UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Saline water intrusion from the Fraser River estuary : a hydrogeological investigation using field chemical… Neilson-Welch, Laurie A. 1999

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

Item Metadata

Download

Media
831-ubc_1999-0130.pdf [ 9.53MB ]
Metadata
JSON: 831-1.0053150.json
JSON-LD: 831-1.0053150-ld.json
RDF/XML (Pretty): 831-1.0053150-rdf.xml
RDF/JSON: 831-1.0053150-rdf.json
Turtle: 831-1.0053150-turtle.txt
N-Triples: 831-1.0053150-rdf-ntriples.txt
Original Record: 831-1.0053150-source.json
Full Text
831-1.0053150-fulltext.txt
Citation
831-1.0053150.ris

Full Text

SALINE WATER INTRUSION FROM THE FRASER RIVER ESTUARY. A HYDROGEOLOGICAL INVESTIGATION USING FIELD CHEMICAL DATA AND A DENSITY-DEPENDENT GROUNDWATER FLOW MODEL by Laurie A. Neilson-Welch A thesis submitted in partial fulfilment of the requirements for the degree of Master of Science i n The Faculty of Graduate Studies Department of Earth and Ocean Sciences We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA 1999 © Laurie A. Neilson-Welch, 1999 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Earth dud Ocean £c fences The University of British Columbia Vancouver, Canada Date Jfaul /5~jm 11 ABSTRACT Field data and a density-dependent groundwater flow model were used to understand the processes affecting saltwater intrusion and saline water distribution at a field site located adjacent to the Fraser River Estuary in Richmond, BC. At the field site saline water from the Fraser River intrudes a confined sand aquifer forming a saline wedge which extends laterally approximately 500 m into the aquifer. Chemical analysis of groundwater samples from three multi-level sampling wells (installed to obtain groundwater samples at closely spaced depth intervals), zone-specific piezometers, a West Bay well, and cone penetrometer tests was completed. The groundwater chemical data were used to delineate salinity contours defining the saline wedge. The chemical data were also analyzed using trilinear diagrams. Four chemically different groundwater facies were identified; shallow fresh groundwater above the saline wedge, brackish groundwater at the transition zone, saline water within the core of the saline wedge, and deep fresh groundwater beneath the saline wedge. Water level data from the zone-specific piezometers at the site were analyzed using a tidal filtering technique to determine average groundwater levels. The average water levels were then converted to equivalent freshwater head and compared for wells screened at the same elevation. The results indicated non-uniform flow conditions within the aquifer. Field data were not sufficient to thoroughly define the flow system. The numerical density-dependent groundwater flow and solute transport model Fracdens (Shikaze et al., 1996) was used simulate site conditions. The model was calibrated to field salinity data and a good match was obtained. The results indicated that the average groundwater flow regime at the site is typical for an area influenced by saltwater intrusion (circulation of dense saline water into the aquifer and then flow of saline, fresh, and mixed water toward the river under a regional freshwater gradient). Calibration of the model indicated that a shallow gradient toward the river exists and that a lower permeability zone Ul (lower than that initially estimated based on available field data) may exist in the area of the saline wedge toe. A simulation to investigate possible effects of seasonal changes in river discharge rate on the river-side model boundary conditions indicated that the groundwater flow regime at the site likely varies due to transient conditions at the river. Seasonal influences caused the dilution of the groundwater salinity near the river-side model boundary due to the migration of fresh river water into the aquifer during high river discharge conditions. The regional groundwater flow regime for the Fraser River delta, developed by Ricketts (1998), was modified to account for the effects of saltwater intrusion from the Fraser River and Georgia Strait. A review of available geochemical data and electrical conductivity measurements indicated that saltwater intrusion into the sand aquifer from the Fraser River and Georgia Strait occurs across the delta. Identification of the different groundwater geochemical facies at the field site and comparison to other groundwater data from the Fraser delta supports the conceptual model for regional groundwater flow. There has been significant industrial development adjacent to the Fraser River Estuary and there is some potential for contamination of soil or groundwater at these industrial sites. For sites influenced by saltwater intrusion, this thesis indicates that a unique density-dependent groundwater flow regime develops that could affect the migration of contaminants. Inadequate site characterization plans, inaccurate contaminant transport predictions, and/or ineffective groundwater remediation designs could result if the effects of saltwater intrusion on the groundwater flow regime are not considered. Characterization of the groundwater flow regime in areas adjacent to the Fraser River is complicated by shallow gradients, tidal effects, and the variable density system. Density-dependent groundwater flow modelling can provide a valuable tool to assess the flow regime. T A B L E O F C O N T E N T S Abstract ii List of Tables vii List of Figures viii Acknowledgments xii 1.0 INTRODUCTION 1 2.0 BACKGROUND - SALTWATER INTRUSION 6 2.1 Saltwater Intrusion in Coastal Areas - General Concepts and Examples... 6 2.2 Numerical Models for Saltwater Intrusion 10 2.3 Saltwater Intrusion in the Fraser Delta, British Columbia 15 3.0 SITE DATA 18 3.1 Site Location and Description 18 3.2 Geological Conditions 21 3.2.1 Regional Geology 21 3.2.2 Local Geology 25 3.3 Hydrogeological Conditions 28 3.3.1 Regional Hydrogeology 28 3.3.2 Site Hydrogeology 30 3.3.2.1 Aquifer Description 30 3.3.2.2 Hydrogeological Parameters 31 3.3.2.3 Horizontal Groundwater Flow Gradient 33 3.3.2.4 Vertical Groundwater Flow Gradient 42 3.3.2.5 Recharge and Drainage Controls 43 3.3.2.6 Tidal/Seasonal Effects on Groundwater Levels 44 3.4 Groundwater Chemistry 45 3.4.1 Regional Distribution of Saline Groundwater 45 3.4.2 Site Groundwater Analytical Results 47 3.4.3 Saline Wedge Position at Field Site 69 3.4.4 Comparison to regional groundwater and surface water chemistry 72 3.5 Hydrological and Geochemical Characteristics of the Fraser River 75 3.5.1 Fraser River Salinity Data 75 3.5.2 River Water Temperature 79 3.5.3 River Water Density -The Salinity-Temperature-Density Relationship 79 3.5.4 Tidal and Seasonal Effects on River Level 81 4.0 MODEL DEVELOPMENT 84 4.1 Governing Mathematical Equations 84 4.2 Boundary Value Problem 87 4.2.1 Model Domain 87 4.2.2 Boundary Conditions for Groundwater Flow 90 4.2.3 Boundary Conditions for Solute Transport 94 4.3 Model Input Parameters 95 4.3.1 Hydrogeological Model Input Parameters 95 4.3.2 Other Model Input Parameters 99 4.4 Numerical Solution 100 5.0 MODEL RESULTS 103 5.1 Calibrated Steady State Solution 103 5.1.1 Model Calibration to Field Salinity Data 103 5.1.2 Discussion of Calibration Results 110 5.1.3 Model Results for Equivalent Freshwater Head 113 5.1.4 Model Results for Groundwater Velocity 113 5.2 Investigation of Transient Effects 117 5.2.1 Modelling of Seasonal Changes at River-Side Model Boundary 117 5.2.2 Discussion of Seasonal Model Results 120 5.2.3 Discussion of Potential Tidal Effects at River-Side Model Boundary 124 5.3 Sensitivity Analysis 126 5.3.1 Dispersivity 126 5.3.2 Aquifer Anisotropy 132 5.3.3 Freshwater Gradient Toward River 135 5.3.4 River Salinity/Density 138 5.3.5 Exit Point Boundary Condition 141 5.3.6 Implications of Sensitivity Analysis 141 6.0 DISCUSSION AND CONCLUSIONS 145 6.1 Groundwater Flow and Saline Water Distribution at the Field Site 145 6.2 Saltwater Intrusion and Groundwater Flow in the Fraser River Delta 147 6.2.1 Conceptual Model of Regional Groundwater Flow 147 6.2.2 Groundwater Geochemical Facies and the Regional Flow System 151 6.3 Implications for Contaminated Site Assessment and Remediation Problems 152 6.3.1 Potential Effects of Groundwater Flow System on Contaminant Migration 152 6.3.2 Groundwater Flow Regime Characterization Strategies for Sites Adjacent to the Fraser River 154 BIBLIOGRAPHY 157 APPENDICES I Well, Piezometer, Borehole, and CPT Logs 161 II Field Water Level Data and Calculation of Equivalent Fresh Water Head.... 192 III Tables of Analytical Results 195 IV Peclet Number Sensitivity Analysis 203 LIST OF TABLES Table 1 Summary of Hydraulic Conductivity Measurements 33 2 Estimates of Model Hydrogeological Parameters 96 3 Dispersivities for Sand Aquifers from Gelhar et al. (1992) 98 4 Other Model Input Parameters 100 5 Summary of Boundary Conditions for the Investigation of Seasonal Influences on the Saline Wedge 119 6 Summary of Sensitivity Simulations - Dispersivity 132 7 Summary of Sensitivity Simulations - Medium Sand Anisotropy 135 8 Summary of Sensitivity Simulations - Freshwater Gradient Toward River.... 13 8 9 Summary of Sensitivity Simulations - River Water Salinity/Density 138 LIST OF FIGURES Figure 1 Typical flow regime for aquifer affected by saltwater intrusion 2 2 Map of the Fraser River Delta 3 3 Saltwater intrusion from Biscayne Bay, Florida (from Kohout, 1960) 7 (a) illustrates extent of saltwater intrusion in plan view, (b) illustrates position of the saline wedge in cross section perpendicular to the coastline. 4 Numerical model results from Frind (1982) showing shape of concentration contours (isochlors) 9 5 A simplified sketch of a coastline in Barcelona, Spain, showing the influence of local features on saltwater intrusion in to an aquifer (from Custodio and Bruggeman, 1987) 9 6 Transient effects on the position of a saline wedge (modified from Custodio and Bruggeman, 1987) 11 7 Comparison of numerical model to field results for concentration contour positions in the Biscayne Aquifer, Florida (modified from Segol and Pinder, 1976) 13 8 Conductivity cross sections, (a) Section parallel to Roberts Bank, (b) Section parallel to Canoe Passage 16 9 Site Plan 19 10 Conceptual model of regional geology and hydrogeology -east-west cross section 22 11 Conceptual model of regional geology and hydrogeology -north-south cross section 23 12 Interpreted geological cross section perpendicular to the river 26 13 Regional groundwater flow in the Fraser River delta (modified from Ricketts, 1988) 29 14a Consecutive Water Level Readings for Deep Piezometers 101, 103, 105, and 107 35 14b Consecutive Water Level Readings for Intermediate Piezometers 102, 104, 106, and 108 36 14c Consecutive Water Level Readings for Shallow Piezometers 111, 112, and 114 37 15a Estimated equivalent freshwater head contours for deep zone-specific piezometers (101, 103, 105, and 107) 39 15b Estimated equivalent freshwater head contours for intermediate depth zone-specific piezometers (102, 104, 106, and 108) 40 16 Estimated contours of depth to saline groundwater using available regional groundwater chemical data 46 17 Groundwater pH for W3 49 18 Salinity versus depth for MLS wells (a) W1, (b) W2, and (c) W3 50 ix 19 Graphs showing the relationship between salinity and (a) electrical conductivity, (b) TDS, and (c) chloride 53 20 West Bay well groundwater chemistry vs. sample depths 55 21 Zone-specific wells 101, 102, and 112 groundwater chemistry vs. sample depths 56 22 Zone-specific wells 103, 104, and 111 groundwater chemistry vs. sample depths 57 23 Zone-specific wells 105, 106, 113, and 114 groundwater chemistry vs. sample depths 58 24 Zone-specific wells 107 and 108 groundwater chemistry vs. sample depths 59 25 K9701 groundwater chemistry vs. sample depths 60 26 K9801 groundwater chemistry vs. sample depths 61 27 K9802 groundwater chemistry vs. sample depths 62 28 W3 groundwater chemistry vs. sample depths 62 29 Ion-chloride relationship for Ca + +, Mg + + , Na+, K + , Sr++, Fe+ +, S04~, and HC0 3" 63 30 Trilinear diagram for site groundwater samples 66 31 Interpreted salinity contours in cross section 70 32 Trilinear diagram for selected groundwater samples from Simpson and Hutcheon (1995) 73 33a Fraser River salinity data - depth profiles over tidal cycle, salinity data Apr. 19-20, 1978 (Ages, 1988), discharge at Hope = 1710 m3/s 77 33b Fraser River salinity data - depth profiles over tidal cycle, salinity data Nov. 23, 1977 (Ages, 1979), discharge at Hope = 780 m3/s 77 34 Map showing the approximate extent of the saline river wedge under different river discharge conditions (modified from Ages and Woollard, 1976) 78 35 Water density variation with temperature and salinity, International Equation of State for Seawater (Unesco, 1981) 80 36 Model-predicted correspondence between highs and lows at Point Atkinson and Fraser Street (from Ages and Woollard, 1976) 83 37 Model domain 88 38a Development of Fraser River salinity stratification, salinity data Apr. 19-20, 1978 (Ages, 1988), discharge at Hope =1710 m3/s 91 38b Development of Fraser River salinity stratification, salinity data Nov. 23, 1977 (Ages, 1979), discharge at Hope 780 m3/s 91 39 Development of river-side model boundary conditions 93 40 Model boundary conditions and hydrogeological parameters for Simulation 1 104 41a Simulation 1 - relative salinity contours 105 41b Simulation 1 - equivalent freshwater head contours 105 41c Simulation 1 - velocity vectors plotted in true scale 106 41d Simulation 1 - horizontal component of velocity vectors plotted in log scale 106 42 Field salinity compared to model results - Simulation 1 107 X 43 Concentration contours for Simulation 1 without the inclusion of the lower permeability zone in the area of the wedge toe 109 44 Equivalent freshwater head contours for the Biscayne Aquifer (fromKohout, 1960) 114 45 Cross section illustrating conceptual groundwater flow regime at the site based on Simulation 1 results 116 46a Simulation 2 - concentration contours after 20 years of seasonal fluctuations at the river-side model boundary 121 46b Simulation 2 - concentration contours prior to initiating simulations of seasonal fluctuations at the river-side model boundary 121 47a Simulation 2 velocity distribution - for conditions of high seasonal river discharge 122 47b Simulation 2 velocity distribution - for conditions of average seasonal river discharge 122 48 Field salinity compared to model results - Simulation 2 125 49 Comparison of model concentration contours for variations in vertical transverse dispersivity 127 50 Comparison of model results in vertical profile for variations in vertical transverse dispersivity 128 51 Comparison of model concentration contours for variations in horizontal dispersivity 130 52 Comparison of model results in vertical profile for variations in horizontal dispersivity 131 53 Comparison of model concentration contours for variations in anisotropy of the Medium Sand unit 133 54 Comparison of model results in vertical profile for variations in the anisotropy of the Medium Sand unit 134 55 Comparison of model concentration contours for variations in freshwater gradient toward the river 136 56 Comparison of model results in vertical profile for variations in freshwater gradient toward river 137 57 Comparison of model concentration contours for variations in river water salinity (and density) 139 58 Comparison of model results in vertical profile for variations in river water salinity (and density) 140 59 Comparison of model concentration contours for variations in exit point elevation 142 60 Comparison of model results in vertical profile for variations in exit point elevation 143 61 Conceptual model of regional hydrogeology and saltwater intrusion -north-south cross-section 148 62 Conceptual model of regional hydrogeology and saltwater intrusion -east-west cross-section 149 63 a Effects of the groundwater flow regime at the field site on the migration of a hypothetical dissolved contaminant plume 153 63 b Effects of the groundwater flow regime for a hypothetical dissolved contaminant plume assuming fresh groundwater flow toward the Fraser River Xll ACKNOWLEDGEMENTS I would like to thank my supervisor, Dr. Leslie Smith, for his support and advice during the completion of this thesis. Thanks also to the other members of my review committee: Dr. Roger Beckie and Dr. Steve Calvert. In addition, I would like to thank the following people for their individual contributions to this thesis. • Steve Shikaze, University of Waterloo, for providing the groundwater flow model Fracdens. • Dr. R. Campanella and Dr. J. Howie, UBC Civil Engineering Department, for providing Cone Penetrometer Test (CPT) data and access to the CPT truck. • Brian Ricketts, Geological Survey of Canada, for funding to install the multi-level sampling wells at the field site. • Ian Hutcheon, University of Calgary, for providing chemical data for groundwater samples from the West Bay well. • Unsaturated Soils Group, University of Saskatchewan for consolidation testing of sediment samples. • Morrow Environmental Consultants Inc. for providing financial assistance, the use of field equipment and a computer. • Motomu Ibaraki for invaluable help with computer problems and questions. • Craig Nichol for installing data loggers at the field site and retrieving data to obtain continuous water level records for tidal filtering. Financial support for this thesis was provided by National Sciences and Engineering and Research Council of Canada and the Geological Survey of Canada. 1.0 I N T R O D U C T I O N Saltwater intrusion is a phenomenon which affects aquifers in coastal areas. The process involves the migration of saline water from the open ocean, inlets or estuaries into an adjacent aquifer system, and is a consequence of density contrasts between fresh groundwater and more dense saline sea water. In areas of saltwater intrusion, saline water is present in the aquifer in the form of a "wedge" which protrudes landward beneath fresh groundwater. Saltwater intrusion occurs under natural conditions and can be increased by aquifer development in coastal areas. A complex density-dependent groundwater flow system develops in areas influenced by saltwater intrusion. The typical groundwater flow regime involves landward flow of saline groundwater towards the toe of the saline wedge at which point the flow reverses and saline water (or mixed water) is carried upward and seaward by a regional freshwater gradient (Figure 1). A transition zone develops at the upper boundary and toe of the saline groundwater wedge across which the water chemistry changes from fresh to saline. This transition zone is formed as a result of dispersion associated with the flow of saline groundwater and may also be influenced by the transient groundwater velocity field which is observed in many coastal areas due to tidal effects (Underwood et al., 1992). Under constant hydrogeological conditions, the system exists in a state of dynamic equilibrium, and the saline water distribution and position of the saline wedge remain constant over time. This state of dynamic equilibrium is maintained by the combination of forces associated with regional seaward flow of fresh groundwater and the opposing landward density-driven flow of saline water. The Fraser River delta in southwestern British Columbia is affected by saltwater intrusion. Here, deltaic sediments extend westward toward the Georgia Strait and are transected by three branches of the Fraser River (Figure 2). Saltwater may intrude the sediments of the Fraser River delta from the ocean (Georgia Strait) or from the Fraser River estuaries at the North Arm, Middle Arm, or Main Channel. Knowledge of the hydrogeological conditions AQUITARD SALTWATER FRESH GROUNDWATER AQUIFER F X r — SALINE GROUNDWATER N AQUITARD SALTWATER WEDGE TOE Figure 1: Typical flow regime for aquifer affected by saltwater Intrusion. 3 4 in these areas of saltwater intrusion would be required for excavation dewatering, contaminant transport studies, or groundwater remediation system design. There has been significant industrial development across the delta and adjacent to the Fraser River (Figure 2), and at some industrial sites contamination may be present within a zone affected by saltwater intrusion. Inadequate site characterization plans, erroneous contaminant transport predictions, or inadequate remediation designs could result if the correct flow regime is not recognized. In this thesis, field data and a two-dimensional, density-dependent flow model are used to understand the processes affecting groundwater flow and saline water distribution at a site located adjacent to the North Arm of the Fraser River (Figure 2). At the site, groundwater data from zone-specific piezometers, multi-level sampling (MLS) wells, and cone penetrometer tests (CPTs) delineate a saline wedge that extends from the estuary into an adjacent confined sand aquifer. A boundary value problem, representing average or steady state conditions as indicated by the field data, is solved for groundwater flow and solute transport using the density-dependent groundwater flow model Fracdens (Shikaze et al., 1996). The model results are calibrated to site groundwater salinity data. Transient effects due to seasonal changes in the river are then investigated. A sensitivity analysis is also completed to understand the system response to variations of selected input parameters. The model parameters which are investigated include: aquifer dispersivity, aquifer anisotropy, groundwater flow gradient towards the river, water density in the river, and exit point position (the exit point is defined in Section 4.2.3). The model results and field geochemical data are reviewed in the context of the regional groundwater flow regime and saline groundwater distribution across the Fraser River delta (Ricketts (1998), Hunter et al. (1994, 1996), and Simpson and Hutcheon (1995)). A conceptual model for regional groundwater flow is developed; which accounts for the possible effects of saltwater intrusion at the delta margin and adjacent to the Fraser River 5 estuary. Implications for contaminated site assessment and remediation problems at areas adjacent to the Fraser River estuary are also examined. 2.0 BACKGROUND - SALTWATER INTRUSION 2.1 Saltwater Intrusion in Coastal Areas - General Concepts and Examples An understanding of saltwater intrusion into coastal aquifers has evolved since the late 19th century (Reilly and Goodman, 1985). Initial work is generally attributed to Bayden Ghyben and Herzberg, who in the late 1800's, developed an equation to predict the position of the saltwater interface (now referred to as the transition zone) (Reilly and Goodman, 1985). Later, saltwater intrusion research involved the development of numerical models for the simulation of groundwater flow in areas of saltwater intrusion, and the collection of groundwater chemical data to provide more detailed information on site specific conditions. As a result of the numerous field studies and modelling studies completed over the past century, extensive knowledge has been gained regarding the unique groundwater flow dynamics and saline water distribution present in coastal areas and the hydrogeological parameters which affect saltwater intrusion. Historically, the general shape of a saline wedge in coastal aquifers was considered to be approximately triangular, and the upper boundary of the wedge was thought to be a sharp saltwater "interface" which could be determined using the Ghyben-Herzberg relation (based on hydrostatic principles). Subsequent research by Cooper (1959), and the development of analytical models which considered hydrodynamic dispersion (e.g. Henry 1964, cited in Cooper et al, 1964), however, indicated that the upper boundary of the wedge was not represented by a sharp saltwater interface but by a thicker transition zone. The collection of field data in coastal areas confirmed the presence of a transition zone. A well studied example is the Biscayne Aquifer, Florida, USA. Groundwater chemical data (chloride concentration) from the Biscayne Aquifer was used by Kohout (1960) to delineate the saline wedge and transition zone within the Biscayne Aquifer, Florida (Figures 3a and 3b). In cross-section the data indicated that a wide transition zone was present (Figure 3b illustrates a vertical transition zone thickness of approximately 40 feet). The thickness of the transition zone in areas of saltwater intrusion is dependent on site-specific hydrogeological and/or geochemical parameters that influence hydrodynamic dispersion (i.e. aquifer dispersivity and groundwater velocity for systems of "usual" velocities where dispersion dominates over diffusion [DeMarsily, 1986]). High dispersivity and/or high groundwater velocity would lead to the development of a thicker transition zone. A transient velocity field due to tidal fluctuations in coastal areas may also influence the thickness of a transition zone. Computer modelling studies completed by Underwood et al. (1992) to investigate the effects of tidal fluctuations on the transition zone beneath atoll islands indicated that the thickness of the transition zone increased with increasing vertical dispersivity and tidal amplitude. With respect to the overall shape of a saline wedge (i.e. shape of the concentration contours delineating a saline wedge), data from the Biscayne Aquifer (Kohout, 1960) indicated that the toe of the wedge was "blunt-nosed" and convex-upward (Figure 3b); not "triangular" as would have been predicted using the Ghyben-Herzberg relation. Numerical density-dependent flow modelling for hypothetical systems (e.g. Frind, 1982a, and Huyakorn et al., 1987) also indicated a blunt toe and convex-upward configuration for the concentration contours at the transition zone (Figure 4). Hydrogeological factors which may affect shape of saline wedge were not investigated in these studies. The relative extent that saltwater intrudes an aquifer under natural (i.e. non-pumping) conditions is dependent on hydrogeological factors such as the permeability of the geological units, the density of the saline water, and seaward freshwater gradients. Figure 5 (from Custodio and Bruggeman, 1987) presents a simplified map-view sketch of a section of coastline in Barcelona, Spain, which shows the variability of the inland penetration of the saline wedge. On Figure 5, the old river channel represents a zone of relatively high permeability and saltwater intrusion extends inland along this former channel. The river and the lake on Figure 5 represent areas where saline surface water extends inland past the coastline and where seaward freshwater gradients are small. In these areas, the saline wedge also penetrates further inland. The dune fields represent an 9 Z n i f i • • i i i i i i i i i i r — i 1 i i i i i i D I S T A N C E (in) Figure 4: Numerical model results from Frind (1982) showing shape of concentration contours (isochlors). Figure 5: A simplified sketch of a coastline in Barcelona, Spain, showing the influence of local features on saltwater intrusion into an aquifer (from Custodio and Bruggeman, 1987). area of freshwater recharge where stronger seaward freshwater gradients are present; here, the extent of saline wedge penetration is reduced. Transient hydrogeological conditions such as seasonal precipitation, changes in seawater level, and changes in water salinity could also affect the position or inland extent of a saline wedge. Data from the Biscayne Aquifer analyzed by Kohout (1960) indicated that an increase in precipitation (during a rainy season) caused the movement of the saline wedge seaward. The general effects of changes in precipitation, sea level, and seawater density on the position of a saline wedge are illustrated on Figure 6 (modified from Custodio, 1987). Under conditions of high seasonal precipitation, low sea level, and or low seawater density, movement of the saline wedge seaward is indicated (a result of increased seaward gradients). Conversely, under conditions of low seasonal precipitation, high sea level, and/or high seawater density the wedge may establish an equilibrium position farther inland. More frequent changes in sea level due to tidal fluctuations were investigated by Underwood (1992) and affected the thickness of the transition zone but did not alter the general position of the saline wedge (as indicated by the 0.5 relative concentration contour). 