Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Trace metals in urban stormwater runoff and their management Li, Tong 2007

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2007-0483.pdf [ 10.83MB ]
Metadata
JSON: 831-1.0100954.json
JSON-LD: 831-1.0100954-ld.json
RDF/XML (Pretty): 831-1.0100954-rdf.xml
RDF/JSON: 831-1.0100954-rdf.json
Turtle: 831-1.0100954-turtle.txt
N-Triples: 831-1.0100954-rdf-ntriples.txt
Original Record: 831-1.0100954-source.json
Full Text
831-1.0100954-fulltext.txt
Citation
831-1.0100954.ris

Full Text

T R A C E M E T A L S IN U R B A N STORMWATER RUNOFF A N D THEIR M A N A G E M E N T by Tong L i B.Eng., Harbin Institute of Technology, 2003 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF APPLIED SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES (Civil Engineering) THE UNIVERSITY OF BRITISH C O L U M B I A April 2007 © Tong L i , 2007 Abstract In the Greater Vancouver Regional District (GVRD), non-point source pollution from an urban watershed and a diesel bus loop was assessed in terms of trace metal contamination in the stormwater runoff. In the Brunette River watershed study, Northwest Hydraulic Consultants (NHC) collected streambed sediment and suspended sediment from selected streams during 7 storm events over 2003. From 1993 to 2003, the major stormwater contamination happened in the most industrialized Still Creek. The streambed Cu, Mn, Fe, and Zn concentration increased by 1.5, 1.7, 1.9, and 1.1 times, respectively. And the suspended Cu, Mn, and Zn increased by a factor of 2.1, 4.2, and 1.5, respectively. The streambed sediment exceeded probable effect level in Still Creek and Stoney Creek to varying degrees with Cu and Zn. The land use is considered to be the origins of these toxicants. Statistically, the magnitude of suspended metal concentration in ug/l is negatively correlated with the drainage areas. While the — concentrations in mg/kg, especially for metal Cu and Zn, showed strongly and positively correlation with the traffic density. Positive correlation existed between the suspended metal loading (kg/yr) and the imperviousness and the catchment area. No apparent trend was observed in terms of export coefficient (g/ha/yr) and land use. 1062 tons of sediments were trapped by Burnaby Lake in 2003. This sediment overloading problem causes serious metal contamination in the lake. Stormwater runoff quality was monitored in 15 storm events from October 2004 to June 2005 in the diesel bus loop in the University of British Columbia. The dissolved Cu and Zn Event Mean Concentration (EMC) exceeded the EPA discharge criteria in 2 and 4 events each, which occurred in the dry season. Diesel bus traffic contributes' a large portion of Cu, Fe, Zn contamination since the average bus loop trace metal levels were much higher than the G V R D urban levels. The runoff trace metal concentrations are strongly related to the antecedent dry period, and are weakly related to the traffic density and the rainfall intensity. From the catch basin filter evaluation, high removal efficiencies on suspended metal/solids were achieved with low particulate loading in the filter chamber. The filter performed well for the dissolved metal removal before the non-reversible saturation was reached. Each kilogram of filter media has an absorption capacity of 52 gram oil and grease, 20 milligram Mn, and 16 milligram Zn. ii Table of Contents Abstract ii Table of Contents iii List of Tables v List of Figures x Acknowledgements : xvii 1 Introduction 1 2 Literature Review 5 2.1 Urban Stormwater Runoff Pollutants and Their Chemistry 5 2.2 Source of Heavy Metal Contamination in the Urban Stormwater Runoff 9 2.2.1 Heavy Metal in the Atmospheric Particulates 9 2.2.2 Heavy Metals in Relation to Land Use Activity 10 2.3 Stormwater Characterization 13 2.4 Stormwater Pollution Control and Management 14 2.4.1 Structural Stormwater Best Management Practice 15 2.4.2 Non Structural Stormwater Best Management Practice 16 2.5 Brunette River Watershed 17 2.5.1 Historical Water/Sediment/Biota Studies on the Brunette River Watershed 17 2.5.2 Change in Land Use Over Time 18 2.5.3 Summary of Spatial Difference in the Water Quality 20 2.5.4 Relation of Land Use to Benthic Invertebrate Community Structure 23 3 Materials and Methods 24 3.1 Experimental Overview -. 24 3.2 Field Methods 24 3.2.1 Brunette River Watershed 24 3.2.2 U B C Bus Loop 26 3.3 Metal Detection Techniques 35 3.4 Quality Control (QA) and Quality Assurance (QC) 36 3.4.1 Q A and Q C in Sample Collection 36 3.4.2 Q A and Q C in Sample Analysis 37 3.4.3 Traffic Analysis 38 iii 3.5 Statistical Analyses 38 4 Results and Discussion 40 4.1 Trace Metals in Brunette River Watershed Stormwater Runoff 40 4.1.1 Land Use Change from 1993 to 2001 40 4.1.2 Suspended Solids and Their Relation to Land Use 42 4.1.3 Suspended Metals and Their Relation to Land Use 46 4.1.4 Sediment Quality Characterization 66 4.1.5 Temporal Changes and the Severity of the Streambed Sediment Contamination 76 4.2 Trace Metals in the Stormwater Runoff from the U B C Diesel Bus Loop 81 4.2.1 Suspended Solids Concentration and Spatial Variation 81 4.2.2 Trace Metal Concentration '. 83 4.2.3 Stormwater Quality Characterization : 95 4.3 Stormwater Runoff Management by Catch Basin Filter 117 4.3.1 Contaminants Removal Efficiency 117 4.3.2 Catch Basin Filter Media Characterization 119 4.3.3 Catch Basin Filter Trapped Particular Material Analyses 127 5 Conclusions and Recommendations •. 131 5.1 Conclusions 131 5.1.1 Extent of Trace Metal Contamination in Aquatic Sediments in the Brunette River Watershed 131 5.1.2 Sediment Budget in the Brunette River Watershed 132 5.1.3 Relationships between Land Use and Trace Metal Contamination 132 5.1.4 Trace Metal Characterization in the U B C Bus Loop Stormwater Runoff. 134 5.1.5 Effectiveness of Catch Basin Filters in Contaminants Removal 136 Bibliography 138 Appendix A Summary of Q A / Q C Data 147 Appendix B Metal Digestion Techniques 149 Appendix C Particle Size Analysis and Its Associated Trace Metals 151 Appendix D Brunette River Watershed Sediment Data 154 Appendix E U B C Bus Loop Stormwater Data 168 Appendix F Transportation and Land Use Maps : 190 iv L i s t of Tables Table 2.1 Sources of contaminants in the urban stormwater runoff 9 Table 2.2 Typical contaminants in the stormwater runoff from different land uses 11 Table 2.3 Land use and traffic changes in the Brunette River watershed between 1973 and 1993 18 Table 2.4 Total impervious area over 60 years in the Brunette River watershed 19 Table 2.5 Spatial differences in the land use and water quality under base flow condition in local watershed 21 Table 2.6 Spatial difference in the land use and water quality under base flow condition in local watersheds (cont.) 22 Table 3.1 Steam sediments sampling stations 25 Table 3.2 Summary of daily bus density in the U B C bus loop 28 Table 3.3 Features of the sampled storm events 29 Table 3.4 Comparison of digestion techniques 35 Table 3.5 Comparison of trace metal detection levels for different analytical techniques 36 Table 3.6 Measurement of method accuracy using the sediment reference material 38 Table 4.1 Land use activities in the Brunette River watershed as a proportion of total area 40 Table 4.2 Population increase in the sub-basins over 40 years 41 Table 4.3 Spearman rank correlation matrix for traffic density, imperviousness, and population in the Brunette River watershed 42 Table 4.4 Spearman rank correlation for suspended solids concentration, imperviousness, and traffic density in the Brunette River watershed : 44 Table 4.5 Suspended sediment loading and budget for the Brunette sub-basins in 2003 45 v Table 4.6 Spearman rank correlation matrix for suspended solids loading, imperviousness, and traffic density in the Brunette River watershed 46 Table 4.7 Spearman rank correlation matrix for catchment area and suspended metals (ug/1) from the Brunette River watershed 51 Table 4.8 Spearman rank correlation matrix for imperviousness and suspended metals (ug/1) from the Brunette River watershed 51 Table 4.9 Spearman rank correlation matrix for traffic density and suspended metals (ug/1) from the Brunette River watershed 52 Table 4.10 Numbers of sampled storm events in dry/wet seasons in 1994/95 and 2003 54 Table 4.11 Spearman rank correlation matrix for catchment area and suspended metals (mg/kg) from the Brunette River watershed 57 Table 4.12 Spearman rank correlation matrix for imperviousness and suspended metals (mg/kg) from the Brunette River watershed 57 Table 4.13 Spearman rank correlation matrix for traffic density and suspended metals (mg/kg) from the Brunette River watershed 57 Table 4.14 Spearman rank correlation matrix for imperviousness and suspended metal annual loading (kg/yr) from different drainage areas of the Brunette River watershed 60 Table 4.15 Spearman rank correlation matrix for catchment area and suspended metal annual loading (kg/yr) from different drainage areas of the Brunette River watershed 61 Table 4.16 Spearman rank correlation matrix for traffic density and suspended metal annual loading (kg/yr) from different drainage areas of the Brunette River watershed 61 Table 4.17 Spearman rank correlation matrix for imperviousness and suspended metal export coefficients (g/ha/yr) from the Brunette River watershed 64 Table 4.18 Spearman rank correlation matrix for traffic density and suspended metal export coefficients (g/ha/yr) from the Brunette River watershed 65 Table 4.19 Mean stream suspended metal export coefficients from the Brunette River watershed in 1994/95 and 2003 65 vi Table 4.20 Emission concentrations and emission rates of the metals contents in the diesel engine exhaust 89 Table 4.21 Spearman rank correlation matrix for conductivity and dissolved metals from catch basin 1 (n=8) 104 Table 4.22 Spearman rank correlation matrix for conductivity and dissolved metals from catch basin2(n=8) 104 Table 4.23 Spearman rank correlation matrix for conductivity and dissolved metals from catch basin 3 (n= 12) 105 Table 4.24 Spearman rank correlation matrix for suspended solids and total metals from catch basin 1 (n=8) 106 Table 4.25 Spearman rank correlation matrix for suspended solids and total metals from catch basin2(n=8) 106 Table 4.26 Spearman rank correlation matrix for suspended solids and total metals from catch basin 3 (n= 12) 107 Table 4.27 Spearman rank correlation matrix for rainfall and total metal E M C in stormwater runoff from the U B C bus loop 108 Table 4.28 Spearman rank correlation matrix for antecedent dry days and total metal E M C in stormwater runoff from the U B C bus loop 110 Table 4.29 Spearman rank correlation matrix for traffic density and total metal E M C in stormwater runoff from the U B C bus loop 111 Table 4.30 Student t-test for the total metals from the U B C bus loop catch basins 115 Table 4.31 Summary of stormwater runoff quality from the U B C diesel bus loop and discharge criteria 115 Table 4.32 Summary of catch basin filter contaminants removal efficiencies 118 Table 4.33 Determination of the filter maintenance period from laboratory adsorption studies 124 Table 4.34 Perlite adsorbed trace metal concentration 125 Table 4.35 Catch basin filter trapped particulate material wet/dry weight and their ratio 128 vn Table 5.1 Spearman correlation between suspended metal and 2003 land use in the Brunette River watershed 133 Table 5.2 Spearman correlation between stream bed metals and 2003 land use in the Brunette River watershed 134 Table 5.3 Spearman correlation between land use and natural factors, and total metal E M C in the stormwater runoff from UBC bus loop 136 Table A-1 Recovery test 147 Table A-2 Duplicate test 148 Table A-3 Spike recovery 148 Table B- l t-test for metal concentration difference between the two digestion methods 149 Table B-2 Precision difference for nitric and aqua regia digestion techniques 150 Table C - l Stream sediment trace metal average concentration in 1973, 1993, and 2003 151 Table C-2 Catch basin filter trapped particulate size associated trace metal concentration during first filter clean out 152 Table C-3 Catch basin filter trapped particulate size associated trace metal concentration during second filter cleanout 152 Table C-4 Perltie particle size and weight distribution 153 Table C-5 Perlite particle size associated trace metal concentration 153 Table D - l Suspended sediment sampling detail for the Brunette River watershed by N H C 154 Table D-2 Suspended sediment trace metal concentration from Brunette River 156 Table D-3 Suspended sediment trace metal concentration from Eagle Creek 157 Table D-4 Suspended sediment trace metal concentration from Still Creek 159 Table D-5 Suspended sediment trace metal concentration from Ramsay Creek 161 Table D-6 Suspended sediment trace metal concentration from Stoney Creek 163 Table D-7 Suspended sediment trace metal concentration from Silver Creek 165 viii Table D-8 Bedload sediment trace metal concentration 167 Table E - l Stormwater quality data from three drains in the bus loop 168 Table E-2 Used perlite particle size distribution and trace metal concentration 186 Table E-3 New perlite particle size distribution 186 Table E-4 Single metal adsorption capacity test data 187 Table E-5 Mn and Zn co-existence adsorption capacity test data 188 Table E-6 Ca and Mg interference test data 189 List of Figures Figure 2.1 Automobiles registration in Burnaby from 1994 to 2004 19 Figure 3.1 Stormwater runoff sampling site selection 27 Figure 3.2 Instrument set-up for metal adsorption test 33 Figure 3.3 Components of a box-whisker plot 39 Figure 4.1 Traffic volumes in the sub-basins of Brunette River watershed in 2005 42 Figure 4.2 Suspended solids concentration in stormwater runoff from Brunette sub-basins 43 Figure 4.3 Locations of suspended sediment sampling stations in the Brunette River watershed 47 Figure 4.4 Suspended Cu (jj.g/1) in stormwater runoff from the Brunette sub-basins 48 Figure 4.5 Suspended Mn (ug/1) in stormwater runoff from the Brunette sub-basins 49 Figure 4.6 Suspended Fe (pg/1) in stormwater runoff from the Brunette sub-basins 49 Figure 4.7 Suspended Zn (ug/1) in stormwater runoff from the Brunette sub-basins 50 Figure 4.8 Temporal changes of suspended Cu concentration (ug/1) in Eagle and Still creeks ...52 Figure 4.9 Temporal changes of suspended Mn concentration (ug/1) in Eagle and Still creeks ..53 Figure 4.10 Temporal changes of suspended Zn concentration (ug/1) in Eagle and Still creeks .53 Figure 4.11 Suspended Cu (mg/kg) in the stormwater runoff from the Brunette sub-basins 55 Figure 4.12 Suspended Mn (mg/kg) in the stormwater runoff from the Brunette sub-basins 55 Figure 4.13 Suspended Fe (mg/kg) in the stormwater runoff from the Brunette sub-basins 56 Figure 4.14 Suspended Zn (mg/kg) in the stormwater runoff from the Brunette sub-basins 56 Figure 4.15 Suspended Cu loading (kg/yr) in stormwater runoff from the Brunette sub-basins .58 Figure 4.16 Suspended Mn loading (kg/yr) in stormwater runoff from the Brunette sub-basins 59 Figure 4.17 Suspended Fe loading (kg/yr) in stormwater runoff from the Brunette sub-basins..59 x Figure 4.18 Suspended Zn loading (kg/yr) in stormwater runoff from the Brunette sub-basins .60 Figure 4.19 Suspended Cu EC (g/ha/yr) in stormwater runoff and total impervious area of the Brunette sub-basins 62 Figure 4.20 Suspended Mn E C (g/ha/yr) in stormwater runoff from the Brunette sub-basins ....63 Figure 4.21 Suspended Fe E C (g/ha/yr) in stormwater runoff from the Brunette sub-basins 63 Figure 4.22 Suspended Zn E C (g/ha/yr) in stormwater runoff from the Brunette sub-basins 64 Figure 4.23 Suspended/bed sediment comparison for Cu in the Brunette sub-basins 67 Figure 4.24 Suspended/bed sediment comparison for Mn in the Brunette sub-basins 67 Figure 4.25 Suspended/bed sediment comparison for Fe in the Brunette sub-basins 68 Figure 4.26 Suspended/bed sediment comparison for Zn in the Brunette sub-basins 68 Figure 4.27 Distribution of Cu in the streambed sediment from the Brunette sub-basins 70 Figure 4.28 Distribution of Mn in the streambed sediment from the Brunette sub-basins 70 Figure 4.29 Distribution of Fe in the streambed sediment from the Brunette sub-basins 71 Figure 4.30 Distribution of Zn in the streambed sediment from the Brunette sub-basins 71 Figure 4.31 Suspended and < 63 um bed sediment comparison for Cu in the Brunette sub-basins 72 Figure 4.32 Suspended and < 63pm bed sediment comparison for Mn in the Brunette sub-basins 72 Figure 4.33 Suspended and < 63pm bed sediment comparison for Fe in the Brunette sub-basins 73 Figure 4.34 Suspended and < 63pm bed sediment comparison for Zn in the Brunette sub-basins 73 Figure 4.35 Seasonal variation for suspended Cu in the Brunette sub-basins 74 Figure 4.36 Seasonal variation for suspended Mn in the Brunette sub-basins 75 Figure 4.37 Seasonal variation for suspended Fe in the Brunette sub-basins 75 Figure 4.38 Seasonal variation for suspended Zn in the Brunette sub-basins : 75 xi Figure 4.39 Monthly rainfall in the Brunette River watershed in 2003 76 Figure 4.40 Mean streambed Cu contamination (mg/kg) change from 1973 to 2003 77 Figure 4.41 Mean streambed Mn contamination (mg/kg) change from 1973 to 2003 77 Figure 4.42 Mean streambed Fe contamination (mg/kg) change from 1973 to 2003 78 Figure 4.43 Mean streambed Zn contamination (mg/kg) change from 1973 to 2003 78 Figure 4.44 Comparison of streambed Cu (mg/kg) and criteria effects level in 2003 80 Figure 4.45 Comparison of streambed Zn (mg/kg) and criteria effects level in 2003 80 Figure 4.46 Suspended solids concentration in stormwater runoff from U B C diesel bus loop ...81 Figure 4.47 Campus catchment and sampling locations for the UBC stormwater monitoring program 82 Figure 4.48 Spatial differences in suspended solids concentration in the UBC campus, bus loop and G V R D areas 83 Figure 4.49 Total Cu concentrations in stormwater runoff from the U B C diesel bus loop 84 Figure 4.50 Total Mn concentrations in stormwater runoff from the U B C diesel bus loop 84 Figure 4.51 Total Fe concentrations in stormwater runoff from the U B C diesel bus loop 85 Figure 4.52 Total Zn concentrations in stormwater runoff from the U B C diesel bus loop 85 Figure 4.53 Spatial difference in total Cu concentration in the U B C campus, bus loop and G V R D areas 86 Figure 4.54 Spatial difference in total Mn concentration in the U B C campus, bus loop and G V R D areas 87 Figure 4.55 Spatial difference in total Fe concentration in the U B C campus, bus loop and G V R D areas 88 Figure 4.56 Spatial difference in total Zn concentration in the U B C campus, bus loop and G V R D areas 88 Figure 4.57 Dissolved Cu concentrations in stormwater runoff from the UBC diesel bus loop..90 Figure 4.58 Dissolved Mn concentrations in stormwater runoff from the UBC diesel bus loop 91 xii Figure 4.59 Dissolved Fe concentrations in stormwater runoff from the U B C diesel bus loop ..91 Figure 4.60 Dissolved Zn concentrations in stormwater runoff from the U B C diesel bus loop..92 Figure 4.61 Suspended Cu concentration in stormwater runoff from the U B C diesel bus loop..93 Figure 4.62 Suspended Mn concentration in stormwater runoff from the UBC diesel bus loop 93 Figure 4.63 Suspended Fe concentration in stormwater runoff from the U B C diesel bus loop...94 Figure 4.64 Suspended Zn concentration in stormwater runoff from the U B C diesel bus loop ..94 Figure 4.65 Total and dissolved Cu pollution pattern in stormwater runoff from the U B C diesel bus loop over a rainfall event on Oct 29, 2004 96 Figure 4.66 Total and dissolved Mn pollution pattern in stormwater runoff from the U B C diesel bus loop over a rainfall event on Oct 29, 2004 96 Figure 4.67 Total and dissolved Fe pollution pattern in stormwater runoff from the U B C diesel bus loop over a rainfall event on Oct 29, 2004 97 Figure 4.68 Total and dissolved Zn pollution pattern in stormwater runoff from the U B C diesel bus loop over a rainfall event on Oct 29, 2004 97 Figure 4.69 Turbidity and conductivity change in stormwater runoff from the U B C diesel bus loop on Oct 29, 2004 98 Figure 4.70 Dissolved and suspended Cu distribution in stormwater runoff from catch basins in the U B C diesel bus loop 99 Figure 4.71 Dissolved and suspended Mn distribution in stormwater runoff from catch basins in the U B C diesel bus loop 99 Figure 4.72 Dissolved and suspended Fe distribution in stormwater runoff from catch basins in the UBC diesel bus loop 100 Figure 4.73 Dissolved and suspended Zn distribution in stormwater runoff from catch basins in the UBC diesel bus loop 100 Figure 4.74 Conductivity and dissolved Cu concentration in stormwater runoff from the U B C diesel bus loop 102 xiii Figure 4.75 Conductivity and dissolved Mn concentration in stormwater runoff from the U B C diesel bus loop 102 Figure 4.76 Conductivity and dissolved Fe concentration in stormwater runoff from the U B C diesel bus loop 103 Figure 4.77 Conductivity and dissolved Zn concentration in stormwater runoff from the U B C diesel bus loop 103 Figure 4.78 Relation between rainfall and total metal E M C in stormwater runoff from the U B C diesel bus loop 108 Figure 4.79 Relation between antecedent dry days and total metal E M C in stormwater runoff from the U B C diesel bus loop 109 Figure 4.80 Relation between traffic density and total metal E M C in stormwater runoff from the U B C diesel bus loop I l l Figure 4.81 Spatial difference of suspended solids from catch basins in the U B C diesel bus loop stormwater runoff 112 Figure 4.82 Spatial variation of Cu from three catch basins in the U B C diesel bus loop stormwater runoff 113 Figure 4.83 Spatial variation of Mn from the catch basins in the U B C diesel bus loop stormwater runoff 113 Figure 4.84 Spatial variation of Fe from the catch basins in the U B C diesel bus loop stormwater runoff 114 Figure 4.85 Spatial variation of Zn from the catch basins in the U B C diesel bus loop stormwater runoff 114 Figure 4.86 Clean perlite particle size distribution curve 119 Figure 4.87 Mn adsorption capacity curve of perlite 120 Figure 4.88 Zn adsorption capacity curve of perlite 121 Figure 4.89 Mn and Zn adsorption capacity curve for perlite 122 Figure 4.90 Interference of Ca and Mg on the adsorption capacity for Mn by perlite 123 xiv Figure 4.91 Interference of Ca and Mg on the adsorption capacity for Zn by perlite 123 Figure 4.92 Used perlite particle size distribution curve 126 Figure 4.93 Particulate size distribution curve for catch basin filter trapped solids (March to May 2005) 128 Figure 4.94 Particulate size distribution curve for catch basin filter trapped solids (May to June 2005) 129 Figure 4.95 Filter trapped particle size associated trace metal concentration 130 Figure F- l Morning peak hour traffic volume in the north half of Vancouver 190 Figure F-2 Morning peak hour traffic volume in the south half of Vancouver 191 Figure F-3 Morning peak hour traffic volume in the north half of Burnaby 192 Figure F-4 Morning peak hour traffic volume in the south half of Burnaby 193 Figure F-5 Morning peak hour traffic volume in Coquitlam 194 Figure F-6 Land use map of the Brunette River watershed in 2001 195 xv A c k n o w l e d g e m e n t s This study was conducted as part of the Brunette Basin Watershed Plan by the Greater Vancouver Sewerage and Drainage District Brunette Basin Task Group. I would like to thank Dr. K. Hall (my supervisor) for providing me this research opportunity, a friendly work environment, and continuing encouragement. His assistance in the preparation for this thesis, his support, suggestions and experiences in sediment quality and stormwater monitoring were very helpful in ensuring the success of this study. H. Schreier is thanked for being a reader on this thesis. My thanks to Mr. Craig Nistor from the North Hydraulic Consultants (NHC) for their sediment collection from the Brunette River Watershed. M y sincere appreciation to Dejon from B.C. Waste Water Doctor Inc. for the catch basin filter supply and installation. Similarly, my gratitude to Paula Parkinson, and Susan Harper of the Environmental Engineering lab (Department of Civil Engineering, UBC) for all their assistance in the laboratory analyses. Finally, 1 would like to thank Zhijin Wang for his support and effort in editing this document, and always being there. xvii 1 Introduct ion In the past few decades, concerns have been rising about pollution problems in watersheds. Pollution from an easily identifiable source, like a factory or a sewage treatment plant that discharges waste products into the water system is called point source pollution. Point source pollution has been vastly reduced by regulations due to its constant and predictable nature. In comparison, pollutant washed off the landscape from a spatially widespread source is called Non-Point Source (NPS) pollution. NPS in the urban area occurs when stormwater washes the impervious surface of the land, carries away natural and anthropogenic pollutants, and deposits them into surface water and ground water. As the name implies, NPS is much harder to predict, identify, isolate, and control than point-source pollution. In the process of rapid urbanization, the land development replaces forests, fields, and meadows with impervious surfaces such as roofs, parking lots and roads. NPS has become the primary source of water pollution in North America. Take a parking lot for example. During the storm event, the rain falling on the highly impervious surface is converted to stormwater runoff. The runoff carries oil and grease from engines, heavy metals from tires, shingles, paints and metal surfaces of the vehicles. Once the increasing runoff overloads the natural drainage system, it is discharged to downstream waters without treatment. On a broader basis, changes of land use can have an adverse impact on individual streams, or on the entire watershed in a cumulative sense. Brunette River watershed, as a typical urban watershed, has been used for case studies on NPS pollution for more than 3 decades. Brunette River watershed is located primarily within the municipality of Burnaby, British Columbia. It also drains parts of the neighboring municipalities of East Vancouver, Port Moody, Coquitlam and New Westminster into the Fraser River. The watershed, with an area of 56 square kilometers, experiences a maritime climate, with a wet autumn-winter season and a comparatively dry spring-summer season. Approximately 75% of annual precipitation occurs between October and March. The hydrology of the watershed is about 80% developed with just 20% remaining as green space (GVRD, 2001). By year 2000, the total impervious area had increased to nearly 50% due to the rapid industrial and residential area development, road networks expansion, and population boom. Reduced land permeability resulted in high peak wet weather to dry weather flow ratios. It is estimated that 75% to 80% of annual precipitation is conveyed by the channel network through the Brunette River watershed, while the remainder is lost to evaporation or groundwater recharge (NHC, 2004). The flashy high flow creates problems such as frequent minor 1 flooding, bank erosion and instability, and increased sediment load in the stream. Low flows in the dry season, which results in high water temperatures and low dissolved oxygen levels in some tributaries of the system, significantly impeded fisheries resources. (GVRD, 2001) Burnaby Lake, located in the central part of Brunette River watershed, is suffering from sedimentation overloading at a rate of 3,300 cubic meters per year from upstream development and bank erosion. The sediment reduces the depth of the lake, chokes out plant species and may be a hazardous to wildlife habitat ( G V R D , 2001). Extensive studies have verified the existence of contaminants and sediments in the water and biota of Brunette River watershed. Due to a wide variety of human activities, contaminants such as trace metals, organic contaminants, fecal coliforms, and suspended solids have become the major concerns in the watershed ( G V R D , 2001). By comparing the pollutant loading from the high storm flow and the base flow, Macdonald et al. (1997) concluded that the majority of pollutants to the watershed are contributed by stormwater runoff. NPS pollution severely impacts the aquatic organisms in the Brunette River watershed by contributing elevated levels of heavy metals, petroleum hydrocarbons, nutrients, and sediments into the aquatic systems. Some of these organic contaminants associated with urban stormwater runoff are carcinogenic and many accumulate in plant and animal tissues. These common organic contaminants include oil and grease, aliphatic and aromatic hydrocarbons, pesticides, anti-sapstain chemicals, plasticizers and polychlorinated biphenyls (PCBs) (Cole et al., 1984; Kobriger and Geinepolos, 1984; Swain and Walton, 1991; Lawson et al., 1985). To mitigate the effects of distorted hydrographs and NPS pollutants from transportation, the G V R D has planned a watershed scale implementation of stormwater Best Management Practices (BMPs), considering that prevention of contamination is the most effective way of resource protection. The completed stormwater B M P consists of structural and non-structural B M P (GVRD, 2001). Structural B M P includes: • Infiltration system • Retention system • Vegetated system • Filtration system • Constructed wetland system • Minimizing impervious surface 2 Non- structural B M P includes: • Source control • Maintenance control • Education • Recycling The stormwater BMPs can reduce the volume of runoff discharged to receiving streams, reduce the pollutant sources and reduce the transfer of urban pollutants to runoff by minimizing directly connected impervious surfaces, providing on-site storage and infiltration and implementing stream buffers, which all help to prevent further degradation and can result in improvements to the streams in the watershed. A few sediment transport studies have been completed in the Brunette River Watershed in the past. In 1997, the Brunette Basin Task Group developed Brunette Basin Watershed Plan, partly to study the existing sedimentation conditions and develop corresponding sediment management options in the Brunette Basin ( G V R D , 2001). The monitoring of suspended sediment and bed load transportation was carried out by N H C during 2003. The studied streams cover Still Creek, Eagle Creek, Ramsay Creek (also known as Robert Burnaby Creek), Stoney Creek, and Brunette River. Suspended sediment samples were collected manually during storm events on January 30, February 20, March 12-13, April 07, and October 16, 2003. There were no summer storm events of sufficient magnitude or duration to be sampled. N H C collected the turbidity data, analyzed suspended sediment concentrations, and calculated the suspended sediment loads (NHC, 2004). After analysis, the dried sediments were provided to Dr. Ken Hall for sediment quality analysis. This became part of my Master thesis project. The method and results of the sediment quality analysis will be discussed in detail in later parts of the thesis. In summary, Chapter 1 introduces the history of urban stormwater runoff contamination in the Brunette River Watershed and states the objectives of the study. Chapter 2 is a literature review, which summarizes different types of contaminants, their sources, transportation or migration patterns, causes of changes in concentration/loading, impact on water quality, and stormwater best management issues. Chapter 3 is materials and methods. It identifies and describes the sampling sites, the laboratory experimental equipments and analytical procedures. Chapter 4, the results and discussions, analyzes the experimental data and compares the current pollution level to other published stormwater studies. Chapter 5, conclusion and recommendations, summarizes the research data and makes recommendations on the use of data for the future. The appendices show the quality control (QC) and quality assurance (QA) records and detailed monitoring data. 3 Study Objectives This study consists of two parts. In the first part, sediment trace metal contamination and loadings in the Brunette River Watershed were analyzed from six sub-basins with different land use and permeability. The second part of the study characterizes the urban stormwater runoff quality from a diesel bus loop and evaluates the performance of a catch basin filter to remove contaminants. In summary, the objects of the research can be expressed as: i. Identify the extent and severity of trace metal contamination in the suspended and bedload sediment in the Brunette River Watershed. ii. Quantify temporal changes in the sediment trace metal concentration and loading rates for the last 30 years. iii. Track the sources of sediment contamination and identify the relations between land use and sediment contamination in the Brunette River Watershed. iv. Characterize trace metal contamination in the urban stormwater runoff associated with a diesel bus loop at the University of British Columbia. v. Identify the relations between trace metal concentration and antecedent dry days, rainfall intensity, and traffic density. vi. Evaluate the effectiveness of a selected BMP, namely a catch basin filter, in terms of contaminants removal efficiency and the filter media adsorption capacity. 4 2 L i te ra ture Review 2.1 Urban Stormwater Runoff Pollutants and Their Chemistry As a Non Point Source (NPS) of water-borne pollutant, urban stormwater runoff has been a source of water quality problems for many years (Bastian, 1986). Effective solutions to controlling this aquatic contamination can only be done with an understanding of contaminant chemistry. Based on research studies in the U.S.A. and Europe (Lorant, 1992), important contaminants in urban stormwater runoff are particulates, organic contaminants, heavy metals, biochemical oxygen demand (B.O.D.), de-icing agents, and nutrients (nitrogen, N and phosphorus, P). Particulates Particulate matter exhibits the highest degree of association with total pollutant quantity in the urban runoff, rivers and lakes (Kobriger and Geinepolos, 1984). The finer solid particulates suspended in the volume of water are known as suspended sediment or suspended load. The larger solid particulates carried along the channel bottom by the flow of water are known as bed load. Sediment is composed of suspended organic and inorganic materials. U.S. Environmental Protection Agency (EPA) identified sediments as the most widespread pollutant in rivers and streams. They can adsorb nutrients, trace metals and hydrocarbons and transport them from land to the aquatic system (APWA, 1981). Once they enter the surface water, excessive sediments can affect aquatic habitat by increasing turbidity and clogging downstream drain ways, thereby reducing flow capacity and causing flooding. Sediments come from construction activities, pavement wear, vehicles, atmospheric fallout, road maintenance, and surrounding land use activities. In the stormwater runoff, sediments increase as a result of the urban development due to the increase of imperviousness. The loose condition of disturbed soils can also lead to increasing down-slope transport of soils by the runoff. For example, excessive soil erosions from highway construction activities can remove the surface layer of soils. Thus, soils are transported into reservoirs, lakes, rivers, and streams, reducing their water holding capacity and quality by contributing to sediment loading and pollutants associated with soils. 5 Organic Contaminants Many of the organic contaminants found in the stormwater runoff are carcinogenic and may accumulate in plant and animal tissues. Common organic contaminants associated with urban stormwater runoff include oil and grease, aliphatic and aromatic hydrocarbons, pesticides, anti-sapstain chemicals for lumber, plasticizers and polychlorinated biphenyls (PCBs) (Cole et al., 1984; Kobriger and Geinepolos, 1984; Swain and Walton, 1991;Lawson etal, 1985). O f great interest among the organic contaminants are the oil and grease from the urban stormwater runoff. Oil and grease contain a wide array of hydrocarbon compounds. Some of the hydrocarbons such as polynuclear aromatic hydrocarbons are known to be toxic to aquatic organisms at low concentrations (Woodward-Clyde, 1990). The main sources of oil and grease are leakage from engines in parking lots and streets, spills at fueling stations and overfilled tanks, restaurant grease traps, and waste oil disposal (Berman etal, 1991). Oil and grease exist in the surface runoff in the forms of free floating, emulsified, or attached to particulates. Driscoll et al. (1990) have reported that 50 to 80 percent of the oil and grease in stormwater runoff may be attached to the sediments. And much of the hydrocarbon load absorbs to particles and settles out. The results of the survey from Richmond, California indicated that oil and grease concentration was highly dependent upon land use, ranging from 4.1 mg/1 in residential areas to 15.3 mg/1 in parking lots (Stenstrom et al., 1984). Qualitative analysis from the same study found that the oil and grease in the stormwater runoff most resembled used automobile crankcase oil. A simulation of management techniques indicated that a 90% reduction in discharge from commercial properties and parking lots, which represented only 9.6% of the total surface area, would result in a 53% reduction in total oil and grease discharged. Heavy Metal Heavy metals come from natural and anthropogenic sources. Lead (Pb), zinc (Zn), cadmium (Ca), and copper (Cu) are the most common heavy metals in stormwater. Chromium (Cr) and nickel (Ni) are also frequently present (EPA, 1983). Heavy metals are of concern because they are toxic to aquatic organisms, and some can be bio-cumulative. It is noted that in the New Zealand aquatic environment, anthropogenic inputs of Cu, Ni , Pb, and Zn exceed natural inputs by a significant margin (Glasby et ah, 1998). Metals can be found in dissolved forms or associated with solids in water. They end up deposited in the streambed sediment and can be picked up by benthic organisms. Some of the insoluble metals can become soluble with increasing acidity, and bound to complexing agents. The soluble form can cause both chronic and acute toxicity to aquatic organisms (Lorant, 1992). 