to follow a straight line rather than a free-fall trajectory (Johnson, 1974) . As the jet plunges into the pool, it diffuses almost linearly (Whittaker and Schleiss, 1984). Because the turbulent plunging jet comprises an expanding, undulating, and aerated outer zone, the initial boundaries of the jet entering the pool are not clearly defined. On account of the irregularity of the flow in the outer region, plunging turbulent jets will produce surface waves in the plunge pool rather than well-defined penetrating shear layers (Ervine and Falvey, 1987). The jet behaviour in the plunge pool is illustrated in Figure 2.3. The jet flow below the water surface may consist of either one or two regions depending on the coherence of the jet (undeveloped or developed conditions) at the moment of impact with the tailwater (Figure 2.2) . A transition zone, known as the flow establishment region, exists when the central core of the plunging jet is still present upon impingement (Figure 2.3). The region of flow establishment is where the shearing action at the edge of the jet decreases the edge velocity, but does not affect the velocity near the centre of the jet (George, 1980). The inner core takes a wedge-like form as turbulence penetrates inwards towards the centreline of the jet. This wedge of undiminished mean velocity is referred to as the potential core. The length of flow establishment varies from 5 to 10 times the thickness of rectangular jets (George, 1980). Beyond this distance, the velocity remains a maximum along the jet centreline but the entire flow velocity field is reduced by diffusion (Hartung and Hausler, 1973). The zone of established flow (or fully developed flow) begins at the point where the central core of the plunging jet is completely diffused. The velocity profile in the established flow region has a very nearly Gaussian distribution (George, 1980). A developed falling jet will diffuse in the plunge pool and form a fully developed flow region (zone of established flow) upon entrance. The diffusion process continues until all the initial energy of the jet is dissipated, or until the influence of a boundary causes an impinging flow region (George, 1980). Ervine and Falvey (1987) summarize results of studies on the behavior of jet diffusion in plunge pools from free-falling circular jets of variable turbulence intensity and air concentration. In the case of rough turbulent impinging jets with air concentrations of the order of 40% (most relevant to prototype behavior), measurements of the outer angle of spread in the zones of flow establishment and established flow yielded average values of 13-14° and 14-15°, respectively (Figure 2.3). The angle of inner core decay has been estimated to 8° based on an approximate momentum balance between the jet at the point of impact and at the end of the zone of flow establishment (Ervine and Falvey, 1987) . Two-phase flow develops when a plunging jet dives through the surface of a plunge pool or tailwater (Wittier et al., Dam Foundation  Erosion, 1995) . As the jet plunges into the pool, a considerable amount of air is entrained, corresponding to an air concentration of 40 to 60% at typical jet velocities of 30 m/s [100 ft/s] at impingement (Bollaert, 2002, as reported by Schleiss, 2002). Bin (1984) and Ervine et al. (1980) describe the mechanisms of air entrainment by a plunging jet at the surface of a plunge pool. The captured air is dispersed into bubbles by shear forces and turbulence and forms an approximately conical biphasic region (Bin, 1984) (Figure 2.3). Air concentration decreases as the jet travels into the pool, thereby increasing jet density (Annandale et al., 1997). Estimates of the mean air concentration with depth can be made by assuming that the mean air flow rate decays approximately linearly from a maximum value at the jet plunge point to zero at the maximum air bubble penetration depth (Ervine and Falvey, 1987). At the maximum air bubble penetration depth, the mean water velocity approximates the bubble rise velocity ending the jet's capacity to retain entrained air. It is reasonable to speculate that the presence of air bubbles in the diffusing plunge pool shear layers will result in a reduction in mean dynamic pressures (Ervine and Falvey, 1987) . However, the presence of free air bubbles was recently shown to be important in the propagation of fully transient water pressures in rock joints (Bollaert, 2002) . The impinging jet induces circulation in the plunge pool (Figure 2.1). Both horizontal and vertical velocities cause rolling actions in the plunge pool, the mechanics of which do not appear to have been fully investigated to date (Sutcliffe, 1985). As discussed by Spurr (1985), the hydraulic action within the plunge pool is largely influenced by the pool boundaries. The formation of a "backroller" (clockwise vortex) between the jet and the upstream boundary is common knowledge (Henderson, 1966). As the submerged jet strikes the bed, a small stream moves upstream in the roller region, while the main stream is deflected upwards to form a downstream surface boil. The difference in upstream/downstream water levels induces the formation of secondary currents at the sides of the pool. A recent study conducted at Colorado State University (Fort Collins, CO) on plunge pool circulation and velocity prediction in a plunge pool basin suggests the formation of a counter-clockwise vortex in the roller region (Hamilton et al. , 1997) . The most significant factor leading to the counter-clockwise rotating vortex is the buoyancy force resulting from the high degree of air entrainment in the plunging jet (ibid.). 2.1.3 Scour Mechanisms Scouring of the rock bed downstream of a spillway is caused by the water pressure or total water load at a particular point (Galvagno, 1998). The three components which make up the total water load are the mean dynamic pressures caused by the impact of the jet on the rock surface; the truly mean dynamic time dependent water pressures (pressures fluctuations); and the Reynolds shear stresses (Otto, 1986, as reported by Galvagno, 1998) (Figure 2.1). The most severe hydrodynamic action on the plunge pool bottom occurs in the impingement region, where the hydrostatic pressure of the free jet region is progressively transformed into fluctuating high stagnation pressures and into an important bottom shear stress (Schleiss, 2002). The combination of forces exerted by the submerged jet varies as the scour hole develops. Initially, the bedrock is subjected to a large pressure gradient in the zone of jet impact. The high dynamic pressures build-up in the bedrock discontinuities and causes the bedrock to fracture. Hydrofracture is the mechanical destruction of the bedrock into blocks by the dynamic action of the jet (Spurr, 1985). As the plunge pool deepens, the mean dynamic pressure of the jet is reduced by the increased travel distance before reaching the bed. However, the fluctuating pressures associated with the turbulent shear flow or developed jet impact ensure the progression of scour. The developed jet impact and the direct jet core impact in the case of small water cushion generate completely different pressure patterns (Schleiss, 2002) . Recently, the propagation of pressure fluctuations through the discontinuities in the rock mass has been considered to be the predominant mechanism in the erosion processes (Bollaert, 2002; Simoes and Vargas JR, 2001; Liu et al. , 1998) . It has been noted that scour can be limited not only by the ability of the jet to dislodge material from the bed, but also by the ability or otherwise of secondary return currents to remove that material to beyond the limits of the scour hole (Hartung, 1959, as reported by Mason, 1989) . Akhmedov (1988) identifies three stages to the development of a scour hole: 1) removal of rock fragments by the hydrodynamic force of the flow, 2) destabilization of rock mass by vibration-induced hydrodynamic pressure fluctuations, and 3) abrasion of the scour hole walls by the recirculating trapped material: Initially, the flow volume will break off fragments of rock. The size of the fragments will be determined by the rock jointing and cleavage. As the scour pit depth increases, the flow velocity near its bed will decrease, and there comes a point where this form of erosion ceases. Subsequent scouring will occur as a result of intensive abrasion. (Akhmedov, 1988) Annandale (1994, 1995) describes the removal of fragments from the rock mass through the process of progressive dislodgment that involves three components: jacking, dislodgement, and displacement. The jacking effect is caused by the instantaneous differential pressures on the upper and lower surfaces of rock blocks. Once the material is destabilized, it is dislodged by the flowing water and displaced in the downstream direction. When the combined effects of the plucking forces normal to the loosened block and the tangential form drag forces induced by the local boundary flows become sufficiently large, the block becomes dislodged and is swept away (Spurr, 1985) . The displaced material is deposited downstream of the scour hole as the flow transport capacity diminishes and a tailrace bar deposit (or mounding) can form (Figure 2.1) . As the scour hole deepens, more energy is required to remove the rock fragments and larger blocks remain trapped inside the hole. The dislodged blocks, recirculating within the live scour-hole, abrade through contact with the other rocks until they are small enough to be ejected as part of the downstream wash load (Spurr, 1985) . The geological conditions of the plunge pool bedrock affect the scour processes. The rate of penetration and build-up of pressure by the jet within the bedrock is influenced by the condition and orientation of its discontinuities (Spurr, 1985). The hydraulic breakage of the rock mass is more efficient in rock with open joints than in sound rock with tight and discontinuous joints. Likewise, the more closely aligned the major bedding planes are to the angle of incidence of the plunging jet, the easier the penetration and the more rapid will be the hydrofracture processes (Spurr, 1985) . The process of dislodgement is affected by the geometry of rock blocks, which is determined by the discontinuity pattern of the rock formation. Smaller size and equi-sided shape provide less resistance to erosion. The orientation of the material relative to the direction of flow also has an impact on its capacity to resist erosion (Annandale, 1994). Rock dipped in the direction of flow is more easily dislodged than rock dipped against it (Annandale et al. , 1996). The strength of rock is determinant in the process of abrasion. The rock susceptibility to abrasion is inversely proportional to its strength (Prochukhan et al., 1971, as reported by Akhmedov, 1988) . The cohesive strength of the rock formation is particularly important at the periphery of the scour hole to resist the erosive shear forces of secondary currents. Ideally, the scour hole is confined laterally to contain the jet and act as a dissipating pool. The non-homogeneity of the rock in a plunge pool may significantly affect its shape (Spurr, 1985). Asymmetrical plunge pool development results from the bedrock heterogeneity. 2.2 PLUNGE POOL SCOUR ASSESSMENT The Dam Foundation Erosion Study Team (DFEST) was formed in 1993 under the management of the U.S. Bureau of Reclamation in an effort to improve the current technology in the assessment of scour downstream of overtopping dams. The need for better analytical tools for analyzing erosion in the foundation and abutment areas of dams is increasing as costly alternatives to overtopping come under consideration (Wittier at al., Spillway  and Foundation  Erosion, 1995). Wittier et al. (1995) explain the basis for the development of new technology: Current methods of predicting and evaluating erosion extents have limited applicability. Existing erosion prediction formulas do not track erosion as a function of time, and have limited application in hard-rock or cohesive foundation materials. (Wittier at al. , Spillway and Foundation  Erosion, 1995) Numerous physical model studies described in the literature have limited application due to the uncertainty of scale effects associated with jet turbulence, jet coherence, jet air entrainment, and foundation material properties. (Wittier et al., Dam Foundation  Erosion, 1995) The main limitations of hydraulic model studies are discussed below, a brief account of existing empirical formulas is made, and the new technology proposed by the DFEST for the prediction of plunge pool scour is presented. 2.2.1 Hydraulic Model Studies Most high-head hydroelectric projects require hydraulic model studies before construction for design optimization. Generally included are downstream scour tests. In practically all energy dissipators, gravity forces predominate, and dynamic similarity is attained by designing the model in accordance with the Froudian relationship (Elevatorski, 1959). The model response to scaled spillway discharges is used to identify possible areas of concern downstream of the hydraulic structure and assess expected plunge pool development. Hydraulic model studies are a common way of predicting plunge pool performance but such predictions can be very incorrect (Whittaker and Schleiss, 1984) . The main limitations of the assessment of plunge pool scour in laboratory are discussed. One of the largest difficulties in modelling plunge pool scour is representing the riverbed geology at a model scale. In order to use model data for predicting scour depths associated with a stilling basin or plunge pool, the model bed material type and size must be chosen carefully to allow scaling (Whittaker and Schleiss, 1984). A typical approach is to examine the rock on site and to estimate, from joint and fissure patterns and from the strength of the rock, the individual block size to which the rock will break down (Mason, 1984) . When a cohesive bed is preferred, the difficulty of choosing the right particle size is further complicated by the choice of the cohesive agent and the proportions of the resulting mixture. In non-cohesive models, the fracturing process of the bedrock is assumed to have already taken place. It follows that the modelled scour hole is more a function of the jet momentum impact and sediment transport capacity (Reynold's stresses), the latter being distorded at a Froude scale. Nevertheless, "if the bed material is chosen carefully, . good predictive results for scour depth can be obtained by using non-cohesive material" (Whittaker and Schleiss, 1984) . The loose structure of the material tends however to exaggerate the areal extent of erosion. Steep slopes similar to those found in rock and a more representative shape of the scour hole are obtained in tests with slightly cohesive material (Whittaker and Schleiss, 1984) . Another limitation of model scour is the failure to reproduce the abrasion process. On prototypes, the material trapped within the scour hole hydraulic action is being tumbled around and disintegrates with impact until it is small enough to be displaced out of the hole. Abrasive erosion which results from loose material being washed against the basin boundaries in a "ball-mill"-type action can be expected to occur to some extent at all sites (Johnson, 1974) . In any case, model tests can not simulate the break-up of the rock blocs by the ball milling effect of the turbulent flow in the plunge pool (Schleiss, 2002). Therefore in the model a mound is formed which is higher and more stable than in the prototype (ibid.). With that respect, model scour underestimates the actual scour development on prototypes. Other limitations include scale effects associated with jet turbulence, jet coherence, and air entrainment. Such phenomena are largely influenced by surface tension and viscous forces which are distorted at the Froude scale. 2.2.2 Empirical Formulas Whereas scour studies began with the first hydraulic structures, plunge pool scour research is a fairly recent development (Hager, 1998) . The depth of rock scour under free falling jets has been the subject of continuing research, starting with the work of Schoklitsch in 1932 (Mason, 1993). Followed experiments by Veronese who in 1937 developed an empirical formula which is recommended by the U.S. Bureau of Reclamation (USBR) as a limiting scour depth. The early investigations were limited to small-scale tests on non-cohesive beds using vertically falling jets. Subsequently, prototype scour data sets were gathered and analyzed and experiments were made using various outlet arrangements and simulating structural discontinuities in bedrock. A large number of empirical equations have been developed from physical model studies in laboratory and from prototype observations. Few investigators have developed expressions for the three-dimensional aspect of scour and instead efforts were directed towards the assessment of scour depth. As well, the time progression of scour was generally omitted and most empirical formulas target the ultimate, or equilibrium depth of scour. The equilibrium scour depth is the depth scoured by a given magnitude of spillway discharge beyond which no significant increase in depth is experienced by more spills of the same magnitude and of significant duration (Spurr, 1985) . Mason and Arumugam (1985) made a thorough review of existing scour depth prediction formulas and tested their performance on comprehensive data sets from 26 prototypes and 4 7 models. The cases included free overfalls, low level outlets, spillway chute flip buckets and tunnel outlets, while the model cases featured both cohesive and non-cohesive, granular bed materials (Mason and Arumugam, 1985). In total, 31 formulas dating from 1932 to 1981 were identified and sorted in five groups. Group I (17 equations) includes a simple form of equations in which the depth of scour below tailwater level, D , is a function of the unit discharge q, the head drop between reservoir and tailwater levels H  , and in some cases the characteristic size of bed material d (Figure 2.1). The equations of Group II (2 equations) add the consideration of tailwater depth h  . Group III (3 equations) consists of highly simplified equations or rules-of-thumb. The equations of Group IV (8 equations) are more complex formulations proposed by Russian authors in the 1960s that include parameters such as energy losses on spillway face, jet entrance velocity and thickness, degree of aeration and disintegration of the jet, angle of impingement, rock strength, and rock discontinuities characteristics. Group V contains only one equation that relates to the time progression of scour. It was found that the most accurate form of expression was the relatively simple formula: d = KOZL dz in which K,  X, y , and Z are constants derived from each author. Mason (1984) contributed to a major improvement in the assessment of ultimate scour depths under free jets by developing two expressions (one for model data and one for both models and prototypes) which are dimensionally balanced and respect the Froude scaling law. The equation for model scour is now considered as an upper bound for prototype scour and recognized as the most practical and state-of-the-art equation for dam design. Even so, Mason (1984) acknowledges that accuracy is limited to 70%. Some theoretical approaches have been proposed for the assessment of scour depth based on the nature of jet diffusion within the plunge pool and the distribution and fluctuation of pressure over a rocky bed. Mason (1989) affirms that "such aspects have been tested and analyzed in the past without, it would appear, producing scour depth formulas any more accurate than the ones already reviewed" . One of the main criticisms addressed towards the mass of empirical formulas is the poor representation of bedrock conditions. The performance of selected empirical formula is evaluated in Chapter V. 2.2.3 The Dam Foundation Erosion Study The Dam Foundation Erosion Study Team (DFEST) came together in 1993 to improve technology for estimating the progressive extent of erosion caused by overtopping flows. The main partners to the study are the U.S. Bureau of Reclamation (USBR), Golder Associates Inc., and Colorado State University (Fort Collins, CO) . A new approach was proposed for quantification of the extent of scour caused by a plunging jet, which applies to the complete range of earth materials. The investigation involved extensive research on a 1:3 scale hydraulic model and near-prototype facility at Colorado State University. The new technology was validated through experiments using simulated fractured rock and granular media (Annandale et al. , 1998; Wittier et al. , Prototype  Validation, 1998). The new technology is based on Annandale's hydraulic erodibility concept (1994, 1995) which gives equal weighting to both hydraulics and engineering geology. Hydraulic erodibility is a threshold condition that defines the conditions when the ability of earth material to resist erosion is exceeded by the erosive power of the water discharging over it or incident to it (Annandale, 1994). The erodibility threshold is a graphical relationship between the rate of energy dissipation of flowing water and a geomechanical classification of earthen materials called the Erodibility Index, based on the analysis of 150 field observations and published data on initiation of sediment motion. Figure 2.4 presents the erodibility threshold for rock and other complex earth materials defined as the boundary between erosion and non-erosion events. Annandale (1995) acknowledges the fact that the fluctuating pressures associated with turbulent flows are mainly responsible for the progressive dislodgment of material in the plunge pool. By definition, an increase in turbulence intensity is associated with a corresponding increase in differential instantaneous pressures and energy loss. Therefore, the rate of energy dissipation (or stream power) was chosen to represent the relative magnitude of the fluctuating disturbance. Theoretical justification of the relationship between rate of energy dissipation and turbulence energy production rate was derived by Yang and Molinas (1982), whereas the relationship between rate of energy dissipation and pressure fluctuation was validated by further analysis of observations by Fiorotto and Rinaldo (1992) (reported by Annandale, 1995). The rate of energy dissipation per unit area or stream power P  can be expressed in terms of velocity Vand energy loss AE such as: P  = yQ&E/A  = yvAE  [KW/m 2 ] where y is the unit weight of water, Q is the total discharge, and A is the area perpendicular to the flow. The Erodibility Index is a replica of Kirsten's (1982) Ripability Index for excavation, which in turn is a modification of the well-known Norwegian Rock Tunnelling Quality Index, Q. The Erodibility Index characterization is essentially a quality assessment of earth material through numerical rating of key parameters. The primary geological parameters that are used to calculate the Erodibility Index are earth mass strength [Ms],  block or particle size [Kb],  discontinuity/inter-particle bond shear strength [ Kd  ] , and shape of material units and their orientation relative to the flow [Js] (Annandale, 1995). The Erodibility Index K  is the product of factors assigned to each constituent parameter: K  = Ms  • Kb  • Kd  • Js The mass strength Ms  is the dominant factor in the resistance to erosion. The other parameters represent, respectively, reducing effects of size, joints conditions, and shape/orientation relative to the flow on eroding the intact rock mass. The paper by Annandale (1995) provides standard tables quantifying these geological parameters and the principal tables are repeated in Appendix I. The Erodibility Index Method (Annandale, 1994, 1995) is evaluated using data collected from four BC Hydro dam sites in Chapter VI. PLUNGE POOL SCOUR LEGEND Vo: initial jet velocity Vi: jet velocity at impact with the plunge pool Vm: maximum jet velocity in the plunge pool H: head difference  between reservoir  level and tailwater surface h: tailwater depth Ds: maximum depth of scour below original bed level 111 free-trajectory  jet | U zone of flow establishment H zone of established flow E l impingement region mean dynamic pressure bottom pressure fluctuations shear stresses hydrodynamic fracturing hydrodynamic uplift surface  boil-up sediment transport Ds Figure 2.1. Main Parameters and Physical Processes Involved in Plunge Pool Scour Figure 2.2. Jet Behaviour in the Atmosphere 10,000 1,000 fM E <u o CL (D QJ I/) C 3 100 10 0.1 ERODIBILITY Rock and Complex Earth Materials • Erosion O No Erosion • • • • • • • • • • $ • * / c o c • • • +1 }.* / o ^ O CD n o o yo o o ° o ° o • • 0,-0 • • - o o O o o o o 0.01 0.1 10 100 Erodibility Index 1,000 10,000 Source.- G.W. Annandale, "Erodibility", Journal  of Hydraulic  Research 33 (April 1995): 471-493, Figure 9. Figure 2.4. Erodibility Threshold for Rock and Other Complex Earth Materials CHAPTER III DAM SITE DESCRIPTIONS The four dam sites selected for this study are Peace Canyon, Seven Mile, Portage Mountain, and Revelstoke, which are owned and operated by B.C. Hydro (BCH), public corporation of British Columbia, Canada. The Seven Mile & Revelstoke projects are part of the Columbia River Basin (southeastern B.C.), whereas the Peace Canyon and Portage Mountain projects are located in the Peace River Basin (central-eastern B.C.). Refer to Figure 3.1 for geographic location. The plunge pool scour investigation requires the knowledge of key factors including spillway design, plunge pool geology, hydraulic conditions, and historical spillway outflows. The sites chosen feature both gated overflow and long chute spillways. Plunge pool geology varies from nearly horizontal sedimentary rock strata to metamorphic argillite, gneiss, and quartzite. The history of spillway discharges is unique to each site. This chapter is a descriptive summary of each of the four dam sites in question given the context of plunge pool performance investigation. For each site, the spillway main design characteristics and the plunge pool geological conditions are defined. The history of spills from initial spillway operation to the end of year 2001 is documented. The chronological plunge pool development is described by means of past sounding surveys performed at each site. 3.1. PEACE CANYON DAM 3.1.1 Project General Arrangement The Peace Canyon Project (PCN) is located at the outlet of the Peace River Canyon in northeastern British Columbia (Figure 3.1), 4 miles southwest of Hudson's Hope, and 80 river-miles upstream from the British Columbia/Alberta border. As the second hydroelectric development of the Peace River, the Peace Canyon generating station reuses the water utilized for power generation at the upstream Portage Mountain Project. The Peace Canyon Project consists of a 1130 ft long by 200 ft high concrete gravity structure built across steep canyon walls, and connected to an embankment saddle dam on the right abutment (Figure 3.2). The saddle dam is approximately 670 ft long with a maximum height of 75 ft. The six-bay spillway arrangement is part of the right half of the dam and the power intake section occupies the left side of the river width. The Dinosaur Reservoir has a surface area of 2200 acres and a total storage volume of 175,000 acre-ft at Maximum Normal Reservoir Level (El. 1650 ft). The powerplant has four units with a maximum sustainable generating capacity of 700 MW. The first two units were commissioned in April 1980 and all four units were in service by November 1980. 3.1.2 Spillway Characteristics The Peace Canyon spillway is a gated six-bay overflow spillway with flip buckets at the toe (Figure 3.3). The main design characteristics are listed in Table 3.1. The spillway discharge is controlled by six 50 ft wide and 41.5 ft high radial gates over an ogee-shaped crest at El. 1612 ft. Spillway bays and gates are assigned numbers 1 to 6 from left to right, with Bay 1 closer to the power intakes. A training wall between Bays 2 and 3 provides lateral flow control. The profile of the two separate chutes follows the downstream face of the dam at a slope of IV: 0.764 7H from the shaped crest to the 55 ft flip bucket radius. Flip buckets for the left chute (Bays 1-2) end with a 30° flip angle and lip at El. 1512.37 ft. The right chute buckets (Bays 3 to 6) have a 20° lip angle and lip at El. 1522.77 ft. To avoid possible damage to the powerhouse and right bank, the six spillway gates should be opened in order 3, 4, 6, 5, 1, 2 and closed in reverse order (BCH Report No. OMSPCN/03, 2001). Spillway Bays 3 and 4 were selected for principal use because they are centered on the plunge pool (BCH Report No. H1742, 1987). The Inflow Design Flood (IDF) at Peace Canyon Dam is the Probable Maximum Flood (PMF) based on the Williston Lake/Probable Maximum Flood 1988 study by B.C. Hydro (BCH Report No. H2003, 1988) . Because of the limited storage in Dinosaur lake, the Williston lake PMF outflow is the Peace Canyon PMF inflow (BCH Report No. H2767, 1994). The Peace Canyon spillway can pass the IDF of 330,200 cfs at reservoir El. 1652.4 ft with all gates fully open. 3.1.3 Plunge Pool Geology Bedrock in the riverbed downstream of the spillway consists of massive silty shale to shaly siltstone that exhibits little or no lamination (BCH Report No. GEO 9/85, 1985). Contacts between interbedded rock types within the foundation shale member are gradational. The lower 3-4 ft of the shale unit is a transition rock to the underlying massive sandstone unit at approximate El. 1430 ft. The sandstone unit is about 30 ft thick and below is another massive shale sequence of unknown thickness. Bedding strikes N60° and dips about 2° implying small components of dip upstream and towards the right abutment (approximately 1.4° or 2.5V:100H). Unconfined compressive strengths of the actual concrete dam foundation range from about 5000 to 13,000 psi in competent shales and about 11,000 to 20,000 psi in the basal sandstone (BCH Report No. H1742, 1987). The most significant engineering feature in the foundation of the project is the presence of a number of bedding planes which vary from intact to separated, with up to an inch or more of infilling which comprises variously decomposed shale, river silt or blast-hole cuttings (BCH Report No. 966, 1978). From the riverbed down to a depth of about 40 feet up to five bedding planes in the shale are visually prominent in that they separate shale from thin beds of sandstone and/or have been opened due to upward relaxation of the beds after removal of the rock load with the erosion of the canyon (BCH Report No. 822 - Appendix A, 1976) . The bedding planes, which are numbered from top down at the site, get tighter towards each abutment. The major discontinuity planes below the spillway foundation are located in Figure 3.4 and defined below: • B.P.2 is the most persistent bedding plane at the site with up to 2 inches of infilling. • B.P.3 is a persistent bedding plane that overlies a thin but very distinct ripple-marked sandstone bed. B.P.3 is generally a frozen contact, but locally open with up to M inch of infilling. • B.P.4 is a 3 ft thick zone of black fissile shale within which four discontinuous bedding cracks are identified (B.P.4A/4B/4C and 4D) . The planes are locally parted but free of infilling. • B.P.X is described as a hairline fracture located about 5 ft above the contact with the basal sandstone. To date, the plunge pool is eroded down to B.P.4 downstream of two of the six spillway bays (Bay 3 and 4). The foundation shale member is essentially free from joints except for a 10 0 ft wide zone oriented subparallel to the river canyon through the left side of the spillway foundation. Jointing is generally sparse with joints striking N30° to 40°E (i.e. at 60° to dam axis) and dipping vertically, generally with joint spacing of 10 to 20 feet (BCH Report No. GEO 9/85, 1985) . Joints are usually tight, discontinuous, and do not extend upward or downward across principal bedding planes. In the Block SI, S2 area of the spillway, there is a fractured zone termed the "hinge zone" [Figure 3.4] where the joints have similar orientation but are more closely spaced with tight joints at 1-3 ft spacing and with open relaxed joints at 5-15 ft spacing (BCH Report No. GEO 9/85, 1985) . Within this weakness zone, a sub-channel was eroded down to bedding plane B.P.4. The fractured "hinge zone" and associated sub-channel trend downstream from Bays 3 and 4 towards the plunge pool area but there is no subsurface information to confirm their extent. Secondary joints are vertical, with a strike of N110°-130°, and occur mostly in the sandstone and siltstone beds . 