2.2 Numerical Models for Saltwater Intrusion Quantitative saltwater intrusion studies were initially completed using a sharp interface approach which approximates the saline-freshwater transition zone as a sharp contact. A number of mathematical solutions to locate the interface position have been developed using this approach. Henry (1959) presented one of the first analytical solutions to determine the interface position, and numerical solutions were developed later by Shamir and Dagan (1971), Mercer et al. (1980), and others. The sharp interface approach, although potentially useful for some water supply problems, does not accurately resolve the changes in groundwater salinity across the saline-freshwater transition zone. Figure 6: Transient Effects on Position of a Sartwater Wedge (modified from Custodio and Bruggeman, 1987). a) steady state conditions b) conditions of raised sea level, increased salinity of sea water, and/or low precipitation cause saltwater wedge to more landward. c) conditions of low sea level, decreased salinity of sea water and/or increased precipitation cause sarrwater wedge to move seaward. In order to more accurately model groundwater flow and saline water distribution in areas of saltwater intrusion, density-dependent flow models were developed. With these models, the groundwater flow and solute transport equations are solved simultaneously for a given time step and are linked by an equation of state which relates groundwater density to solute concentration. Henry (1964) (cited in Cooper et al., 1964) provided the first analytical solution to the density-dependent flow problem. A two-dimensional numerical solution to the problem was first presented by Pinder and Cooper (1970), and later, other two-dimensional numerical solutions were presented (e.g. Frind, 1982 a, b). Three-dimensional solutions have also been developed (e.g. Huyakorn et al., 1987, and Xue et al., 1995). Numerical models have been calibrated to and compared to field data to demonstrate their applicability and accuracy in modelling saltwater intrusion. Lee and Cheng (1974) and Segol and Pinder (1976) used density-dependent groundwater flow models to simulate saltwater intrusion in the Biscayne Aquifer. The model results were compared to field chloride data. For both studies, the modelled chloride concentration contours in cross section exhibited the same general shape as the chloride contours estimated based on field data. The study by Lee and Cheng, however, underestimated the inland extent of the saltwater wedge (Figure 7). In the Segol and Pinder study, the model was calibrated by varying the aquifer permeability and dispersivities to obtain a fit to the field data. The model, however, predicted a longer and vertically thinner saline wedge than was observed based on field data (Figure 7). Other studies that compared model results to field data have included: Xue et al. (1995), Huangheying, China, Birdie et al. (1993), Florida, USA, and Das Gupta and Yapa (1982), Bangkok, Thailand. Xue et al. (1995) simulated saltwater intrusion for an irregular (i.e. not rectangular) model domain and obtained a good match between field data and model results for the position of the saline wedge. The number of data points for model calibration, however, was small compared to that available for the Biscayne Aquifer. As DISTANCE FRCM SHORELINE , IN FEET -1600 -1400 -I20Q -lOOQ -aQQ .600 -400 -200 0 200 400 600 DISTANCE FROM SHORELINE,IN FEET -1600 -1400 -1200 -1000 - 800 - 600 - 400 -10 -20 -30 -40 ti - 5 0 -60 z -• -70 I -80 | -90 UJ j -100 -200 200 400 600 PRESENT ANALYSIS •— AFTER KOHOUT AND KLEIN (1967) -1600 -1400 -1200 -10 y -20 -30 -40 t -50 UJ -60 ? . -70 I -80 % -90 LkJ r> -100 DISTANCE; FROM SHORELINE , IN FEET -1000 - 800 - 600 - 400 - 200 200 400 PRESENT ANALYSIS AFTER KOHOUT AND KLEIN (1967) — AFTER LEE AND CHENG (1974) 600 0.75 (14,200 ppm)isochlors -1600 -1400 DISTANCE FROM SHORELINE . IN FEET -1200 -1000 -800 - 600 - 400 200 400 600 Figure 7: Comparison of numerical model to field results for concentration contour positions in the Biscayne Aquifer, Florida (modified from Segol and Pinder, 1976). Note: Curves labelled "present analysis" refer to the results of Segol and Pinder, 1976. Curves representing field data are labelled as "Kohout and Klein (1967)". such, discrepancies between model results and field data may not be evident. Other studies noted above were also compared to limited field data. The majority of numerical modelling studies of saltwater intrusion that consider site-specific conditions have investigated intrusion into aquifers from the open ocean (Zhang, 1995). Saltwater intrusion also occurs in estuaries. Saltwater enters estuaries providing a source for saltwater intrusion which can extend significant distances from the coastline. Xue et al. (1995) used a three-dimensional numerical model to simulate the effects of groundwater pumping on seawater intrusion at a site in China. Sources for seawater intrusion at the site were the ocean (Bohai Sea) and an estuary (Huangshuihe River). Their numerical simulations indicated that groundwater pumping caused induced seawater intrusion from the estuary and the ocean into a phreatic aquifer. Details of estuary characteristics modelled at the estuary boundary (e.g. salinity stratification, transient effects) were not presented. Two-dimensional numerical modelling in cross-section has been used to investigate seawater intrusion from brackish canals (similar to estuaries) in Florida (Zhang, 1995). This study investigated the effects of variables such as: low-permeability canal bottom sediments; tidal fluctuations; and seasonal groundwater levels on the initiation and progression of saltwater intrusion from the brackish canals. Some of the findings of this study were: low-permeability canal bottom sediments slowed the intrusion of brackish water into the aquifer; tidal fluctuations caused saltwater intrusion to proceed more quickly and with a smoother front than when tidal fluctuations were not modelled; and, under certain conditions, seasonal groundwater level fluctuations caused saltwater to intrude the aquifer as a series of pulses. This study did not examine saltwater intrusion when conditions of dynamic equilibrium had been attained. 2.3 Saltwater Intrusion in the Fraser Delta, British Columbia To date, there have been no studies carried out specifically to delineate the extent of saltwater intrusion in the Fraser delta. Some data characterizing the subsurface electrical conductivity have been collected, however. Borehole logging to obtain electrical conductivity profiles with depth at 22 locations across the lower Fraser River delta has been completed by Hunter et al. (1994). These measurements indicate the electrical conductivity of both the sediments and pore water. Conductivities below 200 mS/m indicate low pore water salinity, 200 to 600 mS/m indicate brackish pore-water, and greater than 600mS/m indicate saline water (Hunter et al., 1994). Additional electrical conductivity data are presented by Hunter et al. (1996). Cross sections presenting some of the data collected by Hunter et al. (1996) are presented as Figures 8a and 8b. Their investigation involved the measurement of ground electrical conductivity at 120 sites on the Southern Fraser delta, south of the Main Channel. Inverse modelling of these ground electrical conductivity measurements was then used to produce a simplified electrical conductivity depth profile at each of the measurement locations. The data from Hunter et al. (1994, 1996) indicate the presence of saline groundwater over an area extending east from the coastline to approximately 10 km inland within the deltaic sediments (Ricketts, 1998). At most measurement locations, a shallow low conductivity layer (indicative of fresh pore water), varying in thickness from 0 to 30 m, was identified overlying a high conductivity layer (indicating saline pore water). At some locations a deeper low conductivity layer was identified beneath the high conductivity layer (the top of this layer was present at depths ranging from approximately 20 to 30 m below the ground surface). The cross sections in Figure 8 from Hunter et al. (1996) are for locations parallel to Roberts Bank (Figure 8a), and parallel to the Fraser River at the western end of Canoe Passage (Figure 8b). These profiles illustrate the conductivity distribution parallel to the coastline and estuary. It is apparent from these cross sections that a complex SOUTHEAST NORTHWEST -500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Figure 8: Conductivity cross sections (from Hunter et al. 1996). (a) Section parallel to Roberts Bank (see Figure 2) (b) Section parallel to Canoe Passage (see Figure 2) Note: darker shading indicates saline groundwater, lighter shading indicates less saline or fresh groundwater. conductivity/salinity distribution exists as the thicknesses of the low and high conductivity layers vary significantly across the sections. Ricketts (1998) states that the conductivity data of Hunter et al. (1994, 1996) do not indicate a simple saltwater wedge configuration, but suggest that seawater intrudes the deltaic sediments in a series of "plumes" which are separated by zones of seaward flowing fresh groundwater. Further analysis of the data presented in Hunter et al. (1994, 1996) and the relationship of these conductivity measurements to saltwater intrusion and regional hydrogeology is provided in Sections 3.4.1 and 6.2.1 of this thesis. Modelling studies of seawater intrusion in the Fraser delta have not been carried out. The hydrogeology of the Fraser River delta has been recently investigated by Ricketts (1998) (discussed in Section 3.5.1 of this thesis). The study, however, was completed to investigate regional fresh groundwater flow systems and saltwater intrusion was not modelled. In addition to saltwater intrusion from the ocean or estuaries, old trapped seawater is another source for saline water in coastal formations (Custodio and Bruggeman, 1987). Regional seawater intrusion studies have identified old trapped seawater, as well as saline water due to saltwater intrusion from the ocean, within deltaic sediments at Bangkok, Thailand (Das Gupta and Yapa, 1982) and Barcelona, Spain (Custodio and Bruggeman, 1987). Simpson and Hutcheon (1995) completed a geochemical study based on sediment, groundwater, and surface water samples from across the Fraser River delta and indicated that old trapped seawater may be present in a shallow sand aquifer and deeper delta front and delta slope deposits. Delineation of areas of saltwater intrusion at the Fraser River delta may be complicated by the presence of saline water from the different sources. 3.0 SITE DATA 3.1 Site Location and Description The field site is located on Lulu Island in Richmond, British Columbia, and is situated on the Fraser River delta (Figure 2). The property is owned by BC Hydro. The BC Hydro Kidd 2 Substation is present on the northern portion of the property and a BC Gas building is present on the southern portion of the property. Hydrogeological and geotechnical research is conducted at the site by the University of British Columbia. The site area (defined as the area where field work was completed) is approximately 500 m long (north to south) and 50 m wide (Figure 9). The local topography is essentially flat, and the elevation of the ground surface at the site is approximately 1.5m above geodetic datum (mean sea level). To the north, the Fraser River (approximately 500 m wide at that point) borders the site. On the north side of the river is the City of Vancouver. There, the delta deposits continue for approximately 500 m (Armstrong and Hicock, 1976), and then the topography rises at a slope of approximately 0.03 (3%) reaching a maximum elevation of about 115 m 4 km north of the river. West of Lulu Island, the Fraser River delta extends to the Pacific Ocean (Georgia Strait), and to the East the delta extends to Surrey Uplands. To the south of the site, the delta continues approximately 15 to 20 km to Boundary Bay. The Main Channel of the Fraser River crosses the delta approximately 10 km south of the site and bounds Lulu Island to the south. The site is equipped with 17 permanent groundwater wells consisting of three multi-level sampling (MLS) wells (Wl, W2, and W3), 12 zone-specific piezometers (101 to 108 and 111 to 114), one West Bay well, and one pumping well (PW). The locations of these wells are identified on Figure 9. Well logs are included in Appendix I and show the construction details for each installation. 19 Fraser River " D W .K9801 River Road BC Hydro Kidd 2 Substation •K98Q2 , W 3 •K9S08 — gl «BH •K9612 • Wl; •K93Q1 i Approximate Area i of BC Gas Building ; (constructed In 1W7) • K 9 3 » • KWill 103.104 05.106 o o Condominium Development Residential Houses Figure 9: Site Plan PW = Pumping Well WB = West Bay Well BH = Borehole K = Cone Penetrometer Test 101-114 = Zone-Specific Wells W1-W3 = Multi-Level Sampling (MLS) Wells 50 100 SCALE (m) The pumping well was installed for testing of hydrogeological parameters (i.e. for pump tests) and has not been used for groundwater sampling or water level measurements. The West Bay well and zone-specific piezometers provided groundwater chemical data and water level data for specific depths or depth intervals. The West Bay well consists of a single well casing with electronically controlled sample/measurement ports at 12 discrete depths or zones. The depths of the sampling ports for the West Bay well are indicated on the West Bay well log in Appendix I. Each of the zone-specific piezometers (101 to 108 and 111 to 114) are screened across a specific depth interval. Piezometer 113 is screened from 1.6 to 2.4 m below the ground surface, piezometers 111, 112, and 114 are screened from approximately 4.25 to 5 m depth, piezometers 102, 104, 106, and 108 are screened from approximately 11.5 tol3 m depth, and piezometers 101, 103, 105, and 107 are screened from 16.5 to 18 m depth. Screen intervals for each of the zone-specific piezometers are indicated on the piezometer logs, Appendix I. Piezometers 101 to 108 were installed in 1995, and piezometers 111 to 114 were installed in 1996 specifically for this study. The three MLS wells (Wl, W2, and W3) were installed specifically for the purposes of this thesis to provide vertically-detailed groundwater chemistry data that could not be obtained from the zone-specific piezometers or the West Bay well (as completed). The MLS wells were used for groundwater sampling but were not used to obtain water level data for the assessment of hydraulic gradients at the site. The MLS wells each consist of a 5 cm diameter PVC casing extending from the ground surface to approximately 22 m depth. Inside the PVC casing, fifteen 0.787 cm outer diameter polyethylene tubes are present. Each of the tubes exits the PVC casing at a different depth providing 15 closely spaced (approximately lm spacing) sampling ports. The well logs (Appendix I) identify the depths of the sampling ports for each of the MLS wells. Note that all of the ports of W3 are operational, however, some of the ports for W l and W2 are not operational. This could be due to damage during the installation of the wells or fine grained silt/clay materials clogging the nylon mesh covering each sampling port. Attempts to make these sampling ports functional were unsuccessful. 21 In addition to the permanent wells/piezometers discussed above, cone penetrometer tests (CPTs) and one continuous-core borehole, providing geological and/or groundwater chemical data, have been completed across the site. Figure 9 identifies the locations of the CPTs and borehole (BH) at the site for which data have been obtained and used in this thesis. The BH was completed to a depth of approximately 22 m and provided a continuous core for stratigraphic information. The stratigraphy based on an examination of the core is presented on a borehole log, Appendix I. The CPTs were completed by the UBC Civil Engineering Department under the direction of Dr. R. Campanella (CPTs completed prior to 1998) and Dr. J. Howie (CPTs completed in 1998), and are identified as K9301, K9308, K9309, K9310, K9311, K9601, K9602, K9603, K9612, K9701, K9801, and K9802 on Figure 9. Detailed friction and resistivity measurements were reviewed for these CPTs to obtain stratigraphic information and relative changes in groundwater salinity (from resistivity measurements) with depth. In addition, groundwater samples were collected at specific depths during CPTs K9701, K9801, and K9802. CPT logs showing the interpreted stratigraphy, resistivity profile, and groundwater sampling depths are included in Appendix I. 3.2 Geological Conditions 3.2.1 Regional Geology The regional geology of the Fraser River delta has been interpreted by Armstrong and Hicock (1976). More recently Ricketts (1998) has compiled a regional geological description for the Fraser River delta based on previous work and newly obtained data from boreholes and CPTs. The stratigraphy of the shallow deltaic sediments has been investigated by Williams and Roberts (1989) and Monahan et al. (1993). A description of the regional geological setting is provided below and Figures 10 and 11 provide East-West and North-South cross sections which illustrate the regional geological conditions. 22 C o CL ZD ! c o tj © 2 o is I to O <D • >• CD O O © CD O J C "D ^ C TJ ^ © i s « 0 O C (D O O O O O c c o c CD O © cq <D <-1! © C o o o o c o CO 1 o O x: CO CM 2 CD co © O fv T3 C o -o o-o o o in TJ c O D) C o t o E CO © O C © (isui) |9Ae| Des UDSUU 04 e/vjpiej LU 23 c o (isui) | G A G | Des UDeuj o; eAupiej U J 24 The Fraser River delta consists of Holocene sediments, deposited over the past 9000 years (Monahan et al., 1993). Surficial floodplain and peat bog deposits, consisting of organic rich sandy to clayey silts, extend across the delta plain (Williams and Roberts, 1989). These deposits are underlain by interbedded silts and sands, and then by a more homogeneous sand unit (Williams and Roberts, 1989). The homogeneous sand unit is fine to coarse grained with a thickness of approximately 8 to 20 m, and is essentially continuous across the delta (Monahan et al., 1993). The sand has a sharp base with a few metres of relief (Monahan et al., 1993), and overlies fine grained (e.g. silt) delta slope deposits (Ricketts, 1998). Beneath the delta slope deposits, which have a thickness of up to 170 m, prodelta silt deposits are present (Ricketts, 1998). Ricketts (1998) and Armstrong and Hickock (1976) provide interpretations of the stratigraphy beneath the deltaic sediments. Ricketts (1998) indicates that the prodelta deposits lie on an unconformity created during the retreat of the Pleistocene glaciation. The unconformity has up to 400 m of relief (from exposure at the Surrey Uplands to the seaward delta margin), and separates the deltaic deposits from older Pleistocene glacial deposits (Ricketts, 1998). The Pleistocene glacial deposits are identified by Armstrong and Hickock (1976) as Capilano Sediments, Vashon Drift, and Pre-Vashon Deposits. For the purposes of a regional groundwater modelling study of the Fraser River delta, Ricketts (1998) subdivides the Pre-Vashon Pleistocene deposits beneath the delta into five alternating layers representing coarser grained sediments (aquifers) and clay, silt, and diamicton (aquitards). The Pleistocene geological units identified by Armstrong and Hickock (1976) as well as the aquifer and aquitard layers of the Pre-Vashon Deposits (Ricketts, 1998) are indicated on Figures 10 and 11. The Pleistocene glacial sediments were deposited over Tertiary bedrock and the contact is marked by a second unconformity. Ricketts (1998) indicates that the unconformity extends up to 800 m deep at some locations. The bedrock outcrops in areas surrounding the delta (e.g. Burrard Uplands) and generally consists of sandstone and mudstone overlying conglomerate and sandstone. 3.2.2 Local Geology The local geological conditions at the site have been interpreted based on core samples obtained during drilling of the borehole, CPTs at the site, and available literature on the Fraser River delta (Monahan et al., 1993). The testing carried out at the site provides stratigraphic information for depths up to 27 m below ground surface (i.e. for the shallow deltaic deposits only). The local stratigraphy is illustrated on Figure 12 for a North-South cross section. The geological deposits observed at the site are consistent with the shallow sediments which have been identified across the Fraser River delta. Based on the site data, five distinct and approximately horizontal geological units can be identified. Descriptions and approximate depths of these units are summarized below. Approximate Depth (m) Geological Unit Description 0-3.5 Clayey Silt: light grey clayey silt, trace fine sand laminations, trace to some organic matter (wood pieces, leaves, etc.). The Clayey Silt unit is topped with a thin (< 1 m thick) layer of fill consisting of sand, gravel, and/or top soil. 3.5-9 Silty Sand: grey, silty, fine and medium grained sand, interbeds of clayey silt, sandy silt, and fine sand. 9-12 Fine and Medium Sand: grey, fine and medium sand, some silt, some interbeds of silt and fine sand. 12 - 22 Medium Sand: grey medium sand, uniform, occasional wood fragments. > 22 Silty Clay: light grey silty clay, thin laminations of silt. *The contact between the Fine and Medium Sand unit and the underlying Medium Sand unit is gradational and has been estimated based on overall change in grain size at borehole or CPT locations. 26 ° E JB r - S | O . o X 0 > 0 o o T3 C 0 Q_ 0 a c o o 0 CO t o CO o b o O a> o o 0 a> 0 0 a 0 CM 0 D U) d £ o co CO (0 "O £ .1 g 8 u CD O CO e co o CO CO 7= O fc m r -a o n I r o i o O ° CD J 3 a> n E W 9 U S OB CO to CO <° c ca £ * _ <o A 8 | c •s .o co o j3 iS o o i «S co O O " £ T 3 CD 2 O J? O coO ro CO S o " T The geological units are approximately horizontal but display some variability across the site (Figure 12). The upper Clayey Silt unit is essentially horizontal maintaining a constant thickness of approximately 3.5m across the site. The underlying units (Silty Sand, Fine and Medium Sand, and Medium Sand) dip slightly to the south across the site. At the southern side of the site, the thicknesses of the Silty Sand and Fine and Medium Sand units increase slightly. The top of the Silty Clay unit is slightly uneven (exhibits approximately 1 to 2m of local relief) and also dips to the south away from the river. The mineralogy of the shallow deltaic deposits has been investigated by Simpson and Hutcheon (1995). Samples of the deltaic sediments were analyzed by X-ray diffraction to determine the mineralogy. Their results indicated an approximate mineral composition in weight percent as follows: quartz 50-60% feldspar 30-40% mica or illite up to 15% calcite up to 11% chlorite amphibole, and pyrite less than 2% The clay sized fraction (< 2um) was analyzed by Simpson and Hutcheon (1995) and the results indicate that the clay fraction contains more than 45% illite, 5 to 10% smectite and chlorite, 0 to 12% kaolinite, and up to 10% quartz and feldspar. Organic matter has been observed in trace amounts within the geological units at the site. This organic matter generally consists of visible pieces of wood fragments, grasses or leaves, etc. in the upper Clayey Silt unit. Wood fragments were also observed within the sand units (Silty Sand, Fine and Medium Sand, and Medium Sand). Gas, predominantly methane (C. Clarkson, personal communication), has been identified within the Silty Clay unit which indicates the degradation of and presence of organic matter within this unit. 3.3 Hydrogeological Conditions 3.3.1 Regional Hydrogeology A hydrogeological study of the entire Fraser River delta has recently been completed by Ricketts (1998). This study involved simulations of groundwater flow in the Fraser River delta area using a numerical groundwater flow model (Modflow). The regional flow regime in the area of the Fraser River delta, as interpreted by Ricketts (1998), is illustrated for east-west and north-south cross sections on Figures 10 and 11 and in plan view on Figure 13. Note that the study did not account for the effects of density-dependent flow in areas of saltwater intrusion. As illustrated on Figure 10 and 11, local flow systems are identified within shallow deltaic deposits. These local flow systems are driven mainly by precipitation on the delta plain. A large portion of the precipitation is captured by an extensive drainage network consisting of ditches, storm drains, and pumps which pump collected drainage water to the river as necessary (City of Richmond, personal communication). Some of the precipitation, however, infiltrates into the deltaic sediments (Ricketts, 1998, estimates a recharge rate of 130 mm/y for the delta), migrates downward through shallow surficial deposits, and then migrates laterally through the more permeable sand aquifer. The local flow systems that are formed, terminate at the Fraser River or at the delta front where groundwater is discharged. Intermediate and regional flow systems indicated by Ricketts (1998) are also illustrated on Figures 10 and 11. The intermediate and regional flow systems are created by groundwater that migrates downward through the elevated areas surrounding the delta (e.g. Burrard Uplands to the north of the site). This groundwater migrates downward through permeable aquifer units within Pleistocene glacial sediments or Tertiary bedrock. Then the groundwater flows upward to the Holocene delta where it is ultimately discharged to the Georgia Strait. 30 The study by Ricketts (1998) also provides information regarding groundwater flow direction in plan view within the sand aquifer of the shallow deltaic deposits (Figure 13). Ricketts' results indicate that the shallow groundwater flow within the sand aquifer is influenced by two east-west oriented groundwater divides; one across Lulu Island and one across the southern delta. The groundwater divide on Lulu Island separates northerly groundwater flow toward the North Arm of the Fraser River from southerly groundwater flow towards the Main Channel. Hydraulic gradients of approximately 0.0001 are indicated for both northerly and southerly flow directions across Lulu Island. 3.3.2 Site Hydrogeology 3.3.2.1 Aquifer Description The aquifer at the site (referenced in this thesis as the sand aquifer) is comprised of the Medium Sand, Fine and Medium Sand, and Silty Sand units. The thickness of the aquifer is variable across the site, ranging from approximately 14.5 to 18.5 m. The site stratigraphy indicates that the aquifer is confined from above by the low permeability Clayey Silt unit and from below by the low permeability Silty Clay unit. Storativity of the aquifer has been estimated through a limited four hour pump test to be approximately 0.003 which also indicates confined aquifer conditions. The confined aquifer may be affected by leakage from below as field data indicate that fresh groundwater flows upward through the lower confining unit. The sand aquifer intersects the river at a depth of approximately 4.5 to 5.5m below geodetic datum (Figure 12). This information was inferred from one CPT (K9301) completed adjacent to the river (for CPT location see Figure 9). The maximum depth of the Fraser River adjacent to the site is approximately 9 m below geodetic datum (Canadian Hydrographic Service, 1996). As such, the Fraser river penetrates the upper 4.5 m of the sand aquifer. For a saltwater intrusion study for which the source of saline water is a river/estuary, it is important to determine whether fine-grained river bottom sediments (i.e. silt and clay) are present. If present, low-permeability river bottom sediments could influence saltwater intrusion into the aquifer from the river (Zhang, 1995). An investigation for another site located on the north shore of the Fraser River (across from the UBC site) has indicated that the sediments on the river bed in the area of the site are sand and gravel (Golder Associates, personal communication). 3.3.2.2 Hydrogeological Parameters Field and laboratory data have been used to estimate the hydraulic conductivities of the geological units. Table 1 presents a summary of the measured hydraulic conductivities for the different geological units, the different test methods, and sources of data. Testing has involved a pump test and slug tests which were completed at the time of UBC hydrogeology field school in May 1996. Hvorslev analysis of the slug test data provided hydraulic conductivities of the Medium Sand (piezometers 101, 103, 105, and 107), Fine and Medium Sand (piezometers 102, 104, and 108), and Silty Sand unit (piezometer 112). Jacob analysis of the pump test data provided a hydraulic conductivity value for the Medium Sand unit. (Note pump test data were not corrected for tidal fluctuations which could affect the calculation of hydraulic conductivity.) Other field pump test and slug test data for the site were analyzed by Wood (1996). Laboratory permeameter tests and grain size analyses were also completed by Wood (1996) for sediment samples representing the different geological units. Consolidation tests for the Silty Clay unit were completed specifically for this study at the University of Saskatchewan. During the early numerical model simulations the upward flux of fresh groundwater at the base of the model was estimated to be 0.05 m/y using the hydraulic conductivity of 1 x 10"8m/s for the Silty Clay determined by Wood (1996) and the upward vertical gradient of 0.17 for the Silty Clay determined using data from the West Bay well (discussed in Section 3.3.2.4). Initial model simulations, however, indicated that a better match to the field data was obtained using a bottom flux of 0.0005 m/y (corresponds to a hydraulic conductivity of 1 x 10"10 m/s). The permeability of the Silty Clay was then re-measured through consolidation tests at the University of Saskatchewan and the hydraulic conductivity of 1 x 10"10 m/s was confirmed by the consolidation tests. Anisotropy in hydraulic conductivity is caused by small or large scale layering of different grain sizes or common grain orientation. Visual observations of the core samples indicate that the Medium Sand unit exhibits little or no visible layering, the Fine and Medium Sand unit exhibits trace to some horizontal layering, and the Silty Sand contains visible horizontal layering. The CPT friction data provide an additional indication of anisotropy (even when visible layering is not present) because frequent changes in grain size with depth can be resolved by measurements taken every 0.025 m. The CPT data suggest that all units comprising the sand aquifer are anisotropic. The medium sand unit contains thin layers of fine sand and coarse sand, the Fine and Medium Sand unit exhibits similar variations in grain size with depth with layers of silty sand, fine sand, and medium sand. The CPT data also confirm layering within the Silty Sand unit. The degree of hydraulic conductivity anisotropy for the geological units at the site has not been measured. As discussed previously, a limited four hour pump test indicates that the aquifer storativity is approximately 0.003. Other hydrogeological parameters such as porosity, aquifer compressibility, dispersivities, tortuosity, and adsorption (Retardation Factor) have not been measured at the site. 33 Table 1: Summary of Hydraulic Conductivity Measurements Geological Unit Average Hydraulic Conductivity (m/s) Source Clayey Silt 5.0 x 10"8 Falling Head Permeability Tests (Wood, 1996) Silty Sand 2 x 10"5 2.8 x 10'5 1.6 x 10"4 'Selected Slug Test Data (Field School Data, 1996) Falling Head Permeability Test (Wood, 1996) Grain Size Analysis (Wood, 1996) Fine and Medium Sand 4.3 x 10"4 1.5 x 10"4 2.8 x 10"4 2.2 x 10"4 'Selected Slug Test Data (Field School Data, 1996) Falling Head Permeability Tests (Wood, 1996) Slug Test Data (Wood, 1996) Grain Size Analysis (Wood, 1996) Medium Sand 4.4 x 10"4 3.9 x 104 1 x 10"4 3.3 x 10'4 3.5 x 10"4 3.6 x 104 'Selected Slug Test Data (Field School Data, 1996) Pump Test (Field School Data, 1996) Falling Head Permeability Tests (Wood, 1996) Pump Test (Wood, 1996) Slug Test Data (Wood, 1996) Grain Size Analysis (Wood, 1996) Silty Clay 1.2 x 10"'0 1.2 x 10"8 Consolidation Tests (University of Saskatchewan, 1998) Falling Head Permeability Tests (Wood, 1996) ' Slug tests were completed during UBC Hydrogeology Field School, May 1996. The average hydraulic conductivities presented above represent averages of selected slug tests which provided the most reliable data for analysis (i.e. water level oscillations during recovery were minimal). 