6 Zn and Cu are common pollutants widely distributed in the aquatic environment. Their sources are mainly from soils, atmospheric deposition, industrial effluents, domestic effluents, and urban storm water runoff. There is extensive literature on the aquatic toxicity of these metals and especially their toxicity to fish. Copper is essential to life despite being as inherently toxic as non-essential heavy metals exemplified by Pb and Hg (Scheinberg, 1991). Cu is toxic at very low concentration in water and is known to cause brain damage in mammals (DWAF, 1996). Zinc is unusual in that it has low toxicity to people, but relatively high toxicity to fish (Alavaster and Lloyd, 1980). Farkas (2003) found different levels of zinc accumulated in muscle, gill, and liver tissue of fishes regardless of their age and size. Several studies have reported that the hydrological conditions, pH, hardness, salinity, metal chemistry, and sediment properties are factors affecting the metal concentration and bioavailability associated with sediment in the sediment-water systems (Straus and Tucker, 1993). For example, Perschbachera (1999) found that the increase in calcium hardness could significantly decrease copper-induced catfish mortalities, while magnesium hardness provided no protection from copper toxicity. In another study of an English catchment, Harrison and Wilson (1985) concluded that different metals exhibited different degrees of partitioning to solid material in urban run-off. The extent of the partitioning of metals between particles and solution will influence the likely efficiency of any treatment applied, and also the bioavailability of the contaminant in the environment. The adsorption of metals to particle surfaces is strongly pH dependent and is a competitive process with other ions in solution (Dzombak and Morel, 1990). Metals such as A l , Cd, Pb, Cu, M n , and Zn are most likely to have increased detrimental environmental effects as a result of a lowered pH (DWAF, 1996). The major ion composition and pH of runoff may change greatly during the course of a rainfall event as a result of progressive dilution from stormwater and the run-off of land derived material (Harrison and Wilson, 1985). Numerous experiments have been undertaken to investigate the heavy metal distribution in the sediment-water system. Research found that metal speciation in sediments had a bimodal distribution over particle-size fractions. Sweeping sediments show that the concentrations of heavy metals are a function of particle size and proportional to the inverse of the particle diameter, i.e. the highest concentrations are found in finest fractions (German, 2002). Sediment levels of Cu were associated with the silt/clay fraction in the range from 0.2 to 35 um (Lin, 2003). The metals whose removals were correlated with the volatile solids fraction were generally associated with a wider range of particle sizes, extending up to 500-1000 um. Thus, potential toxicity to the aquatic ecosystem can be caused by the fine sediments as well as coarse sediments (Lin, 2003). 7 Biochemical Oxygen Demand (B.O.D.) Oxygen demanding substances including plant debris (such as leaves and lawn trimmings), animal excrement, street litter, and organic matter are commonly found in stormwater (EPA, 1992; Woodward-Clyde, 1990). Such substances are washed into the aquatic system by stormwater runoff and can depress the dissolved oxygen levels in streams, lakes, and other water bodies, thereby depriving aquatic life of needed oxygen. B O D is usually a measure of oxygen demand by bacterial activities. The B O D level is generally much higher in the urban runoff than in the treated domestic wastewater discharges, because the B O D in urban stormwater runoff is exerted over a longer period of time than in many other wastewaters (Field and Pitt, 1990). The sediments can store B O D which may become re-suspended and move to the area of dissolved oxygen (D.O.) deficit further downstream (Johnson et al., 1998). In rivers with high B O D levels, aerobic bacteria consume much of the available dissolved oxygen and rob other aquatic organisms of the oxygen they need to live. Organisms that are intolerant of low oxygen levels will not survive. In highly polluted areas with excessive organic contaminants from the stormwater runoff, there is a low diversity of aquatic organisms with high pollution-tolerance (Mitchell and Stap, 1992). Deicing Agents In North America, deicing agents are essential to wintertime road maintenance in terms of ice and snow removing from highways and streets. The deicing chemicals, mainly chlorides, are carried by melting snow and ice onto vegetation along the roadside and eventually to local rivers, streams, and other bodies of water. Depending on the amount of deicing chemicals used, the salts can accumulate in soils along roadsides, affecting trees and other vegetation. At elevated concentrations in soils, deicing agents can inhibit plants' ability to absorb water and nutrients and impede long-term plant growth. Degradation of roadside vegetation can also reduce the ability of these areas to act as buffers to slow runoff of contaminants into the watershed. Salts carried by runoff into aquatic ecosystems can build up to concentrations sufficient to affect aquatic plants and organisms. At high salt levels, some fish species can be affected over a prolonged exposure period, depending on their salt tolerance ability. The reductions of less salinity tolerant organisms, such as macro-invertebrates, will have effects throughout the food web (Environmental Literacy Council, 2002). Nutrients Nutrients such as nitrogen and phosphorus have been known to cause accelerated eutrophication in water bodies. The stormwater runoff transports excessive nutrients from various areas into the aquatic systems. These nutrients can result in accelerated growth of vegetation or algae resulting in impaired use of water in streams, lakes and other water systems. In addition, un-ionized ammonia (one of the nitrogen 8 forms) can be toxic to fish. Special purpose studies, such as characterization of the excessive aquatic plant growth problem, determination of the hydrodynamics of the water body can determine whether the nutrients are causing or significantly contributing to the eutrophication-related water quality use impairment problem (Lee and Jones-Lee 1996; 1997). Some threshold levels have been set for most of these contaminants in the U.S.A. by the Environmental Protection Agency (EPA) (Lorant, 1992) and in the Canada by Canadian Council of Ministers of Environment (1999). 2.2 Source of Heavy Metal Contamination in the Urban Stormwater Runoff Urban stormwater runoff is being recognized as a substantial source of pollutants to receiving waters. These contaminants are generated mainly from anthropogenic activities. An overview of urban stormwater generated pollutants and their chemistry is given in this section. Listed in the Table 2.1 are sources of contaminants in the urban stormwater runoff. Table 2.1 Sources of contaminants in the urban stormwater runoff Contaminants Sources Sediment and Floatable Streets, lawns, driveways, roads, construction activities, atmospheric deposition, drainage channel erosion Pesticides and Herbicides Residential lawns and gardens, roadsides, utility right-of-ways, commercial and industrial landscaped areas, soil wash-off Organic Materials Residential lawns and gardens, commercial landscaping, animal wastes Metals Automobiles, bridges, atmospheric deposition, industrial areas, soil erosion, corroding metal surfaces, combustion processes Oil and Roads, driveways, parking lots, vehicle maintenance areas, gas stations, Grease/Hydrocarbons elicit dumping to the storm drains, direct wash-off from vehicles Bacteria and Viruses Lawns, roads, leaky sanitary sewer lines, sanitary sewer cross-connections, animal waste, septic system (1) Source EPA (1999) 2.2.1 Heavy Metal in the Atmospheric Particulates In urban areas, many economic activities (transportation, domestic waste incineration, mining, etc.) generate atmospheric emission of mostly sub-micron particles. These very small particles may be transported by wind over very long distances to watersheds and their drainage network. The rain carries 9 the heavy metals from the air into the aquatic ecosystem, where it becomes a major trace metal input. Two kinds of depositions may significantly contribute to the contamination of urban runoff and receiving ecosystems. They are wet deposition and dry deposition, both playing an important role in transport of metals. Significant wet and dry fallout occurs locally in urban areas. Dry and wet depositions have different fates inside the urban drainage system due to street cleaning (Garnaud et al., 1999). The atmospheric deposited metal concentration is regionally and heavily dependant on the adjacent land use. For example, Ouellet and Jones (1982) found that major sources of heavy metal deposition in the province of Quebec are caused by the emissions from fossil fuel combustion. Wet Deposition Wet deposition is defined as the deposition of pollutants from atmosphere that occurs during precipitation events. Acid rain is one form of wet deposition. A one-year collection study of the concentrations of metals in air and precipitation reveals that on a weight-weight basis, most metals tend to be concentrated about 1000 times more in precipitation than in air (Struempler, 1979). A study in Germany showed that heavy metals deposited were predominantly (over 90%) dissolved in rain or melting snow. (Nuernberg et al., 1984) This high dissolved metal fraction is related to the lower pH of urban stormwater and to lower SS values compared to river water where most metals are generally bound to particulate matter. Between 1975 and 1982, Lindverg and Turner (1983) collected wet precipitation at forested sites in the southeastern United States and analyzed for Cd, Mn, Pb, and Zn. They observed significant temporal and spatial trends: all metal concentrations were higher during warm weather periods than during winter. Metal concentrations in rain are negatively correlated with rainfall amount, indicating a dilution effect. Dry Deposition Dry deposition refers to the delivery of air pollutants in the gaseous or particle phase to surfaces. Although mostly dissolved in wet deposits, the particulate fraction of metals is of great importance since it allows a comparison between dry and wet deposits. Particulate metal concentrations were not significantly different in dry and wet fallout samples collected at Paris centre. Like bulk metal concentrations in the stormwater, particulate associated metals were not correlated with dry weather duration and the amount of collected particles. (Garnaud et al., 1999). Avila and Alarcon (2003) reported that deposition in bulk precipitation was lower than dry deposition based on research of two sites. 2.2.2 Heavy Metals in Relation to Land Use Activity Urbanization influences the water cycle through changes in flow and water quality. During the last two centuries, heavy metals released by human activities have superimposed new patterns of metal 10 distribution on those that occur naturally. Urban stormwater runoff introduces heavy metals from different anthropogenic sources into the eco-system. Table 2.2 summarizes the concentration ranges of contaminants associated with different land uses. Table 2.2 Typical contaminants in the stormwater runoff from different land uses Toxic Pollutant Concentration (u.2/2) Land Use v ^ 6 5 7 C d C u Pb Zn T P H s P A H s Residential area 0.04-10.7 14-221 120-1,000 47-1,170 15.7-59.8 -Commercial area 0.02-1.06 10.4 160-220 53-1,065 16.4-34.0 ~ City downtown 2.6-7.0 143-390 1,880-2,550 470-534 8.8-51.8 -Industrial area 0.7-3.4 228 488-1,410 655-1,445 61.9-507.0 -Parking lot 1.0-14.6 206 2,000-15,000 1,600 ~ -Street 0.22-3.90 22-220 - 44-480 ~ 0.2-20 Highway 0.6-4.3 90-281 130-4,800 250-336 - — (1) Source from EPA (2004) (2) TPHs: total petroleum hydrocarbons, (3) PAHs: polycyclic aromatic hydrocarbons High levels of trace metals (particularly chromium, copper, lead, and zinc) found in stormwater may come from automobiles, residential and industrial facilities (Woodward-Clyde, 1990). They include tires, brakes, linings, batteries, fluorescent lights, dental amalgams, paints, photographic papers, and photo chemicals (Percy, 2004). If improperly stored and disposed, trace metals from industrial sources can lead to contaminated runoff. US EPA regulates 11 sectors of industrial activities from stormwater discharging. Industries related to mineral, metal, oil and gas manufacture, waste treatment and disposal, recycling, transportation and construction activities are considered to be hazardous and a special permit is required for the stormwater discharge from their industrial areas (Storm Water Permit for Industrial Activities). On the other hand, trace metal pollution from non-point sources, such as vehicles, is much more difficult to quantify and regulate. A recent comprehensive review of urban runoff contamination identified the vehicle traffic to be the primary source of non-point source of trace metals to the urban environment (Johnson, 1997). Pollution potential depends on traffic volume, types of vehicles, way of driving, and maintenance. A publication from University of Wisconsin indicated that concentrations of zinc, cadmium, chromium and lead appear to be directly correlated with the volume of traffic on streets that drain into a storm sewer system 11 (Johnson and Juengst, 1997). Contamination with heavy metals and P A H (polycyclic aromatic hydrocarbons) usually decrease with increasing distance from the road (Johnson and Juengst, 1997). Contaminants contributed from automobiles have been attributed to 5 main sources: exhaust emission, tire wear, oil and grease, corrosion, and breakdown of the road surface (McCallum, 1995). Exhaust emissions have changed in composition over the last 30 years due to regulatory controls on additives in fuels. In the 1970's, Pb was the trace metal of greatest concern, because Tetraethyl Lead (TEL) was used as a gasoline additive to enhance octane ratings. At its peak usage in 1973, Canadian automotive emissions of lead were 14,360 tonnes - approximately 70% of total national lead emissions to the atmosphere (Poon, 1989). The Pb was phased out of gasoline in North America with the substitution of the octane-enhancing gasoline additive M M T (McCallum and Hall, 1998). Starting in 1974, Methylcylopentadienyl Manganese Tricarbonyl (MMT) began to be used in gasoline and is now the primary octane enhancing additive used by Canadian refineries (Forget et al., 1994). M M T is an octane enhancer that forms manganese particles when burned as a gasoline additive (Blumberg and Walsh, 2004). Environmentalists are concerned about manganese from the M M T being emitted into the atmosphere. McCallum (1995) reported a 43% increase of Mn in the street sediment from the Brunette River Watershed from year 1973 to 1993. As an element, manganese is not bio-degradable, so it will accumulate in the soil and air over time. Manganese can be a potent neurotoxin with chronic exposure (ATSDR, 2000), since the transport rate of manganese out of the brain is slower than the transport rate for manganese entering the brain (Yokel and Crossgrove, 2004). Other studies have shown that fine particles containing manganese can be absorbed into the blood through the lungs and transported directly into the central nervous system and brain (Dobson et al., 2004; Zayed et al, 1999). If M M T is widely used as a gasoline additive, it could take decades before the full health consequences are understood (Yokel and Crossgrove, 2004). Water pipes and roofs covered with copper have been identified as large sources of copper contamination in urban stormwater runoff (Westerlund, 2001). A further source could be traffic on streets and roads that, due to wear of brake linings that contain copper, contributes copper to the urban watershed. Westerlund from Stockholm Environment and Health Protection Administration (2001) found that copper and zinc contents are commonly used in the brake linings to conduct heat. The metal content varies significantly among different types of vehicles, such as passenger cars, heavy transport vehicles, and buses. Even in the same type of vehicle, the metal contents can be high in lining for certain models, and low in other models. For example, Cu and Zn content in brake linings for Volvos were 15,000 mg/kg and 14,900 mg/kg. While, only 76.9 mg/kg of Cu and 127 mg/kg of Zn were found in the brake linings of the Scania (Reybekiel, 2001). Tire-tread material has a Zn content of about 1%. Analyses showed that the quantity of Zn released by tire wear in the mid-1990s was of the same magnitude as that released from waste incineration. For 12 1999, the quantity of Zn released by tire wear in the U.S. is estimated to be 10,000 to 11,000 metric tons. Studies suggest that Zn in whole and ground-rubber tires may also be mobile and bio-available. Elevated Zn levels have been demonstrated in leaching experiments using tire rubber and simulated rainwater, synthetic acid rain, lake water, and synthetic seawater (Councell et al., 2004). In addition to tires, there are numerous other sources of Zn emissions to highway runoff waters. These include brake-pad dust, fuel, degradation or lubricating oil and deicing chemicals (McCallum, 1995). The yearly tonnage of Zn consumed in the U.S. that goes to tires is relatively small (4.75% in 1999). Roughly half the Zn consumed in the U.S. goes into galvanization of iron and steel to prevent corrosion (50% in 1999). O f the remaining half, about 17% goes into Zn-based alloys, 13% into brass and bronze, and 20% into non-metallurgical applications (Councell et al., 2004). Automotive emissions of Cd, Cu, Cr, Ni , Pb and Zn along roadway with high traffic density can accumulate in the roadside soil at levels that are toxic to organisms in surrounding environments. In the process of rainfall, urban runoff picks up elevated amount of heavy metal by washing off the accumulated soil from the roadway surface. A study in 1-75, Cincinnati, Ohio demonstrated that heavy metal contamination in the top 15 cm of the roadside soil samples is very high compared to local background level (Turer et al., 2001). Besides depth, the concentration of heavy metal was found to decrease with the distance from highway. Ndiokwere (1984) observed higher accumulations of the metals on the roadside vegetation and soil samples than on crops and vegetables from sites a little farther away. Scanlon (1977) found that metal concentrations decreased to background levels within 48 m from the highway on Virginia highways compared to 50 m on Illinois highways, 30 m on Denver highway (Zimdhal, 1972), and between 30 -35 m on Milwaukee highways (Gupta et al, 1981). The above findings demonstrated that motor vehicle is the direct source of aerial heavy metal deposition into the local soil. These automobile-generated heavy metals are most likely to add to the metal background level of the surrounding land, thereby increasing the heavy metal levels on land adjacent to the highway. There must be a cumulative effect, in which case metals released on the highway become trapped in soil sediment and vegetation adjacent to the highway. However, with proper highway maintenance practices, some of these metals could be removed from the highway surface. 2.3 Stormwater Characterization The purpose of the stormwater characterization is to determine the periodic occurrence, variable quality and quantity of the storm water being discharged from a certain area (Bastian, 1986). Stormwater characterization is essential for the water quality control, monitoring protocol selection, and stormwater best management practice development. 13 First flush and seasonal variations are two important aspects of stormwater characterization. Usually the stormwater that initially runs off an area will be more contaminated than the stormwater that runs off later, after the catchment has been flushed clean. The stormwater containing the high initial contaminant load is called the first flush. First flush is critical in the design of stormwater pollution controls. However, it is not observed in all cases. Factors such as drainage characteristics of the catchment, mobility of the contaminants may prevent the first flush phenomenon from happening (EPA, 2005). Different opinions existed on whether there were seasonal (winter, non-winter) variations in heavy metal contaminants loadings/concentrations. Although no observable seasonal variation was found in Sacramento, California (Kerri et al., 1985) and Organ County and Miami, Florida (Yousef et al., 1985; Mckenzie and Irwin, 1983), other researchers have reported different results. Kobriger and Geinepolos (1984) found that the median lead, zinc and iron concentrations were higher in the winter months, whereas chromium, cadmium, nickel, and copper did not show seasonal variations. Two reasons can explain the higher winter pollutant loading. The first reason could relate to the factors, such as increased surface loading due to lack of sweeping, less atmospheric wind blow-off, more stop-and-go traffic due to dangerous winter conditions, reduced runoff caused by freezing winter roadway conditions, increased automobile body rusting due to winter de-icing agents and a lack of regular highway maintenance to stop surface deterioration. (Kobriger and Geinepolos, 1984; Lorant, 1992). Another explanation by Oberts et al. (2000) is that in the non-winter condition, rainfall induced runoff often results in a quick response, with rapidly ascending flows and high peaks and volumes. These rainfall events have a high intensity, which can transport both soluble and pollutants associated with solids. In the cold winter, however, snowmelt is generally a slower process, which can transport contaminant-associated solids of varying size, depending on the energy of the event. This transport can be dictated by the amount of available contaminants or by the transport capacity of the flows. 2.4 Stormwater Pollution Control and Management During the last decade, treatment of stormwater and road runoff has received increasing attention. As a result of this, different types of treatment methods for stormwater have been implemented in cities and along highways. These methods, often called Best Management Practices (BMPs), change the path and final destination of water, as well as pollutants, in urban areas. However, the available knowledge of the suitability and performance of these BMPs in pollution control is inconsistent and the effects of various BMPs on receiving water quality are poorly understood (Mikkelsen et al., 2001). Storm water BMPs can be grouped into two broad categories: structural and non-structural. Structural BMPs include engineered and constructed systems that are designed to provide water quantity 14 and quality control of storm water runoff. Non-structural BMPs include a range of pollution prevention, education, institution, management and development practices designed to limit the conversion of rainfall to runoff and to prevent pollutants from entering runoff at the source of runoff generation (EPA, 1999). 2.4.1 Structural Stormwater Best Management Practice A wide variety of structural BMPs, including retention, buffer, exfiltration, catch basin inserts, hydraulic separation and filter apparatus, are in use for storm water management. They are designed to accomplish pollutant removal (sedimentation, floatation, filtration, absorption, biological uptake, biological conversion, and degradation), pollutant source reduction, water quality improvement, groundwater recharge, flood control, stream bank erosion control, or some combination of these objectives. As a stormwater best management practice, the catch basin filter will be described in detail, since it was used in this investigation. Catch basin filters are devices installed underneath catch basin inlets to treat stormwater that enters the catch basin. Configurations vary, but the essential element is some type of filtration media (fiberglass, activated carbon, oil absorbent material, etc.) to remove contaminants as the stormwater passes through the filter. Contaminant removal mechanisms include settling, filtration, adsorption, and absorption. (Dayton and Knight, 1999) A catch basin filter is typically designed to remove particulate contaminants, oil and grease, trash and debris right at the storm drain before the stormwater runoff reaches the drainage system. Instead of attenuation of peak runoff flows, catch basin filter serves the water quality improvement purpose only. It can serve as a special purpose B M P : to reduce oil and grease from high traffic sites, to control sediment during municipal construction and unpaved areas with high coarse sediment load (CBIC, 1995). In Seattle area, the recommended maximum catchment area is 465 square meters per unit ( K C , 1998). This represents a design flow of 72 liter/min. Leif (1998) reported that catch basin filters showed a greater than 90% removal efficiency of medium sand. But it is not recommended for dissolved metals or fine particulates such as silt or clay. The filtration media can be specified to remove specific contaminants. One of the popular and effective filter media is a naturally occurring volcanic ash called perlite. Perlite can be used as a stand-alone media or in conjunction with other available media. Although perlite is not chemically active, its highly porous nature, multi-cellular structure, and rough edges make it very effective for removal of fine particles and oil and grease. Laboratory and field-testing have demonstrated that perlite is able to capture even fine silt and clay particles while maintaining a robust resistance to clogging by heavy sediment loads. The perlite's extreme porosity and high surface area allow it to act like a sponge and physically capture free oils and greases as these pollutants flow across its surface. Stenstrom and Lau (1998) compared 5 sorbents in a Continuous Deflective Separation (CDS) unit in terms of oil and grease 15 removal. With a density of 0.1 g/ml, and screen size of 5.7 mm, perlite is the lightest sorbent with large particle size. It was found with relatively low percentage removal efficiency (41%), but a very high adsorption capacity (729 mg/g) without reaching saturation state. 2.4.2 Non Structural Stormwater Best Management Practice In addition to structural non-point source water pollution technologies, there are many non-structural practices. The specific strategic options at the watershed scale include: • Integrated watersheds management planning • Land use control • Polluter-pay principle • Public education • Stream and other environmentally sensitive area stewardship In British Columbia, a major body of knowledge in watershed management has developed during the past 35 years (Toews and Hetherington, 2004). Municipalities have prepared and continually updated Storm Water Management Plans (SWMPs) for each watershed. The overall goal of the watershed management plan is to protect and enhance the integrity of the aquatic and terrestrial ecosystems and human populations they support in a manner that accommodates growth and development ( G V R D , 2001). The preparation of SWMPs involves the analysis of historical rainfall data, estimation of possible rainfall hyetograph and runoff, and estimation of Combined Sewer Overflow (CSO) volume and frequency, a prediction of pollution loadings, and erosion susceptible areas (Lee, 2004). Land use control is often incorporated as part of the integrated watershed management plan. This may include the creation of wide buffer zones around streams, lakes and reservoirs, the preservation and restoration of wetlands, the removal and disallowance of water pollution generating activates in the watershed of drinking and recreational water source, the preservation and enhancement of naturally/ecologically sensitive areas, and the requirements of sustainable development practices (Lee, 2004). Some municipalities have passed municipal bylaws to adopt the polluter-pay principle. These bylaws include: sewer use bylaw, stormwater runoff levy, development cost charges, tree cutting and soil removal bylaws, and aquatic and wild life habitat compensation (Lee, 2004). Public education and environmental stewardship are effective measures in controlling non-point water pollution sources as well. They are usually carried out by government, community and environmental groups. Through voluntary cleanup, habitat restoration, and environmental monitoring, the mitigation of the non-point source pollution can be implemented by residents, farmers, students, and business owners (Lee, 2004). 16 2.5 Brunette River Watershed Brunette River Watershed is the largest watershed in the urban metropolitan area of Greater Vancouver Regional Districts (GVRD). It is an urban watershed with over 175,000 people living or working in it. Over 200 km of open streams and over 140 hectares of wetlands are inhabited by over 200 bird species, and 23 mammal species. Salmon and trout still spawn in some creeks. Brunette watershed has been selected by the G V R D liquid-waste management plan as a case study to develop an integrated management plan that incorporates a stormwater management component (Compass Resources Management Ltd., 1997). 2.5.1 Historical Water/Sediment/Biota Studies on the Brunette River Watershed Accompanying the rapid development of the Brunette River Watershed, changes in the aquatic environment are caused by physical and chemical alterations of the streams and lakes within the watershed (Hall et al., 1999). A record of contaminant history has been traced from sediment core, which adsorbs contaminants and settles out of the water column of Burnaby Lake. Lake core analysis in 1994 showed that the rising contaminant levels are apparently correlated with the early Burnaby residential population increase, with the exception of D D T degradation due to the declining agricultural land use around 1950 and the restriction on pesticide sale in 1973 (Wong, 1996). Declining lake core lead and polychlorinated biphenyls (PCB) contaminant loads, starting between 1970 and 1980, are the probable result of similar restrictions on commercial products (Hall et al., 1999). Since that time, the lead gasoline additive has been gradually phased out and was completely eliminated from gasoline in 1990. Similarly, the manufacture of PCB in the United States was prohibited in 1977 and an accelerated Canadian phase-out program was initiated in 1985 (Environment Canada, 1987). By comparing 1993 surface sediment contaminant concentrations with levels recorded at the same locations during a baseline study conducted in 1973, McCallum (1995) found 140% increase in mercury levels and 81% increase in copper levels in stream sediments. This result contradicts the declining trend observed in the lake core, which may be explained by the past industrial land usage, prior to implementation of waste management regulations (Hall et al., 1999). After the industrial expansion in the Still Creek area in 1947, efforts were made to eliminate the significant point sources discharges from industries during the 1970s (Hall and Wiens, 1976). The recent decline of copper and mercury recorded in the lake core may be the result of the abatement of industrial point-source pollution into Still Creek, while non-point sources throughout the watershed are rising (Hall et al., 1999). Manganese has increased markedly in stream and street sediments since 1973, which is directly related to the introduction of the gasoline additive, Methylcyclopentadienyl Manganese Tricarbonyl (MMT) , as a replacement for 17 tetraethyl lead. The levels of lead, PCB, and D D T related compounds in streambed sediments have all declined in the last 20 years, confirming lake core results (Hall et al., 1999). 2.5.2 C h a n g e in L a n d Use O v e r T i m e The Brunette River Watershed existed in a near pristine state 130 years ago. Early development quickly resulted in the logging of most of the forest in the watershed, allowing agriculture to become the dominant land use (Gardner Dunster Associates Ltd., 1992). The first zoning of industrial land in the Still Creek area in 1947 marked the beginning of a prolonged commercial and industrial expansion (Dawson etal., 1985). McCallum (1995) has quantified the recent changes in land activity, land cover, and traffic density from year 1973 to 1993 with the aid of Geographic Integrated System (GIS) software. Land activity has changed marginally over 20 years from 1973 to 1993. As shown in Table 2.3, changes in the areas of major land use (i.e. residential, commercial, industrial) and impermeable ground covers have been relatively small. The increase in vehicular traffic density throughout the watershed is relatively large. Table 2.3 Land use and traffic changes in the Brunette River watershed between 1973 and 1993 Land Use 1973 1993 (% of the total area) % Change Residential 40.8 • 45.7 + 12 Industrial 11.9 13.2 + 11 Commercial 3.6 4.1 + 14 Institutional 6.6 6.4 -> o Agricultural 1.4 0 -100 Traffic Density (million vehicle-km per day) 3.0 4.3 +44 (1) Source from Hall et al. (1999) The minor changes in land zonings are not likely responsible for the large changes in sediment trace metal levels. The 44% increases in traffic density throughout the watershed may be the main reason of the increase in contaminant levels between 1973 and 1993. Analysis of traffic distribution and spatial contaminant patterns suggest that traffic emissions are a probable source of lead and zinc contamination in the watershed. Copper, chromium and nickel levels are less consistently related to traffic indices. Mn level in the sediment, although increased remarkably in stream and street sediments since 1973, did not exhibit a spatial correlation with traffic due to the unique geochemical properties of the element (Hall et al, 1999). 18 Automobile traffic has rapidly increased throughout British Columbia from 1950 to 1990, particularly in the heavily populated lower mainland region of the province. However, in the last decade, although the population has been growing in a fast pace (22%), the numbers of registered vehicles in the City of Burnaby only has a minor increase of 5% (Figure 2.1) because of the increase in public transportation and car pooling (http://www.gvrd.bc.ca/growth/keyfacts/vehicles.htm). Figure 2.1 Automobiles registration in Burnaby from 1994 to 2004 (http://www.gvrd.bc.ca/growth/keyfacts/vehicles.htm) The total imperviousness for the Brunette River Watershed increased rapidly over the past 30 years. Although there is a fair amount of spatial variation in imperviousness, none of the major sub-watersheds has imperviousness below 27%. The total impervious area covers 45% of the Brunette River Watershed in 2001. At the present rate of increase, it is predicted by G V R D that the over watershed impervious areas will increase to 58% by 2036 (Table 2.4). Unless appropriate management strategies are implemented, the trace metal levels may pose significant risk to aquatic health throughout the watershed within the next 30 years. Table 2.4 Total impervious area over 60 years in the Brunette River watershed Land Cover 1973 1993 2001 2036 Total Impervious Area 34% 41% 45% 58% 19 2.5.3 S u m m a r y o f Spat ia l Dif ference i n the W a t e r Q u a l i t y Water quality may vary considerably over geology, land use, impervious areas, climate, etc. G V R D stormwater characterization studies ( G V R D 1994, 1996, 1998) confirm that water and sediment quality criteria are exceeded frequently in highly urbanized areas (45% imperviousness or higher). In the Brunette River Watershed baseflow study, Macdonald et al. (1997) found that the Pb objective of 4 ug/1 was exceeded in 2 of 65 samples. The zinc objective of 2 ug/1 was exceeded in 22 of 116 samples. The copper objective of 2 p.g/1 was exceeded in 60 of 103 samples. Even within the same watershed, considerable areal difference exists. Copper and zinc levels were higher in the Still Creek than in Eagle or Stoney Creek and the Brunette River. Detailed studies in the Salmon River (a watershed on the urban/rural fringe), with imperviousness ranging from 5 to 15%, suggest that exceedances of water quality objectives for aquatic life are infrequent, despite the presence of localized water quality concerns from specific sources (i.e., nitrate pollution from septic systems and agriculture). Moderate correlations were found between selected water and sediment quality parameters (fecal coliforms, conductivity, zinc and lead in sediments) and imperviousness, but there is insufficient information to identify a threshold for water quality impacts (Zandbergen, 1998). Three case studies were summarized in the following table on the characteristics of Brunette River Watershed, Sumas Watershed, and Salmon Watershed and their baseflow (BF) quality (IRE, 2004). 