3.1.4 Historical Spills The spillway operation at Peace Canyon Dam commenced in late October 1979 while the powerhouse and generating units were still under construction. All river flows were passed over the spillway until commissioning of the first two units in April 1980. Following spills occurred in 1981, 1983, 1984, and 1996. The spillway outflow hydrograph from 1979 to 2001 is presented in Figure 3.5. Each spilling period is defined in terms of discharge magnitude and duration in Table 3.2. During the 1979/80 spillway operation, spilling over Bays 3 and 4 lasted for approximately four months during which Bays 5 and 6 were used discontinuously for about three months. Outflows through Bays 3 and 4 reached a maximum of 30,000 cfs per bay and were maintained between 25,000-27,000 cfs per bay for nine consecutive days. The peak discharge through Bay 5 and Bay 6 reached 26,500 cfs although Bays 5 and 6 were operated mainly at 10,000-15,000 cfs per bay. Total spillway flows did not exceed 68,000 cfs. The following spills of 1981, 1983, and 1984 were all of a shorter duration (maximum of two weeks in 1983) and of a lower magnitude per bay than the 1979/80 spill. However, the previous maximum spillway discharge recorded (68,000 cfs) was exceeded in 1983 with combined flows of 75,000 cfs through Bays 3 to 6. The largest spill event to date at Peace Canyon Dam occurred in 1996, from 23 June to 17 August. Spillway Gate 3 was fully open for the whole month of July discharging approximately 47,000 cfs. Meanwhile, spilling over Bay 4 was maintained at 31,000 cfs on two separate weeks. Outflows through each of Bays 5 and 6 ranged between 15,000-18,000 cfs for most of the spilling period and the peak discharge of 27,000 cfs was held for twenty hours. The maximum spillway discharge of 116,000 cfs (35% of IDF) recorded in 1996 remains the highest discharge recorded on site. 3.1.5. Scour Hole Development The plunge pool at Peace Canyon Dam has been surveyed by echo depth sounding following each spill since reservoir filling in 1979. A grid survey of the plunge pool area before spillway operation was also performed. Drawings of plunge pool soundings from 1979 to 1996 are presented in Appendix II. Table 3.3 summarizes the scour hole development based on each survey data points and interpreted elevation contours. The erosion of the plunge pool confirms the expected erosion from the two distinct zones of flow observed during spilling, the flow from the higher discharge from Bays 3 and 4 and the lower discharge from Bays 5 and 6 (BCH File No. C-PCN-12 06.12, 1997). The scour hole downstream of the spillway flip buckets has formed for the most part during the 1979/80 spillway operation. Prior to the 1979/80 spill a shallow circular hole with invert at El. 1477 ft was centred about 240 ft downstream of the centreline of Bay 4 and plunge pool floor elevations were essentially above El. 1490 ft (Figure 3.6) . Overburden conditions were however unknown. In particular, foundation construction records indicate that an overburden infilled sub-channel (related to the hinge zone) extended downstream from Bays 3 and 4 towards the plunge pool area but its extent was unknown (BCH Report No. H1879, 1986). Remnants of the downstream construction cofferdam could also have been left in place. Following the passage of river flows through the spillway in 1979/80, the volume of material removed below El. 1490 ft approached 1 million cu.ft (Table 3.3) . Tailrace dredging was performed in July 1980 to remove the accumulated debris downstream of the scour hole. The survey of 15 April 1980 showed a scour hole with invert at El. 1458 ft and the El. 1480 ft contour extending over 310 ft in the direction of spillway flows and 220 ft across the spillway bays (Figure 3.7) . Although Bays 3 and 4 were usually operated jointly during the 1979/80 spill, the lowest elevations were recorded downstream of Bay 3. Maximum scour of about 45-50 ft extended some 180-280 ft downstream of the flip buckets. Downstream of Bays 5 and 6, the plunge pool floor was lowered to El. 1482-1484 ft for a maximum scour depth of approximately 3 0 ft. As expected, no clear progression of scour was observed from 1981 to 1987 since spillway discharges per bay were all of a lower magnitude and shorter duration than during the 1979/80 spill. Successive surveys sometimes indicated a change in location of the deepest part of the plunge pool and even a regression of plunge pool depth and/or dimensions (Table 3.3). The shifting location of the plunge pool suggests there are areas in the pool where the bottom intact rock is covered by loose material (BCH Report No. GEO 9/85, 1985) . The following is noted from the surveys performed between 1981 and 1987: • Although spillway Bay 3 was operated alone on five days during the 1981 spill, the plunge pool invert shifted about 60 ft laterally downstream of the centreline of Bay 4 (Figure 3.8). An area of loose rock deposit was mapped downstream of Bays 2 and 3 next to the deepest part of the hole and large rock blocks were also identified over the plunge pool floor (Appendix II, BCH Drawing No. 1007-C14-U8440). The major axis of the scour hole showed a slight inclination with respect to the direction of spillway flows. • From 1983 to 1987, the plunge pool invert location remained relatively constant at some 200 ft downstream of spillway Bay 3. • The minimum plunge pool elevation recorded in April 1983 (El. 1450 ft) was about 10 ft below the lowest data point of the September 1981 soundings (El. 1461 ft) despite the fact that the spillway did not operate between these two surveys. The April 1983 and October 1983 sounding results (before and after the 1983 spill) were similar and the plunge pool invert elevation of both surveys (El. 1450 ft) was the deepest observed to date. Records of the two surveys data points could not be found, and therefore the interpreted elevation contours are questionable. • The 1985 and 1987 soundings of the plunge pool were both obtained on a grid. The plunge pool as surveyed on 9 October 1985 showed infilling by the reduced volume of the scour hole (Table 3.3). Video camera inspection carried out in 1985 indicated that the bottom of the scour hole was covered by cobbles and boulders in the area downstream of Bay 3 and that bedrock was exposed along the right side of the scour hole (BCH Report No. H1879, 1986) . The plunge pool survey following the largest spill to occur at Peace Canyon Dam in 1996 indicated a progression of scour, but longitudinally rather than vertically. The plunge pool as surveyed on 4 August 1996 was no deeper (invert at El. 1459 ft) but the El. 1480 ft contour extended over 410 ft in the direction of spillway flows and 240 ft across the spillway bays (Figure 3.9). The total volume scoured below El. 1480 ft since the beginning of spillway operation in 1979 had reached 581,000 cu.ft. The major axis of the scour hole showed a definite inclination of about 25° to the right of spillway flows. Despite the use of spillway Bay 3 at full capacity during the 1996 spill event, the scour hole maximum depth remained unchanged. Interestingly, the scour hole invert had migrated upstream to be located some 170 ft downstream of the limit between Bays 3 and 4. Downstream of the spillway Bays 3 and 4 typically 4 m [13 ft] of rock had been eroded from the upstream slope of the plunge pool (BCH File No. C-PCN-12 06.12, 1997). The eroded area downstream of Bays 5 and 6 was more defined laterally but not significantly deeper (minimum El. 1481 ft) . As part of a diving inspection, six large rock blocks "the size of a Volkswagen buses" were identified down the base of the slope and at the deepest part of the plunge pool. 3.2. SEVEN MILE DAM 3.2.1 Project General Arrangement The Seven Mile hydroelectric facility (SEV) is located on the Pend d'Oreille River, in southeastern British Columbia (Figure 3.1), approximately 6 miles upstream from its confluence with the Columbia River, 9 miles downstream of the Canada/United States border, and about 10 miles southeast of the City of Trail. Seven Mile is operated as a run-of-the-river plant (no storage) between Seattle City Light's Boundary Dam upstream and Cominco's Waneta Dam downstream. The Seven Mile Project consists of a concrete gravity dam with a crest length of 1138 ft and a maximum height of 262.5 ft above the foundation. Gravity blocks are adjacent to both abutments followed by a four-unit power intake section on the right and a five-bay spillway arrangement on the left (Figure 3.10). The Seven Mile reservoir has a total capacity of 85,000 acre-ft at Maximum Normal Reservoir Level (El. 1730 ft) for a surface coverage of 1000 acres. Currently, three generating units are in service for a maximum sustained capacity of 600 MW. Site preparation for a fourth unit is underway. The powerplant came into service in December 1979 and the third generating unit was commissioned a year later. In May 1988, the reservoir was raised from El. 1715 ft to the current maximum normal operating level of 1730 ft after extensions were added to the spillway gates. 3.2.2 Spillway Characteristics The Seven Mile spillway is a gated five-bay overflow spillway with flip buckets at the toe (Figure 3.11) . Table 3.4 summarizes the characteristics of design. Five 50 ft wide by 54.42 ft high vertical lift gates control the spills over an ogee-shaped crest at El. 1679 ft. Spillway bays and gates are numbered from 1 to 5 from right to left, with Bay 1 next to the power intakes. A training wall between Bays 2 and 3 divides the chute in two independent sections. The left chute profile (Bays 3 to 5) goes from the ogee crest to the dam backslope of 1V:0.8H and merges into a 60 ft radius flip bucket with a lip angle of 30°. The right chute profile (Bays 1 and 2) is similar, but with the addition of a 75 ft long downward sloping section in between the curved section of 60 ft radius. The split chute design was adopted as a compromise between minimizing spillway costs by having as short a bucket as possible and preventing impingement of the spillway jet on the powerhouse (BCH Report No. H1743, 1988). The bucket lip elevations for the left and right chutes are El. 1565.54 ft and El. 1555.54 ft, respectively. Spillway Gates 1, 2, 3, and 4 are to be used under normal conditions with preferential use of Bays 1 and 2 for spills up to 17,400 cfs. Gate 5 may be opened only if there is a danger of the dam or gates being overtopped (BCH S.0.0. 4P-36, 1997). Under normal operation, flows through Bay 3, Bay 4, and Bay 5 are restricted to 27,000 cfs, 13,000 cfs, and 8,000 cfs, respectively. These limitations for spillway gate operation aim at minimizing damages to the left bank. The Inflow Design Flood (IDF) for the Seven Mile Dam is the Probable Maximum Flood (PMF) as established by a recent study (1997) by Morrison Knudsen Corporation for Seattle City Light's Boundary Dam and B.C. Hydro's Seven Mile Dam (BCH Report No. MEP507, 1999). The peak PMF flow (376,900 cfs) can be passed through the spillway at maximum reservoir El. 1732.0 ft with all gates fully open. 3.2.3 Plunge Pool Geology The Seven Mile plunge pool bedrock consists mainly of argillite in sharp contact with granite intrusions. The granite/argillite contact is tight although adjacent argillite is somewhat fractured and altered (BCH Report No. H1743, 1988). Minor diabase and lamprophyre dykes and sills occur. Figures 3.12 and 3.13 show photographs and geological mapping of the plunge pool bedrock as exposed above normal tailwater conditions. The granite intrusion that forms the dam foundation extends 50 to 100 ft downstream of the flip buckets across the width of the spillway. Another granite dyke cuts into the argillite further downstream, just before Church Creek (Figure 3.12). The granite is a massive, medium grained, light colored igneous rock. This is a very sound rock type when fresh with uniaxial compressive strength in the range of 30,000 psi (BCH Report. No. H1743, 1988). The argillite is a grey-black, fine grained, metamorphosed sedimentary rock with weak to well developed foliations parallel to bedding. This is a moderately sound rock type with uniaxial compressive strength of about 20,000 psi in solid rock (BCH Report No. H1743, 1988). As a result of shearing and contact metamorphism, the mass of argillite is locally contorted in the vicinity of granite intrusions. Based on the extent and nature of deformation, two classes of argillite have been identified in the plunge pool area: rehealed and massive (BCH Report No. PSE362, 2001) . The rehealed argillite is characterized by a slight foliation/bedding, a chaotic fabric and numerous healed shears. The massive argillite exhibits a lack of foliation. The structural features of the argillite in the plunge pool area are the result of contact metamorphism and shearing deformation. The mass of argillite is locally contorted in the vicinity of granite intrusions and passive folds are present. The massive argillite is cut by tight and discontinuous randomly oriented joints with spacing from 2 to 12 inches. The rehealed argillite is also affected by random joints and exhibits tight and non-pervasive foliation joints with spacing of 2 to 8 inches. The foliation/bedding of the argillite was mapped at N110-120°/45° over an extended part of the plunge pool (Figure 3.13, Plate 2). Healed shears are a common feature of the argillite and have been mapped throughout the plunge pool area. Healed shears are particularly abundant in the rehealed argillite, between Line 1300 and Line 1400 (Figure 3.12), with a typical orientation of N135° and a dip of 50° to 60°. The features range in width from 0.6 to 4 inches and occur every 5 to 25 ft. Healed shears are also found along the contacts between the granite and argillite. The depth of weathering is generally shallow in the argillite rock mass (BCH Report No. H1743, 1988) . The granite is generally massive with randomly oriented joints and occasional healed shear features. From the 1988 drilling investigations of the dam foundation, the predominant joint patterns were identified as striking northeasterly and having moderate to steep dips to the southeast and northwest (BCH Report No. H2062, 1990). An important joint set oriented N40° and dipping 30° towards the left abutment was identified in the granite intrusion just downstream of the flip buckets (Figure 3.13, Plate 1). Joint spacing varies from a few inches to tens of feet, with the most highly fractured zones occurring near argillite contacts (BCH Report No. PSE401, 2001). In specific locales, moderate to heavy weathering is present along penetrative jointing (BCH Report No. H1743, 1988). No geological information is available on the eroded riverbed downstream of the spillway right chute (Bays 1 and 2) . By continuity, the scour hole probably developed in massive argillite cut by a granite dyke. Past plunge pool performance (Section 3.2.5) suggests that the protruding bedrock downstream of the spillway left chute (Bays 3 to 5) is more scour resistant than the eroded riverbed downstream of Bays 1 and 2. Downstream of Bays 3 and 4, scour occurred along the contact with massive and rehealed argillite. 3.2.4 Historical Spills The Seven Mile spillway became operational in early November 1979 and has been operated every year since. Figure 3.14 presents the spillway outflow hydrograph for the past twenty-three years of operation (1979-2001). The characteristics of the most significant spills relatively to the progression of scour are summarized in Table 3.5 and discussed below. For a detailed record and description of spillway releases from 1979 to 2001, the reader should refer to BCH Report No. N1926, March 2002. From 1979 to 1982, the spill events were increasing in magnitude every year: • Initial spillway discharges, before the powerhouse was functional, ranged between a few thousand cfs to about 3 0,000 cfs. Releases were made through Bays 1 and 2 almost exclusively. On 5 November 1979, the first day of spillway operation, Gates 1 and 2 were fully open for approximately half an hour resulting in a peak outflow of 48,000 cfs . • The first major spill occurred in the spring of 1980, from 20 April to 23 July. Spillway flows were maintained around 70,000 cfs on three consecutive days. Bay 3 was used along with Bays 1 and 2 for a seven-week period at an average discharge of 12,000 cfs. • During the 1981 spring freshet, Bays 1 to 4 were used so that flows through the right chute (Bays 1 and 2) did not exceed the previous year maximum. Average daily flows from Bay 3 ranged between 24,000-26,000 cfs for twelve days and Bay 4 released a steady discharge of 8,000 cfs for nineteen days. During spillway testing on 26-28 May 1981, a peak outflow, of 98,000 cfs through Bays 1 to 4 was maintained for over an hour. • The 1982 spill event exceeded previous spills by magnitude and duration. The spillway was in continuous operation from 28 April to 31 July 1982 during which discharges were maintained around 100,000 cfs for two consecutive days. Bays 1 and 2 spilled a maximum of 34,000 cfs per bay during forty hours. The operation of Bays 3 and 4 was similar in magnitude and duration as in 1981. The subsequent spills up to 1996 were all of a lower magnitude and shorter duration than the 1982 spill. In 1996, the spillway was in operation for nearly half the year and average daily flows reached 100,000 cfs. The maximum spillway discharge of 116,000 cfs (31% of IDF) recorded in 1996 remains the highest discharge recorded on site. The spring freshet of 1997 caused the largest spill event at Seven Mile Dam since completion of the dam. Each of Bays 1, 2, and 3 spilled a minimum daily average discharge of 29,000 cfs (and up to 36,000 cfs for Bays 1 and 2) for four weeks. A steady discharge of 9,000 cfs was released through Bay 4 for nine weeks. Overall, Bays 1 to 4 spilled together continuously for nearly ten weeks. Due to gate restrictions, Gate 5 has never been operated for more than three consecutive hours and the maximum discharge tested was 16,000 cfs in July 1983. 3.2.5 Scour Hole Development The Seven Mile plunge pool and tailrace have been monitored over time, especially during the early years of spillway operation. A total of nine topographic/bathymetric surveys have been conducted between 1979 (prior to spillway operation) and 1997 (following the spring freshet). The plan view drawings of the plunge pool surveys and tailrace soundings are presented in Appendix II. Table 3.6 quantifies the scour hole development based on the elevation data points and interpreted contours of each survey. Successive surveys of the plunge pool have shown the formation of a scour hole gradually increasing in size, although at a decreasing rate. The core of the scour hole has formed within six weeks of initial spillway operation (5 Nov. to 14 Dec. 1979) with relatively low outflows through Bays 1 and 2 (refer to Section 3.2.4). The plunge pool bedrock surface before spillway operation (Figure 3.15) was gradually benched downstream of the left chute, from El. 1565 ft at the left end of the spillway to El. 1550 ft at the limit between the two separate chutes. Downstream of the right chute, the bedrock sloped abruptly in a northwest direction to El. 1495-1500 ft near the tailrace excavation limit (Station 8 + 70) . Photogrammetry of the riverbed based on 1958 and 1972 aerial surveys indicated a shallow depression with invert at El. 1475 ft some 400 ft downstream of the powerhouse. The survey of 14 December 1979 indicated elevation points as low as 1465-1467 ft where the pre-spill survey showed a bedrock surface at approximate El. 1500 ft (Figure 3.16). The maximum scour depth of 35-40 ft was located 325 ft downstream of Bay 1 flip bucket. An estimated 74,000 cu.ft of rock was scoured below El. 1490 ft and the resulting scour hole was relatively narrow in the direction of spillway flows (Table 3.6). The shallow depression in the river channel observed prior to spillway operation appeared to have been filled with shattered rock. Tailrace soundings were performed on 7 August 1980 following the first spring freshet at Seven Mile Dam during which spillway Gate 3 was used along with Gates 1 and 2. The few elevation data points taken from the scour hole area suggested additional scour (roughly 8 ft) in the deepest part of the plunge pool as surveyed on 14 December 1979. No ground survey was performed downstream of the left chute but observations during spilling indicated that flows from Bay 3 were deflected in part by the massive rock slab. Blasting and excavation of a focussing hole down to El. 1535 ft across the spillway left chute was undertaken in the fall of 1980 to contain future spills from Bays 3 to 5. The volume of scoured rock as a result of the 1980 spill cannot be evaluated accurately (poor densification of soundings) but the 1980 soundings showed a tailrace accumulation of material up to El. 152 0 ft (Appendix II, Ref. BCH Drawing No. 224-C17-X7016) . Tailrace dredging was carried out in the fall of 1980 to restore the riverbed to El. 1500 ft. The tailrace sounding survey of August 1982 followed the 1981 spillway tests at large flows and the 1982 spring flood which exceeded previous spills in terms of magnitude and duration. A clear progression of scour was observed, mainly downstream of the existing minimum elevation points. The scour hole as surveyed in August 1982 showed a rather circular configuration with an approximate length of 175 ft and width of 200 ft (El. 1490 ft contour) (Figure 3.17) . The amount of material removed below El. 1490 ft had reached 368,000 cu.ft (Table 3.6). A minimum elevation of 1450 ft was recorded facing the centreline of the right chute at a distance of 380 ft from the buckets. This represents a maximum scour depth of 60-65 ft from the original plunge pool topography. From 1984, the plunge pool surveys showed a rather stabilized scour hole downstream of the spillway right chute. The scour hole surface (El. 1490 ft contour) was approximately 300 ft long and 200 ft wide and the maximum volume of material removed below El. 1490 ft was estimated at 550.000 cu.ft. The scour hole invert (El. 1440-1450 ft) was located 350 to 400 ft downstream of the right chute and confined between the centrelines of Bay 1 and Bay 2. An upstream progression of the scour hole towards the left chute was observed in 1984 (Figure 3.18), which could have resulted from the 1980 blasting and not been covered by the 1982 survey (Figure 3.17). The lowest plunge pool elevation (El. 1440 ft) was recorded in October 1988 following the June 1988 spillway tests during which Gate 1 was fully open. No additional scour downstream of the right chute was observed from the October 1997 survey after the passage of the largest flood (May-June 1997) since construction of the dam (Figure 3.19). The plunge pool bedrock downst ream of the left chute was surveyed in 1984 for the first time after the 1980 remedial works. The 1980 design level of excavation (El. 1535 ft) was still intact in 1984 (Figure 3.18). The following survey of the area was performed in October 1997 after Bays 3 and 4 were used extensively to pass the 1997 spring freshet. The excavated focussing hole (El. 1535 ft) was 25-30 ft deeper over a 1200 sq.ft area downstream of Bay 3 (Figure 3.19). Scour was maximum opposite to the limit between Bays 3 and 4 with a recorded El. 1503 ft. The minimum elevation downstream of Bay 4 was El. 1523 ft. 3.3. PORTAGE MOUNTAIN PROJECT 3.3.1 Project General Arrangement The Portage Mountain Project (PMD) is located at the head of the Peace River Canyon, in northeastern British Columbia (Figure 3.1), 100 river-miles upstream from where the Peace River crosses the British Columbia/Alberta border and 13 miles west of Hudson's Hope. Fourteen miles separates the Portage Mountain Project of the downstream Peace Canyon Dam. The Portage Mountain Project includes the W.A.C. Bennett embankment dam, the Williston Lake, the G.M. Shrum Generating Station, and a long spillway chute on the right abutment (Figure 3.20). The W.A.C. Bennett Dam is one of the world's largest earthfill structures with a 6,700 ft stretch across the head of the Peace River Canyon and a maximum height of 600 ft. At Maximum Normal Reservoir Level (El. 2205 ft), the dam impounds a 684 square mile reservoir, the Williston Lake, for a total capacity of 60.2 million acre-ft. The G.M. Shrum Generating Station consists of a ten-unit underground powerplant of a maximum sustainable generating capacity of 2730 MW. Power generation began in September of 1968 and all ten units were in service by February 1980. 3.3.2 Spillway Characteristics The Portage Mountain spillway consists of an approach channel, headworks including gated overflow bays and gated sluices, a long discharge channel, and a steep chute with a downstream flip bucket (Figure 3.21). Overall spillway length from the ogee crest to the bucket lip is 2400 ft. Details on design are provided in Table 3.7. Spillway releases are controlled by three 50 ft wide by 61 ft high radial gates over ogee crest at El. 2145 ft. Below are nine 6 ft wide by 8 ft high vertical lift gates in sluices with sill at El. 2105 ft. The sluice gates are in groups of three, one group under each overflow bay. From the headworks, the water is conveyed to the main discharge channel via a 350 ft long transition section of gradient 0.0421. The main discharge channel is a concrete lined trapezoidal channel with a base width of 100 ft and a slope of 0.003 over a length of 1460 ft. At the downstream end, it merges to a 380 ft long and 40° steep chute with an asymmetrical spoon-shaped flip bucket. The bucket has a 30° flip angle and a curved lip (135 ft long) at El. 1890 ft. The terminal structure of about 100 sq.ft in plan was developed by trial on a small-scale hydraulic model in order to maximize the jet dispersion and energy dissipation. The differential head between the ogee crest and the flip bucket exit is 255 ft, including a 196 ft drop downstream of the discharge channel. Under normal conditions, there is an additional fall of about 24 0 ft from the bucket lip to the Peace River. Just upstream of the impact zone, a spur dyke projecting out from the left bank separates the tailrace channel from the river (see Section 3.3.5). The radial gates are the primary means of spillway releases. To provide the smoothest flow through the channel, a three-gate operation with the two outside gates higher than the center gate is required (BCH Report No. H2417A, 1992). A ratio of 0.7 to 1, center gate to outside gates is the optimum gate arrangement (ibid.). The nine sluice gates are considered out-of-service and have not been operated since 1984 (BCH Report No. QMSGMS/03, 2001) . The Inflow Design Flood (IDF) for W.A.C. Bennett Dam is the Probable Maximum Flood (PMF) based on the Williston Lake/Probable Maximum Flood 1988 study by B.C. Hydro (BCH Report No. H2003, 1988). The IDF is 992,300 cfs and can be routed with a maximum reservoir level of El. 2209.5 ft for a peak spillway outflow of 307,200 cfs. 3.3.3 Plunge Pool Geology The scour hole at PMD has formed in sedimentary rocks, mainly thick-bedded massive shale separated by thin beds of sandstone. The rock sequence extends down to approximate El. 1500 ft where the proportion of sandstone increases and the shale/sandstone strata become equivalent in thickness. It [zone] contains nearly equal amounts of shale and sandstone and some of the shales are quite sandy in composition, and therefore similar in strength to the sandstones (B.C. and B.B Power Consultants Ltd. - Appendix A2.3, 1959). Below, at approximate El. 1440 ft, is located a 7 to 10 ft thick sequence of coal and shaley interbeds identified as the Peace River Coal Seam. The coal is hard and strong but exceedingly brittle and, therefore, badly broken in some places (B.C. and B.B Power Consultants Ltd. - Appendix A2.3, 1959). The underlying formation is composed principally of thick beds of massive, compact sandstone with a lesser amount of shale interbeds. Minor coal beds occur at about 5%. Figure 3.22 is a photograph of the right cliff rock strata surrounding the spillway bucket whereas the riverbed stratigraphy in the jet impact area is presented in Figure 3.23. In general the contacts between the various beds of the different rocks are tight (B.C. and B.B. Power Consultants Ltd. - Appendix A2.1, 1959). The strata strike relatively uniformly around N132° and dip roughly at 10° in the downstream direction. Overall, the plunge pool bedrock increases in strength with depth and in the upstream direction as the sandstone becomes predominant. Laboratory testing yielded average compressive strengths of 20,200 psi dry to 14,000 psi wet on intact core samples of sandstone and 21,400 psi dry to 2,800 psi wet on intact core samples of shale. The most common discontinuities are the joints and bedding plane fractures typical of bedded sedimentary rocks (BCH Report No. H175S, 1988) . Fracturing in the shales occur at depths less than about a 100 feet. Under relief of pressure they tend to exfoliate along and across their bedding and break down mechanically into hard buttons and cubes (B.C. and B.B. Power Consultants Ltd. - Appendix A2.1, 1959). Also found within shaly and coaly strata are several bedding plane seams of gouge and breccia termed mylonites. The mylonite seams are believed to be planes of weakness along which small shearing strains have taken place (BCH Report No. H1756, 1988) . The seams occur as discontinuous lateral patches with thicknesses from % inch to 4 inches. The underlying sandstones are sparsely jointed rocks. The only joints of any significance in the thick sandstone beds are local discontinuous bedding plane cracks (BCH Report No. H1756, 1988). Cross fracturing is however present in the thinner sandstone beds separating the shale units. The two predominant sets of joints both dip steeply with one striking northwest and the other northeast. A third, weaker set also dips steeply and strikes northerly. 3.3.4 Historical Spills In the thirty-four years since completion of the W.A.C. Bennett Dam in December 1967, excess water from the Williston Lake has been released through the spillway for just over 300 days. The radial gates were operated in 1972, 1974, 1976, 1981, 1983, 1984, and 1996. The spillway outflow hydrograph for the years 1972 to 2001 is presented in Figure 3.24. Table 3.8 summarizes each spill event in terms of magnitude and duration. The first spill event in 1972 was the most important in terms of duration and daily peak outflow. The spillway was in continuous operation for eighty-three days (13 June to 3 Sept. 1972) during which a total of 7.4 million acre-ft of water were spilled. The spillway discharge was increased to 175,000 cfs (3 radial gates opened at 31.2 ft) and maintained for eleven hours as part of spillway performance tests conducted on 10-13 July 1972. This is the highest discharge recorded at the site since the beginning of operation and represents 57% of the radial gates capacity at the Inflow Design Flood Level (El. 2209.5 ft). The latest flood at the W.A.C. Bennett Dam in 1996 caused the largest spill event to date with reference to the volume of water released. From 24 June to 17 August 1996, an approximate 9.3 million acre-ft of water was discharged through the spillway radial gates. The daily average spillway flows were maintained between 100,000 and 120,000 cfs for a total of thirty days. The maximum discharge recorded during the 1996 spill was 124,000 cfs (two outside gates open at 29 ft and centre gate at 20.5 ft). The radial gates are the primary means of spillway release. The sluices gates were tested in June 1972, April 1983, and April 1984. The sluices have never been operated with the radial gates open, nor have more than three sluices ever been opened at the same time (BCH Report No. H1756, 1988). The nine sluice gates are considered out-of-service and have not been operated since 1984 (BCH Report No. OMSGMS/03, 2001) . 