3.3.2.3 Horizontal Groundwater Flow Gradients Water level data have been collected from the zone-specific piezometers at the site to determine the groundwater flow direction. The use of water level measurements for gradient determination at the site is complicated by the influence of the tides on the groundwater levels, as well as the variable density of the groundwater. Due to potentially large tidal wave phase differences between spatially separated piezometers, accurate gradient determination cannot be carried out without an averaging process using numerous consecutive groundwater level readings (Series, 1991). A method outlined by Series (1991) was used to filter the water level data to determine average water levels for the zone-specific piezometers. The procedure involved the collection of 25 or 72 consecutive hourly water level readings for each piezometer and then the calculation of mean water levels. The consecutive water level measurements for deep, intermediate, and shallow zone-specific piezometer sets are presented on Figures 14a, 14b, and 14c. The results illustrate the dynamic nature of the water levels within the piezometers at the site due to tidal influences. Smooth sinusoidal curves for all deep piezometers (101, 103, 105, and 107) and intermediate depth piezometers (102, 104, 106, and 108) are indicated on Figures 14a and 14b. With respect to the shallow piezometers (111, 112, and 114), Figure 14c indicates smooth sinusoidal curves for piezometers 111 and 114. For piezometer 112, however, a deviation from the expected curve is observed. It is assumed that this deviation is the result of temporary disturbances of the water level in piezometer 112 or a malfunction of the data logger. Tidal filtering to determine the average water level was not completed for piezometer 112 as the anomalous readings indicate that the data may not be representative. Tidal filtering was completed for all other deep, intermediate, and shallow piezometers. Once tidal filtering to determine mean water levels for each piezometer was completed, the water levels were converted to equivalent freshwater head. Equivalent freshwater head is defined at a point p within a groundwater system of variable density as "the water level in a well filled with freshwater from p to a level high enough to balance the existing pressure at p" (Lusczynski, 1961). Equivalent freshwater head (h) at any point p is defined by the equation: h = — h - ——— z (z is positive upwards) (1) Po Po where p is the groundwater density at point p [M/L 3], p0 is the density of fresh water [M/L 3], ho is the hydraulic head [m] at point p, and z indicates the vertical elevation of point p. 35 37 (tu) uo!JBA3|3 J3}BM o|japoao 38 The conversion of water levels to equivalent freshwater head is required as hydraulic gradients based on water level data cannot be used directly to determine the groundwater flow direction for a variable density system (Custodio and Bruggeman, 1987). The use of equation 1 requires knowledge of the density of the groundwater which was calculated for each piezometer using field salinity and temperature measurements and the International Equation of State for Seawater (UNESCO, 1981). (Figure 35 shows the relationship between water density, salinity, and temperature as calculated using the International Equation of State for Seawater). A sample calculation showing the conversion of measured water levels to equivalent freshwater head using equation 1 is provided in Appendix II. Average equivalent freshwater head data for the zone-specific piezometers are presented in Table A2-1, Appendix II. (Note that the average equivalent freshwater head was not calculated for piezometer 112 as anomalous water level readings were obtained.) The conversion of the water level data to equivalent freshwater head allows the comparison of water levels, for wells screened at the same elevation, to determine the horizontal groundwater flow direction (Custodio and Bruggeman, 1987). Figures 15a and 15b show the equivalent freshwater head contours for deep piezometers screened in the Medium Sand unit (piezometers 101, 103, 105, and 107) and the intermediate piezometers screened in the Fine and Medium Sand unit (piezometers 102, 104, 106, and 108). With respect to the shallow piezometers, there is insufficient data to determine the groundwater flow direction as the average equivalent freshwater head could not be determined for piezometer 112 (the average equivalent fresh water head for at least three locations must be known to determine a groundwater flow direction). Data from the deep piezometers indicate groundwater flow to the southwest within the Medium Sand unit. Intermediate piezometers completed within the Fine and Medium Sand unit indicate flow is generally to the west. Horizontal equivalent freshwater head gradients are approximately 0.005 based on the data for the deep piezometers and 0.0004 BC Hydro Kidd 2 Substation! E275 -0228 — " - ^ *.103 107 a 3 0° 101 / 1Q5« (0.J94) \ Approximate Gro jndwater Flow Direction •o o o 6 21 39 Residential Houses 50 100 SCALE (m) Figure 15a: Estimated equivalent freshwater head contours for deep zone-specific piezometers (101,103,107, and 107). (0.271) Average Equivalent Freshwater Head (m) Equivalent freshwater head levels are relative to geodetic datum. - 0.250 Estimated Equivalent Freshwater Head Contour (m) BC Hydro Kidd 2 Substatiori Approximate 4 Groundwater Flow Direction o 108 \ 102 , #<ano)\ •(0.117) 106 . (0.120), •104 (0-123) "D D O 40 Residential Houses 50 100 SCALE (m) Figure 15b: Estimated equivalent freshwater head contours for intermediate depth zone-specific piezometers (102,104106, and 108). (0.110) Average Equivalent Freshwater Head (m) Equivalent freshwater head levels are relative to geodetic datum. — - . - - 0.120 Estimated Equivalent Freshwater Head Contour (m) based on the data from the intermediate depth piezometers. It should be noted that measurement errors could influence the ability to accurately measure small horizontal gradients at the site. Errors could be introduced during the survey of piezometer top geodetic elevations or during the measurements of water levels in the piezometers. Errors in salinity measurements could also affect equivalent freshwater head calculations as the salinity measurements are used to determine density which in turn is used to calculate equivalent freshwater head. Salinity measurements were not completed at the same time as the water level measurements and thus may not reflect actual salinity of groundwater at the time of the water level measurements. Based on the flow regime expected for areas of saltwater intrusion (i.e. that depicted on Figure 1), it is expected that flow within the Medium Sand unit has a southern component, flow at the transition zone (at the depths of piezometers 102, 104, 106, and 108) has a northern component, and flow in the shallow Silty Sand has a northern component. This general flow pattern is indicated by data from the deep piezometers in the Medium Sand (flow to the southwest is indicated). For the intermediate piezometers, however, the flow direction indicated by the equivalent freshwater head data is to the west, and there is no clear northerly or southerly component of flow. A possible reason for this result is that the screens (1.5 m long) of the intermediate piezometers are located across the saline-fresh transition zone where there is high variability in groundwater flow direction and groundwater density. The details of groundwater flow at the transition zone would not be resolved using data from these piezometers. Figure 15a indicates the possible presence of a nonuniform flow system (equivalent freshwater head contours are curved). The irregular topography of the top of the Silty Clay unit at the base of the aquifer could influence groundwater flow at depth. The presence of the Fraser River could also affect the flow regime at the site (throughout the aquifer). Groundwater flow within the shallow Silty Sand unit may be influenced by the Lulu Island drainage system. The piezometer network at the site is not sufficient to provide data for a thorough analysis of the details of the groundwater flow regime. 42 The regional flow gradient of fresh groundwater toward the river in the area of the site is one factor which will determine the extent to which saline water from the river intrudes the sand aquifer. The piezometer network at the site is not sufficient to determine the fresh groundwater flow gradient toward the river (i.e. in the area south of the wedge toe). A townhouse development is located to the south of the BC Hydro property, in the immediate vicinity of the toe of the wedge, and no piezometers could be located to obtain the river-ward gradient. It is assumed that the fresh groundwater flow gradient toward the river is shallow (based on the approximately flat topography in the area of the site). 3.3.2.4 Vertical Hydraulic Gradients In a variable density groundwater system, equivalent freshwater head does not represent a "potential" and thus the comparison of equivalent freshwater head values for vertically separated piezometers to assess the groundwater flow directions could lead to inaccurate characterization of the flow system (Oberlander, 1989). Equivalent freshwater head data may be used to assess the direction of the vertical component of groundwater velocity if Darcy's Law in terms of equivalent freshwater head is used (Equation 5, Section 4.1) and if the density variation between two measurement points is small. For systems where the density variation between vertically separated measurement points is significant, however, the use of equivalent freshwater head data to assess flow directions could lead to an erroneous interpretation (Oberlander, 1989). Vertical hydraulic gradients have not been determined using equivalent freshwater head data from the zone-specific piezometers at the site. At the locations where vertically separated piezometers are present, a difference in density between the measurement points was observed (calculated density values for each of the zone specific wells are indicated on Table A2-1, Appendix II). In addition, average equivalent freshwater head data were not obtained in vertically separated piezometers over the same time period which would be required to assess the average vertical direction of groundwater flow within the sand aquifer. Beneath the sand aquifer, fresh groundwater of constant density is present within the Silty Clay unit, and data from the West Bay well is available for the calculation of vertical gradients within this unit. Three measurement ports within the Silty Clay allow for hydraulic head measurements. Data from these measurement ports collected in May 1996 indicate an upward gradient of approximately 0.17 in this unit. Using the measured hydraulic conductivity for the Silty Clay of approximately 1 x 10"10 m/s (measured by consolidation tests at the University of Saskatchewan), the groundwater flux upward through this unit is estimated to be 1.7 x 10"11 m/s (0.0005 m/y). Vertical gradients cannot be determined using water level data from piezometer 113 (screened to approximately 2.4 m depth within fresh groundwater of the Clayey Silt) and piezometers 111, 112, and 114 (screened at approximately 4.5m depth within the freshwater zone of the Silty Sand unit). This is because (a) piezometer 113 is not located directly above any other deeper zone-specific piezometers (which would be required to assess vertical gradients in this complex system) and (b) the water level in piezometer 113 will likely never establish equilibrium with groundwater in the adjacent Clayey Silt unit due to the low permeability of this unit. Hourly water level data were not collected from piezometer 113. 3.3.2.5 Recharge and Drainage Controls The water table at the site is approximately 1.5 m below the ground surface. This level is maintained by a drainage control pumping system which is employed across Lulu Island to prevent flooding. Because of this, water table response to seasonal fluctuations in precipitation over the year is likely minimal. 44 3.3.2.6 Tidal and Seasonal Effects on Groundwater Levels The water level within the sand aquifer at the site is tidally influenced. Data collected at the site from zone-specific piezometers 101, 102, 103, 105, 107, 111, and 114 indicate that there is approximately a 3 hour time lag between high and low tide in the river and high and low water levels at the site (Thomas, 1996; Gaganis, 1996). This time lag, however, was determined using river level fluctuation predictions (obtained from the Institute of Ocean Sciences, Sydney, BC) from a station approximately 4.5 km downstream from the field site. Tidal efficiency factors (TEFs) which indicate the change in water level versus the change in river height over a tidal cycle were also determined. A TEF of 0.27 was determined using data from piezometers 101, 102, 103, and 105, and a TEF of 0.23 was determined using data from wells 111 and 114 (Gaganis, 1996). An average transmissivity of approximately 3 x 10"3 was calculated by Gaganis (1996) for the Medium Sand and Fine and Medium Sand units using the TEF and time lag in an equation provided in Erskine (1991). Using an aquifer thickness of 15 m, the hydraulic conductivity is calculated to be 2 x 10"4 m/s. This result is comparable to the hydraulic conductivities determined for the Fine and Medium Sand and Medium Sand units from other tests (discussed in Section 3.3.2.2). Seasonal or short-term (i.e. tidal or due to precipitation) fluctuations in the water table position at the site cannot be assessed based on the available data. 3.4 Groundwater Chemistry 3.4.1 Regional Distribution of Saline Groundwater Groundwater chemical data for the Fraser River delta is limited. As discussed in Section 2.3, borehole electrical conductivity logging and ground electrical conductivity surveys have been completed at locations across the Fraser River delta by Hunter et al. (1994, 1996). These data can be used to identify areas of saline groundwater (indicated by high electrical conductivity measurements). Groundwater analyses for major ions (also indicate areas of saline groundwater) have also been completed as part of a geochemical study of the Fraser River delta by Simpson and Hutcheon (1995). Figure 16 illustrates an interpretation of the available electrical conductivity and chemical data showing the depth to the top of saline groundwater (as indicated by an increase in electrical conductivity measurements or chloride concentration). Note that the contours presented on Figure 16 have been inferred using available data points as indicated on Figure 16. At locations where no data (or only sparse data) are available (e.g. along the south shore of Lulu Island), contours indicating the depth to saline groundwater are conceptual and have been sketched to indicate areas of possible saltwater intrusion based on the presence of saline water in adjacent water bodies. The available data indicate the widespread presence of saline groundwater across the delta within the sand aquifer and/or within the deeper delta slope sediments. The presence of saline groundwater within the shallow, more-permeable deltaic sediments is inferred to be the result of saltwater intrusion from the Georgia Strait or Fraser River. Saline water within the deeper and finer grained deltaic sediments may be trapped seawater as suggested by Simpson and Hutcheon (1995). Figure 16 identifies locations where available data indicate the presence of fresh groundwater or the presence of fresh groundwater beneath saline groundwater. The presence of fresh groundwater at these locations is inferred to be the result of regional flow of fresh groundwater from upland areas surrounding the delta (See Figures 10 and 11). Further discussion of the regional distribution of saline and fresh groundwater and the regional hydrogeology is provided in Section 6.2.1. 3.4.2 Site Groundwater Analytical Results The geochemistry of the groundwater at the site has been determined through field and laboratory analyses of samples from the MLS wells, West Bay well, zone-specific piezometers, and CPTs. Analyses have included temperature, pH, salinity, electrical conductivity, resistivity, total dissolved solids (TDS), calcium, potassium, magnesium, sodium, strontium, iron, chloride, bicarbonate and sulfate. Salinity, electrical conductivity, temperature, pH, and resistivity were measured in the field. Laboratory analyses of the samples for other environmental water quality parameters were completed at the UBC Environmental Engineering Laboratory or University of Calgary. Table A3-1, Appendix III, summarizes the specific analyses completed for the groundwater samples collected at the site, and identifies the laboratories which analyzed the samples. Tables A3-2 to A3-6 , Appendix III, summarize the analytical results for temperature (Table A3-2), pH (Table A3-3), electrical conductivity (Table A3-4), salinity (Table A3-5), and TDS, cations, and anions (Table A3-6). CPT resistivity measurements are indicated on the CPT logs, Appendix I. The groundwater chemical results allow for the delineation of the saline wedge (i.e. wedge position, shape, and transition zone characteristics) and also provide information regarding the chemical characteristics of fresh and saline groundwater at the site. The following presents and discusses the groundwater chemical data. Temperature: Groundwater temperature was measured immediately upon sample collection from MLS wells W l , W2, and W3 on December 17, 1996 and March 24, 1997. These results are presented in Table A3-2. The average groundwater temperature is estimated based on the December and March results to be approximately 10.7 °C. The sampling results indicate that the groundwater temperature is essentially constant with depth at the sampling locations. PH: The pH of the groundwater samples from wells W l , W2, and W3 was measured on May 1, 1996 and July 30, 1996. The results are presented on Table A3-3. Measurements of pH versus depth for W3 are indicated on Figure 17. On both sampling occasions, the pH of the groundwater sampled from W3 ranged from approximately 6 to 8 pH units. In well W3 a higher pH was measured in groundwater from the deeper measurement ports (i.e. W3-14 and W3-15) than for samples from shallower depths. The pH in the deeper sampling ports at W3 is influenced by the completion of W3 within the Silty Clay unit at depth. Higher pH readings were not observed in the deeper measurements ports for wells W l and W2 as these wells do not intersect the Silty Clay unit. Salinity: Water salinity is defined as "the ratio of mass of dissolved material in seawater to the mass of seawater" (UNESCO, 1981). This ratio, however, is difficult to measure for water samples, and thus salinity is actually measured as the electrical conductivity ratio of the sample and a standard KC1 solution (UNESCO, 1981). An LF 320 Handheld Conductivity Meter, manufactured by Wissenschaftlich-Technische Werkstatten, was used to measure groundwater salinity at the site. Salinity measurements are reported in parts per thousand (ppt). Salinity was measured in the groundwater sampled from W l , W2, and W3 on July 30, 1996, October 1, 1996, December 17, 1996, and March 24, 1997. The salinity measurements are presented in Table A3-4. Figures 18a, 18b, and 18c, show the changes 0 Figure 18a: Salinity Data for W1 Note: No data for W1 at depths of 11 and 12 m below ground surface. July 30, 1996 October 1, 1996 December 17, 1996 March 24, 1997 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Salinity (ppt) 5 f Figure 18b: Salinity Data for W2 Note: No data for W2 at depths of 9,11, 12, 13, 14, 15, 16, 19, and 22 m below ground surface. July 30, 1996 October 1, 1996 December 17, 1996 March 24,1997 0.0 2.0 6.0 6.0 10.0 Salinity (ppt) 12.0 14.0 Figure 18c: Salinity Data for W3 July 30, 1996 October 1,1996 December 17, 1996 March 24, 1997 Note: Data available for all measurement depths. 0.0 2.0 6.0 8.0 10.0 Salinity (ppt) 12.0 14.0 Figure 18: Salinity versus depth for MLS wells (a) W1, (b) W2, (c) and W3. in salinity with depth for wells W l , W2, and W3 respectively. The data indicate that the shallow fresh groundwater above the saline wedge has a salinity ranging from approximately 0.15 to 2.10 ppt and extends to a depth of approximately 11 m at each of the MLS well locations. From approximately 11 to 13 m depth at MLS wells W l and W3, a transition zone is present where salinity increases from approximately 2 to 13 ppt. (Data for W2 is not sufficient to delineate the transition zone.) Beneath this transition zone, the salinity is essentially constant with depth at values ranging from approximately 13 to 15 ppt. In MLS well W3, a deeper transition zone is identified at approximately 18 m depth where the salinity decreases with depth from approximately 14 to 1.5 ppt over a vertical distance of 2.5m. Below 20.5 m depth in W3, fresh groundwater is present within the Silty Clay unit. The other MLS wells (Wl and W2) do not intersect fresh groundwater at depth as they are completed above the Silty Clay. Salinity was also measured in groundwater from the zone-specific piezometers (101 through 114) on July 30, 1996. Piezometers 101 to 108 were installed with 1.5 m long well screens and thus do not provide data to resolve the changes in salinity across the transition zone. The salinity data from the piezometers, however, do confirm the presence of a shallow zone of fresh groundwater above the saline wedge (Piezometers 111 to 114), an intermediate zone of brackish to saline groundwater at the approximate depth of the transition zone (11.5 to 13 m depth in Piezometers102, 104, 106, and 108), and the presence of saline groundwater within the "core" of the saline wedge (piezometers 101, 103, 105, and 107). Electrical Conductivity: Electrical conductivity is defined as "the ability of a substance to conduct an electrical current" and is a field measured parameter which provides a general indication of TDS (Freeze and Cherry, 1979). As such, electrical conductivity indicates the presence of saline groundwater (high TDS). An LF 320 Handheld Conductivity Meter was used to measure groundwater electrical conductivity at the site. The units of measurement for electrical conductivity are mS. Electrical conductivity, while not providing any additional information to that obtained from the salinity measurements, relates linearly to salinity allowing for the development of an electrical conductivity-salinity relationship for groundwater samples at the field site (Figure 19a). Electrical conductivity was measured for MLS wells W l , W2, and W3 on July 30, 1996, October 1, 1996, December 17, 1996, and March 24, 1997. The results are presented in Table A3-5. Electrical conductivity values less than approximately 2 mS were measured for fresh groundwater samples collected above the saline wedge, conductivities between 2 mS and 19 mS were measured for brackish groundwater of the transition zone, conductivities greater than approximately 19 mS are indicative of saline groundwater of the saline wedge, and conductivities less than 3.5 mS (indicating fresh groundwater) were measured in W3 for sampling ports located within the deep Silty Clay unit. Resistivity: During the cone penetrometer tests (CPTs) completed at the site by the UBC Civil Engineering Department, resistivity measurements were obtained over the depth of the CPT. Low resistivity indicates relatively high salinity, and high resistivity indicates low salinity (i.e. fresh groundwater). The resistivity data (in units of ohm-m) or interpretations of resistivity data identifying the position of transition zone (for some CPTs the raw resistivity data were not available but interpretations indicating the approximate depth of the transition zone were available) are presented on the CPT logs in Appendix I. For CPTs numbered K9301, K9308, K9309, K9310, K9311, K9601, K9602, K9603, K9612, K9701, and K9801, the location of the fresh-saline transition zone marking the top of the saline water wedge is indicated by a relatively sharp decrease in resistivity. At the base of the saline wedge, the location of the transition zone from saline groundwater to deep fresh groundwater is indicated by an increase in resistivity. The CPT results indicate the thickness of the transition zone bounding the top of the saline wedge is approximately 2 m and the thickness of the transition zone at the base of the wedge is 2m or greater « •X 15.00 3 •a c o if 10.00 UJ 5.00 0.00 a) Electrical Conductivity vs. Salinity Trendline y= 1.5184x R2 = 0.9976 0.0 2.0 4.0 6. 0 8.0 10.0 Salinity (ppt) 12.0 14.0 16.0 20000 18000 16000 14000 12000 £ . 10000 co P «rmn 6000 4000 2000 b) TDS vs. Salinity 0.0 2.0 Trendline y= 1242.1 x R2 = 0.9969 6.0 8.0 10.0 Salinity (ppt) 12.0 14.0 16.0 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 c) Chloride vs. Salinity Trendline y = 565.59X R2 = 0.9987 0.0 2.0 6.0 8.0 10.0 Salinity (ppt) 12.0 14.0 16.0 igure 19: Graphs showing the relationship between salinity and (a) electrical conductivity, (b) TDS, and (c) chloride (some CPTs were not completed to sufficient depths to define the lower transition zone thickness). Total Dissolved Solids (TDS): TDS ranging from 0 to 1000 mg/L indicates fresh water, TDS ranging from 1000 to 10 000 mg/L indicates brackish water, and TDS ranging from 10 000 to 100 000 indicates saline water (Freeze and Cherry, 1979). TDS, while not providing any additional information to that obtained from the salinity measurements for the wells tested, relates linearly to salinity allowing for the development of a TDS-salinity relationship for the groundwater samples at the field site (Figure 19b). Samples from the zone-specific piezometers (101 to 108 and 111 to 114) were analyzed for total dissolved solids (TDS). As expected, the TDS analyses indicate fresh groundwater in the shallow zone-specific piezometers (111, 112, 113, and 114), brackish to saline groundwater for the intermediate depth piezometers (102, 104, 106 and 108), and saline groundwater in deep zone-specific piezometers (101, 103, 105, and 107). Major Cations and Anions: Major cations (Na+, K + , Mg 2 + , and Ca2+) and major anions (Cl\ HC0 3", and S042") were analyzed to investigate the individual ion contributions to groundwater salinity at the site and to characterize the groundwater geochemistry. The concentrations of one or more major cations/anions were measured for samples from zone-specific piezometers, the West Bay well, and CPTs K9701, K9801, K9802, and MLS well W3 as indicated on Table A3-6. Figures 20 to 28 present the relationship between calcium, magnesium, sodium, potassium, chloride, sulphate, and bicarbonate concentrations (where measured) and sample depth for the West Bay well, zone-specific piezometer groups (identified below), and CPTs K9701, K9801, K9802, and W3. (Additional points labelled SW on Figure 20 are referenced in Section 3.4.4 of this thesis). The West Bay well, CPTs, and MLS well Figure 20: West Bay Well Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix III. SW = Surface water sample from Georgia Strait (Simpson and Hutcheon, 1995) C a l c i u m C o n c e n t r a t i o n (mg/L) 0 100 200 300 400 500 icentralton (mgyL) 0 200 400 600 800 1000 S o d i u m C o n c e n t r a t i o n (mgn.) 0 2000 4000 6000 B00I "1 ' ' 10 J P o t a s s i u m C o n c e n t r a t i o n (mg/L) S t ron t ium C o n c e n t r a t i o n (mg/L) I ron C o n c e n t r a t i o n (mg/L) 0 50 100 150 200 250 300 0 2 4 6 8 10 0 100 200 300 400 500 T 10 J C h l o r i d e C o n c e n t r a t i o n (mg/L) 5000 10000 15000 £ 20-1 Sulfate C o n c e n t r a t i o n (mg/L) 0 500 1000 1500 2000 Bicarbonate Concentration (ntofL) 0 200 400 600 800 1000 1200 Figure 21: Zone-Specific Welis 101, 102, and 112 Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix lit. Calcium Concentration (mg/L) 0 100 200 300 400 500 Magnesium Concentration (mg/L) 0 200 400 600 800 1000 Sodium Concentration (mg/L) 0 2000 4000 6000 8000 Potassium Concentration (mg/L) 0 50 100 150 200 250 300 Strontium Concentration (mg/L) 0 2 4 6 8 10 Iron Not Analyzed Chloride Concentration (mg/L) 0 5000 10000 15000 Sulfate Concentration (mg/L) 0 500 1000 1500 2000 Blcarbonit i Concentration (mg/L) 100 120 0 200 400 600 800 0 0 Figure 22: Zone-Specific Wells 103, 104, and 111 Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix III. Calcium concentration (mg/L) 0 100 200 300 400 500 Magnesium Concentration (mgfL) 0 200 400 600 800 1000 •111 Sodium Concentration (mg/L) 0 2000 4000 6000 8000 Potassium Concentration (mgfL) 0 50 100 150 200 250 300 Strontium Concentration (mgfL) 0 2 4 6 6 10 Iron Not Analyzed Chloride Concentration (mg/L) 0 5000 10000 15000 Sulfate Concentration (mg/L) 0 500 1 000 1500 2000 Bicarbonate Conctntrrtlon (mijrL) 0 200 400 600 800 1000 1200 s »104 I Figure 23: Zone-Specific Wells 105,106,113,114 Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix III. Calcium concentration (mg/L) 0 tOO 200 300 400 500 •113 '•114 Magnesium concentration (mg/L) 0 200 400 600 800 1000 •113 •114 Sodium concentration (mg/L) 0 2000 4000 6000 80( Potassium Concentration (mg/L) 0 50 100 150 200 250 300 •113 • 114 1 I Strontium Concentration (mg/L) 0 2 4 6 8 10 ,113 '•114 . Iron Not Analyzed Chloride Concentration (mg/L) 0 5000 10000 15000 • 105 Sulfate Concentration (mg/L) 0 500 1000 1500 2000 •113 • 105 • Concantrstion (mg/L) 0 200 400 600 800 1000 1200 •113 •114 £ »106 •105 Figure 24: Zone-Specific Wells 107 and 108 Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix III. Calcium Concentration (mg/L) 0 100 200 300 400 500 Magnesium Concentration (mg/L) 0 200 400 600 BOO 1000 Sodium Concentration (mg/L) 0 2000 4000 6000 8000 Potassium Concentration (mg/L) 0 50 100 150 200 250 300 Strontium Concentration (mg/L) 0 2 4 6 8 10 Iron Not Analyzed Chloride Concentration (mg/L) 0 5000 10000 15000 Sulfate Concentration (mg/L) 0 500 1000 1500 2000 Bicarbonate Concintntlon (mg/L) 0 200 400 600 800 1000 1200 25 25 25 Figure 25: K9701 Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix III. Calcium concentration (mgfL) 0 100 200 300 400 500 Magnesium Concentration (mgfL) 0 200 400 600 800 1000 Sodium Concentration (mg/L) 0 2000 4000 6000 BC s i Potassium Concentration (mg/L) 0 50 100 150 200 250 300 Strontium Concentration (mg/L) 0 2 4 6 8 10 Iron Not Analyzed Chloride Concentration (mg/L) 0 5000 10000 15000 Sulfate Concentration (mg/L) 0 500 1000 1500 2000 BtcBrbonat* Concentration (mg/L) 0 200 400 600 800 1000 1200 Figure 26: K9801 Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix III. Calcium Concentration (mg/L) 0 100 200 300 400 S00 Magnesium Concentration (mg/L) 0 200 400 600 800 1000 n Concentration (mg/L) 0 2000 4000 6000 8000 S 8 Potassium Concentration (mg/L) 0 50 100 150 200 250 300 Strontium Concentration (mg/L) 0 2 4 6 8 10 Iron Not Analyzed Chtortde Concentration (mg/L) 0 5000 10000 15000 Sulfate Concentration (mg/L) 0 500 1000 1500 2000 Blcsrbonati Concentration (mg/L) 0 200 400 600 800 1000 1200 E f Figure 27: K9802 Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix III. Chloride Concentration (mg/L) 0 5000 10000 15000 0 1 Figure 28: W3 Groundwater Chemistry vs. Sample Depths Detailed results can be referenced in Appendix III. Chloride Concentration (mg/L) 0 5000 10000 15000 25 Figure 29: Ion-Chloride Relationship for Ca + + , Mg + +, Na+, K+, Sr + +, Fe + +, S0 4 " , and HC03" Chloride Concentration (mg/L) 700 600 Chloride Concentration (mg/L) Chloride Concentration (mg/L) y = 0.0007X R 2 = 0.8335 0 2000 4000 6000 8000 10000 Chloride Concentration (mg/L) 0 2000 4000 6000 8000 10000 Chloride Concentration (mg/L) 0 2000 4000 6000 8000 10000 Chtortde Concentration (mg/L) 0 2000 4000 6000 8000 10000 0 2000 4000 6000 8000 10000 Chloride Concentration (mg/L) Chloride Concentration (mg/L) W3 provide groundwater chemical data for discrete depths at a single site location. The zone-specific piezometer groups consist of nests of 2 to 4 individual piezometers in close proximity which are screened across different depth intervals. The zone-specific piezometer groups are: (101, 102, 112), (103, 104, 111), (105, 106, 113, 114), and (107,108). In general, relatively low concentrations of all cations were observed for samples collected of fresh groundwater at depths above and below the saline wedge, and higher cation concentrations were measured for samples collected from within the saline wedge (i.e. for brackish and saline groundwater). As expected, sodium was the most abundant cation in the samples of the saline groundwater with concentrations as high as 4200 mg/L. Potassium was present at relatively low concentrations in the saline groundwater (less than 100 mg/L) and magnesium and calcium were present at slightly higher concentrations (approximately 300 to 400 mg/L). To illustrate the relationship between the different major cation concentrations and salinity, graphs of cation concentration vs. chloride (relates linearly to salinity as shown on Figure 19c) are presented on Figure 29. Figure 29 shows that all major cations increase with increasing chloride/salinity in the groundwater at the site. For the anion concentrations, chloride provided the greatest contribution of the major anions to the high TDS of saline groundwater (as would be expected); with concentrations as high as 9480 mg/L. (Figure 19c illustrates the linear relationship between chloride and salinity for groundwater samples at the site). Sulfate