20 Table 2.5 Spatial diffeiencesin the Ian Characteristics Brunette Watershed Salmon Watershed Sumas Watershed Location Mainly in Burnaby Mostly in Township of Langley From Whatcom County, USA into Fraser River Type I iban Urban/ 'rural fringe Rural Land, use • • GreenSpace 20% Ind./Com/ 16% Residential 50 • .6 » Agriculture 50%-Green Space 34% ind./Com. 3% Residential 13% • • • • Agriculture 59% Git'en Space 31% Ind./Corn, 4% Residential 6% TIA 41.1% - ,5.7% • Microbial & oxygen problems elevated groundwater • Low DO in summer Storm flows high nitrate (stream « Low DO & high 3'race metals elevated eutrophication) Ammonia in Fall Problem • • Organic contaminants elevated •Unhealthy aquatic ecosystem Microbial problems in summer • lligli coliform counts in Winter • Transportation • Septic Systems • Agricultural Waste Impervious Surfaces • Urban-Rural Conllicts 9 Drainage & Flooding • Street and Storm Runoff • Dri liking-Ground Water (Schreier et al.. 2004) Table 2.6 Spatial difference in (he land use and water quality under base 'flowcondition inlocal watersheds (continued) Base flow Water quality Brunette Watershed (Brunette River) Winter Summer Salmon Watershed (Salmon River) Sumas Watershed (Main Stream) Winter Summer Winter Summer .Nitrate-N ( 1 ) Ammonia '" Fecal Conform® Dissolved 0 , ( l ) Total Cu® Total. Zh® 200-1000 9-11 5-10 100,2000 3-7 3. 15-25 0-3 0-3 700:: 6-11 0-2 0 300-400 4-10 1-2 0-0.2 100-1000 10-12 0-5 Q-0,8 2-12 (Schreier el cti. 2004) T l A: Total Impervious Area Units: (ii mg/1 (2) FC/1 OOrtil (3)ug;! 2.5.4 Relation of Land Use to Benthic Invertebrate Community Structure Stormwater impacts on streams can be determined through an assessment of benthic macro-invertebrate community structure. Macro-invertebrate communities, on the lower end of the trophic level within streams, provide a base for the in-stream food chain health (Judith and Gerlach, 1998). Schueler (1994) reported correlations between macro-invertebrate diversity and imperviousness. Macro-invertebrates were replaced with more pollution tolerant species when imperviousness increased. Sediments are usually chosen as the media of analysis in this study because of the "particle-reactive" nature of many contaminant compounds. Many contaminants entering an aquatic system quickly adsorb to sediments and eventually settle out of the water column. This interaction between contaminants and sediments potentially alters the benthic environment and also leaves a record of contaminant history. Assessments of benthic invertebrate communities were conducted in several studies examining effects of urbanization on small streams in the G V R D area. In Delta, a single site in Cougar Creek canyon has been assessed weekly for the last five years by 5-minute visual examinations of benthic habitat (Rithaler, 1998). Benthic Community Structure (BCS) was assessed in the Brunette, Salmon and Sumas Rivers and in four reference streams (Pepin Creek, Betrand Creek, Anderson Creek, and Little Campbell River) by Richardson (1998) at the University of British Columbia. Invertebrates were collected, identified and enumerated from several sites in each of the Brunette, Salmon and Sumas Rivers, and one site in each reference stream with a surber sampler ( G V R D , 1998). The survey of benthic invertebrates across the lower mainland showed that there were large differences in the organisms present in urban versus rural streams. Species that were largely absent from the Brunette River Watershed were also those species most sensitive to heavy metal exposure in the flow-through experimental stream experiment beside a clean, mountain stream (Richardson, 1998). These results show that contaminants have a large effect on urban stream communities. 23 3 Mate r ia l s and Methods 3.1 Experimental Overview To update the sediment contamination and loading rate in the Brunette River watershed, Northwest Hydraulics Consultant (NHC) collected suspended/bed sediment samples from six sub-watersheds (each with slightly different urban land uses) in 2003. Following this, an urban stormwater runoff quality monitoring and management project was carried out in the U B C diesel bus loop. The monitoring program was undertaken from September 2004 to June 2005. Sediment samples from the Brunette River watershed and stormwater runoff samples from the U B C diesel bus loop were analyzed at the Environmental Engineering Laboratory at the University of British Columbia. The parameters analyzed in the Brunette River watershed sediment were mainly trace metals. The parameters analyzed in the U B C bus loop stormwater runoff included suspended solids, trace metals, and oil and grease. The trace metals of particular concern were Copper (Cu), Manganese (Mn), Iron (Fe), and Zinc (Zn). These metals were chosen because of their reported higher concentrations in urban runoff, their impacts on the environment and their common association with vehicular-related urban stormwater runoff pollution. 3.2 Field Methods 3.2.1 Brunette River Watershed 3.2.1.1 Sediment Collection Suspended Sediment Suspended sediments were sampled on 30 January, 20 February, 12 March, 13 March, 07 April, and 16 October, 2003. Samples were collected manually under storm flow conditions in a sequence rotating from station to station during individual storm events. There were no summer storm events of sufficient magnitude or duration to be sampled. Sample stations are located at stream access points from six urban 24 streams (Still Creek, Eagle Creek, Ramsay Creek, Stoney Creek, and Brunette River) in the Brunette River watershed. These stations are described in Table 3.1. Table 3.1 Steam sediments sampling stations Stream Location Suspended Sampling Bedload Sampling Drainage Area (km 2 ) Brunette River At Lower Stop-log structure • 63 Eagle Creek At Winston Street • 5.9 Ramsay Creek B C Hydro Access Road • • 0.8 Silver Creek Winston Street • 1.7 Still Creek Douglas Road • • 22.1 Stoney Creek Government Street • • 7.1 Flow measurement, along with the water quality monitoring, was used to calculate the loading of suspended sediment in Brunette River watershed. Since discrete grab samples were collected over the rainfall event, the sampled mean concentration method was used to calculate the instantaneous loading rate over the storm event. Bedload Sediment Bedload basins were excavated in Stoney and Ramsay creeks from 15 to 17 September 2003. A n existing concrete sediment chamber in Silver Creek was cleaned out for bedload collection on 1 October 2003. Al l three bedload basins were filled during a large, early season storm event on 16 October 2003. The Stoney Creek basin was excavated on 29 October and 13 November 2003. The Ramsay Creek basin was excavated later on November 20. Due to the sandy bank structure, bedload from Still Creek was dredged up from four different locations during the same period. Depositional volumes were calculated by comparing post-excavation and post-storm bed surfaces in AutoCAD. 3.2.1.2 Parameters o f Interest The contaminants o f interest in this project are the trace metals in the stream sediment from Brunette River watershed. Detailed analyses are focused on four metals: C u , M n , Fe, and Z n . These metals were selected because they are frequently elevated during storms and are common contaminants o f concern in urbanized watersheds (Chapman, 1992; M c C a l l u m , 1995). 25 3.2.1.3 Sample Preparation and Analysis Techniques Laboratory analysis was performed in the Environmental Engineering Laboratory and the Soil Laboratory in the Department of Civil Engineering, U B C . The analysis methods were selected according to the precision and sensitivity required, and after reviewing of existing data, and consultation with the laboratory staff. Suspended Sediment Suspended sediment concentrations from Brunette River watershed were determined by N H C immediately after the samples were collected. The dried sediments attached to membrane filters (0.45um pore size) were sent to the U B C Environment Engineering Laboratory and stored at 4°C prior to trace metal (Cu, Mn, Fe, Zn) analysis. The suspended sediment samples were hot plate digested with aqua regia for total metal content determination. The sediment and filter paper were transferred into a flask. Eight ml of aqua regia (6 ml H N 0 3 : 2 ml HCL) , 30 ml of deionized water, and a few boiling chips were added to the flask as described in Standard Method 3030 (APHA, 1995). Samples were brought to a slow boil and evaporated to less than 5 ml. Additional boiling and H N 0 3 were sometimes needed until digestion was complete, as indicated by a light-colored, clear solution. The acid digest was then diluted with deionized water, filtered through Q8 filter paper (Fisher Brand) and collected in a 25-ml volumetric flask and made to volume. The heavy metal concentrations were analyzed by Flame Atomic Absorption Spectroscopy (Thermo Jarrell Ash 22). The precision and accuracy of the method was determined by performing replicates, standard addition and analyzing certified reference materials (QA/QC data refer to Appendix A) . Bedload Sediment The sediment samples were dried at the Soil Laboratory, then sieved into coarse (> 2mm), medium (63um ~ 2mm), and fine (< 63um) fractions. The medium and fine sediments were taken to the Environment Laboratory and stored at 4 °C until the heavy metal analysis. The heavy metal contents were analyzed by nitric acid digestion and flame A A , following the same procedure as applied for the suspended sediments. 3.2.2 U B C B u s L o o p The U B C diesel bus loop was built over the summer of 2004. It was completed by September when the fall semester started. The bus loop is located at 1950 Wesbrook Mall on the east side of campus. It is intended to be a temporary bus loop until the completion of the new under-ground one. The area 26 occupied by the bus loop is about 7300 m 2 . The stormwater drainage system was divided into 9 catchment areas. During a storm event, surface runoff flows downslope and into the catch basins. Stormwater pipelines are laid underneath each catch basin along the driveway, collecting surface runoff within the catchment area. There are three parallel pipelines to drain the storm flow to the main collection system located at the south end of the bus loop. By reviewing the site selection criteria, 3 storm drains were selected as sampling sites based on drainage basin size, configuration, accessibility, and possible contaminants. A downpipe to each of the monitored storm drains was pre-installed so that stormwater flow could be collected directly after it entered the drain. The down pipe cover could be removed manually. Raw street-runoff was pumped up with a hand pump from the pipeline and placed in a 250-ml plastic bottle. A catch basin filter was installed in one of the storm drains by Bay 7 (No. 44 bus stop, see Figure 3.1) to study its contaminant removal performance. The body of the filter is a mesh box attached with bagged filter media on the side walls and the bottom. The frame of the filter, made of stainless steel, fits well in the openning of the drain. The street runoff flows into the filter chamber, passes through the filter media, and flows into the downpipe, where outflow was collected. The drainage area of Bay 7 is about 318m. 3.2.2.1 Site Selection and Description „ Drain l_p # 3 - N ^  a Drain • # 2 y Si 1 £ ~gj~j Pra ia ' # 1 • i Figure 3.1 Stormwater runoff sampling site selection 27 In Figure 3.1, the circled numbers show the layout of the eleven bus loading areas at the bus loop. The black dots mark the locations of the three catch basins selected for stormwater runoff quality monitoring. They are widely distributed and generally represent the bus loop's land use and traffic conditions. As marked in Figure 3.1, the monitored catch basin 1 is located in the center of the zebra crossing, which connects the bus loop area to main campus. Its catchment area is about 370 m 2 , including the busy unloading area shared by bus routes 25, 41, 43, and 44. Catch basin 2 is located on a driveway, which also serves as a temporary parking area for bus routes 44 and 99. The drainage area of catch basin 2 includes a zebra crossing, which is less traveled than catch basin 1. The drainage area for drain 2 is 410 m 2. As for catch basin 3, it is a loading area used exclusively by bus routes 44. Bus 44 commutes from U B C to downtown during peak hours on weekdays. Detailed daily traffic densities in the U B C bus loop were derived from the Translink standard commute timetable and are listed in the Table 3.2. Table 3.2 Summary of daily bus density in the U B C bus loop Location Bus in service Traffic Density (number of buses / day) Weekday Weekend Bay 1 No. 99 unloading 197 83 (Sat.) 80(Sun.) Bay 2 No. 99 loading 197 83 (Sat.) 80(Sun.) Unloading No. 25, 41, Bay 3 212 88 43, 44, 480 Bay 4 No. 258 . 4 0 Bay 5 No. 480 23 0 Bay 6 No. 995 0 0 Bay 7 No. 44 28 0 Bay 8 No. 43 17 0 Bay 9 No. 41 60 32 Bay 10 No. 49 15 0 Bay 11 No. 25 84 56 (Sat.) 40 (Sun.) (1) Source from http://www.translink.bc.ca/ 28 3.2.2.2 Choice o f Storm Event The timing of each sampling event was reviewed in comparison to Vancouver's rainfall record and the weather forecast. Sampling events can be classified as either a valid sampling event or a questionable sampling event based on the precipitation record. A valid sampling event should satisfy the following criteria (EPA, 1992; Woodward-Clyde, 1995): • Have at least 0.1 inch (2.54 mm) of rainfall; • Occur at least 72 hours following the last measurable rainfall; • Continue sampling for at least three hours; • Have a rainfall amount between 50 and 150% of the long-term mean or median; • Have rainfall duration between 50 and 150% of the long-term mean or median. Sampling extremely large or small events can result in erroneous data collection (Woodward-Clyde, 1995). Table 3.3 summarizes the sampled storm event features in detail. Table 3.3 Features of the sampled storm events Sample Date Sample Duration (Hours) Start - End Time Rainfall (mm) Antecedent Dry Days 6-10-2004 2 15:00- 17:00 2.2 0.5 17-10-2004 2.25 10:15-13:30 4.4 2.8 29-10-2004 8 13:00-21:00 8.6 1.25 14-11-2004 9.75 11:15-21:00 5 0.7 13-12-2004 5.5 10:00-15:30 2 0.3 28-02-2005 _> 10:45-1:45 1.6 17 8-03-2005 13 11:00-24:00 4.8 1.2 16-03-2005 12 11:00-23:00 1.6 7 26-03-2005 4 0:30-4:30 22 5.2 31-03-2005 7 5:30-12:30 25.6 1.8 10-04-2005 4 17:00-21:00 13.8 3.1 2-05-2005 5.3 11:40- 15:00 1.6 14.5 14-05-2005 1.8 3:30-5:20 2.6 4.6 31-05-2005 3.5 10:10-13:40 0.6 7.7 21-06-2005 2 18:20-20:20 5.4 5.4 29 Comparison of the sampled storm events with the criteria listed in Table 3.3 indicates that most of the events were valid by meeting or exceeding the criteria. There are a many storm events during the rain season that cannot meet the requirement for a 72-hour antecedent dry period. However, this consequence is inevitable given Vancouver's climatic pattern. 3.2.2.3 Monitor ing Period and Frequency Baseline monitoring was conducted to determine long-term changes in stormwater runoff composition and the contaminant removal efficiencies of the catch basin filter. The process included stormwater collection and contaminant analysis, catch basin filter installation, and evaluation of overall performance of the catch basin filter in terms of contaminant removal efficiency. Stormwater runoff samples were collected under storm flow condition in a sequence rotating from station to station during individual storm events. The sampling was on 24-hour call. Grab samples were taken manually from storm drains and transported to the laboratory for immediate analysis. Listed are the advantages of using a grab sampling technique in stormwater runoff analysis (Environment Canada, 1993): • Change in concentration during storm events can be better analyzed; • Grab sampling is better for determining changes in stormwater runoff; • Grab sampling provided a better single stormwater runoff characteristic profile at minimal equipment cost/labor; • Data generated from grab sampling can be used for pollutographs. A total of 16 storm events were monitored between October 2004 and June 2005. The monitored period covered the entire wet season (from October to March), and part of the dry season (April to June). The sampling frequency was approximately twice a month. The sample dates were widely distributed to ensure that the samples are statistically independent. The samples taken during the 9-month monitoring period account for both seasonal influences and variability in traffic density, and represent reasonably 'average' flow conditions, from which a degree of certainty regarding the monitoring results were achieved. The total rainfall over these periods was estimated to be more than + 50% of the average annual rainfall. In order to characterize first flush and capture the whole process of runoff quality variation, the sampling period usually started from the beginning of a rainfall event and last for 3 to 8 hours depending on weather conditions. Each sampling period was intended to cover both the increasing and decreasing storm flow, in order to define pollutant loads from urban catchments with a reasonable degree of accuracy. Time-weighted grab samples were taken at certain time intervals (every 20 to 60 minutes) 30 according to the rainfall intensity. There were a few times when the same time interval could not be achieved due to the unpredictable nature of the storm. The objective of time-proportioned sampling is to obtain a representative sample of the entire sampling period, since the constituents of stormwater have been observed to vary widely within the course of a single storm event. For the catch basin filter evaluation, the study lasted for 4 months from early March to late June. The solid particle residues trapped in the filter bag were excavated once in M a y and later in June when the study ended. The reason for cleaning out the trapped particulate materials is to enhance the filter performance, estimate the particulate trapping rates, and to analyse particle distribution and heavy metal contamination from the trapped particulate materials. 3.2.2.4 Parameters o f Interests In the U B C bus loop stormwater runoff quality study, a number of parameters were analyzed to define the stormwater runoff quality. • Suspended solids, • Turbidity • Conductivity • Particulate Metal (Cu, Mn, Fe, Zn) • Dissolved Metal (Cu, Mn, Fe, Zn) • Oil and grease With a limited number of samples, an extensive laboratory analysis was conducted. 3.2.2.5 Sample Preparation and Analysis Techniques Stormwater Runoff Directly after the samples were taken from the site, they were carried back to the Environmental Laboratory to undergo turbidity, conductivity, and suspended solids tests. If parameter analysis could not be performed immediately, the samples were preserved and refrigerated in a walk-in cooler (at 4°C) and the analyses were done within the next few days. Turbidity was measured with the calibrated H A C H 21 OOP Turbidimeter . Well-mixed raw samples were placed in the sample cell and the turbidities were recorded from direct reading. If the turbidity exceeded the upper detection limit, samples were diluted and measured in the same manner. 31 For the conductivity test, the C D M 3 Conductivity Meter was first calibrated with a standard solution. Samples were then measured with a dip-in probe. The conductivity values were recorded from direct reading and adjusted by temperature factors. Suspended solids analysis was performed with a vacuum filtering apparatus. One hundred ml of well-mixed stormwater runoff were passed through a pre-weighted 0.45 urn membrane. The filter was then removed from the filtration apparatus, transferred to an aluminum-weighing dish, and dried over night at 104°C. The aluminum-weighing dish was cooled in desiccators to room temperature then weighed. Suspended metals retained by the 0.45 um membrane filter, and dissolved metals passing through the 0.45 um membrane, were both measured in order to determine the trace metal partitioning between dissolved and particulate in the stormwater runoff. The filtrate from the TSS analysis was preserved for dissolved metal analysis. For the dissolved metal concentration analysis, samples were preserved with H N 0 3 to maintain a pH value of less than 2. In terms of acid digestion, 50 ml of the well-mixed and acid-preserved runoff sample was transferred into digestion tubes. Eight ml of concentrated H N 0 3 and a few boiling chips were added to each digestion tube. The tubes were placed on the block digester and bought to a slow boil at 150 °C until the samples were evaporated down to less than 5 ml. Complete digestion was indicated by a light-colored, clear solution. The solution was filtered by Q8 filter paper and diluted to 25 ml. The total and particle associated trace metals (Cu, Mn, Fe and Zn) were analyzed by the Atomic Absorption Spectrophotometer (Thermo Jarrell Ash 22) with the direct air-acetylene flame method. Suspended metal concentration is calculated by subtracting dissolved metal from the total. Catch Basin Filter-Trapped Particulate Material Trapped particulate materials were removed twice from the filter chamber for periods of 61 days (Mar/2/2005 to May/2/2005) and 52 days (May/3/2005 to Jun/24/2005). They consisted of soil, debris, sand particles, leaves, plastic, paper, cans, hair, and cigarette butts, etc. Particulate materials were sieved into 9 fractions ranging from 9520 urn to 297u,m. Each fraction was dried at 104°C, measured for weight percentages (refer to Appendix C), and FfN0 3 digested for trace metal concentration analysis. The digestion and detection methods were the same for the suspended sediment analysis. Filter Media Adsorption Capacity In the field experiment, perlite was used as a filter media to remove contaminants from the stormwater runoff. When perlite comes in contact with stormwater, the suspended solids, trace metals, and oil and grease attach to its porous surface. Laboratory adsorption capacity tests were conducted to estimate the adsorption capacity of the filter media. 32 fir Figure 3.2 Instrument set-up for metal adsorption test As illustrated in Figure 3.2, the experiment was set up on a burette stand. A 25-ml burette containing metal solution was placed on top of the perlite column. Eight grams of medium-sized perlite (particle size ranged from 4760 um to 6680 um) were used to test the adsorption capacity for trace metals Mn and Zn. Perlite was placed in the column, which is 2.2 cm in diameter and 18 cm in length. M n (10 mg/1) and Zn (10 mg/1) standard solutions were made at around pH 7 and run through the perlite column. For each run, the metal solution dripped down from the burette, contacted the perlite, drained through the column, and collected in a 50-ml erlenmeyer flask. After each 10 ml solution exchange, 40 ml of deionized water was run through the column to wash off the residue metal solution from the perlite surface. The flow rate of the adsorption and washing processes was well controlled at 0.5 ml/min. The same steps were repeated 10 times for the adsorption test of each metal. The well-mixed effluents were then analyzed by A A S for Mn and Zn concentrations. The stormwater may contain various types of trace metals. With the consideration that calcium (Ca) and magnesium (Mg) are the most common ionic metals in urban water, interference of these metal species could affect perlite's complexation with other metals in the storm water runoff. In addition, M n and Zn may also interference with each other to cause perlite adsorption capacity deterioration. So the interference tests of Ca, Mg, Mn, and Zn were carried out in the laboratory to examine their influence on metal complexation with perlite. Procedures were the same as the perlite adsorption capacity test described above. The only difference was the metal solution constituents. Instead of pure M n or Zn solution, combinations of Ca, Mg, Mn, Zn standard solutions were used in the interference test. Constituents of three groups of mixed 33 metal solutions are described below: Mn = 5mg/1, Zn = 5 mg/1; Mn = 5 mg/1, Ca = 5 mg/1, Mg = 2 mg/1; Zn = 5 mg/1, Ca = 5 mg/1, Mg = 2 mg/1. Z(A-B)xC The perlite adsorption capacity was calculated by equation «=io Where A= Influent concentration mg/1 B = Effluent concentration mg/1 C = Volume of metal solution run through = 10 ml n = Times of repetition Oil and Grease Adsorption Capacity The composition of oil and grease in stormwater runoff is very complicated. Due to the laboratory limitations, the synthetic oil was made with motor oil and heavy mineral oil at a ratio of 2:3. One g motor oil and 1.5 grams heavy mineral oil were mixed and dissolved with 300 ml Acetone. Forty-three g perlite was soaked in the Axetone-oil solution, and contacted for 24 hours. The saturated perlite and the oil solution were then put through an extraction process. The n-Hexane and methyl-tert-butylether (80:20) solvent-recover technique was used to recover oil and grease from the perlite. Perlite was freeze-dried at -20°F for 24 hours until water free. A thimble with 20 g freeze-dried perlite was placed in a soxhlet apparatus. A mixture of 160 ml n-Hexane and 40 ml methyl-tert-butylether (80:20) was used as solvent to extract the oil and grease fraction. Each sample was extracted for oil and grease for 4 hours at a rate of 20 cycles per hour. After the extraction was completed, the flask with solvent and extracts was evaporated in a water bath at 8 5 ° C . The clear solvent was recovered in an ice-bath cooled receiver and the evaporation flask was dried at 85°C for 15 minutes. The flask with oil and grease was then cooled in a desiccator at room temperature and weighed. The oil and grease content was determined by the following equation (APHA et al., 1995). ., A , G4 - 7 i ) x l 0 0 0 mg oil and grease/g perlite = — where A = total weight of tared flask (g) B = dried and cleaned flask (g) C = perlite dry weight (g) 34 3.3 M e t a l Detection Techniques Table 3.4 Comparison of digestion techniques Sub sample Applications Preparation Analyses 1 All the samples Refriged at 4 °C Pre-analyses 2 Suspended and Bedload sediment from Brunette River Watershed Microwave digested with aqua regia solution and analyzed by flame A A Cu, Mn, Fe, Zn "i Hot plate digested with aqua regia solution and analyzed by flame A A Cu, Mn, Fe, Zn 4 Hot plate digested with nitric acid and analyzed by flame A A Cu, Mn, Fe, Zn 5 Stormwater runoff from U B C bus loop Block digestion with nitric and analyzed by flame A A Cu, Mn, Fe, Zn 6 Block digestion with aqua regia solution and analyzed by flame A A Cu, Mn, Fe, Zn Listed in Table 3.4 are the digestion techniques carried out before the mass analysis of sediment and stormwater runoff samples. Results from various digestion agents showed similar contaminant levels with slightly different precision. Statistical analyses were conducted on each of the digestion methods (refer to Appendix A). After the digestion process, Cu, Mn, Fe, and Zn solution were measured with the Atomic Absorption Spectrophotometer with air/acetylene flame at the Environmental Engineering laboratory. According to previous research, for metals of Mn, Fe, Zn, the detection limit from ICP-AES is much higher than that from Flame A A . As for Cu, the detection limits from these two techniques are the same level. The detection limits of the two methods were compared in Table 3.5. 35 Table 3.5 Comparison of trace metal detection levels for different analytical techniques Method Flame A A (Varian SpectrAA 2000 SPS-5) Flame A A (Varian SpectrAA 2000 SPS-5) 1CP-AES (Thermo Jarrell Ash ICP 61) 1CP-AES (Thermo Jarrell Ash ICP 61) Site Environmental Engineering U B C Environmental Engineering U B C Chemex Labs Soils Sciences U B C D L (mg/kg dry weight) C u M n Fe Z n 1.5* 0.03* 0.05* 0.05* 0.03- 0.03- 0.05= 0.04= 1 5 100 2 1 0.2 2 0.1 Based on the above comparison, flame A A is adopted in this research for trace metal (Cu, Mn, Fe, Zn) detection in both sediment and stormwater as the sample metal concentrations could be quite low. 3.4 Q u a l i t y C o n t r o l ( Q A ) and Q u a l i t y Assurance ( Q C ) 3.4.1 Q A and Q C i n Sample C o l l e c t i o n The integrity behind sample collection was assured by the collection of representative samples as outlined by Gupta et al. (1981a) and Environment Canada (1993). Stormwater runoff samples were placed in plastic bottles that were washed in accordance with the methods recommended by Marsalek and Greek (1984). • Washing with detergent and tap water; • Rinsing the bottles 2 to 3 times with tap water; • Rinsing the bottles with 10% nitric acid; • Rinsing the bottles 2 to 3 times with distilled/deionized water; • Air-drying the containers and capping them; • Minimum volume of each rinse was 2 to 3% of the container volume. Field blanks were to determine whether sample contamination occurred during field sample collection, handling, and transportation to the laboratory, storage, and sample analysis. According to Environment Canada (1993), field blanks, containing deionzed water, were submitted periodically to 36 determine whether sample bottles were sources of pollutant contamination. The handling procedures for field blanks are the same as with stormwater runoff handling procedures. The sampling bottles were cleaned according to the methods outlined above. One bottle blank was submitted for each batch of the runoff samples from a single storm event in the U B C bus loop. Two method blanks were also submitted for each batch of the samples being analyzed. Other Q A / Q C methodologies such as field replication, and duplicate sampling were also conducted for most of the sampled events. Al l necessary precautions in terms of field sample collection, sample handling, transportation to the laboratory, storages and laboratory analysis were considered. 3.4.2 Q A and Q C i n S a m p l e A n a l y s i s The metal detection levels are provided in Appendix A , as well as a summary of the Q A / Q C program determined from blank replicates and spike recovery. Considering the various methods mentioned in the literature review, aqua regia and nitric acid digestion are two methods with wide application and a reliable history in trace metal detection. Replicates were made from the sieved sediments from Brunette River watershed. The same amounts of sediment were digested by aqua regia and nitric acid. The t-test for the replicates analysis indicated a non-significant difference between these two digestion methods. The t-test data are provided in Appendix B. To further decide on a better digestion method, a more descriptive coefficient of variation (CV) is used. Comparing the C V value from a large group of sediment samples, C V from aqua regia digestion was lower than that of H N 0 3 ' s 3 out of 4 times. The higher degree of precision achieved by aqua regia made it a more reliable digestion method, which was then widely adopted in the sediment analysis. The accuracy of the test was estimated from standard soil reference material analyses. As illustrated in Table 3.6, Cu and Zn are within the 95% confidence limits by aqua regia digestion. Fe was significantly over-predicted, and M n was under-predicted. However, the concentrations of the four target metals were all within the acceptable range of the standard soil, which suggested that near total digestion, was achieved for these elements. 37 Table 3.6 Measurement of method accuracy using the sediment reference material Element Mean Certified Value Acceptable Range Mean/Certified Value Mean within Acceptable Range? Cu 91.2 93.9 74.4-113 0.97 Yes Mn 265.8 320 242-398 0.83 Yes Fe 17180 11600 5500-17700 1.48 Yes Zn 266 246 189-303 1.08 Yes (1) Tabulated metal concentrations in unit of ppm are the average value from 4 replicate aqua regia digestion and flame A A analysis. 3.4.3 T r a f f i c A n a l y s i s Traffic data were obtained from the traffic planning department, City of Burnaby. The traffic analysis was facilitated with the E M M E / 2 transportation model. Five pictures of traffic volumes, including North Vancouver, South Vancouver, North Burnaby, South Burnaby, and Coquitlam, cover different parts of the entire Brunette River watershed (complete data showing traffic volumes are tabulated in Appendix F). Al l the listed volumes are taken from the peak hour - the busiest traffic hour of the morning commute. A reasonable daily volume can be estimated by multiplying the peak hour volumes by a factor of 11. For each route, two volumes are marked on either side of the road. Each number stands for the traffic volume in one direction. The figures in Appendix F also include the shoreline (Burrard Inlet, Fraser River), the municipal boundaries, and names of major streets. Al l volumes in these figures are dated from the year 1999. Within a single catchment, the peak hour traffic volumes from every road were added up to calculate the total auto volume. The 1999 traffic volume data were converted to 2005 volumes through an estimated 10% increase (suggested by Ramsey Stuart, traffic planner for the City of Burnaby). 3.5 Stat ist ical Analyses Non-parametric comparative statistics were used in this study to detect differences in large sample populations because the sample distribution deviated significantly from normality. The box-whisker plot has found many applications in this study. It is a graphical tool which tempers the effects of skewed data without a loss of information. Figure 3.3 labels all of the components that may 38 exist in the plot. They include the parametric range (25 to 75%) of the data, the mean value, and the outliers from a large data set. j WWskers 1 Third'Qaa.rfife. —~| — Median Value First, quartile Figure 3.3 Components o f a box-whisker plot Spearman's rank correlation calculates all possible correlations between the trace metal concentrations with the environmental and land-use variables, it is a good tool to identify sources of contamination. Coefficients (R) fall between -1 and +1. The closer R is to +1 or -1, the stronger the likely correlation. (http://www.revision-notes.co.Uk/revision/l 81 .html) • If R = -1, there is a perfect negative correlation; • If-1 < R < -0.5, there is a strong negative correlation; • If -0.5 < R < 0, there is a weak negative correlation; • If R = 0, there is no correlation; • If 0 < R < 0.5, there is a weak positive correlation; • If R 0.5 < R < 1, there is a strong positive correlation; • If R = 1, there is a perfect positive correlation. * o MCi-Extreme Values 39 4 Results and Discussion 4.1 Trace Metals in Brunette River Watershed Stormwater Runoff Historical sediment contamination studies have been going on in the Brunette River watershed since the 1970's. To evaluate the recent contamination of streams from 1993 to 2001, sediment quality and its associated land uses are analyzed in context of the previous studies. The latest land-use data, including drainage area and total imperviousness, were obtained from G V R D . The traffic density information was obtained from the Municipality of Burnaby City. 4.1.1 Land Use Change from 1993 to 2001 Land use and land cover have been quantified with ArcGIS to examine the relation between trace metal contamination and land use characteristics. Land use has constantly changed over the last 10 years. As listed in Table 4.1, the major changes in land use are reflected by major increases in residential area (+12.9%) and total impervious area (+9.8%), and a significant decrease in open space (-11.2%). The changes in other major land uses, such as industrial, commercial, transportation, and agriculture, have been relatively small (-2%, -0.5%, +0.3%, and 0%, respectively). Table 4.1 Land use activities in the Brunette River watershed as a proportion of total area Land Use Zone 1993 (%) 2001 (%) Change (%) Residential 52.1 65 +12.9 Industrial 13.2 11.2 -2 Commercial 4.1 3.6 -0.5 Transportation 2.7 3.0 +0.3 Agricultural 0 0 0 Open Space ( l ) 28 16.8 -11.2 Total Impervious Area 41 45 +9.8 (1) Includes parks, recreational and conservation areas, open and undeveloped lands, and lakes. (2) Source from 1993 land use data from McCallum (1995); 2001 land use data from GVRD. 40 As shown in Table 4.2, rapid population growth in the Brunette River watershed is possibly the cause of the remarkable increase in residential area, total impervious area, and traffic volume. It is predicted that intense population growth will continue for the next 30 years (GVRD). Table 4.2 Population increase in the sub-basins over 40 years Watershed Name P O P 1996 P O P 1998 P O P 2001 P O P 2036 P O P Percent Increase 1996 to 2036 Still Creek Catchment 92035 99821 135154 150613 64% Eagle Creek Catchment 13798 14989 19232 22329 62% Stoney Creek Catchment 6935 7487 11530 13453 94% Lower Brunette River Catchment 20661 22550 32956 35022 70% The measurement of sediment contaminant in relation to land use is the main focus of this study. Since mere change was found from industrial, commercial, transportation and agricultural land use, the recent changes in stream sediment contamination levels are likely to be attributed to the increasing residential land use and decreasing impermeable ground cover. Major sources of contamination may come from residential lawns and gardens as well as automobile traffic. Lawn and green spaces contribute soils and organic matter to runoff, whereas automobile traffic contributes suspended solids and trace metal (Hall era/., 1999). Horner et al. (1997) found an inverse correlation between the imperviousness of an urban area and the quality of surface runoff. Steady increases in impermeable areas throughout Brunette River watershed have taken place, from 34% in 1973, through 41% in 1993, and reaching 45% in 2001. It is predicted that the total imperviousness will continue increasing to 58% in 2036 ( G V R D data). Based on a synthesis of existing studies relating imperviousness to overall watershed health, the health condition of Brunette River watershed has deteriorated from poor in 1993 to very poor in 2001 (Hall etal., 1999). Intensive urbanization in the past 10 years indicates possible adverse effects on the watershed ecosystem. Spatial variations exist among the six sub-basins, but none of them has imperviousness below 34% (Eagle Creek). Automobile traffic is a well know source of trace metal contamination. From 1993 to 2001, the number of registered vehicles in the city of Burnaby had a minor increase of 5% since the increase in public transportation and car-pooling. During the same period, land use for transportation has only increased by 0.3%, which indicates higher traffic density. Figure 4.1 lists the 2005 morning peak hour traffic volume from the five sub-catchments in the Brunette River watershed. Strong areal differences 41 existed among sub-catchments. 