3.3.5 Scour Hole Development Surveys of the plunge pool at Portage Mountain were made in 1973 and 1996 by echo depth sounding. Drawings of the two plunge pool surveys are included in Appendix II. Table 3.9 summarizes the scour hole development based on each survey data points and interpreted elevation contours. The scour hole has developed mainly during the 1972 spill event and the 1973 and 1996 sounding surveys expose a similar plunge pool bathymetry. Before spillway operation, the riverbed in the zone of jet impact was gently dipping downstream from approximate El. 1640 ft to El. 1630 ft (Figure 3.25) . A spur dyke was built along the left side of the excavated tailrace channel with a rockfill weir extending from its downstream end to the right bank. The dyke was designed to contain spillway flows and minimize the impact of the jet upon the tailrace flow. The weir was provided as a temporary means of maintaining tailwater levels above the cavitation limit of the generating units until the erection of the downstream Peace Canyon Dam. The plunge,pool was surveyed for the first time a year after the first spill event at the site in 1972. The soundings of May 1973 showed a scour hole more than 100 ft deep with the invert located near the centre of the channel (Figure 3.26) . The hole, with the outer edge taken as the El. 1620 ft contour, measured about 750 ft in the direction of spillway flows and 600 ft across, in the direction of river flows. An approximate 260,000 cu.ft of material was removed from the riverbed below El. 162 0 ft including about 12,000 cu.ft below El. 1540 ft (Table 3.9). Much of the scoured material appears to have been deposited between the center and the left bank of the river channel about 500 ft downstream of the center of the scour hole (IPEC Report No. H692, 1973). The 1973 survey did not include soundings of the weir area but photographic records indicate that the weir sustained damage due to the 1972 spill event. The toe of the tailrace channel dyke was, however, eroded during spilling. The plunge pool was surveyed again in August 1996 following the largest spring flood to occur at the site. The 1996 scour hole configuration was essentially the same as in 1972, with the invert roughly 4 ft deeper (Figure 3.27). However, scour had progressed towards the left bank so that the volume of material removed below El. 1540 ft was increased to approximately 36,000 cu.ft. The 1996 soundings indicated that the tailrace weir was for the most part eroded. To date no debris has been removed from the river channel downstream of the scour hole. 3.4. REVELSTOKE DAM 3.4.1 Project General Arrangement The Revelstoke Project (REV) is located on the Columbia River in southeastern British Columbia (Figure 3.1), about 3 miles north of the City of Revelstoke. The powerplant operates as a run-of-the-river plant. The upstream and downstream projects are Mica and Keenleyside, respectively. The Revelstoke Project consists of a concrete gravity dam within the river canyon with a long spillway chute on the right edge, an embankment dam on the right bank terrace, and a downstream four-unit powerhouse (Figure 3.28) . The concrete - dam has a maximum height of 575 ft and a total length of 1550 ft, whereas the earthfill structure extends over 3800 ft and reaches a maximum height of 260 ft above original ground. At Maximum Normal Reservoir Level (El. 1880 ft) , the two main dams impound a 28,500 acres reservoir for a total storage volume of 4.3 million acre-ft. The Revelstoke generating station has an installed capacity of 1980 MW and provisions for two additional generating units. The first three units were in service during the summer of 1984 and the fourth unit was commissioned in January 1985. 3.4.2 Spillway Characteristics The Revelstoke spillway includes headworks comprising two gated overflow bays and two intermediate level outlets, a long chute with a steep initial portion, and a terminal horizontal ski-jump structure (Figure 3.29). Overall spillway length from the ogee crest to the downstream end of the ski-jump is 1300 ft. Design characteristics are summarized in Table 3.10. Spills are controlled by two 45 ft wide by 59 ft high radial gates over ogee crest at El. 1825 ft. Below and between the overflow bays, two 17.5 ft wide by 25 ft high sector gates in outlets with invert at El. 1700 ft allow drawdown of the reservoir below spillway crest. Surface flows follow the downstream face of the dam at a slope of IV: 0.76H from the ogee-shaped crest to a 100 ft radius transition down to the beginning of the chute at El. 1690 ft. The outlets discharge onto the chute directly at this level. The concrete lined chute comprises two straight portions: an approximate 300 ft with a gradient of 0.1715 and another 472 ft at a slope of 0.002. From a minimum width of 120 ft, the chute enlarges to 150 ft and ends with a horizontal invert ski-jump. The differential head between the ogee crest and the terminal structure lip is 205 ft, including a 135 ft drop down the overflow bays. Under normal conditions, there is an additional fall of about 160 ft to the river channel. A pre-excavated plunge pool below the ski-jump was part of the final design based on recommendations of hydraulic model studies (see Section 3.4.5) . The spillway operating gates [radial gates] are the preferred method of release of non-power-related discharge (BCH L.O.O. 3P03-47, 2000). Under normal operating conditions, the spillway radial gates should be opened first and both radial and outlet gates should be opened or closed in pairs by the same amount simultaneously. The Inflow Design Flood (IDF) for the Revelstoke Dam is the Probable Maximum Flood as established in 1975 as part of the preliminary design studies (BCH Report No. 746, 1975) . The spillway (overflow bays and outlets combined) can pass the flood peak of 251,000 cfs at reservoir El. 1885 ft. 3.4.3 Plunge Pool Geology The plunge pool bedrock at Revelstoke Dam consists of a succession of highly metamorphosed sedimentary rocks including mainly micaceous gneiss and quartzite with minor schist and marble. Quartzite contains less than 5% micas and mafic minerals, quartzite gneiss from 5 to 20%, mica gneiss from 20 to 50% and mica schist more than 50% micas (BCH Report No. H1864, 1988). Bedding is pervasive, contacts between beds are generally gradational and compositions may change laterally within a particular bed (BCH Report No. 786, 1976). The main graphic drill logs of the plunge pool area (with respect to the depth of bedrock logged) are presented in Figure 3.30. The unconfined compressive strength of the metasedimentary rocks at the dam site ranges typically between 3000 and 20,000 psi. The structural features of the bedrock result from repeated ductile regional folding followed by brittle fracturing associated with the Columbia River Fault. Geological investigations indicated that the bedrock would likely fracture into 10-foot to 15-foot blocks which would in turn break up to 2 to 8 cubic feet pieces after being dislodged (BCH Report No. H1624, 1983). The main discontinuities affecting the mechanical properties of the rock are here defined: • Bedding, foliation and gneissosity in the rocks of the damsite are generally parallel (BCH Report No. 786, 1976) . In the plunge pool area (as excavated) , the rock present foliation planes oriented in higher concentration at azimuth N56° and with a dip of 26° (Figure 3.30); the main component of dip is oriented perpendicular to spillway flows towards the left hand side of the plunge pool. As a result, some of the unsupported rock composing the right wall of the plunge pool relaxed and became unstable after excavation. • Shears are a prominent feature of the bedrock at the dam site and can be classified as foliation shears (parallel to foliation) , sub-normal (to foliation), or steep. Steep shears in the plunge pool area strike around N80°. The foliation shears, which formed along weak graphite-rich layer parallel to foliation, are the most developed. Shears generally have smooth to rough and undulating surfaces with up to 2 inches of clay infilling. Spacing varies from 10 ft to 100 ft. • Small scale shears and joints occur in a wide variety of orientations throughout the site (BCH Report No. H1864, 1988). Jointing is more important in quartzite and adjacent to steep shears. Conjugate sets, i.e. discontinuities with same dip but opposite dip direction, are common. During construction, it was revealed that bedrock immediately downstream of the plunge pool, particularly near the right side, did not extend as far downstream as expected (BCH Report No. H1864, 1988). Bedrock elevations are shown in Figure 3.30 for different drill holes and seismic lines in the plunge pool/right bank area. The extent of the bedrock is limited by the presence of a buried channel which underlies the right bank terrace. In the plunge pool area, overburden consists mainly of sand, gravel, cobbles, and boulders. 3.4.4 Historical Spills Since the beginning of operation in 1983, the Revelstoke spillway has been used for less than 300 days. The spillway outflow hydrograph for the period of 1983-2001 is shown in Figure 3.31 and a summary of discharges recorded is presented in Table 3.11. The only extended spill goes back to initial reservoir filling. During initial reservoir filling, from October 1983 to April 1984, the outlets were operated at discharges varying from a few hundred cfs to about 35,000 cfs (BCH Report No. H1864, 1988). Discharge was increased up to 48,000 cfs for a short period to test the spray condition in the river downstream of the dam (ibid.). From April 1984, the excess water from the reservoir was released primarily through the overflow bays. Prototype spillway tests were conducted on 11-14 August 1986 to observe the effects of air entrainment on the freeboard along the chute walls. Testing involved a maximum spillway discharge of 70,000 cfs held for about 15 minutes, one spill up to 60,000 cfs within an hour, and two spills up to 50,000 cfs within an hour, all this over a four-day period. The spillway discharge was increased to 70,000 cfs using both the overflow bays (55,000 cfs) and the outlet gates (15,000 cfs). The peak flow of the 1986 tests, which represents 28% of the Inflow Design Flood, was maintained for a few minutes only to let the flow stabilize. The spill of August 1991 did not exceed the 70,000 cfs experienced in 1986 but caused considerable damages to the plunge pool and right bank (see Section 3.4.5). The spillway discharge was increased from 16,000 cfs to 59,000 cfs within a two-hour period, maintained at 59,000 cfs for eight hours, and then cut back to 52,000 cfs for a further thirteen hours. Spilling in 1991 lasted for a maximum of five consecutive days. 3.4.5 Scour Hole Development Echo depth soundings of the Revelstoke plunge pool have been performed following important spills. Drawings of sounding surveys conducted in 1983, 1984, 1986, and 1991 are presented in Appendix II. The scour hole progression is quantified in Table 3.12 based on each survey data points and interpreted elevation contours. The pre-excavated plunge pool lost its original shape soon after the beginning of spillway operation and gradually deepened while the right bank was sloughing into the pool. The plunge pool at Revelstoke Dam was pre-excavated based on recommendations of hydraulic model studies. The pre-excavated plunge pool was provided to control scour hole development on the right bank and to minimize undesirable backwater effects from shoaling in the powerhouse tailrace (BCH Report No. H1864, 1988). The excavation comprised a minimum level at El. 1425 ft covering an approximate 16,500 sq.ft area with a downstream level at El. 1450 ft (Figure 3.32). After an approximate two months of spillway operation, a line survey of the plunge pool topography was performed on 2 0 December 1983 (Appendix II, BCH Drawing No. 212-C14-B4371). A profile survey along the centerline of the plunge pool indicated that the rock sill forming the downstream edge of the pre-excavated plunge pool had been eroded approximately 25 ft down to about El. 1400 ft (BCH Report No. H1864, 1988). The invert of the plunge pool had also been eroded by an average of 5 ft below pre-excavated levels (ibid.). Photographic records indicated that the plunge pool had elongated towards the downstream end with a gravel bar built up immediately downstream of the pool. A first sounding survey covering the bulk of the plunge pool was carried out on 15 May 1984. The lowest elevation recorded was El. 1392 ft some 330-340 ft downstream of the concrete apron at the right of the spillway centreline (Figure 3.33). This represents a maximum scour depth of 58 ft since the beginning of spillway operation. The initial spilling period of October 1983 to May 1984 removed an estimated 5 million cu.ft of material from the plunge pool. Soundings of the spillway plunge pool and tailrace were undertaken following spill tests on 11-14 August 1986. The invert of the plunge pool had reached El. 1380 ft facing the.left half of the spillway at a distance of 300 ft from the concrete apron (Figure 3.34) . Scour had progressed laterally towards the left while the existing invert facing the right half of the spillway was covered with overburden from sloughing of the right bank. The bathymetric survey showed a tailrace accumulation of material up to El. 1450 ft comparatively to the design level of excavation at El. 1440 ft. As part of a tailrace improvement project, the channel downstream of the plunge pool was excavated to nominal invert El. 1432 ft in 1989. During the flood of August 10-11 [1991], the right bank protection work failed resulting in a section of the powerhouse access road collapsed into the plunge pool (BCH Report No. HYD.943, 1991). The following bathymetric survey (22 Sept. 1991) reflected the accumulation of material in the plunge pool with a minimum elevation (El. 1403 ft) more than 20 ft higher than the previous survey lowest point (El. 1380 ft) (Appendix II, BCH Drawing No. 212-C14-C5579). Overburden thicknesses range up to 50 ft opposite the spillway with minimum thickness of 2 0 feet near the spillway and rock shoreline (BCH Report No. N1315, 1992). Results of geotechnical investigations after the August 1991 flood showed that there was no significant change in the depth of the scour hole for most part of the plunge pool since the 1986 spillway tests, except that the downstream corner on the right side of the plunge pool has eroded to significant depth due to the presence of a weak joint in the bedrock at that location (BCH Report No. HYD.943, 1991). It is believed that the deep erosion in the right corner of the plunge pool caused the jet to deflect towards the right bank and undercut the underlying material thus resulting in the collapse of the bank protection work (ibid.). Table 3.1. Peace Canyon Dam / Spillway Characteristics | GENERAL Type of Spillway ! Gated Overflow Spillway Total Spillway Width I 440 ft across Blocks SI to S7 Overall Spillway Length1 | 240 ft (Bays 1-2) I 220 ft (Bays 3 to 6) HEADWORKS Overflow Bays i ! Number of Bays 1 6 Piers Width 10 ft Ogee Crest Elevation El. 1612 ft Gates | 6 Radial Gates Gates Dimension j 50 ft wide x 41.5 ft high TERMINAL STRUCTURE Energy Dissipation Flip Bucket Bucket Radius 55 ft Bucket Invert Elevation El. 1495 ft (Bays 1-2) El. 1510 ft (Bays 3 to 6) Bucket Lip Angle 30° (Bays 1-2) 2 0° (Bays 3 to 6) Bucket Lip Elevation El. 1512.37 ft (Bays 1-2) El. 1522.77 ft (Bays 3 to 6) Bucket Width 116 ft (Bays 1-2 combined) 234 ft (Bays 3 to 6 combined) HYDRAULICS Normal Conditions ! Maximum Normal Reservoir Level El. 1650 ft Tailwater Level with Four Units Rated Discharge2 El. 1518 ft I s Flood Conditions Inflow Design Flood | 330,200 cfs Maximum Flood Level j El. 1652.4 ft Maximum Tailwater Level2 j El. 1532 ft | Notes. 1 1. Estimated from structural drawing of spillway sections (BCH Drawing No. 1007-C14-U4798). i 2. Estimated from current tailwater rating curve (BCH Drawing No. 1007-C14-D4954). i Year Period of continuous spill Bays Total days of operation | Maximum discharge [cfs] Max daily average discharge i W s ] Max daily average discharge sustained 1 week [cfs] Max daily average discharge sustained 4 weeks [cfs] 1979 28 Oct. to 31 Dec. Total f 65 68,000 j 56,000 50,000 f 42,000 3 - 4 | 6 2 - 6 4 21,000 | 15,500 12,500 11,000 5 - 6 52 - 54 | 17,000 ! 14,000 12,500 | 11,000 1980 1 Jan. to 2 Apr. Total 107 | 60,000 I 60,000 49,500 | 36,000 |3 - 4 76 - 78 30,000 27,500 | 24,500 | 18,000 5 - 6 48 - 51 26,500 j 18,000 | 13,500 9, 000 1981 • 24 July to 5 Aug. | Total 12 67,500 [ 47,000 [ 15,000 1 1 3 I 12 25,000 j 22,500 15,000 | 4 7 | 25,000 22,500 1 6 1 2 f 22,500 j 9,500 I i 1983 4-18 July Total 15 | 75,000 f 45,000 | 29,000 1 3 - 4 15 19,000 j 15,000 11,000 1 5 - 6 12 19,000 I 11,000 6,000 1984 10-15 Oct. [ Total 1 6 | 25,000 j 15,000 j | I 3 - 4 1 6 | 13,000 | 7,000 1 1 1996 23 June to 17 Aug. Total | 55 | 116,000 | 116,000 109,000 53,000 i 1 3 [ 55 [ 47,000 j 47,000 47,000 [ 47,000 4 54 31,000 j 31,000 31,000 8, 000 5 - 6 40 - 41 27,000 21,000 18,000 Note. 1. The daily average discharge is the 24-hour average of spillway flows (12:00AM to 11:59PM). 2. "Total" refers to total spillway flows without distinction for the bays in operation. 1" Survey-Date No. of Survey Scour Hole Scour Hole : Maximum Scour Depth2 Scour Hole Scour Hole | Approximate Volume Scoured Below: Data Invert Invert Maximum Maximum | El. El. El. HiiT" Points1 El. [ft] Location [ft] [ft] Length3 [ft] Width4 [ft] i I 1490 ft j [cu.ft] 1480 ft [cu.ft] 1470 ft [cu.ft] 1460 ft [cu.ft] Oct. 1979 176 1477.4 1700R 15 + 2 ON 28,000 100 15 Apr. 1980 165 1458.2 1 1635R 1 15+45N 50 310 220 997,000 318,000 79,000 300 5 Sept. 1981 188 1461 1700R 15+35N 45 300 305 1,080,000 335,000 67,000 Apr. 1983 N/A 1450 -162 OR -14+90 60 310 190 998,000 382,000 137,000 24,000 Oct. 1983 N/A 1450 -1620R -14+95 60 350 220 982,000 388,000 116,000 9, 800 9 Oct. 1985 230 1462 164 OR 14+80N 50 310 160 865,000 261,000 44,000 22 July 1 1987 130 j 1458 164 OR 14+75N 50 330 | 190 1,110,000 373,000 110,000 600 4 Aug. 1996 434 | 1459.3 1665R 14+45N 50 410 | 240 1 1,480,000 581,000 175,000 600 Notes. --1. Within the area of interest (600 x 600 sq.ft): Station 1300R to 1900R (Easting) by Station 14+00N to 20+00N (Northing). 2. Maximum difference in elevation with respect to the original plunge pool topography (Oct. 1979). 3. Outer edge of scour hole taken as the El. 1480ft contour; length in the direction of spillway flows. 4. Outer edge of scour i hole taken as the El. 14 80ft contour; width across spillway bays. GENERAL | Type of Spillway ! Gated Overflow Spillway Total Spillway Width j 322 ft across Blocks SW1 to i SW4 Overall Spillway Length1 [ 330 ft (Bays 1-2) | 2 50 ft (Bays 3 to 5) HEADWORKS , Overflow Bays r i Number of Bays || 5 | Piers Width i 12 ft | Ogee Crest Elevation fEl. 1679 ft | Gates 5 Vertical Lift Gates | Gates Dimension | 50 ft wide x 54.42 ft high TERMINAL STRUCTURE j Energy Dissipation Flip Bucket Bucket Radius 60 ft Bucket Invert Elevation El. 1545 ft (Bays 1-2) El. 1555 ft (Bays 3 to 5) Bucket Lip Angle 30° (all bays) Bucket Lip Elevation El. 1555.54 ft (Bays 1-2) El. 1565.54 ft (Bays 3 to 5) Bucket Width 115.5 ft (Bays 1-2 combined) 17 7.5 ft (Bays 3 to 5 combined) HYDRAULICS Normal Conditions Maximum Normal Reservoir Level El. 1730 ft Tailwater Level with Three Units Rated Discharge2 El. 1523 ft Flood Conditions Inflow Design Flood ["376,900 cfs Maximum Flood Level | El. 1732 ft Maximum Tailwater Level2 j El. 1556 ft Notes. 1. Estimated from structural drawings of spillway sections (BCH Drawing Nos. 224-C14-D677/D678). 2. Estimated from current tailwater rating curve (BCH Drawing No. 224-C14-B1740). Table 3.5. Seven Mile Dam / Historical Spills Year Period of continuous spill | Bays | 1 Total days of operation { Maximum | discharge [cfs] Max daily average discharge [cfs] {Max daily | average {discharge ! sustained | 1 week I [cfs] Max daily average discharge sustained 4 weeks [cfs] 1979 5 Nov. to 2 8 Dec. Total1 30 48,000 29,000 26,000 15,000 1980 20 Apr. to 23 July ! Total 188 77,000 71,000 47,000 31,000 1 163 34,000 26,000 j 18,000 9,000 I 2 116 33,000 33,000 | 17,000 11,000 | 3 I 4 7 15,000 12,000 j 12,000 11,000 1981 1 May to 14 July Total [ 124 98,000 76,000 71,000 37,000 j 1 | 124 30,000 24,000 16,000 7, 000 1 2 I 85 26,000 25,000 19,000 7, 000 | 3 41 40,000 27,000 J 24,000 I 4 28 13,000 10,000 I 8,000 1982 2 8 Apr. to 31 July | Total 141 | 100,000 | 64,000 j 40,000 I 1 - 2 | 141 - 93 34,000 33,000 j 24,000 | 11,000 1 3 | 41 | 26,000 25,000 ! 21,000 r ! I 4 i | 39 j 8,000 8, 000 | 8,000 1996 12 Feb. to 12 July | Total 175 | 116,000 99,000 ! 82,000 i ' | 60,000 ! i [ 174 52,000 34,000 27,000 J 21,000 2 j 159 48,000 31,000 27,000 21,000 3 | 132 31,000 30,000 | 18,000 9,000 [ 4 [ 88 10,000 9,000 9,000 9,000 1997 15 Apr. to 17 July 1 Total 1 128 i | 114,000 112,000 j 107,000 F 99,000 ! I 1 | 127 40,000 36,000 | 34,000 | 30,000 _ j. 103 37,000 36,000 J 34,000 j 30,000 j 3 r~ " 1 86 f | 31,000 31,000 i 31,000 29,000 ! i 4 1 75 | 10,000 9,000 | 9,000 | 9,000 Notes. 1. Gate Operations Record incomplete for the year 1979. Spilling occurred over Bays 1 and 2 primarily. 2. Flows through Bay 1 and Bay 2 reached 64,000 cfs and 47,000 cfs, respectively, during spillway testing on 30 May 1990. 3. The daily average discharge is the 24-hour average of spillway flows (12:00AM to 11:59PM). 4. "Total" refers to total spillway flows without distinction for the bays in operation. Table 3.6. Seven Mile Dam / Scour Hole Development Survey-Date No. of Survey Data Points1 Scour Hole Invert El. [ft] Scour Hole Invert Location [ft] Maximum Scour Depth2 [ft] Scour Hole Maximum Length3 [ft] Scour Hole Maximum Width4 [ft] ! Approximate Volume Scoured Below: f El. 1490 ft [cu.ft] El. 1480 ft [cu.ft] El. 1470 ft [cu.ft] El. 1460 ft [cu.ft] Before 30 Oct. 1979 N/A 1475 -Line 1730 -Sta 8+50 . . . 29,000 2 , 100 14 Dec. 1979 71-45 1465 Line 1600 Sta 9+30 35-40 90 150 74,000 17,000 700 7 Aug. 1980 25 1461.9 Line 1615 Sta 9+05 33-38 190 >175 196,000 43,000 3,200 11-12 Aug. 1982 42 1450 Line 1660 Sta 9+50 60-65 175 200 368,000 171,000 52,000 5,200 20 Sept. 1984 N/A 1455 Line 1680 Sta 9+60 55-60 290 160 418,000 201,000 75,000 8,400 15 Oct. 1986 55 1448 Line 1660 Sta 9+55 62-67 250 160 339,000 156,000 53,000 7,000 Oct. 1988 61 1440 Line 1635 Sta 9+35 65-70 280 200 550,000 301,000 145,000 53,000 21-26 Oct. 1997 65-209 1450.1 Line 1670 Sta 9+25 55-60 280 210 520,000 269,000 116,000 31,000 Notes. 1. Within the area of interest (350 ft x 200 ft): Line 1450 to 1800 by Station 8+00 to 10+00 (bathymetry) (if second number) (600 ft x 200 ft): Line 1200 to 1900 by Station 10+00 to 12+00 (ground survey) 2. Maximum difference in elevation with respect to the original plunge pool topography (Before 30 Oct. 1979). 3. Outer edge of scour hole taken as the El. 1490 ft contour; length in the direction of spillway flows. 4. Outer edge of scour hole taken as the El. 1490 ft contour; width across spillway bays. 5. The survey performed on 30 April 1981 was not considered because taken after plunge pool blasting and biased by debris accumulation. | GENERAL 1 Type of Spillway j Long Chute Spillway Total Spillway Width [2 05 ft across Headworks | Overall Spillway Length1 2400 ft HEADWORKS | Overflow Bays Number of Bays 3 Piers Width 15 ft j Ogee Crest Elevation El. 2145 ft Gates | 3 Radial Gates 1 Gates Dimension [50 ft wide x 61 ft high | Sluices | Number of Sluices 3 Piers Width 8 ft Sill Elevation El. 2105 ft ! | Gates [ 9 Vertical Lift Gates Gates Dimension 6 ft wide x 8 ft high TERMINAL STRUCTURE Energy Dissipation Asymmetrical Flip Bucket Bucket Radius Variable j j Bucket Invert Elevation El. 1876 ft ! | Bucket Lip Angle 30° 1 — — 1 Bucket Lip Elevation El. 1890 ft i Bucket Width 135 ft (lip length) | | HYDRAULICS ! | Normal Conditions Maximum Normal Reservoir Level El. 2205 ft j Tailwater Level with Ten Units Rated Discharge2 El. 1655 ft j Flood Conditions l Inflow Design Flood 307,200 cfs ! Maximum Flood Level El . 2209.5 ft I Maximum Tailwater Level2 El. 1664 ft Notes. 1. Estimated from structural drawings of 1006-C21-U16/U32). 2. Estimated and extrapolated from Power spillway profile (BCH Drawing Nos. 1006-C14-U812, Records data for the years 1996 to 2001. Year Period of continuous spill Bays Total days of operation | Maximum discharge [cfs] Max daily average discharge [cfs] Max daily average discharge sustained 1 week [cfs] Max daily average discharge sustained 4 weeks [cfs] 1972 13 June to 3 Sept. Total 85 175,000 167,000 85,000 15,000 1974 23-31 July 3-29 Aug. Total 36 27,000 25,000 18,000 197 6 7-16 July 3-27 Aug. Total 36 36,000 36,000 30,000 r i98i 24 July to 5 Aug. Total 13 49,000 45,000 14,000 1983 10-30 May 2 8 June to 11 Aug. Total 68 87,000 80,000 60,000 7,000 r 1984 7-13 Aug 7-11 Sept. 10-15 Oct. Total 25 23,000 23,000 1996 2 4 June to 17 Aug. Total 56 124,000 122,000 ! 117,000 59,000 Notes. 1. Spillway discharges were released through the overflow bays primarily. 2. The daily average discharge is the 24-hour average of spillway flows (12:00AM to 11:59PM). 3. "Total" refers to total spillway flows without distinction for the bays in operation. 1 Survey Date No. of Survey Scour Hole Scour Hole Maximum Scour Depth2 [ft] 1 Scour 1 Hole Scour Hole Approximate Volume Scoured Below: 1 Data Points1 Invert El. [ft] Invert Location [ft] j Maximum Length3 [ft] Maximum Width4 [ft] El. 1620 ft [cu.ft] El. 1600 ft [cu.ft] El. 1580 ft [cu.ft] El. 1560 ft [cu.ft] El. 1540 ft [cu.ft] Dec.1967 May 1969 N/A 1634 15-16 May 1973 103 1519.6 160,150E 5,22 ON -115 750 600 260,000 171,000 116,000 66,000 12,000 4 Aug. 1996 609 1516.2 160,190E 5, HON -118 700 650 280,000 174,000 110,000 70,000 36,000 Notes. 1. Within the area of interest (1000 ft x 1000 ft): 159,700E to 160,700E by 4,800N to 5,800N 2. Maximum difference in elevation with respect to the original plunge pool topography (Dec. 1967 & May 1969) . 3. Outer edge of scour hole taken as the El. 1620 ft contour; length in the direction of spillway chute (Northing). j 4. Outer edge of scour hole taken as the El. 1620 ft contour; width perpendicular to spillway chute (Easting). Table 3.10. Revelstoke Dam / Spillway Characteristics | GENERAL Type of Spillway [Long Chute Spillway Total Spillway Width | 224 ft across Blocks SI to S3 Overall Spillway Length1 1300 ft | I HEADWORKS Overflow Bays Number of Bays 2 Pier Width Ogee Crest Elevation El. 1825 ft Gates 2 Radial Gates | Gates Dimension 45 ft wide x 59 ft high Outlets | Number of Outlets 2 | Pier Width 1 10 ft | Sill Elevation El. 1700 ft Gates [ 2 Outlet Sector Gates 1 Gates Dimension 17.5 ft wide x 25 ft high TERMINAL STRUCTURE J Energy Dissipation Horizontal Ski-Jump | Bucket Radius 0° Bucket Invert Elevation El. 1620 ft | Bucket Lip Angle 0° Bucket Lip Elevation El. 1620 ft Bucket Width fiio ft HYDRAULICS 1 Normal Conditions Maximum Normal Reservoir Level El. 1880 ft Tailwater Level with Four Units Rated Discharge2 El. 1459 ft Flood Conditions | Inflow Design Flood 251,000 cfs [ Maximum Flood Level El. 1885 ft | Maximum Tailwater Level2 El. 1481 ft Note. 1. Estimated from structural drawings of U11/U4 7/U4 8) . 2. Final design values based on improved spillway profile (BCH Drawing Nos. 212-C21-tailrace channel. Year Period of continuous spill 1 Surface 1 Bays/ | Outlets | Total I days of | operation Maximum discharge [cfs] 1 Max daily average discharge [cfs] Max daily average discharge sustained 1 week [cfs] Max daily average discharge sustained 4 weeks [cfs] 19831 Oct. to 31 Dec. Outlets 1 -90 48,000 r i N/A N/A N/A 19842 1 Jan. to 9 May Outlets 124 40,100 30,500 29,600 20,300 ? — ' — — ! Surface | Bays 1 3 2 24,200 24,200 8,000 1985 14-19 Feb. Surface j Bays ! 6 33,600 31,100 1986 25-28 July Spillway Testing: 11-14 Aug. Total 8 70,000 | ... | Outlets j 5 15,000 i 10,800 f Surface Bays 1 l | 3 55,000 1990 14-18 June Surface Bays ! j 5 ! 18., 700 17,100 1991 9-14 Aug. 18-21 Aug. j Surface 1 Bays ! ! 12 59,200 35,500 1997 2-5 Oct. Surface Bays | 4 j 18,400 | 18,300 1 Notes. 1. Gate Operations Record missing for the year 1983. 2. The spillway discharges were released through the overflow bays from 17 April to 9 May 1984 . 3. The daily average discharge is the 24-hour average of spillway flows (12:00AM to 11:59PM). 4. "Total" refers to total spillway flows without distinction for the bays/outlets in operation. , Survey Date No .of Survey Scour Hole Scour Hole Maximum Scour Depth2 Scour Hole Scour Hole Approximate Volume Scoured Below: Data Invert Invert Maximum Maximum El. El. El. El. Points1 El. [ft] Location [ft] [ft] Length3 [ft] Width4 [ft] 1425 ft [cu.ft] 1410 ft [cu.ft] 1400 ft [cu.ft] 1390 ft [cu. ft] As Excavated 114 1425 Extended Area 110 210 | -15 May 1984 |"N/A 1392 44,52 0E 42,060N 58 N/A 290 >1,345,000 310,000 50,000 ; Following spill tests on Aug.1986 j N/A I 1380 44,63 0E 42,010N 70 . 480 370 1,158,000 293,000 107,000 30,000 22 Sept. 1991 141 1403 44,590E 42,170N 60 350 430 442,000 5, 000 ! — Notes. 1. Within the area of interest (700 ft x 500 ft): 44,300E to 45,000E by 41, 900N to 42,40ON. 2. Maximum difference in elevation with respect to the original plunge pool topography (As Excavated). 3. Outer edge of scour hole taken as the El. 1425 ft contour; length in the direction of spillway chute 4. Outer edge of scour hole taken as the El. 1425 ft contour; width perpendicular to spillway chute. NORTH T ERR11 PORTAGE MOUNTAIN PROJECT ( W.A.O. BENNETT DAM) •PEACE CANYON DAM t Figure 3.1. Geographic Location of Dam Sites EL.1522.77 FT figure .3.3. Peace Canyon Spillway Sections A & B: 1007-C14-D4243 Section C: 1007-C14-U4321 SECTION © 19+00R 18+00R 17+00R 16+00R 15+00R Station [ft] Figure 3.4. Peace Canyon Dam / Spillway Foundation Geology E! /6/2.00 k Im 1600 § Uj Uj 1500 SECTION © 1400 B.R2 El. 1612.00 k [H £ o £ ki SECTION © 1600 I500 I400 H-IQ £ i-i (t> 13 m OJ n cu o pj P •c o 3 a si) 3 CO >d H-M M s: pj ^ o c r t l-h M 0 S ffi o 113 ii SD •O 13* vo <1 vo r t O to o o Spillway Discharge [cfs] 28 Oct. 1979 ltO_2Ap Qmax != 68,000 cfs Q = 50,000 60,000 r. 1980 cfs for 37 days 25 July to 5 Qmax i= 67, Q = 40,000 -Aug. !. 1981 cfs (Bays 3 47,000 cfsfor 500 4-6) 4 days _(Bays 3-4-6) 4-18 Qmax A July 40, 1983 75,000 cfs .000 - 45.00C cfsfor 3 day; 10-15 Oct. 1984 Qmax 1= 25,000 cfs for  3 hr (Bays 3-4) o 20 Jan 1987 Qmax = 25,000 cfs tor 1 hr (Bay 3) (jd oo o ISJ o o U5 CI 24 June Qmax Q > 110,000 to 17 Aug. 115,000 -cfsfor M June 2001: Qmax |= 18,1 Low 000 Tests Flow Spillway cfs (Bays 3-4 and Bay 3 only) PEACE CANYON DAM I PLUNGE POOL BEFORE SPILLWAY OPERATION I 1 so# 3 1 53 14Sn s^v -V • Survey Data Points Ref.  BCH Drawing No. 1007-C14-X8423 ^  < PEACE CANYON DAM I PLUNGE POOL SURVEY OF 15 APRIL 1980 Ref. BCH Drawing No. 1007-C14-X8423 PEACE CANYON DAM I PLUNGE POOL SURVEY OF 5 SEPTEMBER 1981 > • Survey Data Points Ref.  BCH Drawing No. 1007-C14-U8440 Q) UJ 145q PEACE CANYON DAM / PLUNGE POOL SURVEY OF 4 AUGUST 1996 • Survey Data Points Ref.  BCH Drawing No. 1007-C11-D1983 Figure 3.10. Seven Mile Dam / General Arrangement Figure 3.11. Seven Mile Spillway Figure 3.12. Seven Mile Dam / Plunge Pool Bedrock - Geological Mapping 1 1 1 1 1 1 1200 1300 1400 Line ROCK TYPES ARGILLITE - REWORKED, MASSIVE TO SLIGHTLY LAMINATED ARGILLITE - REWORKED, HEALED SHEARS, GRAPHITIC GRANITE - PYRITIC, MED-GRAINED DIABASE - ALTERED LAMPROPHYRE - ALTERED SEVEN MILE DAM PLUNGE POOL BEDROCK - GEOLOGICAL MAPPING 1200 1100 c o -t—> 03 CO 1000 1500 [ft] Notes. 1. Geological Mapping Completed in August & October 1999 (Ref.  BCH Drawing No. 224-C14-C1599) 2. Topography As Surveyed in October 1997 SYMBOLS JOINT; INCLINED, VERTICAL \ V FOLIATION OR BEDDING HEALED SHEAR (APPROXIMATE) - S , GEOLOGICAL BOUNDARY; V DEFINITE, APPROXIMATE 120/45 ORIENTATION STRIKE / DIP (RIGHT HAND RULE) 1600 1700 1800 SEVEN MILE DAM PHOTOGRAPHS OF PLUNGE POOL BEDROCK PLATE 2 -ARGILLITE (MASSIVE) Facing Right Abutment Figure 3.13. Seven Mile Dam / Photographs of Plunge Pool Bedrock •8 <u cn u tn >-(D J '5. CO SEVEN MILE DAM Spillway Outflow Hydrograph - 1979 to 2001 Years of Spillway Operation Ref. BCH Drawing No. 224-C11-D7025 • Survey Data Points Ref.  BCH Drawing No. 224-C11-D117 Ref. BCH Drawing No. 224-C11-D119 Ref. BCH Drawing No. 224-C11-U217 SPILLWAY PLUNGE POOL WAC BENNETT DAM Figure 3.20. Portage Mountain Project / General Arrangement Figure 3.21. Portage Mountain Spillway U/S ELEVATION PLAN AND PROFILE 0 200 400 FT SCALE: fa - I I H — I ROCKFILL GROINS EL.2145 FT_ HEADWORKS 1890 FT CHUTE SCALE: SECTIONS 0 100 200 FT fa - I I I 1 Figure 3.22. Portage Mountain Project / Right Cliff Rock Strata PORTAGE MOUNTAIN PROJECT 6.000N 5.000N PLUNGE P OOL PLAN UJ o o o <o Note. Reference  BCH Drawing Nos. Plan: 1006-C11-U6 Section A: 1006-C11-U14 Section B: 1006-C11-U19 LEGEND Shale Sandstone Peace River Coal Seam Water Line (~EI. 1650 ft) Figure 3.23. Portage Mountain Project / Plunge Pool Geology SECTION A SECTION B -J oo 27 Aug. 1976 24 July to 5 Aug. 1981 45,000 cfs for  1 Spillway Discharge [cfs] cfs for  11 days day . 28 June to H Aug. 1983 :Q = 75,000-Sept., & 10-15 Oct. 1984 80,000 cfsfor  3 day;; 24Jun Qmax •Ts-ui o u> o e to 17 Aug. 1996 : 124, Q > 100,000 JOOcfs; cfs for  30 days CO •g I «< -o o g 0 CD § m 1 o 3 l IQ H S £ , "O K 2 m o ^ N) O O PORTAGE MOUNTAIN PROJECT I PLUNGE POOL ~W BEFORE SPILLWAY OPERATION Ret. BCH Drawing Nos. 1006-C11-U6/U144 1006-C14-C1261 PORTAGE MOUNTAIN PROJECT / PLUNGE POOL SURVEY OF 15-16 MAY 1973 • Survey Data Points Ref.  BCH Drawing No. 1006-C14-B1262 ^ 1650 •S 160Q U 155(K UJ G o 0 PORTAGE MOUNTAIN PROJECT I PLUNGE POOL SURVEY OF 4 AUGUST 1996 • Survey Data Points Ref.  BCH Drawing No. 1006-C11-D1107 °Oo PLUNGE CHUTE iL HEADWORKS-SURFFACE BAYS El.1700 FT PLAN AND PROFILE „ 0 80 160 FT SCALE: U.I I I =J TERMINAL STRUCTURE El. 1425 SECTIONS „„„ ,, 0 80 160 FT SCALE: h- I I | t HEADWORKS-OUTLETS CHUTE Figure 3.29. Revelstoke Spillway 44,300 44,400 44,500 Easting [ft] 44,600 44,700 REVELSTOKE PLUNGE POOL - GRAPHIC DRILL LOGS 44,800 44,900 45,000 42,400-42,300-42,200-cn a 42,100 42,000-41,900-z o t-Sond, grovel, cobbles 8 boulders (GW) Sandy grovel with some cobbles Ouorlzite gneiss Z a: id i-UJ a: iiif UJ a: 3 ^ _i a. o >-1 tr o UJ o Grovel, cobbled and boulders with some sond M p i Grovel B sond Cobb^grovel If") send i  with som^  >bbies, gravel boulders a sand Ouarl H l ! ^ SuortzTte V gneiss 12 feldspar I Quartzite i ^ 3 I- >-UJ cc a. uj 9 § o UJ WL s w M& I 'Hl'lU'I'I'll OH 74-2 DH 7 3 - 9 oz o < <r> o „ 50 IOO 1538 Overburden Broken, loose, weathered vCv-3 Quartzite gneiss 150 1437 a: o CO UJ o Thick bedded Quartzite gneiss •Altered 'Quortzite gneiss t h i n bedded Quartzite gneiss \Morb le Thick bedded Quartzite gneiss Thin bedded Quartzite gneiss O u. > IE UJ > o o 0 IOO 1 2QO D.H. 7 5 - 18 Note. Reference  BCH Drawing Nos. Plan: 212-C14-U5499 Drill Logs: 212-C14-D499/D500 D502/D508 Figure 3.30. Revelstoke Dam / Plunge Pool Geological Information l-1-IQ C i-i n> Spillway Discharge [cfs] (1) < fD M 01 rt O W fD D 0) 3 to p-£ flj O e rr l-h I-1 O £ W >< & O IQ H (1) tr vo CO u> r t O to o o - < fD Q) -1 LT) •g. W ai -< O T3 fD O =3 25-28 July 1986: Outlets Q = 10,800 14-18 June 1 Q = 17,000 f Qmax :fs for 1 Jan. 17 Ap Oct. - pec. 1983: Outlets Qmax = 48,000 cfs Apr. 1984: Ou lay 1904: Overflow June & July 1984: Outlets, to 16 . to 9 14-19 Qmax Feb. 1^85: Overflow  Bays 33,000 - 34,000 cfs for  30 hrs '3 da^fe 390: Overflow )r 2 days 2-5 Oft.  1997: Overflow  Bays 18,000 cfs for  48 Jets Overflo Says w Bay: Bays 9-14, Qmax Q-5 18-21, & 25 = 59,000 cfs 2,000 cfs for hrs 11-14 Qmax Aug. = 70, 1986 ,000 Aug . 