was present in the saline groundwater in relatively moderate amounts (approximately 1000 mg/L). The relationship of increasing sulfate concentration with increasing chloride concentration or salinity is indicated on Figure 29. With respect to bicarbonate, relatively low concentrations were observed for the saline groundwater (generally at or less than 100 mg/L), and higher bicarbonate concentrations were measured in the fresh groundwater samples. The highest bicarbonate concentration was measured for samples of the deep fresh groundwater (up to approximately 1000 mg/L in samples from the West Bay well). The bicarbonate results (as indicated on Figure 29) do not indicate a trend of increasing bicarbonate with increasing salinity as was observed for the other ions. Simpson and Hutcheon (1995) suggest that the concentration of bicarbonate in the groundwater of the Fraser River delta is affected by diagenetic reactions and is not related to salinity. The relative major cation and anion concentrations (in meq/L) for groundwater samples analyzed from the site (for which all major cations and anions were analyzed) have been plotted on a trilinear diagram (Figure 30). The trilinear diagram provides a convenient graphical method to distinguish between different geochemical groundwater facies by groupings of data points on the diagram (Freeze and Cherry, 1979). On Figure 30, samples of saline groundwater, brackish groundwater, shallow fresh groundwater (above the saline wedge), and deep fresh groundwater (within the Silty Clay unit) have been plotted using different symbols to illustrate the chemical differences between these groundwater facies. For the saline groundwater samples analyzed from the site, the relatively high sodium concentration (compared to magnesium and calcium) causes these samples to cluster in an area on the lower right portion of the cation plot on Figure 30. With respect to the anion plot of Figure 30, the saline groundwater samples also plot at the lower right side due to relatively high proportion of chloride compared to other ions. As groundwater becomes less brackish (i.e. at the transition zone) the relative sodium and chloride concentrations decrease. The brackish groundwater samples plot in a zones to the upper-left of the saline groundwater samples on the cation and anion plots of Figure 30. The points on the trilinear diagram representing the shallow fresh groundwater plot in a widespread zones in the upper central and central portions of the cation and anion plots respectively. Samples of deep fresh groundwater within the Silty Clay unit contain relatively high concentrations of sodium compared to magnesium and calcium and thus 66 100 Figure 30: Trilinear diagram for site groundwater samples. © Shallow Fresh Groundwater 4- Brackish Groundwater + Saline Groundwater <> Deep Fresh Groundwater plot in a distinct zone at the lower right corner of the cation plot. The deep fresh groundwater samples plot in a zone at the bottom left side of the anion plot which is characterized by relatively high bicarbonate, low to moderate chloride, and low sulphate. The combined cation-anion plot on Figure 30 further illustrates the differences between the bulk chemistry of shallow fresh, brackish, saline, and deep fresh groundwater at the site. Each of the different groundwater types or facies (shallow fresh, brackish, saline, and deep fresh) plots in a discernible zone on the combined cation-anion plot. The trilinear diagram indicates that the deep fresh groundwater exhibits a characteristic chemistry that is distinguishable from shallow fresh groundwater samples. As discussed in Section 3.3, the shallow fresh groundwater at the site is recharged by precipitation and forms local groundwater flow systems within the shallow geological units at the site. The deep fresh groundwater represents groundwater at the downstream end of the intermediate or regional groundwater flow system (source for this water is inferred to be recharge to the upland areas to the north of the field site). Due to the longer flow path for the deep groundwater and the flow of this water through different geological units, more geochemical reactions would be expected to occur to change the chemistry from the initial recharge water chemistry. The shorter flow path of shallow fresh groundwater would be expected to limit the geochemical changes to the groundwater prior to sampling. Strontium: Seawater contains higher concentrations of strontium than freshwater and thus the strontium concentrations at the site can be used to identify areas of saltwater intrusion from the Fraser River. Strontium was analyzed in groundwater samples from the West Bay well, zone-specific piezometer groups, and CPTs K9701 and K9801 and the strontium concentration versus depth at these locations are presented on Figures 20 to 26. Relatively high strontium concentrations are observed for samples of saline groundwater when compared to fresh groundwater at the site (the positive relationship between strontium and chloride is indicated on Figure 29). The relatively low strontium concentrations in fresh groundwater sampled from the deeper Silty Clay unit (samples from the West Bay well), confirm the non-marine origin of water in the Silty Clay. Iron: Iron was analyzed to investigate elevated iron concentrations which were indicated by the formation of iron precipitate at the time of groundwater sampling. Iron was analyzed in selected samples from the West Bay well. Figure 20 presents the iron concentrations versus depth for the West Bay well. Figure 29 illustrates the week relationship between iron and chloride which suggests that iron concentration is not related to salinity. Chemical results for a samples of saline water from the Georgia Strait and from the west side of Vancouver Island (Simpson and Hutcheon, 1995) indicate very low concentrations of iron (0.2 mg/L). The low iron concentrations in seawater (the source for saltwater intrusion at the site) further indicate that the observed iron concentrations at the site are not related to salinity but are due to geochemical reactions in the subsurface environment. In general, the iron levels at the site can be considered relatively high due to the observed formation of iron precipitate immediately upon sampling. The dissolved iron in the groundwater indicates anaerobic or reducing conditions in the subsurface possibly due to the degradation of organic matter within the sediments. This is also suggested by low dissolved oxygen concentrations in the order of 0.1 mg/L which were measured in the aquifer (L. Smith, personal communication). Note that the charge balance errors on Table A3-6 are calculated based on concentrations of the major ions and the iron concentrations are not included in the charge balance error calculations. Iron increases the charge balance error by up to 11% for the samples analyzed from the West Bay well. Iron concentrations had a much greater influence on the charge balance error for fresh groundwater samples of low ionic strength than for brackish or saline groundwater samples. Note that for low ionic strength solutions, higher charge balance errors can be tolerated as they are caused by small measurement errors of a few mg/L (Simpson and Hutcheon, 1995). 69 3.4.3 Saline Wedge Position at Field Site Using the groundwater chemical data obtained at the field site, the position of the saline wedge in cross-section perpendicular to the Fraser River can be delineated. Figure 31 presents an interpretation of the saline groundwater wedge position based on the available chemical data. Salinity contours on Figure 31 were estimated based on the following data. • Salinity measurements from W l , W2, and W3. • The chloride concentration in groundwater sampled at CPTs K9701 K9801, and K9802. These chloride data were converted to salinity using the chloride-salinity linear relationship indicated on Figure 19c. • Resistivity data for CPTs K9309, K9301, K9612, K9602, K9601, K9603, K9308, K9311, and K9310 was used to identify the location and thickness of the upper and lower transition zones bounding the saline wedge. • Groundwater chemical data from the zone-specific piezometer groups and the West Bay well. As discussed previously, the groundwater chemical data indicate that a saline wedge extends approximately 500 m inland (i.e. to the south) from the Fraser River (Figure 31) into the Sand Aquifer beneath the field site. Above the saline wedge, fresh groundwater is present (within the Silty Sand unit) and the saline wedge is bounded at the base by the Silty Clay unit where fresh groundwater is present. Salinity contours depicted on Figure 31 indicate that the shape of the wedge is oblong; the transition zone slopes gently to the south (inland) and then dips more steeply to form a blunt toe. Note, however, that the vertical scale of Figure 31 is approximately 10 times greater than the horizontal scale which affects the visualization of the shape of the saline wedge. The position of the saline wedge at the site may change slightly over the year due to seasonal changes in precipitation and river conditions. The groundwater salinity data for W l , W2, and W3 was collected on four different sampling occasions and thus provide an initial indication of the seasonal changes in saline wedge position at the site. The data are presented on Figures 18a, 18b, and 18c, and indicate that, in general, slightly lower salinities were measured for each depth in the MLS wells when sampled in December (1996) and March (1997) (winter and early spring), and higher salinities were observed during July (1996) and October (1996) (summer and fall). The data indicate that the saline water wedge may change position slightly over the year due to seasonal changes in the river (river level and salinity would be affected as discussed in Section 3.5) and/or seasonal precipitation across Lulu Island. As discussed in Section 2.1 and illustrated on Figure 6 an increase in precipitation over the winter months could cause the wedge to move slightly toward the river (resulting in lower salinities measured at the site in winter and early spring), less precipitation in the summer and fall could conversely cause an increase in observed salinity measurements at the site. Note, however, that drainage controls employed across Lulu Island (discussed in Section 3.3.2.5) could reduce the effect of seasonal changes in precipitation on the saline water wedge position by limiting recharge to the sand aquifer in the area of the site. Seasonal changes in river salinity and water level are complex. As discussed in Section 3.5, in general during low river flow conditions (during the fall and winter months), lower water levels and an overall higher salinity would be observed. During high river flow conditions (spring and summer), higher water levels and overall lower river water salinity would be expected. Tidal action also influences river water level and salinity over the year and further complicates the generalization of seasonal influences. Available field data are not sufficient to analyze the effects of seasonal (or tidal) changes in the Fraser River conditions on the transient behaviour of the saline wedge at the site. Over time, saltwater intrusion from the estuary may establish steady state or average salinity levels in the aquifer and the saline wedge position may not exhibit transient behaviour that can be measured using field chemical data. 72 3.4.4 Comparison to Regional Groundwater and Surface Water Chemistry The study by Simpson and Hutcheon (1995) involved the collection and analysis groundwater samples from boreholes and surface water samples from the Fraser River and the Georgia Strait. The chemistry of selected groundwater and surface water samples from Simpson and Hutcheon (1995) is presented on Table A3-7, Appendix III. The samples from Simpson and Hutcheon (1995) are plotted on a trilinear diagram (Figure 32). Samples A, B , and E on Table A3-7 represent samples from a borehole (D4on Figure 16) at the eastern end of Sea Island (adjacent to the North Arm of the Fraser River). Sample A represents shallow fresh groundwater, Sample B represents saline groundwater, and Sample E represents deep fresh groundwater. Samples B and E plot in the same respective zones on the trilinear diagram (Figure 32) as the samples of the different groundwater facies from the field site (Figure 30). This suggests that similarities exist between the geochemistry at the field site and that at the eastern end of Sea Island (approximately 1.5 km to the west of the field site). Sample A, however, does not plot in the same area as the shallow fresh groundwater samples from the field site as it contains slightly a slightly lower proportion of sodium. Simpson and Hutcheon present chemical data for another borehole located on the west side of Lulu Island (D5 on Figure 16). The chemical results from this borehole did not identify saline groundwater within the sand aquifer. Saline groundwater was encountered in this borehole, however, at deeper depths (greater than 19m depth) within the finer grained delta slope and prodelta deposits. The chemical results for two saline groundwater samples from 32 m (Sample P) and 45 m (Sample Q) depth in borehole D5 are plotted on Figure 32. These saline groundwater samples plot in a slightly lower zone on the anion plot of Figure 32 than saline groundwater samples from the site due to lower 100 Figure 32: Trilinear diagram for selected groundwater samples from Simpson and Hutcheon (1995). CD Shallow Fresh Groundwater A shallow fresh groundwater from 2.87 m depth from a well (D4 on Figure 15) at the eastern end of Sea Island. +• Saline Groundwater B saline groundwater from 19.2 m depth from a well (D4 on Figure 15) at the eastern end of Sea Island. P saline groundwater from 31.1m depth from a well (D5 on Figure 15) atthe western side of Lulu Island. Q saline groundwater from 44.7 m depth from a well (D5 on Figure 15) at the western side of Lulu Island. <J>. Deep Fresh Groundwater E deep fresh groundwater from 36.27m depth from a well (D4 on Figure 15) at the eastern end of Sea Island. Surface Water rwV9 a surface water sample of fresh river water from a location near the mouth of the Middle Arm of the Fraser River. SW a seawater sample collected from the Georgia Strait west of Sea Island (inferred to be approximately 2.5 km off shore). Marine a seawater sample from the west coast of Vancouver Island _ sulfate concentrations. The samples also contained significantly higher concentrations of ions than the samples of saline groundwater from the field site (chloride concentrations in samples P and Q are approximately two times greater than the saline groundwater samples from the field site). The data suggest that there are slight differences in chemistry between saline groundwater within the sand aquifer at the site and deeper saline groundwater within the delta slope deposits. The surface water samples (samples rwV9, SW, and Marine) from Simpson and Hutcheon (1995) referenced on Table A3-7 have also been plotted on Figure 32. The seawater samples (SW and Marine) plot in the same zone on Figure 32 as the saline groundwater samples on Figure 30. This shows that the saline groundwater at the site contains the same general proportions of major ions as seawater. Surface water sample rwV9 (fresh river water) plots in the same general zone as the shallow fresh groundwater samples. This indicates that the chemistry of the shallow fresh groundwater may be more similar to the fresh river water than to the deep fresh groundwater (although the shallow fresh groundwater and fresh river water may not be physically connected). The individual ion concentrations for the surface water sample SW have been compared to the concentrations measured in the saline groundwater in the West Bay Well (Figure 20). The concentrations of potassium, magnesium, sodium, strontium, and chloride in the saline groundwater at the West Bay well are less than those measured for seawater sample SW. This would be expected as, in general, the concentration of ions within the river adjacent to the site (source for saltwater intrusion) would be lower than that of seawater in the Georgia Strait (due to the mixing of seawater and fresh river water in the river). Dilution factors were calculated as 100% minus the percentage of the concentration of an ion in groundwater divided by the concentration of the same ion in sample SW. Based on the available site data and data for sample SW, the dilution factor for chloride (considered a conservative or non-reactive ion) is calculated to be approximately 38-45%. Dilution factors for other ions are calculated to be approximately: 14-27% (for Sr), 45-50% (for Na), 43-62% (for Mg), and 68-77% (for K). The differences in the dilution factors for the different ions are likely due to geochemical reactions in the subsurface (e.g. ion exchange, precipitation, etc.). Note that the calcium, sulfate, and iron concentrations at the site exceeded sample SW, If the calcium, sulfate, and iron concentrations in the saline groundwater at the site were purely the result of seawater intrusion from the Fraser river, it would be expected that the concentrations of these ions would be less than sample SW (as discussed previously). The observed higher calcium and sulfate concentrations in the saline groundwater at the site are attributed to geochemical reactions in the subsurface. Simpson and Hutcheon (1995) also suggest that subsurface geochemical reactions influence the concentrations of sulfate and calcium in the groundwater of the Fraser River delta. 3.5 Hydrologic and Geochemical Characteristics of the Fraser River The Fraser River is subjected to many influences which cause continued changes in the river's physical and chemical state. An understanding of the physical and chemical properties of the river and how they change with time is important to study the hydrogeology in areas of saltwater intrusion adjacent to the Fraser River. This is because the river represents a system boundary. Conditions at this boundary determine the features of the saline groundwater wedge where saltwater intrusion from the.river occurs. The following subsections present a summary of the available information regarding the hydrogeologic and geochemical characteristics of the Eraser River near the.field site. 3.5.1 Fraser River Salinity Data Seawater from the Georgia Strait intrudes the Fraser River forming a "wedge" of saline water beneath the discharging fresh river water. Data have been published providing salinity measurements in the Fraser River for a station at the Oak Street Bridge; approximately 800m downstream.ffom the field site. .Eraser.River salinity time series measurements were recorded on November 23-24, 1977 (Ages, 1979), and on April 19-20, 1978 (Ages, 1988). Measurements were recorded approximately hourly and were obtained at 1 m depth intervals. The salinity profile in the river near the site changes over the tidal cycle, and seasonally with river discharge, as the saline water wedge migrates up and down the river. This is illustrated in Figure 33a for one tidal cycle using data collected on November 23 and 24, 1977 by Ages (1979), and in Figure 33b using data collected on April 19 and 20, 1978 (Ages, 1988). At high tide the saline wedge migrates up the river increasing river water salinity near the site. As the tide lowers, the saline wedge in the river moves seaward causing a decrease in river salinity adjacent to the site. At low tide, only fresh river water is present adjacent to the site. The influence of seasonal.river discharge rate on the salinity in the Eraser River adjacent to the field site is evident by comparison of Figures 33a and 33b. At low river discharge (Figure 33a, discharge at Hope is 780 m3/s) the saline water wedge intrudes the river a greater distance and thus generally higher salinity values are observed adjacent to the field site. At higher river discharge (Figure 33b, discharge at Hope is 1710 m3/s), the saline water does not intrude the river to the same degree and lower values of salinity are observed near the field site. Figure 34 further illustrates the relationship between river discharge rate and the high-tide extent of saline water. Figure 34 indicates that at discharge rates greater than approximately 2800 m3/s, only fresh water is present in the river adjacent to the field site. The combination of the individual influences of the river tidal fluctuations and the changes in seasonal river discharge rate described above will determine the position of the saline river wedge at a given time (and thus the salinity in the river adjacent to the field site). In general at high tide and low river discharge the wedge intrudes the river further producing 77 0.0 - • 1.0 - • 2.0 ti 3.0 ti 4.0 n 5.0 i i 6.0 i i 7.0 n 8.0 u 9.0 i i 0.0 n 1.0 n 2.0 Figure 33a: Fraser River Salinity Data - Depth Profiles Over Tidal Cycle Salinity Data Apr. 19-20,1978 (Ages, 1988) Discharge at Hope = 1710 m3/s -High Tide Salinity Profile •Mid Tide Salinity Profile • Low Tide Salinity Profile 1— 10.0 15.0 Salinity (ppt) 0.0 20.0 25.0 0.0 1.0 2.0 3.0 ^ 4.0 5.0 IA 6.0 U 7.0 8.0 9.0 10.0 11.0 12.0 Figure 33b: Fraser River Salinity Data - Depth Profiles Over Tidal Cycle Salinity Data Nov. 23,1977 (Ages, 1979) Discharge at Hope = 780 m3/s -High Tide Salinity Profile -Mid Tide Salinity Profile - Low Tide Salinity Profile -t- -t-0.0 5.0 10.0 15.0 Salinity (ppt) 20.0 25.0 78 higher salinity near the site than at low tide and high river discharge. The Fraser River data indicate, however, that the river salinity adjacent to the .field site changes over time in a complicated manner. Even at a given time, the salinity in the river adjacent to the field site is not constant over the depth of the river (as indicated on Figures 33a and 33b). In addition, the salinity is not uniform laterally across the river as the toe of the saline wedge in the river advances further and faster in the centre of the river channel than at the edges (Ages, 1988). 3.5.2 River Water Temperature Temperature data for the Oak Street Bridge station was also collected at the time of the salinity time series measurements on November 23-24, 1977 (Ages, 1979), and April 19-20, 1978 (Ages, 1988). The data indicate that the temperature of the river water near the site changes over time and over depth depending on the migration of the saline river wedge. The temperature range for the measurements collected on November 23-24, 1977 was approximately 1.5 °C to 6.1 °C. Over this period, higher temperatures were generally present within the saline wedge and at the base of the river, and lower temperatures were measured for shallow seaward flowing fresh water. In April 1978, the temperature range was 6.7°C to 8.0°C, with little difference between fresh and saline river water. 3.5.3 River Water Density - The Salinity-Temperature-Density Relationship The water density in the river is a factor which influences the density-dependent groundwater flow system and saltwater intrusion at the field site. Water density can be calculated using salinity and temperature measurements and the International Equation of State for Seawater (UNESCO, 1981). Figure 35 shows the relationship between density, salinity and temperature for water as determined using the International Equation of State Figure 35: Water Density Variat ion wi th Temperature and Sal ini ty International Equat ion of State fo r Seawater (Unesco, 1981) 1020 - I - . Salinty (ppt) for Seawater. The density of water varies significantly with changes in water salinity. From Figure 35, a 5 ppt change in water salinity results in approximately a 4 kg/m3 change in water density. Temperature also affects water density, however, the effect is less dramatic. For a given salinity, the density calculated at 5°C is approximately 0.5 kg/m3 higher than the density calculated at 10°C. It is apparent based on the Fraser River salinity data presented in Section 3.4.1, that there is significant variability of the salinity (and thus density) in the river water. As indicated on Figures 33a and 33b, fresh river water (salinity less than approximately 2 ppt) was identified at shallow river depths and during periods of low tide. Fresh river water has a density of approximately 1000 kg/m3. The maximum salinity measured in November 1977 was approximately 23 ppt at the base of the river. This salinity value corresponds to a density of approximately 1018 kg/m3 (at 10°C) using the International Equation of State for Seawater. 3.5.4 Tidal and Seasonal Effects on River Level The Fraser River experiences continuous changes in water level due to tidal fluctuations and seasonal river discharge rates. Tides in the Fraser River Estuary are semidiurnal and the mean tidal range at the mouth of the Fraser River is approximately 3 m (Kostaschuk and Atwood, 1990). The water level in the river rises in response to high tide conditions and falls in response to low tide conditions. The water level in the river is also affected by the river discharge rate, with higher river levels observed for higher discharge rates. In general, a peak river discharge rate is observed in late May or early June and low discharge rates are observed in winter months (i.e. December, January, and February) (Ages and Woollard, 1976). A hydrograph illustrating the changes in discharge rate over the year for the Fraser River is included on Figure 34. 82 A model was developed by Ages and Woollard (1976) which predicts the changes in river stage due to the individual or combined effects of tidal fluctuations and seasonal river discharge rate. A graph which illustrates changes in model-predicted water levels for a station located approximately 2 km upstream from the field site is provided as Figure 36. The graph indicates that seasonal changes in river discharge rate (from 1400 to 8500 m3/s) cause the river water level to change approximately 0.6 m in elevation (without tidal fluctuations considered). As illustrated on Figure 36, the combined effects of tidal fluctuations and changes in river discharge rate lead to constant changes is river water level adjacent to the field site. In general, the lowest low water level would be expected as a result of the combination of low tide and low river discharge conditions. Highest high water levels would be expected at high tide and during periods of high river discharge. Minimum and maximum instantaneous water levels reflecting these conditions have been measured from 1971 to 1984 at a station along the North Arm of the Fraser River approximately 2 km upstream from the field site (Environment Canada, 1997). A minimum instantaneous water level of ^ approximately 1.5 m below geodetic elevation, and a maximum instantaneous water level of 3 m above geodetic datum were measured. This suggests that the river level fluctuation adjacent to the field site due to tidal and seasonal influences does not exceed 4.5m. 83 jr//// jr//// ////// ////// s/// / / 4.0 7.0 ' ' T — r r -e.o -s.o FRRSER ST. - r — r — • 1— -2.0 1.0 HT. ( FT. WRT GEO. Figure 36: Model-predicted correspondence between highs and lows at Point Atkinson and Fraser Street (from Ages and Woollard, 1976). . 4.0 MODEL DEVELOPMENT 4.1 Governing Mathematical Equations The density dependent groundwater flow regime at the field site has been simulated using the model Fracdens (Shikaze et al., 1996). The groundwater flow equation and the solute transport equation are linked by Darcy's Law and an equation of state relating groundwater density to solute concentration. A non-linear problem results as the solution of the groundwater flow equation depends on the solution of the transport equation and vice versa. A.Picard iteration scheme is used to solve the non-linear system of equations whereby the equations are solved in sequence until convergence is attained for a given time step. The program proceeds through a number of time steps until the desired output time is reached. The following section presents the equations used to solve the density dependent flow problem. Groundwater Flow Equation The two-dimensional groundwater flow equation written in terms of equivalent freshwater head (h) is: ( - \ M V.£*, J da. = S - y = l , 2 (2) where k ; j is the permeability tensor [L 2], g is the acceleration due to gravity [L/T 2], // is the dynamic viscosity [M/L/T], and the term rj is equal to 1 for the vertical direction and equal to 0 for the horizontal direction. Equivalent freshwater head (h) was defined by equation 1 in Section 3.3.2.3. 85 The relative density pr is defined as: P - 1 (3) Pr = Po S s is the specific storage which is calculated in terms of porous media compressibility (or [LT 2 /M] ), fluid compressibility (J3 [LT 2 /M] ), and porosity (n) with the equation: The groundwater flow equation (2) is solved to obtain the distribution of equivalent freshwater head. The solution to equation (2) requires knowledge of the spatial distribution of density (which is unknown initially). Thus, equation (2) is first solved using an estimated density distribution, and then using updated density distributions as the model proceeds through the simulation. Dairy's Law Following the solution of equation (2), Darcy's Law (in terms of equivalent freshwater head) is used to solve for groundwater flux (q [L/T]). This form of Darcy's Law is written: S s •=/?0g(a+ n $ (4) KfPo8\ ch (5) q, = -Transport Equation Once the groundwater flux is determined using equation (2), the transport equation is solved to calculate solute concentrations throughout the model domain. The transport equation can be written: d nD.