180000 -</) u E 150000 3 > 120000 -o £ 90000 -O 60000 -z ro 30000 -v 0 . 0 -Still Eagle Stoney Ramsay Brunette Figure 4.1 Traffic volumes in the sub-basins of Brunette River watershed in 2005 Spearman rank correlations between traffic density and imperviousness and population were calculated in Table 4.3. Traffic density showed a perfect positive correlation with the population, and a strong positive correlation with the imperviousness. Table 4.3 Spearman rank correlation matrix for traffic density, imperviousness, and population in the Brunette River watershed Traffic Density Imperviousness Population Traffic Density 1.00 Imperviousness 0.6 1.00 Population 1.00 0.6 1.00 4.1.2 Suspended Sol ids a n d T h e i r R e l a t i o n to L a n d Use 4.1.2.1 Suspended Solids Concentration The suspended sediment concentration from six tributaries in the Brunette River watershed are presented in the box whisker plot in Figure 4.2. A wide distribution of the median values from as low as 15 mg/1 to 300 mg/1 was found within the sub-basins. Ramsay Creek had the highest median suspended sediment concentration of 296 mg/1, with large variation among the sampled storm events. The second 42 highest concentration was 127 mg/1 at Stoney Creek. As for Brunette, Eagle, Still, and Silver creeks, their median suspended sediment concentrations were an order of magnitude less than that of Ramsay Creek. And the variations among sampled storm events were considerably smaller. The median suspended sediment concentration of Still Creek, which attributes half the flow to Burnaby Lake, had approximately twice the concentration of Brunette River. This is an indication of sediment trap in Burnaby Lake, which will be discussed later in this chapter. Brunette Eagle Still Ramsay Stoney Silver Sub-Watersheds Figure 4.2 Suspended solids concentration in stormwater runoff from Brunette sub-basins The measurement of stream suspended solids concentration in relation to land use is indicated by the Spearman rank correlation. As listed in Table 4.4, the suspended solids concentration had a strong negative correlation with catchment area, imperviousness, and traffic density. A possible explanation is that the storm flow dilution factor has a stronger influence over the suspended solids concentration than the sources of non-point pollution. During a storm event, the sub-basin with large catchment area and high imperviousness tends to generate much higher storm flow than the small catch basin. The high stream flow dilutes the suspended sediment concentrations, which could easily overcome the effect of sediment non-point source pollution. There were considerable variations in the suspended sediment concentration among the sampled storm events due to different storm intensities. 43 Table 4.4 Spearman rank correlation for suspended solids concentration, imperviousness, and traffic density in the Brunette River watershed Suspended Solids Catchment Imperviousness Traffic Concentration Area Density Suspended Solids Concentration 1.00 -0.90 -0.90 -0.90 Catchment Area -0.90 1.00 1.00 0.80 Imperviousness -0.90 1.00 1.00 0.80 Traffic Density -0.90 0.80 0.80 1.00 4.1.2.2 Suspended Solids Loading and Sediment Budget in the Burnaby Lake The suspended sediment loadings in the Brunette River watershed are presented in Table 4.5. The wide range in suspended sediment loadings is striking. Brunette River had the highest loading of 3214 t/yr. The loading of Stoney Creek accounted for 87% of Brunette River, given that it is expected to be the largest contributor of sediment to the river. Still Creek, with sediment loading of 1102 t/yr is the largest contributor of sediment to Burnaby Lake, which is reasonable given its large drainage area. Eagle and Silver creeks contributed an order of magnitude less sediment loading due to their smaller drainage areas and lower sediment yield rates. Ramsay Creek, on the other hand, produced 20% as much suspended sediment as Still Creek from a watershed of less than 4% the area. 44 Table 4.5 Suspended sediment loading and budget for the Brunette sub-basins in 2003 Suspended Stream Discharge Inflow vs. Sediment Sampling Station , , Sediment Loading Name ( l O W / y r ) Outflow (t/yr) Budget (t/yr) (t/yr) At Winston Street Eagle Creek 4.0 93 At Douglas Road Still Creek 24.3 1102 1406 At BC Hydro Access Road Ramsay Creek 0.9 211 At Winston Street Silver Creek 1.5 70 2870 1062 At Government Street Stoney Creek 7.7 2800 At Lower Stop-Logs Brunette River 79.1 3214 3214 (1) The sediment budget data is derived from NHC monitored flow data. (2) Silver and Stoney Creeks bypass the Burnaby Lake and contribute sediment to the down stream Brunette River directly. To give insight into the sedimentation situation of Burnaby Lake, the sediment budget was calculated. By subtracting the suspended sediment loading of downstream Brunette River from the sum of suspended sediment loadings of the input streams, 1062 tonnes of sedimentation was trapped by Burnaby Lake in 2003 (Table 4.5). Burnaby Lake appears as a sediment trap since 25% less suspended solids were measured at Brunette River than from the upstream rivers. Most of the sediment loads could be attributed to the development surrounding Still Creek and the stream-bank erosion caused by high flows during storms and. Excessive sedimentation is a significant source of contaminants to Burnaby Lake, which leads to water quality deterioration and endangers aquatic life (Sekela et al, 1998). Spearman rank correlation coefficients were presented in Table 4.6 to reflect the effect of land use on the suspended sediment transportation. The suspended sediment loading had a strong positive correlation with the catchment area and imperviousness. And it had a weak positive correlation with traffic density. Thus, imperviousness is identified as a major influence on elevated stream suspended sediment levels. The increasing impervious surface area throughout Brunette River watershed has 45 provided an effective environment to collect and accumulate solids from atmospheric deposition and vehicular traffic. Table 4.6 Spearman rank correlation matrix for suspended solids loading, imperviousness, and traffic density in the Brunette River watershed Suspended Solids Catchment Imperviousness Traffic Loading Area Density Suspended Solids Loading 1.00 0.80 0.80 0.30 Catchment Area 0.80 1.00 1.00 0.80 Imperviousness 0.80 1.00 1.00 0.80 Traffic Density 0.30 0.80 0.80 1.00 4.1.3 Suspended Metals and Their Relation to Land Use This study compares the 2003 suspended sediment associated trace metal level in the Brunette River watershed with historical levels in 1973 and 1993. Increases in residential land use, imperviousness, and traffic density are suspected to be major contributors to the increasing level of trace metal throughout the watershed. A secondary purpose of this study is to evaluate the risk to the aquatic environment by comparing 2003 trace metal levels in the stream sediment to Canadian sediment guidelines. The sediment sampling details, including the sampling location, period of sampling, number of storm events and sample intervals, are listed in Table D - l in Appendix D. Figure 4.3 is a map showing the sampling stations within the Brunette River watershed. 46 Still Creek: at D o u g l a s Ro-ad KOarrwsters Eagle Creek: at W i n s t o n Street Si lver Creek : at Wins ton Street Stehev Creek: at Gove'romant Street Brunette RiVen a t . t o w e r s t o j H o g struetijre Ramsay C r e & f c B C . H y d r o A c c e s s Road Figure4,3 Locations of suspended sediment sampling stations in the Brunette River watershed 4.1.3.1 Suspended Metal Concentration Trace metal concentrations from the six sub-basins in the Brunette River watershed were calculated in units of ug / l and mg/kg. These two units convey the trace metal concentration in different ways. The ug/1 stands for the amount of suspended metal per unit volume of stream flow. Flow rate, which is affected by rainfall intensity, plays an important role in trace metal transportation. For a tributary with high imperviousness and a large catchment area, a sudden rise in the flow rate from a storm event can significantly increase the stream sediment loading by bringing in street dirt and causing bank erosion. However, the trace metal concentration in ug/1 may decrease due to the dilution effect of the increasing flow volume. On the other hand, mg/kg illustrates the amount of metal per unit weight of suspended sediment. Suspended Metal Concentration in ug/1 Figures 4.4 to 4.7 compare the suspended metal concentrations in ug/1 from six sub-basins. 80 I : 60 • -* O 3- 40-Brunette Eagle Still Ramsay Stoney Silver Sub-Watershed Figure 4.4 Suspended C u (fig/1) in stormwater runoff from the Brunette sub-basins 48 1200 1000 300 6i) 3 600H 400 200 Brunette Eagle Still Ramsay Stoney Silver Sub-Watershed ure 4.5 Suspended M n (trg/1) in stormwater runoff from the Brunette sub-basins 40000 Sub-Watershed mre 4.6 Suspended Fe (ug/1) in stormwater runoff from the Brunette sub-basins 49 160 140 120 100 M l 3 80 c N 60i 40 20 O O Brunette Eagle Still Ramsay Stoney Silver Sub-Watershed Figure 4.7 Suspended Z n (ug/1) in stormwater runoff from the Brunette sub-basins The median suspended metal (Cu, Mn, Fe, Zn) concentrations (on a ug/1 base) were highest in Ramsay Creek with large variation among the sampled storm events. The metals Cu and Zn were more evenly distributed through the watershed than Mn and Fe. Although the median Cu concentration in Stoney Creek was at the same level as Ramsay Creek, the mean value was visibly lower. Cu levels were fairly high in Still Creek (10.2 ug/1) with high imperviousness and in Brunette River (8.6 ug/1) with a large drainage area. In the less developed Eagle and Silver creeks, Cu was at trace levels (< 5 ug/1). Zn followed the same pattern of distribution as Cu, with the exception of an extremely low concentration in Brunette River (6.9 ug/1). As for M n and Fe, the median concentration from Ramsay Creek was a magnitude higher than other streams. Although Eagle and Still sub-watersheds had different land use characteristics, their Mn and Fe concentrations were at about the same levels (53-64 ug/1 and 1085-1258 Among the measured suspended metals, Fe had the highest concentration in suspended sediment on a scale of 104 ug/1. The Mn level was two orders of magnitude smaller than Fe. Cu and Zn concentrations were about three orders of magnitude smaller than Fe. These results are found reasonable since Fe and Mn occur naturally. Fe is the fourth most abundant element in the earth's crust, and M n 0 2 makes up about 0.14% of the earth's crust (http://chemistry.about.com/library/blfe.htm). In addition, Mn is 50 abundant in the soils of Burnaby Mountain and Ramsay watershed (Hall et al, 1999). Cu and Zn are generated largely from anthropogenic activities, which makes Cu and Zn better indicators of urban non-point source pollution. Spearman rank correlation coefficients are listed in Table 4.7 to 4.9 to reflect the effect of land use on suspended metal concentrations. The suspended Mn, Fe and Zn concentrations show strong negative correlations with drainage area, imperviousness and traffic density. And Cu exhibits a weak negative correlation with these land use features. Table 4.7 Spearman rank correlation matrix for catchment area and suspended metals (ug/1) from the Brunette River watershed Catchment Area C u M n Fe Z n Catchment Area 1.00 C u -0.3 1.00 M n -0.9 0.1 1.00 Fe -0.8 0.3 0.9 1.00 Z n -0.9 0.5 0.8 0.9 1.00 Table 4.8 Spearman rank correlation matrix for imperviousness and suspended metals (ug/1) from the Brunette River watershed Imperviousness C u M n Fe Z n Imperviousness 1.00 C u -0.3 1.00 M n -0.9 0.1 1.00 Fe -0.8 0.3 0.9 1.00 Z n -0.9 0.5 0.8 0.9 1.00 51 Table 4.9 Spearman rank correlation matrix for traffic density and suspended metals (ug/1) from the Brunette River watershed Traffic Density C u M n Fe Z n Traffic Density 1.00 C u -0.1 1.00 M n -0.7 0.1 1.00 Fe -0.9 0.3 0.9 1.00 Z n -0.8 0.5 0.8 0.9 1.00 Since the suspended solids concentration is negatively correlated with land use, it is reasonable that the suspended solids associated metal follow the same pattern. During a storm event, a sub-basin with large catchment area and high imperviousness tends to generate much higher storm flow than small catch basin. The high stream flow dilutes the suspended sediment associated metal concentration, which can easily overcome the effect of trace metal non-point source pollution. A comparison has been made based on the available suspended metal data in this study and the previous study by Macdonald in 1994 to 1995 in Figures 4.8 to 4.10. 200 Eagle Figure 4.8 Temporal changes of suspended Cu concentration (ug/1) in Eagle and Still creeks 52 1 500 1200 O 900 OB 3 600 300 Eagle Still Figure 4.9 Temporal changes of suspended Mn concentration (ug/1) in Eagle and Still creeks Eagle Still Figure 4.10 Temporal changes of suspended Zn concentration (ug/1) in Eagle and Still creeks 53 The suspended metal (Cu, Mn, Zn) levels in 2003 were significantly lower than levels found in the 1994/95 study. In Eagle Creek, the 2003 median suspended Cu, Mn, Zn concentrations accounted for 40, 50, and 47% of the 1994/95 concentrations, respectively. And the median suspended Cu, Mn, Zn concentrations in Still Creek accounted for 33, 60 and 25% of the concentrations from 1994/95. Since seasonal effects have been verified to influence suspended sediment quality, the sampled storm events were categorized into dry and wet seasons in Table 4.10 Table 4.10 Numbers of sampled storm events in dry/wet seasons in 1994/95 and 2003 Number of Sampled Storm Events Year Dry Season Wet Season (Apri l to September) (October to March) 1994/95 7 5 2003 2 5 In the 1995/94 study, suspended metal samples were collected during the dry season 7 out of 12 times, whereas in the 2003 study, samples were collected only 2 of 7 times during the dry season. The dry season is generally associated with longer antecedent dry periods and higher suspended sediment concentrations. It is reasonable that the 1994/95 suspended sediment had a higher trace metal content than in 2003. Suspended Metal Concentration- in mg/kg The particle associated trace metals in concentrations of mg/kg are listed in Figures 4.11 to Figure 4.14. The metal concentrations varied widely among sub-basins. The median Cu and Zn concentrations were highest in the highly industrialized Still Creek, which has a high impervious area and high traffic density. The downstream reaches of Brunette River had median Cu and Zn concentrations 24 and 26% lower than Still Creek. For Eagle Creek, which collects runoff from less urbanized Burnaby Mountain, Cu and Zn concentrations were 62 and 26% lower than Still Creek. On the other hand, M n and Fe levels were highest in the Eagle Creek drainage area, which is most likely attributed to the erosion of Mn/Fe enriched soil from green spaces. Exceptionally, suspended Mn appeared higher in Brunette River than in Still Creek. This may be related to the release of manganese from anoxic lake sediments (Hall et al., 1999). Ramsay and Stoney creeks had the suspended metal concentrations at a fairly low level. 54 2000 1500 3 u 1000 5001 Brunette Eagle Still Ramsay Stoney Silver . Sub-Watershed ure 4.11 Suspended Cu (mg/kg) in the stormwater runoff from the Brunette sub-basins 30000 250001 20000 1 00 g 15000 s 10000 5000 Brunette Eagle Still Ramsay Stoney Silver Sub-Watershed ure 4.12 Suspended Mn (mg/kg) in the stormwater runoff from the Brunette sub-basins 55 200000 150000 1 100000 50000 0 Brunette Eagle Still Ramsay Stoney Silver Sub-Watershed Figure 4.13 Suspended Fe (mg/kg) in the stormwater runoff from the Brunette sub-basins 3000 2500 2000 1 c N 1500 1000 500 Brunette Eagle Still Ramsay Stoney Silver Sub-Watershed Figure 4.14 Suspended Zn (mg/kg) in the stormwater runoff from the Brunette sub-basins 56 Spearman rank correlations (Table 4.11 to 4.13) show that suspended metal contamination in mg/kg is strongly and positively correlated with traffic density, and weakly and positively correlated with imperviousness and catchment area. Table 4.11 Spearman rank correlation matrix for catchment area and suspended metals (mg/kg) from the Brunette River watershed Catchment Area C u M n Fe Z n Catchment Area 1.00 C u 0.6 1.00 M n 0.5 0.8 1.00 Fe 0.5 0.8 1.00 1.00 Z n 0.5 0.8 1.00 1.00 1.00 Table 4.12 Spearman rank correlation matrix for imperviousness and suspended metals (mg/kg) from the Brunette River watershed Imperviousness C u M n Fe Z n Imperviousness 1.00 C u 0.6 1.00 M n 0.5 0.8 1.00 Fe 0.5 0.8 1.00 1.00 Z n 0.5 0.8 1.00 1.00 1.00 Table 4.13 Spearman rank correlation matrix for traffic density and suspended metals (mg/kg) from the Brunette River watershed Traffic Density C u M n Fe Z n Traffic Density 1.00 C u 1.00 1.00 M n 0.8 0.8 1.00 Fe 0.8 0.8 1.00 1.00 Z n 0.8 0.8 1.00 1.00 1.00 57 4.1.3.2 Suspended Metal Loading The annual suspended metal loading in kg/year provides an estimation of the magnitude of pollutants accumulated in the sub-basins of the Brunette River watershed. The annotations in Figure 4.15 indicate the sub-catchment area in square kilometers. 3000 • 2700 • 2400 • a 2100" D >, 60 1800 • *-—' 60 1500 • _c -a 1200 • 3 Cu 900 • 600 • 300 • 0 63 5.9 O.i 1 .7 22.1 Brunette Eagle Ramsay Silver Still Stoney Sub-Watershed Figure 4.15 Suspended Cu loading (kg/yr) in stormwater runoff from the Brunette sub-basins 58 25000 20000 as u ~Sb 1 5 0 0 0 an c 3 c 10000 5000 Brunette Eagle Ramsay Silver Still Stoney Sub-Watershed Figure 4.16 Suspended Mn loading (kg/yr) in stormwater runoff from the Brunette sub-basins 600000 500000 £ 400000-M 300000-_c -a C3 Brunette Eagle Ramsay Silver Still Stoney Sub-Watershed Figure 4.17 Suspended Fe loading (kg/yr) in stormwater runoff from the Brunette sub-basins 59 4000 3500 i 3000 2500 60 2000 c -a cd ^ 1500 C N 500 Brunette Eagle Ramsay Silver Still Ston ey Sub-Watershed Figure 4.18 Suspended Zn loading (kg/yr) in stormwater runoff from the Brunette sub-basins In Figures 4.15 through 4.18, the trace metal (Cu, Mn, Fe, and Zn) loadings from Brunette River sub-basins were the highest, with large variations among sampled storm events. Still Creek had the second highest trace metal loading, followed by Stoney Creek. The loadings in Ramsay and Silver Creek were quite low, with small variations. Spearman rank correlation coefficients are tabulated below to reflect the effect of land use on trace metal loadings. A perfect positive correlation is found existing between all metal loadings and the imperviousness and catchment area (Table 4.14 to 4.15). The correlation between metal loading and traffic density is strongly positive. (Table 4.16). Table 4.14 Spearman rank correlation matrix for imperviousness and suspended metal annual loading (kg/yr) from different drainage areas of the Brunette River watershed Imperviousness C u M n Fe Z n Imperviousness 1.00 C u 1.00 1.00 M n 1.00 1.00 1.00 Fe 1.00 1.00 1.00 1.00 Z n 1.00 1.00 1.00 1.00 1.00 60 Table 4.15 Spearman rank correlation matrix for catchment area and suspended metal annual loading (kg/yr) from different drainage areas of the Brunette River watershed Area C u M n Fe Z n Area 1.00 C u 1.00 1.00 M n 1.00 1.00 1.00 Fe 1.00 1.00 1.00 1.00 Z n 1.00 1.00 1.00 1.00 1.00 Table 4.16 Spearman rank correlation matrix for traffic density and suspended metal annual loading (kg/yr) from different drainage areas of the Brunette River watershed Traffic Density C u M n Fe Z n Traffic Density 1.00 C u 0.6 1.00 M n 0.6 1.00 1.00 Fe 0.6 1.00 1.00 1.00 Z n 0.6 1.00 1.00 1.00 1.00 It is not surprising that Brunette River had the highest trace metal loadings given the large area (63 1cm2) of its sub-basin to generate and catch pollutants. In addition, the higher the imperviousness, the more pollutants that could be generated per unit area, thus leading to high trace metal loading. 4.1.3.3 Suspended Metal Export Coefficients Normalizing the annual loading rate by catchment area provides an areal loading rate referred to as an export coefficient (EC). E C is a runoff weighted and areally normalized presentation of concentration data. Instead of demonstrating the total amount of trace metals drained from the watershed, annual E C reflects the distribution of trace metals per unit area. Units of E C value are expressed in g/ha/yr. Compared to trace metal loading in kg/yr, the differences in EC values among the sub-basins are demagnetized by normalizing catchment area. As illustrated in Figure 4.19 to 4.22, trace metals showed different patterns of distribution in the sub-basins. The annotations in Figure 4.16 indicate the total impervious area (TIA) for each sub-basin. Highest exports of Cu and Zn were found in Stoney Creek sub-basin. The reason for this is not apparent, though there is a possibility of some point sources within the non-industrial area. The E C levels of Cu in Brunette, Still, and Ramsay creek sub-basins were about 61 15 to 20% lower than those from the Stoney Creek sub-basin. The E C in less developed Eagle and Silver creek sub-basins were two to three times lower than the Still Creek sub-basin. Mn and Fe exports were found to be highest in Ramsay Creek, and their E C levels in other sub-basins were considerable lower. 600 • 3 3% 500 • 47% 63% 3 3% >^  400 • ~c5 300 • 40% 3 4% 200 • 3 100 • Brunette Eagle Ramsay Silver Still Stoney Sub-Watershed Figure 4.19 Suspended Cu E C (g/ha/yr) in stormwater runoff and total impervious area of the Brunette sub-basins 62 7000 t 6000 H 0 5 5000 ^ Brunette Eagle Ramsay Silver Still Stoney Sub-Watershed Figure 4.20 Suspended Mn E C (g/ha/yr) in stormwater runoff from the Brunette sub-basins 300000 250000 a " 200000 -| ~5 " 150000 SD _n -5 ca 3 100000 50000 Brunette Eagle Ramsay Silver Still Stoney Sub-Watershed Figure 4.21 Suspended Fe E C (g/ha/yr) in stormwater runoff from the Brunette sub-basins 63 1000 800 H Brunette Eagle Ramsay Silver Still Stoney Sub-Watershed Figure 4.22 Suspended Zn E C (g/ha/yr) in stormwater runoff from the Brunette sub-basins Statistically, the EC value of Cu is weakly and positively related to imperviousness and traffic density. The E C values of Mn and Fe are strongly and negatively correlated with imperviousness and traffic density (Table 4.17 to 4.18). And for Zn, the E C value is weakly and negatively correlated with all land use factors. Table 4.17 Spearman rank correlation matrix for imperviousness and suspended metal export coefficients (g/ha/yr) from the Brunette River watershed Imperviousness C u M n Fe Z n Imperviousness 1.00 Cu 0.4 1.00 Mn -0.9 -0.2 1.00 Fe -0.5 0.3 0.7 1.00 Zn -0.4 0.3 0.3 0.7 1.00 64 Table 4.18 Spearman rank correlation matrix for traffic density and suspended metal export coefficients (g/ha/yr) from the Brunette River watershed Traffic Density C u M n Fe Z n Traffic Density 1.00 C u 0.3 1.00 M n -0.7 -0.2 1.00 Fe -0.8 0.3 0.7 1.00 Z n -0.5 0.3 0.3 0.7s 1.00 The export coefficients for stream suspended Cu, Mn, and Zn were also been calculated by Macdonald in the 1994/95 storm water study of Still and Eagle creeks. Table 4.19 shows the comparison of average E C values from two studies. Table 4.19 Mean stream suspended metal export coefficients from the Brunette River watershed in 1994/95 and 2003 Average Still Creek Eagle Creek Suspended Metal E C C u M n Z n C u M n Z n 1994/1995 90 940 340 -180* 2050 40 2003 182 621 246 78 1428 217 (1) 1994/1995 Still Creek sampled at Gilmore Station (2) 2003 Still Creek sampled at Douglas Road (3) Units in g/ha/yr * Dissolved value is greater than total value. This is due to variability of the data and the statistical estimation, (see pp. 3-33 of Macdonald et al, 1997) The export coefficients of the toxic metals (Cu and Zn) from the Still Creek sub-basin, which has a high impervious area and high traffic density, were higher than the values calculated for the Eagle Creek sub-basin, which collects runoff from less urbanized Burnaby Mountain. From 1994 to 2003, Cu increased in Still Creek by 100%. This could possibly result from the increasing traffic activity throughout the Brunette River watershed, since a significant source of Cu contamination is from automobile brake pads and fuel consumption. The trend is not available for Eagle Creek due to analytical error (negative value) in Macdonald's (1997) study. Mn E C from Eagle Creek was higher than Still Creek because Eagle Creek has its natural mineral source in ground water and soil. During non-storm 65 periods, ground water seeping into the stream becomes oxidized, causing manganese accumulation in the sediment. During a storm event, this manganese from sediment and soil erosion is carried downstream and increases the suspended Mn E C in Eagle Creek. Despite a slight decrease in Still Creek, Zn export had a significant increase in the Eagle Creek, which might be linked to the increase in automobile traffic. 4.1.4 Sediment Q u a l i t y C h a r a c t e r i z a t i o n 4.1.4.1 Streambed Sediment and Suspended Sediment In addition to suspended sediment, bed sediments from the three sub-basins in the Brunette River watershed were analyzed for particule size and trace metal content. One streambed sediment sample was collected from Ramsay Creek, two from Stoney Creek at different times, and four from Still Creek at different locations. The trace metal contents from the suspended sediment and streambed sediment are compared in box whisker plots in the following figures. Even though samples were taken at different times, each sample is averaging over a long enough time period to be characteristic of its receiving watershed. In each box-whisker plot, the metal concentrations associated with suspended sediments contain the overall data collected during 7 storm events over a 10-month period. Since very limited numbers of bed sediment were sampled, the laboratory replicates were used in the box whisker plots, for the purpose of comparison. 66 1000 800 ^ 600 400 200 • Bed I ' I S i i s p R n d e H Ramsay Still Stoney Sub-Watershed ure 4.23 Suspended/bed sediment comparison for Cu in the Brunette sub-basins • Bed • Suspended Ramsay Still Stoney Sub-Watershed ure 4.24 Suspended/bed sediment comparison for Mn in the Brunette sub-basins 67 80000 60000 ^4 5? 40000 20000 0 X • Bed L_iiSuspended Ramsay Still Stoney Sub-Watershed ure 4.25 Suspended/bed sediment comparison for Fe in the Brunette sub-basins a N 600 • Bed E 3 S u s p e n d e d Ramsay Still Stoney Sub-Watershed ure 4.26 Suspended/bed sediment comparison for Zn in the Brunette sub-basins 68 Figures 4.23 to 4.26 illustrate the trace metal distribution in stream suspended sediment and streambed sediment. Suspended sediments are a bigger concern than the bed sediment not only because they are high in trace metal concentration, but also because they are potentially hazardous to downstream water quality. In Ramsay Creek, the average suspended Cu, Mn, and Zn concentrations were 1.1, 1.3 and 0.9 times higher than the average bedload metals, respectively. In Still Creek, the average suspended Cu, Mn, Fe and Zn concentrations were 5.3, 4.8, 1.0, and 4.9 times higher than the average bedload metals, respectively. And in Stoney Creek, the average suspended Cu, Mn, and Zn concentrations were 0.5, 1.4, 1.0, and 0.3 times higher than the average bedload metals, respectively. Comparing among streams, spatial differences existed in bedload-to-suspended trace metal ratios in the studied sub-basins. Still Creek had a much higher bed-to-suspended ratio than Ramsay Creek and Stoney Creek due to its extremely high suspended metal concentration. This fact is partially attributable to land use, since Still Creek receives greater amounts of contamination from its highly industrialized catchment area. Besides, bank erosion added additional suspended solids into Still Creek due to its high flow dynamic. 4.1.4.2 Particle Size Associated Trace Metal Concentration in Streambed Sediments Figure 4.27 to 4.30 show the partitioning of metals between fine (<63 um) and coarse (63pm to 2 mm) streambed sediments. Generally, trace metals have a tendency to associate with fine particles rather than coarse ones because the relatively high surface area of fine particles provide greater chance for the trace metals to be adsorbed. Cu and Zn had the highest percentage attachment with fine particle sediment from Ramsay Creek. For Stony and Still creeks, the percentage metal concentration associated with fine particles varied from different places and times within the same stream. The different hydraulic conditions and surrounding land uses of different creeks may explain why the trace metal concentration variations existed. 69 Figure 4.27 Distribution of Cu in the streambed sediment from the Brunette sub-basins Figure 4.28 Distribution of M n in the streambed sediment from the Brunette sub-basins 70 £_ a; u. 160000 140000 120000 100000 80000 60000 40000 20000 0 c ^ c ^ t n 1 ^ * * * * o O o - ' - ' P ' - 1 = = = = 4-» 7-H W H J - t-1 — CO CO 2 to co co to Diameter<63um • 63um to 2mm Figure 4.29 Distribution of Fe in the streambed sediment from the Brunette sub-basins 800 > ~ < y > > ~ m > ~ o h rsi n >sf O O 0 ^ £ = = = = S I S r i « ! H hi S « s Diameter<63um B S S u m to 2mm Figure 4.30 Distribution of Zn in the streambed sediment from the Brunette sub-basins Figures 4.31 to Figure 4.34 compare trace metal concentrations in the suspended sediment and the finer fraction (< 63um) of bed sediment in the form of box whisker plots. The purpose of this comparison is to observe stream sediment partitioning between suspended and fine bed sediment. 71 1000 800 oo 600 00 U 400 200 M i ]bed • suspended Ramsay Still Stoney S u b - W a t e r s h e d s Figure 4.31 Suspended and < 63 um bed sediment comparison for Cu in the Brunette sub-basins 3000 2500 20001 oo 1500 1000 500 Ramsay Still Stoney S u b - W a t e r s h e d s • bed CZHsuspended Figure 4.32 Suspended and < 63pm bed sediment comparison for Mn in the Brunette sub-basins 72 80000 60000 40000H 200001 Ramsay Still Stoney S u b - W a t e r s h e d s I Inert I I suspended Figure 4.33 Suspended and < 63pm bed sediment comparison for Fe in the Brunette sub-basins 1500 Ramsay Still S u b - W a t e r s h e d s Stoney • bed LlLJsuspended Figure 4.34 Suspended and < 63um bed sediment comparison for Zn in the Brunette sub-basins The sediment transportation is a constant settle/re-suspension process controlled by flow regime. The suspended sediment settles on the streambed in steady flow, whereas the bed sediment is re-73 suspended by turbulence and scour. In Still Creek, the trace metal (Cu, Mn, Fe, Zn) concentrations in the finer fraction of bed sediment were below those of the suspended sediment. The results suggest that bed sediment re-suspension is caused by highly dynamic flow. The opposite situation was found in Stoney Creek, which experienced a predominance of settling processes. High Mn levels were found in the bed sediment in Stoney Creek. This is due to natural high levels in marine clap deposition in watershed. In Ramsay Creek, bed Cu, Mn and Zn were higher than suspended sediment, and bed Fe was lower than suspended sediment. It is not obvious in Ramsay Creek that which process is in control. 4.1.4.3 Seasonal Effects o f Suspended Sediment Contamination Seasonal variation is analyzed to determine how storm runoff affects the quality of stream flow in the Bruntte River watershed. Figures 4.35 to 4.38 illustrate the average winter (W) and summer (S) suspended metal concentrations from six sub-basins. N H C collected the suspended sediment sampled from December 2002 to October 2003. Figure 4.35 Seasonal variations for suspended Cu in the Brunette sub-basins 74 c 3 0 0 c 2 0 0 + o ° 100 c co co LU co co LU n co E co a. E CO or c o o CO CO co co Figure 4.36 Seasonal variations for suspended Mn in the Brunette sub-basins =§> 14000 c 12000 { •B 10000 CO I £ 8000 -j-8 6000 o 4000 " 2000 0 Figure 4.37 Seasonal variations for suspended Fe in the Brunette sub-basins 60 -r 50 --D C O 40 --~tz 30 --Q) O c 20 --o O c 10 --N — i - T i—a—j—i - -1 | i— i | I.._L.T_L . -i i i i i i—i-,-1 m | | I-I | w » § co co § w ^ § J » 5 ? § J m § » i ^ U J g> jj= 55 E E ° 2 §! > Figure 4.38 Seasonal variations for suspended Zn in the Brunette sub-basins 75 The summer mean stream flow trace metal concentrations were higher than the winter levels in the sub-basins, except for Still Creek. This is likely attributed to the fact that the summer rainfall events are less frequent but more intense (such as thunderstorm events) than winter events, and the longer build up in summer due to longer antecedent dry periods. And thus higher pollutant loads tend to be washed into steams during a storm event in the summer. Moreover, larger concentration variations existed during the summer period. This finding agrees with Macdonald's 1994/95 stormwater seasonal variation study of the Brunette River watershed. Monthly rainfall,data in 2003 were obtained from the Environment Canada website (http://www.climate.weatheroffice.ec.gc.ca /climateData/monthlydatae.html). 300 -250 -200 -150 -100 -50 -n W S / W u - Dec 02 Jan 03 Feb 03 Mar 03 Apr 03 May 03 Jun 03 Jul 03 Aug 03 Sep 03 Oct 03 Depth 140 151 27 134 140 49 13 20 4.1 40 248 Sampling Period Figure 4.39 Monthly rainfall in the Brunette River watershed in 2003 As shown in Figure 4.39, the 2003 total winter rainfall (839 mm) was 6.6 times of the total summer rainfall (126 mm). Statistically, summer rainfall variation is higher than winter, since the coefficient of variation in summer (0.75) is higher than winter's (0.5). This explains the lower winter stormwater contamination and the higher variation in the quality of summer stormwater runoff. 4.1.5 Temporal Changes and the Severity of the Streambed Sediment Contamination 4.1.5.1 Comparison of 2003 and the Historical Contamination Level (1973/1993) The watershed contamination history has been determined by two approaches. The first approach compares current bed sediment trace metal contamination with the data retrieved from nearby locations in the 1973 (Hall et al., 1976) baseline study. Listed in Figures 4.40 to 4.43 are the mean trace metal concentrations for 1973, 1993 (McCallum, 1995) and 2003 stream bed sediment in the fine (< 63pm) stream sediment. The numbers of samples for the 1973, 1993, and 2003 studies is 16, 58, and 37, 76 respectively. 