1991: Overflow  Bays "or 8 hrs hrs 13 : Spillway cfs for  < 15 min ;reeboard Te: [Overfl ui o NJ o o o ts 3w Bays & Outlets) cn T3 CD o c 3 to o m * £ f  5 5 § QJ _ -a P IT > UD 00 U) NJ O O S 8 8. m 1 s I S REVELSTOKE DAM I PLUNGE POOL BEFORE SPILLWAY OPERATION End Ski-Jump • Survey Data Points Ref.  BCH Drawing Nos. 212-C21-D7126/7127 REVELSTOKE DAM I PLUNGE POOL SURVEY OF 15 MAY 1984 End Ski-Jump • Survey Data Points Ref.  Drawing Nos. 212-C21-D7125/7127 Figure 3.34. Revelstoke Dam / Plunge Pool Topography As Surveyed Following Spillway Tests on August 1986 REVELSTOKE DAM / PLUNGE POOL FOLLOWING SPILL TESTS ON AUGUST 1986 • Ref. BCH Drawing No. 212-C21-D133 CHAPTER IV COMPARATIVE  ANALYSIS OF FACTORS AFFECTING PLUNGE POOL SCOUR Plunge pool scour is dependent upon the interaction of many factors specific to a particular dam site. Most authors have put the emphasis on hydraulic parameters (Mason and Arumugam, 1985), while others insist on the importance of plunge pool geological conditions (Spurr, 1985; Annandale, 1995) . To assess pool scour accurately, the influence of the spillway layout, the spill duration, the geology and the relative rates of scour development between the outer pool and the live-scour hole must be taken into account (Spurr, 1985). In this chapter, the main factors susceptible to affect plunge pool scour are examined and the importance of each is established through a comparative analysis of plunge pool performance at the sites of study. 4.1 SPILLWAY CHARACTERISTICS One would expect that spillways with different degrees of energy dissipation on the spillway itself and different arrangements for the free-falling jet to exhibit different plunge pool scour depth for otherwise equal hydraulic parameters (Novak, 1985). The ways in which the change in dissipator could affect scour are principally by causing variations in jet dispersion laterally, in aeration and impact angle and also by causing varying amounts of head loss through or along the conveyancing structure (Mason, 1985) . Key design features of the four spillways at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam are compared in Table 4.1. The main spillway characteristics likely to affect scour development are examined through comparative observations between sites of study. Empirical data from experiments and field observations indicate that the two main parameters affecting the scour depth are the difference in elevation between the upstream reservoir and the downstream tailwater level (or total head H) and the discharge per unit width of flow (or unit discharge q) . These two parameters are closely related to the spillway design. The total head is associated with the size of the project and the unit discharge depends on the final width of the conveyance structure. At Portage Mountain Project and Revelstoke Dam, the spillway is separated from the dam and located high on the right abutment of the canyon as opposed to Peace Canyon Dam and Seven Mile Dam where the spillway is part of the concrete dam. The total head is considerably higher in the first case with the maximum head at Portage Mountain Project (550 ft) and the minimum head at Peace Canyon Dam (132 ft) . Accordingly, scour depth is maximum at Portage Mountain Project and minimum at Peace Canyon Dam. Under normal operation, spillway flows at Peace Canyon Dam and Seven Mile Dam are routed through four bays each about 60 ft wide. At Portage Mountain Project and Revelstoke Dam, flood flows are released through a long spillway chute with a terminal width of 135 ft and 150 ft, respectively (Table 4.1). Flows through the Portage Mountain spillway have the particularity of being contracted from a clear span of 150 ft at the crest to 135 ft at the bucket lip. For a same spillway discharge, the flow is generally more concentrated at the exit of the Portage Mountain spillway, which could explain the greater scour capacity of the jet. The consensus for the last, 60 years has been that q is far more dominant than H in assessing scour depth (Mason, Discussion, 1989). The basis for this argument is that for any given dam, the differential elevation between reservoir and tailwater levels (H) remains relatively constant and yet, major scour advances occur in response to increased spillway discharges (q). However, the variation in scour depth between sites is probably best explained by the relative head of each project with respect to one another. The importance of head drop from reservoir to tailwater level is particularly noticeable when plunge pool development in similar bedrock is compared, for instance at Peace Canyon Dam and Portage Mountain Project. During the 1996 spring freshet, daily average spill at the upstream Portage Mountain Project reached 122,000 cfs (q=900 cfs/ft) and scouring progressed slightly to a maximum depth of 118 ft while at the downstream Peace Canyon Dam, spillway Bay 3 spilled 47,000 cfs for a month (q=810 cfs/ft) and the maximum scour depth remained stable at 50 ft. Considering the two very distinct spillway layouts in this study (gated overflow vs. long chute), both unit discharge and total head drop are important parameters in the comparative analysis of scour development at the four dam sites. Head losses on the spillway face and air entrainment of high-velocity flows are intimately related. Clearly, the more energy lost along the conveyance structure, the less is available for scour processes. Most authors agree that aerated water will produce a lesser scour than unaerated, "solid" water (Mason and Arumugam, 1985). Energy losses on the downstream face of a spillway occur during development of the turbulent boundary layer and in the fully developed turbulent flow. The air entrainment starts at a point where the boundary layer from the bottom intersects the top water surface (Chaudhry, 1993) . For long spillway chutes, the conditions of fully developed turbulent flow predominate and hence the air entrainment process is maximized. Although the Revelstoke spillway chute is shorter than the one at Portage Mountain Project, the air slots configuration results in the entrainment of a large amount of air and excessive bulking of flow occurs. Based on the fact that plunge pool scour is greater at Portage Mountain Project and Revelstoke Dam, the energy losses and cushioning effect of air entrained on the spillway surface can be considered small relatively to the remaining energy that dissipates into the pool. The spillway terminal structure is a key element in the dissipation of flow, energy before the jet impinges the pool surface. Energy dissipation is achieved by promoting jet dispersion in the atmosphere. The fall height adds to the energy dissipation through air resistance. At the Portage Mountain Project, the final spillway bucket geometry was developed by trial on a small-scale hydraulic model in order to maximize jet dispersion and energy dissipation. The maximum depth of scour was less than 1/3 of that obtained with the preliminary design with the resultant scour volume reduced seven times (Johnson and Alam, 1969) . From the model studies, it was anticipated that for prototype spill of approximately 100,000 cfs, the flow would be so well dispersed that practically all of the energy would be dissipated in the air, with the water reaching the riverbed in the form of a heavy spray (ibid.). If the high spillway discharges which caused deep scour at the Portage Mountain Project did reach the riverbed as a fully aerated jet, than the attenuating effect of air entrainment on the scouring capacity of free jets is questionable. Some authors (Taraimovich, 1978; Yildiz and Uziicek, 1994) consider that the impact angle affects the depth of scour; the steeper the impact angle, the greater the depth of scour. The impact angle is dependent upon the flow energy at the bucket exit, the bucket lip angle, and the fall height to the tailwater level. When flip buckets are near the tailwater level, bucket lip angle and impingement angle are similar. Conversely, the angle of impingement of a jet issued from a horizontal ski-jump (e.g. Revelstoke spillway) can be near vertical if the fall distance is sufficiently long. With a 235 ft fall from the bucket to the river, flows issued from the Portage Mountain spillway impact the tailwater surface at an approximate angle of 49° with the horizontal. At Peace Canyon Dam, with the small difference in elevation between the spillway buckets and the tailwater level, the free jet projected upwards at 20° (Bays 3 to 6) strikes the water surface at about the same 20° angle (downwards). Accordingly, the scour hole at Portage Mountain Project is the deepest and the plunge pool at Peace Canyon Dam the shallowest. Spillway layout is observed to affect plunge pool scour development to some extent. The influence of head drop between reservoir and tailwater levels and jet impingement angle with the tailwater surface is particularly noticeable when scour development at Peace Canyon Dam and Portage Mountain Project is compared. The effects of energy losses and air entrainment on the spillway structure, which are more important at Portage Mountain Project and Revelstoke Dam, were not perceived. As well, the extent to which jet dispersion and aeration through free fall attenuate the scour processes is uncertain. The impact of spillway discharge per unit width on the progression of scour is discussed further in Section 4.3. 4.2 PLUNGE POOL GEOLOGY Comprehensive reviews of scour evaluation methods show that attempts to understand the dynamic interaction of spillway hydraulics with the plunge pool geology have been rare (Mason and Arumugam, 1985; Whittaker and Schleiss, 1984; Schleiss, 2002). Mason (1985) recognizes that "it is a widely held view that ultimate scour depth has little to do with rock strength". Certainly, major scour holes have developed in massive hard rocks such as granite and gneiss, as well as in weaker sedimentary rocks (Mason, 1993) . Interestingly, the hole shape tends to depend more on hydraulic factors than on geological ones (ibid.). From Spurr's (1985) point of view, "it is evident that geometrically similar plunge pools with identical jets will scour to different lengths over the same spill period when the bedrocks have different erosion strengths". Both Mason (1985) and Spurr (1985) agree that variable geology in the plunge pool can lead to unconfined conditions and asymmetrical development. The latest plunge pool configuration at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam is presented in Figures 4.1 and 4.2. From the review of plunge pool development at each site, the effects of geological conditions on the scour development were clearly observed. At Peace Canyon Dam, the flat bedding planes appear to be limiting the depth of erosion in favor of lateral expansion of the bottom of the plunge pool (BCH Report No. HY296, 1985). Following increased spillway discharges of the 1996 spill, the surveyed plunge pool had extended over about 80 ft in the flow direction but was not found to be deeper. The steeped profile and flat base of the scour hole are indicative of the influence of bedding planes and steep joints (Figure 4.1) . As well, the scour hole orientation clearly defined from the 1996 soundings is closely associated with the plunge pool geological features (Figure 3.4) . Orientation of the plunge pool axis is similar to jointing in the foundations of Blocks SI to S4 and of the closer jointed and previously eroded "hinge zone" under parts of Block SI [refer to Figure 3.4] (BCH Report No. H1742, 1987). The Peace Canyon plunge pool started to form in the bedrock weakness zone downstream of spillway Bays 3 and 4 and developed in accordance with joints and bedding planes features. The main scour hole at Seven Mile Dam (downstream of spillway Bays 1 and 2) is confined and rather symmetrical in the upstream/downstream direction, but its left side slope is much steeper than the right one (Figure 4.1) . Were the plunge pool development to occur independently of bedrock conditions, one would expect some lateral symmetry because spillway Bays 1 and 2 have typically been operated together. In fact, it appears that the protruding bedrock downstream of Bays 3 to 5 is more resistant to scour than the riverbed downstream of Bays 1 and 2. In the first two weeks of spillway operation, flows of 25,000-29,000 cfs through Bays 1 and 2 combined (q=220-250 cfs/ft) scoured a hole roughly 35-40 ft deep. Contrastingly, flows through Bays 3 to 5 were deflected by the massive rock slab and the plunge pool had to be blasted and excavated across Bays 3 to 5 to contain future spills (1980 remedial works). In 1997, the excavated level was scoured down by 25-30 ft after Bay 3 was operated at approximately 30,000 cfs (q=500 cfs/ft) for a month. Scour downstream of the spillway left chute is localized, whereas the main scour hole from the right chute spills extends over an approximate 150 ft radius. The scoured plunge pool at Portage Mountain Project is deep and fully confined (Figure 4.2), as expected by the lateral homogeneity of the sedimentary rock and the similarity in strength of the shale/sandstone strata. The scour hole is essentially circular in shape but with somewhat straight segments oriented northeast and northwest, equivalent to the predominant joint patterns. The side slopes of the hole are irregular but the effect of the downstream dip of the stratification is observed towards the left bank. Along the dipping direction of bedding planes (Figure 3.23), the plunge pool wall is steeper on the downstream face than on the upstream face. The Revelstoke plunge pool is a typical case of problematic asymmetrical development caused by the low resistance of the peripheral rocks with respect to the bottom rock directly impinged by the spillway jet (Figure 4.2) . In this case, the plunge pool bedrock is limited by an adjacent buried channel and the right bank is composed mainly of sand, gravel, and cobbles. During spilling, the flow energy is deflected by the more resistant bedrock and return currents are formed that undermine the toe of the right bank slope inducing failure. Although it is clear that the geological conditions at each site have influenced the plunge pool development, other factors, mainly the spillway hydraulics, might have had a greater impact. Whereas plunge pool bedrock at Peace Canyon Dam and Portage Mountain Project is comparable in geology and structural features, the shallow plunge pool with a stepped profile and flat base at Peace Canyon Dam contrasts markedly with the confined Portage Mountain pool of more than 100 ft deep. In this example, the jet impingement angle and total head associated with the spillway layout are believed to be important factors. 4.3 SPILLWAY DISCHARGES Plunge pool scour depths on prototypes are usually associated with spillway outflow hydrographs and the general assumption made is that the observed scour depth corresponds to the peak flow of previous spills. Considering the intermittent and variable nature of prototype floods, this simplification should be taken with care. For instance, large flows associated with hydraulic spillway tests are generally of a duration too short (less than an hour) to be indicative of the depth of scour. However, they do promote hydraulic fracturing of the rock mass and enhance the scour progression by smaller spills. It must be remembered that scouring is a dynamic process, and so magnitudes, frequencies and durations of spilled discharges need to be taken into consideration (Whittaker and Schleiss, 1984) . Figure 4.3 presents the spillway flow duration curve for each site (up to the last plunge pool survey date) with information on the duration of main spills and historical peak flows. A comparison of spill history at each dam site in terms of frequency, duration, and magnitude is made and the respective effects on plunge pool development are observed. The spill occurrence at dam sites depends on the hydrologic conditions of the river basin and on storage regulations (Seven Mile is a run-of-the-river plant as opposed to Portage Mountain Project which has a large storage capacity). The Seven Mile spillway has been operated every year since the beginning of operation in 1979 and has certainly been the most frequently used spillway of the four dam sites. In spite of this, the Seven Mile plunge pool downstream of the preferred spillway bays (Bays 1 and 2) has remained stable since 1984. On the other hand, the Revelstoke spillway has been used in rare instances but each spill event seemed to have activated the scour progression. At Revelstoke Dam, part of the right bank protection work was eroded throughout each spillway operating period, leaving the right slope progressively more vulnerable to failure. The frequency of spills at Peace Canyon Dam corresponds to that of the upstream Portage Mountain Project since there is limited storage between the two sites. The short-term but more frequent spills do not appear to affect the plunge pool development at Peace Canyon Dam. The same statement probably applies to Portage Mountain Project but could not be verified because the only two surveys of the plunge pool have been made after major spills. Each of the four dam sites has experienced at least one extended spill: • At Peace Canyon Dam, river flows were routed through the spillway for about five months before commissioning of the generating units. The major spill event of 1996 lasted for eight consecutive weeks. • At Seven Mile Dam, continuous spill exceeding a month in duration has been common during spring freshets of the 1979-2001 period. In 1996, the spillway was in operation for nearly half the year. In 1997, total spillway discharges were maintained above 100,000 cfs for about four weeks. • Three spill events at Portage Mountain Project have lasted more than a month and occurred in 1972, 1983, and 1996. The 1996 spill event was the most important in terms of continuous operation and volume spilled. • The only extended spill at Revelstoke Dam goes back to initial reservoir filling, from October 1983 to April 1984. Subsequent spills lasted at most six consecutive days. The effect of spilling duration on the progression of scour is closely related to the issue of scour rate. As discussed in the next section, plunge pool performance data of the four studied sites seem to indicate that scour progresses rather quickly and that extended spills are not a condition to plunge pool development. For each dam site, the spillway discharge is recorded every hour and a daily spill value, averaged over a twenty-four-hour period, is derived. The recorded magnitudes of peak flow and maximum daily average spill with the associated plunge pool response are reviewed: • The highest magnitude of spillway flows was observed at Portage Mountain Project in 1972 with 175,000 cfs following which a scour hole more than 100 ft deep had formed. The peak discharge was held for eleven hours before which flow was maintained at 160,000 cfs from which the daily average discharge of 167,000 cfs. • The maximum flood discharge to be passed over the spillway at Peace Canyon Dam and Seven Mile Dam is the same with 116,000 cfs (4 bays) during the 1996 spring freshet. The discharge was maintained for twenty-four hours at Peace Canyon as opposed to Seven Mile for which the daily average spill was smaller. Plunge pool scour depths are similar at both sites (50-60 ft) but were attained before 1996 following smaller spillway flows. In fact, the plunge pool bottom elevation at Peace Canyon Dam was reached following the 1979/80 spill with a maximum recorded discharge of 68,000 cfs (4 bays). At Seven Mile Dam, average spillway flows of 25,000-29,000 cfs through Bays 1 and 2 combined scoured the bedrock down to 35-40 ft depth in 1979. • The maximum discharge to be released from the Revelstoke spillway was 70,000 cfs (held for about fifteen minutes) during the 1986 spillway tests. This is less than half the peak flow experienced at Peace Canyon and Seven Mile dams and yet, the scour depth increased and reached 70 ft following the tests. The maximum daily average spill at Revelstoke Dam was 35,500 cfs in 1991, suggesting that spillway flows greater than 40,000 cfs have been rather unusual and of short duration. Along with the magnitude of spillway flows, the width of flow should be considered. Large spillway flows at Peace Canyon Dam and Seven Mile Dam were released through four bays of about 60 ft width whereas spills at Portage Mountain Project and Revelstoke Dam were routed through a single chute with a terminal width of 135 ft and 150 ft, respectively. In addition, variable spillway discharges from one bay to another are common at Peace Canyon Dam and Seven Mile Dam due to the preferential use of certain spillway bays. In any cases, the maximum spillway discharge per width of flow (or unit discharge) was witnessed at Portage Mountain Project. In Chapter V, the unit discharge is used to calculate scour depth. The magnitude of spillway flows seems to have a major impact on plunge pool scour compared to the frequency and duration of spills. However, as seen in the case of Seven Mile Dam, scour can result from relatively low discharges, perhaps suggesting that flow velocity (associated with the head drop from reservoir to bucket lip) is more important than the actual discharge in the early stages of plunge pool development. At each site, the scour hole was formed following a single spill event and not by repetitive spills; scour progression corresponded to an increase in spillway flows. Daily peak flows were observed to be more effective in initiating scour than extended spills. 4.4 SCOUR RATE As mentioned in Chapter II, most equations for the prediction of plunge pool scour ignore time as a parameter. In fact, the time progression of scour is one of the most conflicting aspects of plunge pool formation. As early as 1950, Hunter Rouse commented that "scour was proportional to the geometrical progression of time and as such a final equilibrium depth could not be expected" (Mason and Arumugam, 1985). To support the concept of a maximum, ultimate scour depth, Mason and Arumugam (1985) state that: 1) Many of the major prototype scour developments analyzed reached their measured depths after only a few days of flood discharge; and 2) Results of the data analyses suggest that a similar order of accuracy is available when analyzing prototype data as when analyzing model data for which measurements are generally made after a long period of constant flow. Spurr (1985) insists that "prototype scour data seldom relate to the equilibrium scour condition, either because of insufficiently long spill durations or due to a lack of progressive surveys from which to recognize equilibrium condition." The successive plunge pool surveys performed at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelsotke Dam allow a certain assessment of the time progression of scour at each site. There is a difficulty in assessing the rate of plunge pool scour at Peace Canyon Dam because the two distinct scour episodes occurred following an extended spill with sustained high discharges. The major plunge pool development resulted from initial spillway operation which extended from October 1979 to April 1980. No indication regarding the rate of formation of the scour hole as observed on April 1980 is available, although Bays 3 and 4 were used more extensively over a two-month period. The following spills of 1981, 1983, 1984, and 1987 were all smaller in magnitude and duration, and plunge pool scour was not observed to progress until 1996 following the largest spill to occur on site. During the 1996 spill event, high flows were relatively constant for about a month and scour cannot be associated with a shorter time interval. The time progression of plunge pool scour at Peace Canyon Dam remains uncertain. The progression of scour downstream of the spillway right chute at Seven Mile Dam was observed in the early years of spillway operation (1979-1982) after which the scour hole appeared to have stabilized. The core of the scour hole was formed within six weeks of initial operation (5 Nov. to 14 Dec. 1979) of which only the first two were effective since spillway discharges were then cut back. Plunge pool scour progressed further following the 1982 spill event which exceeded the magnitude of previous spills on a few days only. In 1982, the maximum average daily spill through the right chute reached approximately 65,000 cfs on two consecutive days. In 1997, when discharges of this magnitude were released through the right chute for more than two weeks (3 0 May to 14 June) , the scour hole remained stable. These observations support the concept of equilibrium conditions and suggest that the maximum scour depth for a given discharge can be attained within days. The time progression of scour downstream of the left chute is unknown because thirteen years of spillway operation separate the two reference surveys performed in September 1984 and October 1997. At Portage Mountain Project, a scour hole more than 100 ft deep has formed following a single spill event in the early years of spillway operation. The spill event lasted from 13 June to 3 September 1972 but daily average spillway discharges greater than 100,000 cfs occurred on five days only (non-consecutive). High flows spillway tests were performed on one day during the 1972 spill and a peak discharge of 175,000 cfs was held for eleven hours. During the 1996 spring flood, the historical peak discharge (175,000 cfs) was not exceeded but spillway flows were more constant and daily average discharges were maintained above 100,000 cfs for thirty days. The 1996 scour hole configuration was essentially the same as in 1972, with the exception that the scour hole invert had progressed towards the left bank and scour depth was roughly 4 ft deeper. This suggests that even a massive scour hole can form within a single spill event and that scour can approach equilibrium after a few daily peak flows. Plunge pool scour at Revelstoke Dam was seen to progress gradually throughout the years of spillway operation. The first and only prolonged spill at Revelstoke occurred during reservoir filling (October 1983 to May 1984) with spillway discharges varying from a few hundred cfs to about 35,000 cfs. The line survey performed on December 1983 is consistent with the soundings taken on 14 May 1984 indicating no significant change in topography along the centreline of the plunge pool after December 1983, although spilling lasted until May 1984. The next survey which showed a plunge pool invert about 10 ft deeper and shifted to the left was performed after the 1986 spillway tests. Such testing involved a maximum spillway discharge of 70,000 cfs held for about 15 minutes, one spill up to 60, 000 cfs within an hour, and two spills up to 50,000 cfs within an hour, all this over a four-day period. During testing, erosion was noted by the colour of the water which became muddy at the increase of discharge. At the end of the second test up to 50,000 cfs, it was noted that "the river downstream did not appear to be very muddy, indicating that the scouring of river banks and riverbed in the vicinity of the plunge pool might be nearing equilibrium for this discharge" (BCH Report No. H1907, 986). Finally, the collapse of the right bank into the plunge pool in 1991 occurred within eight hours of spilling at 59,000 cfs. The Revelstoke plunge pool was seen to respond rapidly to increases in spillway discharges but equilibrium conditions were also observed when spill was constant. From the available information, it appears that plunge pool scour can progress rather quickly and that equilibrium conditions can be reached. The concept of ultimate or equilibrium scour depth is well represented by the Seven Mile plunge pool. Experience at Seven Mile Dam and Portage Mountain Project also suggests that this scour limit can be attained within a few days. Plunge pool scour can progress even more quickly (within hours) as observed following the 1986 spillway tests at Revelstoke dam. 4.5 TAILRACE IMPROVEMENTS Loose material removed from the scour hole may accumulate at the downstream margin of the scour hole and form a bar deposit. Accumulation of debris in the tailrace area following important spills was observed at all four sites of study, especially in the early stages of plunge pool development. The tailrace bar can raise the tailwater level and hence limit the depth of scour for a given spillway discharge. Once the debris are removed and the riverbed restored to its original elevation, scour may progress further without increase in spillway flows. A summary of observations on plunge pool development relating to tailrace improvements at each dam site is presented: • At Peace Canyon Dam, the accumulation of material at the downstream end of the plunge pool following the 1979/80 spill increased the tailwater level by approximately 4 ft. Tailrace dredging was performed in the summer of 1980 to restore the riverbed to its original elevation. In the following years, further scour of the plunge pool was not apparent until spillway discharges were greatly increased to pass the 1996 spring freshet. After the 1996 spill event, the tailwater level was raised by about 1.3 ft. • At Seven Mile Dam, dredging of the tailrace area was performed for the first time during the fall of 1980, after the first important spring flood. A subsequent progression of scour in the plunge pool ,was observed from the 1982 soundings, but spills had also increased in magnitude and duration. Tailrace improvements were undertaken for a second time in the summer of 1989. The scour hole has remained stable since, even after the largest spill ever to occur on site (in 1997) . • At Revelstoke Dam, the sounding survey of 1986 showed a mound of material in the middle of the tailrace channel downstream of the plunge pool. The tailrace channel was excavated in 1989 and the tailwater level was lowered by approximately 5.6 ft. Unfortunately, plunge pool scour following the 1991 spill event could not be assessed accurately due to the important thickness of overburden caused by the collapse of the right bank into the pool. • At Portage Mountain Project, the accumulation of scoured debris is confined between the center and the left bank of the river channel some 500 ft downstream of the centre of the scour hole. Tailrace improvements were not performed to date. Although the accumulation of scoured debris downstream of the plunge pool caused an increase in tailwater level at all sites, no further scour progression in direct response to tailrace improvements was perceived. 1 j | PEACE I CANYON } | SEVEN MILE | PORTAGE j MOUNTAIN ! 1 |REVELSTOKE 1 j Bays I 1-2 Bays 3-6 1 Bays | 1-2 | Bays | 3-5 | Spillway Characteristics | Total spillway length L [ft] | 240 I 220 j 330 | 250 i | 2400 i | 1300 Clear span at crest ! w0 [ft] i 100 i | 200 1 | 100 s | 150 s ! I 150 i i | 90 1 Clear span at bucket lip 1 W£ [ft] 1 116 j 234 j 116 1 178 | 135 | 150 ] Differential head between | MNRL1 and bucket lip I hQ [ft] 138 127 j 174 164 315 i | 260 1 Total head between MNRL1 j and tailwater level I H0 [ft] [ 132' i I 2 07 550 i j 421 I Bucket lip angle 1 e [ ° ] 1 3 0 | 20 | 30 r [ 30 I 1 | 0 I Jet impingement angle^ | 9' [°] ! 3 i I 23 i 1 3 7 | 39 | 49 \ ! 