„ c\ at q ^ = nd y = 1'2 ( 6 ) where t is time [T] and c is the solute concentration [M/L3] expressed as the relative solute concentration (c=Cij/Co) where cnis the maximum solute concentration and C y is the solute concentration at point (i,j). in equation (6) is the hydrodynamic dispersion tensor [L2/T]. For two dimensional problems, the dispersion tensor can be expressed as two dispersion coefficients (D x and D z) which correspond to the principal directions relative to the local orientation of the flow line (de Marsily, 1986). The dispersion coefficients D x (in the direction of flow) and D z (perpendicular to the direction of flow) are defined by the following equations: IV=<+«A (7a) Bz = Td0+az\^\ (7b) where d c is the free-solution diffusion coefficient [L2/T] and x is the tortuosity of the porous media. Equation of State The solute concentrations as determined with the transport equation are used to re-calculate the spatial variation in groundwater density. This procedure is done with an equation of state which linearly relates density with solute concentration. The equation relating density to solute concentration used by the model Fracdens is written: f P=Po 1.0 + Po V Po (8) where Pmax is the maximum density which is associated with the maximum solute concentration (i.e. at the solute source). The density distribution is then used as an initial condition to solve the groundwater flow equation. This cycle of equations 2, 5, 6, and 8 is repeated, with each cycle representing one Picard iteration, until convergence is attained (when density is approximately equal to the previously calculated density at each point in the domain). Then, the final density distribution is used as an initial condition to solve equation (1) to initiate the next time step. 4.2 Boundary Value Problem 4.2.1 Model Domain The model domain represents a rectangular two-dimensional north-south cross section through the site and is illustrated on Figure 37. The right hand side of the model corresponds to the bank of the Fraser River and will be referenced as the river-side model boundary. (Note that attempts were made to model the region beneath the river within the model domain, however, a stable solution was not obtained.) The left hand side represents II x: o II N < 8 IM o Ii o l"i 88 II -o C N ^ i i LU LU CM CN "* ",. E e * - CM dl dl TJ c <8 E J2 TJ CD E £ •— oo i i UJ LU It £ TJ c D CO c I * l 5 >> O en I _ X t -8 o CT II CT o II _ E o c "0 E o Q Q) "D O co <D D U> O E o a point 650 m south of the Fraser River on Lulu Island, Richmond, BC, and will be referenced as the island-side model boundary. The island-side boundary of the domain was selected to be beyond the area of saltwater intrusion as extrapolated from the field data (actual data to confirm the precise position of the toe of the saline wedge could not be obtained due to housing development to the south of the field site). Vertically, the domain extends 19m and includes the three geological units comprising the aquifer at the site: the Silty Sand, Fine and Medium Sand, and Medium Sand units. In addition, the top few metres of the underlying Silty Clay unit are included within the model domain. It was necessary to include the upper part of this lower confining unit for the investigation of the effects of the upward flux of fresh groundwater observed within the Silty Clay. The inclusion of the Silty Clay unit at the base of the model domain also allows for the investigation of the effect of slight relief of this low permeability layer. The upper confining layer (the Clayey Silt) has not been included within the model domain and is represented by an impermeable boundary above the aquifer materials (boundary conditions are discussed in Sections 4.2.2 and 4.2.3). The advantage of this model arrangement is a smaller domain which leads to shorter computing times due to fewer nodes. The Clayey Silt unit can be excluded from the model domain as the permeability contrast between the Clayey Silt (relatively low permeability) and the Silty Sand (higher permeability) indicates that vertical recharge from the water table is locally negligible compared to lateral flow within the sand aquifer. The configuration of the geological units is based on the site stratigraphy interpreted from boreholes and CPTs completed at the site (Section 3.2.2). In order to model the site geological conditions using Fracdens, the stratigraphy must be simplified and represented by a series of rectangular blocks with uniform hydrogeological properties (Figure 37). The simplified geological/hydrogeological conditions can be modified (e.g. geological unit thickness or permeability) during model calibration if required. The origin of the domain is located at the bottom left corner in Figure 37. As such, the vertical direction (z) is positive upward and the horizontal direction (x) is positive to the right (i.e. toward the river). The field measurements (depths below ground surface and geodetic elevations) are adjusted for this coordinate system. 4.2.2 Boundary Conditions for Groundwater Flow The boundary conditions for flow are illustrated on Figure 37. The top boundary of the model is specified as impermeable to groundwater flow to represent confined aquifer conditions. At the bottom boundary, a constant upward flux of 0.0005 m/y is imposed (the measured upward vertical gradient of 0.17 at the West Bay well multiplied by the hydraulic conductivity of the Silty Clay - 1 x 10'10 m/s). This condition allows for the investigation of the effects of the small upward flux of fresh groundwater through the Silty Clay unit which was observed at the site. The boundary condition for flow at the river-side of the model is specified equivalent freshwater head. Equivalent freshwater head is calculated for the river-side model boundary using equation (1) and a simplification of the water density stratification in the river. The density of the water in the river is dependent on salinity and is represented by two zones (see Figures 38a and 38b): a shallow layer of fresh water of density 1000 kg/m3, and a zone at the bottom of and beneath the river over which the density is constant and equal to the maximum density. For the river-side boundary segment which is adjacent to the fresh water zone in the river, a constant equivalent freshwater head is applied which is equal to mean sea level (e.g. h=21 m for the model domain). For the segment of the river-side boundary which is adjacent to saline river water, and the segment which extends below the base of the river, the density is a maximum and is assumed to be constant with depth (pm,*)- The equivalent freshwater head is calculated using equation (1) for this segment, and the resulting equivalent freshwater head values increase linearly 91 Figure 38a: Development of Fraser River Salinity Stratification Salinity Data Apr. 19-20, 1978 (Ages, 1988) Discharge at Hope =1710 m3/s Fresh Water Zone (from water surface to 7 m below geodetic) - High Tide Salinity Profile -Mid Tide Salinity Profile - Low Tide Salinity Profile Saline Water Zone (7 m below geodetic to river bottom) 10.0 15.0 Salinity (ppt) 20.0 25.0 E 3 B TJ O % •a o <u O) I I g U 3 o.o 1.0 2.0 3.0 4.0 5-0* 6.0 7.Q 8.0 9.0 10.0 11.0 12.0 0.0 Figure 38b: Development of Fraser River Salinity Stratification Salinity Data Nov. 23,1977 (Ages, 1979) Discharge at Hope = 780 m3/s — • — High Tide Salinity Profile — B — Mid Tide Salinity Profile — A — Low Tide Salinity Profile Fresh Water Zone (from water surface to 7m below geodetic) Saline Water Zone (7 m below geodetic to river bottom) 5.0 10.0 15.0 Salinity (ppt) 20.0 25.0 with depth. Figure 39 illustrates the development of the river-side model boundary conditions. The boundary condition at the island-side of the model is also specified equivalent freshwater head. At this boundary, fresh groundwater is assumed to be present within the aquifer units and thus the equivalent freshwater head is a constant value over the vertical depth (i.e. h=ho when p= 1000 kg/m3). Thus, horizontal (seaward) flow across the island-side boundary is maintained. The specific boundary conditions at the island and river sides of the model (i.e. values for equivalent fresh water head) were determined during model simulation through trial and error calibration to field salinity data for W l , W2, W3, K9701, K9801, and K9802. First, a river water salinity at the base the river was assumed. The river water salinity must be at least 16.5 ppt, which was the salinity measured at 16m depth in K9801 at the field site, and must be less than 22.8 ppt which is the approximate salinity of seawater in the Georgia Strait (estimated using chloride data for sample SW from Simpson and Hutcheon, 1995, and the chloride-salinity relation illustrated on Figure 19c). Then, p^x was calculated accordingly for each elevation (z) along the river-side boundary. An equivalent freshwater head value at the island-side model boundary was then assumed. A transient simulation was carried out until the saline wedge reached an equilibrium position and the salinity results were compared with field data. The equivalent freshwater head at the island side model boundary and/or the salinity at the river-side model boundary were then adjusted, Pmax and the equivalent freshwater head were recalculated, and the model was simulated to obtain a better match to the field salinity data. This procedure was repeated as necessary with the goal of matching model output to field salinity data. Note that the above trial and error calibration procedure was necessary to determine the boundary equivalent freshwater head conditions at the island-side and river-side model boundaries. Field data were not available to determine average equivalent freshwater head 93 8 o CD > CD CO o 1 s-i fc-o x o -lO o © t o o CO C < £ =D ? C § O £ o s D "D C D O ^ A S © i O 0 0) 1 TJ | ^ - COO M— D O ~ /-CO c CD lb€Z CCD a O « ~ 0 © > -9 Q co D O) T J D CD _c 1111 Pi l += n O i s i I I II I I i l O 1— >i Nk-S i t c <d T J I I q . o . values at these boundaries (due to the temporal variations) which would be required for dynamic equilibrium simulations. 4.2.3 Boundary Conditions for Solute Transport The boundary conditions for solute transport are indicated on Figure 37. The top of the model is impermeable and as such the boundary condition is specified as zero flux. The bottom boundary is specified a relative concentration of zero. The condition at the island-side boundary of the model is specified concentration; the salinity at this boundary is also specified as 0 ppt (relative concentration = 0). The transport boundary conditions at the river-side of the model are based on simplification of the salinity distribution with depth in the river (Figures 38a and 38b). Along the river-side model boundary, the transport boundary condition adjacent to the zone of saline water in the river and beneath the river is prescribed concentration. The relative concentration over this depth interval is equal to 1. Saline water enters the model domain along this boundary. Above the segment of prescribed concentration, the condition imposed for the segment of the river-side boundary which is adjacent to the freshwater zone of the river is specified as a Cauchy condition. This boundary condition allows for fresh and mixed groundwater to exit the model domain to the river. The development of the river-side model boundary conditions is indicated on Figure 39. The z-coordinate above which groundwater flow exits the domain and below which saline water enters the domain is termed the exit point. In other saltwater intrusion modelling studies where the saline water body is not stratified, the exit point was determined through a trial and error procedure (e.g. Huyakorn et al., 1987). This procedure involves the solution of the model using an initial estimated exit point, and then a review of the velocities at the boundary to determine whether the velocities above the exit point are directed out of the domain and velocities below the exit point are directed into the domain-The exit point is adjusted and the velocity distribution is re-calculated until this condition is attained. For this saltwater intrusion problem, the exit point was initially estimated to correspond to the top of the saline water in the river as indicated on Figures 38a and 38b. Upon completion of the simulation, the velocity vectors were reviewed to confirm their directions above and below the exit point (no adjustment of the exit point position was required). For this investigation, the water salinity has been used as the "solute" (dependent variable) for the solute transport problem. Salinity is used so that field measurements can be easily compared to model results. Typically, chloride is used for the modelling of salt water intrusion as it is considered a conservative tracer. The site data indicate a linear relationship between chloride concentration and salinity (Figure 19c) and this supports the use of salinity as a solute for modelling. 4.3 Model Input Parameters The input parameters used for the modelling study are discussed below. All model parameters are expressed in units of metres, kilograms, and years. 4.3.1 Hydrogeological Model Input Parameters For each geological unit the following parameters must be specified: permeability in x and z directions (kx and kz), longitudinal dispersivity (ax), vertical transverse dispersivity (az), porosity (n), aquifer compressibility (a), and tortuosity (T ) . These parameters and the parameter values which have been used for the modelling study are discussed below. Estimated parameter values are summarized in Table 2 . 96 Table 2: Estimates of Model Hydrogeological Parameters PERMEABILITY Horizontal Permeability Mm2) Vertical Permeability Mm2) Anisotropy Ratio kJK Silty Sand 2 x IO'12 2 x IO-14 100:1 Fine and Medium Sand 4.4 x l O 1 1 4.4 x 10"13 100:1 Medium Sand 4.4 x 101 1 4.4 x IO 1 2 10:1 Silty Clay 1.0 x l O 1 7 1.0 xlO"1 7 1:1 DISPERSIVITY Longitudinal Dispersivity a, (m) Vertical Transverse Dispersivity a, (m) Silty Sand 1.0 0.01 Fine and Medium Sand 1.0 0.01 Medium Sand 1.0 0.01 Silty Clay 1.0 0.01 POROSITY Porosity Silty Sand 0.3 Fine and Medium Sand 0.3 Medium Sand 0.3 Silty Clay 0.4 COMPRESSIBILITY Compressibility (m yi^ /kg) Silty Sand 1 x IO"23 Fine and Medium Sand 1 x IO"23 Medium Sand 1 x IO"23 Silty Clay 1 x IO"22 TORTUOSITY Tortuosity Silty Sand 0.7 Fine and Medium Sand 0.7 Medium Sand 0.7 Silty Clay 0.1 Permeability: Horizontal permeability estimates used for the hydrogeological model are presented in Table 2. The values for horizontal permeability (kx) have been established for each of the aquifer units (Medium Sand, Fine and Medium Sand, and Silty Sand units) based on slug tests completed during UBC Hydrogeology Field School, 1996, for zone-specific piezometers. The permeabilities (units of m2) were converted from hydraulic conductivity (units of m/s) using a conversion chart from Freeze and Cherry (1979). Note that the small difference in measured permeabilities of the Fine and Medium Sand and Medium Sand units is considered negligible and these geological units are assigned equal horizontal permeabilities. The horizontal permeabilities derived from slug tests completed during field school 1996, were selected as the most reliable permeability estimates. Other available hydraulic conductivity measurements presented in Table .1, though close to the values selected as model input parameters, were not specifically selected for this modelling study. Pump tests were not corrected for tidal fluctuations and as such may not provide data for accurate determinations of permeability. Grain size analyses and falling head permeability tests completed by Wood (1996) were not considered representative of the larger scale of the aquifer which would be modelled in this study. The slug test data analyzed by Wood (1996) were not available for independent analysis. The vertical permeability (kz) has been determined based on estimates of the anisotropy of each aquifer unit. In a recent study by Bair and Lahm (1996), the authors suggest that some degree of anisotropy is present even for visually homogeneous aquifer materials. In their study Bair and Lahm (1996) tabulated anisotropy measurements for a number of sand or sand and gravel aquifers. The horizontal to vertical anisotropy ratio for the two sand aquifers noted in the paper are 3:1 and 15:1. Based on this information, as well as the visual observations of the core samples (discussed in Section 3.3.2.2), ranges of reasonable anisotropy ratios have been estimated for the geological units comprising the aquifer, and the vertical permeabilities have been calculated accordingly. The estimates of vertical permeability used for the hydrogeological model are presented on Table 2. The hydraulic conductivity of the Silty Clay unit (underlying the aquifer units) was determined by consolidation tests at the University of Saskatchewan to be 1 x 10"10 m/s. This hydraulic conductivity value was calculated based on the estimated in-situ load for the Silty Clay (approximately 20 m depth). The corresponding permeability of 1 x 10 is used for the hydrogeological model (Table 2). Although some visible layering was observed in core samples obtained from the Silty Clay unit, isotropic conditions are assumed for this study. Dispersivity: The longitudinal and vertical transverse dispersivities used for this modelling study are presented on Table 2 and have been estimated based on dispersivity values reported in the literature for sand aquifers. Gelhar et al. (1992) compiled a listing of dispersivity values measured for 59 field sites and evaluated their reliability. High reliability dispersivity measurements are presented for sand aquifers at four sites, and these are presented in Table 3. Gelhar et al. (1992) also suggest, based on their analysis of available dispersivity measurements, that as a first approximation, the vertical transverse dispersivity is approximately two orders of magnitude lower than the longitudinal dispersivity. Based on the ranges of dispersivity measurements for sand aquifers presented by Gelhar et al. (1992), the longitudinal dispersivity for the sand aquifer (Medium Sand, Fine and Medium Sand, and Silty Sand units) at the site is estimated to be 1.0 m and the vertical transverse dispersivity is estimated to be approximately 0.01 m. The longitudinal and vertical transverse dispersivities of the Silty Clay unit are set to be equal to those of the sand aquifer units. Table 3: Dispersivities for Sand Aquifers from Gelhar et al. (1992) Site Name/Location Aquifer Material Average Aquifer Thickness (m) Hydraulic Conductivity (m/s) Longitudinal Dispersivity (m) Vertical Transverse Dispersivity (m) Borden glaciofluvial sand 9 7.2E-5 m/s 0.43 to 0.5 0.0022 Mobile, Alabama layered mediumsand 21.6 - 4.0 -Bonnaud, France sand 3 2.gE-4to3.7E-4m/s 1.6 (average of six measurements) Porosity: Porosity values for different types of geological deposits are presented in Freeze and Cherry (1979). Based on this information, a porosity of 0.3 is estimated for the sand aquifer units. The Silty Clay unit is estimated to have a higher porosity of 0.4. Aquifer Compressibility: Aquifer compressibility values for different types of geological deposits are presented in Freeze and Cherry (1979). Using these data, values for compressibility of the different geological units at the site have been estimated and are presented on Table 2. Using equation (4), the estimated compressibility, the estimated porosity, an approximate thickness of 10 m the Medium Sand unit, and input parameters presented in Section 4.3.1, the storativity of the aquifer is calculated to be 0.001. This value is close to the storativity of 0.003 which was measured as a result of a short term pump test completed at the site. Tortuosity: As discussed in Section 4.1, an estimate of tortuosity is required to calculate the diffusive component in the coefficient of hydrodynamic dispersion D x and D z . DeMarsily (1986) reports tortuosity values for sand to be approximately 0.7 and for clays to be approximately 0.1. These values have been used for the sand and clay geological units at the site. 4.3.2 Other Model Input Parameters Other parameters not related to the geological conditions must also be specified for the solution of the boundary value problem. These are: water viscosity, freshwater density, water compressibility, acceleration of gravity, and free-solution diffusion coefficient. The values used for these parameters are presented in Table 4. 100 Table 4: Other Model Input Parameters Parameter Symbol Value Water Viscosity It 35470.7 kg/m yr Freshwater Density P o 1000.0 kg/m3 Water Compressibility 3 4.42 x IO"25 m yr2/kg Acceleration of Gravity g 9.77 x 1015 m/yr2 Free-Solution Diffusion Coefficient d o 3.15 x 10"2 m2/yr 4.4 Numerical Solution Fracdens is based on the Galerkin finite element method and the time derivatives are approximated using the finite difference method. The equivalent freshwater head distribution, solute concentration distribution, and velocity distribution are calculated by Fracdens as the program steps through time and are recorded in output files at specified simulation times. The field site is modelled using the boundary value problem specified previously. For each simulation, at t=0 the saline water begins to migrate into the model domain from the river-side model boundary. Saltwater intrudes the aquifer moving toward the island-side boundary. The system ultimately reaches a dynamic equilibrium or steady state; at which time there is essentially no further change in the solute concentration distribution within the model domain. Saltwater intrusion at the field site is assumed to be at steady state (i.e. in a state of dynamic equilibrium). The deltaic sediments forming Lulu Island in the vicinity of the field site are estimated to have been deposited approximately 5000 years ago (inferred from Williams and Roberts, 1989). Thus, once the model reaches a steady state concentration distribution, the results can be compared to the site data. The numerical accuracy of the model is dependent on the Peclet number (P) and Courant number (C). These are defined as: 101 \Al Al vAt P = and C = D a, A/ where Al is the distance across the element in the direction of the velocity (v), cti is the dispersivity in the direction of the velocity, and At is the length of the time step. As a general guideline, in order to minimize the effects of numerical dispersion and overshoot on the results, the Peclet number should not exceed 2 and the Courant number should not exceed 1 for any gridblock (Frind et al., 1987). The Courant number is affected by the grid spacing and the time step. Fracdens uses an adaptive, variable time-stepping scheme that either increases or decreases the size of the time step depending on the temporal rate of change of either concentration or equivalent head (Shikaze et al., 1996). Thus numerical instabilities due to violations of the Courant criterion are minimized. In order to satisfy the Peclet criterion of 2, a small vertical grid spacing is required which generates large numbers of nodes for the model domain (resulting in a lengthy computing time). To justify the use of larger Peclet numbers (i.e. to reduce computing time), a series of simulations was completed for a simplified model domain to investigate the effect of grid spacing on model results. Simulations were completed for a saltwater intrusion problem using a homogeneous rectangular domain (x=50m, z=10m) with hydrogeological properties equal to those estimated for the Medium Sand unit at the field site (Table 2). Details of the simulations are provided in Appendix IV and the results are summarized below. The results of the test simulations for variations in vertical grid spacing (Pz) are presented on Figures A4-2 and A4-3, Appendix IV. Variations in Pz did not significantly affect the position (elevation) or thickness of the transition zone up to Pz = 100. The effects of 102 overshooting, however, were observed for Pz greater than approximately 50. Variation in Pz also had some effect on the position of the saline wedge toe as the inland extent of the wedge toe was slightly farther for simulations with lower Pz numbers. The results for variation of horizontal grid spacing are illustrated on Figures A4-4 and A4-5, Appendix IV. In the x-direction, numerical dispersion was evident for Peclet numbers greater than 2 as the thickness of the transition zone generally increased with increasing Px. Based on these preliminary simulations, the P z criterion was relaxed to 50 and the P x criterion was maintained at 2 for the modelling of saltwater intrusion at the field site. The resulting grid spacing was 2.0 m in the x-direction and 0.5 m in the z-direction. 5.0 MODEL RESULTS 5.1 Calibrated Steady State Solution 5.1.1 Model Calibration to Field Salinity Data Model calibration involved the trial and error variation of hydrogeological input parameters and boundary conditions to obtain model results which best fit the field salinity data. The resulting model domain, boundary conditions, and permeability parameters are indicated on Figure 40 for the best fit simulation (referenced as Simulation 1). The model was simulated to a final time of 300 years to attain dynamic equilibrium. Relative concentration contours for Simulation 1 are presented on Figure 41a. A series of graphs showing the comparison between field salinity measurements and model output for K9701, W l , W2, W3,K9802, andK9801 are presented as Figure 42. The estimates of hydrogeological input parameters in Table 2 (Section 4) were obtained from field measurements or literature sources and these were not modified to obtain the best fit result (except for hydraulic conductivity as discussed below). Hydraulic conductivity in the area of the saline wedge toe and other factors that had a significant effect on model output were varied to obtain a best fit to the field data and are discussed below. 1) The hydraulic conductivity distribution in the area of the wedge toe was modified from that indicated on Figure 37. It was necessary to decrease the permeability of the Medium Sand and Fine and Medium Sand units in the area of the wedge toe in order to more closely match the shape of the saline wedge indicated by the field salinity data (Figure 31). The zone where permeability was adjusted from original estimates is referenced as the "lower permeability zone" in this thesis. E J CN CN I I o £ C O CN II X t _o o II o E 8 o o CD II CJ E CM _ E o © © E g o a o o a> o o o O) o > JZ "D C o c O TJ C o o ^ . li o 0 I D G> o CM II .c 105 0 100 200 300 400 500 600 X (m) Figure 41a: Simulation 1 - relative salinity contours. Contour Salinity 0.1 1.9 0.2 3.8 03 5.7 0.4 7.6 0.5 9.5 0.6 11.4 0.7 133 | 0.8 15.2 0.9 17.1 1.0 19.0 0 100 200 300 400 500 600 X(m) Figure 41b: Simulation 1 - equivalent freshwater head contours (m). Contour interval 0.025m. 106 0 100 200 300 400 500 600 X(m) Figure 41c: Simulation 1 - velocity vectors plotted in true scale Velocity vector scale: —• = 5 m/y 0 100 200 300 400 500 600 X(m) Figure 4 Id: Simulation 1 - horizontal component of velocity vectors plotted in log scale. 107 K9701 E 5 | •a c S 10 (9 I » 15 to £ -s. « 20 25 4 Field Salinity K9701 (x=155m) Model Results (x=154m) 10 Salinity (ppt) 15 20 W 3 25 4 + Field Salinity W3 (x=373m) — • — Model Results C*=374m) 10 Salinity (ppt) 15 20 5 10 15 Salinity (ppt) 20 25 K9802 4 Field Salinity K9802 (x=500m) Model Results (x=500m) 10 Salinity (ppt) 15 20 W2 I 5 1 25 + Field Salinity W2 (x=343m) — - — Model Results (x=342m) 5 10 Salinity (ppt) 15 20 O i 25 K9801 + Field Salinity K9801 (x=608m) — — Model Results (x=608m) 10 Salinity (ppt) 15 20 Figure 42: Field Salinity Compared to Model Results - Simulation 1 108 Through the trial and error calibration process, the areal extent of the lower permeability zone was determined, and the zone was assigned horizontal and vertical permeabilities of 8 x 10"12 m2 and 8 x 10"14m2. (The horizontal permeability of the Medium Sand and Fine and Medium Sand units was reduced by a factor of approximately 5 times and the anisotropy of the Medium Sand unit was increased to 100:1). The result of modifying the permeability in the area of the wedge toe was to increase the slope of the concentration contours at the wedge toe and raise the position of the transition zone in the central portion of the saline wedge (provided a better match to the field data). (For comparison, Figure 43 illustrates the concentration contours for the saline wedge simulated without the modification of permeability). The lower permeability zone at the southern end of the site was not originally included in the model domain (Figure 37); which was based on the interpreted stratigraphy shown on Figure 12). Slightly lower permeability sediments may be present at the southern end of the site, however, as indicated by the thickening of the Silty Sand and Fine and Medium Sand units and the thinning of the Medium Sand unit to the south on Figure 12. As such, the inclusion of the lower-permeability zone at the island-side of the model is reasonable. No piezometers have been completed at the southern end of the site to confirm the presence/absence of a lower-permeability zone. No other hydrogeological factors or boundary conditions modified during the calibration procedure affected the shape of the saline wedge. It is possible that three-dimensional effects of an east-west component of groundwater flow or other non-uniform flow conditions could affect the shape of the saline wedge. These effects were not investigated. 2) The river-ward freshwater gradient (unknown initially) was estimated and then modified to obtain the desired position or inland extent of the toe of the saline wedge. The river-ward freshwater gradient is controlled by the equivalent freshwater head boundary condition at the island-side model boundary. The best fit to field results was obtained 109 X(m) Figure 43: Concentration contours for Simulation 1 without the inclusion of the lower permeability zone in the area of the wedge toe. 110 using an equivalent freshwater head value at the island-side model boundary of 21.205m. This value produced a small river-ward freshwater gradient of 0.000315. A small river-ward gradient would be expected for the area of the site based on the relatively flat topography. The gradient of 0.0003 is comparable to the regional gradient across Lulu Island of 0.0001 estimated by Ricketts (1998). 3) The average salinity at the base of the river was also determined through model calibration. Changes in river salinity affected the modelled salinity or concentration values within the saline wedge as well as the inland extent of the saline wedge toe. A river salinity of 19 ppt (density = 1014.3 kg/m3 at 11°C) was determined through model calibration. As discussed in Section 4.4.2, the river salinity must be less than 22.8 ppt which was the salinity measured in a seawater sample from the Georgia Strait (sample SW from Simpson and Hutcheon, 1995). The average river salinity must also be greater than 16.5 ppt which was measured in the sand aquifer adjacent to the river at K9801. The calibrated salinity values falls within the expected range. 5.1.2 Discussion of Calibration Results In general, a good agreement between model results and field data was obtained. The shape of the modelled saline wedge is similar to that indicated by the field data (compare Figures 41a and 31). In addition, the modelled concentrations at the individual sampling points fit reasonably well to the field salinity measurements (Figure 42). The good match between model results and field data indicates that the model is representative of average hydrogeological conditions and boundary conditions at the field site. It should be noted, however, that a unique solution is not expected as the conditions at both the river- and island-side model boundaries are not well constrained and there is some degree of uncertainty associated with other model input parameters. I l l Minor differences between the field and Simulation 1 results are unavoidable due to the difficulties in representing actual site conditions in terms of a simplified mathematical model. Some more distinct discrepancies between model results and field data can be observed in Figure 42 and by comparison of Figures 41a and 31. These discrepancies are discussed below. The model results underestimate the salinity at the wedge toe and overestimate the salinity near the river-side model boundary. Variations in different hydrogeological parameters and boundary conditions during the calibration process could not correct the results for this discrepancy. The differences between model results and field data are attributed to (a) complex transient effects at the river-side model boundary which were not modelled for Simulation 1, .