1200 1973 Still 1993 2003 Ramsay 1973 1993 2003 Stoney 1973 1993 2003 968 217 319 18 125 86 20 289 275 Figure 4.40 Mean streambed Cu contamination (mg/kg) change from 1973 to 2003 Still Ramsay Stoney 1973 1993 2003 1973 1993 2003 1973 1993 2003 436 427 711 562 640 559 645 1603 1642 Figure 4.41 Mean streambed Mn contamination (mg/kg) change from 1973 to 2003 77 100000 90000 oi 80000 cn 70000 E 60000 •a a <fl T3 <U CO 50000 40000 30000 20000 10000 0 Still l 'J/3 1993 Ramsay 2003 1973 1993 2003 Stoney 1973 1993 2003 • Cone. 32318 23733 44527 17063 21185 36654 25877 30679 89378 Figure 4.42 Mean streambed Fe contamination (mg/kg) change from 1973 to 2003 at E c N C £ v i/> •a o m 450 400 350 300 250 200 150 100 50 0 1973 Still 1993 2003 Ramsay 1973 Stoney 1993 2003 1973 1993 2003 I Cone. 399 354 372 60 209 107 69 127 211 Figure 4.43 Mean streambed Zn contamination (mg/kg) change from 1973 to 2003 The above figures show the net changes in streambed sediment contamination during the past 30 years in the Brunette River watershed. Each of the three monitored streams showed a unique pollutant transition pattern. In Still Creek, the Cu level had a major decrease (more than 50%) from 1973 to 1993 due to the ceased operation of extensive industry in the sub-basin (McCallum, 1995). Cu levels then increased by 50% from 1993 to 2003 due to the constant input of non-point source pollution. In Ramsay and Stoney creeks, Cu increased more than 10 times from 1973 to 1993, and slightly decreased after that. Zn levels in Still Creek were constantly elevated, with a 11% decrease from 1973 and a 5% increase 78 from 1993. The high Zn concentration could be attributed to the high impervious area and extensive traffic density in the Still Creek sub basin. In Ramsay Creek, Cu, Mn, and Zn reached peak levels in 1993. The streambed trace metal concentrations decreased slightly after 1993, which suggests well-controlled non-point source pollution. As for Stoney Creek, the major increase in Cu (by 13 times) and Mn (by 1.5 times) occurred from 1973 to 1993 and in Fe (by 1.9 times) from 1993 to 2003, whereas Zn increased steadily in the past 30 years. The gasoline addictive, methylcyclopentadienyl manganese tricarbonyl (MMT), is a major anthropogenic source of Mn in the environment, which may have caused a significant increase in the stream bedload sediment after 1973. In the late 1990's, many fuel suppliers started removing M M T from gasoline due to environmental contamination concerns (Loranger et al, 1994, Lyttle et al, 1995, Brault et al, 1994). 4.1.5.2 Comparison to Canadian Sediment Guidelines The second approach of the sediment study evaluates the current risk to the aquatic environment by comparing contaminant levels in stream bed sediment throughout the watershed to recently developed interim Canadian sediment guidelines (Smith et al. 1995). The guidelines define the threshold effects level (TEL), which indicates the possibility of adverse biological effects, and the probable effects level (PEL), which indicates a high probability of biological effects,.for most contaminants of concern in urban runoff. The threshold effects levels (TEL) and probable effects levels (PEL) for Cu and Zn in freshwater bed sediments are marked in the figures below. Since Fe and Mn are abundant in soil, no limitation was proposed. The bars in Figures 4.44 and 4.45 represent the trace metal concentrations extracted from the < 2 mm bed sediment from Stoney Creek, Still Creek, and Ramsay Creek. 79 oi 01 E c o c V u c o u 3 200 150 100 50 PEL Stoney 10.29 TEL Stoney 11.13 J M Mm It; Ramsay 11.20 Still #1 Still #2 Still #3 Still #4 ICu Cone. 116 19 73 75 89 13 Figure 4.44 Comparison of streambed Cu (mg/kg) and criteria effects level in 2003 oi E c o c V u c o u c N 350 300 250 200 150 100 50 0 Wis. 1 Stoney 10.29 PEL TEL • • Stoney 11.13 Ramsay 11.20 Still #1 Still #2 Still #3 Still #4 iZn Cone. 219 48 51 156 127 155 37 Figure 4.45 Comparison of streambed Zn (mg/kg) and criteria effects level in 2003 The 2003 streambed sediment contamination evaluation indicates that the PEL criteria are not exceeded in any of the studied streams in the Brunette River watershed. Cu concentrations exceeded the T E L 4 out of the 7 sampled events, in Still Creek and Stoney Creek. Both temporal and spatial differences exist in the bed sediment trace metal levels. The October 29 Cu and Zn concentrations from Stoney Creek were twice that of the T E L level. However, two weeks later on November 13, the concentration dropped below the T E L . The Still Creek bed sediments were 80 excavated at 4 locations on the same day. Al l Cu concentrations from Still Creek exceeded the T E L , except for dredge 4. The less developed Ramsay Creek had bed sediment quality below the T E L value for both Cu and Zn. The stream contamination situation has not changed much since 1993. Hall et al. (1999) reported that most exceedances of the T E L criteria for Cu and Zn happen in the Still Creek area, but the exceedance is rarely seen in stream sediments downstream of Burnaby Lake. 4.2 Trace Metals in the Stormwater Runoff from the UBC Diesel Bus Loop The catchment of the U B C diesel bus loop has some unique features. First, it is a genera! service loop area with substantial volumes of through flow traffic. Also, there are no large industrial, manufacturing, or warehouse activities around the bus loop and so large truck traffic is minimal and limited to deliveries. The campus bus loop traffic activities result in unique stormwater characteristics. 4.2.1 Suspended Solids Concentration and Spatial Variation 1600 1400 __ 1200 1000 800 » 600 1 400 200 Sample Date Figure 4.46 Suspended solids concentration in stormwater runoff from U B C diesel bus loop Figure 4.46 shows the suspended solids concentrations from 15 monitored storm events dated from October 2004 to June 2005. In each box-whisker, there are suspended solids data from three catch basins 81 in one storm event. The suspended solids concentration varied largely from 10 to 1500 mg/1. A general seasonal effect is observed from the monitored storm events. The average suspended solids concentrations from the wet season (October to March) was 198 mg/1, which is 24% lower compare to the average of 245 mg/1 from the dry season (April to June). During the wet season, the monthly average rainfall amount was twice as much as the dry season. The high frequency of rainfall provided the least time for solids to accumulate. Stormwater quality sampling has been undertaken at several locations in U B C catchments through 2001 to 2003. The project captured storm events from the south campus catchment, which drains to Booming Grounds Creek and ultimately to Musqueam Marsh, the north campus catchment, which drained by a vertical drain ("Spiral Drain") into a deepwater outfall in English Bay, and a reference stream in Pacific Spirit Regional Park (Figure 4.47). Figure 4.47 Campus catchment and sampling locations for the U B C stormwater monitoring program To illustrate the spatial variation in storm runoff, the suspended solids concentration from the U B C diesel bus loop is compared with previous studies conducted on the U B C campus and in the G V R D urban area in Figure 4.48. The G V R D urban area dataset covers Still Creek watershed in Burnaby (Macdonald et al., 1997), Serpentine River in Surrey, and Wagg Creek in North Vancouver (Kavelaars, 1998). 82 250 o 200 -o E •g 100 -c a m D " 50 -0 m South Catchment North Catchment Bus Loop G V R D Urban Area Figure 4.48 Spatial differences in suspended solids concentration in the U B C campus, bus loop and In Figure 4.48, the event average suspended solids concentration from U B C bus loop was about 5 times higher than the average concentration gathered from the 2001 campus study and the G V R D urban area. This indicates that the traffic-dominated land use with high imperviousness can significantly increase the suspended solids yield compared to comprehensive urban land use. Since good correlation between total metal concentration and TSS has been found in many studies (Hallberg, 2006 and Sansalone et al., 1995), suspended solids reduction is the most expedient method for stormwater management. 4.2.2 Trace Metal Concentration 4.2.2.1 Total Metal Concentration and Spatial Variation Total metal includes the amount of metal attached to the suspended solidsand the amount of metal dissolved in the stormwater runoff. Each box in the following figures contains the total metal concentrations from three catch basins in a single rainfall event. G V R D areas 83 350 300 250 JT 200 3 o 2 150 1 o H 100 50 1 o o T o * T T I 1 L J /n /n /n ' s 'j> ^ ^ ^ ^ ^ ->> <£-Sample Date Figure 4.49 Total Cu concentrations in stormwater runoff from the U B C diesel bus loop 1400 Sample Date Figure 4.50 Total M n concentrations in stormwater runoff from the U B C diesel bus loop 84 90000 80000 70000 60000 • 50000 • 40000 • 30000 • 20000 10000 0 O T x _ I t g t * o O r1" <r> /n yn <j ^ ^- ^ ^ ^ ^ ^ <£-Sample Date Figure 4.51 Total Fe concentrations in stormwater runoff from the U B C diesel bus loop yn /n /n / , 'j> " i - ^ °^ ^ ^ ^ °> y 57 57 57 Sample Date Figure 4.52 Total Zn concentrations in stormwater runoff from the U B C diesel bus loop 85 Based on Figures 4.49 to 4.52, the inter-storm variation was high. Total metal (Cu, Mn, Fe, Zn) concentrations were lower at the end of 2004, then started to increase in the early months of 2005. There are two events, which were sampled on February 28, 2005, and May 2, that stand out with extremely high total metal concentrations. This result was caused by long antecedent dry periods, 17 days and 15 days respectively, compared to the average antecedent dry period of 5 days within the monitoring period. A seasonal effect was also found in the distribution of total metal concentrations. The concentration in November and December were the lowest due to high monthly rainfalls of 200 mm and 188 mm, respectively, compared to the average monthly rainfall of 124 mm within the monitoring period. To illustrate the spatial variation in stormwater quality, the total metal concentrations from the U B C bus loop are compared with previous studies (2001 to 2003) conducted at the U B C north/south campus catchment and the G V R D (Macdonald, 2005) urban area in Figures 4.53 to 4.56. South North Catchment Catchment Bus Loop GVRD Urban Area Figure 4.53 Spatial difference in total Cu concentration in the U B C campus, bus loop and G V R D areas As shown in Figure 4.53, the average total copper concentrations from the four locations were highly variable. The bus loop total copper concentration was about twice the average concentration of the north catchment and 3.5 times the south catchment. The G V R D regional Cu level was halfway between the north/south campus catchment. The non-point sources of copper contamination are largely related to bus activities, such as brake ware and corrosion (Reybekiel, 2001). The high intensity and frequent stops by diesel buses is likely the cause of the elevated Cu levels in the U B C diesel bus loop. 86 250 South North Bus Loop GVRD Urban Catchment Catchment Area Figure 4.54 Spatial difference in total M n concentration in the U B C campus, bus loop and G V R D areas As shown in Figure 4.54, the bus loop Mn concentration was the highest among the studied areas. Unlike Cu, Mn concentrations from the south catchment and G V R D urban areas were close to the bus loop level. Natural and anthropogenic sources both contribute to Mn contamination. In the south catchment and G V R D urban areas, Mn could be associated with groundwater, soils and the gasoline additive methylcyclopentadienyl manganese tricarbonyl (MMT), which generally replaced lead-based additives as an anti-knocking agent. But in the U B C bus loop, with nearly 100% imperviousness, M M T is likely to be the only source of Mn contamination, since manganese dioxide is commonly used as an additive in up-graded fuel to improve diesel engine performance. In addition, other possible sources of M M T contamination are from the parking lot south of the bus loop and atmosphere deposition. 87 South North Bus Loop GVRD Urban Catchment Catchment Area Figure 4.55 Spatial difference in total Fe concentration in the U B C campus, bus loop and G V R D areas Mean concentrations of total Fe are compared in Figure 4.55. The bus loop stormwater runoff had 4, 13, and 9 times greater Fe levels than the U B C south catchment, north catchment, and G V R D urban areas, respectively. High Fe level in the bus loop could be related to the corrosion of metal components from buses. Fe is not usually studied as a toxicant because it occurs so commonly in nature and it is not particularly toxic. 350 South North Catchment Catchment Bus Loop GVRD Urban Area Figure 4.56 Spatial difference in total Zn concentration in the U B C campus, bus loop and G V R D areas 88 Figure 4.56 shows that the average Zn concentrations in the north and south campus catchments and G V R D areas were at the same level, which was about 4 times lower than the average Zn concentration from the bus loop. The extremely high Zn concentration in the U B C bus loop indicates that traffic is a primary source of urban Zn contamination. Johnson and Juengst (1997) found that zinc appeared to be directly correlated with volume of traffic on streets in Wisconsin, with its presence in automobile tires, paints, automotive engine oil, and vehicle exhaust. From the above comparison, the contribution of Fe from diesel buses is the highest, followed by Zn, Mn and Cu. This result is supported by results from Wang et al. (2003). Wang conducted vehicle engine exhaust tests under transient-cycle conditions and found that the emitting concentrations of crust 0 elements (such as Fe) were much higher than those of anthropogenic elements (Cu, Mn and Zn) from diesel engine exhaust. In addition, the emission rate of the metal contents in vehicle exhaust is 0.99 linear regression related to the consumption rates of metal contents. Table 4.20 shows the emission concentrations (C e) and emission rates (ER) of the metals Cu, Mn, Fe and Zn in diesel engine exhaust, the concentrations of these metal contents in diesel fuel (C f ) and their consumption rates (CR) for the tested diesel engine under transient-cycle testing conditions. Table 4.20 Emission concentrations and emission rates o f the metals contents in the diesel engine exhaust Elements Engine Exhaust Diesel Fuel C e (ug/m 3 ) E R (ug/min) Cfftig/1) C R (ug/min) Cu 55.4 396 2780 417 Mn 21.0 150 1040 156 Fe 543 3880 27800 4180 Zn 111 794 5630 845 (1) The engine is Mitsubishi-6D 14-2AT, manufactured in 1990 (2) The concentrations of metal contents contained in the diesel fuel used in this study is Taiwan standard and could be different from those used in other countries. As a result, the emission of metal contents in vehicle exhaust could mostly be attributed to the consumption of metal contents in diesel fuel. The high metal contents found in diesel vehicle exhaust suggest that the management of metal contents in diesel fuel would be of great importance for eliminating metal emissions from diesel engines in the future. 89 4.2.2.2 Dissolved Metal As the most hazardous part of stormwater runoff, dissolved metal concentration was measured from the stormwater runoff in the U B C bus loop. Comparing Figures 4.57 through 4.60, dissolved metal concentrations were highly variable during the monitored period. The average concentrations of dissolved metal in the dry season (April to June) were 1.5, 3.8, 3.1, and 3.8 times higher than the average concentrations in the wet season (October to March) for Cu, Mn, Fe, and Zn, respectively. The dissolved Cu, Mn, and Zn on February 28 were exceptionally high compared to the other storm events in the wet season after a long period of dry weather. 240 Sample Date Figure 4.57 Dissolved Cu concentrations in stormwater runoff from the U B C diesel bus loop 90 550 500 450 400 « 350 I 3 0 0 -a £ 250 \ 200 5 150 " 100 • 50 • 0 /n /n yn yy yj> ^ J^ ^ ^ ^ v^ W ^ V " . ^  ^ ^  ^ ' S a m p l e Date Figure 4.58 Dissolved Mn concentrations in stormwater runoff from the U B C diesel bus loop 8000 6000 4000 H 2000 o ° T o & a g s & <o X^ ^- - i - ^ ^ ^ ^ °> <2- ^ e ^ V ^ ^ ^ Sample Date Figure 4.59 Dissolved Fe concentrations in stormwater runoff from the U B C diesel bus loop 91 I 800 c N 1500 H 1200 i 900 n 600 300 ' <r> 'n yn <3 ^ ^ ^ ^ ^ ^ °> ^ -v <, ^ ^ ^ ^ *"* ^ ^ ^ ^ ^ Sample Date Figure 4.60 Dissolved Zn concentrations in stormwater runoff from the U B C diesel bus loop 4.2.2.3 Suspended M e t a l In stormwater runoff, total metals are transported in dissolved form and in particulate form associated with suspended solids. In this study, the concentration of suspended metal was calculated by subtracting the dissolved metal concentration from the total metal concentration. As illustrated in Figures 4.61 through 4.64, there is no obvious trend in suspended metal levels that can be related to season. There were a few times when "dissolved" metal was estimated to exceed "total" metal concentrations. This is a physical impossibility and could be due to the small sample size, methodological processes and handling errors, such as a contaminated digestion flask or spill from the overheated digester. In most cases, especially for the samples collected during the first several hours of the rain, the concentration of dissolved metal was found to be higher than that of total metal. This phenomenon suggests that the dissolved metal level is very close to the total level. Thus, during the early period of rainfall, the dissolved metal dominates the metal partition, and the amount of suspended metal could be trivial. 92 600 500 400 O •o 300 200 100 O O T 6J X T /n /n yn y , <l ^ ^ *^  -E- ^ <£-Sample Date Figure 4.61 Suspended Cu concentration in stormwater runoff from the U B C diesel bus loop 2000 1800 1600 1 Sample Date Figure 4.62 Suspended Mn concentration in stormwater runoff from the U B C diesel bus loop 93 100000 90000 80000 1 70000 60000 50000 40000 30000 20000 • 10000 • 0 o o I JL O O Ml X I yn /n /n / y <j> ^ ^ °^ 3s Sample Date ure 4.63 Suspended Fe concentration in stormwater runoff from the U B C diesel bus loop 1600 1400 1200 3 1000 \ N •o 800 H c a. 600 400 200 * "T I 3 fk - 4 ^ ^ <o ^ ^ °^ >C- °> °> °> °> ^ ^ ^ e- - 3 -7 O ^ O Sample Date ure 4.64 Suspended Zn concentration in stormwater runoff from the U B C diesel bus loop 94 4.2.3 Stormwater Quality Characterization 4.2.3.1 First Flush First flush is defined as the first portion of storm water runoff containing the main proportion of the pollutant load. Many research papers have reported first flush as a typical phenomenon in a small watershed with high imperviousness. A stormwater study in Sweden showed that 50-60% of the total mass of Zn, Cu, Pb, "Ni, Co, Cd was transported with the first 30% of runoff (Lind et al., 2001). Sansalone and Buchberger (1997) found that dissolved Cu, Zn fractions exhibited a strong first flush in pavement flow in Cincinnati. In this study, the first flush phenomenon was observed in 70% of the monitored rainfall events. However, in a few events, especially with discontinued rainfall, first flush was not obvious. Figures 4.65 to 4.78 illustrate the change in metal concentration over a single storm event as an example of first flush in this study. The storm event on October 29, 2004, was selected, since the sampling period lasted for 8 hours, which was sufficiently long to cover the first flush, peak flow and the tailing clean runoff. The rainfall over the storm event is not shown in the figures due to the absence of hourly rainfall data. 95 180 o o o o LO LD LD O o n o ro i - l O ro 66 66 66 66 Cri O y-t i-t T-t TH •iH (N (N Total Cu * -Dissolved Cu ure 4.65 Total and dissolved Cu pollution pattern in stormwater runoff from the U B C diesel bus loop over a rainfall event on Oct 29, 2004 ure 4.66 Total and dissolved M n pollution pattern in stormwater runoff from the U B C diesel bus loop over a rainfall event on Oct 29, 2004 96 25000 "i o o o O m m m O o ro o n TH O m m LD 00 6i 66 o i-H i-H (N IN Total Fe * -Dissolved Fe , ure 4.67 Total and dissolved Fe pollution pattern in stormwater runoff from the U B C diesel bus loop over a rainfall event on Oct 29, 2004 97 During the storm event, metal (Cu, Mn, Fe, Zn) concentrations first increased quickly until they reached a peak value by the second hour. They then gradually declined over the course of the storm event. When the amount of surface runoff increased at the beginning of rainfall, the particulate bounded metals as well as the dissolved metals, which accumulated on the road surface for days, were picked up by the first flush. The higher the surface flow, the more metals that are washed into the stormwater collection system. The timing of the first flush could be influenced by rainfall intensity and the antecedent dry period. First flush is likely to appear earlier in a heavy storm with high surface flow. The first flush phenomenon was also verified by the turbidity and conductivity data from the same storm event. Turbidity and conductivity reached peak values at the same time, which shows a greater emission rate of pollutants within the first two hours of the rainfall event (Figure 4.69). 1200 1000 800 4 600 "O I- 400 200 --\ v \ i i i i i X v 300 250 200 § 150 -I o 100 § o 50 13:00 13:30 15:00 18:30 19:15 19:45 20:15 21:00 » Turbidity * -Conductivity Figure 4.69 Turbidity and conductivity change in stormwater runoff from the U B C diesel bus loop on Oct 29, 2004 4.2.3.2 Dissolved Metal to Suspended Metal Ratio Generally speaking, the amount of metal associated with the suspended particulates in the U B C bus loop stormwater runoff was greater than the amount of metal in dissolved form. The ratios of dissolved/suspended metal concentration to total metal concentration varied in metal species and sampling locations. The percentages marked in Figures 4.70 to 4.73 represent the percentage of dissolved metal and the percentage of suspended metal in the total metal concentration. Each box contains the metal concentrations sampled from a single catch basin through all monitored storm events. 98 300 250 1 200 H 6 0 3 150 100 50 40% 60% nr 35% 65% j ' imnyn'1 33% 67% ^ D i s s o l v e d i__JSuspended Catch Basin ure 4.70 Dissolved and suspended Cu distribution in stormwater runoff from catch basins in the U B C diesel bus loop 1000 500 CT- 600 3 400 i 200 25% 75% • 1 6% 84% 20% 80% - J Disso lved • S u s p e n d e d 1 2 Catch Basin ure 4.71 Dissolved and suspended M n distribution in stormwater runoff from catch basins in the U B C diesel bus loop 99 50000 40000 30000 DO 3 20000 10000 5% 95% 3% 97% 1 2 Catch Basin 4% 96% Dissolved S u s p e n d e d Figure 4.72 Dissolved and suspended Fe distribution in stormwater runoff from catch basins in the U B C diesel bus loop 1000 • 800 600 c N 400 200 48% 52% 4 1 % 5 9% 3 8% 62% Dissolved S u s p e n d e d Catch Basin Figure 4.73 Dissolved and suspended Zn distribution in stormwater runoff from catch basins in the U B C diesel bus loop 100 As shown in the above figures, each metal had a unique pattern of distribution between dissolved and suspended froms. In addition, the distribution was found to be slightly different for individual catch basin. For Cu, the dissolved part accounted for 33 to 40% of the total metal in the three monitored catch basin. For Mn, the range was.from 16 to 25%. For Fe, the range was from 3 to 5%. And the range for Zn was from 38 to 48%. The metal's partitioning in dissolved and suspended forms is determined by chemical properties and the composition of stormwater runoff. Fe is predominantly associated with particles in stormwater runoff, as the dissolved fraction formed only a small percentage of the total fraction. Previous studies have reported similar transport and partitioning behavior for Fe (Yousef et al., 1984; Harrison and Wilson, 1985). In contrast, a significant portion of Cu and Zn are found in the dissolved fraction. A number of previous studies have demonstrated that Cu and Zn exhibited an intermediate partitioning behavior between the particulate and dissolved fractions (Morrison et al., 1984; Harrison and Wilson, 1985; Revitt et al., 1990). The metal concentrations could vary largely from diffeent sites with the partitioning remain similar. Glenn (2002) reported trace metal partitioning in snow water from urban roadways in Cincinnati, Ohio. Total Cu was approximately 150 mg/L with approximately 63% of the Cu mass in dissolved form. Zn was approximately 4,500 mg/L with approximately 89% of the Zn mass in dissolved form. Since the dissolved metals have higher bioavailability and are more difficult to remove by storm-water treatment, Zn and Cu are considered to be more hazardous to the aquatic environment than Fe and Mn. Catch basin 1 had a higher dissolved metal to total metal ratio for Cu, Mn, and Zn than the other two catch basins. The reason may be related to the speed of the diesel buses driving through the different catchment areas. Catch basin 1 is located in the middle of the driveway where buses pass at reduced speeds, while the catchment areas of basins 2 and 3 are mainly for bus loading/unloading purposes. Wang (2003) found that the emission concentration of particulate matter from diesel vehicle engines is inversely proportional to the specified engine speed. To the contrary, the increase in engine speed results in an increasing fraction of particulate metal contents. The frequent stop and start-up of diesel buses from catch basins 2 and 3 is likely to introduce higher levels of particulate metal emission, which would have made the dissolved to total metal ratio lower than catch basin 1. 4.2.3.3 Relation between Dissolved Metal and Conductivity The field data from October 29, 2004, was plotted in Figure 4.74 to 4.77 in order to examine the correlation between dissolved metal and conductivity. A visibly positive correlation can be observed from the Cu, Mn, Fe, Zn concentrations and their conductivities, although other substances in storm water runoff are also contributers to conductivity. 101 13:00 13:30 15:00 18:30 19:15 19:45 20:15 21:00 - - • - -Conduc t i v i t y—*—Disso lved Cu Cone. Figure 4.74 Conductivity and dissolved Cu concentration in stormwater runoff from the U B C diesel bus loop Figure 4.75 Conductivity and dissolved Mn concentration in stormwater runoff from the U B C diesel bus loop 102 13:00 13:30 15:00 18:30 19:15 19:45 20:15 21:00 - - • - -Conductivity • Dissolved Fe Cone. Figure 4.76 Conductivity and dissolved Fe concentration in stormwater runoff from the U B C diesel bus loop 300 140 13:00 13:30 15:00 18:30 19:15 19:45 20:15 21:00 - - • - -Conductivity » Dissolved Zn Cone. Figure 4.77 Conductivity and dissolved Zn concentration in stormwater runoff from the U B C diesel bus loop To verify the correlation between dissolved metal concentration and conductivity, spearman rank correlation coefficients were calculated from the data set sampled on October 29 and listed in Tables 103 4.21 to 4.23. Statistical analysis shows that all the metals are positively correlated with the conductivity except for Zn in catch basin 2. Stronger correlations existed in the catch basin 1 than the other two basins. Table 4.21 Spearman rank correlation matrix for conductivity and dissolved metals from catch basin 1 (n=8) Conductivity Cu Mn Fe Zn Conductivity 1.00 Cu 0.54 1.00 Mn 0.91 0.55 1.00 Fe 0.81 0.41 0.71 1.00 Zn 0.88 0.24 0.88 0.74 1.00 Table 4.22 Spearman rank correlation matrix for conductivity and dissolved metals from catch basin 2 (n=8) Conductivity Cu Mn Fe Zn Conductivity 1.00 Cu 0.78 1.00 Mn 0.40 0.25 1.00 Fe 0.21 -0.73 0.53 Zn -0.48 -0.15 0.72 1.00 0.79 1.00 104 Table 4.23 Spearman rank correlation matrix for conductivity and dissolved metals from catch basin 3 (n=12) Conductivity Cu Mn Fe Zn Conductivity 1.00 Cu 0.38 1.00 Mn 0.54 0.90 1.00 Fe 0.51 0.92 0.93 1.00 Zn 0.58 0.78 0.91 0.87 1.00 The spearman rank correlations are also applied to total metal concentration and suspended solids in Table 4.24 to 4.26. Total metal shows a weak positive correlation to suspended solids in catch basin 1, and a stronger positive correlation in catch basins 2 and 3. Westerlund and Viklander also found that total suspended solids were highly correlated with total concentrations of Cu and Zn from snowmelt in 2006. As a result, the removal of suspended solids from storm water runoff could be an efficient way to reduce heavy metal pollution. 105 Table 4.24 Spearman rank correlation matrix for suspended solids and total metals from catch basin 1 (n=8) Suspended Solids Cu Mn Fe Zn Suspended Solids 1.00 Cu 0.48 1.00 Mn 0.38 0.93 1.00 Fe 0.5 0.91 0.93 1.00 Zn 0.19 0.71 0.74 0.62 1.00 Table 4.25 Spearman rank correlation matrix for suspended solids and total metals from catch basin 2 (n=8) Turbidity Cu Mn Fe Zn Turbidity 1.00 Cu 0.81 1.00 Mn 0.95 0.79 1.00 Fe 0.95 0.79 1.00 1.00 Zn 0.81 0.64 0.91 0.91 1.00 106 Table 4.26 Spearman rank correlation matrix for suspended solids and total metals from catch basin 3 (n=12) Turbidity Cu Mn Fe Zn Turbidity 1.00 Cu 0.88 1.00 Mn 0.92 0.90 1.00 Fe 0.96 0.94 0.97 1.00 Zn 0.84 0.81 0.70 0.75 1.00 4.2.3.4 Relation between Total Metal EMC and Rainfall Amount, Antecedent Dry Period, and Traffic Density Stormwater runoff is a primary source of trace metals in the environment. The potential effects of rainfall amount, antecedent dry period, and traffic density to trace metal concentration in stormwater runoff are investigated in this study. Rainfall levels have been reported to affect stormwater quality in the pollutant wash-off process in the past studies (Gnecco et al., 2005). Figure 4.78 shows the relation between metal event mean concentration (EMC) and rainfall during the sampling period. In most cases, high metal concentrations were associated with low rainfall and vice versa. 107 1200 1000 800 600 400 200 Sample Date I i^i Rainfall Amount - — - Cu — • - M n — A -Fe /100 -Zn Figure 4.78 Relation between rainfall and total metal E M C in stormwater runoff from the U B C diesel bus loop The above observation is verified by statistical analysis, since the total metal E M C for Cu, Mn, Fe and Zn all showed a weak negative correlation with amount of rainfall (Table 4.27). In another parking lot runoff study using simulated rainfall, the intensity and duration of rainfall were inversely related to degree of toxicity, which was primarily caused by dissolved Zn (Greenstein et al, 2004). Table 4.27 Spearman rank correlation matrix for rainfall and total metal E M C in stormwater runoff from the U B C bus loop Rainfall Cu M n Fe Zn Rainfall Cu Mn Fe Zn 1.00 -0.40 -0.32 -0.22 -0.30 1.00 0.85 1.00 0.15 0.39 1.00 0.90 0.84 0.13 1.00 108 Antecedent dry period, which controls the pollutant building-up process, is a crucial factor affecting the level of stormwater runoff contamination. As shown in Figure 4.79, longer antecedent periods between storms are associated with the presence of increased metal concentration in stormwater runoff. During precipitation, surface runoff acts as a rinsing agent in removing dry deposition from the road surface. Given a sufficient amount of rainfall, all pollutants, suspended or dissolved, could be washed off the road surface and into the stormwater drainage system. Figure 4.78 reveals a strong correlation between antecedent dry period and trace metal E M C . 1200.00 CM CM CM CM Sample Data a A n t e c e d e n t Dry D a y s Cu - M n • F e / 1 0 0 - • Z n Figure 4.79 Relation between antecedent dry days and total metal E M C in stormwater runoff from the U B C diesel bus loop From the spearman rank analysis, total metal E M C shows a positive correlation with antecedent dry days. As calculated in Table 4.28, the E M C for Cu, Mn, and Zn were strongly correlated with the antecedent dry period. And the E M C for Fe was weakly correlated with the antecedent dry period. A highly significant correlation was also found between the amount of dissolved Cu and the length of antecedent dry period by Hewitt and Rashed's (1992) study of rural highway runoff. 109 Table 4.28 Spearman rank correlation matrix for antecedent dry days and total metal E M C in stormwater runoff from the U B C bus loop Antecedent Dry Days Cu Mn Fe Zn Antecedent Dry Days 1.00 Cu 0.76 1.00 Mn 0.71 0.85 1.00 Fe 0.13 0.15 0.39 1.00 Zn 0.73 0.90 0.84 0.13 1.00 A strong seasonal effect could also be observed in Figure 4.79. In the dry season (April to June) the total metal event mean concentrations were 1.5, 1.3, 2.5 times those of the rain season (October to March) concentrations for Cu, M n , and Zn, respectively. The reason for high pollutant levels in the dry season can be attributed to the longer antecedent dry days, which allows more pollutants to build up on the road surface before the rainfall washes them off. Figure 4.80 shows trace metal E M C and traffic density in the bus loop. Traffic density is defined as the total number of buses arriving/leaving the bus loop during the monitoring period. Traffic density during each sampling date varied. According to the bus schedule, the number of buses in service on weekdays is about 1.5 times higher than on the weekend. Even within the same day, buses operate more frequently during rush hours. Traffic density was observed to be proportional to trace metal concentration for only a few times in the wet season. For most storm events, especially in the dry season, no apparent trend was found. 110 1200.00 "5, 1000.00 ~ 800.00 200 c 0) o c o O 3 o 3 o 600.00 CD CO 00 CO CD CD o CM CM i CO CM x— o o T— CM CM in o co co 1 T— 1 1 i IT) in in o O o o o o o o O O O CM o o CM O CM o CM O CM o CM CM CM CM CO I CO in o o CM o • in o o CM CM i in in o o CM T— I m • in o o CM co in in o o CM CM CD I in o o CM Sample Date H Number of the Buses — Cu • Mn • Fe/100 • Zn Figure 4.80 Relation between traffic density and total metal E M C in stormwater runoff from the U B C diesel bus loop Statistically, traffic density shows a weak positive correlation with Cu, Mn, and Fe concentrations, and a weak negative correlation with Zn concentration (Table 4.29) Table 4.29 Spearman rank correlation matrix for traffic density and total metal E M C in stormwater runoff from the U B C bus loop Traffic Density Cu Mn Fe Zn Traffic Density 1.00 Cu 0.03 1.00 Mn 0.18 0.85 1.00 Fe 0.30 0.15 0.39 1.00 Zn -0.11 0.90 0.84 0.13 1.00 111 Based on the above analysis, antecedent dry period appears to be the dominant factor in stormwater runoff quality for this study. Rainfall and traffic density also have the potential to affect trace metal concentrations. But their influences are comparatively weak. 4.2.3.5 Spatial Differences in Stormwater Runoff Spatial difference in stormwater runoff were not only present on a broader scale within the G V R D area, but also, within different stormwater catch basins in the U B C bus loop. In Figure 4.81, each box-whisker contains the suspended solids (SS) concentration from a single catch basin for all sampled events. 1600-O O 1200-800 -400 • o I _4 i 1 I Catch Basin 1 Catch Basin 2 Catch Basin 3 Figure 4.81 Spatial difference of suspended solids from catch basins in the U B C diesel bus loop stormwater runoff Noticeable differences in suspended solids concentration existed among catch basins. Catch basin 3, located at the No.44 bus stop, had a higher median SS than catch basins 1 and 2, which were 32% and 41% lower. The reason that catch basin 3 was high in SS could be attributed to its relatively low runoff receiving volume. Catch basin 3 has a similar catchment area and lower traffic density than the other two basins. However, due to its uneven pavement surface, a great amount of stormwater either accumulated in the pothole or ran into other adjacent catch basins. The first flush was prolonged in catch basin 3, so when one takes a median value, the SS from catch basin 3 becomes the highest. 112 Variations in total metal concentration are also found in the three catch basins, as shown in Figures 4.82 to 4.85. Similar to the suspended solids, catch basin 3 had the highest metal concentration followed by catch basins 1 and 2. 350 Catch Basin 1 Catch Basin 2 Catch Basin 3 Figure 4.82 Spatial variation of Cu from three catch basins in the U B C diesel bus loop stormwater runoff 1400 Catch Basin 1 Catch Basin 2 Catch Basin 3 Figure 4.83 Spatial variation of Mn from the catch basins in the U B C diesel bus loop stormwater runoff 113 100000 80000 60000 o 40000 H 20000 O O illll Catch Basin I Catch Basin 2 Catch Basin 3 Figure 4.84 Spatial variation of Fe from the catch basins in the U B C diesel bus loop stormwater runoff 1400 1200 1000 H 800 B 600 o H 400 200 -x o iiiji liilli iiiii 1111 Catch Basin 1 Catch Basin 2 Catch Basin 3 Figure 4.85 Spatial variation of Zn from the catch basins in the U B C diesel bus loop stormwater runoff A n independent sample t-test was applied to the total metal concentrations in Table 4.30. For total Cu and Zn, there are statistically significant differences existing in catch basins 2/3. For Mn, significant differences existed in catch basins 1/3 and 2/3. For Fe, significant differences existed in catch basins 1/3. These results show that major differences in total metal concentration occurred in catch basin 3 as 14 compared to the other two. Similar to the suspended solids, the constantly high total metal concentration in catch basin 3 is likely due to the smaller surface flow volume into catch basin 3. Table 4.30 Student t-test for the total metals from the U B C bus loop catch basins Sample Groups (Catch Basin) C u M n Fe Z n t value P value t value P value t value P value t value P value 1/2 1.851 0.066 0.150 0.881 -1.542 0.125 0.150 0.156 1/3 -0.705 0.481 -3.856 0.000 -J .JJJ 0.001 -0.487 0.626 2/3 -2.619 0.009 -3.912 0.000 -1/449 0.149 -0.022 0.045 4.2.3.6 Comparison o f Stormwater Runof f Quality with Discharge Criteria Table 4.31 Summary of stormwater runoff quality from the U B C diesel bus loop and discharge criteria Oct. 6 Oct. 17 Oct. 29 Nov. 14 Dec. 13 Discharge Criteria T D T D T D T D T D Dissolved C u 46 26 68 44 49 17 68 44 21 7 100 M n 63 ^ j 49 9 77 29 49 9 84 11 500 Fe 2068 288 984 362 4370 269 984 362 17156 83 ~ Z n 83 42 62 28 130 58 62 28 57 9 200 Feb. 28 M a r . 