38 f I Notes. | 1. MNRL: Maximum Normal Reservoir Level i 2. Theoretical value; no allowance for energy losses on spillway face, flow aeration, and j air retardation. j 3. The bucket structure is submerged at maximum normal tailwater level (four units rated 1 discharge). SEVEN MILE POWERHOUSE SPILLWAY i i BAY 1!BAY 2 PEACE CANYON SPILLWAY BAY 6i BAY 5! BAY 4! BAY 3 BAY 2iBAY 1 BAY 3 BAY 4 BAY 5 Figure 4.2. The Latest Plunge Pool Scour Configuration at Portage Mountain Project and Revelstoke Dam PORTAGE MOUNTAIN SPILLWAY n Approximate Water Line ~ REVELSTOKE DAM 200,000 150,000 I Q) O) co xz o c/> CD Q 100,000 50,000 200,000 150,000 •6 <u O) o <2 b 'S Q 100,000 50,000 PEACE CANYON DAM 1979/80 spill: 172 days 1981 spill: 12 days 1983 spill: 15 days 1984 spill: 6 days 1996 spill: 55 days Note. Data from  28 Oct. 1979 to 4 Aug. 1996 —i—i—i—r 100 200 300 400 500 Number of days discharge is equalled or exceeded MOUNTAIN PROJECT 1972 spill 1974 spill 1976 spill 1981 spill 1983 spill 1984 spill 1996 spill 85 days 36 days 36 days 13 days — 68 days 25 days 56 days Note. Data from  13 June 1972 to 4 Aug. 1996 —I—I—I—r 100 200 300 400 Number of days discharge is equalled or exceeded 500 200,000 150,000 •e CD o> o (0 b >. to Q •8 CD O) O <2 b TO • 100,000 50,000 200,000 150,000 100,000 50,000 -. Maximum " Average = 116,000 cfs = 112,000 cfs -1—i—i—r SEVEN MI LE DAM The spillway has operated almost every year. 1980 spill: 188 days 1981 spill: 124 days 1982 spill: 141 days 1996 spill: 175 days 1997 spill: 128 days Note. Data frcim  5 Nov. 1979 to 21 O ± 1997 —i—i—r 100 200 300 400 500 Number of days discharge is equalled or exceeded Maximum • Average = 70,000 cfs : 35,500 cfs REVELSTOKE DAM 1983/84 spill: ~200 days 1985 spill: 6 days 1986 spill: 8 days 1990 spill: 5 days 1991 spill: 12 days Note. Data frfam  1 Jan. 1984 to 22 S;pt. 1991 100 200 300 400 Number of days discharge is equalled or exceeded 500 Figure 4.3. Spillway Flow Duration Curve of Each Site of Study CHAPTER V CONVENTIONAL ASSESSMENT OF PLUNGE POOL SCOUR Conventional methods for the prediction of plunge pool scour consist of physical model studies and empirical equations that target the ultimate scour depth. In the laboratory, hydraulic models that respect the Froude law scaling are used to simulate scour downstream of overflow structures. Empirical formulas are derived from experiment and/or prototype observations. Spurr (1985) addresses the limitations of conventional methods for the assessment of plunge pool scour: The use of empirical formulae, mostly developed from physical models for predicting downstream scour in the prototype, has traditionally been proven unreliable. Later attempts to include both rock and hydraulic characteristics using hydraulic models with weakly cohesive bed materials have met with partial success, although the difficulties in calibrating the model bed materials to the prototype rock characteristics still remain. (Spurr, 1985) In this chapter, the reliability of hydraulic model studies and empirical power formulas in the assessment of plunge pool scour is questioned. First, the predicted plunge pool performance from small-scale model testing of Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam is compared with observed plunge pool development on site. Then, the accuracy of a number of empirical formulas is examined when applied to past scour conditions identified at the four dam sites. 5.1 HYDRAULIC MODEL STUDIES Hydraulic model studies were conducted for each site of study and plunge pool scour tests were part of the investigation. The results of model studies for Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam are summarized and modelled scour patterns are compared with observed plunge pool scour on site. The aim of this section is not to identify the flaws of each model study, but to highlight that great care must be taken in order to achieve reasonable results in the prediction of plunge pool scour. 5.1.1 Peace Canyon Dam Small-scale model studies were conducted for the Peace Canyon Project which included downstream scour investigation. The model was built to a 1:60 linear scale, respecting the Froude similitude criteria (BCH Report No. H715, 1970) . However, the Project design was not final and only part of the spillway was built in the laboratory. The first spillway bays next to the powerhouse were reproduced to simulate the Inflow Design Flood (IDF) (363,000 cfs at the time) considering a seven-bay, six-bay, and five-bay spillway arrangement and using three different bucket shapes. Crest elevation on the model was 2 ft higher (El. 1614 ft) than the as-built prototype structure. Scour conditions were tested on both cohesive and non-cohesive erodible beds representing prototype materials between 2 in. and 5 ft in diameter. For each assessment, the non-cohesive bed was initially used to model the equilibrium scour depth. Then testing was carried out to develop semi-cohesive stratification of the material, such that a similar scour depth was obtained but with more realistic slopes. One series of tests conducted on the small-scale model approaches prototype scour conditions of the 1996 spill event. In 1996, Bay 3 was operated at full capacity for over a month with the reservoir fluctuating around El. 1649 ft for an average discharge of 47,000 cfs (Section 3.1.4) . This is similar to the simulated IDF through the seven-bay spillway arrangement (51,900 cfs per bay) at reservoir El. 1653 ft with a 20° bucket lip at El. 1517.8 ft (5 ft lower than as-built). Figure 5.1 illustrates the scour profiles downstream of Bay 3 centreline as surveyed after the 1996 spill and as modelled in laboratory. The major differences between prototype and hydraulic model conditions are also listed. The plunge pool bottom elevation as surveyed in 1996 (El. 1459 ft) was about 30 ft shallower than what was predicted from the model studies (El. 1430 ft) and the scour hole invert was located some 12 0 ft closer to the bucket structure on the prototype. Upstream and downstream slopes of the scour hole on both model and prototype are similar. The model prediction was conservative with respect to scour depth, but not in terms of upstream progression of scour. 5.1.2 Seven Mile Dam Comprehensive hydraulic model studies of the Seven Mile Project were conducted on a 1:100 scale model (BCH Report No. N68, 1976). Scour downstream of the as-built spillway arrangement was investigated using non-cohesive and cohesive bed materials. The erodible riverbed was first modelled with loose 0.75 in. crushed gravel and then using a weak sand-cement-gravel mixture. The gravel bed scour tests provided information on general scour patterns and potential areas of undermining. The sand-cement-gravel bed helped to better define the areal extent of scour and boundary conditions. Two gate settings were tested on each bed and one run of sequential flows up to 370,000 cfs was performed on the non-cohesive gravel bed. The scour testing program first involved spilling at Maximum Normal Reservoir Level (El. 1730 ft) with all gates open at 13.3 ft for an approximate discharge of 24,000 cfs per bay. To date, Bays 1 to 3 on the prototype have all been used to release such discharge (and up) for a prolonged period of time. During the 1997 spring flood, Bay 3 released a steady discharge of approximately 30,000 cfs for a month while Bays 1 and 2 spilled daily average flows between 30,000 and 36,000 cfs (Section 3.2.4). Plunge pool soundings of October 1997 showed a depression with El. 1503-1509 ft over a 1200 sq.ft area downstream of Bay 3 and a scour hole with invert at El. 1450 ft downstream of Bays 1 and 2 (Figure 3.19). The predicted scour patterns on the 1:100 scale non-cohesive and cohesive models corresponding to a total spill of 120,000 cfs (24,000 cfs per bay) are presented in Figures 5.2 and 5.3. As expected, model scour is more confined in the sand-cement-gravel mixture, but the scour depth is about 100 ft deeper than simulated on the loose gravel bed. Neither model predicted the formation of a major scour hole downstream of the right chute (Bays 1 and 2) as seen on the prototype. On the non-cohesive model, the scour processes were likely hindered by the accumulation of deposits whereas the long downstream leg observed on the cohesive model was caused by scour surfacing through the bed as it sloped down towards the tailrace excavation. As for prototype scour downstream of Bay 3, it corresponded in depth to the bottom elevation modelled on the gravel bed (El. 1501-1509 ft) and the observed location of maximum scour was consistent with the cohesive model prediction. Other simulations at higher flows resulted in modelled plunge pool elevations above the existing topography on site. Overall, the model scour patterns were not considered representative of prototype scour. 5.1.3 Portage Mountain Project Plunge pool scour at Portage Mountain Project was investigated using a 1:96 scale model (Johnson and Alam, 1969). The "spoon-like" shape of the spillway flip bucket was designed from physical modelling in a way to minimize scour depth by enhancing energy dissipation in the atmosphere. The outfall model included the steep portion of the discharge chute (starting about 1900 ft downstream of the headworks), the flip bucket, and the river/tailrace area. Model scour tests were carried out using gravel having a narrow range of grain size distribution. Plunge pool scour patterns were investigated for spillway discharges of 260,000 cfs and 390,000 cfs. Figure 5.4 illustrates the two scour configurations obtained from model testing. The scour hole invert was located near the center of the channel in both cases and reached El. 1580 ft and El. 1570 ft, respectively, in the lower and higher spillway flows simulations. No reasoning for the small change in scour depth with large discharge increase was given in the referenced report (BCH Report No. H1756, 1988) . The side slopes of the plunge pool were measured on the model to be approximately 2.4H:1V (ibid.) . The existing plunge pool at Portage Mountain Project was scoured down to El. 1520 ft (Figure 3.26) following a single spill event (1972) during which a maximum discharge of 175,000 cfs was released through the spillway (Section 3.3.4). Hence, prototype scour depth exceeded model study expectations by 5 0 ft for a maximum spillway discharge that was less than half that tested on the model. As such, the side slopes are steeper on the prototype scour hole than on the model. 5.1.4 Revelstoke Dam Extensive hydraulic model studies were conducted to assess plunge pool development at Revelstoke Dam. As part of the design studies, a model that comprised the entire spillway and a 3300 ft reach of the Columbia River with provision for powerhouse discharges was built to a 1:100 scale (BCH Report No. 2278, 1983). Scour tests were performed on the final design of the excavated plunge pool using two different types of bed. For the first test series, the bedrock directly impacted by the spillway jet (plunge pool floor and apron) was modelled in a sand, pea gravel, and bentonite mixture while the bedrock outside this area was modelled using a scour resistant mixture of sand and cement. In the second test series, the bedrock strata were simulated by placing 1.25 to 1.75 in. (10 to 15 ft on prototype) gravel rocks flat side down in layers held together with bentonite clay and oriented to the prescribed foliations of the rock. Overburden was represented by fine sand. Testing in both test series involved progressive flows up to failure of the right bank armour protection. The maximum scour depth for a given discharge was generally greater in the second test series but the difference lay in the thickness of one block (15 ft on prototype) . Figure 5.5 illustrates the model scour configuration for spillway discharges of 40,000 cfs and 60,000 cfs. In the first case, the simulation involved increasing spillway flows in steps, from 0 to 40,000 cfs, using the low level gates only. This test anticipated the expected initial operating conditions during the period of filling the prototype reservoir (BCH Report No. 2278, 1983). The scour patterns in the layered-gravel bed showed a plunge pool floor down to El. 1380 ft some 170-190 ft downstream of the concrete apron along the spillway centreline. On the prototype, the plunge pool had reached El. 13 92-13 93 ft over an extended area further downstream and to the right than predicted by the model (Appendix II, BCH Drawing No. 212-C21-D7125) . With an increase in spillway discharge to 60,000 cfs, the modelled scour hole was 20 ft deeper (min. El. 1360 ft) in both the layered-gravel bed and the pea gravel-bentonite mixture. In the layered-gravel bed, the increase in simulated flow caused a progression of scour downstream and towards the right bank. Scour in the pea gravel-bentonite mixture was limited downstream by a bar of deposited material. Spillway flows of 60,000 cfs were tested in 1986 on the prototype for about half an hour. Plunge pool soundings following the spillway tests showed a scour hole with minimum El. 1380 ft facing the left half of the spillway at a distance of 300 ft from the concrete apron (Appendix II, BCH Drawing No. 212-C21-D134) . Had the discharge of 60,000 cfs been held for longer, the observed scour depth and extent on the prototype could have been greater. Overall, the model scour predictions were conservative compared to observed scour on site. The scour patterns modelled in the layered-gravel bed were more representative of prototype scour configuration. In both test series, launching of the right bank apron started at 40,000 cfs and continued in an orderly fashion for the remainder of the tests. A weak return eddy formed along the right bank at a spillway discharge of 75,000 cfs and increased in strength at each subsequent increase in spillway discharge (BCH Report No. 2278, 1983). This return eddy was the main factor in the attack on the right bank for spillway flows greater than 90,000 to 95,000 cfs and lead to the failure of the right bank at flows greater than 105,000 cfs (ibid.) . Launching of the riprap was expected to be slower and more organized on the prototype so that failure of the right bank armour would occur at higher flows than exhibited in the model studies. During the flood of August 10-11 [1991], the right bank protection work failed resulting in a section of the powerhouse access road collapsed into the plunge pool (BCH Report No. HYD.943, 1991). Spillway flows were increased from 16,000 cfs to 59,000 cfs within a two-hour period, and failure occurred within eight hours of spilling at 59,000 cfs. The model was valuable in predicting the general flow patterns in the plunge pool and subsequent failure of the right bank, but somehow the erosion processes were underestimated. 5.2 SCOUR DEPTH EMPIRICAL FORMULAS Case studies show wide variances in the accuracy of the predicted and actual depths of scour (Wittier et al., Dam Foundation  Erosion, 1995). The reasons for the lack of accuracy are model specific formulas, site specific application, fragmented results from multiple studies, and the factors of geology and cohesive material properties (ibid.). The objective of this section is to present some of the most common empirical formulas for the assessment of maximum scour depth and evaluate their accuracy in predicting past scour development at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam. First, the empirical formulas selected for this study are listed with brief comments of their derivation. Then, the accuracy of each equation is evaluated through a back analysis of plunge pool development at the four sites of study. 5.2.1 Descriptive List of Equations The list of scour depth formulas found in the literature that were developed empirically from laboratory experiments and/or prototype observations is exhaustive. A comprehensive review of 31 expressions was presented by Mason and Arumugam (1985) . Ten equations were selected for this study, based on the conditions of their derivation and published performance on case studies. Preferences were made to well-documented equations derived from prototype scour observations. The scour depth formulas are listed in Table 5.1 and brief comments on each of them follow: 1. Veronese (1937) (reported by Hager, 1998) performed experiments on scour processes from a vertically falling jet in a 20 in. wide flume. For grain sizes smaller than 0.2 in., the scour depth was found to be smaller than expected given the original trend observed for larger particles (Whittaker and Schleiss, 1984). Veronese proposed an equation independent of the bed material size, which is endorsed by the U.S. Bureau of Reclamation (1977) as a limiting scour depth. 2. Based on the work of Veronese, Jaeger (1939) (reported by Hager, 1998) suggested a different scour-depth equation in which the Froude similarity law was satisfied. Jaeger also considered the addition of tailwater depth as a significant factor. 3. Damle (1966) (reported by Mason and Arumugam, 1985; Breusers and Raudkivi, 1991) considered both model data and field observations from dams in India with ski jumps and presented three expressions: a mean value for model scour, a mean value for prototype scour, and a maximum worst case for both. The prototype version is shown in Table 5.1 and used in this study. 4. The Chian Min Wu (1973) expression is similar to the limiting equation of Veronese and specific to ski-jump spillways (Whittaker and Schleiss, 1984). Chian Min Wu (1973) (reported by Breusers and Raudkivi, 1991) used model and prototype data from dams in Taiwan. 5. Martins' (1975) (reported by Breusers and Raudkivi, 1991) formula presented in Table 5.1 was derived from prototype observations specific to ski-jump spillways. One particularity of the Martins equation is the preferential use of head drop between reservoir level and bucket lip elevation (Z) instead of total head between reservoir and tailwater levels (H). 6. Taraimovich (1978) based his work on detailed data sets of prototype and model scour at hydrodevelopments in narrow canyons (arch dams) and with a wide lower pool in the USSR. The original equation proposed by Taraimovich includes many factors such as bucket flip angle, upstream slope of the scour hole, rock strength, settling velocity, and velocity coefficients, but which approach unity as a whole. Therefore, a simplified version by Mason and Arumugam (1985) is presented in Table 5.1 and used in this study. 7. Mason (1984) and Mason and Arumugam (1985) presented two equations for the prediction of ultimate scour depth by a plunging jet based on the analysis of 26 sets of scour data from prototypes and 47 from models. The two equations, which are dimensionally balanced and respect the Froude scaling law, are generally considered state of the art. The Mason A formula (7A) was developed as a best fit for all of the model data and the Mason B formula was derived using both model and prototype observations (7B) . The model data best fit is now considered to be an acceptable upper bound for prototype scour. 8. Wang Shixia's (1987) formula was developed from a regression analysis of 50 data sets of scour from prototypes in China with trajectory bucket type energy dissipators. The equation is dimensionally balanced and incorporates a coefficient for the rock condition. 9. Yildiz and Uziicek (1994) proposed a modified version of the Veronese equation to account for the different behaviour of free-trajectory jets as opposed to vertically falling jets. When discharge is issued from spillway flip buckets, it is suggested that the effective scour depth estimated by the Veronese formula be measured along the tangent to the jet entering the tailwater. In all formulas the ultimate depth of scour is measured from the tailwater surface. The central parameters of the empirical power formulas are essentially common between authors and the difference between equations lies in the choice of exponents. 5.2.2 Performance of Empirical Formulas In this section, the scour depth empirical formulas listed in Table 5.1 are applied to the four sites of study. From the review of plunge pool performance at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam (Chapter III), specific spill events were linked to different phases of plunge pool development. The hydraulic conditions associated with the observed scour depths are listed in Table 5.2 for a total of 14 scour data sets. Each equation of Table 5.1 was used to process the assembled data sets and the main findings are discussed. The accuracy of the selected empirical formulas can be assessed by comparing calculated scour depths with observed plunge pool depths on prototypes. Figures 5.6 to 5.15 illustrate the performance of each equation in the predicted versus actual plunge pool floor elevation for each of the 14 scour data sets. Clearly, the Veronese (Figure 5.6) and Mason B (Figure 5.13) empirical equations are too conservative, while the formulas by Damle (Figure 5.8) and Wang Shixia' (Figure 5.14) underestimate the scour depth in 50% of the cases. The Mason B formula is particularly unsuitable to Portage Mountain and Revelstoke plunge pools with calculated elevations such as El. 1054 ft instead of observed El. 1520 ft and El. 1178 ft instead of El. 1380 ft (not shown in Figure 5.13). The equation by Mason which is characterized by dependent exponentials (Table 5.1) is in fact increasingly sensitive to larger head between reservoir level and tailwater surface. The five empirical equations specific to ski-jump spillways - Damle (Figure 5.8), Chian Min Wu (Figure 5.9), Martins (Figure 5.10), Wang Shixia (Figure 5.14), and Yildiz & Uziicek (Figure 5.15) - give the best predictions. The most accurate expressions are the ones by Damle (Figure 5.8) and Chian Min Wu (Figure 5.9) with a standard error of estimate of 16.4 ft and 18.0 ft., respectively. Interestingly, the Damle equation gives equal weight to both unit discharge q and total head H (Table 5.1). Scour downstream of the spillway left chute at Seven Mile Dam was overpredicted by all formulas, whereas most equations underestimated the plunge pool depth at Revelstoke Dam. The results of the back analysis using empirical formulas are combined in Figure 5.16 in terms of scour depth below tailwater level. Differences of 60-80 ft between computed depths for a unique set of hydraulic conditions are common. The accuracy of each formula was assessed by the ratio of calculated scour depth over observed depth, or relative error x e, and the resulting coefficient of variation V  •. ^calculated  I  ^ observed V  = Sjx e where D is the maximum scour depth below tailwater level, S is the standard deviation of the relative error, and x e is the mean relative error. The results of the statistical analysis compared with data from a previous study by Mason and Arumugam (1985) are presented in Table 5.3. Based on consistency, the best forms of equation are those with the lowest coefficients of variation. From the present study, the Jaeger and Mason A empirical formulas are identified as such. Both are expressions which include the tailwater depth. However, the equations are rather conservative with a mean ratio of calculated versus observed scour depth around 1.5. The equations for which the estimates of scour depth are on average closer to the observed depths on prototypes are the Damle and Wang Shixia formulas, but with a coherence of about 60%. As noticed previously, the plunge pool scour depths predicted by the Veronese and Mason B equations are well beyond the observed ones, by more than a factor of two on average. The Mason B formula was nonetheless the most accurate expression to fit 26 sets of scour data from prototypes and 47 from models based on a thorough analysis by Mason and Arumugam (1985) . From the same study, the best form of equations identified for prototype scour besides the one proposed by the authors was the one by Damle, with a coefficient of variation of 33%. In conclusion, the reliability of empirical equations as a means of predicting plunge pool scour depth is doubtful. Although all equations provide on average conservative values for design, the variability in results from a single data set and inconsistency of a given formula from one site to another are problematic. The Veronese limiting equation should not be used as suggested by the U.S. Bureau of Reclamation and the version B of the equations by Mason should also be rejected. The Damle empirical formula was seen to give the best combination of precision and accuracy in the prediction of plunge pool scour depth at the four sites of study despite a tendency for underestimation. The Wang Shixia formula could be calibrated to each site by changing the bedrock condition parameter and used as a predictive tool for future scour. The Mason A equation should be considered as a limiting scour depth formula to assess the worst-case scenario. Table 5.1. List of Empirical Formulas for Ultimate Scour Depth Prediction 1 ! Veronese - 1937 i (U.S.B.R.) Ds + h = 1.32 q0'54 H°'225 2 Jaeger - 1939 Ds + h = 0.446 g°-5°H0-25 (h/dj 0 3 3 3 3 Damle - 1966 (prototype data) Ds + h = 0.20 q0'50 H0'50 4 Chian Min Wu - 1973 i Ds + h = 0.872 q0'51 H 0 2 3 5 5 | Martins - 1975 Ds + h = 1.05 q0'60 Z°-10 6 Taraimovich - 1978 (simplified form by Mason & Arumugam, 1985) Ds + h = 0.314 q0 6 7 H0'25 7A Mason A - 1984 (model data) Ds + h = 3.27 q0'60 H0-05 h0'15 g0.30 dm0.10 7B j Mason B - 1984 (prototype & model data) Ds + h = K [0 . 0929q"|x |"0 . 3048H"|y h 0 1 5 g0.30 dm0.10 K = 28.35 - 12 .14H0'10 x = 0.60 - H/984 y = 0.15 + H/656 8 Wang Shixia - 1987 Ds + h = 2.44 Kr [(q2/g)0'33]0-89 H 0 1 1 9 Yildiz and Uziicek - 1994 (Modified Veronese) Ds + h = 1.32 q0'54 H0'225 sin0' Jsin2 8 + (H-Z)/Z tan 0'= V " cos Q Note. Units are imperial. Nomenclature dm [ft] = Median particle diameter/Mean block size d90 [ft] = Size of bed material of which 90% by weight is smaller Ds [ft] = Maximum depth of scour below original bed level g [ft/s2] = Gravitational acceleration h [ft] = Tailwater depth H [ft] = Head difference between reservoir level and tailwater surface Kr = Coefficient reflecting the capacity of rock strata to resist scour (Wang Shixia, 1987) = 0.70 - 1.10 (avg. 0.90) for solid rock = 1.10 - 1.40 (avg. 1.25) for medium rock = 1.40 - 1.80 (avg. 1.60) for soft or broken rock q [cfs/ft] = Spillway unit discharge (discharge per unit width) 9 [°] = Angle of flip bucket exit 9' [°] = Impact angle of the jet to the horizontal with the tailwater surface Z [ft] = Head difference between reservoir level and bucket lip elevation Scour Data Set1 Survey Date Scour Hole Invert El. [ft] Maximum Scour Depth2 [ft] 1 Effective Spill Daily Average Discharge [cfs] Width of Flow [ft] Unit Discharge, q [cfs/ft] Reservoir El. [ft] Tailwater El. [ft] Total Head, H [ft] jTailwater | Depth, h j [ft] PCN80-3/4 Apr. 1980 1458 50 1979/80 55,000 118 (Bays 3-4) 470 1639 1519 120 j 14 1 PCN80-5/6 Apr. 1980 1482 30 1979/80 36,000 116 (Bays 5-6) 310 1639 1519 120 | 14 PCN96-3 Aug. 1996 I 1459 50 1996 47,000 58 (Bay 3) 810 1649 1519 130 14 PCN96-5/6 Aug. 1996 1481 30 | 1996 42,000 116 (Bay 5-6) 360 1649 1519 130 14 SEV79-1/2 Dec. 1979 i 1465 i 1 35-40 1979 29,000 116 (Bays 1-2) 250 1710 1516 195 11 SEV80-1/2 | Aug. I 1980 j 1462 33-38 1980 59,000 116 (Bays 1-2) 510 1714 1527 189 17 SEV82-1/2 1 Aug. ! 1982 \ 1450 I 1 60-65 j 1982 66,000 116 (Bays 1-2) 570 1713 1528 185 23 SEV97-1/2 I Oct. 1 1997 1 1450 I | 55-60 I 1997 i 72,000 116 (Bays 1-2) 620 1728 | 1535 192 30 | Oct. j 1997 1503 30 1997 31,000 60 (Bay 3) 172 8 1535 192 SEV97-4 | Oct. j 1997 1523 10 1997 9, 000 903 (Bay 4) 100 1728 1535 192 ) PMD72 j May 1 1973 1520 115 1 1972 1 167,000 135 (Chute) 1240 2206 1651 555 17 PMD96 | Aug. i 1996 | 1516 118 i 1996 122,000 135 (Chute) 900 2190 j 1660 530 _ REV84 j May [ 1984 i 1392 i 58 1983/84 29, 0004 150 (Chute) 190 1837 1448 389 | 8 REV86 I Aug. i 1986 : 1380 70 j 1986 50,000s 150 (Chute) | 330 i 1879 1460 419 p r > -j ! Notes. i 1. Labeled after the site abbreviation followed by the year of effective spill and the spillway bays downstream of which scouring occurred. 2. Maximum difference in elevation with respect to the original plunge pool topography. 3. Includes flow divergence in Bay 5. 4. Slightly lower than the maximum daily average discharge recorded (30,500 cfs), but the total head was greater by about 14 ft. 5. Observations near the end of the 1986 spillway tests showed that no further erosion would be expected for spillway discharges up to 50,000 cfs (BCH Report No. H1907, 1986). Table 5.3. Statistical Analysis of Scour Depth Empirical Formulas Performance PRESENT STUDY MASON & ARUMUGAM*  (1985) No. EMPIRICAL FORMULAS 4 DAM SITES (14 scour data sets) PROTOTYPES (26 scour data sets) | Maximum error1 Minimum error1 Mean error1 Coefficient of | variation2 [%] Mean error1 Coefficient of variation2 [%] ! j 1 I Veronese - 1937 | 4 .32 1.48 2 .19 | 40.8 1.4171 41. 63 j. 2 | Jaeger - 1939 j~ 1. 94 1. 05 1.52 17 . 7 j~ 1.3973 39.00 [ 3 i 1 Damle - 1966 i i j 2.32 0 . 78 1.14 39 . 7 | 0.7064 33 .16 1 4 i i | Chian Min Wu - 1973 i | 2 . 62 0 . 85 1.28 42 .1 5 Martins - 1975 | 2.33 0 . 75 1.26 41. 5 0 . 8538 47 . 19 6 Taraimovich - 197 8 j 2.42 f 0 . 85 1.30 36 . 3 0.8644 44 . 13 | 7 A | Mason A - 1984 2 . 07 0 . 91 1.44 26 . 5 1. 003 25 .43 j 7B 1 | Mason B - 1984 i 4 .56 1. 54 2 . 54 40 . 6 1. 07 30.1 1 8 jWangShixia - 1987 1.78 0 . 72 1.14 37.4 | 1 9 1 Yildiz & Uziicek - 1994 | 2 . 62 | 0.67 1.23 46 . 9 f 1 Notes. i 1. Relative error (jt ) = calculated scour depth (Dcalculaled) / measured scour depth (Dobservetl) j 2. Coefficient of variation (K) = standard deviation (5 ) / mean error (jf ) [%] ! 3. Mean error and coefficient of variation obtained using the model data only (47 scour data sets). i •Source: P.J. Mason and K. Arumugam, "Free Jet Scour Below Dams and Flip Buckets", Journal  of Hydraulic  Engineering,  ASCE, Vol. Ill (Feb. 1985): 220-235, Table 3. PEACE CANYON DAM COMPARISON OF PROTOTYPE SCOUR AND MODEL SCOUR 1600 - | 1550 -Profile Along Spillway Bay 3 Centreline Major Differences  Between Prototype and Hydraulic Model Conditions MODEL PROTOTYPE Crest El. [ft] 1614 1612 Reservoir El. [ft] 1653 1649 Bucket lip El. [ft] 1517.8 1522.77 Discharge per bay [cfs] 51,900 47,000 Unit discharge [cfs/ft] 885 810 Minimum bed El. [ft] 1430 1459 c o J-> ru > V UJ 1500 1450 -1400 t—|—i—i—r 1300 Prototype - October 1979 Prototype - August 1996 Hydraulic Model (Ref.  BCH Report No. H715, 1970, Figure 20) n—I—I—I—|—I—i—I—i 1600 Station N [ft] 'l 1 1 1 1400 1500 'I 1 1 1 1 1700 T~r r 1800 ~i i i i i r 1900 Figure 5.1. Peace Canyon Dam / Comparison of Prototype Scour and Model Scour SPILLWAY DISCHARGE = 120,000 CFS (24,000 CFS PER BAY) Note. Ref.  BCH Report No. N68, 1976, Figure 27. 7 MILE PROJECT PEND D'OREILLE RIVER IUOO SCALE MODEL GRAVEL SCOUR PATTERN POST 123-36-I20-I3.3 A RECOMMENDED SPILLWAY Figure 5.2. Seven Mile Dam / Model Scour Patterns in Non-Cohesive Bed SPILLWAY DISCHARGE = 120,000 CFS (24,000 CFS PER BAY) 7 MILE PROJECT PEND D'OREILLE RIVER IMOO SCALE MODEL Note. Ref.  BCH Report No. N68, 1976, Figure 32. SAND CEMENT SCOUR PATTERN POST 123 - 36 - 120 - I3.3 A RECOMMENDED SPILLWAY Figure 5.3. Seven Mile Dam / Model Scour Patterns in Cohesive Bed PORTAGE MOUNTAIN PROJECT PLUNGE POOL SCOUR FROM MODEL TESTS SPILLWAY DISCHARGE = 260,000 CFS Figure 5.4. Portage Mountain Project / Plunge Pool Scour From Model Tests fwi 4,500 y SPILLWAY DISCHARGE = 390,000 CFS Note. Ref.  BCH Drawing Nos. 1006-C14-B305/306 REVELSTOKE DAM SCOUR CONFIGURATION FROM MODEL STUDIES Note. Ref.  BCH Report No.2278, 1983, Figures 6, 14 & 16. Scour Patterns #1 Scour Patterns #2 Scour Patterns #3 LAYERED GRAVEL-BENTONITE DISCHARGES Spillway: 40,000 cfs Powerhouse: 59,200 cfs (after 1/2 hr) LAYERED GRAVEL-BENTONITE DISCHARGES Spillway: 60,000 cfs Powerhouse: 59,200 cfs PEA GRAVEL-BENTONITE MIXTURE DISCHARGES Spillway: 60,000 cfs Powerhouse: 0 cfs Figure 5.5. Revelstoke Dam / Scour Configuration from Model Studies d o 4-> fU > a) "O a» U CL E o u 1540 1520 1500 H 1480 1460 1440 1420 1400 1380 1360 1340 Notes. 1. Refer to Table 5.1 for the mathematical expression. 2. Refer to Table 5.2 for details on the scour data sets. 3. Constant value of dm: 0.82 ft 4. Equation not applicable to the data sets SEV97-3 & SEV97-4 'On • <T> o Q 00 z u IE a. Q. * * * * vO LD i rvl rv <~o Q cn z o Z CL CL rv en 21 CO I l l l | l l l l | l l l ! " T " I J ' I I I " ' I ' | I I I I | I I I I | I 1 I I j I I I I j ! 1 I I " J I I I ! | T T T T j I I I i | H " l " l | " T 1340 1360 1380 1400 1420 1440 1460 1480 Observed Elevation [ft] 111111111111111 1500 1520 1540 1560 c o U-> 03 > CD T3 <u 1540 1520 - d 1500 1480 - d 1460 - d 1440 Q. E o 1420 1400 Notes. 1. Refer to Table 5.1 for the mathematical expression. 2. Refer to Table 5.2 for details on the scour data sets. 1380 - d 1360 - d 1340 r M I I | I II I | ! I! I | I I I I | I I I I | I I I 1340 1360 1380 1400 i i i I i 11 i i i i i i i i i i 1420 1440 1460 1480 Observed Elevation [ft] I I II | I I II | I I M | I I I 1500 1520 1540 1560 1540 —d 1520 - d c o 4-J > <u T3 CU 4-> Q. E o u Notes. 1. Refer to Table 5.1 for the mathematical expression. 2. Refer to Table 5.2 for details on the scour data sets. 1500 1480 - d 1460 - d 1440 1420 1400 1380 - d 1360 - d 1340 r M i i i i II i i i i i i i i i I I i i i i i i i i i i i i i i i i i i i i i i i II i i i i i i i i M i i i i i | i i i i | i i i i | i i i i | i i i i | II i i | I I i i | i i i i | i i M | i II i | i i i i 1340 1360 1380 1400 1420 1440 1460 1480 1500 1520 1540 1560 Observed Elevation [ft] c o '•*-> > _QJ LU T3 C1J CL E o u Notes. 1. Refer to Table 5.1 for the mathematical expression. 2. Refer to Table 5.2 for details on the scour data sets. 1540 - d 1520 - d 1500 - d 1480 - d 1460 1440 1420 - d 1400 - d 1380 1360 1340 [ i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i M i[ i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i | i i i i | i i i i | i i M | I i i i j i i i i 1340 1360 1380 1400 1420 1440 1460 1480 1500 1520 1540 Observed Elevation [ft] 1560 1540 - d 1520 -3 1500 c o 4-> ro > cu TJ O) O U Notes. 1. Refer to Table 5.1 for the mathematical expression. 2. Refer to Table 5.2 for details on the scour data sets. 1480 1460 1440 1420 1400 1380 1360 1340 [ I I I ! | I I I I | I I I I | i I I I | I I I I | : I :< j I ! 1340 1360 1380 1400 1420 1440 1460 1480 1500 Observed Elevation [ft] c o '•<-> ro > cu T3 CU 4—> CL E o u Notes. 1. Refer to Table 5.1 for the mathematical expression. 2. Refer to Table 5.2 for details on the scour data sets. 3. Wang Shixia coefficient: Kr = 1.25 for PCN Kr = 0.90 for SEV Kr = 1.25 for PMD Kr = 1.10 for REV 1540 1520 -E 1500 1480 1460 1440 1420 1400 1380 1360 1340 i i i i i i M i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i M i i i i M i i i i i i M i i i i i M i i i i i | i M i | i i i i | i M i | i i M | M M | i i i i | i i i i 1340 1360 1380 1400 1420 1440 1460 1480 1500 1520 1540 1560 Observed Elevation [ft] 1540 1520 1500 - d 1480 ~d a o j - J ro > _gj LU T3 <3J J-J CL E o 1420 1460 1440 - d 1400 - d 1380 - d 1360 - d Notes. 