(b) geological heterogeneities which were not identified at the site based on the available site data, and/or (c) three-dimensional effects which could not be modelled. Further discussion of the effects of seasonal influences on the river-side model boundary is provided in Section 5.2. Another discrepancy between model results and field data is observed at the transition zone. The model results predicted a thicker transition zone at the top of the saline wedge (approximately 5 m thick) than that indicated by detailed field salinity measurements across the transition zone at W3 (approximately 2 m thick) and resistivity data from CPTs (indicated a transition zone of approximately 2 to 3 m). (At other locations, the thickness of the transition zone is not precisely defined by the available field salinity data). Factors that may have caused the difference between model results and field data include: overestimation of horizontal and vertical aquifer dispersivity values used in the model, underestimation of the anisotropics of the geological units, and/or numerical dispersion associated with the grid spacing used in the model. Numerical dispersion is not expected to have a significant effect on the thickness of the transition zone as discussed in Section 4.4. A discussion regarding the sensitivity of the model results to changes in dispersivity and anisotropy is given in Section 5.3.1 and 5.3.2. 112 A final discrepancy between model results and field data is observed in the position of the transition zone between fresh and saline water near the river-side model boundary. The model results indicate that the transition zone near the river is higher than that indicated by the field data. This discrepancy could be a result of inaccurate representation of field conditions at the river-side model boundary. The salinity (and density) stratification in the river was simplified to two layers of constant salinity, but the actual river salinity varies gradually from fresh to saline with depth across a transition zone. Another difference between field conditions and the model domain is that the true shape of the river bed could not be modelled; the model boundary condition used for Simulation 1 represents an aquifer with a vertical boundary that is in full contact with saline water in the river (which is not the true condition). The simplification of the complex transient behaviour of conditions at the river-side model boundary to represent one constant or average condition could also lead to discrepancies between model results and field data at locations near the river. The discrepancies between model results and field data for Simulation 1 illustrate the difficulty in modelling a saline intrusion. The study by Segol and Pinder (1976) provides another example which demonstrates the difficulties of modelling saltwater intrusion in coastal aquifers. Like Simulation 1, the Segpl and Pinder study calibrated model results to concentration contours in cross section that were delineated by detailed field data (see Figure 7). As indicated on Figure 7, the most notable discrepancies between the Segol and Pinder model results and field data include: the position of the wedge toe, the elevation of the transition zone in the central portion of the model, and the shape (bluntness) of the wedge toe. Segol and Pinder did not obtain an exact match between model results and field data but indicate that their results are show a satisfactory agreement. 5.1.3 Model Results for Equivalent Freshwater Head 113 Figure 41b illustrates equivalent freshwater head contours for Simulation 1. The pattern of equivalent freshwater head contours is typical for areas of saltwater intrusion and is similar to that obtained by Kohout (1960) using field data from the Biscayne Aquifer, Florida (Figure 44). The model results for equivalent freshwater head were not calibrated to match field equivalent freshwater head data due to the difficulties in obtaining accurate measurements of equivalent freshwater head in the field and the lack of equivalent freshwater head measurement points along the model domain. Figure 41b illustrates the very small lateral differences in equivalent freshwater head within the saline wedge (which would be difficult to measure in the field). Note that on Figure 41b, flow directions cannot be determined by the construction of flow lines orthogonal to the contours of equivalent freshwater head in a variable density system (Jorgensen et al., 1982). 5.1.4 Model Results for Groundwater Velocity The velocity vectors (vector components v x and vz) are calculated by Fracdens using Equation 5 (Section 4.1) divided by porosity (v; = q/n). The velocity vectors for Simulation .1. are plotted on Figure 41c. Figure 41d presents the horizontal component of the velocity vectors in log scale for an improved visualization of horizontal groundwater flow directions across the model domain. The velocity vectors in Figures 41c and 41d indicate the general flow regime expected for areas of saltwater intrusion which is illustrated on Figure 1. Flow of saline water toward the wedge toe and then a reversal of flow as saline water and mixed water are directed toward the river under the regional freshwater gradient is illustrated by the velocity vectors. At the river-side model boundary (x=650m), flow above the exit point (z=14m) is directed out of the system and flow below the exit point is directed into the aquifer as would be expected. Also as expected, vertically upward flow of groundwater is observed Figure 44: Equivalent freshwater head contours for the Biscayne Aquifer (from Kohout, 1960). 115 in the Silty Clay unit at the base of model domain (upward groundwater velocities are approximately 1.25 x 10"3m/y). Due to the large range of velocities, however, the directions of the vectors within the Silty Clay are not discernible on Figure 41c. Figure 45 illustrates a conceptual model of the average flow regime at the field site as indicated by the Simulation 1 velocity results. The model results indicate that the horizontal groundwater flow velocity at the base of the Medium Sand unit (flow of saline groundwater toward the island-side model boundary) ranges from approximately 0.5 to 6.3 m/y in the central portion of the saline wedge. These velocities correspond to horizontal equivalent freshwater head gradients of 0.00001 to 0.0001. In the area of the wedge toe, where lower permeabilities were assigned to the Medium Sand and Fine and Medium Sand units, the horizontal velocities are somewhat lower; ranging from approximately 0.4 to 1.4 m/y (for flow toward the island-side model boundary). Note that the horizontal equivalent freshwater head gradients at the base of the Medium Sand unit indicated by the model results are lower than the gradient of 0.005 indicated by the field data for the deep zone-specific piezometers. The difference is attributed to the difficulties in measuring/calculating small gradients at the site using field data. At some locations at the base of the model domain, minor anomalies in the groundwater flow directions are indicated on Figures 41c and 41d. At approximately x=345 and x=590m, the velocity vectors form small "convection cells" near the base of the model. The small convection cells are formed at the downgradient side of the higher relief permeability blocks representing the deep Silty Clay unit (refer to Figure 40). As such, the observed convection cells at these locations are attributed to abrupt changes in permeability in the model domain and are not inferred to represent true flow conditions at the field site where smoother relief of the Silty Clay layer is expected. Another location of anomalous velocity vectors is at the river-side model boundary within the permeability block which represents the Silty Clay unit (x=640 m, z=0 to 6 m). Here, 116 117 the horizontal component of the velocity vectors is not directed into the model domain as would be expected below the exit point, but is directed toward the river-side model boundary. It is possible that this is the result of the small convection cell at approximately x=590 m which may have some affect on the flow within the Silty Clay unit to the right. Note the velocities in the Silty Clay unit at the river-side model boundary are very small (approximately 1 x 10'4 to 1 x 10'3 m/y) and slight numerical errors could contribute to the observed anomalous result. The small scale anomalies in the velocity field are not anticipated to significantly affect the model results. 5.2 Investigation of Transient Effects 5.2.1 Modelling of Seasonal Changes at River-Side Model Boundary As discussed in Section 3.5, the relationship between river level, river salinity, tidal fluctuations, and seasonal changes in river discharge rate is complex. The development of river-side model boundary conditions to accurately represent the frequently changing conditions in the Fraser River adjacent to the site would be difficult, and the application of such boundary conditions is not possible using the model Fracdens. Thus, in order to investigate the possible influences that the changing conditions in the river could have on the saline wedge at the field site, the transient effects at the island-side model boundary-were simphfied to represent seasonal changes and applied to the modeL In general, from approximately September to February, relatively low river discharge rates occur. During this time, the water level in the river is lower than average (approximately 0.3 m lower than average based on Figure 36) and the salinity at the base of the river is higher than average. From approximately March to August, the river discharge rates are relatively high. During this time, discharge rates are generally high enough to prevent the 118 saline river wedge from migrating up the North Arm of the Fraser River to the location of the field site (see Figure 34). Also during periods of high river discharge, the water level in the river is higher than average (approximately 0.3 m higher than average). The changing salinity and water level conditions in the Fraser River affect the hydrogeological conditions and could influence the saline water distribution within the wedge in the sand aquifer beneath the site. During periods of high river discharge, only freshwater is present in the base of the river, and the high river water levels (increased head) could create strong enough gradients to cause freshwater from the base of the river to enter the aquifer. Under conditions of average or low river discharge, more dense saline water is present in the base of the river and the saline water enters the aquifer due to density effects. Over time the changing river level and salinity conditions at the river will produce a saline water distribution within the aquifer that has reached a state of dynamic equilibrium. The resulting saline water distribution of Simulation 1 was modelled to dynamic equilibrium without consideration of the changing conditions at the river-side model boundary. The investigation of the effects of seasonal changes in river salinity and river level for this thesis involved alternating simulations representing high river discharge conditions and average conditions (i.e. Simulation 1). Conditions of low river discharge were not simulated (simulation of low river discharge conditions would have required an increase in salinity from Simulation 1 which could not be accommodated by Fracdens). The following summarizes the simulation methodology: 1) . A steady state simulation (to 300 years) was completed using the average river-side boundary conditions determined for Simulation 1. 2) . Using the output from 1) as input, a period of 6 months was simulated using river-side boundary conditions to represent high river discharge conditions (indicated on Table 5)-119 3) Using the output from 2) as input, a period of 6 months was simulated using river-side boundary conditions to represent average river discharge conditions (indicated on Table 5). 4) Sequential simulations 2) and 3) were completed until the desired output time was reached. A total time period of 20 years was simulated. 5) The equivalent freshwater head value at the island-side model boundary was then adjusted (if required) to keep the saline wedge within the model domain. (The effect of high river discharge conditions was to cause the toe of the saline wedge to advance further into the model domain. River-ward gradients used for Simulation 1 were not sufficient to maintain the wedge position within the model domain In order to obtain a realistic solution where the wedge position after 20 years of seasonal simulations was similar to that indicated by the field data, the equivalent freshwater head value at the island-side model boundary required adjustment). Then the simulation procedure was re-initiated with 1). Table 5: Summary of Boundary Conditions for the Investigation of Seasonal Influences on the Saline Wedge. River Discharge Water Level Salinity Time Period High Flow High Ni l 6 months h=21.3m Oppt (Mar. to Aug.) Average Flow Average Average 6 months h=21.0 19 ppt (Sept. to Feb.) The above procedure does not accurately represent changes in river-side boundary conditions. It is also likely that the output after 20 years of seasonal fluctuations has not reached steady state. The 20 year time period was selected to provide a reasonable indication of the effects of the modelled seasonal changes within a manageable computing time period. The above simulation methodology is assumed to provide a good indication as to the possible influences of seasonal changes on the saline wedge. 120 5.2.2 Discussion of Seasonal Model Results The final result for the simulation of 20 years of seasonal fluctuations is referenced as Simulation 2 and the relative concentration contours for Simulation 2 are illustrated on Figure 46a. For comparison, Figure 46b presents concentration contours for Simulation 2 prior to the initiation of seasonal fluctuations. The velocity distributions for Simulation 2 during the final "high" river discharge cycle and "average" river discharge cycle are presented as Figures 47a and 47b. One effect of the modelled seasonal fluctuations was to cause the saline wedge to migrate further into the model domain. This is due to the reversal of flow direction into the model domain from the river-side model boundary during periods of high river discharge conditions (illustrated by velocity vectors on Figure 47a). In order to prevent the saline wedge from reaching the island-side model boundary, the equivalent freshwater head value applied to the island-side model boundary was increased slightly from that calibrated for Simulation 1. The resulting freshwater gradient toward the river (during average seasonal discharge conditions) was 0.000338 (approximately 7% higher than the gradient of 0.000315 for Simulation 1). The concentration contours for Simulation 2 illustrate that under conditions of high river discharge, fresh water from the river enters the model domain at the river-side model boundary. The inflow of fresh river water causes a decrease in the salinity values adjacent to the river over the simulation period of 20 years. Dilution of saline water in the aquifer near the river due to the inflow of fresh water is apparent when the concentration contours of Figure 46a are compared to Figure 46b. Areas in the aquifer near the river which were bounded by the 0.9 relative concentration contour on Figure 46b are crossed by the 0.7 and 0.5 relative concentration contours on Figure 46a. The results of Simulation 1, which did not consider seasonal effects at the river-side model boundary, indicated a higher groundwater salinity near the river than was observed at the field site (Figure 42, for 121 X(m) 0 100 200 300 400 500 600 X (m) Figure 46: Simulation 2 concentration contours. (a) after 20 years of simulated seasonal fluctuations at the river-side model boundary. (b) initial conditions prior to initiating simulation of seasonal fluctuations at the river-side model boundary. (Boundary conditions at the top, bottom, and river-side are those of Simulation 1. Equivalent freshwater head at the island-side boundary was increased from 21.105m in Simulation 1 to 21.22m.) 122 X(m) Figure 47: Simulation 2 velocity distribution. Velocity vector scale. — = 5 m/y (a) for conditions at high seasonal river discharge (i.e. high river level and freshwater in river). Velocity distribution illustrates conditions after 19.5 years of seasonal simulations. (b) for conditions at average seasonal river discharge. Velocity distribution illustrates conditions after 20 years of seasonal simulations. K9801 and K9802). Dilution of salinity within the aquifer near the river due to the inflow of fresh water under conditions of high seasonal river discharge explains this discrepancy. The seasonal fluctuations caused the transition zone to increase in thickness in the area near the river (between x=500 m and x=650 m). The transition zone thickness before and after the seasonal simulations can be observed by comparison of Figures 46a and 46b. The cause for the increased thickness of the saline wedge transition zone near the river is due to advection of freshwater into the aquifer (rather than dispersive effects which also affect transition zone thickness). Simulated seasonal changes at the river-side model boundary had a major effect on the velocity distribution. The velocity vectors for high river discharge conditions (after 19.5 years of seasonal simulations) and average river discharge conditions (after 20 years of seasonal simulations) are presented on Figures 47a and 47b. As illustrated on Figure 47a, during periods of simulated high river discharge, the general horizontal flow direction within the model domain was through the aquifer; toward (and out of) the island-side model boundary. The flow of groundwater throughout the aquifer toward the island-side model boundary does not indicate that freshwater is flowing across the entire aquifer, but is a pressure effect that drives saline water further landward. During periods of average river discharge conditions the circulation of saline water into the aquifer (as was observed for Simulation 1) occurred (Figure 47b). On Figure 47b, however, convection cells are formed between approximately x=420 and x=550 that were not observed for Simulation 1. The convection cells are attributed to changes in the density of the groundwater near the river and the effects of density-dependent flow. The simulated seasonal fluctuations cause a "slug" of freshwater to enter the aquifer (during average river-discharge conditions) followed by a "slug" of freshwater (during high river discharge conditions). Thus during average river-discharge conditions, a "slug" of saline water may enter the aquifer above freshwater. Convection cells could be formed as the slug of saline groundwater "sinks" to the base of the aquifer. Note that conditions of low river discharge (low water level and high salinity in the base of the river) were not modelled, which could cause water of greater salinity (density) to enter the aquifer than observed for average river discharge conditions. Calibration for Simulation 2 was limited to the variation of the river-ward freshwater gradient. Figure 48 presents graphs that compare the results for Simulation 2 to the field data. The results indicate a similar match between model results and field data as was obtained for Simulation 1 (Figure 42). The effects of salinity dilution at the river-side model boundary are illustrated by the Simulation 2 results for K9801. 5.2.3 Discussion of Potential Tidal Effects at River-Side Model Boundary The influence of tidal fluctuations at the river-side model boundary was not investigated for the field site. Under high tide conditions, an increase in the flow of saline water into the aquifer may be observed. Under low tide conditions, flow of saline water and fresh water toward the river could occur. The short term tidal effects at the river-side model boundary, however, are not expected to have a significant effect on the position of the saline wedge. The high frequency tidal oscillations would not provide a long term change in the system which would be required to affect the saline wedge position (based on the long time to obtain an equilibrium wedge position for Simulation 1). Over the long term, tidal fluctuations may cause the transition zone to increase in thickness (Underwood, 1992). A relatively thin transition zone is indicated at the site by the field data, and the thin transition zone could not be simulated even without consideration of seasonal or tidal fluctuations. As such it is anticipated that the effects of tidal fluctuations on the thickness of the transition zone at the field site are minimal. 125 Salinity (ppt) Salinity (ppt) W2 51 « Field Salinity W2 (x=343m) — • — Model Results (x=342m) 25 5 10 Salinity (ppt) 15 20 25 K9801 * Field Salinity K9801 (x=608m) — — — Model Results (x=608m) 10 Salinity (ppt) 15 20 Figure 48: Field Salinity Compared to Model Results - Simulation 2 (after 20 years of simulated seasonal fluctuations in river) 126 5.3 Sensitivity Analysis 5.3.1 Dispersivity The sensitivity of model concentration results to changes in dispersivity was investigated by separately varying the horizontal dispersivity and vertical transverse dispersivity at values greater than and less than those used in Simulation 1. Table 6 summarizes the simulations completed to investigate the sensitivity of model results to changes in dispersivity. The effects of the variation in vertical transverse dispersivity are indicated by comparison of the results for Simulations 1, 3, and 4. Figure 49 presents a comparison of the concentration contours for these simulations, and Figure 50 compares the vertical salinity profiles for these simulations at four x-coordinates within the model domain. A decrease in vertical transverse dispersivity to 0.005 m (from 0.01 m in Simulation 1) caused a very slight decrease in the thickness of the transition and had a slight impact on the position of the toe of the saline wedge (the 0.1 relative concentration contour advanced slightly further into the aquifer for when lower dispersivity was used). When the vertical transverse dispersivity was increased to 0.02 m (Simulation 4), the transition zone thickness increased slightly and the toe of the wedge advanced slightly less far into the aquifer than for Simulation 1. As discussed in Section 5.1.2, the transition zone for Simulation 1 was thicker than that indicated by the field data. The results of this sensitivity analysis indicate that a reduction in the vertical transverse dispersivity could provide for a slightly better match between model results field data (with respect to transition zone thickness). The relatively small decrease in transition zone thickness with the change in vertical transverse dispersivity from 0.01 m to 0.005 m indicates, however, that reducing the vertical transverse 127 0 100 200 300 400 500 600 X(m) Figure 49: Comparison of model concentration contours for variations in vertical transverse dispersivity. 128 0 5 10 15 20 25 0 5 10 15 20 Salinity (ppt) Salinity (ppt) x=373m ^^^™*Disp-z = 0.01 m — A — Disp-z = 0.005 m — • — D i s p - z = 0.02 m a. 20 U~—— r . 25 10 15 Salinity (ppt) 20 25 E. 5 •a e I 10 a I * 15 CO Q. % 20 x=608m Dixp-z = 0.01 m ~ & — Disp-z = 0.005 m - • — Disp-z = 0.02 m 5J 25 10 15 Salinity (ppt) 20 25 Figure 50: Comparison of model results in vertical profile for variations in vertical transverse dispersivity. Horizontal Dispersivity (Disp-x = 1.0 m, Vertical Transverse Dispersivity (Disp-z) varied as indicated. 129 dispersivity alone may not sufficiently reduce the thickness of the transition zone to closely match field data Vertical transverse dispersivities less than 0.005 m were not simulated as this would require excessively long simulation times to meet grid spacing constraints. Note that for Simulation 3 the Peclet number in the z direction (Pz) was equal to 100. As discussed in Section 4.4 , at Pz numbers of 100 the effects of overshooting on the model concentration results were observed. The use of Pz equal to 100, however, did not cause numerical dispersion to increase the thickness of the transition zone. The effects of variation of horizontal dispersivity values can be observed by comparison of Simulations 1, 5, and 6. Figure 51 presents the modelled concentration contours for simulations for which the horizontal dispersivity was varied. Figure 52 illustrates the variability in vertical salinity profile for changes in horizontal dispersivity at four locations across the model domain. An increase in horizontal dispersivity to 1.5 m (from 1.0 m used in Simulation 1) did not affect the thickness of the transition zone and the salinity depth profiles for Simulation 6 are almost identical to Simulation 1 (Figure 51). The decrease of horizontal dispersivity to 0.5 m also did not cause a significant change to the position of the wedge or the variation of salinity with depth (Figure 51). It is anticipated that effects of increasing/decreasing the horizontal dispersivity would be observed in the 0.01 and 0.001 relative concentration contours near the toe of the wedge. Low salinities beyond the 0.1 concentration contour were not investigated as field data was not available in this region of the system to provide a meaningful constraint for model calibration. 130 0 100 200 300 400 500 600 X(m) 0 100 200 300 400 500 600 X (m) Figure 51: Comparison of model concentration contours for variations in horizontal dispersivity. 131 o i x= 155m JL 5 •o c 2 10 Si o m •-•-VV B P - * * 25 Disp-x = 1.0 m —A— Disp-x = 0.5 m — • — Disp-x = 1.5 m 10 15 Salinity (ppt) 20 25 E. 5 •a c S 10 (9 8 "5 15 a £ 20 25 x=S00m Disp-x = 1.0 m A— Disp-x = 0.5 m - • — D i s p - x = 1.5 m 10 15 Salinity (ppt) 20 x=373m E 5 5 I ft I 10 + (9 I , Q. S 20 Disp-x = 1.0 m —A— Disp-x = 0.5 m — • — Disp-x = 1.5 m 25 I -10 15 Salinity (ppt) 20 25 g 5 •a c i 10 (9 1 , 2 1 5 + xs tL 2 20 x=608m Disp-x = 1.0 m -A— Disp-x = 0.5 m - • — D i s p - x = 1.5 m 25 L 10 15 Salinity (ppt) 20 25 Figure 52: Comparison of model results in vertical profile for variations in horizontal dispersivity. Vertical Transverse Dispersivity (Disp-z = 0.01 m, Horizontal Dispersivity (Disp-x) varied as indicated. Table 6: Summary of Sensitivity Simulations - Dispersivity Simulation Name horizontal dispersivity (m) vertical transverse dispersivity (m) Simulation 1 1.0 0.01 Simulation 3 1.0 0.005 Simulation 4 1.0 0.02 Simulation 5 0.5 0.01 Simulation 6 1.5 0.01 5.3.2 Aquifer Anisotropy Preliminary model simulations using a simplified model domain indicated that an increase in the anisotropy of the aquifer caused a decrease in the thickness of the transition zone. As the transition zone for Simulation 1 was thicker than that indicated by the field data, the effect of increasing the anisotropy of the Medium Sand unit within the model domain was investigated. A simulation (Simulation 7) was completed using a Medium Sand anisotropy ratio (kx:kz) of 100:1. (The anisotropy ratio of the Medium Sand was 10:1 for Simulation 1). Note that observations of the core samples did not indicate an anisotropy as high as 100:1 for the Medium Sand. The permeabilities assigned to the Medium Sand unit for Simulations 1 and 7 are summarized in Table 7. The relative concentration contours for Simulations 1 and 7 are illustrated on Figure 53. Figure 54 provides graphs showing the variation in salinity with depth for Simulations 1 and 7 at four locations across the model domain. Increasing the anisotropy of the Medium Sand unit did not significantly affect model results. Only very slight differences in concentration profiles are observed on Figure 54. 133 0 100 200 300 400 500 600 X(m) Figure 53: Comparison of model concentration contours for variations in anisotropy of the Medium Sand unit. 134 Figure 54: Comparison of model results in vertical profile for variations in in the anisotropy of the Medium Sand unit. Variations in ratios are indicated. 135 Table 7: Summary of Sensitivity Simulations - Medium Sand Anisotropy Simulation Name horizontal permeability (m2) vertical permeability (m2) Simulation 1 4.4 x 10"n 4.4 x 10'12 Simulation 7 4.4.X.10"11 4.4 xlO" 1 3 5.3.3 Freshwater Gradient Toward River The sensitivity of model concentration results to changes in freshwater gradient toward the river was investigated by increasing and decreasing the equivalent freshwater head on the island-side model boundary. Table 8 summarizes the simulations completed. Figure 55 presents a comparison of the concentration contours for these simulations, and Figure 56 compares the vertical salinity profiles for these simulations at four x-coordinates within the model domain. The results indicate that the position of the saline wedge toe (as indicated by the 0.1 relative concentration contour) is very sensitive to small changes in the freshwater gradient toward the river. An increase in the gradient from 0.000315 to 0.000331 (an increase in equivalent freshwater head at the island side model boundary of 1 cm) caused the wedge toe to be present at a position approximately 30 m closer to the river. A decrease in the gradient from 0.000315 to 0.000300 (a decrease in equivalent freshwater head of 1 cm) caused the toe to advance approximately 40 m further into the aquifer. Significant effects of concentration profiles at locations in the central and river side portions of the saline wedge were not observed for the small changes in gradient simulated. The concentration profile in the area of the wedge toe (x=155m) was affected due to dispersion in the horizontal direction under conditions when the toe advanced further inland. 136 0 100 200 300 400 500 600 X(m) 0 100 200 300 400 500 600 X(m) X(m) Figure 55: Comparison of model concentration contours for variations in freshwater gradient toward the river. 137 x=155m " " " G r a d i e n t = 0.000315 - A — Gradient = 0.000323 - • — Gradient = 0.000308 10 15 Salinity (ppt) 20 25 10 15 Salinity (ppt) 20 25 Figure 56: Comparison of model results in vertical profile for variations in freshwater gradient toward river. Freshwater gradient varied as indicated. 138 Table 8: Summary of Sensitivity Simulations - Freshwater Gradient Toward River Simulation Name Equivalent Head at Island-Side Boundary (m) River-Ward Gradient Simulation 1 21.205 0.000315 Simulation 8 21.215 0.000331 Simulation 9 21.195 0.000300 5.3.4 River Salinity/Density The sensitivity of model concentration results to changes in river water salinity (and corresponding density) was investigated. Table 9 summarizes the simulations completed. Figure 57 presents a comparison of the concentration contours for these simulations, and Figure 58 compares the vertical salinity profiles for these simulations at four x-coordinates within the model domain. The results indicated that the position of the wedge toe was sensitive to changes in the river water salinity; the wedge migrated further into the aquifer for conditions of higher river water salinity. A change in river salinity of 1 ppt caused 30 to 40 m difference in the position of the toe of the saline wedge (as indicated by the 0.1 relative concentration contour). Horizontal dispersion in the area of the wedge toe is evident for conditions where the wedge advances further into the aquifer. Table 9: Summary of Sensitivity Simulations - River Salinity/Density Simulation Name River Salinity (PPt) Corresponding River Water Density (kg/m3) Simulation 1 19 1014.3 Simulation 10 18 1013.5 Simulation 11 20 1015.1 139 X(m) 0 100 200 300 400 500 600 X(m) X(m) Figure 57: Comparison of model concentration contours for variations in river water salinity (and density). 140 Figure 58: Comparison of model results in vertical profile for variations in river water salinity (and density). 