8 M a r . 16 M a r . 26 M a r . 31 Discharge Criteria T D T D T D T D T D Dissolved C u 178 58 89 40 96 38 55 25 121 39 100 M n 602 71 234 27 266 15 193 16 295 35 500 Fe 23907 232 13737 317 16693 175 13178 228 13683 537 -Z n 587 176 293 97 248 81 165 77 298 108 200 115 Table 4.31 Summary of stormwater runoff quality from the U B C diesel bus loop and discharge criteria (cont.) A p r . 10 May 2 M a y l 4 M a y 31 Jun.21 Discharge Criteria T D T D T D T D T D Dissolved C u 49 30 158 108 158 79 187 117 116 58 100 M n 162 37 518 343 262 48 280 102 192 46 500 Fe 4168 203 6717 3904. 15900 457 9995 429 7471 324 -Z n 250 138 1015 669 542 251 558 404 449 282 200 (1) T — total metal event mean concentration (2) D — dissolved metal event mean concentration (3) All data listed in the table in ug/1 (4) Discharge criteria for end of pipe discharges to the environment or to storm sewers from the Pacific place site (5) Source:http://www.env.gov.bc.ca/epd/epdpa/contam_sites/standards_criteria/standards/end_of_pipe.html The event mean concentrations (EMCs) in the sampled storm events are listed in Table 4.31. The last column is the discharge criteria from provincial government legislation. Comparing the bus loop stormwater runoff quality with the discharge criteria, dissolved Cu exceeded the discharge criteria in 2 out of 15 cases; Dissolved Zn exceeded the criteria in 4 out of 15 cases, whereas dissolved M n is below the criteria in all sampled events. Al l hazardous discharges occurred in the dry season, which suggests that stormwater runoff from the U B C bus loop is not always safe to be discharged directly in to the receiving water. Especially during the dry season, stormwater runoff should be properly treated before going into the drainage system. 116 4.3 Stormwater Runoff Management by Catch Basin Filter 4.3.1 Contaminants Removal Efficiency Nine storm events were sampled after the installation of catch basin filters in early March 2005. The removal efficiencies of turbidity, conductivity, suspended solids, and metal concentrations are calculated and summarized in Table 4.32. The filter showed high and stable performance for suspended solids removal for the first month and then dropped sharply afterwards. Removal efficiencies of suspended and total metals were found higher than the dissolved metals. The adsorptive material worked moderately well for the dissolved metal removal before the filter media reached a saturation state, which is non-reversible. An evaluation of the filter media for stormwater runoff treatment was conducted by DeBusk and Langston in 1997. They found that mesocosm sand filter has a consistently high suspended solids removal performance (65%~75%) under low loading. For the same media, dissolved Cu removal initially was high, but the percentage removal rate gradually declined from 76 to 26%. The metal removal efficiencies were found to be variable to the different metals. The filter showed a better removal performance for Fe and M n then Cu and Zn. Different filter media have unique features for selected urban runoff treatment applications. For example, pear tree bark is the most effective in terms of metal removal, and synthetic wollastonite material works best for total phosphorous removal. Kellems et al. (2003) found that the adsorbent organic medium (compost filtration) can remove about 95% of dissolved copper and zinc from heavily industrialized shipyards, where the stormwater can be characterized as containing relatively high concentrations of metals, intermediate concentrations of solids, and low concentrations of oil and grease. 117 Ta'b]e-:4..'3.2*Summary. .of..catch basin filter contaminants removal efficiencies Sample Catch Basin Filter Contaminants Removal Efficiency (%) Data Turbidity Conductivity Suspended Total Dissolved Suspended (2005) Solid Cu M n Fe Zn Cu M n Fe Zn Cu Mn Fe Zn March 8 65 18 77 64 73 72 52 49 48 39 27 .66 75 73 52 March 16 66 ,6 63 52 72 76 7.6 30 25 33, 10 71 75 78 53, March 26. 80 6 61 66 83 93 50 46 52 47 3.2; 78 87 95 75 March 31 78 -38 93 64 83 89; 68 40: 34 52 31 :83 87 91 89, April 10 34 "" -25 73 " '77 75 95: 95' 32. 27 63 58 .9.8 93 92 94 May 02 31 -2 49 25 M; 35 28 15 23': 33 -11 .51 47 36 34 May 14 33; 10 61 54 80 9.4 40 m -18 -92 -33 95 96 96 91 May 31 57 8 62 29: 48: 86: 40 -3 16 ' -2.7 3 64 72 93 92 June 21 58 4 76 53 74 83" 21 -14 -3 -67 .-21 79 92 88 61 (l.)The;particulate material The filter drained well during the first month of the study due to the low solids loading and some decline in filter draining rates occurred later. The coarse material was excavated on May 3 from the catch basin insert, which improved the filter performance substantially. The performance suggests that regular maintenance, e.g. monthly excavation, needs to be conducted to ensure high solids removal performance. But the catch basin filter is not an efficient way to remove dissolved metals from stormwater runoff. A two-month period at this site resulted in saturation of the filter media since total metal removal dropped drastically in May. Further investigation of the perlite character and adsorption capacity will be discussed in the next section. 4.3.2 C a t c h B a s i n F i l t e r M e d i a C h a r a c t e r i z a t i o n To estimate the maintenance period for the catch basin filter, it is important to understand the characteristics of perlite as the filter media used in this study. Perlite was analyzed for particle size distribution, and adsorption capacity of trace metals and oil and grease under laboratory conditions. 4.3.2.1 Perlite Particle Size Distribution Particle size distribution is an important physical property of perlite. A series of seven sieves was used to classify the clean perlite particles according to their physical size. The sieve opening covers a wide range from 500 um to 9520 um. Figure 4.86 illustrates the perlite particle size distribution curve. 1 0 0 Log Particle Size (mm) Figure 4.86 Clean perlite particle size distribution curve 119 The middle section of the curve, with the highest slope, shows that more than 60% weight percentage of the particle sizes have a diameter at around 4760 um (detailed weight percentage data are listed in Table C-4 in Appendix C). As a result, the median perlite particle size of 5 mm is selected for the following adsorption capacity test. Literature shows that 5 mm is a commonly adopted perlite size for filter media, since in Stenstrom and Lau (1998)'s study of CDS unit for oil and grease removal, they used the screen size of 5.7 mm perlite for the experiment. 4.3.2.2 Perlite Adsorption Capacity for Trace Metals Single metal adsorption Capacity The adsorption capacity curve for trace metals Mn and Zn is shown in Figure 4.87 and Figure 4.88. The same shape of curve is observed from the adsorption capacity tests of both M n and Zn. The adsorption efficiency slope is high during the first half of the run, and gradually tapers off until a saturation state is reached. It is calculated that the adsorption capacity of M n and Zn are 48 mg Mn per kg perlite and 56 mg Zn per kg perlite. Figure 4.87 Mn adsorption capacity curve of perlite 120 0 10 20 30 40 50 60 70 80 90 100 110 120 130 Volume (ml) Figure 4.88 Zn adsorption capacity curve of perlite In terms of metal removal, perlite has not demonstrated to be an effective adsorbent in many studies. Blais et al. (2003) tested metal adsorption from acid effluent using various natural adsorbents. Perlite demonstrated the weaker metal adsorption than oyster shells, cedar bark, vermiculite, cocoa shells and peanut shells. The Zn adsorption capacity by activated carbon (20mg/g) from metal contaminated wastewater was three orders of magnitude higher compared to the perlite in this study (Kurniawa et al., 2006). Some less costly industrial by-products or natural materials such as banana peel and coconut shell have Zn adsorbtion capacity of 5.8 mg/g and 2.92mg/g with the influent strength double to this study (Kurniawa et al., 2006). Compared to the conventional adsorbents, permeable reactive filter materials have also been tested as a new method for heavy metals removal. Polonite (trademark) showed the highest metal treatment efficiency, where Mn, Fe, Zn and Cu concentrations from landfill leachate were removed by 99, 93, 86, and 67, respectively. Blast-furnace slag (BFS) also demonstrated good removal efficiency, where Cu and Zn were removed by 66 and 62%. The sand-peat mixture did not show a promising removal capacity for any of the elements studied, with the exception of Cu (25%). The removal of different elements was suggested to be a combination of several factors, e.g., precipitation, ion exchange and adsorption (Kietlinska and Renman, 2005). However, the comparisons in the above studies vary, depending on the characteristics of the individual adsorbent and the initial concentration of the adsorbate. 121 Interference Test for Metal Co-existence Interference from other metal species may affect the complexation of perlite with the target metal in stormwater runoff. The following section aims to examine the interference of target metals on each other, and the target metal with the most common ionic metals, such as Calcium (Ca) and Magnesium (Mg), in urban stormwater runoff. Figure 4.89 M n and Z n adsorption capacity curve for perlite In the first phase of the interference test, a mixed Mn and Zn solution was used in the interference test. Figure 4.89 demonstrates the metal's adsorption process. Mn and Zn showed similar adsorption patterns of equilibrium concentration throughout the test. Although the same Mn (5 mg/1) and Zn (5 mg/1) influent concentrations were used, the Zn effluent concentration was always below the Mn effluent concentration, which indicates a slight better adsorption tendency of Zn. For the mixed metal solution, the adsorption capacity for Mn is 27 mg/ kg perlite and 40 mg Zn/ kg perlite, which is substantially lower than the single metal adsorption (48 mg/kg and 56 mg/kg for Mn and Zn, respectively). The combined adsorption of Mn and Zn is 67 mg/kg perlite, which is higher than the individual adsorption. This is likely due to the increase in the initial ion strength in the metal solution (Dogan and Alkan, 2003). In the second phase of the interference test, Ca and Mg, as the most commonly existing elements in water, were introduced as interference agents. Two types of synthetic stormwater were made with a combination of Mn (5 mg/1), Ca (5 mg/1) and Mg (2 mg/1), and a combination of Zn (5 mg/1), Ca (5 mg/1), and M g (2 mg/1). The adsorption curves are shown in Figure 4.90 and 4.91. 122 5 0 H 1 1 1 1 1 1 1 1 1 1 1 1 0 10 20 30 40 50 60 70 80 90 100 110 120 Volume Run Through (ml) Figure 4.90 Interference of Ca and M g on the adsorption capacity for Mn by perlite Figure 4.91 Interference of Ca and M g on the adsorption capacity for Zn by perlite 123 The second interference test followed exactly the same step of interference test one. And the following adsorption capacities were found: Group One: 20 mg Mn / kg Perlite 20 mg C a / k g Perlite 6 mg M g / kg Perlite Group Two: 16 mg Zn / kg Perlite 16 mg Ca / kg Perlite 6 mg Mg / kg Perlite Results show that the adsorption capacity of both Mn and Zn were reduced by 26% and 60%, respectively, in the presence of Ca and Mg. In the natural environment where stormwater runoff contains various types of trace metals, the perlite adsorption capacity might vary from laboratory results. The metal adsorption from the field will be discussed later of this chapter. Research from CB1C (1995) concluded that the main purpose of using perlite as a filter media is to reduce oil and grease from high traffic sites, and to control sediment during municipal construction and in unpaved areas with high coarse sediment load. However, perlite adsorption is not an efficient technique for dissolved metal removal. This conclusion agrees with laboratory results from the bus loop study, in which dissolved metal adsorption is low on a mg metal/kg perlite basis. 4.3.2.3 Determination o f the Filter Maintenance Period As listed in Table 4.33, the catch basin filter's life cycle, in terms of dissolved metal removal, could be determined by the volume of stormwater running through the filter before a saturation state is reached. The volume of stormwater runoff is estimated by dividing the absorption capacity by the average trace metal concentrations in the stormwater runoff (Table 4.33). Table 4.33 Determination of the filter maintenance period from laboratory adsorption studies Typical Dissolved Metal Cone, in Filter Adsorption Capacity Volume to Reach Storm Runoff (u.g/1) (mg metal / kg perlite) Saturation (1) Mn Zn M n Zn Mn Zn 50 147 20 16 400 108 (1) Typical storm runoff dissolved metal concentration is the average event mean cone. (EMC) from the monitored storm events at the UBC bus loop. 124 A catch basin filter with 1 kg perlite can take 400 liters of stormwater runoff from the U B C bus loop before the Mn ions reach saturation state; and it takes 108 liters of stormwater runoff before the Zn ions t reach saturation state. To verify the above laboratory result, this theoretical value is compared to an estimation of what actually happened in the field. With an area of 318 m 2 and a runoff coefficient of 0.6, the bus loop generated 23.09 m 3 of stormwater runoff before the catch basin filter reached a state of saturation. Upon saturation, 404 mg of Mn/ kg perlite and 1086 mg of Zn/ kg perlite were estimated to be adsorbed by the filter media. These field values are considerably higher than the laboratory test results since the perlite bag also traps soluble metals adsorbed by solids in the stormwater runoff. Other reasons could possibly be related to the inaccurate estimation of the runoff coefficient, the difference of adsorption processes between lab conditions and the field, and the unpredictability of the natural environment. 4.3.2.4 Used Perlite Test After 110 days in service, the catch basin filter was pulled out and the used perlite was analyzed for water content, particle size and metal concentration. The perlite dry weight accounted for 42% of the wet weight. The perlite was placed in six meshed bags, which were either laid on the bottom of the chamber or tied against the sidewalls. The perlite from each bag was analyzed for heavy metal concentration. The perlite adsorbed metal concentrations are listed in Table 4.34. Table 4.34 Perlite adsorbed trace metal concentration Bag Size and Position C u Metal Content ( M n mg/kg) Fe Z n Long 1 / Bottom 6 39 3080 25 Long 2 / Bottom 8 46 4266 29 Short 1 / Side 4 26 1724 15 Short 2 / Side 5 27 1767 16 Short 3 / Side 4 35 2251 18 Short 4 / Side -> j 25 1721 14 (1) Data presented are the averages from triplicate test The longer perlite bags, which were located on the bottom of the filter chamber, had 70% higher heavy metal contents than the shorter perlite bags located vertically along the sidewalls. This suggests 125 that the perlite bags at the bottom, which had a longer contact time with stormwater runoff, functioned better than the sidewall bags. The bagged perlite was then mixed and measured for its particle size distribution analysis. o.i 1 10 Log Particle Size (mm) Figure 4.92 Used perlite particle size distribution curve Figure 4.92 shows that most of the used perlite particle sizes are within the range of 2000 urn to 4750 pm, which is of the same range as clean perlite. The trace metal concentrations of the grouped particles of perlite are listed in Table C-5 in Appendix C. The analysis shows that about 60% of the heavy metals were associated with perlite particles sizes of less than 600 um, which only constitute 8.6% of the total perlite particles. The finer perlite particles, with more surface area per unit weight, have a higher adsorption capacity, which suggests a better removal of heavy metals. However, this result may be overstated: during the sieving process, fine material (< 600 urn) that passed through the sieve consisted not only of small perlite particles, but also of surface dust (very fine material) with high metal content. 4.3.2.5 Perlite Adsorption Capacity for Grease and O i l Laboratory Oil and Grease Test Berman et al. (1991) found that oil and grease leakage from engines in parking lots and streets is a main source of organic contamination in stormwater. Since the physical properties of perlite filter media 126 render it good choice for oil and grease removal from stormwater runoff, an experiment was carried out to test the oil and grease adsorption capacity of perlite. Through a series of extraction processes, 5 grams of grease and oil were fully recovered from the saturated perlite and oil solution. Perlite adsorbed 2.22 g oil. And the rest of 2.92 g oil was recovered from the oil solution. From the lab analysis, 1 gram of perlite can reach saturation concentration by adsorbing 51.7 mg oil and grease. This adsorption capacity for oil and grease is about 1000 times better than the adsorption capacity for trace metals (Mn, Zn). Used Perlite Oil and Grease Test The oil and grease saturation concentration (51.7 mg/g) of perlite under the laboratory conditions is much higher than the oil and grease concentration (20.3 mg/g) recovered from used perlite in the field experiment. Two explanations are possible for the lower field adsorption. One is that the field perlite hasn't yet reached the saturation state. The second is that the co-existence of interference material, such as trace metals and suspended solids in stormwater runoff, reduce the adsorption capacity of perlite under field conditions. Driscoll et al. (1990) of U C L A have reported that 50 to 80% of the oil and grease in stormwater runoff may be attached to sediment and settles out. In field conditions, oil and grease removal is partially facilitated by perlite adsorption, and partially associated with the removal of suspended solids. A longer observation period of catch basin filters is suggested for future study. In addition, the used perlite from the field could be tested to determine if more oil and grease can be adsorbed under laboratory conditions. 4.3.3 C a t c h B a s i n F i l t e r T r a p p e d P a r t i c u l a r M a t e r i a l Analyses During storm events, particulate materials were washed into the filter chamber by stormwater runoff. The catch basin filter screen was gradually filled with particulate materials, which substantially reduced the filter's performance. During 110 days in service, the filter chamber was manually cleaned twice. The particulate material removed from the filter was analyzed for particle size and heavy metal concentration. As described in Table 4.35, the trapped particulate materials had a water content of 30%. On a dry weight basis, the particulate trapping rate was 1.75 kg/month for the first phase (March 2 to May 2) and 1.25 kg/month for the second phase (May 3 to June 24). The reason for the higher solids loading in the first phase is related to more frequent stop-and-go traffic under wet conditions (Kobriger and Geinepolos, 1984; Lorant, 1992). 127 Table 4.35 Catch basin filter trapped particulate material wet/dry weight and their ratio Service Period Mar/2/2005 ~ May/2/2005 May/3/2005 ~ Jun/24/2005 Total Wet Weight (kg) 3.50 2.18 Total Dry Weight (kg) 2.52 1.50 Water Content 28% 31% 4.3.3.1 Particulate Size Analys is Figure 4.93 is a plot of particle sizes with their relative weight percentages from the first excavation (March - May 2005). The distribution curve is nearly a straight line, which means the particle sizes of the particulate material were evenly distributed among the eight different openings of the sieve. CD a: 10 o -I 1 1—I—I I I I I I 1 1—I—I I I I I 0.1 1 10 Log Particle Size (mm) Figure 4.93 Particulate size distribution curve for catch basin filter trapped solids (March to May 2005) A similar particulate size distribution (Figure 4.94) was measured for the filter trapped particulate material during the second phase of the study (May - June 2005). 128 0.01 0.1 1 10 Log Particle Size (mm) Figure 4.94 Particulate size distribution curve for catch basin filter trapped solids (May to June 2005) In the second phase, the weight percentage of very fine particles (76 pm to 250 um) was relatively lower and the particle sizing from 420 urn to 9420 pm was evenly distributed. This result is likely due to the rapidly ascending flows and high intensity of the dry season storms, in which more coarse materials are washed into the catch basins (Oberts et al., 2000). 4.3.3.2 Heavy Meta l Concentration in the Catch Basin Filter Trapped Particulate Material In the analysis of catch basin filter trapped particulate materials, 600 um was selected as the segregation point between fine and coarse particles. More than half of the metal contents were associated with fine particles, which only account for 42% of the total weight of the particulate materials. As shown in Figure 4.95, the bars represent metal concentrations in the trapped particulate material from two recovery events, marked as 1 (March - May) and 2 (May - June). The top part of the bar, in black color, indicates the metal concentration associated with the fine particles and the lower part of the bar, in grey color, depicts the metal concentration associated with the coarse particles. 129 Figure 4.95 Filter trapped particle size associated trace metal concentration Most of the heavy metal contents were associated with the fine particles. This relationship is attributed to the higher surface area of the fine particles (the particulate size associated heavy metal concentrations are listed in Table C-2 and Table C-3 in Appendix C). The second phase of the study had 16, 62, 71, and 38% higher trace metal concentrations than the first phase for Cu, Mn, Fe, and Zn respectively. The higher trace metal concentrations in the dry season are a result of less particulate material loading and longer contact time between the contaminant and the particulate materials during long antecedent dry periods (Kobriger and Geinepolos, 1984; Lorant, 1992). 130 5 Conclusions and Recommendations 5.1 Conclus ions In the Brunette River watershed sediment study, the 2003 trace metal (Cu, Mn, Fe, Zn) levels associated with suspended sediment and stream bed sediment were compared with historical databases for 1993 and 1973 to illustrate temporal changes. The magnitude of trace metal contamination was than statistically related with land use (land cover, land activity, and traffic) to identify the sources of non-point trace metal pollution in urban stormwater runoff. In the U B C bus loop stormwater runoff study, the magnitude of trace metal contamination from 15 storm events was analyzed and compared with the provincial discharge guidelines. Stormwater runoff quality was strongly correlated to antecedent dry days, and weakly affected by rainfall amount, and traffic density. To best manage stormwater runoff quality, the performance of a catch basin filter was evaluated in terms of suspended solids and trace metals removal. For an indication of the filter media properties, the adsorption capacity of trace metals, oil and grease were tested in the laboratory and compared with field value. Suggestions are given for better maintenance of the catch basin filter. 5.1.1 Extent o f T r a c e M e t a l C o n t a m i n a t i o n i n A q u a t i c Sediments i n the B r u n e t t e R i v e r W a t e r s h e d In common with many other urban areas, the 2003 stream sediments in the Brunette River watershed were contaminated to various extents with Cu and Zn. With the exception of Ramsay Creek, the average bed load sediment concentrations of Cu (61 mg/kg and 63 mg/kg) and Zn (134 mg/kg, and 119 mg/kg) from Stoney Creek and Still Creek both exceeded the threshold effects level (TEL) where there are impacts on aquatic organism. Considerable temporal changes of trace metals were found throughout the watershed. In Still Creek, the stream bed metal concentrations experienced a drop (1973 to 1993) and an increase (1993 to 2003) of various extents among the metals of interest. The Cu level had a major decrease from the 1973 level and remained constantly high over the last decade. This observation suggests that although sources of industrial copper pollutant have been controlled, additional non-point Cu sources persist in the Still Creek watershed. In Ramsay Creek, Cu, Mn, and Zn reached peak levels in 1993, and then started to decrease by a factor of 31, 13, and 49% respectively. As for Stoney Creek, the major increase in Cu (by 131 13 times) and Mn (by 1.5 times) occurred between 1973 and 1993. Their levels remained steady afterwards. The Mn increase is mainly attributed to the widespread usage of the gasoline additive M M T . The increasing number of automobiles in the watershed could be a good explanation for the increase of Mn input in the steam sediment. The Zn level has increased slowly but steadily since 1973. The ratios of suspended metals to the corresponding bedload trace metals, in descending order, are 4, 3.5, 2.5, and 1.5 for Cu, Mn, Zn, and Fe, respectively. These ratios imply that the fine sediment particles in suspension have higher hazardous potential to the watershed health than coarse bedload sediment particles. This finding has also been verified by the particle fraction study of bedload sediment. The fine particles (less than 63 um) in the bedload sediment had concentrations 3.9, 2.8, 1.9, 1.5 times higher than those of coarse particles (less than 2 mm) for Cu, Mn, Fe, Zn, respectively. Seasonal variation is observed in the stream suspended sediment quality. The dry season (May to September) average trace metal levels associated with suspended sediment were significantly higher than those from the wet season (October to April). For the wet season, the total rainfall was 6.6 times the dry season, and the antecedent dry day days were much shorter. These are the possible explanations for the elevation of trace metal levels in stormwater during the summer. 5.1.2 Sediment Budget in the Brunette River Watershed Burnaby Lake sediment budget calculation was carried out using the suspended sediment loading dataset provided by N H C . In 2003, there was a total of 1406 tonnes of sediment transferred into Burnaby Lake from five upstream tributaries (Still, Eagle, and Ramsay creeks). Only 344 tonnes was discharged from the lake through the Brunette River, as the only outlet to the lower Fraser River. It is calculated that 1062 tonnes of sedimentation was retained by Burnaby Lake in 2003. In terms of metal deposition, 510 kg Cu, 3300 kg M n , 4820 kg Fe, and 628 kg Zn were deposited in the sediment of Burnaby Lake in year 2003. 5.1.3 Relationships between Land Use and Trace Metal Contamination Land use, traffic density and demographic data were collected for 2003 and compared with historical data to explore possible causes for spatial and temporal differences in trace metal non-point source contamination. Major changes in land use over the last ten years are illustrated by the significant increase in residential land use (increase in area by 12%) and decrease in open areas (reduction in area by 11%). Relatively minor changes occurred in industrial (reduction in area by 2%), commercial (reduction in area by 0.5%), transportation (increase in area by 0.3%) and agricultural land uses. The increase in 132 population (by 22%) and traffic activity has potentially contributed to the rise in trace metal levels in stormwater runoff. Stream suspended and bedload sediment quality and land use information have been analyzed to explore the relationships between stream sediment contamination and land use characteristics. Table 5.1 is a summary of correlations found between suspended metal and land use characteristics in 2003. Both trace metal concentration (mg/kg) and loading show a strong positive correlation with land use characteristics. However, no apparent correlation was observed between trace metal concentration (ug/1) or export coefficient and land use characteristics. Table 5.1 Spearman correlation between suspended metal and 2003 land use in the Brunette River watershed Suspended Metal Land Use Concentration (pg/l) Concentration (mg/kg) Loading (kg/yr) Export Coefficient (g/ha/yr) Cu Mn Fe Zn Cu Mn Fe Zn Cu Mn Fe Zn Cu Mn Fe Zn Impervi--ousness X K ) K ) y y V y y y y X ( ' ) X Traffic Density X K ) (/) K ) y y y y y y y y X K ) Catchment Area X K ) K ) ( ' ) y y y y y y y y -- - -- -(1) {/) - Strong negative correlation (-0.5 < r < 0) (2) y — Strong positive correlation (0.5 < r < 1) (3) x - Weak or no correlation (-0.5 < r < 0.5) Table 5.2 is a summary of correlations between trace metals associated with stream bed sediment and land use characteristics in 2003. The stream bed Cu and Zn concentrations are found to be positively related with land use, which reflects of their anthropogenic origins. Mn and Fe are negatively related with land use, which demonstrates the importance of natural sources. 133 Table 5.2 Spearman correlation between stream bed metals and 2003 land use in the Brunette River watershed Land Use Stream Bed Metal Concentration (mg/kg) C u M n Fe Z n Imperviousness K ) X X Traffic Density K ) K ) Catchment Area K ) ( ' ) (1) (<0- Strong negative correlation (-0.5 < r < 0) (2) ^ ~ Strong positive correlation (0.5 < r < 1) (3) x - Weak or no correlation (-0.5 < r < 0.5) 5.1.4 T r a c e M e t a l C h a r a c t e r i z a t i o n i n the U B C B u s L o o p S t o r m w a t e r R u n o f f Fifteen storm events were monitored from the U B C diesel bus loop in order to characterize the quality of stormwater runoff. In 70% of the cases, the trace metal levels as well as the suspended solids increased rapidly during the first several hours of the storm events until a peak value was reached, which is demonstrated as the "first flush". The peak values then gradually decreased to trace levels given a long enough period of rainfall. The peak of dissolved metal usually occurred slightly after the peak of total metal because of the time delay from the dissolving and mobilization processes. Positive correlations were found between dissolved metal concentration and conductivity, and also between total metal concentration and suspended solids from each of the monitored catch basins. In terms of stormwater quality, the dissolved metal concentrations were compared with the discharge criteria from the provincial government. Dissolved Cu exceeded the discharge criteria limit 2 out of 15 cases, dissolved Zn exceeded the criteria 4 out of 15 cases, and dissolved Mn was below criteria in all sampled events. This finding suggests that stormwater runoff from U B C bus loop is not always safe to be discharged directly into the receiving water. Since all hazardous discharges occurred in the dry season, stormwater runoff should be properly managed, at least seasonally, before being discharged into the drainage system. The study of storm water runoff composition shows that the metals are distributed differently in of dissolved and suspended forms due to their chemical properties. Dissolved Cu and Zn accounted for 36 and 45% of the total concentration, while Mn and Fe only accounted for 20 and 4% of their total concentration. Since dissolved metal is more mobile and has a higher bioaccumulation potential, Zn and Cu are considered to be more hazardous to the aquatic environment than Fe and Mn. 134 To illustrate spatial and temporal variations, the bus loop storm water quality is compared with previous studies conducted at the U B C south and north catchments and across the G V R D urban area. The average suspended sediment concentration from the U B C bus loop was about 5 times higher than the average concentrations from the 2001 U B C campus study and the G V R D urban area, which indicates that traffic dominated land use with high imperviousness has the potential to generate more suspended solids than comprehensive urban land use. Previous research showed that the Cu, Mn, Fe, and Zn concentrations from U B C south and north catchments were slight above or at the same level as the G V R D urban area concentrations. With high imperviousness and intensive traffic, the U B C diesel bus loop trace metal concentrations were 3, 0.7, 9, and 3.2 times higher than the G V R D urban area for Cu, Mn, Fe, and Zn, respectively. This finding suggests that diesel buses, as non-point pollution sources, induce a great amount of hazardous metal into the environment, which could severely degrade water quality by contributing trace metals into the stormwater drainage system. The emission of Fe from dissolved fuel was much higher than Cu, Mn, and Zn, since the emitted concentrations of crust elements are much higher than those of anthropogenic elements from diesel vehicle engine exhaust. Extensive variations were found in suspended solids trace metal concentrations from the sampled storm events. Similar to the result from the Brunette River watershed sediment, the storm runoff from the U B C bus loop showed seasonal changes. The trace metal event mean concentration (EMC) from the wet season (October to March) accounted for 54, 77, 109, and 79% of the event mean concentration from the dry season for Cu, Mn, Fe, Zn, respectively. The reason for the seasonal effect could be related to natural factors such as antecedent dry days, rainfall duration, and rainfall amount. As summarized in Table 5.3, positive correlations existed between antecedent dry days and trace metal event mean concentrations. Rainfall was negatively related with the trace metal concentration for most of the storm events in 2005. This suggests that the dilution effect could be significant enough to overcome the toxic effects of the runoff given a heavy enough storm event. Although diesel buses are likely the predominant source of trace metal contamination in stormwater runoff from the U B C bus loop, weak correlation was found between traffic density and trace metal concentration. 135 Table 5.3 Spearman correlation between land use and natural factors, and total metal E M C in the stormwater runoff from U B C bus loop Event Mean Total Metal Cone, (u.g/1) C u M n Fe Z n Rainfall X X X X Antecedent D r y Days • X Traffic Density X X X X (1) {/) - Strong negative correlation (-0.5 < r < 0) (2) S — Strong positive correlation (0.5 < r < 1) (3) * - Weak or no correlation (-0.5 < r < 0.5) 5.1.5 Effectiveness of C a t c h B a s i n F i l ters i n C o n t a m i n a n t s R e m o v a l To prevent stormwater with hazardous levels of contaminants from being discharged into the urban drainage system, a stormwater catch basin filter was installed. The filter performance was evaluated by examining its contaminate removal efficiency. In addition, filter media adsorption capacities of the trace metal and oil and grease were studied for better maintenance of the catch basin filter. The filter showed high and stable efficiencies in total metal (Cu 62%, Mn 75%, Fe 83%, Zn 62%), dissolved metal (Cu 39%, M n 37%, Fe 47%, Zn 32%), turbidity (72%), and suspended solids (74%) removal for the first month in use. Then, a gradual degradation started in the following month. The catch basin filter was cleaned after two months of service. And its performance with the suspended solids and total metal removal significantly improved. The perlite filter media showed poor performance in terms of dissolved metal removal in the second study period, which is likely attributed to limited adsorption capacity. The use of a catch basin filter is an effective stormwater management practice to control suspended solids loading from the stormwater runoff. But regular maintenance should be conducted to ensure high performance. The catch basin filter is not as effective at removing dissolved metals. Once the filter media reaches saturation, its ability to absorb dissolved metal is lost permanently. Particulate size analysis and adsorption tests were designed to characterize perlite particles. From the single trace metal adsorption capacity test, 1 kg perlite has an adsorption capacity of 48 mg Mn or 56 mg Zn. However, the adsorption capacity could be substantially reduced by the presence of other metals in the solution. Mn adsorption capacity decreased by 44% in the presence of Zn and decreased by 58% in the presence of Ca and Mg. Zn adsorption capacity decreased by 29% in the presence of Mn and decreased by 71% in the presence of Ca and Mg. According to the (Ca, Mg) interference test, a catch basin filter with 1 kg perlite can take up to 400 liters of stormwater runoff before the Mn ions reaches 136 saturation state; and it takes 108 liters of runoff before the Zn ions reach saturation state, with the average dissolved metal concentrations from the U B C bus loop. -Laboratory oil and grease adsorption tests showed that 1 kg of unused perlite reaches saturation by adsorbing 51.7 g oil and grease. This result is much higher compared to the oil and grease extracted (20.3 g/kg) from the used perlite in the catch basin filter. It is possible that the oil and grease adsorption capacity is also subject to interference from other substances in stormwater runoff. Based on the above adsorption capacity study, perlite is a very effective adsorbent to remove oil and grease from stormwater runoff. However, in terms of heavy metal removal, since perlite performed three orders of magnitude less effectively than oil and grease, alternative methods need to be explored. Recommendations • Quantify the effects of impervious land surfaces on stream contamination. • Take action to reduce the amount of sedimentation into Burnaby Lake by applying sediment control strategies. • Investigate the actual effects that contaminants might be posing to aquatic organisms in the Brunette River watershed. • Propose remediation response, such as stormwater BMPs, to the heavily contaminated streams with trace metals that are hazardous to aquatic health. • Monitor flow rates from the U B C bus loop storm catch basin in order to determine pollutant loading and its relation to stormwater quality from the U B C campus. • Carry out perlite oil and grease adsorption capacity in the presence of competing interferences (trace metals). • Investigate the impacts of the U B C bus loop stormwater runoff contamination on receiving waters quality and aquatic organisms (Bioassay studies would informative). • Application of the stormwater catch basin filter research to other locations with a different mix of vehicles, traffic patterns, and surrounding land uses. • Compare perlite media adsorption to possible other adsorption materials. 