1. Refer to Table 5.1 for the mathematical expression. 2. Refer to Table 5.2 for details on the scour data sets.' 1340 m o oo z rsi u i—1 CL 1 cn # > # to ro # VO cn tN lJ-r* CL i O CO >LU m II M i i i M i i i i i i II i i i M IIiiiiiiiiiIIiii i | i i i i i i II i I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I ! I I I 1340 1360 1380 1400 1420 1440 1460 1480 1500 1520 Observed Elevation [ft] 1540 1560 220 — 200 — 180 — 160 — 140 — CL oj 120 T3 i _ Z ! O u in TJ <u 4—» =3 CL E o u 80 100 60 40 — 2 0 — * + O X • LEGEND 1 - Veronese 2 - Jaeger 3 - Damle 4 - Chian Min Wu 5 - Martins 6 - Taraimovich 7A - Mason A 7B - Mason B 8 - Wang Shixia 9 - Yildiz & Uziicek • A Notes. 1. Scour depth below tailwater level. 2. Refer to Table 5.1 for the list of empirical equations and Table 5.2 for the scour data sets. 3. Constant value of dm: 0.82 ft 4. Wang Shixia coefficient: Kr = 1.25 for PCN Kr = 0.90 for SEV Kr = 1.25 for PMD Kr = 1.10 for REV 5. Outside values: Veronese: (131, 256) Mason B: (131, 597) Mason B: (144, 564) 20 40 60 80 100 Observed scour depth [ft] 120 140 160 CHAPTER VI THE ERODIBILITY INDEX METHOD When formed in 1993, the main objective of the Dam Foundation Erosion Study Team (U.S. Bureau of Reclamation, Golder Associates Inc., Colorado State University, and other partners) was to develop a state-of-the-art procedure for predicting the extent of plunge pool scour caused by overtopping dams. The Erodibility Index Method (or hydraulic erodibility) by Annandale (1994, 1995) was chosen as the working basis for the study. Extensive research was conducted at the Hydraulic Laboratories of Colorado State University (Fort Collins, CO) to supplement the existing knowledge of plunging jet hydraulics and validate the method on different erodible media. This chapter presents an evaluation of the Erodibility Index Method in the assessment of observed plunge pool scour at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam. First, the conceptual approach of the new technology is described. Then, the methodology is applied to the four sites of study using past scour conditions. A sensitivity analysis of key parameters is also performed to complete the evaluation of the Erodibility Index Method. 6.1 CONCEPTUAL APPROACH The Erodibility Index Method relates the relative ability of earth material to resist scour, expressed in terms of a geomechanical index known as the Erodibility Index, to the relative magnitude of the erosive power of water, expressed in terms of stream power, to define an erosion threshold for an extensive range of earth materials (Annandale, 2 002b). The Erodibility Index K  was defined in Chapter II (Section 2.2.3) as the product of four coefficients, each representing a geomechanical property of the earth material: where Ms  defines the mass strength, Kb  the block/particle size, Kd  the discontinuity/inter-particle bond shear strength, and Js  the shape and relative orientation to the flow. Kb  and Kd  can be further defined as: K  = Ms  • Kb  • Kd  • Js (6.1) Kb  = RQD/Jn Kd  = JrjJa ( 6 . 2 ) where RQD  is the Rock Quality Designation, Jn  the joint sets number, Jr the joint roughness number, and Ja  the joint alteration number. The constituent parameters of the Erodibility Index are rated with the use of standard tables (Appendix I) . The generic expression for the stream power P  or rate of energy dissipation per unit area was presented in Chapter II (Section 2.2.3) in terms of velocity V and energy loss AE such as: P  = yQlsEj  A = yvAE  [KW/m 2]  (6.4) where y is the unit weight of water, Q the total discharge, and A the area perpendicular to the flow. Specific equations can be derived to compute the available stream power in various flow conditions. For flow in a plunge pool, the power available P A to scour is a function of the jet hydraulics (see Section 6.2.2). The erodibility threshold is a rational correlation established by relating the Erodibility Index and the calculated stream power for 150 field observations and published data on initiation of sediment motion. From Annandale (1995) , the power required P R to initiate scour of bedrock characterized by an Erodibility Index K is given by: P r=K 015 (6.5) The graphical threshold relationship for rock and other complex earth materials was presented in Chapter II (Figure 2.4). The essence of the method that was developed to calculate the scour depth by using the Erodibility Index Method entails a comparison between available stream power [PA] and stream power that is required to initiate scour (Wittier et al. , New  Technology,  1998). Hydraulic erodibility occurs when the erosive power of water exceeds the power required to scour the riverbed (P A > P R) . Figure 6.1 illustrates the conceptual approach of the Erodibility Index Method in the assessment of maximum scour depth in a plunge pool. First, the plunge pool bedrock profile is characterized using the Erodibility Index (Section 6.2.1). The power required to scour the bedrock material to a certain depth is determined from the erodibility threshold relationship (equation 6.5). Then, the rate of energy dissipation in the plunge pool (power available for scour) is computed along the centreline of the submerged jet (Section 6.2.2) . The rate of energy dissipation, or available power, is a discretized function of the total head at various elevations along the centerline of the submerged jet (Wittier et al., New  Technology,  1998). Bed material strength normally increases as a function of elevation below the surface, whereas the erosive power of a jet decreases in the same direction (Annandale et al., 1997) . The estimated elevation of maximum scour depth is determined from the intersection of the two stream power curves (Section 6.2.3) . 6.2 PRACTICAL APPLICATIONS The Erodibility Index Method is applied to estimate plunge pool scour at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam. The conceptual approach described in the previous section is resumed in details and in a methodological fashion with the geomechanical characterization of the plunge pool bedrock at first, the spillway jet hydraulics computations in second, and finally the combination of the two resulting stream power curves to determine the maximum scour depth elevation. 6.2.1 Erodibility Index Characterization Erodibility Index characterization of plunge pool bedrock at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam is presented in Tables 6.1 to 6.4. The assessment of geomechanical properties was based on laboratory rock testing results, drill logs, and general observations from geotechnical investigations. Rating of geomechanical parameters was based on standard tables presented in Appendix I. To illustrate the existing range of rock mass parameters at each site, the bedrock classification was divided in three categories: low, typical, and high\ The "typical" classification is considered to be representative of the average condition of the plunge pool bedrock at each site. The "low" and "high" indexes should be taken as lower and upper bounds, respectively. According to the "typical" Erodibility Index, the investigated plunge pools can be classified from the most vulnerable to scour to the least in this order: Revelstoke, Seven Mile, Portage Mountain, and Peace Canyon. Despite an average rock mass strength, the shale of the Peace Canyon dam foundation is rated as the most scour resistant because of the high RQD, the discontinuous character of the bedding fractures, and the unfavorable orientation of discontinuities with respect to spillway flows. Rock in the identified "hinge zone" at Peace Canyon Dam is better represented by the "low" classification index. The thick-bedded shale of the Portage Mountain plunge pool is stronger due to its high content in sand, but the presence of open rebound fractures reduces its resistance to erosion. The strong argillite at Seven Mile Dam is more vulnerable to hydraulic erodibility because of the positive orientation of main discontinuities relatively to the flow. The average plunge pool bedrock at Revelstoke Dam is rather susceptible to the erosive power of water considering the low mass strength and poor RQD. In each of the four cases of geomechanical classification, the extended range of Erodibility Index values (low -high) marks the heterogeneous character of bedrock. Generally, the bedrock quality increases with depth. For instance, the rebound fractures formed along bedding planes and relaxation joints present in the Peace River Canyon (Peace Canyon Dam and Portage Mountain Project) vanish with depth. Also, weathering of the argillite composing the Seven Mile plunge pool is more important near the surface than at depth. In this analysis, the rock mass quality is assumed constant with depth in reason of a lack of information on the subsurface distribution of geological parameters. The "typical" Erodibility Index is considered to be representative of the bulk of the plunge pool bedrock. 6.2.2 Jet Hydraulics - Computational Methods Prediction of the ultimate depth of scour by the Erodibility Index Method requires that the erosive power of water be quantified along the jet trajectory. The power (KW/m2) available to erode material is a function of the jet hydraulics (Annandale, 2002c) . The rate of energy dissipation in the plunge pool (power available for scour [PA~\) can be expressed as a discretized function of the total head at various elevations ( j  , j  +1, etc...) along the centerline of the submerged jet (ibid.) : dPAj y v j M j dz  1000(zy -z j+ l) ( 6 . 6 ) The change in energy AEj  between points j and j  +1 represents the sum of changes in velocity head, pressure head, and elevation head such as: f..2 AE j = .2 A 2g + Pj ~Pj+1 r (6.7) where V is the jet centreline velocity [m/s], p the static pressure [Pa], Z the elevation [m] , g the gravitational acceleration [m/s2] , and y the unit weight of water [Pa/m]. The second and third terms of equation 6.7 cancel out and the change in energy AEj is equal to the change in velocity head (first term). The dynamic pressure of the jet above hydrostatic converges towards hydrostatic pressure as the jet travels deeper under the plunge pool surface (Golder Associates, 2001). The vertical distribution of power available for scour (P A) in the plunge pool is essentially related to the submerged jet velocity profile. The level of accuracy in the prediction of the latter is determinant in the evaluation of the Erodibility Index Method. Most past research work on the diffusion of jets has been confined to submerged entry cases (Ervine and Falvey, 1987). Little attention has been paid to the more complex problems arising from the influence of impinging free turbulent jets and air entrainment on the diffusion process (ibid.) . Theoretical expressions on computation of jet velocity decay in a plunge pool presented here are derived from two distinct studies: the work by Ervine and Falvey (1987, 1997) and the research conducted at Colorado State University as part of the Dam Foundation Erosion Study (Lewis et al., 1996; Bohrer et al., 1998). Ervine and Falvey (1987) investigated the characteristics of a turbulent circular jet issuing horizontally and plunging through the atmosphere and diffusing into a pool. The free-fall jet was treated as a continuous mass with a solid core (undeveloped jet) dissipating most of its energy in the plunge pool. Research conducted at Colorado State University focused on simulating dam overtopping conditions (rectangular jet) to characterize the behavior of a free-falling jet in both undeveloped and developed conditions. Ervine and Falvey (1987) suggested a relatively simple expression for the calculation of velocity along the centreline of a plunging jet in a pool, based on an estimated 8° of inner core decay and the assumption of linear velocity decay with depth in the zone of established flow. Bohrer et al. (1998) considered an empirical approach using dimensional analysis. From Bohrer et al. (1998), the velocity decay of the jet as it progresses towards the bottom of the plunge pool is a function of the jet velocity at impact with the pool surface, the angle of impingement, the air concentration of the jet at impact, and the gravitational acceleration. Table 6.5 summarizes the computational steps in the prediction of the jet centreline velocity profile in a plunge pool, and subsequently the rate of energy dissipation. First, the flow depth and velocity at the exit of the flip bucket are estimated using the Bernoulli equation. The required basic hydraulic data are the spillway unit discharge, the energy head at the exit of the flip bucket, and the flip angle. Once the initial jet velocity is known, the free jet velocity upon entrance in the plunge pool can be calculated from the ballistic equations knowing the head drop from the bucket lip to the pool surface. However, this overlooks the fact that the jet loses its coherence with the fall and is affected by air resistance. A Dimensional Equation Technique (DET) was developed by Lewis to estimate the velocity of a turbulent, developed jet by incorporating an aerodynamic drag deceleration term into the Ervine and Falvey (1987) expression for undeveloped jets (Lewis et al., 1998). The coherence of the jet (undeveloped or developed) at impact with the pool surface was identified by Bohrer et al. (1998) to affect the submerged jet velocity decay. The determination of the jet breakup length is thus an important step in the computation of the jet velocity profile. Once the jet velocity profile is known, equations 6.6 and 6.7 are used to derive the distribution of the power available to scour. 6.2.3 Maximum  Scour Depth The Erodibility Index Method defines the maximum scour depth as that elevation where the stream power available to scour converges with the power required to scour the bedrock. The power required to scour the plunge pool bedrock at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam was assessed in Section 6.2.1 and a constant "typical" value was assigned to each site (Tables 6.1 to 6.4) . The power available to scour decreases as the jet travels down the plunge pool and for each site varies in function of the spillway discharge and the reservoir and tailwater levels (Section 6.2.2). Derivation of the rate of energy dissipation curve involves a series of computational steps presented in Table 6.5. The Erodibility Index Method was used for each site of study to estimate the maximum scour depth elevation corresponding to each of the 14 scour data sets compiled in Chapter V (Table 5.2) . The results are discussed and compared with the actual scour depth elevations observed on sites. The basic hydraulic data of the 14 scour data sets and the computed parameters for the derivation of the stream power profiles are summarized in Table 6.6. For each site, the jet velocities are similar and do not reflect the changing magnitude of spillway discharges. The jet velocities at impact with the plunge pool surface increase with the drop height from the bucket exit to the water surface. The expression proposed by Lewis (Lewis et al. , 1998) to estimate the impact jet velocity in the case of developed jets is problematic since the retardation factor RF accounting for air resistance exceeds the maximum jet velocity at impact (undeveloped conditions) . In general, the free jet issued from the Peace Canyon and Seven Mile spillways comprises a solid inner core upon entry in the plunge pool, whereas the drop height from the Portage Mountain and Revelstoke spillways is sufficient for the jet to reach the developed state. Therefore, the impact jet velocity might be overpredicted by the ballistic equations in the last two cases. The fall height from the bucket lip to the river enhances the air concentration of the jet which enters the plunge pool. Higher air concentrations are calculated for the hydraulic conditions identified at Portage Mountain Project and Revelstoke Dam. The computed angles of impingement of the jet with the plunge pool surface are also steeper due to the large fall height resulting in the highest values at the Portage Mountain spillway. The angle of impingement determines the submerged jet centreline trajectory along which the velocity and stream power are calculated. The basic parameters presented in Table 6.6 for the derivation of the jet centreline velocity and stream power profiles are based on theoretical considerations and empiricism. The turbulent nature of free-trajectory jets brings a degree of uncertainty to the calculations. The estimates of jet thickness are particularly questionable considering the jet is comprised of two phases (solid inner core and aerated outer shell) and the presence of surface waves. The computed profiles of jet centreline velocity and stream power corresponding to each of the 14 scour data sets are presented in Figures 6.2 to 6.5 for the four distinct plunge pools. The submerged jet velocity profiles were computed using the equations developed as part of the Dam Foundation Erosion Study (Bohrer et al., 1998) . The power required to scour the bedrock at each site is identified on the stream power profiles and the point where the two power curves intersect corresponds to the predicted maximum scour depth elevation. In the case of Peace Canyon Dam (Figure 6.2), the four distinct sets of hydraulic conditions merge into a singular curve of jet centreline velocity and stream power. The Erodibility Index Method predicts a minimum plunge pool elevation around El. 1510 ft, which is actually above the original riverbed. The jet hydraulic computations suggest that the erosive power of water is essentially dissipated within the first ten feet of the plunge pool. The same observation is applicable to the Seven Mile data sets (Figure 6.3). For each profile, the stream power drops drastically in the early portion of the submerged jet trajectory. The observed differences in jet velocity profiles and stream power curves for Seven Mile result mainly from the variation in tailwater levels between the scour data sets. Two of the data sets (SEV79-1/2 and SEV97-4) also represent the behaviour of a developed jet as opposed to undeveloped. The rate of velocity decay is slower in the case of developed jets, but the effects on the stream power profiles are not important (see Section 6.3.2) . The scour depth predictions are acceptable for two sets of hydrauli cs conditions (SEV97-3 and SEV97-4) corresponding to the spillway left chute observations. The assessment of maximum scour downstream of the spillway right chute is however inadequate with underpredictions ranging from 34 ft to 60 ft. The Erodibility Index Method also underestimates the extent of plunge pool scour at Portage Mountain Project (Figure 6.4). by as much as 100 ft. In the case of Revelstoke Dam (Figure 6.5), the predicted maximum scour depth elevations are similar for the two separate sets of hydraulic conditions and approach the observed elevations of the scour hole invert. However, as will be discussed further in the next section, the scour depth predictions at Revelstoke Dam are extremely sensitive to uncertainties in the estimated power required to scour the plunge pool bedrock. Figure 6.6 combines the results for the four sites of study and illustrates the variance between predicted and observed plunge pool floor elevation corresponding to each set of hydraulic conditions. The overall performance of the Erodibility Index Method in the assessment of maximum scour depth for four distinct plunge pools can be characterized by a generalized tendency for underestimation and a standard error of estimate of 53 ft. 6.3 SENSITIVITY ANALYSIS The Erodibility Index Method was applied to the four sites of study based on the best information available at the time. Considering the extended range of Erodibility Index that characterizes each plunge pool bedrock and the uncertainties involved in the jet hydraulics computations, a sensitivity analysis of key parameters involved in the scour depth analysis is performed. The parameters which are further investigated based on their respective influence on scour depth predictions are the Erodibility Index, the jet velocity profile, the jet impact velocity, and the jet air concentration at impact with the plunge pool surface. 6.3.1 Erodibility Index The characterization of plunge pool bedrock accounts for half of the Erodibility Index Method. As mentioned in Section 6.2.1, the "typical" Erodibility Index was considered to be representative of the bulk of the plunge pool bedrock for the maximum scour depth analysis (Section 6.2.3). However, the heterogeneous character of the rock mass and uncertainties in the assessment of geological parameters requires that the lower and upper bounds defined for the Erodibility Index (Tables 6.1 to 6.4) be taken into account. The effects of a changing Erodibility Index (K) on the scour depth predictions are evaluated. The predicted elevations of maximum scour depth for the expected range of Erodibility Index (low, typical, and high rock mass quality) are presented in Table 6.7. For each site, given the extended range of Erodibility Index that characterizes the plunge pool bedrock, the power required to scour the rock mass (PR ) varies over three orders of magnitude, and up to four in the case of Revelstoke. For the Revelstoke plunge pool, a stream power of 1.53 KW/m2 is required to scour the zones of weaker rock as opposed to 1480 KW/m2 for the sounder rock. Accordingly, the estimates of maximum scour depth elevation range between El. 1278 ft and El. 1443 ft. However, the difference in predicted scour extent when considering the low versus high Erodibility Index is not always representative of the gap that separates the two indices. For instance, the plunge pool bedrock at Peace Canyon Dam and Portage Mountain Project shows a similar range of Erodibility Index but the corresponding variation in elevation is more considerable in one case than in the other. For the Peace Canyon plunge pool, uncertainties in the Erodibility Index value result in less than 2 0 ft variation in elevation. In the case of Portage Mountain, the difference in elevation reaches 70-80 ft. For all sites, the sensitivity of the Erodibility Index Method to an uncertainty in the Erodibility Index value is increased for small values of Erodibility Index (K  < 500, P R < 100). The Revelstoke scour depth predictions are the most affected by the uncertainties regarding the Erodibility Index value and the Peace Canyon results are the least sensitive. A significant range of estimated scour depth elevations for each site was to be expected from the variation between lower and upper bound values of Erodibility Index, but the changing sensitivity of the Erodibility Index Method with respect to the Erodibility Index is problematic. However, this characteristic is directly attributed to the rather sharp profile of stream power available for scour which is derived from a jet velocity decay function (see next section). For most cases (ten scour data sets out of fourteen), the Erodibility Index Method still underestimates the scour depth even when considering the lower bound Erodibility Index assigned to the plunge pool bedrock. 6.3.2 Jet Velocity Profile As seen in Section 6.2.2, the vertical distribution of power available for scour in the plunge pool is derived from the jet velocity decay along its centreline trajectory. In the maximum scour depth analysis (Section 6.2.3), the jet centreline velocity was calculated using the equations developed as part of the Dam Foundation Erosion Study (Bohrer et al., 1998) and a function of the jet condition (undeveloped or developed) at impact with the pool surface. The equation proposed by Ervine and Falvey (1987) was presented in Table 6.5 along with the equations by Bohrer et al. (1998), but was not used because it refers to circular jets. The two forms of expression were here applied to a unique data set of each site to observe the influence of the jet velocity profile on the Erodibility Index Method. The extent to which the scour depth predictions are affected by the equation used to compute the jet velocity decay in the plunge pool is discussed. The velocity profiles as given by the different equations and the corresponding stream power profiles are illustrated for each of the four study sites in Figures 6.7 to 6.10. The two equations developed by Bohrer et al. (1998) for undeveloped and developed jets provide similar jet velocity decay trends, although the rate of velocity decay is slower in the case of developed jets. The difference between the two velocity profiles by Bohrer et al. (1998) reflects on the stream power profiles with a maximum variation in elevation of 11 ft in the case of Portage Mountain (Figure 6.9). As expected, the equation for undeveloped jets gives the most conservative scour depth estimates. The Ervine and Falvey (1987) equation also refers to undeveloped jets but the resulting stream power profiles show no consistent trend from one site to another relatively to the computed profiles using Bohrer et al. (1998). The scour depth predictions when using the velocity decay equation by Ervine and Falvey (1987) as opposed to Bohrer et al. (1998) are more conservative for Peace Canyon, similar for Seven Mile, and underestimated for Portage Mountain and Revelstoke. The maximum range of estimated scour depth elevations covered by the three different equations for a unique data set varies from 5 ft in the case of Seven Mile to 43 ft in the case of Revelstoke. In general, the two forms of expression investigated for the computation of jet velocity decay in the plunge pool result in similar stream power profiles with moderate vertical discrepancy. Apart from Peace Canyon, the use of the equation by Ervine and Falvey (1987) over the Dam Foundation Erosion Study equations (Bohrer et al. , 1998) does not improve the accuracy of the Erodibility Index Method. The performance of the Erodibility Index Method lies in the description of the jet velocity profile underwater and the need for an improved expression of jet velocity decay in a plunge pool is obvious. The Bohrer et al. (1998) equations are based on empiricism and were derived from a limited range of test simulations whereas the equation by Ervine and Falvey (1987) refers to circular jets and depends mainly on the estimated jet diameter at impact with the plunge pool surface. 6.3.3 Jet Impact Velocity The impact velocity of the jet with the plunge pool surface is the starting point to the computation of the submerged jet velocity profile. In the maximum scour depth analysis (Section 6.2.3), the jet impact velocity was calculated using the ballistic equations. The velocities thus computed represent maximum values that don't account for air resistance. The deceleration caused by air drag becomes important when the plunge length is such that the jet loses its coherence (e.g. Portage Mountain and Revelstoke spillways). The equation proposed by Lewis et al. (1996) (Table 6.5) failed to estimate the air retardation factor that lowers the impingement velocity of a developed jet (Section 6.2.3). To account for the uncertainties involved in the calculation of the jet impingement velocity, the velocity and stream power profiles were recalculated for one scour data set of each site using a series of jet impact velocities below the original one. The variation in scour depth predictions caused by each diminution of jet impact velocity is examined. The velocity profiles and associated stream power profiles derived from a regression of jet impact velocity at each of the four sites of study are presented in Figures 6.11 to 6.14. For all sites, the divergence in stream power profiles originating from the values of jet impingement velocity increases as the stream power decreases. Therefore, the scour depth predictions using the Erodibility Index Method are more sensitive to uncertainties in the calculation of jet impingement velocity when the stream power required to scour the plunge pool bedrock is low. The vertical discrepancies in stream power profiles resulting from the regression of jet impact velocity with constant increments are also lessened at each increment. For the Peace Canyon and Seven Mile plunge pools, the estimated maximum scour depth is reduced by at most 2 ft and 5 ft, respectively, when the calculated jet velocity at impact with the water surface is lowered by increments of 10 ft/s. As for the Portage Mountain and Revelstoke plunge pools, the maximum variation in elevation for each increment of 20 ft/s reaches 10 ft and 15 ft, respectively. To a 50% reduction in maximum jet impact velocity (calculated from the ballistic equations) corresponds a difference in elevation of about 10 ft for Peace Canyon, 20 ft for Seven Mile, 40 ft for Portage Mountain, and 50 ft for Revelstoke. The conservative approach when using the Erodibility Index Method is to assume the jet that impinges the plunge pool surface as compact and use the ballistic equations to calculate the impact velocity. The sensitivity analysis confirms that the jet velocity at impact with the plunge pool surface is an important parameter in the assessment of maximum scour depth using the Erodibility Index Method. However, in the cases of Peace Canyon and Seven Mile, the effects of a diminution in jet impingement velocity are rather minor on the scour depth predictions. In this study, the uncertainties regarding the estimation of jet impact velocity were primarily directed toward the sites of Portage Mountain and Revelstoke because of the lengthy plunge of the spillway jet. Surprisingly, the Erodibility Index Method underestimates the maximum scour depth elevation of both plunge pools although using the maximum theoretical value of jet impact velocity. 6.3.4 Jet Air Concentration at Impact The air concentration of the jet at impact with the plunge pool surface is an important parameter in the computation of the jet velocity profile when using the equations by Bohrer et al. (1998) . In Section 6.2.3, the air/water ratio of the jet at the moment of impact with the water surface was estimated using an expression by Ervine and Falvey (1987) which relates to the initial jet thickness and plunge length through the atmosphere. Uncertainties arise from the. fact that the expression was initially derived from small-scale experiments on compact circular jets. To evaluate the importance of the parameter on the Erodibility Index Method, the velocity and stream power profiles were computed using different air concentrations of the impinging jet for a unique scour data set of each site. The influence of the estimated jet air concentration on the scour depth predictions is observed. The velocity profiles as function of the jet air concentration at impact and the corresponding stream power profiles are shown for each of the four study sites in Figures 6.15 to 6.18. For all sites, the divergence in stream power profiles caused by the different values of air concentration increases as the stream power decreases. Therefore, the scour depth predictions using the Erodibility Index Method are more sensitive to errors in the estimation of jet air concentration when the stream power required to scour the plunge pool bedrock is low. For the Peace Canyon and Seven Mile plunge pools, an uncertainty of 20% in the value of jet air concentration would translate to a maximum error in scour depth elevation of 2 ft and 7 ft, respectively. The same increment of air concentration produces a variation in scour depth elevation of approximately 20 ft for Portage Mountain and 30 ft for Revelstoke. The conservative approach when using the Erodibility Index Method is to consider the jet that impinges the plunge pool surface as non-aerated (air concentration of 0%). Due to the uncertainties involved in the estimation of air concentration of the spillway jet at impact with the plunge pool surface, the level of sensitivity of the Erodibility Index Method to this parameter is unacceptable. When using the equations developed as part of the Dam Foundation Erosion Study (Bohrer et al. , 1998) to compute the submerged jet velocity profile, the constituent value of jet air concentration at impact should be estimated within 10% of accuracy. In the four study cases, the Erodibility Index Method still underestimates the maximum scour depth for Peace Canyon, Seven Mile, and Portage Mountain even when assuming a maximum jet density (air concentration of 0%) . 