141 5.3.5 Exit Point Position As discussed in Section 5.1.2, one of the discrepancies between Simulation 1 results and the field data was the elevation of the transition zone near the river-side model boundary. The transition zone was modelled to be higher in elevation than that indicated by the field data. Simulation 12 was completed to investigate the effect of a lower exit point on the position of the transition zone near the river-side boundary. The exit point for Simulation 12 was set to z=13 m (compared to z=14 m for Simulation 1). The results of Simulation 12 show that the elevation of the exit point did not have a noticeable effect on the position of the transition zone near the river-side model boundary. This result is illustrated on Figure 59 (presents concentration contours for Simulations 1 and 12) and Figure 60 which illustrates the salinity profiles with depth (for Simulations 1 and 12) at four points along the model domain. The model results for Simulation .1. and Simulation 12 are very similar except in the area of the wedge toe. Simulation 12 produced lower concentrations at the wedge toe than Simulation 1. 5.3.6 Implications of Sensitivity Analysis The sensitivity of model results (Simulation 1) to variations in dispersivity, aquifer anisotropy, freshwater gradient toward the river, river salinity/density, and exit point position was investigated. The parameters selected for the sensitivity analysis were those that, based on preliminary simulations, were the most likely to affect the position of the saline water wedge and/or the saline water distribution within the wedge and at the transition zone. The results of the analysis did not indicate significant model sensitivity to variations in horizontal or vertical transverse dispersivity, or aquifer anisotropy, or exit point elevation. Changes in the freshwater gradient toward the river and changes in river salinity/density affected the position or inland extent of the toe of the saline wedge but did 142 Figure 5 9 : Comparison of model concentration contours for variations in exit point elevation. 143 Figure 60: Comparison of model results in vertical profile for variations in exit point elevation. not significantly affect the saline water distribution at the transition zone (except at the wedge toe where variability in horizontal dispersion was observed). 144 The sensitivity analysis does not indicate that modification of the model input parameters investigated could significantly decrease the discrepancies between Simulation 1 results and field data. Thus the sensitivity analysis confirms that model calibration was completed adequately (within the constraints of the model) to obtain a close match between model results and field data. The model domain was constructed based on available and reliable site geological, hydrogeological, and chemical data. The sensitivity analysis indicates that additional field investigation could aid in the characterization of site conditions to provide a basis for the development of a model domain that is an even closer representation to site conditions than Simulation 1 (e.g. 3-dimensional model, model that accounts for transient effects at river-side boundary, model that incorporates small scale geological heterogeneities at the site). 145 6.0 DISCUSSION AND CONCLUSIONS 6.1 Groundwater Flow and Saline Water Distribution at the Field Site The analysis of field data and the simulation of a numerical density-dependent flow model provided valuable insight to understand the saline water distribution and hydrogeological conditions adjacent to the Fraser River at the field site. Chemical data allowed for the delineation of the saline wedge within the sand aquifer at the site. Water level data were analyzed but were not useful to assess the groundwater flow regime. Even when the 72-hour average water levels were calculated and then converted to equivalent freshwater head, a definite groundwater flow direction was not indicated by the field data at intermediate or deep depths within the sand aquifer. The difficulty in detennining the groundwater flow regime at the field site based on water level measurements was compounded by the limited number of available piezometers and the presence of very small lateral gradients which are sensitive to small field measurement errors. Numerical density-dependent flow modelling provided valuable information regarding the groundwater flow conditions at the site. The good fit between the model results and the field salinity data for Simulation 1 indicates that the model is a good representation of field conditions. The model results indicate that a shallow gradient toward the river exists (the calibrated gradient for Simulation 1 was 0.000315) and that the average salinity at the base of the river is approximately 19 ppt. Note, however, that due to unknowns in the island- and river-side model boundary conditions these calibrated values are not unique. The groundwater velocity results for Simulation 1 indicate that the average groundwater flow regime at the site is similar to that which is typical for areas of saltwater intrusion (circulation of saline water into and then out of the aquifer forming a saline wedge below river-ward flowing fresh groundwater). The model results for Simulation 2 indicate that the groundwater flow regime at the site likely varies due to seasonal changes in river discharge rate which affect river salinity (and 146 density) and river level. Tidal fluctuations could also affect the groundwater flow regime. During periods of high seasonal discharge, river levels may be high enough to cause fresh water in the river to intrude the aquifer and cause groundwater to flow in a southerly direction (away from the river). During periods of low seasonal discharge rate and/or low tide, flow toward the river of both saline and fresh groundwater could occur (indicated on Figure 6). Due to the complex nature of the transient water level and salinity conditions in the Fraser River, details of the changing flow regime at the site in response to changing river conditions were not fully characterized. The calibration of the numerical model indicated that there may be slight variations in aquifer permeability across the site that were not identified based on the original interpretation of the geological conditions. Model calibration indicated that the Medium Sand and Fine and Medium Sand units may have a lower permeability in the area of the wedge toe (lower than that estimated based on the field data for these units at other locations in the aquifer). The inclusion of this lower permeability zone in the model domain was necessary to obtain a better match between field salinity data and model results. A subsequent review of the available field data suggested that the presence of a zone of slightly lower permeability in the area of the wedge toe is reasonable. Although a close match between Simulation 1 results and field data was obtained, some minor discrepancies were evident. The sensitivity analysis does not indicate that modification of the model input parameters investigated could significantly improve the match between Simulation 1 results and field data. Thus the difficulty in precisely modelling saltwater intrusion at the field site is illustrated. Additional field investigation could aid in the characterization of site conditions to provide a basis for the development of a model domain that is an even closer representation to site conditions than Simulation 1 (e.g. 3-dimensional model, model that accounts for transient effects at river-side boundary, model that incorporates small scale geological heterogeneities at the site). 147 6.2 Saltwater Intrusion and Groundwater Flow in the Fraser River Delta 6.2.1 Conceptual Model of Regional Groundwater Flow This section presents a conceptual model of regional groundwater flow for the Fraser River delta. This conceptual model is based on the regional groundwater flow model for the Fraser River delta developed by Ricketts (1998), available chemical data for the Fraser River delta which indicates areas of saline groundwater (Figure 16), and the local groundwater flow regime at the field site as indicated by the results of Simulation 1. Figures 61 and 62 illustrate the conceptual model in north-south and east-west cross sections. The sand aquifer has been identified by Monahan et al. (1993) to be essentially continuous across the delta plain. As such, it is assumed that saltwater intrusion into the sand aquifer occurs at locations adjacent to the Fraser River estuaries (Main Channal, North Arm, and Middle Arm) where saline water is present in the river. Saltwater intrusion from the Georgia Strait into the sand aquifer at the western margin is also expected. At locations where saltwater intrusion into the sand aquifer occurs, a saline wedge, similar to that identified at the field site, is inferred to be present. Local density-dependent flow regimes would develop in these areas; involving the migration of saline groundwater in the sand aquifer toward the wedge toe and then circulation back toward the river/ocean under the influence of river-ward/seaward freshwater gradients (see Figures 41c and 41d). The flow system in these areas may be complicated by the transient effects due to changing salinity and water level in the Fraser River. As indicated on Figure 16, deeper saline water within the delta slope and prodelta deposits was also identified at locations tested by Hunter et al. (1994). At locations near the Georgia Strait, saline water in the deeper sediments is inferred in the conceptual model to be the result of saltwater intrusion from the ocean. The typical density-dependent flow system for areas of saltwater intrusion is inferred to develop adjacent to the Georgia Strait 148 149 (Figure 62). At locations inland from the Georgia Strait where saline water was identified in the deeper, fine-grained deltaic sediments, the saline water is inferred to be trapped seawater (old seawater present at the time of sediment deposition) as suggested by Simpson and Hutcheon (1995). The effects that trapped saline water could have on the regional flow regime (due to density-dependent flow), are not considered here. According to the regional groundwater flow model developed by Ricketts (1998), fresh groundwater from the upland areas surrounding the delta is eventually discharged at the delta margins and/or to the Fraser River. This fresh groundwater is inferred to migrate upward beneath the delta to layers of higher permeability sediments where it flows seaward. The flow of fresh groundwater through higher permeability zones within the delta slope deposits may circumvent zones of fine grained deposits where trapped saline groundwater is inferred to be present (Figures 61 and 62). (Note that data are not available to specifically locate zones of fresh/saline groundwater within the deeper delta slope deposits). Ultimately, this fresh groundwater undergoes submarine discharge through the delta slope deposits or, at some locations, through the sand aquifer (Ricketts, 1998). Ricketts (1998) suggests that at locations of submarine discharge of fresh groundwater, the seaward or river-ward freshwater gradient may be sufficient to inhibit saltwater intrusion from the ocean/estuary. Available chemical data for the Fraser delta identified four locations near the mouth of the Fraser River (identified on Figure 16) where saline groundwater was not present at the depths investigated. At these locations, the discharge of fresh water to the Georgia Strait may inhibit saltwater intrusion. Figure 62 further illustrates this effect where fresh groundwater is shown to flow seaward through the deep deltaic sediments and inhibit saltwater intrusion at the delta margin. The available data indicate that the regional groundwater flow regime for the Fraser River delta is the result of a complex combination of density-dependent flow (due to trapped saline water, saltwater intrusion from the ocean, and saltwater intrusion from the Fraser 151 River estuaries) as well as the effects of upward and lateral flow of fresh groundwater beneath the delta and shallow freshwater flow systems within the sand aquifer. Further research is required to more thoroughly understand the details of the regional hydrogeology of the Fraser River delta. 6.2.2 Groundwater Geochemical Facies and the Regional Flow System Based on the conceptual model for regional groundwater flow, shallow fresh groundwater (derived from recharge to the delta plain and flow through the sand aquifer) would be expected to have a chemistry that is characteristically different from the deep fresh groundwater (derived from regional flow from the upland areas surrounding the delta). Similarly, saline groundwater within the sand aquifer adjacent to the Fraser River (derived from saltwater intrusion from the Fraser River) would be expected to have a chemistry that is different from saline groundwater within the deeper deltaic sediments (inferred to be trapped seawater within the delta slope deposits). The analysis of limited geochemical data from the field site and from Simpson and Hutcheon (1995) (Section 3.4.4) indicated distinct chemical differences between samples of shallow fresh groundwater, deep fresh groundwater, saline groundwater within the sand aquifer, and saline groundwater within the deeper deltaic sediments. Should further research confirm distinct chemical characteristics for the different groundwater facies (i.e. shallow fresh, deep fresh, saline water within the sand aquifer, and saline water within the delta slope deposits), groundwater chemistry may be a useful tool to aid in the assessment of the regional groundwater flow system for the Fraser River delta. 152 6.3 Implications for Contaminated Site Assessment and Remediation Problems 6.3.1 Potential Effects of Groundwater Flow System on Contaminant Migration As illustrated on Figure 2, there has been significant industrial development adjacent to the Fraser River. At these locations, there is a potential for groundwater contamination due to spills or slow releases of products such as fuel, oil, solvents, wood preservation chemicals, and other substances associated with historical or existing industrial operations. Should these contaminants reach the subsurface environment, a dissolved contaminant plume would form. The size, extent, and migration of the dissolved contaminant plume would depend in part on the groundwater flow regime in the area of contamination. Thus, in order to effectively characterize and/or predict the extent and migration of a dissolved contaminant plume (required for contaminated site assessments and site remediation projects), the groundwater flow regime must be well understood. In the area of the field site, this thesis has shown that saltwater intrusion from the Fraser River estuary occurs and forms a local density-dependent groundwater flow system which is illustrated by Simulation 1 (Figures 41c and 41d). It is hypothesized (as discussed in Section 6.2) that at other areas adjacent to the Fraser River estuary, similar flow systems are present within the sand aquifer. Figure 63 a illustrates the anticipated effect of the density-dependent groundwater flow regime determined for the field site on a hypothetical dissolved contaminant plume. Figure 63 a shows that landward flow of groundwater at the base of the aquifer causes the plume to migrate away from the river, and river-ward flow at the top of the aquifer causes the plume to migrate toward the river. Without knowing the possible effects of saltwater intrusion on the local groundwater flow system, fresh groundwater flow toward the river may be incorrectly assumed for a contaminated site based on common hydrogeological principles (e.g. rivers typically act as 153 Fraser River MonltoAig Wei 0 50 100m Horizontal Scale (m) Figure 63a:Effects of the groundwater flow regime at the field site on the migration of a hypothetical dissolved contaminant plume. Dissolved Contaminant Plume .^U^ Dense Nonaqueous Phase Uqdd CONAPL) 2.S— Estimated Salinity Contour (ppt) Fraser River Monitoring Wei 0 50 100m Horizontal Scale (m) Figure 63b: Effects of the groundwater flow regime on a hypothetical dissolved contaminant plume assuming fresh groundwater flow toward the Fraser River. <^~> Dfcsotvod Contaminant f V r r i e Dense Nonaqueous Phase liquid CDNAPO 154 discharge points for groundwater). Figure 63b illustrates a hypothetical dissolved contaminant plume formed within a groundwater flow system where saltwater intrusion is not considered (i.e. fresh groundwater flows toward the river). On Figure 63b, the contaminant plume is shown to migrate toward the river at all depths forming a shape that is significantly different from that depicted on Figure 63 a. The design of a monitoring well network or a groundwater remediation system is based on knowledge or assumptions of the groundwater flow regime. Thus effective designs are dependent on accurate characterization of the groundwater flow regime. For example, monitoring wells installed to sample at points A and B (Figure 63 a) may be located to assess contamination at locations that are thought to be downgradient of a dissolved contaminant plume (based on the assumption of fresh groundwater flow toward the river). Sampling at points A and B, however could provide data that underestimate the depth extent of contamination (well B would not intersect contamination at depth). With respect to the design of a remediation system, a pumping well at location B (Figure 63 a) would require a sufficient pumping rate to reverse local gradients to draw the dissolved contaminants toward the well. At the field site small landward gradients are indicated at the base of the saline wedge. At other locations adjacent to the Fraser River that are seaward of the field site (where salinity/density in the river is higher), however, higher landward gradients may be present. Inadequate pumping rates may be designed if the effects of saltwater intrusion on the groundwater flow system are not considered. 6.3.2 Groundwater Flow Regime Characterization Strategies for Sites Adjacent to the Fraser River Section 6.3.1 demonstrated the importance of an accurate determination of the groundwater flow regime for contamination studies at sites located adjacent to the Fraser River estuaries. The collection of field data is one tool which can be used to understand 155 the hydrogeological conditions. Water level data can be used to assess the groundwater flow directions in areas of saltwater intrusion (variable density systems) if the water levels are converted to equivalent freshwater head (discussed in Section 3.3.2.3). (Note that if fresh groundwater is incorrectly assumed and actual water levels are used to infer groundwater flow direction, erroneous predictions could result). For areas affected by tidal fluctuations, averaging of consecutive water level measurements may be required to determine average groundwater flow directions. Seasonal changes in the river may also influence the groundwater flow regime in areas adjacent to the Fraser River estuaries as indicated by Simulation 2 presented in this thesis. Although it is possible to collect and analyze field data to characterize site hydrogeological conditions in areas adjacent to the Fraser River, transient effects, very small horizontal gradients, and unavoidable measurement errors make characterization difficult. The difficulties in using field data to determine the groundwater flow regime for a site adjacent to the Fraser River (even when density and tidal effects are considered) were demonstrated by the field data analysis completed for this thesis (discussed in Section 6.1). A numerical model can provide insight into the groundwater flow regime; particularly for sites where the collection/analysis of field data is complex or does not yield abundant information regarding the hydrogeological conditions. For the site investigated for this thesis, the field data was not adequate to thoroughly assess the groundwater flow conditions, however, the modelling procedure confirmed the presence of a flow system typical for areas of saltwater intrusion. The modelling also provided valuable information regarding the complexities of the system (e.g. seasonal effects and the changes at the river-side model boundary) which could be useful to guide future field investigation. In general, the determination or confirmation of the groundwater flow regime through numerical modelling can be used as a decision making tool for groundwater contamination assessments or remediation projects at areas adjacent to the Fraser River. Numerical model results could provide data to aid in the design of a monitoring well network or groundwater remediation pumping system and could be useful for other hydrogeological problems. 157 BIBLIOGRAPHY Ages, A. 1979. The Salinity Intrusion in the Fraser River: Salinity, Temperature and Current Observations, 1976, 1977. Pacific Marine Science Report 79-14. Institute of Ocean Sciences, Patricia Bay, Sidney, BC, 192 pp. Ages, A. 1988. The Salinity Intrusion in the Fraser River: Time Series of Salinities, Temperatures and Currents 1978, 1979. Canadian Data Report of Hydrography and Ocean Sciences 66, Institute of Ocean Sciences, Department of Fisheries and Oceans, Sidney, BC. Ages, A. and A. Woollard. 1976. The Tides in the Fraser River Estuary. Pacific Marine Science Report 76-5. Institute of Ocean Sciences, Patricia Bay, Victoria, BC, 100 pp. Armstrong, J. E. and S. R. Hicock. 1976. Map 1486A, Surficial Geology, Vancouver, British Columbia. Geological Survey of Canada. BC Department of Agriculture. 1974. Climate of British Columbia, Tables of temperature and precipitation, Climatic Normals 1941-1970, extremes of record. Birdie, T., N . Blandford, and B. McGurk. 1993. Application of a Density-Dependent Flow and Transport Model to a Coastal Aquifer System. Engineering Hydrology, C. Y. Kuo, editor, pp. 1165-1170. Bair, E. S. and T. D. Lahm. 1996. Variations in Capture-Zone Geometry of a Partially Penetrating Pumping Well in and Unconfined Aquifer. Ground Water, v. 34, no. 5, pp. 842-852. Canadian Hydrographic Service. 1996. Marine Chart 3491, Fraser River, BC. Cooper, H. H. 1959. A Hypothesis Concerning the Dynamic Balance of Fresh Water and Salt Water in a Coastal Aquifer. Journal of Geophysical Research, v. 64, no. 4, pp. 461-467. Cooper, H. H., F. A. Kohour, H. R. Henry, and R. E. Glover. 1964. Sea Water in Coastal Aquifers, US Geological Survey Water-Supply Paper 1613-C, 84 pp. Custodio, E. and G. A. Bruggeman. 1987. Groundwater Problems in Coastal Areas. United Nations Educational, Scientific and Cultural Organization (UNESCO), Belgium, 596 pp. De Marsily, G. 1986. Quantitative Hydrogeology. Academic Press Inc., USA, 440 pp. Das Gupta, A. and P. N . D. D. Yapa. 1982. Saltwater Encroachment in an Aquifer: A Case Study. Water Resources Research, v. 18, no. 3, pp. 546-556. Environment Canada. 1995. Hydronometric data for discharges of the Fraser River at Hope, personal communication. Environment Canada. 1997. Minimum and maximum instantaneous water elevations, North Arm Fraser River at Vancouver, personal communication. Erskine, A. D. 1991. The Effect of Tidal Fluctuation on a Coastal Aquifer in the UK. Ground Water, v. 29, no. 4, pp. 556-562. Freeze, R. A. and J. A. Cherry. 1979. Groundwater. Prentice-Hall, Inc., USA, 604pp. Frind, E. O. 1982a. Simulation of long-term transient density-dependent transport in groundwater. Advances in Water Resources, v. 5, June, pp. 73-88. Frind, E. O. 1982b. Seawater intrusion in continuous coastal aquifer-aquitard systems. Advances in Water Resources, v. 5, June, pp. 89-97. Frind, E.O., E. A. Sudicky, and S. L. Schellenberg. 1987. Micro-scale modelling in the study of plume evolution in heterogeneous media. Stochastic Hydrology and Hydraulics, v. 1, pp. 263-279. Gaganis, P. 1996. Field Techniques in Groundwater Hydrology, Hydrogeology Field School Report, University of British Columbia. Gelhar, L. W., C. Welty, and K. R. Rehfeldt. 1992. A Critical Review of Data on Field-Scale Dispersion in Aquifers. Water Resources Research, v. 28, no. 7, pp. 1955-1974. Greater Vancouver Regional District (GVRD). 1974. Land Use Map, Georgia Strait Urban Region. Henry, H. R. 1959. Salt Intrusion into Fresh-Water Aquifers. Journal of Geophysical Research, v. 64, no. 11, pp. 1911-1919. Hunter, J. A., J. L. Luternauer, M. C. Roberts, P. A. Monahan, and M . Douma. 1994. Gorehole geophysics logs, Fraser River delta (92G), British Columbia. Geological Survey of Canada Open File 2841, 30 pp. Hunter, J. A., J. L. Luternauer, and P. J. Kurfurst. 1996. Regional electromagnetic soundings with the Geonics EM-34 conductivity meter in the southern Fraser River Delta, British Columbia. Geological Survey of Canada Open File 3359, 141 pp. 159 Huyakorn, P. S., P. F. Anderson, J. W. Mercer, and FL O. White, Jr. 1987. Saltwater Intrusion in Aquifers: Development and Testing of a Three-Dimensional Finite Element Model. Water Resources Research, v. 23, no. 2, pp. 293-312 Jorgensen, D. G., T. Gogel, and D. C. Signor. 1982. Determination of Flow in Aquifers Containing Variable-Density Water. Ground Water Monitoring Review, Spring, pp. 40-45. Kohout, F. A. 1960. Cyclic Flow of Salt Water in the Biscayne Aquifer of Southeastern Florida. Journal of Geophysical Research, v. 65, no.7, pp. 2133-2141. Kostaschuk, R. A., M . A. Church, and J. L. Luternauer. 1989. Bedforms, bed material, and bedload transport in a salt-wedge estuary: Fraser River, British Columbia. Kostaschuk, R. A. and L. A. Atwood. 1990. River discharge and tidal controls on salt-wedge position and implications for channel shoaling: Fraser River, British Columbia. Canadian Journal of Civil Engineering, v. 17. Pp. 452-459. Lee, C. and R. T. Cheng. 1974. On Seawater Encroachment in Coastal Aquifers. Water Resources Research, v. 10, no. 5, pp. 1039-1043. Lusczynski, N . J. 1961. Head and Flow of Ground Water of Variable Density. Journal of Geophysical Research, v. 66, no. 12, pp. 4247-4256. Mercer, J. W., S. P. Larson, and C. R. Faust. 1980. Simulation of Salt-Water Interface Motion. Ground Water, v. 18, no. 4, pp. 374-385. Monahan, P. A., J. L. Luternauer, and J. V. Barrie. 1993. A Delta Plain Sheet Sand in the Fraser River Delta, British Columbia, Canada. Quaternary International, v. 20, pp. 27-38. Oberlander, P.L. 1989. Fluid Density and Gravitational Variations in Deep Boreholes and Their Effect on Fluid Potential. Ground Water, v. 27, no. 3, pp. 341-350. Pinder, G. F. and H. H. Cooper. 1970. A Numerical Technique for Calculating the Transient Position of the Saltwater Front. Water Resources Research, v. 6, no. 3, pp. 875-882. Reilly, T. E. and A. S. Goodman. 1985. Quantitative Analysis of Saltwater-Freshwater Relationships in Groundwater Systems - A Historical Perspective. Journal of Hydrology, v. 80, pp. 125-160. 160 Ricketts, B. D. 1998. Groundwater Flow Beneath Fraser River Delta, British Columbia; A Preliminary Model. Unpublished Manuscript, Geological Survey of Canada, 43 pp. Segol, G. and G. F. Pinder. 1976. Transient Simulation of Saltwater Intrusion in Southeastern Florida. Water Resources Research, v. 12, no. 1, pp. 65-70. Series, M. E. 1991. Determining the mean hydraulic gradient of ground water affected by tidal fluctuations. Ground Water, v. 29, no. 4, pp. 549-555. Shamir, U. and G. Dagan. 1971. Motion of the Seawater Interface in Coastal Aquifers: A Numerical Solution. Water Resources Research, v. 7, no. 3, pp. 644-657. Shikaze, S. G., E. A. Sudicky, R. G. McLaren, and J. E. VanderKwaak. 1996. A Two-Dimensional Numerical Model for Transient, Density-Dependent Flow and Solute Transport in Discretely-Fractured, Saturated Porous Media. Groundwater Simulations Group, Waterloo Centre for Groundwater Research, University of Waterloo, Waterloo, Ontario. Simpson, G. and I. Hutcheon. 1995. Pore-Water Chemistry and Diagenesis of the Modern Fraser River Delta. Journal of Sedimentary Research, v. A65, no. 4, pp. 648-655. Thomas, D. 1996. Hydrogeology Field School Report, University of Calgary. Underwood, M. R., F. L. Peterson, and C. I. Voss. 1992. Groundwater Lense Dynamics of Atoll Islands. Water Resources Research, v. 28, no. 11, pp. 2889-2902. UNESCO. 1981. Tenth report of the joint panel on oceanographic tables and standards. Unesco technical papers in marine science 36, Sidney, B.C., Canada, 25 pp. Williams, H. F. L. and M. C. Roberts. 1989. Holocene sea-level change and delta growth: Fraser River delta, British Columbia. Canadian Journal of Earth Sciences, v. 26, pp. 1657-1666. Wood, E. A. 1996. Comparison of Hydraulic Conductivity Test Methods for the Kidd2 Site in Richmond, B.C. BASc. Thesis, University of British Columbia, 34 pp. Xue, Y., C. Xie, and J. Wu. 1995. A three-dimensional miscible transport model for seawater intrusion in China. Water Resources Research, v. 31, no. 4, pp. 903-912. Zhang, G. 1995. Numerical Models of Saltwater Intrusion From Brackish Canals in Southeast Florida. PhD. Thesis, The Florida State University College of Arts and Sciences, 140 pp. 161 APPENDIX I PIEZOMETER, WELL, BOREHOLE A N D CPT LOGS 162 PIEZOMETER 101 Site Location: UBC Field Site, Richmond, BC Drilling Date: 1995 Borehole Diameter: 0.17 m Borehole Depth: 18.0 m Weill Material: PVC Well Diameter: 0.05 m WEII Screen Length: 1.5 m Well Depth (below grade): 18.0 m Top of Well Elevation: 1.370 m Depth (m) 0 1 2 -3 -4 -5 -6 -7 -8 -9 -1 0 -1 1 -1 2 -13 — 1 4 -1 5 -16 17H 18 19H 20 101 Backfill Bentonlte Seal \ Sand Backfill Ground Surface £ 3 — 15.58 m — 15.88 m — 16.5 m 18.0 m 163 PIEZOMETER 102 Site Location: UBC Field Site, Richmond, BC Drilling Date: 1995 Borehole Diameter: 0.17 m Borehole Depth: 12.80 m Weill Material: PVC Well Diameter: 0.05 m WEII Screen Length: 1.5 m Well Depth (below grade): 12.80 m Top of Well Elevation: 1.371 m Depth (m) 0 1 -2 -3 -4 -5 -6 -7 -8 -9 -1 0 -1 1 -12 13 14H 15 1 6 -1 7 -18 19 H 20 102 Ground Surface Backfill Bentonlte Seal -— Sand Backfill S3 — 10.38 m — 10.68 m — 11.30 m 12.80 m 164 PIEZOMETER 103 Depth (m) 103 n Ground Surface Site Location: UBC Field Site, Richmond, BC u 1 -Drilling Date: 1995 Borehole Diameter: 0.17 m 2 -Borehole Depth: 18.05 m 3 -Weill Material: PVC Well Diameter: 0.05 m 4 -WEII Screen Length: 1.5 m 5 -Well Depth (below grade): 18.05 m Top of Well Elevation: 1.366 m 6 — 7 -8 -9 -1 0 -1 1 -1 2 -13 — 1 4 -Backflll 1 5 -1 6 -Bentonlte Seal — 15.63 m 15.93 m 1 7 -w — 16.55 m 1 8 -Sand ^ Backfill r — 18.05 m 1 9 -2 0 -165 PIEZOMETER 104 Site Location: UBC Field Site, Richmond, BC Drilling Date: 1995 Borehole Diameter: 0.17 m Borehole Depth: 12.84 m Weill Material: PVC Well Diameter: 0.05 m WEII Screen Length: 1.5 m Well Depth (below grade): 12.84 m Top of Well Elevation: 1.362 m Depth (m) 0 1-2 -3 4 H 5 6 -7 -8 -9 -10-11 12 13 14-15-16 17H 18 19-20-104 Backfill Bentonlte Seal — Sand ' Backfill Ground Surface ^ 3 ^3 — 10.42 m — 10.72 m — 11.34 m — 12.84 m 166 PIEZOMETER 105 Depth (m) 105 n Ground Surface Site Location: UBC Field Site, Richmond, BC u 1-Drilling Date: 1995 Borehole Diameter: 0.17 m 2 -Borehole Depth: 18.11 m 3 -Weill Material: PVC Well Diameter: 0.05 m 4 -WEII Screen Length: 1.5 m 5 -Well Depth (below grade): 18.11 m Top of Well Elevation: 1.355 m 6 -7 -8 -9 -10-11-12-13-14-Backfill 15-16-Bentonlte Seal — 15.69 m — 15.99 m 17-• — 16.61m 18-Sand ^ Backfill V — 18.11m • -:• ~ 19-20-— 167 PIEZOMETER 106 Depth (m) 106 n Ground Surface Site Location: UBC Reld Site, Richmond, BC u 1-Drilling Date: 1996 Borehole Diameter: 0.17 m 2 -Borehole Depth: 12.74 m 3 -Weill Material: PVC Well Diameter: 0.05 m 4 -WEII Screen Length: 1.5 m 5 -Well Depth (below grade): 12.74 m Backfill Top of Well Elevation: 1.368 m 6 -7 -8 -9 -10-11-Bentonlte Seal 10.32 m 10.62 m — 11.24 m 12- Sand Backfill • — 12.74 m 13-14-15-16-17-18-19-20-— 168 PIEZOMETER 107 Depth (m) 107 n Ground Surface Site Location: UBC Field Site, Richmond, BC u 1 -Drilling Date: 1995 Borehole Diameter: 0.17 m 2 -Borehole Depth: 17.99 m 3 -Weill Material: PVC Well Diameter: 0.05 m 4 -WEII Screen Length: 1.5 m 5 — Well Depth (below grade): 17.99 m Top of Well Elevation: 1.486 m 6 — 7 -8 -9 — 1 0 -1 1 -1 2 -1 3 -1 4 -Backfill 1 5 -1 6 -Bentonlte Seal — 15.57 m — 15.87 m 1 7 -— 16.49 m 1 8 -Sand ^ Backfill T — 17.99 m 1 9 -2 0 -— 169 PIEZOMETER 108 Depth (m) 108 n Ground Surface Site Location: UBC Field Site, Richmond, BC KJ 1-Drilling Date: 1995 Borehole Diameter: 0.17 m 2 -Borehole Depth: 12.99 m 3 -Weill Material: PVC Well Diameter: 0.05 m 4 -WEII Screen Length: 1.5 m 5 — Well Depth (below grade): 12.99 m Top of Well Elevation: 1.486 m 6 -7 -8 -9 -Backfill 10-11-Bentonlte Seal — 10.57 m — 10.87 m — 11.49 m 12-13 — Sand ^ Backfill — 12.99 m 14-15-16-17-18-19-20-— 170 PIEZOMETER 111 Site Location: UBC Field Site, Richmond, BC Drilling Date: March 27,1996 Borehole Diameter: 0.152 m Borehole Depth: 5.18 m Weill Material: PVC Well Diameter: 0.05 m WEII Screen Length: 0.76 m Well Depth (below grade): 4.88 m Top of Well Elevation: 1.286 m Depth (m) 0 1-2 -3 -4 5 6 7 -8 -9 10-11-12-13-14-15-16 17-1 18 19-20-111 Backfill • Bentonite Seal Sand Backfill Ground Surface 73 — 1.2 m 3.66 m 4.11 m •4.88 m 5.18 m 171 PIEZOMETER 112 Site Location: UBC Field Site, Richmond, BC Drilling Date: March 27,1996 Borehole Diameter: 0.162 m Borehole Depth: 4.88 m Weill Material: PVC Well Diameter: 0.05 m WEII Screen Length: 0.76 m Well Depth (below grade): 4.88 m Top of Well Elevation: 1.326 m Depth (m) 0 IH 2 3 -4 -5 6 H 7 8 -9 -10-11-12-13-14-15-16 17H 18 19-20-112 Backfill • Bentonlte Seal 4 Sand Backfill Ground Surface — 1.2 m — 3.66 m — 4.11 m — 4.88 m 172 PIEZOMETER 113 Site Location: UBC Reld Site, Richmond, BC Drilling Date: March 27,1996 Borehole Diameter: 0.152 m Borehole Depth: 2.36 m Weill Material: PVC Well Diameter: 0.05 m WEII Screen Length: 0.76 m Well Depth (below grade): 2.36 m Top of Well Elevation: 1.298 m Depth (m) 0 1 2 -3 -4 5 6 -7 -8 -9 i o n n 12 H 13—1 14 1 5 -1 6 -17 18-1 19 20—1 113 Bentonlte Seal — Sand Backfill S3 Ground Surface 1.6 m 2.36 m 173 PIEZOMETER 114 Site Location: UBC Held Site, Richmond, BC Drilling Date: March 27,1996 Borehole Diameter: 0.152 m Borehole Depth: 5.1 m Weill Material: PVC Well Diameter: 0.05 m WEII Screen Length: 0.76 m Well Depth (below grade): 5.1 m Top of Well Elevation: 1.343 m Depth (m) 0 1 -2 -3 -4 -5 6 -7-8 9 -10 -1 1 -12 13 — 14 -15-1 6 -17-18-1 9 -20 114 Backfill • Bentonlte Seal Sand Backfill Ground Surface — 1.2 m — 3.66 m — 4.34 m —5.1 m 174 PUMPING WELL Site Location: UBC Field Site, Richmond, BC Borehole Diameter: 0.25 m Borehole Depth: 20.03 m Weill Material: PVC Well Diameter: 0.15 m WEII Screen Length: 6.03 m Well Depth (below grade): 20.03 m Depth (m) 0 1-2 -3 -4 -5 -6 7 -8 -9 10-11-12-13-14 15-16-17-18 19-20-PW Backfill Bentonlte Seal Sand Backfill Ground Surface 12.2 m 13.1 m 14.0 m — 20.03 m 175 WEST B A Y W E L L Deptr (m) W B n — Ground Surface Site Location: UBC Field Site, Richmond, BC \j 2 - — 2.5 m Casing Diameter: 0.108 m Borehole Depth: 38.57 m 4 -6 -— 4.0 m 8 -1 0 -1 2 -Packers Inflated between measurement ports — 8.0 m — 11.0 m 1 4 - — 14.0 m 1 6 - — 16.5 m 1 8 -2 0 -— 19.5 m 2 2 -Measurement Port — 22.0 m 2 4 -— 23.5 m 2 6 -2 8 - — 28.0 m 3 0 -3 2 -— 31.0m 3 4 -3 6 -— 35.5 m 3 8 -4 0 -— 176 MONITORING WELL LOG Site Location: UBC Field Site, Richmond, BC Drilling Date: 1996 03 26 Drilling Method: Sonic Drill Borehole Diameter. 