137 Bibl iography Alavaster, J.S. and Lloyd, R., 1980. "Water quality criteria for fish". 2nd Edition, London: Butterworths. A P W A , 1981. "Urban storm water management". American Public Works Association. A P H A , 1995. "Standard methods for the examination of water and waste water". 19th edition, American Public Health Association, Washington, DC. ATSDR, 2000. "Toxicological profile for manganese". Department of Health and Human Services, Public Health Service, Atlanta, G A . Avila, A . , Alarcon, M . , and Rodrigo, A . , 2003. "Heavy metal deposition and cycling at two Mediterranean holm-oak (Quercus ilex L.) forests in northeastern Spain". Journal De Physique. IV, 107: 75-78. Bastian, R.K., 1986. "Potential impacts on receiving waters". Urban Runoff Quality Impact and Quality Enhancement Technology Proceedings of an Engineering Foundation Conference. 157-169. Berman, L . , Hartline, C , Ryan, N . , and Thome, J., 1991. "Urban runoff: water quality solutions". American Public Works Association, special report No. 61. Blais, J.F., Shen, S., Meunier, N . , and Tyagi, R.D., 2003. "Comparison of natural adsorbents for metal removal from acidic effluent". Environmental Technology, 24(2): 205-215. Blumberg, K. and Walsh, M.P., 2004. "Status report concerning the use of M M T in gasoline". International Council on Clean Transportation. Brault, N. , Loranger, S., Courchesne, F., Kennedy, G. , and Zayed, J., 1994, "Bioaccumulation of manganese by plants: Influence of M M T as a gasoline additive". Science of Total Environment, 153: 77-84. Canadian Council of Ministers of the Environment, 1999. "Canadian environmental quality guidelines". Catch Basin Insert Committee (CBIC), 1995. "Evaluation of commercially available catch basin inserts for the treatment of stormwater runoff from developed sites". King County Surface Water Management Division, Port of Seattle, WA. Chapman, D., 1992. "Water quality assessments: A guide to the use of biota, sediments and water in environmental monitoring". Chapman & Hall, London. Cole, R.H. , Frederick, R.E. , Healy, R.P., and Rolan, R.G. , 1984. "Preliminary findings of the priority pollutant monitoring project of the nationwide urban runoff program". Journal of WPCF, 56: 898-908. Compass Resources Management Ltd., 1997. "Brunette Basin task group report", for Greater Vancouver Regional District, Burnaby, B C . Councell, T .B. , Duckenfield, K . U . , Landa, E.R., and Callender, E . , 2004. "Tire-wear particles as a source of zinc to the environment". Environmental Science & Technology. 38: 4206-4214. 138 Dawson, L . , Flaherty, M . , and Gang. M . , 1985. "Still Creek inventory". Part f$ Rupert Cassiar Neighborhood Association. Dayton & Knight, 1999. "Best management practices guide for stormwater". for Greater Vancouver Sewerage and Drainage District. Dayton & Knight Ltd., North Vanvouer, B C . DeBusk, T .A. , Langston, M . A . , Schwegler, B.R., and Davidson, S., 1997. "An evaluation of filter media for treating stormwater runoff. Fifth Biennial Stormwater Research Conference. Department of Water Affairs & Forestry (DWAF), 1996. "Water quality guidelines". 1st edition, D W A F , Pretoria, W A Dobson, A.W. , Erikson, K . M . and Aschner, M . , 2004. "Manganese neurotoxicity". Annals of the New York Academy of Sciences, 1012: 115-128. Dogan, M . and Alkan, M . , 2003. "Adsorption kinetics of methyl violet onto perlite". Chemosphere. 50(4):517-28. Driscoll, E.D., Shelley, P.E. and Strecker, E.W., 1990. "Loadings and impacts from storm water runoff. VIII: Analytical investigation and research report. Federal Highway Administration. Dzombak, D.A. and Morel, F . M . , 1990. "Surface complexation modeling". John Wiley & Sons Inc. Environment Canada, 1987. "Managing PCB wastes". Fact sheet published by Minister of Supplies and Services Canada. Environment Canada, 1993. "CSO and U R investigative assessment guidelines". Fraser Pollution Abatement Office, Environment Canada, North Vancouver, B C . Environmental Literacy Council, 2002. "Environmental impacts of deicing salts". http://www.enviroliteracy.org/article.php/709.html E P A , 1983, "Results of the nationwide urban runoff program". Vol.1 Final Report, Water Planning Division, U.S. Environmental Protection Agency, NTIS PB84-185552. Washington, DC. E P A , 1992a. "Draft storm water pollution prevention for industrial activities". Office of Wastewater Enforcement and Compliance, U.S. Environmental Protection Agency, Washington, D C . E P A , 1992. "Storm water pollution prevention for industrial activities". Draft report U.S. Environmental Protection Agency, Washington, D C . E P A , 1994. "EPA's polluted brochure" EPA-841-F-94-005. U.S. Environmental Protection Agency, Washington, D C . E P A , 1999. "Preliminary data summary of urban storm water best management practices" U.S. Environmental Protection Agency, Washington, DC. E P A , 2004. "Sewer sediment and control: A management practices reference guide". U.S. Environmental Protection Agency, 600/R-04/059. Washington, DC. 139 EPA, 2005. "Stormwater first flush pollution". http://www.epa.nsw.gov.aU/mao/stormwater.htm#whatis Farkas, A . , Salanki, J., and Specziar. A . , 2003. "Age- and size-specific patterns of heavy metals in the organs of freshwater fish Abramis brama L. populating a low-contaminated site". Water Research 37(5): 959-64. Field, R. and Pitt, R.E. , 1990, "Urban storm-induced discharge impacts". Water Environment & Technology 2(8): 64-67. Forget, E . , Courchesne, F., Kennedy, G . , and Zayed, J., 1994. "Response of blue spruce to manganese pollution from M M T " . Water, Air, Soil Pollution 73: 319-324. Gardner Dunster Associates Ltd., 1992. "The nature of Burnaby: An environmentally sensitive areas strategy", prepared for the City of Burnaby. Garnaud, S., Mouchel, J .M. , Chebbo, G . , and Thevenot, D.R., 1999. "Heavy metal concentrations in dry and wet atmospheric deposits in Paris district: Comparison with urban runoff. Science of the Total Environment, 235: 235-245. German, J. and Svensson, G . , 2002. "Metal content and particle size distribution of street sediments and street sweeping waste". Water Science & Technology, 46: 191-198. Glasby, G.P., Stoppers, P., Walter, K.R. and Davis, K.R., 1998. "Heavy-metal pollution in Manukau and Waitemata harbours in New Zealand". New Zealand Journal of Marine and Freshwater Research, 22: 595-611. Glenn, D.W. and Sansalone, J.J., 2002. "Accretion and partitioning of heavy metals associated with snow exposed to urban traffic and winter storm maintenance activities." Journal of Environmental Engineering, 128: 2-167. Gnecco, I., Berretta, C , Lanza, L . G . , and La Barbera, P., 2005. "Storm water pollution in the urban environment of Genoa, Italy". Atmospheric Research, 77: 60- 73. Greenstein, D., Tiefenthaler, D., and Bay, S., 2004. "Toxicity of parking lot runoff after application of simulated rainfall". Arch. Environ. Contam. Toxicol. 47: 199-206. Gupta, M . K . , Agnew, R.W., and Meinolz, T . L . , 1981. "Constituents of highway runoff. Volume II, Final report to Federal Highway Administration, Washington, D C . G V R D , 1993. "Choices: Growth management and transportation". G V R D Communications and Education. Burnaby, B.C. G V R D , 1998. "Design of a monitoring program to evaluate environmental effects of stormwater discharges". Prepared for District Burnaby, B.C. Prepared by E V S environment consultants, North Vancouver, B C . G V R D , 1998. "Still Creek - Brunette River floodplain mapping". Technical Report. Burnaby, B C . G V R D , 2001, "Brunette Basin Watershed plan". Prepared for Brunette Basin Task Group, Prepared by G V R D . Policy and Planning Department, Burnaby, BC. 140 G V S D D , 1999. "Best management practices guide for stormwater", Prepared by Greater Vancouver Sewerage and Drainage District. Hall, K.J . , Kiffney, P., Macdonald, R., and Larkin, G. , 1999. "Non-point source contamination in the urban environment of Great Vancouver: A case study of the Brunette River Watershed". Department of Environment, Fraser River Action Plan, Publication No.: D O E F R A P 1998-11. Hall, K.J . , Larkin, G.A. , Macdonald, R.H. , and Schreier, H . , 1998. "Water pollution from urban stormwater runoff in the Brunette River Watershed, B.C.". Prepared for Environment Canada. Hall, K.J . and Wiens, J.H., 1976. "The quality of water in tributaries of the lower Fraser and sources of pollution". Westwater Research Centre, University of British Columbia, Vancouver, B C . Hallberg, M . , 2006. "Suspended solids and metals in highway runoff: implications for treatment systems". Licentiate thesis. Department of Land and Water Resource Engineering. Stockholm, Sweden. Harrison, R . M . , Wilson, S.J., 1985. "The chemical composition of highway drainage waters - major ions and selected trace metals". Science of the Total Environment 43: 63-77. Hewitt, N . C . and Rashed, M . B . , 1992. "Removal rates of selected pollutants in the runoff waters from a major rural highway". Water Research. V26(3): 311-319. Horner, R.R., Booth, D.B. , Azous, A . , and May, C.W., 1997. "Watershed determinants of ecosystem functioning". American Society of Civil Engineering. New York, N Y . Johnson, B., Yandora, K. , and Bryant, S., 1998. "A comprehensive approach to urban stormwater impact assessment". Greensboro Storm Water Services, Greensboro, N C . Johnson, C D . and Juengst, D., 1997. "Polluted urban runoff: A source of concern". University of Wisconsin-Extension, Madison, WI. Judith, A . and Gerlach, O., 1998. "The Influence of land use and stream morphology on urban stream water quality". Virginia Department of Forestry, VI. Kavelaars, T., 1998. "Characterization of stormwater runoff from three catchments". Final report Vol 1& 2 prepared for Greater Vancouver Regional District, Burnaby, B C . Kellems, B .L. , Johnson, R., and Sanchez, F., 2003. "Design of emerging technologies for control and removal of stormwater pollutants". Submitted for presentation at Water World & Environmental Resources Congress, Philadelphia, PA. Kerri, K.D. , Racine, J.A., and Howell, R.B., 1985. "Forecasting pollutant loads from highway runoff. Transportation Research Record, 1017: 39-46. Kietlinska, A . and Renman, G. , 2005. "An evaluation of reactive filter media for treating landfill leachate". Chemosphere, 61(7): 933-940. King County (KC), 1998. "Surface water design manual", King County Surface Water Management Division, Washington, W A . 141 Kobriger, N.P. and Geinepolos, A . , 1984. "Sources and migration of highway runoff pollutants". Vol . 1 Research report for Federal Highway Administration, Environment Division, Washington, DC. Kurniawan, T .A. , Chan, G.Y.S. , Lo, W., and Babel, S., 2006. "Comparisons of low-cost adsorbents for treating wastewaters laden with heavy metals". Science of the Total Environment 366: 409-426 Lawson, E . M . , Mitchell, G.B. , and Walton, D.C. , 1985. "Stormwater monitoring in an industrial catchment basin located in Burnaby, B.C". British Columbia Ministry of Environment report. Lee, G.F. and Jones-Lee, A . , 1996a. "Evaluation monitoring for stormwater runoff monitoring and B M P development". Report of G . Fred Lee & Associates, El Macero, C A . Lee, G.F. and Jones-Lee, A . 1997d. "Evaluation monitoring as an alternative for conventional stormwater runoff water quality monitoring and B M P development". Report of G . Fred Lee & Associates, El Macero, C A . Lee, Y .D . , 2004. "Adaptation of North American technologies and practices for stormwater management and non-point water pollution source control to Korean condition". Report prepared at IRE, U B C . Leif, W.T., 1998. "Sediment removal in catch basins and catch basin inserts". Snohomish County Department of Public Works, Everett, W A . Lin, J.G., Chen, S.Y., and Su, C.R., 2003. "Assessment of sediment toxicity by metal speciation in different particle-size fractions of river sediment". Water Science and Technology, 47: 233-241. Lind, B.B. , Backatrom, M . , and Geisler, E . , 2001. "First flush effect of metals and anions in stormwater runoff from roads in mid-Sweden". Swedish Geotechnical Institute, Sweden. Lindberg, S.E. and Turner, R.R., 1983. "Trace metals in rain at forested sites in the eastern United States". CEP Consultants Ltd, 107-114. Loranger, S., Zayed, H. , and Forget, E . , 1994. "Manganese contamination in Montreal in relation with traffic density", Water, Air and Soil Pollution, 32:142-152. Lorant, F.I., 1992a. "Effects of piped median vs. open ditch drainage on stormwater quality". Research and Development Branch, Ontario Ministry of Transportation, O N . Lorant, F.I., 1992b. "Highway runoff water quality - Literature review". Research and Development Branch, Ontario Ministry of Transportation, O N . Lyttle, C M . , Smith, B.N., and McKinnon, C .Z . , 1995. "Manganese accumulation along Utah roadways: A possible indication of motor vehicle exhaust pollution". The Science of the Total Environment, 162: 105-109. Macdonald, R., Hall, K. and Schreier, H . , 1997. "Water quality and stormwater contaminants in the Brunette River Watershed, British Columbia, 1994/95". Westwater Research Unit, Institute for Resources and Environment, University of British Columbia, Vancouver, B C . 142 Macdonald, R., 2005. "Estimated stormwater character and loadings in runoff from U B C drainage catchments". Submitted to: Alek Pedaruski, Utilities Department, University of British Columbia, B C . Marsalek, J. and Greek, B., 1984. "Toxic substances in urban runoff in the Niagara River Area". Unpublished Report Series. Environmental Hydraulic Division, National Water Research Institute, Burlington, O N . McCallum, D.W., 1995. "An examination of trace metal contamination and land use in an urban watershed". M A S c Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, BC. McCallum, D.W. and Hall, K.J. , 1998. "Tracing the contaminant history of an urban watershed through an examination of aquatic sediments". Prepared for Environment Canada, D O E F R A P 1998-24. McKenzie, D.J. and Irwin, G.A. , 1983. "Water quality assessment of stormwater runoff from a heavily used urban highway bridge in Miami, Florida". U.S. Geological Survey Water Resources Investigations Report 83-4153. Tallahassee, F L . M E L P (now W L A P ) , 1999. "Tackling non-point source water pollution in British Columbia: An action plan". Prepared by Ministry of Environment, Lands and Parks, British Columbia (now called Ministry of Water, Land and Air Protection), Victoria, B C . Mikkelsen, P.S., Viklander, M . , Linde, J.J., and Malmqvist, P.A., 2001. "BMPs in stormwater management in Denmark and Sweden". Watershed Protection Techniques, 1: 55-61. Mitchell, M . K . and Stapp, W.B. , 1992. "Field manual for water quality monitoring". 8th Ed. Ann Arbor, MI. Morrison, G.M.P. , Revitt, D . M . , and Ellis, J.B., 1984. "Variations of dissolved and suspended solid heavy metals through an urban hydrograph". Environmental Technology 7: 313-318. Ndiokwere, C . L . , 1984. "Study of heavy metal pollution from motor vehicle emissions and its effect on roadside soil, vegetation and crops in Nigeria". Environmental Pollution, Series B: Chemical and Physical, 7: 35-42. Northwest Hydraulic Consultants Ltd (NHC), 2004. "Brunette basin sedimentation monitoring and analysis program". Draft Report prepared for G V R D , North Vancouver, B C . Nuernberg, H.W., Stoeppler, M . , and Brainica, M . , 1984. "Bioaccumulation of heavy metals by bivalves from Lim Fjord (North Adriatic Sea)". Mar. Biol. (Berlin) 81(2): 177-188. Oberts, G .L . , Marsalek, J., and Viklander, M . , 2000. "Review of water quality impacts of winter operation of urban drainage". Water Quality Resource, 35: 781-808. Ouellet, M . and Jones, H.G. , 1982. "Historical changes in acid precipitation and heavy metals deposition originating from fossil fuel combustion in eastern North America as revealed by lake sediment geochemistry". Water Science and Technology, 15: 115-130. Percy, J.A., 2004. "Contaminant concerns: Heavy metals and the bay of Fundy". Sponsored by Environment Canada (Atlantic Region), Issue No. 25. 143 Perschbachera, P.W. and Wurtsbaugh, W.A. , 1999. "Effects of calcium and magnesium hardness on acute copper toxicity to juvenile channel catfish". Aquaculture 172: 275-280. PPRB, L H D , and M E L P S , 1991. "Standards for end of the pipe water based discharges from the Pacific place site", Pollution Prevention and Remediation Branch, Environment and Lands Headquarters Division, and Ministry of Environment, Lands and Parks Site. Victoria, B C . Poon, D., 1989. "Annual report: Leaded and lead-free gasoline regulations monitoring program". Environment Agency, Chicago, IL. Revitt, D . M . , Hamilton, R.S., and Warren, R.S., 1990. "The transport of heavy metals within a small urban catchment". Science of the Total Environment, 93: 359-373. Reybekiel, A . C . , 2001. "Metal emissions from Stockholm traffic - wear of brake linings". Air Quality and Noise Analysis on behalf of the Stockholm Environment and Health Protection Administration and the Swedish Environmental Protection Agency, Stockholm. Richardson, J.S., 1998. "Benthic invertebrate community structure in small streams in the Lower Fraser Basin" - unpublished data. University of British Columbia, Vancouver, B C . Rithaler, R., 1998. "Benthic invertebrate monitoring in Cougar Creek" - unpublished data. Corporation of Delta, BC. Sansalone, J.J. and Buchberger, S.G., 1997."Partitioning and first flush of metals in urban roadway storm water". Environmental Engineering, 123(2): 134-143. Sansalone, J.J., Buchberger, S.G., and Koechling, M.T. , 1995. "Correlations between heavy metals and suspended solids in highway runoff: implication for control strategies". Transportation Research Record, Issue No. 1483. Scanlon, P.F., 1977. "A study of lead, cadmium, nickel and zinc levels in soil, vegetation, invertebrates and mammals associated with highways of different traffic densities". Report to U.S. Department of Transportation by Department of Fisheries and Wildlike Sciences, Virginia Polytechnic Institute and State University, Blacksburg. Scheinberg, H. , 1991. "Metals and their compounds in the environment: Occurrence, analyses and biological relevance". V C H , N Y . 803-851. Schreier, H. , Hall, K.J. , Brown, S., Lavkulich, L . , and Zandbergen, P., 2004. "Integrated watershed management". 2nd edition. University of British Columbia, Vancouver, B C . Schueler, T., 1994. "Feature article". Watershed Protection Techniques 1(3): 104. Sekela, M . , Brewer, R., Tuominen, T. , Sylvestre, S., and Moyle. G. , 1998. "Effect of a rainfall event on contaminant levels in the Brunette River Watershed". Aquatic and Atmospheric Sciences Division, Environment Canada, Pacific and Yukon Region, Vancouver, B C . Smith, S., MacDonald, D., Keenleyside, K. and Gaudet. C , 1995. "The development and implementation of Canadian sediment quality guidelines". Ecovision World Monograph Series. Stenstrom, M . K . and Lau, S.L., 1998. "Oil and grease removal by floating sorbent in a C D S device". 144 University of California Los Angles, C A . Stenstrom, M . K . , Silverman, G.S. and Bursztynsky, T .A . , 1984. "Oil and grease in urban stormwaters". Journal of Environmental Engineering, 1 10: 58-72. Straus, D.L. and Tucker, C.S., 1993. "Acute toxicity of copper sulfate and chelated copper to channel catfish". J. World Aquaculture Soc, 24(3): 390-395. Struempler, A .W. , 1979. "Aerosol metal relationships in air and precipitation". Environment International, 2: 63-65. Swain, L . G . and Walton, D.G. , 1991. "Quality of stormwater/urban runoff. B .C. Ministry of Environment, Victoria, B.C.". Paper presented to B.C. W N A Conference in Vancouver, B C 1-15. Toews, D. and Hetherington, E . , 2004. "A brief history of forest hydrology in British Columbia". Watershed Management Bulletin, 8: 1. Turer, D., Maynard, J. B., and Sansalone, J. J., 2001, "Heavy metal contamination in soils of urban highways: Comparison between runoff and soil concentrations at Cincinnati, Ohio". Water, Air and Soil Pollution, 132: 293-314. Wang, Y . F . , Huang, K . L . and L i , C.T. , 2003. "Emissions of fuel metals content from a diesel vehicle engine". Atmospheric Environment, 37: 4637-4643. W E F and A S C E , 1998. "Urban runoff quality management". W E F Manual of Practice No. 23. A S C E Manual and Report on Engineering Practice No. 87. Alexandria and Reston, V A . Westerlund, C. and Viklander, M . , 2006. "Particles and associated metals in road runoff during snowmelt and rainfall". Science of the Total Environment, 362: 143- 156. Westerlund, K . G . , 2001. "Metal emissions from Stockholm traffic - wear of brake linings". The Stockholm Environment and Health Protection Administration, Stockholm. Wong, M . , 1996. "Personal communication". Environmental Protection Branch, Environment Canada. Woodward-Clyde Consultants, 1990. "Urban targeting and B M P selection". United States Environmental Protection Agency, Water Division, Chicago, IL. Woodward-Clyde Consultants, 1995. "Stormwater quality monitoring guidance manual". Prepared for Washington Department of Ecology, Water Quality Program. Yokel, R.A. and Crossgrove, J.S., 2004. "Manganese toxic kinetics at the blood-brain barrier". Health Effects Institute Research Report, No . l 19. Yousef, Y . A . , Harper, H . H . , Wiseman, L.P. , and Bateman, J .M. , 1985a. "Consequential species of heavy metals in highway runoff. Transportation Research Record 1017. Transportation Research Borad, Washington, D C , 56-62. 145 Yousef, Y . A . , Wanielista, M.P., and Harper, H.H. , 1985a. "Removal of highway contaminants by roadside swales". Transportation Research Record 1017. Transportation Research Board, Washington, D C , 62-68. Yousef, Y . A . , Wanielista, M . P., Hvitved-Jacobsen, T. , and Harper, H.H. , 1984. "Fate of heavy metals in stormwater runoff from highway bridges". Science of the Total Environment 33: 233-244. Zandbergen, P. and Schreier, H . , 1998. "Urban watershed assessment: Watershed health indicators and management." For Environment Canada C D - R O M @ Institute for Resources and Environment, The University of British Columbia, Vancouver, B C . Zayed, J., Vyskocil, A . , and Kennedy, G. , 1999b. "Environmental contamination and human exposure to manganese — contribution of methylcyclopentadienyl manganese tricarbonyl in unleaded gasoline". International Archives of Occupational and Environmental Health, 72:7-13. Zimdahl, R.L. , 1972. "Impact on man of environmental contamination caused by lead". In Interim Report, National Science Foundation Grant, Colorado State University, Fort Collins. 98-110. 146 A p p e n d i x A S u m m a r y of Q A / Q C D a t a The Q A / Q C program included field and laboratory blanks, replicates, and spike addition. Accuracy Test To confirm the accuracy, a known amount of the metal of interest is reanalyzed every ten sample. In this test, mid-point check standards were used. Table A - l Accuracy Assessment for Trace Metals Metals Amount added (mg/1) Amount found (mg/1) Recovery Rate (%) Cu 0.5 0.6 112 0.1 0.1 94 Mn 0.5 0.5 97 0.1 0.1 124 Fe 5.0 4.8 95 1.0 1.1 112 Zn 0.5 0.5 101 0.1 0.1 95 As shown in Table A - l , good accuracies are observed in this test. 75% of the analyzed metals have recovery rates within the acceptable range of 87 to 110%. Replicate Test To establish method precision, sample duplicates were split and went through the same digestion and analysis process. The results in Table A-2 show good duplication results. 147 Table A-2 Duplicate test Metals Duplicate 1 (mg/1) Duplicate 2 (mg/1) Difference (%) Cu 0.23 0.23 0.21 0.23 11.6 3.5 Mn 0.27 0.25 0.27 0.26 2.2 5.4 Fe 8.85 6.25 8.10 6.50 8.4 3.8 Zn 0.21 0.22 5.5 0.38 0.48 20 Table A-3 presents the results of detection limits, replicates and spike recoveries as well as metal detection limits stormwater runoff from U B C bus loop. Table A-3 Spike recovery Sampling & Detection Spike Recovery Parameter Analysis Precision Limits N % N M i n (%) M a x (%) Ave (%) Cu 3xl0 - 2 18 4.00 5 80 159 104 Mn 3x10~2 17 3.75 5 66 115 84 Fe 5xl0 - 2 18 5.73 5 85 132 103 Zn 4xl0"2 18 4.34 5 65 130 112 (1) All metal concentration units are in mg/1 (2) Sample precision: N = number of grab sample analyzed (3) The Detection Limit (DL) is calculated as 2 times of the average standard deviation. 148 A p p e n d i x B M e t a l Digest ion Techniques Digestion Comparisons A set of bed load sediment samples were digested using both nitric and aqua regia acid digestion in order to compare their recovery. Each type of digestion was then analyzed by flame A A spectroscopy. Results from the t-test (Table B-l) show non-significant difference between two digestion methods for Cu, Mn, Fe, and Zn. Table B - l t-test for metal concentration difference between the two sediment digestion methods Element No. of Metal Concentration mg/kg H N 0 3 : t-test Replicates H N 0 3 Aqua Regia Aqua Regia Cu 4 229 230 1.00 tcalculate^ttable 4 86 85 1.01 difference non-4 390 401 0.97 significant 4 48 52 0.92 •> 663 672 0.99 Mn 4 2603 2587 1.01 tcalculate^ttab|e 4 559 556 1.01 difference non-4 543 546 0.99 significant 4 347 324 0.56 -i 1452 1464 0.99 Fe 4 135 134 1.01 tca!culate^tta5ie 4 37 37 1.00 difference non-4 40 40 1.00 significant 4 35 40 0.88 64 66 0.97 Zn 4 209 207 1.01 tcalculate^tjable 4 107 107 1.00 difference non-4 449 449 1.00 significant 4 73 85 0.86 "i 637 638 1.00 149 In comparison of the precision for the nitric and aqua regia digestion methods, a large number of sediments are analyzed. In Table B-2, results were compared in terms of standard deviation and coefficient of variation. Aqua Regia showed slight higher precision than nitric digestion method. Table B-2 Precision difference for nitric and aqua regia digestion techniques for sediments Digestion Method Aqua Regia (n=10) H N O s ( n=8) Metal C u M n Fe Z n C u M n Fe Z n 4.2 24.4 945.6 18.4 9.7 J J . J 1358.6 18.7 4.7 29.1 1249.6 15.9 4.1 28.5 1210.2 12.7 6.0 25.6 1537.7 19.7 5.1 44.0 1964.8 14.1 3.4 20.8 903.8 12.9 4.0 34.4 1436.5 14.7 Cone. 5.1 28.5 1200.7 20.5 3.4 22.2 1091.5 10.9 mg/kg 4.5 24.7 1040.6 18.1 4.3 32.0 1524.6 15.9 4.7 24.6 995.4 14.9 5.1 34.7 1825.2 29.6 4.7 27.3 1087.8 16.9 4.3 32.6 1373.7 16.1 6.8 19.6 909.3 16.5 ~ - - -6.2 22.7 1158.6 17.0 - - - ~ Average 5.0 24.7 1102.9 17.1 5.0 32.7 1473.1 16.6 Standard Deviation 1.0 3.1 194.6 2.2 2.0 6.1 294.6 5.8 Coefficient Variation 2 0 % 10% 2 0 % 10% 40% 2 0 % 2 0 % 3 0 % 150 A p p e n d i x C P a r t i c l e Size A n a l y s i s a n d its Associated T r a c e M e t a l s Bedload Sediment from the Brunette River Watershed The trace metal data collected from 1973 and 1993 were from less than 180 pm fraction. During the latest years, many river and lake sediment studies started to use less than 63 pm fraction for the trace metal analysis (Forstner 1990). So the less than 63 um sediment fraction was chosen to perform trace metal analyses on the 2003 stream sediments. In order to compare the results over the 30 years of the Brunette River Watershed study, the 1973 and 1993 data were converted to less than 63 pm values based on the sediment fraction analysis by Donald McCallum (1995). From the 14 samples, McCallum found linear relationships between the ratios of metal concentrations in the two size fractions (<180 pm: <63 pm). Table C - l Stream sediment trace metal average concentration in 1973, 1993, and 2003 <180um Bedload Fraction mg/kg <63 um Bedload Fraction mg/kg Location Year C u M n Fe Z n C u M n Fe Z n Still Creek 1973 750 323 28600 307 968 436 32318 399 1993 195 346 23219 262 217 427 23733 354 2003 - - - - 319 711 44527 372 Ramsay 1973 14.2 416.5 15100 46 18 562 17063 60 Creek 1993 97 474 18748 161 125 640 21185 209 2003 - - - - 86 559 36654 107 Stoney 1973 15.5 478 22900 53 20 645 25877 69 Creek 1993 224 1188 27149 98 289 1603 30679 127 2003 _ _ 275 1642 89378 211 Catch Basin Filter Trapped Particulate Material The metal concentrations of each range of the particle sizes from the first excavation are listed in the following Table C-2. 151 Table C-2 Catch basin filter trapped particulate size associated trace metal concentration during first filter clean out Sieve # Diameter Weight Metal Content Distribution (m g/kg) (pm) % C u M n Fe Z n #48 297 18.29 85.6 303.4 27011.6 197.2 #45 355 6.54 4.4 21.4 1643.4 12.7 #35 500 10.22 5.4 32.9 2461.7 15.7 #30 600 7.16 3.8 25.6 1970.0 13.4 #20 841 10.46 4.7 36.0 2914.1 13.1 #10 2000 22.74 8.9 64.0 4594.9 23.4 #4 4760 15.72 4.0 34.5 703.8 12.3 #2 9520 8.69 0.5 7.0 148.6 2.7 (1) Particles retaining period: Mar/2/2005 ~ May/2/2005 The metal concentrations of each range of the particle sizes from the second excavation are listed in the following Table C-3. Table C-3 Catch basin filter trapped particulate size associated trace metal concentration during second filter cleanout Sieve # Diameter (u.m) Weight % Metal Content Distribution (m C u M n Fe g/kg) Z n #200 76 0.99 114.7 244.9 33490.6 180.4 #100 147 6.66 63.2 204.1 21297.9 133.6 #60 250 9.02 34.8 187.1 14044.0 107.8 #40 420 15.72 25.6 165.8 13136.6 92.3 #30 595 8.83 33.6 177.9 14445.7 91.5 #20 841 8.36 27.0 200.3 15365.1 93.7 #10 2000 17.88 22.8 205.5 23895.2 70.4 #4 4750 20.8 17.1 127.0 7848.3 42.3 #3.5 5660 2.91 23.2 138.7 14674.2 51.7 #2 9420 8.83 19.6 101.9 9895.8 42.6 (1) Particles retaining period: May/3/2005 ~ Jun/24/2005 152 Catch Basin Filter Media - Perlite Unused perlite: Fresh perlite was analyzed for particle size and weight distribution in Table C-4. Table C-4 Perltie particle size and weight distribution Sieve # Diameter (u.m) Perlite Weight Percentage (%) 2 9520 1.6 6680 12.8 4 4760 61.6 10 2000 12.0 20 841 9.4 30 595 2.4 35 500 0.1 Used perlite: used perlite from catch basin filter after 110 days in service was analyzed for particles size associated trace metals (Table C-5). Table C-5 Perlite particle size associated trace metal concentration Sieve # Diameter Weight Metal Content Distribution (m g/kg) (P-m) % C u M n Fe Z n ••60 250 5.51 38.6 ...32.7" 51.8 Vv6 Billjjl 3.06 2(1.6 l|IIJi|lB 17.0 •30 5'J5 1.4 3.9 8.7 6.2 5.9 #20 841 1.31 4.1 7.5 8.3 4.2 #10 2000 1.92 2.9 3.9 2.5 3.1 #4 4750 69.4 20.3 21.4 13.9 31.4 #3.5 5660 7.96 5.1 4.4 2.3 2.3 #2 9420 9.09 4.5 5.0 2.9 2.5 (1) The particulate in grey shading are mainly soil shake off from perlite. 153 Table D - l Suspended sediment sampling detail for the Brunette River watershed'by XIR' Brunette Eagle Still Ramsay Stone} Silver Date Tinie Date Time Date Time Date Time Date Time Date Time 12-Mar-03 1.2:3,0 19-Dec-02 14:5,0 19-Pec-02 9:30 :19-Dec-,02 13:30 16-Jan-03 14:30 l2-Mar-03 15:25 13-\iar-03 11.:00: 19-Dec-02 15:00. 19-Dec-02 11:30 30-Jan-03 16:27 16-Jan-03 14:45 l?-.\lar-03 15:30 13-Mar-03 11:15 30-Jan-()3 17:30 19-Dec-,02.. 11:35 30-.Ian-03 16:30 30-Jan-03 10:15 12-Mar-03 15:40 13-\Iar-03 11:20, 12-Klar-03 14:55 19-Dec-02 11:45 30-Jan-03 .16:45 3.0-Jan-03 10:35 12-Mar-03 15:45 13-\Iar-03 11:25 12-Mar-03 14:57 16-.lan-03 10:30 12-Mar-03 13:30 30-Jan-03 .17:00 7-Apr-03 ,9:15 13-Mai-()3 11:30 12-Mar-03 14:59 16-Jan-03 10:45 7-Apr-03 8:15 30-Jan-03 17:10 7-Apr-03 9:25 7-Apr-03 10:25, 12-Ivl.ar.-03 15:00 16-Jan-03 10:50 7-Apr-03 15:30 30-.Tan:-03 17:15 7-Apr-03 13:15 7-Apr-03 10:30 13-Mar-03 13:30 16-Jan-03 10:55 7-Api-03 15:45 12-Mai--03 16:00 7-Apr-Q3 13:20 7-Apr-03 10:35, 7-Apr-03 12:00 16-Jan-03 11:05 7-Apr-03 16:00 12-Mar-03 16:10 7-Apr-03 13:25 7-Apr-03- 11:15 7-Apr-03 12:02: 12-Mar-03 11:15 17-Jun-03 13:15 12-Mar-03 16:15 7-Apr-03 14:45 7-Apr-03 17:15. 7-Apr-03 12:05 12-Mar-03 14:18 16-Qct-03 '9:00 12-.Mar-03 16:30 7-Apr-03. 14:50 18-Jun-03 9:45 7-Apr-03 12:45. 12-Mar-03 .14:20 16-<)ct-03 9:15 13-Mar-03 12:00. 7-Apr-0'3 1.7:40 16-Oct-03 7:4.5: 7-Apr-03 12:46 12-.\Iar-03 14:25 16-Oet-03 9:30 13-.\lar-03 1:2:10 7-Apr-03 17:42 16-Oct-03 8:00. 7-Api-03 12:48: 12-Mar-03 14:30 16-CXI-03 13:15 13-Mar-03 12:15 7-Apr-03, 17:45 16-Oct-03 11:45 7-Apr-03 12:50 l3-Mar-()3 14:00 16-Oct-03 13:30 13-Mar-03 12:30 18-Jun-03 11:45 16-Oct-03 15:15 7-Apr-03 13:02 13-Mar-03 14:03 16-Gct-03 17:00 13-Mar-03 13:00 16-C)ct-03 15:05 16-Oct-03 16:00 7-Apv-03 13:03 13-Mar-03 14:12 18-Jun-03 11:30 7-Apr-03 13:05 7-Apr-0a 16:30 16-Qc.t-0.3 8:00 18-Jun-03 12:15 7-Apr-03 16:33 16-Oct-03 11:00 16-Oei-03 8:15 7-Apr-03 16:36 16-()ci-03 11:30 > n 3 a. S' O W s s n n n a. xn n a, 3 ft Table D-LSuspended sediment sampling detail for the Brunette Kver watershed by NHG (cont.) 16-Oct-03 8;30 7-Apr-03 16:40 16-Oct-03 15:15 16-Oct-Q3 8:33 7-Apr-03 i6:42 16-Oct-03, 15:30 !6 -0Gt -O3 8:35 7-Apr-03 16:45 16-Oct-03 10:30 17-Jun-03 16:00 16-()et-()3 10:45 16-Oct-03 10:00 16-Oct-03 14:30. 16-Oct-03 1,0:15 16-Oct-03 14:00 16-Oc,t-03 14:15 Table D-2'Suspended sediment trace metaTconcentrato al lower slop-log structure Sample Date Sample Time , SS^  Metal Cone, in mg/kg Metal Cone, in ug/'|. Cone, (mg/1) Cu Mn Fe Zn Cu Mri Fe Zn 7-Apr-03 12:30 9.9 7484. 3718 50.425 877 74 37 498 9 7-Apr.-03 12:30 6.0 8509 5695. 57332. -975. 51 34 346 6 7-Apr-03 12:30 7.6 4390 5589 54925 863 33 42 415 7 18-JM-03 1.2:30 5.7 1757 6045 6Q044 1209 10 34 342 7 16-Ocl-03 11:00 41.2 120 838 16529 270 5 35 681 11 16-Oct-03 11:15 11.7 607 1257 35888 464 7' 15 419 -5, 16-Dct-03 11:20 9,4 381 1495 43495: 559. 4 14 4.09 •5 16-Oet-03 11:25 124 304 1231 30182 528 4 1.5 374 •7 16-Oct-03 11:30 123; 6483 2366 3.4296 821 80 29 422 10 16-Oet-03 10:25 7.8 1959 .3689. 43898 79.1 . 15 29 343 6 16-Oct-03 10:30 10.7 619 2189 35530 313 7 23 380 3 B-Mar-03 10:30 .6:1 '517 4489 47738 516 3 '27 289 3 T3-Mar-03 10:35 '5V4 211 5087 53382 848 1 27 .287 5 13-Mar-03 11:15 6.2 328 397-7 52567 382 2 .24 324 2 13-Mar-03 17:15 31.2 129 2314 33189 598 4 72 1035 19 I2-Mar-03 9:45 .6.0 56 4504 122057 676 0 27 731 4 12-Mar-03 7:45 112.2 135 1803: 31473 258. 15 202 3532 29 13-K4ar-03 8:00 111.6 148 1703 25483 332 17 190 2844 37 12-Mar-03 8:00 292.2 69 1048 21762 211 20 306 6359 62 12-\lar-()3 11:45 185.1 7.2 1097 22182; 239 13 203 4106 44 Table D-3 Suspended sediment trace.metal concentration from Eagle:Creek Eagle Creek Suspended MetatConc. in mg/kg Metal Cone, in mg/1 Sam pie Date Sample Time Sediment Cone, (mg/1) Cu Mil Fe Zn Cu Mn Fe Zn 19-Dec-02 14:30 1:3; 13252 27703: 89675 528 1.7 35 113 1 19-Dec-02 14:50 2.1 307 45395 56082. 655 1 33 119 1 19-Dec-Q2 14:50 18 62.5. 195.43 62302 .535 1 35 112 1 19.Dec-02 15:00 1.7 642 20155 69071 2394 1 34 116 4 30-Jan-03 17:30 30.4 198 1603 36209 690 6 49. ioo 21 30-Jah-G3 17:30 33.5 151 1412 30013 .523 5 47 10.06 18 30-Jan-03 17:30 33.7 214 1478 1079 -84 7 50 36 -3 30-Jah-03 17:30 42.0 • 236 1755 36302. 638 8 62 1291 23 12-Mar-03. 14:55 51.7 128 2494. 35697 486 7. 129 1845 25 l2-Mar-03 14:58 65.8 132 2165; 31087 424 9 ..143; 2647 28 12-.\lar-03 14:58 59.7 301 2353; 5:1358:' 501 18 141 3068 30 12,3Vlar-03 15:00 62.6 140 2204 35381 446 9 138 2214 2.8 13-\iar-03 13:30 10.6 2209. 3268 43720, 9.70 23 3.5 464 10 7-Apr-03 12:00 8.4 116 4382 48211 1632' 1 37 406 14 7-Apr-03 12:00 10.3 1.8 4142: 44670 322 o. 43 462 3 7-Apr-03 12:18 11.3 35 3529. 42240. 427 .0 40 478 •5 ?-Apr-03 12:05 10.3 152 2904 4530? 685 2 30 466 7 7-Apr-03 12:45 25.7 134 2465. 47705 692 3 .63 1224 18 7-Apr-03 12:46 :24.5 83 2168 44420, 525 2: :53 1090 13 7-Apr-03 12:48 28.6 129 2258 41375 633 4 .65 1185 18 7-Apr-03 12:30 27.3 79 2441, 44523. 565 2 67 1216 15 7-Apr-03 13:00 28.1 226 2307 50584 5,97 6 65 1421 .17 7-Apr-03 13:02 28.0 '230 2306 51848. 703 6 65 1451 20 7-Apr-03. 13:03 '34.9 154 2014 43951 518 5 70 1535 18 7-Apr-03 13:04 28.3 194 2214 52710 616 6 63 1493 17 18-Jun-03 12:15 3.2 413 9690 80362 .180 1 31 255 1 16-Oct-03 8:15 192,6 97 3043 41870: 372 19 583. .8022 71 16-Oct-03 8:30 181.7 76 3126 4043.8. 393 14 5.68 7349. 