6.4. EVALUATION The concept of the Erodibility Index Method is appealing because it gives equal weighting to both jet hydraulics and plunge pool geology. The Erodibility Index, which is used to characterize the geological conditions of the plunge pool, has well established roots in the engineering classification of rock masses for tunnel support (the Rock Tunnelling Quality Index, Q) and excavation (Kirsten's (1982) Ripability Index). However, by incorporating a new set of parameters through the use of a geomechanical index, a new series of uncertainties is added to the estimation of maximum scour depth. This is especially true since engineering and geological investigations don't usually cover the plunge pool area. Besides, the geological conditions of the plunge pool are commonly heterogeneous and cannot be represented adequately by a single factor. This situation was portrayed in the analysis with the expected power required to scour the bedrock of each plunge pool extending over at least three orders of magnitude. With regard to the jet hydraulics, there is an obvious need for improved analytical tools in the description of two-phase jet velocity decay in a plunge pool. For the majority of computed stream power profiles, the bulk of the energy was dissipated within the initial few feet of jet submergence. Therefore, the maximum scour depth was underestimated even when considering the maximum jet impact velocity, the maximum jet density, and the minimum Erodibility Index value. However, the main weakness of the Erodibility Index Method is that the vertical distribution of power available for scour in the plunge pool is essentially related to the submerged jet velocity profile which does not reflect the changing magnitude of spillway discharges; jet velocities are similar for a given site and vary as a function of the reservoir and tailwater levels. In addition, of the 150 field observations that form the basis of the erodibility threshold relationship (Figure 2.4), 137 were made from auxiliary earth spillways. The processes by which these spillway channels resist the erosive attack of flowing water should not be compared to the scouring action of a spillway jet. So even with an accurate bedrock characterization and improved equations for the computation of jet velocity profile, the use of the Erodibility Index Method for the assessment of plunge pool scour is questionable. The overall performance of the Erodibility Index Method in the assessment of maximum scour depth for four distinct plunge pools was characterized by a generalized tendency for underestimation and a standard error of estimate of 53 ft. As part of the Dam Foundation Erosion Study, The Erodibility Index Method has been computer coded and a first test version of the software was available in June 2002. The project has however been shelved indefinitely and the prospects of the Erodibility Index Method remain unknown. Table 6.1. Peace Canyon/Erodibility Index Characterization BEDROCK - SHALE Low ] Typical High Unconfined Compressive Strength (UCS)1 5,000 psi 10,500 psi 17,000 psi Ms 34 72 117 Rock Quality Designation2 70% 90% 95% RQD 70 90 95 Number of Joint Sets 3 + random 3 3 Jn 3 .34 2 . 73 | 2 . 73 Roughness of Joint Open Discontinuous i Discontinuous Jr 1 . 0 4 . 0 [4.0 Alteration of Joint Crushed rock filling Unaltered j Unaltered Ja 4 . 0 1. 0 i 1. 0 -Main Discontinuity N30-40 °/90 ° N60°/2° N60°/2° -Dip Direction With flow With flow -Apparent Dip in Direction of Flow 90° 1° i 1° -Ratio of Joint Spacing 1:1 1:2 1 : 1 Js 1 . 14 1 . 33 1 . 50 K=Ms»(RQD/Jn)•(Jr/Ja).Js | 203 12,600 24,400 r PR = K0'75 [KW/m2] 53 . 8 1190 1950 Notes. 1. Based on Laboratory Rock Testing Program of 1959 5 tests-dry), 1967 (4 tests-dry), and 1975 (2 tests-dry) (BCH Report No. 354, 1959; BCH Drawing Nos. 1007-C14-C3854 & 1007-14-D424 9). 2. Based on 502 ft of core logging from 1977-78 post- grout drilling of the dam foundation (hole series STH); the statistical distribution is presented below. RQD Distribution 400 350 300 250 200 150 100 50 0 £ £ § £ £ § RQD [%] BEDROCK - ARGILLITE Low Typical High Unconfined Compressive Strength (UCS)1 Ms 5,000 psi 34 20,000 psi 138 33,000 psi 228 Rock Quality Designation2 RQD 60% 60 90% 90 95% 95 Number of Joint Sets3 Jn 4 4 . 09 3 + random 3 . 34 3 2 . 73 Roughness of Joint3 Jr j Smooth planar | 1.0 Smooth undulating 2 . 0 Discontinuous 4 . 0 Alteration of Joint3 Ja Graphitic 4 . 0 Slightly altered 2 . 0 Unaltered 1.0 -Main Discontinuity -Dip Direction -Apparent Dip in Direction of Flow -Ratio of Joint Spacing Js N110-120°/4 0-4 5° With flow 40° 1:4 0.46 N110-120°/40-45° With flow 40° 1:2 0.49 N110-120°/40-45° With flow 40° 1:1 0 . 53 K=Ms»(RQD/Jn)•(Jr/Ja)»Js 57 . 4 1, 820 16,800 P R = K 0' 7 5 [KW/m2] 20 . 8 279 1480 Notes. 1. Based on Laboratory Rock Testing Program of 1973 (8 tests) (BCH Report No. 750, 1975). 2. Based on 414 ft of core logging from 1973 exploration drilling in the riverbed argillite (DH73-1A/3/8/10); the statistical distribution is presented below. 3. Reference BCH Report No. PSE362 - Appendix G, 2001. RQD Distribution 200 £ 150 o> g 100 o o 50 0 =F 5s $ § 9 £ . £ S § £ £ RQD [%] BEDROCK Low 1 Typical High Unconfined Compressive Strength (UCS)1 Ms 10,000 psi 69 1 f 18,000 psi j 124 22,000 psi 152 Rock Quality Designation RQD 50% 50 | 80% ; 80 95% 95 Number of Joint Sets Jn 4 4 . 09 3 + random 3 . 34 3 2 . 73 Roughness of Joint Jr Open 1. 0 Open 1. 0 Discontinuous 4 . 0 Alteration of Joint Ja Crushed rock filling 4 . 0 Surface staining 1. 0 Surface staining 1. 0 -Main Discontinuity -Dip Direction -Apparent Dip in Direction of Flow -Ratio of Joint Spacing Js N132°/10° With flow 10° 1:4 0 . 98 N132°/10° With flow 10° 1:1 1.25 i N132°/10° With flow 10° 1 : 1 1 .25 K=Ms»(RQD/Jn)•(Jr/Ja)«Js 207 3 , 710 26,400 PR = K0'75 [KW/m2] 54 . 5 476 2070 Note. 1. Based on Laboratory Rock Testing Program of 1958 (4 tests-dry), 1959 (6 tests-dry), and 1965 (33 tests-dry); the statistical distribution is presented below (BCH Report No. 150, 1958; BCH Report No. 31, 1959; BCH File No. 003805546, 1965). UCS Distribution - Shale 10 8 6 4 2 0 to V <b O <V V- Co % iv > UCS [x1,000 psi] UCS Distribution - Sandstone 10 8 6 4 2 0 <b ^ <6 <v V- Co 'b is iV > A) ib is UCS [x1,000 psi] Table 6.4. Revelstoke/Erodibility Index Characterization ( BEDROCK SOIL | Low Typical | High j Typical Unconfined Compressive Strength (UCS)1 Ms 3,000 psi 21 9,000 psi 62 20,000 psi 138 SPT=30-50 0 . 19 Rock Quality Designation2 RQD 5% 5 10% 10 80% 80 5% 5 Number of Joint Sets Jn r " 5 . 00 3 + random 3 .34 3 2 . 73 5 5 . 00 Roughness of Joint Jr Smooth planar 1 . 0 Rough planar 1 . 5 Rough undulating 3 . 0 Jr/Ja=0.6 (<f> = 30°) Alteration of Joint Ja Graphitic,>5 mm 13 . 0 Slightly clayey 3 . 0 Surface staining 1. 0 j -Main Discontinuity -Dip Direction -Apparent Dip in Flow Direction -Ratio of Joint Spacing Js N56°/26° With flow 5° 1:4 1 . 09 N56°/26° With flow 5° 1:2 1 .23 N56°/26° With flow 5° 1:1 1 .39 1 . 00 K=Ms*(RQD/Jn)•(Jr/Ja)*Js 1 . 76 114 16,900 "oTTi P R = K0-75 [KW/m2] 1 1.53 34 . 9 1480 0 . 183 Notes. 1. Based on Laboratory Rock Testing Program of 1972 (16 tests), 1976 (42 tests), and 1978 (28 tests); the statistical distribution is presented below (BCH Report No.664, 1973; BCH File No.15-2-57, 1976; BCH Report NO.H1204, 1980). 2. Based on 905.5 ft of core logging from exploration drilling in the plunge pool surroundings (DH73-8/9, DH74-1A/2, DH75-17/18/19/20/21, and DH76-29/32/36); the statistical distribution is presented below. 3. For cohesionless granular material, P„ = 0.480K0'44 [KW/m2] (Wittier et al. , 1998). (0 20 in « 15 10 o 5 o z 0 o UCS Distribution to <V $ $ UCS [x1,000 psi] <V RQD Distribution o O 300 200 1 0 0 -0 ^ "> Is y d C5 e> C5 ,ci K tV ^ fr <0 £ Q .O CO ,S> ,§> lb Q> £ £ RQD [%] Table 6.5. Jet Hydraulics - Computational Steps Computational Step Equation Source 1. Estimate the flow depth at the exit of the spillway bucket (jet thickness), do. d0= y  COS 9 n 2 * E = y + -2gy 2 Bernoulli Equation 2. Calculate the jet velocity at the exit of the spillway bucket, Vo . V 0=q!y 3. Determine the jet breakup length, Lb. The jet breakup length is the fall distance before the free jet loses its coherence and becomes a conglomeration of discrete water particles. The breakup length marks the limit between undeveloped jet and developed jet. C2 =-yd0F02 2Lh \2 + 1 - 1 d0F: C = \.\A-Tu-Fg Fo=Vol4gd~o Ervine et al., 1997 4. Compute the jet velocity at impact with the plunge pool surface, Vi. Undeveloped jet V , = ^ + 2 g H d Developed jet V ;=^+2gH d -RF RF = ^Cd(palpjH dl<S>)vZ J Ervine et ! al., 1997 Lewis et al., 1996 5. Estimate the jet thickness at impact with the plunge pool surface, di. d, = d„ Ervine et al., 1997 6. Estimate the mean air concentration of the jet at impact, Ci. c,  =-B, l + 5; where B: = 0.2 H„ Ervine and Falvey, 1987 7. Determine the underwater velocity at all points along the jet centreline trajectory, V. Undeveloped jet V  = AVid;/L V  = V,  for L < Ad, Undeveloped jet — = 0.0675 V; + 0.1903 Ervine and Falvey, 1987 V  = V:  for Developed jet In V  = V,  for In = 0.638 In 0.638 Bohrer et al., 1998 Notes. * Assumed energy loss on spillway surface of 5% for Peace Canyon and Seven Mile and 10% for Portage Mountain and Revelstoke. ** Estimated turbulence intensity (Tu) of 5% for flip bucket jets. *** Calculated from gravitational considerations only. Nomenclature Bi [--] = Ratio of air to water flow rate entering the plunge pool C [--] = Jet turbulence parameter Cd [--] = Drag coefficient Ci [--] = Air concentration of the jet at impact with the plunge pool surface do [ft] = Initial jet thickness (at the exit of the flip bucket) di [ft] = Jet thickness at impact with the plunge pool surface E [ft] = Specific energy available at the exit of the flip bucket Fo [--] = Initial Froude Number of jet g [ft/s2] = Gravitational acceleration Hd [ft] = Head drop from the bucket lip to the plunge pool surface q [ft2/s] = Spillway unit discharge (discharge per unit width) L [ft] = Distance along the jet centreline trajectory beneath the water surface Lb [ft] = Jet breakup length RF [ft/s] = Retardation factor (aerodynamic drag deceleration term) Tu [--] = Turbulence intensity of jet Vo [ft/s] = Initial jet velocity (at the exit of the flip bucket) Vi [ft/s] = Jet velocity at impact with the plunge pool surface V [ft/s] = Underwater velocity along the jet centreline trajectory y [ft] = Elevation head at the exit of the flip bucket 0 [°] = Angle of flip bucket exit ® [ft] = Diameter of a sphere having the same volume as a water drop pa [lbm/ft3] = Air density pw [lbm/ft3] = Water density Table 6.6. Jet Hydraulics - Basic Parameters Scour Data Set 1 q [ c f s / f t ] E [ f t ] 9 [°] Vo [ f t / s ] do [ f t ] Hd [ft] ( Lb [ f t ] Hd/Lb ( j e t coherence) Vi [ f t / s ] RF [ f t / s ] di [ f t ] 1 C i | [%] 1 6 ' | [°] PCN80-3/4 4 7 0 | 1 1 0 2 0 | 82 . 1 5 . 3 6 . 6 j 82 0 . 0 8 (undeveloped) 8 4 . 7 I 5 . 6 I 1 8 2 4 PCN80-5/6 [ 3 1 0 1 1 0 2 0 I 8 2 . 9 i 3 . 5 5 . 6 5 6 0 . 1 0 (undeveloped) ! 8 5 . 0 i r~ 3 . 7 | 2 0 2 4 PCN96-3 | 8 1 0 1 2 0 2 0 84 . 3 9 . 0 8 . 6 | 1 3 3 0 . 0 6 (undeveloped) | 8 7 . 5 | 9 . 4 1 6 2 5 PCN96-5/6 | 3 6 0 | 1 2 0 2 0 | 8 6 . 3 3 . 9 5 . 9 | 62 0 . 0 9 (undeveloped) | 8 8 . 5 | i 4 . 1 2 0 2 3 SEV79-1/2 | 2 5 0 1 4 7 3 0 1 9 6 . 3 2 . 3 4 0 . 8 3 7 1 . 0 9 (developed) | 1 0 9 5 9 . 9 2 . 4 | 4 6 4 0 SEV80-1/2 | 5 1 0 1 5 1 3 0 | 9 6 . 7 I 4 . 6 3 1 . 2 I 7 3 s J 0 . 4 3 (undeveloped) 1 0 7 5 . 0 | 3 4 3 8 SEV82-1/2 | 5 7 0 1 5 0 3 0 J 9 6 . 2 I 5 . 1 3 0 . 5 !8 1 0 . 3 8 (undeveloped) 1 0 6 | 5 . 7 3 3 3 8 SEV97-1/2 | 6 2 0 1 6 4 3 0 | 1 0 1 I 5 ' 4 2 3 . 6 i 8 5 0.2 8 (undeveloped) 1 0 8 j 1 6 . 0 , 3 0 3 6 SEV97-3 5 2 0 1 5 4 3 0 f~98 . 0 § 4 . 6 3 3 . 2 1 7 3 0 . 4 6 (undeveloped) | 1 0 8 j s | 5 . 0 3 5 3 8 SEV97-4 1 0 0 1 5 4 3 0 I 9 9 . 4 0 . 9 3 1 . 0 | 1 5 2 . 0 8 (developed) 1 1 0 9 ? | 1 2 5 1 . 0 5 4 3 8 PMD72 [ 1 2 4 0 2 8 4 3 0 j 1 3 3 8 . 0 2 4 4 | 1 2 9 1 . 8 9 (developed) j 1 8 3 j 5 1 9 | 7 . 9 5 2 5 1 PMD96 1 9 0 0 2 7 0 3 0 1 3 0 6 . 0 | 2 3 3 s 9 8 2.3 9 (developed) | 1 7 9 5 3 3 5 . 9 ! 5 5 5 1 REV84 I 1 9 0 1 9 5 0 | 1 1 2 1 . 7 5 1 7 3 2 9 I 5 . 8 9 (developed) 1 5 4 4 3 1 1 . 5 | 6 7 4 3 REV86 J 3 3 0 j 2 3 3 1 0 1 2 2 2 . 7 J" 1 6 1 " | 4 6 j 3 . 5 1 (developed) 1 5 9 | 4 3 0 2 . 4 1 6 1 4 0 | Notes. | 1. Refer to Table 5.2 for a complete description of hydraulic conditions in each of the scour data sets. | 2. Refer to Tables 6.5 and 5.1 for a description of parameters. Table 6.7. Sensitivity Analysis of Erodibility Index SCOUR DATA SET PREDICTED PLUNGE POOL FLOOR ELEVATION OBSERVED Low Typical I High i '1 PEACE CANYON ! K = 2 03 PR = 53.8 K = 12,600 PR = 1190 K = 24,400 PR = 1950 PCN80-3/4 El. 1500 ft El. 1512 ft No scour El. 1458 ft PCN80-5/6 El. 1501 ft El. 1512 ft No scour El. 1482 ft PCN96-3 El. 1498 ft El. 1510 ft El. 1512 ft El. 1459 ft PCN96-5/6 El. 1499 ft El. 1511 ft El. 1512 ft El. 1481 ft j SEVEN MILE K = 57 .4 PR = 2 0.8 K = 1,820 PR = 279 K = 16,800 PR = 14 8 0 SEV79-1/2 El. 1474 ft ! i El. 1499 ft El. 1507 ft El. 1465 ft SEV8 0-1/2 El. 1474 ft | El. 1504 ft El. 1513 ft El. 1462 ft SEV82-1/2 El. 1474 ft I El. 1504 ft El. 1514 ft El. 1450 ft SEV97-1/2 El. 1479 ft j El. 1510 ft El. 1520 ft j El. 1450 ft SEV97-3 El. 1480 ft 1 ! El. 1511 ft El. 1520 ft El. 1503 ft SEV97-4 El. 1501 ft El. 1522 ft El. 1527 ft | El. 1523 ft i 1 " :.: 1 1 :: : v:/.:.:.:.:.:..: v..:.: i PORTAGE MOUNTAIN K = 207 Pr = 54.5 K = 3,710 PR = 476 K = 26,400 PR = 2074 PMD72 El. 1544 ft El. 1600 ft El. 1621 ft El. 1520 ft PMD96 El. 1565 ft j El. 1615 ft El. 1633 ft El. 1516 ft ' i I REVELSTOKE K = 1.76 PR = 1.53 I K = 114 PR = 34.9 K = 16,900 PR = 14 8 0 REV84 El. 1296 ft El. 1397 ft El. 1434 ft El. 1392 ft REV86 El. 1278 ft El. 1399 ft El. 1443 ft El. 1380 ft Notes. K: Erodibility Index. Pr: Power required to scour the bedrock; function of K. -2. THE ERODIBILITY INDEX METHOD - CONCEPTUAL APPROACH J j GEOTECHNICAL Stream Power 2fj JET HYDRAULICS Stream Power SCOUR DEPTH Stream Power 0.01 0.1 1 10 100 1,000 10,000 Erodibility Index, K Figure 6.1. Conceptual Approach of the Erodibility Index Method in the Assessment of Plunge Pool Scour 1540 -1530 -1520 -1510 -1500 -1490 -1480 .9 1470 (O jj 1460 1450 1440 1430 1420 1410 1400 1540 1530 1520 1510 1500 1490 PEACE CANYON Submerged Jet Velocity Profile Scour Data Set PCN80-3/4 PCN80-5/6 PCN96-3 PCN96-5/6 Tailwater Elevation El. 1519 ft El. 1519 ft El. 1519 ft El. 1519 ft Impact Velocity Vi: 84.7 ft/s Vi: 85.0 ft/s Vi: 87.5 ft/s Vi: 88.5 ft/s I I I II I I I I I I I I I I I II I I l I II II I I I I I I II I I I I I I I I I I II I II I I I II II I I I I I I I I ! I I I II I I I I I I I I I I I 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Jet Centreline Velocity [ft/s] 1480 — .1 1470 (O jy 1460 1450 1440 1430 1420 1410 1400 Submerged Jet Stream Power Profile Scour Data Set PCN80-3/4 PCN80-5/6 PCN96-3 PCN96-5/6 Maximum Scour Depth El. Predicted . Observed El. 1512 ft El. 1512 ft El. 1510 ft El. 1511 ft El. 1458 ft El. 1482 ft El. 1459 ft El. 1481 ft | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | II i i | i 0 100 200 300 400 500 600 700 800 Stream Power [KW/m2] i |II i i | i i i i | i i i i 900 1000 1100 1200 Figure 6.2. The Erodibility Index Method Applied to the Peace Canyon Plunge Pool c o 1540 1530 1520 1510 1500 1490 1480 _ 1470 a> 1460 LLI 1450 1440 1430 1420 1410 1400 1540 1530 1520 1510 1500 1490 1480 _ 1470 J3 1460 a c o SEVEN MILE Submerged Jet Velocity Profile Scour Data Tailwater Impact :— Set Elevation Velocity E— SEV79-1/2 El. 1516 ft Vi 109 ft/s E_ SEV80-1/2 El. 1527 ft Vi 107 ft/s E SEV82-1/2 El. 1528 ft Vi 106 ft/s E— SEV97-1/2 El. 1535 ft Vi 108 ft/s E SEV97-3 El. 1535 ft Vi 108 ft/s SEV97-4 El. 1535 ft Vi 109 ft/s i 111111111111111111111111111111111111111 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Jet Centreline Velocity [ft/s] Submerged Jet Stream Power Profile 1450 1440 1430 1420 1410 1400 Scour Data Maximum Scour Depth El. Set Predicted Observed SEV79-1/2 El. 1499 ft El. 1465 ft SEV80-1/2 El. 1504 ft El. 1462 ft SEV82-1/2 El. 1504 ft El. 1450 ft SEV97-1/2 El. 1510 ft El. 1450 ft SEV97-3 El. 1511 ft El. 1503 ft SEV97-4 El. 1522 ft El. 1523 ft i | > i i i | i i i i | i i i i | i i i i | i i II | II i i | i M i | i 100 200 300 400 500 600 700 800 Stream Power [KW/m2] i | i i i i | i i i i | i i i i 900 1000 1100 1200 Figure 6.3. The Erodibility Index Method Applied to the Seven Mile Plunge Pool PORTAGE MOUNTAIN Jet Centreline Velocity [ft/s] Stream Power [KW/m2] Figure 6.4. The Erodibility Index Method Applied to the Portage Mountain Plunge Pool REVELSTOKE 1460 — _— 1450 — Submerged Jet Velocity Profile 1440 1430 E— 1420 E— 1410 - - E-1400 —5 :— ET - i 1390 —z E— a o '4-t (U 1380 > dJ UJ 1370 1360 1350 1340 1330 -Scour Data Tailwater Set Elevation REV84 El. 1448 ft Impact Velocity Vi: 154 ft/s 1320 z REV86 El. 1460 ft Vi: 159 ft/s -— i— 1310 E— 1300 i M i i |M i i |M 11 | 1 1 1 1 | 1 1 1 1 | 1 1 1111 1 111 1 111 1 | 1 1 ! 1 | 1 1 1 1 | 1 1 1 1 | 1 1 1 1 | 1 1 1 1 | 1 1 1 1 | 1 1 1 1 1 111 1 1 11111 11 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Jet Centreline Velocity [ft/s] 1450 — - , : Submerged Jet Stream Power Profile i 1440 - E 1430 i 1420 E— 1410 :— 14 00 i g 1390 E c o 1380 -o 73 r <o > 1370 - II Ul E 33 z— LU 1360 —i lo -1350 ; ^ II Scour Data Maximum Scour Depth El. -1340 ii i—* Set Predicted Observed E— 1330 REV84 El. 1397 ft El. 1392 ft ~ . . 1320 —E REV86 El. 1399 ft El. 1380 ft - — 1310 — E— 1300 I i i i i | I I i i j i i i i i i i i | II i i | i i i i | M i i | i M i | i i i i | i II i | i i M | M i i c 100 200 300 400 500 600 700 800 900 1000 1100 1200 Stream Power [KW/m2] Figure 6.5. The Erodibility Index Method Applied to the Revelstoke Plunge Pool 1620 - q 1600 - = 1580 -1560 -1540 1520 c .2 1500 ro > QJ m 1480 T3 oi Q. E o U 1340 Notes. 1. Refer to Figures 6.2 to 6.5 for estimates of maximum scour depth elevation. 2. Refer to Table 5.2 for details on the scour data sets. I I I I | I I II | I I I I | I II I | II II | I I I I | I I I I | I I I I | I I I I | I I I I | I II I | I I I I | I I I I | I II I | I I I I | I I I I | I I I I | I I II | I I I I j I I I I | I I I I | I 1340 1360 1380 1400 1420 1440 1460 1480 1500 1520 1540 1560 Observed Elevation [ft] 1580 1600 1620 Jet Centreline Velocity [ft/s] Stream Power [KW/m2] Jet Centreline Velocity [ft/s] Stream Power [KW/m2] PORTAGE MOUNTAIN c o +-» fU & 1660 — | I I I I | I II I | M I I | I I I I| I I I I | I I I I| I I I I | 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Jet Centreline Velocity [ft/s] 1660 - q 1640 — 1620 -1600 -1580 -1560 -1540 -1520 -1500 — 1480 — Predicted Maximum Scour Depth El. Ervine and Falvey 1987: El. 1625 ft Bohrer et al. 1998 (Undeveloped): El. 1604 ft Bohrer et al. 1998 (Developed): El. 1615 ft Observed El. 1516 ft 1460 I I | I I I I | I I I I | II I I | I I I! | I I I I | I I I I | I I II | I II I | I II I | I II I | I I I 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 Stream Power [KW/m2] c o (D 5 c o fD £ 1460 1450 1440 1430 1420 1410 1400 1390 1380 1370 1360 1350 1340 1330 1320 1310 1300 1460 1450 1440 1430 1420 1410 1400 1390 1380 1370 1360 1350 1340 1330 1320 1310 1300 DATA SET: REV86 DEVELOPED JET JET VELOCITY PROFILE Ervine and Falvey 1987 Bohrer et al. 1998 (Undeveloped) Bohrer et al. (Developed) 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Jet Centreline Velocity [ft/s] Observed El. 1380 ft Predicted Maximum Scour Depth El. Ervine and Falvey 1987: El. 1432 ft Bohrer et al. 1998 (Undeveloped): El. 1389 ft Bohrer et al. 1998 (Developed): El. 1399 ft i i i i i i i i i i i i i i i i i i i i i i 100 200 300 400 500 600 700 800 Stream Power [KW/m2] 900 1000 1100 1200 Jet Centreline Velocity [ft/s] Stream Power [KW/m2] c o 1540 1530 1520 1510 1500 1490 1480 _ 1470 JS 1460 LU 1450 1440 1430 1420 1410 1400 1540 1530 1520 1510 1500 1490 1480 1470 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Jet Centreline Velocity [ft/s] c o <u J5 1460 1450 1440 1430 1420 1410 1400 i i i i | II i i | i II 100 200 Vi = 108 ft/s: El. 1510 ft Vi = 100 ft/s: El. 1514 ft Vi = 90 ft/s: El. 1519 ft Vi = 80 ft/s: El. 1523 ft Vi = 70 ft/s: El. 1526 ft Vi = 60 ft/s: El. 1529 ft Observed El. 1450 ft-i i i I | M ii|i ii i| i i M | i i i i | i i i i | i i i i | M i i | i M 300 400 500 600 700 800 900 1000 1100 1200 Stream Power [KW/m2] PORTAGE MOUNTAIN Jet Centreline Velocity [ft/s] Stream Power [KW/m2] Jet Centreline Velocity [ft/s] Stream Power [KW/m2] Jet Centreline Velocity [ft/s] Stream Power [KW/m2] Jet Centreline Velocity [ft/s] Stream Power [KW/m2] PORTAGE MOUNTAIN 1640 — 1620 1600 — £7 1580 -.1 1560 -1540 —| 1520 -1500 — 1480 — (D > CL> 1460 1660 ii 11 II 1111111111111111111111111111111111111111111111111111111111111111 111111111111II11111111111 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Jet Centreline Velocity [ft/s] 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 Stream Power [KW/m2] Jet Centreline Velocity [ft/s] Stream Power [KW/m2] CHAPTER VII CONCLUSION As part of this study on plunge pool scour, the plunge pool performance at four of B.C. Hydro dam sites was reviewed with respect to the spillway design, the plunge pool geology, and the spilling history. Conventional methods of assessment of maximum scour depth such as hydraulic model studies and empirical formulas were compared with a new approach called the Erodibility Index Method. The main limitations, conclusions, and recommendations of this study are discussed. 7.1 LIMITATIONS When analyzing prototype observations with respect to plunge pool scour, many factors are considered so that various uncertainties are introduced. In particular, the representative spillway discharge to be associated with a given episode of plunge pool development, the estimated progression of scour from successive sounding surveys, and the extrapolation of geological information to characterize the plunge pool bedrock. In general, the hydraulic records of spillway operation are available and reliable, but uncertainties arise from the selection of representative parameters to account for the observed scour. The quantitative analysis of plunge pool scour requires the identification of a prevailing discharge which formed the scour depth. Such discharge is ambiguous since the observed plunge pool scour configuration, at any given time, is the result of a unique combination of frequency and discharge magnitude/duration of all prior spills. For the quantitative part of this study on plunge pool scour, the observed maximum scour depth was associated with the preceding most important spill and the maximum daily average discharge (discharge averaged over a twenty-four hour period) of the spillway outflow hydrograph was chosen as representative of the scour depth. Although this approach is consistent with qualitative observations of scour rate on prototypes (Section 4.4), the scouring power of the maximum daily average discharge could be underestimated had the plunge pool equilibrium conditions not been reached. Such might be the case for the 1972 spill at Portage Mountain Project where the maximum daily average spill was an isolated peak caused by high flows spillway testing. The value of daily average discharge can also be deceptive when the gate operations are irregular. For instance, the maximum daily average spillway discharge of 167,000 cfs recorded at Portage Mountain represents in fact a sustained peak flow of 175,000 cfs for eleven hours preceded by thirteen hours of continuous spill at 160,000 cfs. For most spills considered, the maximum daily average discharge was representative of at least a few days of spilling near this magnitude. The data of maximum spillway discharges sustained for progressive periods of time in Tables 3.2, 3.5, 3.8, and 3.11 are a good indication of the fluctuations in spillway outflows. The other hydraulic parameters of importance, i.e. the reservoir and tailwater levels, were more constant with time and the mean elevation of the day was used with confidence. In a few instances when the tailwater elevation had not been recorded, the site tailwater rating curve (function of total outflows) was used. The selection of representative hydraulic parameters for the analysis of prototype scour development is a subjective task, but decisions were based on careful observations and consistency. For each plunge pool, the scour hole development was analyzed by means of sounding surveys which involve some uncertainties. Although the depth measurement from an echo sounder can be quite precise (within a foot), the plan positioning accuracy can be more challenging depending on the method used. The Global Positioning System technology was not available when the first surveys were carried out and is not always functional in deep canyon area. When a tag line with shore based control points is used, measuring errors of approximately ± 3-6 ft can be encountered. The existing plunge pool surveys for each site were often performed by different crews with different instrumentation and means of positioning. The signal recorded by the sonar beam is a reflection of the river bottom, without distinction between sound bedrock and overlying loose rock or soil. The amount of loose material trapped within the scour hole, if any, is unknown. Finally, the sounding survey coverage can be limited by turbulent waters because the excess of air bubbles interferes with the technology. Some limitations with regard to plunge pool surveys are characteristic to each site in this study: • The accuracy of the plunge pool surveys performed at Peace Canyon Dam is limited by the presence of loose material trapped inside the scour hole and redistributed according to the spillway operation. From the successive soundings, the plunge pool bottom elevation was not shown to progress but it is likely that the maximum scour depth to bedrock has. • At Seven Mile Dam, none of the sounding surveys performed has covered the main scour hole (downstream of the right chute) satisfactorily in order to capture the entire geometry of the depression and perhaps the lowest elevation. As for scour downstream of the spillway left chute, the thirteen years that separate the two available surveys prevent any evaluation of scour progression. • The first sounding survey of the Portage Mountain plunge pool was carried out almost nine months after the end of the initial spilling period so that infilling of the scour hole was enhanced. No pre-spill bathymetric survey of the plunge pool was undertaken at Portage Mountain and the available riverbed topography was obtained before the tailrace channel works. • For many of the plunge pool surveys at Revelstoke Dam, the original soundings record points were not available and the accuracy of the interpreted contours maps is unknown. The survey of 1991 following the most important spill on site was ineffective in the assessment of scour depth because the plunge pool was filled with overburden. For the monitoring of scour progression, sounding surveys of the plunge pool should be carried out following each important spill, and cover the entire area susceptible to scour. Since scouring is usually maximum in the early years of spillway operation, a reference pre-spill survey of the plunge pool is fundamental. Densification of the survey should allow a single and accurate interpretation of elevation contours. Ideally, a plan grid would be defined and followed for any subsequent survey. The use of suitable geophysical techniques to determine the depth to bedrock is necessary for a thorough assessment of plunge pool scour. Information on the geology and geomechanical characteristics of the plunge pool bedrock is limited. In general practice, engineering and geological investigations are confined to the location area of the structures to be erected. From preliminary design studies to dam safety investigations, the Peace Canyon dam foundation has been drilled and investigated extensively but the downstream riverbed subjected to the spillway jet impact was omitted. As such, the downstream extent of identified weakness features (the "hinge zone" and associated sub-channel) remained uncertain and the geological condition of the plunge pool bedrock was assumed to be that of the dam foundation. In the case of Seven Mile Dam, extrapolation from foundation drill core logs is problematic since the granite intrusion on which the dam is founded vanishes downstream of the flip buckets and the bulk of the plunge pool is composed of argillite. At Portage Mountain Project, the plunge pool geology was interpreted from the site bedrock stratigraphy and the nearest drill hole of the riverbed was located about 1,000 ft upstream of the spillway axis. Geological information on the Revelstoke plunge pool was obtained from post excavation inspection which revealed important weakness features overlooked by the few exploratory drill holes. The plunge pool bedrock is subject to intensive hydrodynamic action and the rock resistance to scour is dependent on its properties. The plunge pool area should be included in the geological investigations. This study on plunge pool scour was based on the best information available at the time. The limitations discussed above are common when it comes to investigate plunge pool scour on prototypes. Complete reviews of plunge pool performance with time and with respect to bedrock geology and spilling history are seldom done because accurate details of spillway discharge, head drop, tailwater depth, plunge pool geology, and scour depth are difficult to come by. This study is very valuable in that regard. 