0.152 m Borehole Depth: 22.07 m Well Material: PVC Glued Pipe with Polyethylene sample tubes Well Screen Length: N/A Screen Slot Size: N/A Well Diameter: PVC pipe = 0.0508 m Sample Tubes = 0.478 cm ID, 0.787 cm OD Well Depth (below grade): 22.07 m Top of Well Elevation (from survey): Notes: Multi-level sampling well constructed with PVC pipe. Fifteen holes were drilled in the PVC pipe for the sampling ports. Fifteen polyethylene tubes were threaded inside the PVC pipe and through each hole. A nylon mesh was fastened to the end of each sampling tube. Depth| (m) O 1-2~ 3-4~ 5-6~ 7 8-9~ 10-11-12" 13-14-15-16-17-18-19-20-21 22—1 Ground Surface Native Sand Collapse W E L L NO: W l - l m Bentonite. Grout 6 m Sample p o r t Depth Number (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 8.08 9.07 10.08 11.08 12.08 13.07 14.08 15.08 16.17 17.08 18.08 19.08 20.06 21.07 22.03 MONITORING WELL L O G Site Location: UBC Field Site, Richmond, BC Drilling Date: 1996 03 26 Drilling Method: Sonic Drill Borehole Diameter: 0.152 m Borehole Depth: 22.1m Well Material: PVC Glued Pipe with Polyethylene sample tubes Well Screen Length: N/A Screen Slot Size: N/A Well Diameter PVC pipe = 0.0508 m Sample Tubes = 0.478 cm ID, 0.787 cm OD Well Depth (below grade): 22.1 m Top of Well Elevation (from survey): Notes: Multi-level sampling well constructed with PVC pipe. Fifteen holes were drilled in the PVC pipe for the sampling ports. Fifteen polyethylene tubes were threaded inside the PVC pipe and through each hole. A nylon mesh was fastened to the end of each sampling tube. Depth, (m) 0 1" 2~ 3-4~ 5-6-7~ 8-9-10-11" 12-13-1 14 15-1 16-17 18-19-20-21 22—I Ground Surface Native Sand Collapse 177 WELL NO: W 2 - lm Bentonite Grout - 6m Sample P 0 I t Depth Number (m) 1 8.04 2 3 4 5 6 7 8 9 10 11 12 13 14 15 9.04 10.05 11.05 12.05 13.07 14.07 15.07 16.20 17.09 18.09 19.09 20.09 21.09 22.09 MONITORING WELL LOG Site Location: UBC Field Site, Richmond, BC Drilling Date: 199603 27 Drilling Method: Sonic Drill Borehole Diameter 0.152 m Borehole Depth: 21.5 m Well Material: PVC Glued Pipe with Polyethylene sample tubes Well Screen Length: N/A Screen Slot Size: N/A Well Diameter: PVC pipe = 0.0508 m Sample Tubes = 0.478 cm ID, 0.787 cm OD Well Depth (below grade): 21.5 m Top of Well Elevation (from survey): Notes: Multi-level sampling well constructed with PVC pipe. Fifteen holes were drilled in the PVC pipe for the sampling ports. Fifteen polyethylene tubes were threaded inside the PVC pipe and through each hole. A nylon mesh was fastened to the end of each sampling tube. Depthj (m) 0 1-2-3~ 4-5-6 7~ 8-9~ 10-11 12-13-14-15-16-17-18-19-20-21-22-Ground Surface Native Sand ^ Collapse m~ WELL NO: W3 lm Bentonite Grout - 6 m Sample Depth Port ,S\ Number (m) 7.50 8.50 9.51 10.53 11.52 12.52 2 3 4 5 6 7 8 9 10 11 12 13 14 15 13.50 14.50 15.61 16.50 17.50 18.49 19.49 20.49 21.47 DEPTH (m) INTERPRETED STRATIGRAPHY 179 1 - 2 - 3 - 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 CLAYEY SILT Clayey silt,trace fine sand, trace to some organic matter (leaves.wood, etc.), light grey. SILTY SAND Laminated silty fine sand Laminated medium and fine sand, some silt ! Z Z Z Z Z Z _ Z Z Z -M^T"TTi~anH Tifjp."^mrT sbfieZsiltZ Laminated silty fine sand _ Medium sand ' T ^ r r i T n a t f d s i T I y T T n i T ' i a n r T ' FINE AND MEDIUM SAND Medium sand, some fine sand, grey, occasional silty sand lenses MEDIUM SAND Medium Sand, uniform SILTY CLAY Clayey silt to silty clay, thinly laminated, light grey SOIL LOG: BOREHOLE LOCATION: Approx. 320 m South of River DRILLED: March 1995 METHOD: Mud rotary drill and core barrel DEPTH INTERPRETED STRATIGRAPHY RESISTIVITY DATA 1 FILL 1 1 1 1 1 1 1 10 20 30 40 50 60 70 CLAYEY SILT Resistivity (ohm-m) — 2 Sandy or clayey silt to silty clay and clay, trace fine sand. DATA NOT AVAILABLE — 3 3.5m — 4 SILTY SAND - 5 Silty sand with some layers of fine sand, some clayey and sandy silt lenses 6 7 8 6.0m - FINE AND MEDIUM SAND Medium and fine sand, uniform, some fine sand lenses, unit thickness unknown - 9 MEDIUM SAND Medium Sand, uniform — 10 - 11 - 12 - 13 -14 15 16 .... Fresh-Saline Transition Zone or "Interface" - 17 - 18 - 19 - 20 - 21 22 23 22.0m Saline-Fresh Transition Zone - SILTY CLAY Clayey silt to silty clay SOIL LOG: CPT K9301 LOGGED BY UBC Civ. Eng. DRILLED: 1993 LOCATION: Approx. 375m south of River METHOD: CPT DEPTH (m) INTERPRETED STRATIGRAPHY RESISTTVITY DATA 181 FILL 1 1 1 1 1 1 1 10 20 30 40 50 60 70 - 1 - 2 - 3 CLAYEY SILT Sandy or clayey silt to silty clay and clay, trace fine sand. 3.5m Resistivity (ohm-m) • DATA NOT AVAILABLE - 4 - 5 SILTY SAND Silty sand with some layers of fine sand, some clayey and sandy silt lenses - 6 - 7 - 8 8.7m - 9 - 1 0 - 11 - 12 - 13 FINE AND MEDIUM SAND Medium and fine sand, uniform, some fine sand lenses, unit thickness unknown MEDIUM SAND Medium Sand, uniform r Fresh-Saline Transition Zone or "Interface" - 14 - 15 - 16 - 17 - 18 - 19 - 20 20.5m Saline-Fresh Transition Zone - 21 - 22 SILTY CLAY Clayey silt to silty clay - 23 SOIL LOG: CPT K9308 LOGGED BY UBC Civ. Eng. DRILLED: 1993 LOCATION: Approx. 288m south of River METHOD: CPT DEPTH INTERPRETED STRATIGRAPHY RESISTIVITY DATA Sensitive Fines 1 1 1 10 20 30 1 I I 1 40 50 60 70 1 Sand, jravefly (fill} CLAYEY SILT Resistivity (ohm-m) 2 Silty sand to silty clay and clay, - 3 some organic matter 3.6m - 4 SILTY SAND 43.3 «, - 5 Silty sand with some layers of fine sand \ \ - 6 \ 54.5 • — 7 *** - 8 f 29.1 9 9.1m / 1 MEDIUM SAND 25.0 f — 10 10.4m Medium Sand, uniform - 11 SILTY SAND 11.7m 20.2 ? \ — 12 FINE AND MEDIUM SAND 23.9 * - 13 Medium and fine sand, some fine sand lenses 13.9m / / / 14 18.7? • - 15 MEDIUM SAND i 1 1 — 16 Medium Sand, uniform 21.4* 1 - 17 1 1 - 18 18.9 j — / Possible •mm 19 / Fresh-Saline / Transition Zone - 20 20.5m 9.89 / — 1 - 21 SILTY CLAY 1 1 1 — 22 Clayey silt to silty clay • 7.65 1 — 23 1 1 1 6.53 SOIL LOG: CPT K9309 DRILLED: Jan 19,1993 LOCATION: South end of site near No.4Rd. METHOD: CPT DEPTH INTERPRETED STRATIGRAPHY RESISTIVITY DATA 1 2 CLAYEY SILT Sandy or clayey silt to silty clay and clay, trace fine sand. 1 10 I l l l l l 20 30 40 50 60 70 Resistivity (ohm-m) - 3 3.6m -4 5 SILTY SAND Silty sand with some layers of fine sand, 5 7 m some clayey and sandy silt lenses « 19.8 \ \ \ -6 7 8 FINE AND MEDIUM SAND Medium and fine sand, some fine sand lenses / «J 22.9 1 / P 22.4 Fresh-Saline Transition Zone - 9 4 l 2.67 - 10 10.4m i t 1.75 — 11 MEDIUM SAND f 1.68 - 12 ' Medium Sand, uniform I • l 1.47 - 13 I I - 14 i 1.54 - 15 i i i - 16 16.9m • 2.25 -17 18 19 SILTY CLAY Clayey silt to silty clay Possible Saline-Fresh Transition Zone - 20 - 21 - 22 — 23 SOIL LOG: CPT K9310 - DRILLED: Jan 19,1993 LOCATION: North of River Dr., 6 m South of dike METHOD: CPT DEPTH INTERPRETED STRATIGRAPHY RESISTIVITY DATA FILL 1 1 1 1 1 1 1 10 20 30 40 50 60 70 1 CLAYEY SILT Sandy or clayey silt to silty clay and clay, 2.0m trace fine sand. Resistivity (ohm-m) -2 3 4 SILTY SAND Silty sand with some layers of fine sand, some clayey and sandy silt lenses DATA NOT AVAILABLE - 5 - 6 - 7 - 8 9 10 9.0m Fresh-Saline Transition Zone or "Interface" -FINE AND MEDIUM SAND -11 12 Medium and fine sand, uniform, some fine sand lenses, unit thickness unknown MEDIUM SAND 13 Medium Sand, uniform - 14 - 15 - 16 - 17 - 18 - 19 - 20 21 21.0m Saline-Fresh Transition Zone - 22 SILTY CLAY Clayey silt to silty clay 23 SOIL LOG: CPT K9311 LOGGED BY UBC Civ. Eng. DRILLED: 1993 LOCATION: Approx. 210m south of River METHOD: CPT ^Iti INTERPRETED STRATIGRAPHY RESISTIVITY DATA M • 1 CLAYEY SILT 1 10 1 1 20 30 1 1 1 1 40 50 60 70 - 2 Sandy or clayey silt to silty clay and clay, trace fine sand. Resistivity (ohm-m) - 3 3.4m — 4 5 SILTY SAND Silty sand with some layers of fine sand, some clayey and sandy silt lenses \ \ \ 26.8 - 6 * / 31.7 7 8 8.5m / / / • 30.0 — 9 10 FINE AND MEDIUM SAND Medium and fine sand, uniform, some fine 10.5m s a n d l e n s e s 1 1 1 27.4 4 . / — 11 12 MEDIUM SAND Medium Sand, uniform * 2.03 / - > 19.5 Fresh-Saline Transition Zone - 13 - 14 1 0.69 - 15 — 16 17 k 0.69 - 18 P 0.84 — 19 20 • 0.91 - 21 21.5m -22 23 SILTY CLAY Clayey silt to silty clay * 2.9 \ \ \ Vi 11.2 Saline-Fresh Transition Zone SOIL LOG: CPT K9601 DRILLED: Feb. 8,1996 LOCATION: Approx. 3 m North of Pumping Well METHOD: CPT DEPTH (m) INTERPRETED STRATIGRAPHY 186 RTiSISTIVTTY DATA 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 S ill, xlayey. an<Lsandy. -Sand, gravelly (fill) CLAYEY SILT Sandy or clayey silt to silty clay and clay, trace fine sand. 4.05m SILTY SAND Silty sand with some layers of fine sand, some clayey and sandy silt lenses 9.45m FINE AND MEDIUM SAND Medium and fine sand, uniform, some fine sand lenses 14.80m MEDIUM SAND Medium Sand, uniform 21.25m SILTY CLAY Clayey silt to silty clay i — i — i — i — i — i — r 10 20 30 40 50 60 70 Resistivity (ohm-m) 19.3 t 31.6 J / / / ^ 28.1 I I 27.3 4 / / 15.9 ^ 2.24 0.56 0.50 0.75 0.65 Fresh-Saline Transition Zone Saline-Fresh Transition Zone SOIL LOG: CPT K9602 LOCATION: Approx. 3 m South of Pumping Well DRILLED: Feb. 8,1996 METHOD: CPT ' H K I H INTERPRETED STRATIGRAPHY RESISTIVITY DATA (m) - 1 - 2 Sand (fill) 1 1 1 10 20 30 1 1 1 1 40 50 60 70 CLAYEY SILT Sandy or clayey silt to silty clay and clay Resistivity (ohm-m) - 3 - 4 - 5 - 6 4.0m Id S SILTY SAND Silty sand with some layers of fine sand, some clayey and sandy silt lenses >. V . > S \ \ \ * 352 - 7 21.0 - 8 ® = Groundwater Sampling Location *>22.8 - 9 9.6m ® 9.61 m ^ 35.4 - 10 FINE AND MEDIUM SAND 31.0 f l 27.8 * -s 15.7 y s - 11 - 12 Medium and fine sand, some fine sand lenses ® 11.0m Fresh-Saline Transition Zone - 13 f 2.73 - 14 14.8m ® 14.63 m l ' 1.67 1 - 15 - 16 - 17 MEDIUM SAND Medium Sand, uniform ® 16.43 1 l * 1.45 1 1 1 - 18 » 1.36 - 19 1 1 i - 20 1 * 1.51 I - 21 21.2m 1 I - 22 - 23 SILTY CLAY Clayey silt to sDty clay 1 1 i 2.66 — \ \ \ * 8.6 Saline-Fresh Transition Zone SOIL LOG: CPT K9603 DRILLED: May 1996 LOCATION: Approx. 15m N, 20m W of Pumping Well METHOD: CPT DEPTH INTERPRETED STRATIGRAPHY RESISTIVITY DATA 1 FILL 1 1 1 1 1 1 1 10 20 30 40 50 60 70 1 CLAYEY SILT Resistivity (ohm-m) — 2 Sandy or clayey silt to silty clay and clay, trace fine sand. . DATA NOT AVAILABLE 3 4.0m 4 - 5 6 SILTY SAND Silty sand with some layers of fine sand, some clayey and sandy silt lenses - 7 - 8 - 9 - 10 11 12 11.0m Fresh-Saline -FINE AND MEDIUM SAND Transition Zone or "Interface" - 13 Medium and fine sand, uniform, some fine sand lenses, unit thickness unknown - 14 - 15 MEDIUM SAND Medium Sand, uniform - 16 - 17 - 18 - 19 - 20 21 22 23 23.0m SILTY CLAY Clayey silt to silty clay SOIL LOG: CPT K9612 DRILLED: 1996 LOCATION: Approx. 341m south of River METHOD: CPT 189 DEPTH (m) INTERPRETED STRATIGRAPHY RESISTIVITY DATA — 1 — 2 — 3 — 4 5 6 7 — 8 — 9 10 I— 11 12 13 — 14 — 15 — 16 — 17 — 18 — 19 20 r - 2 1 22 I — 2 3 L _FIII CLAYEY SILT Sandy or clayey silt to silty clay and clay 3.5m SILTY SAND Silty sand with some layers of fine sand, some clayey and sandy silt lenses 8.5m m MEDIUM SAND 9.5m SILTY SAND 11.0m FINE AND MEDIUM SAND Medium and fine sand, some fine sand lenses 14.5m MEDIUM SAND Medium sand, uniform »15.0m 19.0m 21.0m SILTY CLAY Clayey silt to silty clay i i i r n r 10 20 30 40 50 60 l I 70 \ f I t I / Fresh-Saline Transition Zone I •I 1 Saline-Fresh Transition Zone SOIL LOG: CPTK9701 LOCTION: Southern End of Site ® Groundwater Sampling Location DRILLED: 1997 METHOD: CPT 190 DEPTH (m) INTERPRETED STRATIGRAPHY RESISTIVITY DATA t i l l 1 I I 10 20 30 40 50 60 70 -1 2 3 •4 5 • 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Fill CLAYEY SILT Sandy or clayey silt to silty clay and clay 4.1m SILTY SAND Silty sand with some layers of fine sand, 5.6m some clayey and sandy silt lenses FINE AND MEDIUM SAND ® 7.40m Medium and fine sand, some fine sand lenses 10.25m MEDIUM SAND Medium sand, uniform ® 11.0m ® 16.0m 18.4m SILTY CLAY Clayey silt to silty clay 1 Fresh-Saline Transition Zone Saline-Fresh Transition Zone SOIL LOG: CPTK9801 LOCTION: 42 m south of Fraser River dike. ® Groundwater Sampling Location DRILLED: 1998 METHOD: CPT DEPTH (m) INTERPRETED STRATIGRAPHY RESISTIVITY DATA — 1 I I I I I I I 10 20 30 40 50 60 70 — 2 — 3 No Stratigraphic Data Available No Resistivity Data Available — 4 — 5 — 6 — 7 — 8 - 9 — 10 — 11 ® 10.65m — 12 — 13 — 14 - 15 — 16 - 17 ® 16.65m — 18 — 19 — 20 — 21 — 22 — 23 SOIL LOG: CPT K9802 DRILLED: 1998 LOCTION: 17.2m South of River Dr. METHOD: CPT ® Groundwater Sampling Location APPENDIX II FIELD WATER L E V E L D A T A A N D CALCULATION OF EQUIVALENT FRESHWATER HEAD 193 Table A2-1: Average Equivalent Freshwater Head Calculat ions Piezometer 2z 4 h Top X Average 1-Aug-96 Equivalent Piezometer Mid-Screen Geodetic Water Screen Salinity 3Density Freshwater Head Number Depth (m) Elev. (m) Elev. (m) Elev. (m) (PPt) (kg/m3) (m) Deep Piezometers (Data Set 1) 101 16.940 1.370 0.005 -15.583 14.2 1010.6 0.170 103 16.990 1.366 0.105 -15.583 14.1 1010.6 0.271 105 17.050 1.355 0.030 -15.583 14.0 1010.5 0.194 107 16.930 1.486 0.015 -15.583 13.4 1010.0 0.171 Deep Piezometers (Data Set 2) 101 16.940 1.370 0.004 -15.583 14.2 1010.6 0.169 103 16.990 1.366 0.105 -15.583 14.1 1010.6 0.271 105 17.050 1.355 0.014 -15.583 14.0 1010.5 0.178 107 16.930 1.486 0.014 -15.583 13.4 1010.0 0.170 Intermediate Piezometers (Data Set 1) 102 11.740 1.371 0.066 -10.386 6.8 1004.9 0.117 104 11.780 1.362 0.064 -10.386 7.7 1005.6 0.123 106 11.680 1.368 0.028 -10.386 11.8 1008.8 0.120 108 11.930 1.486 0.100 -10.386 1.8 1001.0 0.110 Intermediate Piezometers (Data Set 2) 102 11.740 1.371 0.066 -10.386 6.8 1004.9 0.117 104 11.780 1.362 0.064 -10.386 7.7 1005.6 0.123 106 11.680 1.368 0.028 -10.386 11.8 1008.8 0.120 108 11.930 1.486 0.100 -10.386 1.8 1001.0 0.110 5Shallow Piezometers (Data Set 1) 111 4.420 1.286 -0.078 -3.038 0.5 1000.0 -0.078 112 4.270 1.326 -3.038 0.4 1000.0 114 4.380 1.343 -0.066 -3.038 0.3 1000.0 -0.066 5Shallow Piezometers (Data Set 2) 111 4.420 1.286 -0.080 -3.038 0.5 1000.0 -0.080 112 4.270 1.326 -3.038 0.4 1000.0 114 4.380 1.343 -0.068 -3.038 0.3 1000.0 -0.068 1 h0 = average water elevation determined using a tidal filtering of 72 or 25 consecutive hourly water level readings as described in Serfes (1991). 2 z = average mid-screen elevation of the piezometer group. Note z must be the same for each piezometer in a given group in order to compare calculations of equivalent freshwater head to assess groundwater flow direction. 3 Density calculated using the International Equation of State of Seawater (Unesco, 1980) for each well using the salinity noted above, a temperature of 11 C. Density for shallow piezometers 111,112, and 114 is assumed to be 1000 kg/m3 4 Average equivalent freshwater head presented in m with respect to geodetic datum. 5 h0 and h not calculated due to anomalous water level readings for piezometer 112 194 Sample Calculation of Equivalent Freshwater Head The equation for equivalent freshwater head (h) is written as follows. h = —«„ z Po Po 3n where p is the groundwater density at point p [M/L ], p 0 is the density of fresh water [M/L 3], ho is the hydraulic head [m] at point p , and z indicates the vertical elevation of point p . For piezometer 101: p = 1010.6 kg/m3 p 0 = 1000kg/m3 h0 = 0.005 m z = -15.641 m 1010.6 h = 0.005-1000 1010.6-1000 1000 (-15.641) 0.0051 -(-0.166) = 0.171 m APPENDIX III TABLES OF FIELD AND LABORATORY A N A L Y T I C A L RESULTS 196 TABLE A3-1 : Summary of Water Analyses Completed Well Number Field Measurements Laboratory Analyses Temp PH Salinity EC CPT Res TDS Ca++ K+ Mg++ Na+ Sr++ Fe++ Cl- HC03- S04-1 101 X X X X X X X X X X 1 102 X X X X X X X X X X 1 103 X X X X X X X X X X 1 104 X X X X X X X X X X 1 105 X X X X X X X X X X 1 106 X X X X X X X X X X 1 107 X X X X X X X X X X 1 108 X X X X X X X X X X 1 111 X X X X X X X X X X 1 112 X X X X X X X X X X 1 113 X X X X X X X X X 1 114 X X X X X X X X X X 2 West Bay X X X X X X X X X K9301 X K9308 X K9309 X K9310 X K9311 X K9601 X K9602 X K9612 X 1 K9701 X X X X X X X X W1 (all ports) X X X X W2 (all ports) X X X X 1 W3 (all ports) X X X X X 1 K9801 X X X X X X X X X X 1 K9802 X 1 Laboratory Analyses Completed at UBC Environmental Engineering Laboratory 2 Laboratory Analyses Completed at University of Calgary TABLE A3-2: Groundwater Temperature Measurements Well Number Mid-Screen Depth (m) 17-Dec-96 T(°C) 24-Mar-97 T(°C) W1-1 8.08 10.8 11.3 W1-3 10.08 10.4 11.1 W1-6 13.07 10.8 10.9 W1-7 14.08 10.6 11.2 W1-8 15.08 10.5 10.8 W1-9 16.17 10.7 10.8 W1-10 17.08 10.6 10.7 W1-11 18.08 10.5 10.8 W1-12 19.08 10.3 10.8 W1-13 20.06 10.3 10.8 W1-14 21.07 10.4 10.5 W1-15 22.03 10.1 10.7 W2-1 8.04 10.4 10.7 W2-3 10.05 10.6 10.7 W2-10 17.09 10.6 10.8 W2-11 18.09 10.5 11.0 W2-13 20.09 10.5 10.8 W2-14 21.09 10.5 11.0 W3-1 7.50 10.1 10.7 W3-2 8.50 10.5 10.3 W3-3 9.51 10.6 10.3 W3-4 10.53 11.0 10.5 W3-5 11.52 11.0 10.8 W3-6 12.52 11.0 10.6 W3-7 13.50 11.0 10.6 W3-8 14.50 11.0 10.9 W3-9 15.61 10.8 10.8 W3-10 16.50 10.8 10.8 W3-11 17.50 10.8 10.8 W3-12 18.49 10.8 10.8 W3-13 19.49 10.8 10.8 W3-14 20.49 10.8 10.6 W3-15 21.47 11.1 10.8 Temperature results expressed as degrees celcius TABLE A3-3: Groundwater Chemistry - Field pH Measurements Well Number Depth (m) 1-May-96 PH 30-JUI-96 pH W1-1 8.08 7.08 6.58 W1-3 10.08 6.92 6.56 W1-6 13.07 7.03 6.68 W1-7 14.08 7.15 6.51 W1-8 15.08 7.02 6.54 W1-9 16.17 6.93 6.58 W1-10 17.08 6.91 6.63 W1-11 18.08 6.99 6.60 W1-12 19.08 6.92 6.54 W1-13 20.06 7.00 6.60 W1-14 21.07 7.00 6.64 W1-15 22.03 6.97 6.52 W2-1 8.04 - 6.30 W2-3 10.05 - 6.30 W2-10 17.09 - 6.73 W2-11 18.09 - 6.42 W2-13 20.09 - 6.33 W2-14 21.09 - 6.65 W3-1 7.50 7.07 6.75 W3-2 8.50 7.07 6.70 W3-3 9.51 7.51 6.66 W3-4 10.53 6.95 6.36 W3-5 11.52 6.75 6.29 W3-6 12.52 6.70 6.50 W3-7 13.50 - 6.63 W3-8 14.50 6.85 6.68 W3-9 15.61 6.86 6.71 W3-10 16.50 6.81 6.91 W3-11 17.50 6.75 6.96 W3-12 18.49 6.81 7.09 W3-13 19.49 7.04 7.06 W3-14 20.49 7.36 7.70 W3-15 21.47 8.42 8.00 pH results expressed as pH units - = not measured TABLE A3-4: Groundwater Chemistry - Field Electrical Conduct iv i ty (EC) Measurements Well Number Depth (m) 30-JUI-96 EC (mS) 1-Oct-96 EC(mS) 17-Dec-96 EC(mS) 24-Mar-97 EC (mS) Estimated Yearly Average EC (mS) W1-1 8.08 2.22 3.26 1.55 1.26 2.07 W1-3 10.08 3.77 5.40 3.30 2.56 3.76 W1-6 13.07 19.31 19.27 19.16 18.07 18.95 W1-7 14.08 19.22 19.31 19.31 19.00 19.21 W1-8 15.08 19.94 20.10 20.20 19.20 19.86 W1-9 16.17 22.00 22.00 21.90 20.50 21.60 W1-10 17.08 21.50 22.40 22.10 22.10 22.03 W1-11 18.08 21.90 22.60 22.30 22.10 22.23 W1-12 19.08 22.10 22.70 22.30 22.00 22.28 W1-13 20.06 21.90 22.50 22.40 22.20 22.25 W1-14 21.07 22.20 22.40 22.30 22.20 22.28 W1-15 22.03 22.50 22.30 22.30 22.10 22.30 W2-1 8.04 1.05 1.04 1.02 0.41 0.88 W2-3 10.05 1.38 1.34 1.34 0.51 1.14 W2-10 17.09 21.50 21.10 21.20 20.80 21.15 W2-11 18.09 21.60 21.20 21.40 20.90 21.28 W2-13 20.09 22.80 23.40 23.60 23.10 23.23 W2-14 21.09 21.10 23.20 22.60 22.10 22.25 W3-1 7.50 0.98 0.94 0.42 0.21 0.64 W3-2 8.50 0.95 0.93 0.61 0.25 0.69 W3-3 9.51 1.00 0.95 0.73 0.19 0.72 W3-4 10.53 1.26 1.22 1.15 0.77 1.10 W3-5 11.52 6.96 7.81 6.51 5.93 6.80 W3-6 12.52 17.79 18.23 18.35 17.79 18.04 W3-7 13.50 21.10 20.40 20.70 20.30 20.63 W3-8 14.50 20.90 20.90 21.00 20.50 20.83 W3-9 15.61 21.10 21.10 21.40 21.10 21.18 W3-10 16.50 20.70 20.90 21.10 20.80 20.88 W3-11 17.50 21.90 23.10 21.40 21.10 21.88 W3-12 18.49 18.82 18.72 17.61 16.76 17.98 W3-13 19.49 15.53 13.79 11.62 11.98 13.23 W3-14 20.49 3.62 3.53 3.52 3.54 3.55 W3-15 21.47 2.90 2.88 2.87 2.87 2.88 Electrical Conductivity results expressed in mS (mS/m) Measurements taken using an LF 320 Handheld Conductivity Meter manufactured by Wissenschaftlich-Technische Werkstatten. TABLE A3-5: Groundwater Chemistry - Field Sal ini ty Measurements Well Number Depth (m) 30-Jul-96 Salinity (ppt) 1-Oct-96 Salinity (ppt) 17-Dec-96 Salinity (ppt) 24-Mar-97 Salinity (ppt) Estimated Yearly Average Salinity (ppt) W1-1 8.08 1.1 1.8 0.6 0.5 1.0 W1-3 10.08 2.1 3.2 1.8 1.3 2.1 W1-6 13.07 12.8 12.7 12.6 11.8 12.5 W1-7 14.08 12.8 12.7 12.7 12.4 12.7 W1-8 15.08 13.3 13.2 13.3 12.6 13.1 W1-9 16.17 14.8 14.6 14.5 13.5 14.4 W1-10 17.08 14.4 14.8 14.6 14.5 14.6 W1-11 18.08 14.7 15.0 14.8 14.6 14.8 W1-12 19.08 14.8 15.1 14.7 14.5 14.8 W1-13 20.06 14.8 15.0 14.9 14.7 14.9 W1-14 21.07 15.0 14.9 14.8 14.8 14.9 W1-15 22.03 15.1 14.8 14.8 14.6 14.8 W2-1 8.04 0.4 0.3 0.3 0.0 0.3 W2-3 10.05 0.6 0.5 0.5 0.0 0.4 W2-10 17.09 14.3 14.0 14.0 13.7 14.0 W2-11 18.09 14.4 14.0 14.1 13.7 14.1 W2-13 20.09 15.6 15.6 15.7 15.3 15.6 W2-14 21.09 14.3 15.4 15.0 14.6 14.8 W3-1 7.50 0.3 0.3 0.0 0.0 0.2 W3-2 8.50 0.3 0.3 0.1 0.0 0.2 W3-3 9.51 0.3 0.3 0.1 0.0 0.2 W3-4 10.53 0.5 0.5 0.4 0.2 0.4 W3-5 11.52 4.2 4.8 3.9 3.5 4.1 W3-6 12.52 11.7 11.9 11.9 11.6 11.8 W3-7 13.50 13.3 13.5 13.6 13.4 13.5 W3-8 14.50 13.9 13.9 13.8 13.5 13.8 W3-9 15.61 14.0 14.0 14.1 14.0 14.0 W3-10 16.50 13.7 13.9 14.0 13.7 13.8 W3-11 17.50 14.6 15.4 14.1 13.9 14.5 W3-12 18.69 12.4 12.3 11.5 10.9 11.8 W3-13 19.49 10.1 8.8 7.3 7.6 8.5 W3-14 20.49 2.0 2.0 2.0 2.0 2.0 W3-15 21.47 1.6 1.6 1.5 1.5 1.6 Salinity results expressed as parts per thousand (ppt) Measurements taken using an LF 320 Handheld Conductivity Meter manufactured by Wissenschaftlich-Technische Werkstatten. 201 TABLE A3-6: Groundwater Chemistry - Laboratory Analyses Well Number Mid-Screen Depth (m) TDS Ca++ K+ • Mg++ Na+ Sr++ Fe++ Br- Cl- HC03- S04 -Charge Error (%) 101 16.94 17348 294 84.8 340 4200 5.3 - - 8000 30 1140 4.7 102 11.74 9028 223 55.5 190 1610 4.2 - - 3880 0 670 11.4 103 17.22 17348 306 89.5 360 4050 5.3 - - 7880 28 1150 5.0 104 11.78 9604 230 61.5 190 1880 4.5 - - 4230 0 700 9.6 105 17.05 17428 300 95.4 360 4100 5.4 - - 7730 16 1140 3.6 106 11.68 14568 306 93.3 250 3100 5.8 - - 6650 16 910 8.9 107 16.93 16708 317 96.2 320 3780 6.1 - - 7570 85 890 5.5 108 11.93 2968 76 22.1 80 399 1.3 - - 903 30 350 8.1 111 4.47 1060 24 7.2 58 106 0.3 - - 98 203 270 4.2 112 4.27 948 19 4.8 43 95 0.3 - - 106 183 210 8.5 113 1.79 1408 93 9.8 54 148 1.0 - - 22 268 570 3.4 114 4.39 852 17 4.5 52 75 0.3 - - 37 343 140 5.9 West Bay Z11 4.20 - 48 8.0 108 118 0.6 28.00 4 158 174 334 7.6 West BayZIO 8.00 - 44 2.8 50 101 0.3 77.70 0 101 181 215 2.0 West Bay Z9 11.00 - 66 5.4 61 124 0.5 96.60 2 206 108 321 1.5 West Bay Z8 14.00 - 439 70.4 423 3179 5.9 357.00 52 5863 135 816 3.2 West Bay Z7 16.50 - 423 71.4 454 4111 5.8 150.00 22 7415 108 809 2.4 West Bay Z6 19.50 - 401 68.2 514 3845 5.1 65.00 55 7121 77 979 1.9 West Bay Z5 22.00 - 159 69.9 424 3406 4.5 0.05 56 6102 212 775 0.3 West Bay Z3 28.00 - 9 18.2 10 475 0.1 0.84 3 250 955 15 1.3 West Bay Z1 35.50 - 9 14.7 8 459 0.1 0.92 3 200 1047 43 5.0 K9701 15.00 - 45 22.5 67 155 0.6 - - 300 265 210 6.5 K9701 19.00 - 325 90.0 505 2510 4.2 - - 5630 442 530 2.3 W3-1 7.50 - - - - - - - - 50 - - -W3-2 8.50 - - - - - - - - 45 - - -W3-3 9.51 - - - - - - - - 46 - - -W3-4 10.53 - - - - - - - - 93 - - -W3-5 11.52 - - - - - - - - 2330 - - -W3-6 12.52 - - - - - - - - 6740 - - -W3-7 13.50 - - - - - - - - 7660 - - -W3-8 14.50 - - - - - - - - 7800 - - -W3-9 15.61 - - - - - - - - 7940 - - -W3-10 16.50 - - - - - - - - 7920 - - -W3-11 17.50 - - - - - - - - 8365 - - -W3-12 18.49 - - - - - - - - 7070 - - -W3-13 19.49 - - - - - - - - 5680 - - -W3-14 20.49 - - - - - - - - 960 - - -W3-15 21.47 - - - - - - - - 706 - - -K9801 7.40 1590 27 31.0 55 323 0.6 - - 370 315 308 3.0 K9801 11.00 15600 205 189.0 508 3550 3.3 - - 8090 98 850 7.9 K9801 16.00 19300 283 223.0 631 4680 4.3 - - 9330 177 1350 3.3 K9802 10.60 - - - - - - - - 4160 - - -K9802 16.60 - - - - - - - - 9480 - - -Results Expressed as mg/L - = not analyzed 202 TABLE A3-7: Surface Water and Groundwater Samples from Simpson and Hutcheon (1995). Well Number Mid-Screen Depth (m) TDS Ca++ K+ Mg++ Na+ Sr++ Fe++ Br- Cl- HC03- S 0 4 -Charge Error (%) Groundwater Samples A 2.87 - 37 9.8 53 13 2.0 9.85 nd 27 261 35 9.7 B 19.20 - 295 156.0 532 4640 4.4 3.69 14 8797 136 444 0.9 E 36.27 - 9 22.0 8 330 0.5 0.27 1 322 538 7 5.9 P 31.09 - 173 240.0 945 8050 8.5 3.03 20 15313 2101 nd 2.6 Q 44.75 - 197 281.0 948 8655 8.3 1.66 22 16428 1993 nd 2.5 1Surface Water Samples rwV9 0 - 18 5.2 5 18 0.0 0.20 nd 34 63 10 0.51 SW 0 - 296 295.0 900 7630 6.9 0.20 58 12889 277 675 5.69 Marine 0 - 404 394.0 1146 9908 8.2 0.02 69 18000 130 2455 0.49 Results Expressed as mg/L nd=concentration below laboratory detection limit - = not analyzed 1 Results reported in Simpson and Hutcheon (1995): rwV9 = fresh river water, mouth of Fraser R. Middle Arm SW = seawater, Georgia Strait near Sea Island Marine = Marine water, west coast of Vancouver Island Samples A, B, and E are from a well at the eastern end of Sea Island. Samples P and Q are from a well at the western side of Lulu Island. 203 APPENDIX IV EFFECTS OF GRID SPACING ON MODEL RESULTS APPENDIX IV - Effects of Grid Spacing on Model Results 204 The following describes the simulations completed to investigate the potential effects of grid spacing on model results. A boundary value problem was developed consisting of a model domain that was simplified from that applied to the field site. The model domain was 50 m long and 10 m high, and was comprised of a homogeneous aquifer exhibiting the same hydrogeological properties as the Medium Sand unit at the field site (Figure A4-1). Boundaries for flow and transport were initially estimated to obtain a preliminary indication of the effects of grid spacing on model results. Then, following completion of Simulation 1 presented in this thesis, salinity (density) at the right hand side of the model was adjusted to reflect that calibrated for Simulation 1. 10m-dc/&z=o oh/az=o c=0 h=10.14 0-lg»4.4xlff" <x,=1.0 k=4.4xl0'2 oc=0.01 Ibc/lx-0 z=7 »h=10m p=1014.3kg/m3 *h=10.148m I 0 ac/az=Odh/az=o 50m Figure A4-1: Simplified model domain for investigation of grid spacing on model resuits. Five simulations were completed as indicated on Table A4-1. Simulations A, B, and C were completed with Px = 2 and Pz = 10, 50, and 100. The simulation results for the variation of Pz are illustrated on Figures A4-2 and A4-3. Variations in Pz did not significantly affect the position (elevation) or thickness of the transition zone up to Pz = 100. The effects of overshooting, however, were observed for Pz greater than approximately 50. Variation in Pz also had some effect on the position of the saltwater wedge toe as the inland extent of the wedge toe was slightly further for simulations with lower Pz numbers. Based on these simulations, it was determined that a Pz = 50 would provide accurate model results while allowing for a slightly coarser grid (compared to using Pz = 2). 205 Simulations B, D, and E were completed with Pz = 50 and Px = 2, 3, and 4. The results for variation of horizontal grid spacing are illustrated on Figures A4-4 and A4-5. In the x-direction, numerical dispersion was evident for Peclet numbers greater than 2 as the thickness of the transition zone generally increased with increasing Px. Based on these simulations, the P x criterion was maintained at 2 for the modelling of saltwater intrusion at the field site. Table A4-1 : Summary o f Simulat ions fo r Invest igat ion o f Grid Spacing on Model Results Simulat ion Name a„ a z delta x delta z Px Pz A 1 0.01 2 0.1 2 10 B 1 0.01 2 0.5 2 50 C 1 0.01 2 1 2 100 D 1 0.01 3 0.5 3 50 E 1 0.01 4 0.5 4 50 206 Figure A4-2: Comparison of modelled concentration contours for variations in vertical Peclet number. 207 10 15 Salinity (ppt) 10 15 Salinity (ppt) Figure A4-3: Comparison of model results in vertical profile for variations in vertical Peclet number (Px=2, Pz varied as indicated). 208 0 10 20 30 40 50 X ( m ) 0 10 20 30 40 50 X ( m ) 0 10 20 30 40 50 X ( m ) Figure A4-4: Comparison of modelled concentration contours for variations in horizontal Peclet number. 209 E. •o c s 2 (9 I 5 » CD a a a 10 \ x=15m B (Px=2) —A—D (Px=3) — • — E (Px=4) ~ ~ i i *^ "Tfr i 10 15 Salinity (ppt) 20 25 x=35 m B (Px=2) ^— D (Px=3) • — E (Px=4) T J C 3 O w i 5 O ca £ •s. 2 10 4 10 15 Salinity (ppt) 20 E. C 3 s a. « Q 10 !• x=2Sm B (Px=2) A D (Px=3) — • — E (Px=4) •"TV 0 » 10 15 Salinity (ppt) 20 25 10 15 Salinity (ppt) Figure A4-5: Comparison of model results in vertical profile for variations in horizontal Peclet number (Pz=50, Px varied as indicated). 

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.831.1-0053150/manifest

Comment

Related Items