7.2 CONCLUSIONS This study involved a review of plunge pool performance at four dam sites of British Columbia with respect to site-specific factors such as spillway design, plunge pool geology, and spilling history. One of the objectives was to establish the relative influence of each factor on the plunge pool development (Chapter IV) . The main• conclusions are repeated here: • The spillway layout does affect the plunge pool scour development, mainly through the design unit discharge and total head drop for a given outflow and reservoir level. The energy losses and cushioning effect of air entrained on the spillway surface are negligible relative to the remaining energy that dissipates into the plunge pool. The spillway jet impingement angle induced by the design of the terminal structure seems to affect the depth of scour, whereas the effects of enhanced jet dispersion on the limitation of scour depth are questionable. • It is clear that the plunge pool scour configuration is affected by the prevailing geological conditions. Scouring tends to initiate in the weakness zones of the bedrock and can be limited by scour resistant rock. The low resistance of peripheral rock with respect to bottom rock directly impinged by the spillway jet is likely to cause problematic conditions of plunge pool development (unconfined conditions and/or asymmetry). The scour hole side slopes are generally influenced by the jointing pattern of the bedrock. • The magnitude of spillway flows seems to have a major impact on plunge pool scour compared to the frequency and duration of spills. Extended and/or repeated spills are not a condition to plunge pool development. The bulk of the scour hole is generally formed following the initial significant spill. In fact, the flow velocity (associated with the head drop from reservoir to bucket lip) might be predominant over the actual discharge in the early stages of plunge pool development. Scour progression corresponds to an increase in spillway flows. • Observations support the concept of equilibrium conditions and suggest that the maximum scour depth for a given discharge can be attained within days. • In any case, the plunge pool scour configuration results from the spillway jet erosive power and geometry coupled with the resistance and geological discontinuities of the bedrock. The main objective of this study was to evaluate the Erodibility Index Method for estimating plunge pool scour and see if there is an improvement in accuracy when compared to conventional methods such as empirical formulas and hydraulic model studies. The results of downstream scour tests from small-scale model studies were comparable to prototype observations in one of the four study cases (Section 5.1). The main difficulty in modelling plunge pool scour in laboratory is the appropriate scaling of bed material. The problem when using empirical equations to estimate the maximum scour depth is the variability in results for a single data set and the inconsistency of a given formula from one site to another. Ten well-known formulas were tested and the standard error of estimate varied from 16 ft for the Damle equation to 70 ft for the Veronese expression (Figures 5.6 to 5.15). Despite a tendency for underestimation, the Damle empirical formula was seen to give the best combination of precision and accuracy in the prediction of plunge pool scour depth at the four sites of study; accuracy of more than 60% should however not be expected. The performance of the Erodibility Index Method in the assessment of maximum scour depth was disappointing and did not outclass the conventional methods. Even supposing an accurate bedrock characterization and improved equations for the computation of jet velocity profile, the use of the Erodibility Index Method is questionable. The main weakness of the approach is that the vertical distribution of power available for scour in the plunge pool is essentially related to the submerged jet velocity profile which do not reflect the changing magnitude of spillway discharges. The erodibility threshold relationship, which forms the basis of the approach, was also derived from a majority of observations on auxiliary earth channel spillways. The overall performance of the Erodibility Index Method in the assessment of maximum scour depth for four distinct plunge pools was characterized by a generalized tendency for underestimation and a standard error of estimate of 53 ft. 7.3 RECOMMENDATIONS  AND EXPECTATIONS At this time, no new technology has been proven reliable for the assessment of plunge pool scour. The Erodibility Index Method, even with a refinement of equations to describe the jet hydraulics, is questionable. Current effort is directed towards the numerical analysis of hydraulic transients propagating through the discontinuities in the rock mass. Simoes and Vargas (2001) proposed a hydromechanic, coupled analysis, using the finite element method for the hydraulic problem, involving flow and wave propagation in the fractures, and discrete element method for the mechanical problem of block equilibrium. Bollaert (2002) (reported by Schleiss, 2002) based his work on fully transient water pressures in rock joints and developed a new model for the evaluation of the ultimate scour depth which includes the hydrodynamic fracturing of closed-end rock joints and dynamic uplift of rock blocks. Until these new approaches are tested independently on an extensive database of known prototype scour conditions, one must rely on the conventional methods of prediction. In the feasibility design stage of a project, the scour tests results from the hydraulic model studies should be used in conjunction with empirical equations. Considering the great variability offered by the existing equations, a careful selection should be based on the applicability of the formula (ski-jump spillway vs free overfall) and its performance in predicting scour depth in similar geological conditions. The engineering decision could be based on a statistical analysis of results obtained from the chosen empirical formulas, reinforced by the intuitive knowledge of hydraulic action in the plunge pool gained from the small-scale model studies. For an existing project with unforeseen plunge pool development, complementary hydraulic model studies in which the bed material is calibrated to the observed scour configuration are suggested to evaluate future scour progression in response to higher spillway flows. Even after a thorough review of plunge pool performance at Peace Canyon Dam, Seven Mile Dam, Portage Mountain Project, and Revelstoke Dam, and a careful analysis of different means of evaluation of maximum scour depth, the plunge pool response to the Inflow Design Flood (IDF) can only be assessed qualitatively: • Since initial spilling at Peace Canyon Dam (1979/1980), the plunge pool scour depth has not increased, although the largest spill on site occurred in 1996. The maximum spillway discharge was recorded in 1996 and reached 116,000 cfs, which represents 35% of the Inflow Design Flood. The Peace Canyon plunge pool was seen to develop laterally instead of in depth even with an increase in spillway flows. The scour configuration downstream of spillway Bay 3 (El. 1459 ft) should be representative of the expected scour development from the Inflow Design Flood since Gate 3 was fully open for one month in 1996. Scour downstream of spillway Bays 1 and 2 is likely to be deeper because of the steeper angle of the flip buckets (30°) . The maximum scour depth elevation corresponding to the Inflow Design Flood predicted by the Damle empirical formula (El. 1465 ft) is above the current elevation. • Since 1984, the major scour hole at Seven Mile Dam did not progress either in depth or extent despite the reach of new historical limits from the 1997 spill. The maximum spillway discharge recorded on site reached 116,000 cfs which represents 31% of the Inflow Design Flood. The 1997 survey of the Seven Mile plunge pool indicated minimum elevations of 1450 ft and 1503 ft downstream of the right and left chutes, respectively. The major scour hole is confined and effective in containing spillway flows from the spillway right chute, whereas the plunge pool configuration downstream of the spillway left chute is not fully developed to contain a full discharge from the three bays. The expected plunge pool bottom elevation to result from the Inflow Design Flood using the Damle equation (El. 1461 ft) is above the current low. The small-scale model studies suggest a minimum bedrock elevation around El. 1450 ft downstream of the spillway left chute. The Portage Mountain plunge pool was submitted to large spillway flows of 175,000 cfs (57% of the IDF) in the early years of operation (1972). The plunge pool has remained relatively stable since and the lowest elevation recorded from the 1996 soundings is El. 1516 ft. The plunge pool is deep (more than 100 ft) and confined and is therefore efficient in containing spillway flows and dissipating the excess of energy. The Damle empirical formula predicts a maximum scour depth at El. 1441 ft for the Inflow Design Flood conditions. The plunge pool at Revelstoke Dam is still immature since the spillway has been used with restraint over the years. The spillway discharge reached a peak of 70,000 cfs (28% of the IDF) for a few minutes as part of the 1986 spillway tests, after which the plunge pool was scoured down to El. 1380 ft. The Revelstoke plunge pool is known to respond rapidly to increases in spillway discharges and its performance is jeopardized by unconfined conditions. The maximum scour depth associated with the Inflow Design Flood is estimated at El. 1317 ft according to the Damle empirical formula. The hydraulic model studies included scour tests up to a spillway discharge of about half the Inflow Design Flood and the maximum equilibrium depth of scour was observed near El. 1310-1320 ft. SELECTED BIBLIOGRAPHY PAPERS Akhmedov, T.Kh. 1988. Calculation of the Depth of Scour in Rock Downstream of a Spillway. Water  Power  and Dam Construction 40 (December): 25-27. Annandale, G.W. 1994. Taking the Scour Out of Water Power. International Water  Power  and Dam Construction 46 (November): 46-49. Annandale, G.W. 1995. Erodibility. Journal  of Hydraulic  Research 33 (April): 471-493. Annandale, G.W., S.P. Smith, R. Nairns, and J.S. Jones. 1996. Scour Power. Civil  Engineering,  July, 58-60. Annandale, G.W., T.M. Lewis, R.J. Wittier, S.R. Abt, and J.F. Ruff. 1997. Dam Foundation Erosion: Numerical Modeling. In Energy  and Water: Sustainable Development: Proceedings  of the XXVII  IAHR  Congress  Held  in San Francisco,  California,  10-15 August 1997, Theme D, edited by F.M Holly and A. Alsaffar, theme editor J.S. Gulliver, 429-434. New York (NY): ASCE. Annandale, G.W., R.J. Wittier, J.F. Ruff and T.M. Lewis. 1998. Prototype Validation of Erodibility Index for Scour in Fractured Rock Media. Proceedings  of the 1998 ASCE  International Water  Resources Engineering Conference  Held  in Memphis,  Tennessee, 3-7 August 1998, 1096-1101. Annandale, G.W. 2002b. Quantification of the Relative Ability of Rock to Resist Scour. In Rock Scour Due to Falling  High-Velocity  Jets: Proceedings  of the International Workshop  on Rock Scour Held  in Lausanne, Switzerland,  September 25-28, edited by A.J. Schleiss and E. Bollaert, 201-212. Lisse (Switzerland): Swets and Zeitlinger. Annandale, G.W. 2002c. Quantification of Extent of Scour Using the Erodibility Index Method. In Rock Scour Due to Falling  High-Velocity Jets:  Proceedings  of the International Workshop  on Rock Scour Held  in Lausanne, Switzerland,  September 25-28, edited by A.J. Schleiss and E. Bollaert, 63-72. Lisse (Switzerland): Swets and Zeitlinger. Bin, A.K. 1984. Air Entrainment by Plunging Liquid Jet. In IAHR  Symposium on Scale Effects  in Modeling  Hydraulic  Structures Held  in Esslingen, West  Germany,  September 3-6, edited by H. Kobus, 5.5-1-5.5-6. Bohrer, J.G., S.R. Abt, and R.J. Wittier. 1998. Predicting Plunge Pool Velocity Decay of Free Falling, Rectangular Jet. Journal  of Hydraulic Engineering  124 (October): 1043-1048. Bollaert, E. 2002. Transient Water Pressures in Joints and Formation of Rock Scour Due to High-Velocity Jet Impact. Communication No.  13 of the Laboratory of Hydraulic  Constructions (LHC),  EPFL, Lausanne. Breusers, H.N.C., and A.J. Raudkivi. 1991. Scouring. IAHR Hydraulic Structures Design Manual 2. Rotterdam (Netherlands): A.A. Balkema. Chaudhry, M.H. 1993. Open-Channel Flow.  New Jersey (NY): Prentice-Hall. Chian Min Wu. 1973. Scour at Downstream End of Dams in Taiwan. Proceedings of the IAHR  Symposium on River Mechanics  Held  in Bangkok, Thailand, paper A13. Da Cunha, L.V., and Lencastre, A.C. 1966. La dissipation de l'energie dans un evacuateur en saut de ski. Observation de 1'erosion. Laboratorio Nacional de Engenharia Civil, Lisboa, Portugal. Publication No. 288. Damle, P.M. et al. 1966. Evaluation of Scour Below Ski-Jump Buckets of Spillways. Proceedings  of the Golden Jubilee  Symposium Held  in Poona, India, 154-163. Davies, T.W., and M.K. Jackson. 1982. Optimum Conditions for the Hydraulic Mining of China Clay. Proceedings  of the First  U.S. Water  Jet Conference  Held  in Golden, Colorado,  7-9 April 1981, 129-146. Golden (CO): Colorado School of Mines Press. Davies, T.W., and M.K. Jackson, n.d. Nozzle  Design for  Coherent Water  Jet Production.  Exeter (U.K.): Exeter University, Department of Chemical Engineering. Elevatorski E.A. 1959. Hydraulic  Energy  Dissipators. New York (NY): McGraw-Hill Book Company, Inc. Ervine, D.A., E. McKeogh, and E.M. Elsawy. 1980. Effect of Turbulence Intensity on the Rate of Air Entrainment by Plunging Water Jets. Proceedings  of the Institution of Civil  Engineers, Part 2, 69 (June): 425-445. Ervine, D.A., and H.T. Falvey. 1987. Behaviour of Turbulent Water Jets in the Atmosphere and in Plunge Pools. Proceedings  of the Institution of Civil  Engineers, Part 2, 83 (March): 295-314. Ervine, D.A., H.T. Falvey, and W. Withers. 1997. Pressure Fluctuations on Plunge Pool Floors. Journal  of Hydraulic  Research 35 (February): 257-279. Fiorotto, V. and A. Rinaldo. 1992. Turbulent Pressure Fluctuations Under Hydraulic Jumps. Journal  of Hydraulic  Research 30 (April): 499-520. Galvagno, G. 1998. A Study of Scour Below Ruskin Dam Spillway Using a Non-Cohesive Bed Hydraulic Model. M.A.Sc. Thesis, University of British Columbia. George, R.L. 1980. Impinging  Jets.  Denver (CO): Engineering and Research Center, Water and Power Resources Service. REC-ERC-80-8. Golder Associates. 2001. GoldDFE 1.0 - User Manual. Hager, W. H. 1998. Forum Article - Plunge Pool Scour: Early History and Hydraulicians. Journal  of Hydraulic  Engineering  124 (December): 1185-1187 . Hamilton, K.J., S.R. Abt, R.J. Wittier, and G.W. Annandale. 1997. Erosion at Dam Foundations: Plunge Pool Circulation. Proceedings  of the Association of State Dam Safety  Officials  Annual Conference  Held  in Pittsburgh,  Pennsylvania,  7-10 September 1997, 693-702. Hartung, F. 1959. Die kolkbildung hinter iiberstromten wehren im hinblick auf eine bewegliche sturzbettgestaltung. Die Wasserwirtshaft  49(1): 309-313. In German. Hartung, F., and E. Hausler. 1973. Scours, Stilling Basins and Downstream Protection Under Free Overfall Jets at Dams. Transactions of the 11th International Congress  on Large Dams Held  in Madrid,  Spain, 11-15 June 1973, 39-56. Henderson, F.M. 1966. Open Channel Flow.  New York: Macmillan. Jaeger, C. 1939. Uber die Ahnlichkeit bei flussbaulichen Modellversuchen [On Similarity for River Engineering Experiments]. Wasserkraft  und Wasserwirtschaft  34(23/24): 269-270. In German. Johnson, R.E and S. Alam. 1969. Spillway - Part 2. Engineering  Journal, October, 82-87. Johnson, P.L. 1974. Hydraulic  Model  Studies of Plunge  Basins for  Jet  Flow. Denver (CO) : Engineering and Research Center, Bureau of Reclamation. REC-ERC-74-9. Kirsten, H.A.D. 1982. A Classification System for Excavation in Natural Materials. The Civil  Engineer in South Africa,  July, 293-308. Lewis, T.L., S.R. Abt, J.F. Ruff, R.J. Wittier, and G.W. Annandale. 1996. Erosion at Dam Foundations: Predicting Jet Velocities. Proceedings  of the Association of State Dam Safety  Officials  Annual Conference  Held  in Seattle, Washington,  8-11 September 1996, 437-446. Liu, P.Q., J.R. Dong, and C. Yu. 1998. Experimental Investigation of Fluctuation Uplift on Rock Blocks at the Bottom of the Scour Pool Downstream of Three-Gorges Spillway. Journal  of Hydraulic  Research 36 (January): 55-68. Lowe III, J., P.C. Chao, and A.R. Luecker. 1979. Tarbela Service Spillway Plunge Pool Development. Water  Power  and Dam Construction 31 (November): 85-90. Manitoba Hydro. 1970. Stabilization of Spillway Scour Hole at Grand Rapids, Manitoba. Presented  to the Hydraulic  Power  Section - Canadian Electrical Association in Winnipeg,  Canada, September 1970. Martins, R.B.F. 1975. Scouring of Rocky River Beds by Free Jet Spillways. Water  Power  and Dam Construction 27 (April): 152-153. Mason, P.J. 1984. Erosion of Plunge Pools Downstream of Dams Due to the Action of Free-Trajectory Jets. Proceedings  of the Institution of Civil Engineers, Part I, 76 (May): 523-537. Mason, P.J. 1985. Discussion on "Erosion of Plunge Pools Downstream of Dams Due to the Action of Free-Trajectory Jets" by P.J. Mason, 1984. Proceedings  of the Institution of Civil  Engineers, Part I, 78 (August): 991-999. Mason, P.J., and K. Arumugam. 1985. Free Jet Scour Below Dams and Flip Buckets. Journal  of Hydraulic  Engineering  111 (February): 220-235. Mason, P.J. 198 9. Effects of Air Entrainment on Plunge Pool Scour. Journal of Hydraulic  Engineering  115 (March): 385-399. Mason, P.J. 1989. Closure of Discussion on "Effects of Air Entrainment on Plunge Pool Scour" by P.J. Mason, 1989. Journal  of Hydraulic Engineering:  262-265. Mason, P.J. 1993. Practical Guidelines for the Design of Flip Buckets and Plunge Pools. Water  Power  and Daw Construction 45 (September/October): 40-45. Novak, 1985. Discussion on "Erosion of Plunge Pools Downstream of Dams Due to the Action of Free-Trajectory Jets" by P.J. Mason, 1984. Proceedings of the Institution of Civil  Engineers, Part I, 78 (August): 991-999. Otto, B. 1986. Study of Scour Potential at Burdekin Falls Dam Due to Rock Stresses. Queensland Water Resources Commission. Prochukhan, D.P., S.A Fried, and L.K. Domansky. 1971. The Rock Foundation of Hydraulic Structures. Stroiizdat, Moscow. Quintela, A.C., J.S. Fernandes, and A.A. Da Cruz. 1979. Barrage de Cahora-Bassa. Problemes poses par le passage des crues pendant et apres la construction. Transactions du Treizieme Congres  des Grands  Barrages tenu a New  Delhi, India, 713-730. Ramos, C.M. 1982. Energy Dissipation in Free Jet Spillways. Transaction of the International Symposium on the Layout of Dams in Narrow  Gorges, I.C.O.L.D.,  Brazil. Schleiss, A.J. 2 002. Scour Evaluation in Space and Time - The Challenge of Dam Designers. In Rock Scour Due to Falling  High-Velocity  Jets: Proceedings  of the International Workshop  on Rock Scour in Lausanne, Switzerland,  September 25-28, edited by A.J. Schleiss and E. Bollaert, 3-22. Lisse (Switzerland): Swets and Zeitlinger. Shixia W. 1987. Scouring of Riverbeds below Sluices and Dams. In Design of Hydraulic  Structures: Proceedings  of the International Symposium on Design of Hydraulic  Structures Held  at Colorado  State University  August 24-27 1987, edited by A.R. Kia and M.L. Albertson, 295-304. Simoes, G.F., and E.do A. Vargas JR. 2001. Analysis of Erosion Processes Downstream of Spillways in Large Dams. Proceedings  of the 38th U.S. Rock Mechanics  Symposium Held  in Washington,  D.C., 7-10 July  2001, 959-966. Spurr, K.J.W. 1985. Energy Approach to Estimating Scour Downstream of a Large Dam. Water  Power  and Dam Construction 37 (July): 81-89. Sutcliff H. 1985. Discussion of "Erosion of Plunge Pools Downstream of Dams Due to the Action of Free-Trajectory Jets", by P.J. Mason, 1984. Proceedings  of the Institution of Civil  Engineers, Part I, 78 (August): 991-999. Taraimovich, I.I. 1978. Deformations of Channels below High-Head Spillways on Rock Foundations. Hydrotechnical  Construction No. 9 (September): 917-923 . U.S. Bureau of Reclamation. 1973. Design of Small Dams, 2nd Edition. Denver: Water Resources Technical Publication. Veronese, A. 1937. Erosioni di fondo a valle di un scarico [Bottom Erosions Downstream of a Dam] . Annali dei Lavori Pubblici 75(9) : 717-726. In Italian. Whittaker, J.G., and A. Schleiss. 1984. Scour Related to Energy Dissipators for  High  Head  Structures. Zurich (Switzerland): VAW Laboratory of Hydraulics, Hydrology, and Glaciology/Federal Institute of Technology ETH. Mitt. Nr. 73. VAW/ETH. Wittier, R.J., J.F. Ruff, S.R. Abt, G.W. Annandale, B.W. Mefford, K. Adhya, and D. Morris. 1995. Dam Foundation  Erosion: 1994 Year  End Summary Report. Denver (CO) : US Bureau of Reclamation Water Resources Research Laboratory. Wittier, R.J., B.W. Mefford, S.R. Abt, J.F. Ruff, and G.W. Annandale. 1995. Spillway and Foundation Erosion: Estimating Progressive Erosion Extents. Proceedings  of Waterpower'95  Held  in San Francisco, California,  25-28 July  1995, 1706-1714. Wittier R.J., G.W. Annandale, S.R. Abt, and J.F. Ruff. 1998. New Technology for Estimating Plunge Pool or Spillway Scour. Proceedings  of Association of St-citZG D3.ni S3.fs ty Off  i. ci 3 2. s t D3.n\ Ss.f s t y' 98, Hq  2. c? J.11  L3.S Vegas,  Nevada,  11-15 October 1998, 755-766. Wittier R.J., G.W. Annandale, J.F. Ruff, and S.R. Abt. 1998. Prototype Validation of Erodibility Index for Scour in Granular Media. Proceedings  of the 1998 International Water  Resources Engineering Conference  Held  in Memphis,  Tennessee, 3-7 August 1998, 1090-1095. Yang, C.T. and A. Molinas. 1982. Sediment Transport and Unit Stream Power Function. Journal  of the Hydraulics  Division, ASCE,  108, HY.6: 774-793. Yildiz D., and E. Uziicek. 1994. Prediction of Scour Depth from Free Falling Flip Bucket Jets. International Water  Power  and Dam Construction 46 (November): 50-56. BC HYDRO REPORTS Peace Canyon Dam BCH Report No. 354: Report of Laboratory Testing on Rock Core Samples / Peace River Project (by Ripley and Associates). June 1959. BCH Report No. H715: Peace River Site 1 Development / Detail Spillway Hydraulic Model Studies (by Lasalle Hydraulic Laboratory Ltd.). December 1970. BCH Report No. 822 - Appendix A: Site 1 Dam / Engineering Geology of Foundation Bedrock of Dam-Powerhouse-Spillway (Phase A) (by Dolmage Campbell & Associates Ltd.). December 1976. BCH Report No. 966: Peace River / Site 1 Project / Gravity Dam and Spillway Foundation Report. December 1978. BCH Report No. HY296: Peace Canyon Project / Review of Spillway Scour Potential. February 1985. BCH Report No. GEO 9/85: Peace Canyon Project / Memorandum on Plunge Pool Development. July 1985. BCH Report No. H1879: Peace Canyon Project / Deficiency Investigations / Spillway Scour / Status March 1986. November 1986. BCH Report No. H1742: Peace Canyon Dam / Design Report. December 1987. BCH Report No. H2 003: Williston Lake / Probable Maximum Flood. June 1988. BCH Report No. H2767: Peace Canyon Dam / Comprehensive Inspection and Review 1993. November 1994. BCH File No. C-PCN-1206.12: Peace Canyon Plunge Pool, Spillway and Tailrace Channel Inspection. June 1997. BCH Report No. OMSPCN/03: Peace Canyon Dam / Operation, Maintenance and Surveillance Manual for Dam Safety. February 2001. Seven Mile Dam BCH Report No. 750: Pend d'Oreille River / Seven Mile Project / Summary of Information for Advisory Board Meeting No.1. June 1975. BCH Report No. N68: Addendum to Final Report / Detailed Comprehensive Hydraulic Model Studies of Seven Mile Project / Pend d'Oreille River (by Western Canada Hydraulic Laboratory Ltd). November 1976. BCH Report No. H1743: Seven Mile Dam / Design Report. August 1988. BCH Report No. H2062: Seven Mile Dam / Deficiency Investigations. March 1990 . BCH System Operating Order No.4P-36: Seven Mile Project. May 1997. BCH Report No. MEP507: Seven Mile Dam / Deficiency Investigations / Spillway Adequacy for Extreme Floods (by Klohn-Crippen Integ and NHC) . March 1999. BCH Report No. PSE362: Seven Mile Dam / Dam Safety Improvements / Auxiliary Spillway / 1999-2000 Investigations Report. May 2001. BCH Report No. PSE401: Seven Mile Dam / Dam Safety Improvements / Working Dam Foundation Summary (Draft). October 2001. BCH Report No. N1926: Seven Mile Dam / Plunge Pool Scour Report. March 2002 . Portage Mountain Project BCH Report No. 150: First Report of Laboratory Testing on Soil and Rock Samples / Peace River Project (by Ripley and Associates). June 1958. BCH Report No. 31: Report of Laboratory Testing on Rock Core Samples / Peace River Project (by Ripley and Associates). July 1959. B.C. and B.B. Power Consultants Limited. Report on the Peace River Hydro-Electric Project Vol. II - Appendix A2.1: Geological Report #3 / Portage Mountain Site (by D. Campbell). June 1959. B.C. and B.B. Power Consultants Limited. Report on the Peace River Hydro-Electric Project Vol. II - Appendix A2.3: Geology of Portage Mountain Dam Site (by V. Dolmage). June 1959. BCH File No. 003805546: Laboratory Testing of Rock Cores / Portage Mountain Project (by Coast Eldridge). June 1965. IPEC Report No. H692: Portage Mountain Development / W.A.C. Bennett Dam / 1972 Spillway Tests. July 1973. BCH Report No. H1756: Portage Mountain Development / Design Report. October 1988. BCH Report No. H2417A: W.A.C. Bennett Dam / Comprehensive Inspection and Review 1990 / Hydrotechnical. October 1992. BCH Report No. OMSGMS/03: W.A.C. Bennett Dam / Operation, Maintenance and Surveillance Manuel for Dam Safety. January 2001. Revelstoke BCH Report No. 664: High Revelstoke Versus Downie - Low Revelstoke / Feasibility Study. January 1973. BCH Report No. 746: Revelstoke Project / Spillway Design Flood. March 1975 . BCH Report No. 786: Revelstoke Project Preliminary Design Study / Foundation Investigations. April 1976. BCH File No.15-2-57: 1976 Testing Program / Revelstoke Project (Rock) / Laboratory Test Results - Vol II (by Thurber Consultants Ltd.). 1976 BCH Report No. H1204: Memorandum on In Situ Rock Investigations 1978-1979. March 1980. BCH Report No. H1624: Summary of Information for Advisory Board Meeting No. 11. June 1983. BCH Report No. 2278: Report on Revelstoke Project / Spillway Plunge Pool Model Study (by Western Canada Hydraulic Laboratories Ltd.). August 1983 . BCH Report No. H1907: Revelstoke Project / Spillway Freeboard Tests. October 1986. BCH Report No. H1864: Revelstoke Project / Design Report. February 1988. BCH Report No. HYD.943: A Fact Finding Report on the Failure of Revelstoke Powerhouse Access Road During the August 1991 Flood - Draft. November 1991. BCH Report No. N1315: Report on Overwater Acoustic Profiling and Seismic Refraction Survey / Revelstoke Dam - Access Road Replacement Project. January 1992. BCH Local Operating Order No. 3P03-47: REV - Outlet Works and Spillway Operation (Non Power Water Discharges). March 2000. Appendix I Erodibility Index Charts ERODIBILITY INDEX CHARTS Erodibility Index. K K  = Ms-  Kb  • Kd  • Js  where Kb  = RQD/Jn Kd  = JrjJa T h e Rock Qual i ty Des ignat ion ( R Q D ) is a s tandard parameter in drill co re logging and is ca lcu la ted a s the ratio (in percent) of the cumulat ive length of all p ieces of co re greater than 10 c m to the total length of the core run. T h e other parameters are def ined in the fol lowing tab les from Annanda l e (1995). Tab le 1. M a s s Strength Numbe r for Rock s (Ms) Hardness Identif ication  in Profile Unconf ined Compressive Strength (MPa) Mass Strength Number (Ms) Very soft rock Material crumbles under firm (moderate) blows with sharp end of geological pick and can be peeled off  with a knife;  is too hard to cut triaxial sample by hand. < 1.7 1.7-3.3 0.87 1.86 Soft rock Can just be scraped and peeled with a knife;  indentations 1 mm to 3 mm show in the specimen with firm (moderate) blows of the pick point. I 3.3-6.6 6.6-13.2 3.95 8.39 Hard rock Cannot be scraped or peeled with a knife;  hand-held j specimen can be broken with hammer end of geological I pick with a single firm (moderate) blow. I 13.2-26.4 r 17.70 Very hard rock Hand-held specimen breaks with hammer end of pick under j more than one blow. 26.4-53.0 53.0-106.0 35.0 70.0 Extremely hard rock Specimen requires many blows with geological pick to break through intact material. >212.0 280.0 Note. The value of Ms for rock can be determined by equating it to the unconfined compressive strength in MPa if the strength is greater than 10 MPa, and equal to 0.78*UCS105 when the strength is less than 10 MPa. Tab l e 2. M a s s Strength Numbe r for Granu la r So i l s (Ms) Consistency Identif ication  in profile SPT Blow Count Mass Strength Number (Ms) ^ e r y loose Crumbles very easily when scraped with geological pick. 0-4 0.02 Loose Small resistance to penetration by sharp end of geological pick. 4-10 0.04 Medium dense Considerable resistance to penetration by sharp end of geological pick. 10-30 0.09 Dense Very high resistance to penetration by sharp end of geological pick - requires many blows of pick for excavation. 30-50 0.19 Very dense High resistance to repeated blows of geological pick -requires power tools for excavation. 50-80 0.41 Note. Granular materials in which the SPT blow count exceeds 80 to be taken as rock - see Table 1. Table 3. Joint Set Number (Jn) Number of Joint Sets Joint Set Number (Jn) Intact, no or few joints/fissures 1.00 One joint/fissure set 1,22 One joint/fissure set plus random 1.50 Two joint/fissure sets 1.83 Two joint/fissure sets plus random 2.24 Three joint/fissure sets 2.73 Three joint/fissure sets plus random i 3.34 Four joint/fissure sets j 4.09 Multiple joint/fissure sets ! 5.00 Note. For intact granular materials take Jn = 5.00. Table 4. Relative Ground Structure Number (Js) Dip direct ion of closer spaced jo in t set O Dip angle of closer spaced jo in t set 0 Ratio of Joint Spacing, r 1:1 1:2 1:4 1:8 180 /0 90 1.14 [ ]  20 1.24 | "L26 In direction of 89 0.78 0.71 0.65 0.61 stream flow 85 0.73 0.66 0.61 0.57 80 0.67 0.60 0.55 0.52 70 0.56 0.50 0.46 0.43 60 0.50 0.46 0.42 0.40 50 0.49 0.46 0.43 0.41 40 0.53 0.49 0.46 0.45 30 0.63 0.59 0.55 0.53 20 0.84 0.77 0.71 0.67 10 1.25 1.10 0.98 0.90 5 1.39 1.23 1.09 1.01 1 1.50 1.33 1.19 1.10 0 / 1 80 0 1.14 I 1.09 1.05 1.02 Against 1 0.78 0.85 0.90 0.94 direction of 5 0.73 0.79 0.84 0.88 stream flow 10 0.67 0.72 0.78 0.81 20 0.56 0.62 0.66 0.69 30 0.50 0.55 0.58 0.60 40 0.49 0.52 0.55 0.57 50 0.53 0.56 0.59 0.61 60 0.63 0.68 0.71 0.73 70 0.84 0.91 0.97 1.01 80 1.26 1.41 1.53 1.61 85 1.39 1.55 1.69 1.77 89 1.50 1.68 1.82 1.91 180 /0 90 1.14 1.20 | 1.24 Notes. 1. For intact material take Js = 1. 2. For values of r greater than 8 take Js as for r = 8. Table 5. Joint Roughness Number (Jr) Joint Separation Condit ion of Joint Joint Roughness Number (Jr) Joints/fissures tight or closing during excavation Discontinuous joints/fissures 4.0 Rough or irregular, undulating 3.0 Smooth undulating 2.0 | Slickensided undulating 1.5 I Rough or irregular, planar 1.5 | Smooth planar 1.0 i Slickensided planar 0.5 Joints/fissures open and remain open during excavation Joints/fissures either open or containing relatively soft gouge or sufficient  thickness to prevent joint/fissure wall contact upon excavation 1.0 Shattered or micro-shattered clays 1.0 Notes. 1. For intact granular material take Jr = 3.0. 2. For granular materials, the quotient Jr/Ja crudely approximates tan (<f>r), where 4>r is the equivalent residual friction angle. Table 6. Joint Alteration Number (Ja) Description of Gouge Joint Alteration Number (Ja) for  Joint Separation (mm) <1.01 1 . 0 - 5 . 0 2 > 5.03 Tightly healed, hard, non-softening impermeable filling 0.75 — — Unaltered joint walls, surface staining only 1.0 — — Slightly altered, non-softening, non-cohesive rock mineral or crushed rock filling 2.0 2.0 4.0 Non-softening, slightly clayey non-cohesive filling 3.0 r 6.o 10.0 Non-softening, strongly over-consolidated clay mineral filling, with or without crushed rock 3.0 6.04 10.0 Softening or low friction clay mineral coatings and small quantities of swelling clays 4.0 8.0 13.0 Softening moderately over-consolidated clay mineral filling, with or without crushed rock 4.0 8.04 13.0 Shattered or micro-shattered (swelling) clay gouge, with or without crushed rock 5.0 10.04 18.0 Notes. 1. Joint walls effectively  in contact. 2. Joint walls come into contact after approximately 100 mm of shear. 3. Joint walls do not come into contact at all upon shear. 4. Also applies when crushed rock occurs in clay gouge without rock wall contact. 5. For granular materials, the quotient Jr/Ja crudely approximates tan (4»r), where <j>r is the equivalent residual friction angle. Appendix II Plunge Pool Surveys List of Drawings Attached Peace Canyon 1007-C14-X8423 1007-C14-U8440 1007-C14-D44 71 1007-C14-D44 72 1007-C14-D4 54 4 Unlabeled 1007-C11-D1983 Dam Soundings Taken in Plunge Pool Area - April 15/80 Soundings Taken in Plunge Pool Area - Sept 5/81 Plunge Pool Elevations & Contours Before 1983 Spill -April 1983 Plunge Pool Elevations & Contours After 1983 Spill-Oct 1983 Depth Sounding and Video Camera Survey of Plunge Pool Oct 7-9, 1985 Depth Sounding Survey of Plunge Pool - July 22, 1987 Plunge Pool Bathymetric Survey After Spilling on August 4 1996 Seven Mile Dam 224-C11-E7024 224-C11-D192/ D7025 Unlabeled 224-Cll-D117 224-Cll-D119 224-C22-D2 -C11-D116 Unlabeled 224-C11-U217 Spillway Plunge Pool / Rock Contours Before Spillway Operation (Survey taken before 30 Oct 1979) Spillway Plunge Pool / Elevations After Spilling to Dec 14, 1979 Survey of Tailrace Channel on 7 Aug. 1980 Tailrace Soundings Survey /August 1982 Tailrace Soundings Survey / September 2 0th, 1984 Tailrace Soundings Survey / October 15th, 1986 Tailrace Soundings - Oct. 1988 / Plan 3D Topography / River Soundings - Surveyed Oct 1997 / Plan Portage Mountain Project 1006-C14-B1262 GMS Spillway Scour Hole After First Spill Event 13 June 1972 to 3 September 1972 - Underwater Contour Plan 10 06-C11-D1107 Spillway - Plunge Pool Bathymetric Survey - 1996 212-C21-D7126 Spillway Plunge Pool / As Excavated by CR-6 and CR-4 Prior to Use of Spillway 212-C14-B4371 Section Along Centreline of Spillway Plunge Pool 212-C21-D7125 Spillway Plunge Pool / Sounding Survey Taken 15 May 1984 Following Initial Spillway Operation 212-C21-D134 Survey of River Channel Immediately Downstream of Plunge Pool Following Spill Tests 11 to 14 August 1986 212-C14-C5579 Revelstoke Plunge Pool / Bathymetry Data / September 22nd 1991 212-C14-C5580 Revelstoke Plunge Pool / Overburden Isopach Plan N to o (_n / X TOE OF RANiOOM ROCK.FIU-DAAWI\G Nc. '.V4S B2: A/or*: OCT  30j~19 n 1 -.1™ 8KT-SH COLUMSIA aOW UH'POWp'AU7H0WT —•! . • S U C H Mitt ' .'.S Pi wb&Ar. Pt-UOKlt^Fioi. ' . ;, JJ ' • ' OitiiimM. •• • •• -j — i 1 •Mii'-'-j: - f c r - i — : I - . - 224 —CI  1 - E 7 0 2 4 J I I L 1200 -1100 -1000 -900 800 V) CD o' -700 - 6 0 0 -500 2100 Line [ft] SEVEN MILE DAM SURVEY OF TAILRACE CHANNEL ON 7 AUG. 1980 Reference Drawing No. 224-C17-X7016 1000 1200 900 800 cn Q) i—t-o" 13 700 600 500 1700 Line [ft] 2200 SEVEN MILE DAM TAILRACE SOUNDINGS - OCT. 1988 PLAN Reference Drawing Nos. 224-C22-D6 to D9 Very  shallow, rocks and boulders above Water  Lint t TOO 1630 % 1600 a I* 13 SO ISOO Reference  line mi SECTION  LA NOTES /.  Sounding  undertaken  on  Hay  IS  and  Uay  16,  1973 by  Burnett  Resource  Surveys  Ltd. (  Owg  No.  IOOS  -  C26-  D!  ) 2.  Dotted  contours  taken  from  IPEC Dwg  No.  U-3040-II35  (June  1969  ) Stele too too 200 300 reel KL0HN-CR1PPEN 1NTEQ BRITISH COLUMBIA HYDRO AND POWER AUTHORITY GMS TAILRACE WEIR AND DOWNSTREAM RIVER CHANNEL IMPROVEMENT - RESOURCE SMART GMS SPILLWAY SCOUR HOLE AFTER FIRST n SPILL EVENT 13 JUNE 1972 TO 3 SEPTEMBER 1972 - UNDERWATER CONTOUR PLAN OATE FEB 1993 D. Y. /HCP FIG. 2 - 3 OWO Ma I 0 0 6 - C I 4 - B I 2 6 2 REPORT No. KCt 139 (H2709! _ 215 E 48600.000 + E 48700.000 + Stc 33+43. 42 E 48800.000 E 48900.000 + E 49000. 000 + E 49100.000 N 1800. 000 0*00 r 8 -)-N 1700.000 A I P t 6 M B E R M -j- N 1600. 000 TMs Survey was done by Survey & Photogrammetry on 4 Aug 1996. Contour Interval 2.0m. Water Elevation 500.2m (approx). Coordinates are on tne Dam Construction Datum. Elevations are Geoaetic. + N 1500.000 I *«JNCTI0NAI CvCE 1 2. 07 C«*»N C. 0. B. PPOJECT NO S2 BChydro SLRvEr anO PhOTOCRaumCTRi CEPaATwCnT CMCCICO INSPECTED G. M. SHR'JM GENERATING STATION Spillway - Piange Pool Ba+hyme+ric Survey - 1996 l*SPeCT£0 :.NSP£CTE0 inspected NTS fi£?. 93016 OAIE 5 C ' t £ i : 2 o o o " 1 0 0 6 - C t l - D i l O T i" DCS nico BRITISH COLUMBIA HYDRO AND POWER AUTHORITY CCA MW (TtCK) C O L U M B I A R I V E R - R E V E L S T O K E  P R O J E C T SPILLWAY PLUNGE POOL SOUNDING SURVEY TAKEN 15 MAY 1984 'FOLLOWING INITIAL SPILLWAY OPERATION CMKD WfO I I 1 I I 1 1 1 1 1 IMP AWO MO OATI 18 MAY 1984 m v i t i O N t •McnoriLMce jmno. °°'SS*CT acALi I ' - 2 0 ' M O W 212- C2I - 07125 R 0 aid 1443-6 PLAN Sco/ei/'-  too' Horixonto/  sco/e /"-/OO Vertico/  SCO /a /'-SO' -| /SOO £ -]  /480  k 1460  $ ^ I4Z0  ^ 1 S 1400  ki -I  /3SO  ^ 1, JOB SSCZIQ1S THB00G8 EIP SAP FS0TBCT10B SXB DUG. 212—C2V-D133 REFERENCE PRAWlNO» MMCI PRAWINOR REMARKS I. iff. Contours redrawn Section A notes chongec/ DA I /8 \DAJ\ A 9Y  [ cw. \rrml  pate BRITISH COLUMBIA HYDRO aad POWER AUTHORITY APPROVALS B . C . H Y D R O - C I V I L I N S P E C T I O N S E C T I O N OKS. DATE DATE R E V E L S T O K E G. S. SURVEY OF RIVER CHANNEL IMMEDIATELY DOWNSTREAM OF PLUNGE POOL FOLLOWING SPILL TESTS II TO 14 AUGUST 1986 DM Oft. DAJ FCB'tn DR. CM. A! i=t»'<7 SCALC: MICROFILM ID . M . » T _ . , _ | • " •« • » 2 1 2 - C 2 I — D 1 3 4 |"2 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0063412/manifest

Comment

Related Items