71 16-Oct-03 8:35 180.6 130 3123 44087 398 23 564 •7963 72 16-Oct-G3 8:35 191.0 87 2962 38000 348 17 566 7257 66 16-Oct-03 8:35 188.6 7.1 3181, 41085 388: 13 600 7.750 73 l6-Oct-G3 10:30 173.8 326 249? 2672.0 345' 57 434 4645 60 16-Qct-03 10:45 153.1 .154 2305: 30841 290 .24 353 4722' 44 16-Oct-()3 14:30 238.7 49 1851. 25589 208 12 442 6108 50 TableD-4 Suspendedsediment trace mcial concentration.fromStill Creek Suspended Metal Cone, in mg/kg Metal Cone, in mg/kg Sample Date Sample Time Sediment Cone, (mg/1) Cu Mn Ee Zn Cu Mn Fe Z11 19-Dec-02' 9:30 2:3 .8556 34953 995745 15851 20 80 2284 36 19-l)ec-()2 11:30 124.7 135 740 38932 303 17 92 4856 3.8 19-Dec-02 11:35 14.8 266 4544 53335 769 4 67 789 11 19-Dec-02: 11:40 2.3 907 23860 9.8212 1313 .2 54 221 3 16-Jan-03 10:30 3.0 649 20458 13678.4 1302 2 62 413 4 16-Jari-03 10:45 3.2; 573 20951 153963 2122 2 67 495. 7 .16-Jan-03 10:50 3:4: 826 .21289 157192. 2352. 3 73 541 8 16-.Tan-03 10:55 3.1 690 18719 186367 882 2 59 585 3 16-Jan-03 11:48 4.0 5.61 14356 138432 1130 2 58 556 5 20-Eeb-03 1630 21.9 395 1637 49644 770 9 36 1086 17 2O-Feb-03 16:30 23.0 408 1580 46032 679 9 36. .1060 16 20-Feb-03 16:40 22.2 628 177.6 54.674 672 1.4 39 1211 15 20-I-eb-03 16145 22.1 414 1746 53990 982 9 :39 11.91 22 20-Eeb-03 16:45 .257 442 1680 4534.8 858; 11 43 1166 22 20-Feb-03 17:00 ie;6 336 2406 52209. 921. 6 40 866 15-12-Mar-03 14:15 65.8 352 1204 36914 717' 23 79 2428 47 12-Mar-03 14:15 63.5 375 11.28 33794 578: 24 72 2147 37 12-Mar-03 14:18 72.7 337 1220 36289 729 24 89 2638 53 12-Mar-03; 14:20 67.0 445 1280 38781 7.04 30 86 2597 47 12-Mar-03 14:25 75.7 468 1254 38500 641 35 95 2913 48 12-.\lar-03 14:30 63.9 554, 1075 32317 632. 35 69 2066 40 13-Mar-03. 14:00; 19.1. 4825 1778. 39209 956 92 34 747 18 13-Mar-03 14:00 201 1715 1506 35467 766 34 30 711 15 13-Mar-03 14:05 12.9 55.74 2154; 42114 975 72 28 542, 13 13-Mar-03 14:05 22,6. 452 856 33187 610 10 19 751 1* 13-Mar-03 14:10 21.1 915 1673 36078; 810 19 35 760, 17 7-Apr-03 14:30 54.5 222 959 25094 .534 12 52; 1367 29 7-Apr-03 14:55 41.0 332 1217 30825 639 14 50 1264. 26 7-Apr-OJ: 16:35 38.0 239 1298 29533 619 9 49 11.22 24 7-Apr-03 16:40 35.4. 200 1124 30656 551 7 40 1085 19 7-Apr-03 16:42 43.1 201 981 26038 .484 9 42 1123 21 7-Apr-03 16:45 53.2 155 .877 23946 459 8 47 1274 24 17-.lun-03 16:00 8.5 290: 7203 94964 700: .2 61 809 6 16-.Oct-Q3 10:00 128.8: 17? 707 19787 387 23 91 .2548 50 16-Oel-03 10:15 123.7 196 827 20025 461 24 102 2478 57 16-Oet-03 14:00 39.4 258 847 24076 409 10 33 950 16 T6-©ct-03, 14:15 107.4 1.70 727 18585 357 18 78 1995 38 Table D-5 Suspended sediment trace metal concentration from Ramsay Creek" Suspended Metal Cone, iii mg/kg Metal Cone, in fig/1 Sample Date Sample Time Sediment Cone, (hig/1) Cu Mil Fe Zn Cu Mn Fe Zn l?-Dec-02 13:3(1 4.5 159 .3043 2357? 673 I. 14 105 3 19-Dec-02 13:30 6.6 3797 2549 .24731 450 25 17 162 .3 30-Jan-03 16:28 268.2 33 648 1.8739 99 9 174 5026/ 27 30-Jan-03. 16:30 101.8: 59 847 19470 127' 6 86 1983 13 30-Jan-03 16:30 382.5 25 39.7 12882 61 9 152 • 4928 23 30-Jan-03 16:45 400.3 3:6 508 15806 73 12 203 6327 29 12-Mar-03 13:3,0 129.3 90 904 26132: 125 12 11.7 3380 16 12-.\lar-03 13:30 162.3 150 719 21496; 116 24 117 3489 19 7-Apr-03 8:15 33.4 8.1 1277 30.018 164 3 43 1003 5" 7-Apr^03 8:15 34.0 4.1 1174 33516. 239 1 40. 11.39 8 7-Apr-03 1.5:30 218.6 34 647 11633: 98 7 142 2542 21 7-Apr-03 15:30 84.3 89 754 17.194 161 7 63 1449 14 7-Apr-()3 15:45 133.6 38 548 11871 149 5 73 1586 20 7-Apr-03 15:45 8.1.4: 87 843 17397 221 7 69 1417 18 7-Apr-()3 16:00 119.3 .44 648 10233 124. 5' 77 1221 .15 7-Apf-03 16:00 62.4 60 768 19100 .202. 4 48 1193 13 17-Jun-03 13:15 5.1 454 3028 53417 48 2 16 274 0 17-.Iun-03 13:15 5.5 988> 2188 44627 531 5; 12 .245 3 16-Oct-03 9:0.0 380.4 53 1545 4212? 190 20; 588 16027 72 16-Oct-03 9:00 561.2 16 16-Oct-03 9:15 344.0, 32 16-Oct:-03: 9:15 703.6 21 K-Oct-03 9:30 322.9 35 16-Oct-03: 9:30 1376.3: 284: 16-OcM)3 13:15 1376.3 28 16-Oct-03 13:15 3096.7 14 16-Oct-03 13:30 2479.0 16 re-ect-03 13:30 1305.6 7.6 16-Oet-03 16:50 507.2 43 16-Oct-03 16:50 1304.4 16 16-Oct-03 17:00 1187.9 44 16-Oct-03 17:00: 520.2 49. 479 12729 62 9 269 7143 35 1061. .27345 113 11 365 94,08 39 661 21203 80. 15 465 14918 56 945 29842 95 1.1. 305 9637 31 673 23763 .62. 119 282 9956 26 709 19582 81 38 976 26951 112. 360 11936 45 44 11.16 36962 13.8 369 11005 56 41 914. 27281 138 811 19302 70 .34 1058 25199 92 444- 10546 65 ,22 225: 5349 33; 250 11035 31 21 327 1.4.394 40 422" 13265 52 52 501 15757 62 528 17799 76 26 275 9260. 39 Table D-6 Suspended sediment .trace metal concentration from Stoney Creek Suspended Metal Cone, in mg/kg Metal Cone,in ug/1 Sample Date Sample Time Sediment Cone, (mg/1) Cu Mn Fe Zn Cu Mil Fe Zn 16-.Ian-()3 14:30 10.2 161 1327 46905. 369 '2 14 481 4 16-Jan-03 14:40 4.1 182. 1759 52507 206: 1 7- 213 1 16-.Tan-03 14:40, 3'.7 175; 2105 51652 200. 1 8: 193 1 30-Jan-03 10:15 5,7 311 2097 41367 .514 2 12 236 .3 30-Jan-03: 10:35 49.6 28 115. 9454 92 1 6 469 5 30-Jan-03 10:35 4.3 •365. 2493 398.85 .974. 2 11 1.73 4 3.0-Jan-03 15:00 120.8 81 659 16230 199 10 80 1961 24 30-.Ian-03 15:10 117.8 92 646 .17208 236 11 76 .2027 28 3()-Jan-(.)3 17:15 1.85,9 55; 429 .12397 1.32 10. 80 2304 25 12-Mar-03 16:00 374.9. 58 615 37894 140 22 231, 14208 •••53 12-Mar-03 16:10 194.5 93 •828 23738 191 18 161. 4618 37 12-.Vlar-03 16:15 359.7 49 .516 .32357 124 17 186 11640 44, 12-Mar-03 16:30 273.3 7.9 639 36796 142 22 175 10056 39 13-Mar-03. 12:00 27.8 1309 927 1.5933 282 36 26 442 .8 13-Mar-03 12:10 44.0 379. 662 13611 224, 17 29 599 10. 13-Mar-03 12:15 55.5 102 654 16.08,5 236 6 36. 893 13 13-Mar-03 12:30 157.1 86 665 18408 189 14 .104 2892 30 13-Mar-()3 12:55 194.7 75- 48.0 15,129 111 15 93 2946 27 13-Mar-Q3 12:55 .132.6 113 .622 17165 185 15 82 2275 25': 13-\lar-()3 13:00 155.1 4.79 618 17298 189 74 96 2682 29 18-Jun-03 11:30 1.8 652. 3913 605.16 1433 1 7 10.7 3 16-Oct-03 8:00 57.3 94 777 16279 183 • 5 44 928 10 16-Gct-03 11:00 437.0 28 507 15617 94 12 222 6825 41 16-Oct-G3 11:30 960.8; 23 48.4. 14:722 89 22 465 14145 86 16-()cl-03 15:15 1457.2 ,21 382 11208 75 30. 55:7 16331 no. 16-Oct-03 15:30. 1133.0. 21 3.81 ;2494 81 21 43.1 .2826 .92 Table D-7 Suspended sediment trace metal concentration from Silver Creek Suspended Metal Cone, hi mg/kg Metal Cone, in ug/1 Sample Date Sample Time Sediment Cone, (mg/1) Cu Mn Ee Zn Cu Mn Fe Zn 12-\<1ar-03 15:25 73.4 119 1062 24735 344 9 78 1815 25-12-Mar-03 15:30 122.7 130 846 16594 272 16 104 ;20.35 33 l2-Mar-03 15:40 68.6 134 1202 28281 418 9 82 1939 .29 12-Mar-03 15:45 75,6 213 1045 24526; 329. 16 78 1840 25 7-Apr-03 9:15 22.9 219 1964 30627 891 2 22 346 10 7-Apr-03 9:25 6.8 40 2468 35459 946 6 17 240/ 6 7-Apr-03 9:25 9.7 41 1780 31558. 656 0 17 305 6 7-Apr-03 9:25: 8.0 54 1595 36261 .1.017 0 13 290 8 7-Apr-03 13:15 26:7 139 1117 31614 •624 4 30, 843 17' 7-Apr.-03 13:20 22.5 167 1006 29706 579 4 23 .670 13 7-Apr-03 13:25: 22.9 131 978 29095 570 3 22 667 13 7-Apr-03 13:25 23.7 168 1015 33118 672 4 24. 78.5. 16: 7-Apr-03 14:45 33.7 156 1057 '2618.1 585 5' 36 882 .20 7-Apr-03 14:50 31.0 124 1025 28102 474 4 32 8.70 15 7-Apr-03 14:50 29,6 283 1060 28681 541 8 31 849 15. 7-Apr-03 14:50 27.8 91 1041 25610 420: 3 29 712 12 7-Apr-03 17:38 38.5 75 973 24372 542 .3: 3,7 938 21 7-Apr-03 17:45 46.0 68 1022 27020 421 3 47 1242 19 7-Apr-03 17:45 48.5 214 885 ,21858 .350 10 43 1061 17 7-Apr-G3 17:45 36.3 151 1094 2.7831 41.2 f 40 1Q10 15 18-Jun-03 H;45 53.8 37 1245 1030.77 1457 2, 67 5549 78 18-Jun-03 11:50 38.7 .679 1164 91501 1221 26. 45 3545 47 1.6-Oct-03. 15:04 219.7 59 865 23946 170 13 190 5262 37 Bedload Sediment Trace Metal Table D-8 Bedload sediment trace metal concentration Location & Date < 63 p.m (mg/kg) < 2 mm (mg/kg) C u M n Fe Z n C u M n Fe Z n Stoney 10.29 228 2603 135008 209 116 281 21355 219 Stoney 11.13 322 680 43748 212 5 238 15301 48 Ramsay 11.20 86 559 36654 107 19 300 23020 51 Still #1 175 503 39225 327 73 195 15796 156 Still #2 663 1452 64052 637 75 296 25570 127 Still #3 390 543 40279 449 89 252 19948 155 Sampling date is not known for Still Creek 167 Appendix E U B C Bus Loop Stormwater Data Stormwater Quality I able H-l Storniwaler quality/data from three drains in the bus loop Catch Basin #1 Date Sample Turbidity Conduc. SS Total Metal (ug/1) Dissolved Metal (Ug/1) Metal/TSS (mg/kg) Time (NTLI); (us/cm) (mg/1) Cu Mn Fe Zn Cu Mn Ee Zn Cu Mn Ee; Zn o 15:50 148 132. 54 SO 90 213 199 34 * 128 24 1540 6.603 7876 1623 7 'er 6, 16:10 95.1 127 47.5 .5,5 82 3407 •9.1 18 * 26 28 26.48 4759 243047 4552 'O 16:30 58.5 117; 45.8 80 75 1812 75 29 39 * 32 1319 929 47652 1103 o 16:50 61.6 132 40 14 57 1794 94 13 33 484 55 104 2150 116843 3531 10:15 988 124 '5 30 • 263 4.2. 32 * 38 1541 - 66833 1450 11:15 12.9 122. 14:7- 77 * 422. 66 3:6. * 23 1937 73550 8164 -S' 11:45 17 138 13.9 46 •'* 513 25 15 * 3 2416 - 118034 4.841 11:45 18.4 13.8 111 26 .* 475 19: 23 * 31.2 9 4490 -- 450.60 2794 'o .£> • o 11:45 17.9 138 18.1 39 36 1786 60 110 * 292 36 -- -- - --Octi 13:30 " 25,7 193 15.8 163 2.8 979 45 25 * * 30 6217 2407 166575 2748-13:3.0' 26.2 193 15,3. 61 * •534 41 20 * 40 2087 -- 91425 2300 13:30 26.3 193 20.5 31 * 549 103. :40 31.3 59 6277 28691 5300 691 December 13, 04 Ki)vetiiber 14, 04 CO to to I—' to •—S P' .!—; P t o i—» :00 1—•': ON. 45, (5-i CO t o '-;' i i o Vl V» to .© 45-: vi P P O' P. CO p 45. w »—• v o p Lo p Vl v*. © : '-O C O ,45-•so, SO so Vi .45. O, CO 45. t o Vi OS -o. ON: Vl 'ON -g © ! Vl :so Vl (—• Vl •ON Oi. 45. 45-45. CO 4 i 00; ; o o -CO 00 45. 45. : C 0 -•oo-to 45V - O :tO '-j to: P, 45-,t—'. 1 45-P o c ,ui :so' h-i' CO p P ON 45-t o 4^ 00 UP j—' 'Co' . <° Os to .©-'00 r P OS t o 45-p 45-c o to ON '•50 l--» t o t o o o - o . OS' 45-. <1 t o 45-- O t o to b o to ON to ON V? |5-. -45>-i—• - J to sp t o t o 45-OY o b CO co:: 'ON-CO 45-CO CO s o 45. <!' tO-CO AO •42. to 4> 4 i SO C o VI SO 45. VI to to ,45.. l iO: to Ol ,ON .Ui: to 45-t o 45. to SO 13101 ••>« 00 ON Vl oo-t o . so o\ - J OO OS OS :O0; 45r t o : t o OS to t o o b ,vi o to Vi* Vi .Vi Vi :ON SO P . SO' OS ;t° xo-X f t : OO-tO:: 45. -4 -J P t o ON to ON - J Vi p -O. 45-vi < V*: ; V I .Cis-c o OS-P t o t o : >J Vl ii>i Vi : t0 OS N'' Cu. 00 t o •h— 45* t o 'S .Vl ON O' p O: p P p. o to 45- P Vi: Ul ON; P 00 p p p o- © P OS 45- ON O * CO so P >4 ON -so to 45-1-5-. OO. Vl CO Vi CJ t o to to SO Vi b o p .SJJ: Ul to OS ON CO t o P b o : Vl-. p •o- 45-to 45-* * to to •o'-to 45. 45, CO o Vl Ct» CV o :CO © 'Co: P. o b o © ' o. NO P. P .—., to o o Vi 45. LO h-• so ON CO.' 45. 00 Vl 45-p t o to p t 0 ' .00 o SO p IO-. P t o . SO p t o v> :P t o 45-to SO ,45-. CO; to ON CO t o 45. t o 45. to so CO-to o © CO o t o -c/l. AO' ' © ' CO OS .to Or P 45. : t o P P. P CO 45-P Vi o : co Vi 45. OS p t o OS t o : CO OS -4 V» P •'SO' •p! so OS CO t o :CO. L/i-CO. to.. 45. - O - O P t o OS to CO ON -4 t/i ip 00 o : tO Vl O 45. o c -p to, 45-O to NO-O : CO H-» o ,co ;p h-' 45-yi t—1 VI OS • to co. 00 to .45-' ON to 45. Vi Vl ON October 204 to h-5 p sp •so. oo 1—', Vl CO CO O P Vl'. 4i L/i -Vl. to P :o: ,p to p p p so' OO to to '45-Cs ON' VI to to .45-Vl to 45. 00: to --4 Vl 15.7 -4 to h-« O-. OO to 45.'. :Cs-to SO •so: to. to p CO Vl eg so to. V» bo oo to bp to ;bo OS 4=5 so OS so oo to SO v« ON to oo .to -j tO: -o •i4 ON so :ob -45.->—»: i—i P -J o to p to •-o 'to-' so-so p to 415.5 p-r*. tw CO. to to so Vl :45-SO ;Os Vl •o 45-to 45.' to p ids to. i—» Co to to o-. 45. to-00 t—' -o. 45. to Vl ON-45. Vi ON to © -o to -* to to-to •to to to to so to 45- ON to to to to ,cs. i.o 45. so t*N o-45-so to 00 to 45. .45. .O. ~J p oo to' :Q\. OS 45-;vi p ; to VI oc Vl so ..* v » CO ON -I—1 45.-so p OS '<4-os OS-45-; SO .-o. to NO p '<r w p so .so o to .4^  oo ~J o Vl o-•54 NO Vi 45-p -so p tu: P- p -Vi:. O' CO to, o to-• © . -I—'• os p bo p 5.190 ON-•P to OS 45-o OS to-•:'© so to 00 p to .ON. to-p cc so p 45-to p to -O o ^ * -P V. p ;fO so . © . ~4 Vt p so p-45-p to to p 0L\ March 16,05 to u i to to I—* ;cs u» 4* : t o 1—* o o '••© o O : , o Ul i i i : o : .4* Ui s O : p U i 4* :~4 i—> os si so- OS UI. OC 42. tO: © os: u> OS OS SO' •o. t—' so UI u> Ui UI o i—i to CO •bo' SI Ul o o Os-l o : U i U i '©' ;to <©• ! © oo'' bo o ; Ui © to; I2J. • o ; U i o o OS so. to: t—. to. to •si 42. uv U i : !Sl rSO to', to; UI so to U l u i : OS. to o Hi :sO SO OS ©: i—* O' to >o U l ' to © ,4^ 17231 U I ' u i 42.: OS U) oo UJ :Ui •os; ;Ui: :oc-u i : es-se Ui OS Os so to - u i : :©; 42. *—' to. 00 SI OS to o Ul to - o to © so to U i - j to SO: •^ 4 oo to '•fe U i to 42-to U l © Ui oo 42. U i to U> 4=. - o o >2-l © -42. so to I—* ;Ui. •if- to to ' to U i so J—' oo. o ; t o Ui , to to .42. OO '42.; •Ui: to o p to •so; so 1^ os Os Sl © os OS • o o 42. to U i 42. Ul to .to '.OS SI Ul 00. Ui'' o -to OS 42. tO: SO to OS 2»-121.4 42. to as. to so I—* to 42., -o u i 42. U i U i so IO U)' 42-O 86156 to . to so-Ul tO. U i 42, © 42. 42. •42. 42. to O : SO: oo: to to U i © po to .OS so; 42." to 42. Ui SO 1.017 oo-'42. OO to SO oo :Ui SO Ui SI U i Ui SO .ui Ui. so 4=. Ui Ui Ui March 8.05 Febmaiy 28, 05 0:30 23:30 22:30 21:45 21:00^ 11:00 12:00 10:45 9:45' 1—1 4-Vi, 12:15; 11:30 h-i. Ui Ui u i © 14:30 so SI U i 42. l o i—> Ui. to o 42. so to :OQ 1046 1114 1418 - o :© . •so; CO os' i—i •~4 OS OS U l os • © .OS S| •OS. 1.74 i-T 1 ' Ui AT-;H*„ LZZ 23.8 to U i . oo : © 4=* •^ 4 ,-0: SO: OS Ui Os 3.577 3456 3693 ob OO: OO :00 U l 42. :S|-- i l to Ui to Ui: Cl-42. Ul -4 4^ SO . O 4^ so o CS to © U l 4=-© U l CO © 4* O © . to : © : ©: SI 00 i—1 © . © 42. to u i Ui OS so-os-es :.u> 42-U> U l to; p : to 4i' OS tO: U i : S|. to U l so t o OS oo -OS. •CO :QS Ui © : s i U i h-i:-- o .42-ob Si: 4^ U i 42S to ; UI OS Ui Si •OS Ui: to oo Ul so 4 i , 00-•ui Ul oo Ul to Ul 445 so so' .4455 3964. 5087 42-00 4* ob. 19308 4082 2833.0 29006 386.11 1912:1 21939 22013 108.30 3459 3329 to. U i : to Ui to U i 42; © oo :4i Ul\ ~j Ui to CO S4 SO 00 U l Ui 4 i : £ ob to OS 00 Ul to SO Ui i—» ~-4 .5-S t - j t o Ui '2 to" sp-U l - o . to I—1 u i U l to sb : OS: - J 00 Ul }—* 00 Ul p p tO: 4 i Ui O. to to - U i OS tO oo U l to U) OS; to 4i- to so •1—i ;i-r-; :© . © • 1—1 Ui 4 ,^ Ui SO to-© • c s . to i—• to U ) OS SI >— -OS to to Ul SI son t o . ISO Ui to so 4i-• © ; U l OS : l — U l -os OO-:oC, Ui U i OS, SI o . © o so oo. to to 4^ © U l --4 - - J I—1 • so OS os OS so to U i so to .42. to Ui -~1 to U i SO 42. i — oo; U i SI 00 42-o to Ui o 4^ to U i S) to Ui. to U l SO 4 i M , ob U i --4 OO 4 i o to Ul •©: 4=. - J o .Ul, © -4 SO: 00' ob u i 857 00 Ul o : 7,8.8 i— U i U i 4 ,^ oo . © -4 •SO' CO. oo U l oo '.Ul :S|-ob Ul © SI oo oo 1334 to 42-© so: •©• 35517 37366 33743 34641 33020 32491 34690 •35517 37366 33743 34641 33020 32491 44250 •33290 1076 oo o 00 oo. 1—' SO 1283 1028; 00 00 o so to 1076 oc 00 oo so 1283. 1028 oc 'bo O to to © .I—*'" O ILl April a os t o . t o p SO J—l-00 OO -4 1—i - J i-o. : o y i vo p 45. o . 3 45. Vi-t o © © co VO o vo ON 45.. OO S I o c :6o: .ON V 0 S I >-V I Vl' S I O S I ?^ SO. so: i t o '45-ON t o t—I 45-VO t o Vl 45. • o o O o o 45. S I V l . 00, o -:vo 00 45. SO: .45- - o O : S t O ; Vi; t o . so V i .ON vo os Vl ;-o. -O. I— o o 6c Vl Vl: <sj; S I ON' Co S i VO so' Vi" t o © Vl ;l-5.-. vi; SO; so 45. , -o . C O 3489 V i OO OO t o t o 1—I Vl :so o Vl: 4>' ON: 45. Vl o © Vl O0: •© Vl co - o V0 - o 00 t o 00 ;SO OO so" vo- t o o o t o S ) ••si 45. t o - O VO t o 45. 45.. .©: v» © vo •oo 45.' o o vo t o so t o 'OS. Vl 45. t o Vi t o ON. d o t o N t o ; VO 45. ;vi NO' so ON VI 3.63 vo V l - o . .45-o o ON: t o t o VO ' t o ' S t t o o t o SO Vl SIX Vi o t o V l .so H-i* ©' p Vl ;co so -4 45. :SO t o • o o o 45. V) t o o VO i—1 o c o 45-.45. :.VO t o VO S J 45. ' S 4 i - J •as VO oo; VO vo VO C O t o V5J so-so t O : v i - o . t o so ON t o so: © S 4 t o t o so 45. p - o - 4 .1—. © o o 45. © t o t o S I CO V I Vi © VO •vo-45-t o t—' - 4 © ©. ON © © 1612, ON VO o o NO NO VISI-ON o o 45. o o ON ' t o •45.. SO 45->4 45. t o ' H -;so Match 31; 05 12:30 c o : 10:30 9:30 8:40 7:30 6:50 6:10 45. ISO .SO C O -VO' V l to; -si S I •vo-- S O ' 00 V i so 45. NO 45-to )—' © VO V l CO, 4> ~4 Os: - P -t—1 V l i—'. to >5-i : 00: V l VO O to-O0: -co to ON' S I ' V l 45. © - V r so • © • i—I 45, h^, OO P . 45.; 00. © to ON vo-' © tO' © >4 to V l 00 45--oo, ON vo V l ' S . V l © -to V l O v to so-VO ON 45. - J . V O 1 , © •so 1—1. to 00 :©• ; © ; to: 00 45. VO VO 45. .NO.-NO ; V O oo to, S ) 45-NO 45-to 'so' -it * V i S / i So 45. NO h-' 45-:ON V I (0 f—* so NO. 45-© to. © to so NO C J ON' ON ON to V l Ito' o v V l to 45. oc vo. 45. to. 00' ON co to 45-45. 45-s i 45-ON V l -4 to oc co (iO; so; to to to i V i to oc 45- .45.-to to 00 Co -4 V i NO CO © ,C0' 00--o.; vo, vo 45. -4 -4 -o. ; O v •ON: to V i 45. VO. VO © ' -4 to © ' NO CO to © to V I so 0' to ON V i SO v i to ON to V0 C O ' 'Jj' OS 45-oo :'to; v » to to 45. V I NO- V i ON v i vo C O ON 45.: 45.: t5-»' vo ON C O S I to ON. to 45.: V i -0. to, OO to SO: V i ON NO - o vo V l to S I V l to, V i VO' NO s | so Oi © C O 'ON •54 45-to vo CO oc NO v>. 'to-NO to to 0 s t 45. 00 oo oo h-' V l oo VO-. .©' NO 00 00 -4 -st vo to S O 00' M a r c h 2 6 , 0 5 v « 45.- co to: ; , ' . © vo 0 vo © ' V I vo © . V l © to p ' © © ' vo p 45. S J ON NO ;ob' co: • - © t o to VO' ON-t o ON t o 45. to :SO 45. 00 ON; •00 ' V l so VO: I—' •1—1 NO - J .00 ON :l—• 45-. © to t o , ' P 45. so t o . s t ob 00 t o 45; ON' 'VO. Wi © :6b t o ON .45.- 45. V l ON 45. VO. ON? 45. .so i—' to so o c 00 to 0 S4 tO: i—'' V l 00 s i ~4 0 ' to 342.0 V i 45.: ON. : v i to •to-V l 00' s t 1-5. to 45.' 45. to V i vo 4^ ON ON vo NO' vo '00 45. .00 oc SO. P : to V l to to,, vo -.si. vo vo 45. 00 to 45. 00 4^  :VO :oo to: 45. to to: P-ON ! ° 45., vo-i5-l' 45. to .0 ON 45. :P> vo. i-T' ON to © Co I—? SO V l K—' ON' CO :tO 45. -00 'si vo to to so ON to to V l : to i-ri CO NO to. : © V l 4=t p ->J t o . ON I—'. 45. .00 to to v< t o vo to • S to so; 45. V l to (—' :oS, NO-to s ; ;SO S ) .00 45. ON •—' 45., © : V i : NO -so - v .00 V l 00 45. to 0 t o . ON 45. t o Vo 12214 to. 45.: ON O 45. •h—' to : V l 45. V l S I , t O ; '45.. to 45. 45. OS :45. '45. S I vo' W V l V i :ON so V l :oo ON ON C O 45. © ON to SO '©" -1 A O . vo vo to vo ON 00: oc OS 00 vo 11:40 338 220 225 224 661 ,??83 1609 230 .520 5874 1601 623 1836 27730 446? 12:00 ,137 387 125 198 526: 7579 i;19'7 134 373 .5975 1019 9.91. 2630 37897 5983 >/-, 12:30 86 403 125 188 507 5858 .1497 168 417 3331 969 942 2533 29290 5983 13:00 77 345 100 158 474 4910 .1033 105 342 2996 644 985 2964 30690 6458 14:00 58 256 75 144 456 3838 902 93 296 2347 526 1196 3796 31984 7514 15:00 56 187 63 131 445 2249 856 76 289 1573 • 477 1310 4451 22490 8564 3:3(3 26 103 40 75 5.6 481 297 57 46' 322 .2.79 7.52 55.6 4812 2972 4:00 17 91 20 195 60 573 870 58 47 184 303 3910 1192* 11458 1739-4:40 38 91 35,3 73 66 2359 253 67 47 297 289 83 75 2670 287 5:20 11 77 33 269 28 872 1241 37 14. 95 120 3225 335 10460 2890 10:10 131 109 57 120 ,94 1909 414 125 72 335 330 1494 1169 23863 5169 o 10:40 130 109 214 180 179 6696 414 142 79 346 .387 601 595 22320 1381 11:10 183 116 .157 195 181 4525 449 138 .90 481 443 ,887 824 .20566 2043 ! 12:00 147 118 300 227 34:6 .13235 642 .1.48 ,145 ,429 455 540 824 31511 1528; 13:40 375 136 500 277 524 21569 1164 190 194 631 1054 396 748 ,30813 1663 18:20 189 187 147 153 273 5975 963 106 118 308 610 6.95 1239 27.159 4377 . © ' ' 1.8:40 .126; 161 80 153 181 3311 588 .205 99 34? 588 1272 1511 •27595 4902 19:00, 70 87 133 78 108 3503 535 •53 46 401 174 392 538 17514 2674 •eft 19:20 42 76 •93 117 190 13684 474 38 14, 248 169 836 1360 97743 3388 :M- 19:50 47 60 160 85 111 6285 .275 33 1: 219 117 354 463; 26188 1144 20:20 38 65 67 84 81 35.42 177 26 8 438 96 836 806 35423. 1772 (1); * s t a n d s fo r < de tec t io i v l im i t £L\ Os © October 29, 04 to . to I—' o so © o Ul. t O U i OS' so SO © AO -O I— to CO to tO 1*4. CO © , © -: © co o © •©• 42- U l .S0'; *»; •SO.- o to . © to to. to •<«• tO :'— S) (O 42.. CO U i S g ©. •© Ss. to U l © ,00 so. - 4 ' l - ' r to s i © , © OS 00 O - J . © © 'SO ; CO 86.3 oo' to to • © : -'00 to: 1—1 ©" •>©• •oo. i—' to •+¥ to 42. : © to CO op OS to • bo. •OO; 0 0 .©' OS. 'S| ob CO •.•>! SO oo- - o to to to .to S I :bo-• s i : .si1 OS 42-to o -tO Ui U l CO 42. U l Cs .Ui 'so; to-co OS ,so S I SO SI : C 0 . :CU :m - o to. co i—' .SO U i :OS OS OS 42. CO OS Ul - © . 2-4 t o 0 0 S4 42. ••<*»• oo U) to so, CO OS to . U i bo! Os -*, OS. to CO to CO to - o to <*! :co U i to. C^O SO to 42-. OS -1 Cs to Ui CO o U i 42. oo to to to CO to ;ui , © ;to-U l : CO SI OS © CO. 42. OS ;Os •00 42.. Co. Os ,Ui co 42. 42-U l o Co 42. © 42. to. © • CO OO © •CO 42. © ; OS S I © CO CO © o OC o to . © . © CO Ui . ' © to S I © •co'-© © to OC o :.CO © SO © 42.-o October 17, 04 t—•• r—1 1— 1—' _ H- ' I—* ,_t I—1 ; p : ON u» >-* u» :-UV'. J-V U i p P u» U l ;uv • o O ' P p to co. .oo so; oo co s i co 42-u . SO to SO to Op-OS' CO. CO I—' to so os to Ui * . CO so :so to to 42. os 42-U i ; © Ul OS .SO so 42-to OS OS o 42. U l ^4 CO to 42- CO Os OO Cs U i © to' U i •o to :tO Co so © SO os to 0 0 1 OS . © © SO to 42- Ui to cs U l ; © oo ob 1—1 • to Ul' OS to - o co: 42- so-42. © OS .to October 6, 04 :00 : H - I--CO P • © • tO: to 42-'. 42-42. SI Co OS' O0 ,11; to-s i : CO tO: •o •0s: 42- O : U i •OS; -to: .Ul. CO 42. Ul co; SO Co CO 42. 42. Cft, © '••si. ' , © Ui: 42. 42.. 42. :.~-J Ul •os- © ^ ' tO co; 42. s i C \ •42,. 4 i U i 00 tO; SO S i OS to: 0 0 to bo Os U l . os' to to CO to CO •00 o so ; Co .42.' 42.: U l • o S| © .Ul 42. -42. 42- ,Si: U i OS © 1 1 to •:co 42. 'SJ; OS -42> to to. !i—' 42. o . ' U l . 42. oo to to -Cs. sO : 0 0 42.. " © • © : 42. i—. OO 1—>. Ui o \D i—i i—i. Ui )m~L. HI to to © • February 28, 05 o to ON' to 42. to to o to s i to to 42. :tO NO to-• o 42- Ii. 00 to 00, 42. 42. O £ to to 42. ON V l VI 42.-, 42. to lift. 42- 42» 42. s i to 42. V l •ON ON s i S I 42- o - 0 0 :ON tO. >-? tO .SI o o to o to to to to 42. 42. tO :ON o to to. 42- to oo V l CO V l V I to o ON; V l >2-< to 42. S I t O to ON 4^  to l—» > J O -oo OO'. W\ •oo 'OO 1—' - 4 .ON O . ON -^>' V" NO to to CO i—• 1—. ON. v> o vo; to 42-ON. to -~1 V i to 42.. 42. to ' ' S I -~I ON s j NO ,42. oc s | to 0 0 to to 0 0 o .so OO © to :to-: to V l 'ON; -j to 4V ON ••Si- i—' 42. to S | December 13, 04 to. o ;•-•: to J _ >—' i _ . t o to : © • o oc ON VI, 42. to to. o i-i. •Co 42. o , o to 42. :*-!. P CO ©: V l • p V l . p . P p . VI V l V i . P p . ,c>. CO •(__., to H-» S | .;—i to s 4 OO - p . CO V i 42. to to ON to ON O N to 42. ON NO s ) to NO-to to o ' s | NO £ OS-to P 42. 0 0 NO P CO- 42., 42. CO V i NO ON VI'. V NO: NO; i d , S | S I —' ••si to '.<£>• V| : 42-p : V l o . : fo ,'h-i to: V l ob to. 42-: t d : OS' :Co 'to- 42-. to to- to ON • N0. to :(—' ON, 0 0 to ON. NO tO NO to to to V l 42-S I s i oc to to to p ' O P November 14, 04 CO: to V I ON. NO: P : to 42. NO 1— NO 4a. CO: to co . P . to. to ON NO. i_4: ^ :NO- ^ to V i v,: V i 42,. 4 ¥ 1—* >—' 42. i—» ON 'SQ' V r —4 to ;ON to so oo OO p ' S i 42-:tO ON- 0 0 1844 • to to iOV 2693 3933 3021 6088 4333 4987 4=. 42. IT- -:Ui: V l to I—1 (-T- -tO- V l ,h - - s i SO OS- CO * •«•• -O * OS: t- to o N' co; V i to ' o ON 42. ON w (—1 ;to to- f_ 42-ON : V l V i V i to o P : . © • OO to NO v i O O O . O 0 0 : S l , -to O ' to tft tii' p © : 0 o o • © : • P to; 42. to to © 42. to; s i ON' ON' S J NO. so oo OO - P - J ON. P ; oo- V l P > P p P . i o ••p p i—. V-*' 42-- to :tO to to ON ON, s i p •'Si" to P ,to- NO to ' 1— to p : i—' to :.Vi'- V l AO 42.: to oo .so 42. V l . 42. VI. to: to P : V l p • o , p ' O ' P ' :.° P : p h-1 1—I l_ so 42. to to- Vt . 42- to p P p • p P O p to to •tO ir-i to so 42-42. ON to i— :?o SO 42. 42. : p 42.; to p P : p : P o : o March 26, 05 42- C(J.- to 1—' , i - - - 1-^  co ' © :'i-^' 'Ui: CO o :ui © , to ' © . © ;© to ;co. .42. s 1—'• O S CO oo so u i C O :NO bo' CO So Ul •.CO y i ; s i U l co U l vo :ON. CO to so OS OS CO •OS .4>, NO S I to. oo to to t o to, CO C O ON t O to to to ;co ::00,-:O0'i to CO 3016 b -- o ; m s i 1—1 ISO :4i :OS U i Ui-' U l U i O S C O 0 0 :oo NO: to CO 42. NO ;oo to co: OS to to' to O : to s o to c to NO - o CO -c/r U l to o ItO' 'OS. 1—i .ON to 1—' -CO. O' . CO to CO u> CO 4i CO 4V -ON 1—I 0 0 4i' CO S I ON : o CO to. to :© to: t O : ' S O so o 6b CO :'-0 .1--. 4s.' o to to :0 COO': OO '.42.- 42. 42. o S I ON ICIJ; U l , co-co 1J 2X •00 to SO o U l >J • o SO 1—' V—* oo ,to, -NO-•SI C J so o : OS CO OS :4n-'. CO 1-21 to to NO S ) :CO-•NO". . © ' • OS CO Os' to .Ul 4a-CO i—»'• . to CO o to March 16, 05 0:30 23:00 22:00 t-—• si o o , 16:15 15:30 14:45 42-CO CO 0 0 Co co CO 42- 42. CO Ul U i 4* S I :so os - j t—' NO • o U) '42-' so 122*' CO to 1—I to OS .i—'• •oo-,1—' so p ; •CO © U l © ' 42. 42. © i—i: o O : U l © •CO 42i 42. 0 0 CO GO; 42-. U l 42. so-42-Ui: - J oo O .42. os o Ul- U i • © : © OS O ' U l oo o> •CO' iOS' to SO. so CO. 1—-CO CO' oc to U i u> U i : 0s. to S i S I C s ' CO to co Ul' o OS © ' to to 1—' :i-2-.. •SI" to CO so' cii o •CO . o to -O. to •oo Co oo CO s-O. CO 42-to • • p NO to to Ui s i S | ; U i .1—1 -as CO. 1—2. oo p I—* CO' © 42. - j *—I so . © U l OS CO © NO o •ui 'OS S I U l U i so to - o 42-to ON U I - o 42-to oc U i -OS to to CO to U l NO 16147 CO CO U l SO' - o U l i—1 CO:' to so SO •NO U l Co © to 'Os to U i to • © ' oo to 42. OS •CO; ! co CO oc oo © ' o S I U i CO :co U l 0 0 oo CO to NO U l OS so Ul U i o March 8; 05 !—• to . © - to CO to to: to 1—» to I-- I-"-© • O -o p : CO: P'; CO © •CO: o 42. u> o o © ©:. - o o S I ' 1—» 1—' 42-C o ; 42S' OS CO • p ^1 NO. 42;-J—1 • ;to b o o b b s -co 122-U l CO O N O s , 12-U l 42S 'O0. -20 to. 42. i— os :cv, p SO. U i os; • o o : © os o -•'os'- S I :'l—' :Oi 0 0 to »-*, , U l to - J so 1—1 to 1—1 • OS: C o S i :UI-: . U i : CO to : s j . 0 42, NO CO .-©•• NO, i—f O s U l CO: ,4V: U i ' o :Os S I o O s •so CO 42. 0 to ' © ' . © 1—1 o 0 0 U l ~o to, CO: SO CO to 42.. : o s . to to s i 4=2 CO OS U i . U l OS OS 42-S l 0 to .42-©' o o o to •OO: o b s o i—I o c to to so - 0 ,0 © to OS; -42-;. to 42. o NO 42. ,OS; OS NO t O : OO: ;to; CO-to ,'h-• S I - o c 1— o 42.. to © to © to to to to ••si. OS C o —1 to-co Co to U l to C J OO' © ' o o o b OS ' © ' : U l ' S I - o U i '•*>: - J 0 0 U i OS NO CO 42. NO 42. CO - o NO to to OS • S I OS to 42. - o i—* 0 0 © ' s i so ,1934 U l U l - U l U i >J. 42. os o o so - o o 42-U l 0 0 42. 0 U ) 125401 CO CO S I SO S I to : CO. to: ;'42>, '•Os; C s 42. o •CO : U l OS . : o ' : - Q 0 0 to NO-CO to-ut S O ' ; ^ i U i CO CO U i U l to 1502 U i OS 1—1 S O 1 o c to - J 42. to - o o c to o o o U i 42-0 0 ' 1—* SO 42. 9LI © o May 2 05 45. O J -to to © © .iii- o o - ©. -© ©. to to to to '•>— © to © VO: os t O 45. - 4 tO. 42. 6: 'OV :OY to op © 42. © 42. os Oh :i—' so 42-o o o ; © OS' op ON © t O : Oh SO to "42. s t O 0 : 42. • 0 J : O J O J O J ' O h ; Oh 0 0 OC to O J s J OC' oc O J O C 42-, V « Ov OS. ' © Oh .42. O J SO OS O J : Oh VO to O •42.', VO to © oc oc 0 0 O V oc . v i S J V l •VO' to © • 0 vo Os 42-i— to to. O J : O to s J to Ov to O J O J -OS, © O J •oo; 42.' Os' s i vo to : © . ' ©: s . • ov 06 O J to S J © O J to s i O J : o v vo O J S i O J 42. to 42. 42. os Oh s i Oh Oh t—- 1-2-© to Os >OJ- O J to to to •SO April 10, 05 to to IT- © O l± © V l O J ' © ; 42. O s i 42. to © s i ' so. \zZ .1-21 © ; s i o h t O ' © >--to to -4>: ; O J :to : 42... • Os :tO. 42. so to so to s i O J 42. 42. Os so - O J : :Os OS O J 0 0 O J OS 'OS so S I 42-V I : tO V i S I S J SO 42.' o s i os to S J . :OS OS 42. to '42-Cs: >--SO. >--' O J 42. 42. :0C •00 O J so. S i to S | to 42. V i . to so OC; Oh so- s j :42. 0 0 >—" , I—; O J to to .SO •os' 0 0 to oc OS OS © ' : © , 4 i to !.*f . to • O J 42. •OV so' :.Sb. so O J © i—•' : O J ; OS S I OO: S J s i ; O J 1—I t-l. :OJ. (-2 to to ' S i ' to S J : SO Oh O J V . Os S J . s i O J oc : O J • O J : S J © J> Oh SO Oh 42. I--S J sp. VP so •SI. . . s i Os: 42.: S I so o S J •OS : v i -42. ,42-'© Os 42. to 'I—l: )LJ. to: to "to to S I s i to i V i . so OS to O J 0 VO OS 0 S I OS OS 42.' ir-; r -O J V l so to 0 0 42. 42. March 31 , 05: to I— O J ' O J © o so 00 S I . OS OS V l O J ' O :.4.0. O J , © ' V i ' © , ..© O J © ' 42. SO; sp °> 42. 42. to :>--V i >2 to s j ^ • • © O' so 42. S l 0 0 42. :2) ss • ^ S J O J to O J s i s i Oh S J O J O0 . © ' : *—•• ° V l 0 0 O0 . © " so 42. O J so to 'SO S J . S i to •SO 42.. O J : ' © S I s j to © O J © 42. to S J so to. SI: O J : S I to "si .42.. , 00 SP Oh s i Op ' . s i : vo-©: . O J 42-V l V I Oh O J V l VP S J 0 0 O J '00 0 0 r-. to. to s | •SO •so s i to, O J 42. to to 0 0 OS to , V i to- to Oh OS 42- V I :so 42. O J O J to to , .. V i 42. OA © SO- 0 0 vo,- © 42. oc SO' VP SO V l 42. oc. O J 1— •42-os O J OS S I 0 0 :O0: OS •so so " ^ ' © . 00 V i O J - ' so O J 00' os .OS O J 42., to. s i Oh 0 0 © OS O J O J 42. os to Oh. to so ;o Oh: op' vp: sp 42. V l ., O J O J O J .00 OS ,OJ S | to OS 42. OS 0 . sp , ' © - J vo H-• 0 0 42. 0 0 t—» S J 42. Oh S J 'I—.' sp to JZ v i Oh SO .1—- O J 4>-0 0 1560 6 0 -so Oh s i Oh 0 0 :6s-'42--1560 Oh © s i O J © S J 0 J V l 1560 ; © OS s i to .OJ to V i LLl 4>; -4 June 21. 05 to o so SO so 'P? ! oo. :<*>. to. to o v i p to o -o p 4a-; © to ©; •4>. P : © p 1— OJ so - o os; os to O J p ,4 i OS' ''>-> OJ to 42- OJ i-4>.. OJ oo to P , OJ • P : W • © ;so to-to to. j •• :OJ:- ,42- ob v> t—l' to ; P 42.. OJ :oo p Os .1-2.. to to to >-4- 'da oc -oo-. •SI i—, OJ', to •p.- to 42- OJ ;oc SJ OJ : -g to 42., , V i to .OS 42. to' Oi' OS' V l O V so OS- H2< :oo p Os ; OS 1—1 1—I ;0s SO SI O J s i OS V i : SI •so" 42. ; O J V I OS to OS to to . V l so ' .Sl : to.' 42. OS to SO: oo 42.. OJ to V l p .04; to OJ- OS OJ- OJ p - . ' to 42. to I—* O J V l O J O J O J 42. so to OS 42. O J oo V l p to >2^  42. 42. 42. : oo oo OS V i ,oo oc • , to . O J , s i O J ; O J O J to O J o O J V l so J > 42. . © so Os' 0 0 V l : OO so SO' V t 42. SI OS OJ OJ •42.. '42S . © 42. V l SI SO © SI V l • os l_i 42. 42.. . V l OS ,42. OS . 0 V l OJ,' s i .os 0 0 OS. V l OJ' OS' to OJ SO SI to V i -to OS 0 0 I—' to: OS to OS V so; os © 42. :to so to May 31, 05 to ;vo 0 0 OS so 42. i—i 0 0 OJ so )—,' t—i ! p s i ,SI OS, - 0 1—» OJ so 0 0 to. SI OJ OJ V l V l to © -os V i ,OJ o -0. to SO OJ'-SO o OS s i 42.. May 14,05 to p . to-©. 42. V l to 42. SO © 0 p - s i 6 0 O J V i so' SI -si O J . V i p O J V l to p S | OS . O J 42. to. .00 O J V l so -to Ov 42. OS 42.- O J , © © . 0 0 OS o . ©'. OJ OS: SI so to 0J to s i © V i so .SO s i 0 0 42. s 4 Catch Basin #3 Date Sample Turbidity Conduc. SS Total Metal (ug/1), Dissolved Metal (ug/1) Metal,/TSS (mg/kg) Time ( N i l ) (Us/cm) (mg/.) Cu Mil Fe Zri Cu Mn Fe Zn Cu Mil Fe. Zh •o 10:15 62 122 54 27 53 2219 90 19 *: 34 686 3327 175472 4606 11:15 63.7 845 47.5 60 50 2307 57 22 * .* 50 5549 4590 329001 1089 11:45 79.9 159 .45.8 16 * 123 67 13 135 25 555 0 1854 6346 £> Q 12:30 73.3 125 40 21 50 2535 49 17 * •* t 714 5248 423135 6774 c 13:30 94.8 145 53.3 '.30i 63 2839 41 16 64 2545 21 1887 223: 38309 2697 13:00 799 174 338 101 268 10520 356 20 23 201 19 300 430 13150 1200 13:30 727 256' 216 89 233 10151 358 12 31 359 73 410 370 12690 77.0 15:00 1000 279 •572. 155 454 20760 422 24 232 23 270. 730 25950 2410 18:30 •••1000 265 66 121 464 22433 393. 17 31 293 120 .1830 740 28040 260 p" 19:15 635 245 338 -98: 31.4 12827 205 8 26 122 59 290 500 16030 690 •CV <N 19:45 602 199 .232 65 236 9040 224 37 46 391 106 280 380: 11300 520 -g 19:45 606 218 292 90 243 9919 209 44 40; 405. 76 270 160 .12400 690 19:45 608 169 274 81 265 9660 232 49 45 413 112 290 420 12080 640 20:15 400 165 180 37 189 7,233 1.14 3 16 46 7' 200 300 9040 210 20:15 406 191 182 29 163 .6532 171 6 17. 81 •m 180 260 8160 310 20:15 415 162: 178 37 204. 7799 190 1 9 70 7 210 330 9750 340 21:00 187 • 129 121 1-2 138 3452 34 •* 13 33 * 100 110 4320 70 6LI February 28, tO ' O : U l 1—' •Ui •'Ul U l o CA. ' © • o ;os si >J .U l ' 4> ~0 SI Ul' I— 42>: 42. CN-lUli U l :to 4=- 'HH Ul: U i oc ::0' •U) U l Ul . to U l U i ..42S. ; © • 0 0 P ' '2-1 _ to. - J to 42. " O : l — * : 00; oo U i •Ul- U l U l ON .©: •SI ON NO ,Cs Ul; • o ON ^ to: 1—• :bo- © • -so NO U r U l ON ND-- G \ 1—' :to U l 0 ;tO :42.- ON •0 42. ' S O - to to ob '42. :U1 to to U i •M- 4>i to • . © 1 t . h - •to O to to to U l : © ' 42. to to •NO- NO - J to U l . U)' •Ui U l U l NO :+i, 42.' Ui .ON -O 'GO Ul 42. • ON U l U i ON ON to 00 U l •to' to Ui s i 0 • 0 ;NO =8 o . ON. SI December 13, 04 to U i ON :tO U l -P. to tO: U l s i S I U l to to to NO ON NO: to 00' to .ON: O to to 42. S4 -o SI ON :NO •NO to NO O ON to 4> U l 0 to 0 0 0 ON © O 42. o ON .42. :ON :to -0 4a. © NO to . © ' o 42.' SI © " .1—< O - : © • u i 0 42. U i 8 42. 1-2 0 •to U l 00 •SI'' u i to to ON - J U>' •O p to U l :.l—i-0 . ON •Uv • 1—1 . * * U l U i v« :'oo; ob r .1—» to 42i. bo 'NO-to :00 -NO U i 0 ON U) I—1 ^ -O :4i: V 1 O ^ i 1—j to 0 • 0 : to S I © 10 o 42. 0 Nov. 14, 04 8 1— 2_,' (-^ i_. ••)—'' t—* •1-" 1—* op;. '9> 'ON U i to ^ Li* 1—*' 1—*: 8' Ik) 'h-p U i U i U l u* U i U i U V O U i I—1 IT—1 i-j' i—2 1—2 to to to to ON .'ON ON 42; ON -.SI 0 . p . ~o. U i : :oo 'ON: O N 42. NO Ul O N Ul ~ J 0 to NO O O V © NO o ON u i U l •si .•St 00 ON ON OO :oo: 0b 42. p . ON iU>: p 00 NO .to U l ON bo; NO 00 ~4 U i : to U l : 42. NO • o :to 4=2 U l : to. to tO; 42. to; U ; '.Ul 42-'SI' .ON •Ul; S>l to. SI i—'' SO P ' 42.' 4^ •ON' .NO IOC-' S i U i U l U i ;:S4 ^ l : 42. :-I— •SO; ••0 -0 U l -4 OO' ON w to to to; to I—. ^ '1— U i :4i U i :;ON- U l r— O N 00 U l S 4 '<—* to to-U l 42.. 42. SO' 42. to S i s i :00 U) SO es-U l to ON ;to U i 42; to U l :S i to :i-'. NO 1—1 : o o ON to SI 0 ON NQ: 42. OC Os j— •:Ui: S i ' U i u i S i U) u i : to :Os pb :©:• U l SI 42. 00 ;U>: si so; O . , to to to' o v u i 'SO- 0 u 10 -00 U l 2 j 0 U i '-2| 00 00 -0. to * >*•• * r -U l -5 ••>—! to :^-j.. Ir-i '1—. • OS 42. -SJ: U i p .tO. 42. ' O , U l OS SO. SO Ul to- ON ^ ' so •to o U l to. 42. Ul' U l OS U i 0 , 0 ' O ' 0 OS NO ON 42. U i 0 . 0 0 to 42. O u OS. ON so: 42. U i U l to i—1 U l U i 0 to 00 U l U) u i U l to -0 2-i to U l ON 0 0 0 0 O 00 O0' •OS NO 42. .—> 421. U l O P O • O 42. 42. SO' Ul 00 h -42. U i Ul. Ui: U i S | O ' p . . . i-i' to :cs U l 42. so U l 0 0 i—' itO U i 42. 42. to to _ to: U 42. 42-U l 1—' Os •o to •Ul U i >-*• OS' ' S i ; to 42. '•-j OS .OS: to so U i 42. -0 ON .ON, to 1—• ,tO : 00 .00 42. NO 42. 1—* to OC .00 42. OS U l OS -OS S i 0 IJi ON U). OS ON SI 0 • O 0 0 0 p 0 :0,- O : •O : 0 42> OS U l 42s O O ' S i io '5J . 0 w 42. ON U l 42. O O . to 42. to 9:45 .921 705 370 106 554 23822. 500 9 4 241. 90 173 898 38630 811 10-45 708 487 260 76 343 16965 338 11 0 207 133 177 793. 39150 780 11:00 .393 1271 46 117 393 13008 378 49 184 .2053 320 86 256 23.848 1,95 ' A 21:00' '752 390 132 157 649 29742 496 44 .82 256 95 137 393 18048 736 © 00" 21:45 171 243 26 ,4:6 145 7315 :29.0 6 8 126 59 270 503 19829 1581 "o- 22:30 414 258 64 118 539 26892 542 8 8 125 77 323 357 14483 .1072 23:30 326 160 42' 97 396 13195, 441 12 7 243. 61 322 436 16631 772 0:30 288 171 39 .61 300 8665 268 9 11 240 .85 91 365 .18465 1105 11:00: 148 657 16 15 81 1024 130 8 56 570 59 71 295 13519 225 >r-i o. oo" 21:00 '275 298 20 62 180 ,8501 246 9 14 192 84 152 508 16821 489 21:45 l is •238 8 30 92 4155 184 2 6, 71 49 106 333 16.752 421 Marcli 22:30 128 229 1.4 58 93 .74.07 218 6 .4: 96 61 110 503 25094 505 Marcli 23:30 92 .144 10 24. 71 3405 243 9 6 166 •57 139 565 18849 630 0:30 42, 154 5; 16 41 2006 120 8 6 151 67 90 532 1301.6 326 1.1:00 306 1.42. 600 124 351 20444 451 32 13 181 68 103 292: 17037 376 12:00 339 189 560 93 195 28768 170 35 10 111 89 83 175: 25686 152 14:45 157 108 680 156 522 37230 311 32 6 92 58 114 383: 57375 228 so" 15:30 249 128 710 185 804 47221, 347 25 6 163 53 130 566 332.54 244 o 16:15 80 162. 250 108 373 2443.4 27.8. 27 9 .119 63 216 746 48869 557 17:00 183 151 ' 780 282 1453 60564 716 30 8: 127 48 181 932 38823 459 22:00 106 83 130 124 347 21111 313 22 4 75 84 477 1335. 81196 1202 23:00 103 73 430 81 270 23201. 368 20 6 77 65 94 314 26978 428 181 M a r c h 2 6 , 0 5 4=. CO to .i— i-- P 4=-: CO to t— OJ p .Oh CO p ; Oh P P p . w o OJ. P . Oh OJ p Oh p to p O P C O •:>— oo to: ON' oo s i :SI ON' 00 so: s i 42- NO -OJ-' 00 Oh p Oh 1—1 OJ' • o : O ' ; ' © op P CO OJ: P .Vi Oh CO CO to CO to SI, 42. P OJ V l , 42- 00 p . CO JS.; OJ : O J SI 42. oo -4 O : to ON Oh OJ • i—' to 42-: J > -OJ- so p so 42. to Vt OO' s o . SO s i so: .ON s i oo -42- 42- SJ 42- ON NO: O J . CO. so. 1—i • r-i: Oh .to V i :Oh: to to < © 00 to :.so O S to s i ' s i O J :.OJ - J . OJ 42. 42. OO O ' O J to 1034 o OJ' 1—I 42.: (—. .ON so :00, V l •OJ, - p . 'OJ. to h-i oo : O J . oo Oh p . NO to iON-ON, S I S I OJ NO 00 CO ON OS s i I— P s | 42. 00 NO & Oh _ OJ 00 SI to 42. p M a r c h 26 , . 0 5 42-to Co p to p s | so i—. OS. I— SO: 42. OS. to; ^ I O ' S O oo ^ Oh so o V« - p . co +; •Oh: ON ^ S O so • p tO: SI to to oo to' p 42. Oh: :00: s i V i OJ to OJ to o t—i 42. :S4" I—1 to- to V i ' S O .to.. 0J OS to :to to: so tJ |2^ ' P s t—» p w-* Oh SI •sc. to OJ t—* oo to OJ S O to V l oo C0-to p : P O S OJ oo 42-: OJ. to ON OJ • OS; •42.: OJ 42, to to to 42-to O ' s | r-* P 42-to t—i -h-' 0 0 OJ SO 42.. I—;, to to V l t_Sl •so to-p V l •oT ! ° 8 to 42. f ° - M to V i P 42. S O O J 42-: J > so p . OS-si to, a-ON: to p to si M a t c h 16 , 0 5 to to s J ON :Oh p • P : OJ p P V i O 42. 42.. to NO ON so O P OJ to CO to :OS-P . OJ ' P -sl • © • p S D O S ON 'NO ON Os p 42-CO OJ 42. OJ •VO: CO CO •42. to OS t—i p 42-OS P V l co :oj; f p . to to OJ: S i : Ov OV 06 so 324.4 .2748 1368 1223: 1861 '3272 4210 6786 Oh OJ 42. . i - -tO OJ vi p : .r-f tO to to • O :co . p - ~ ' :OS. • ON •42.; tO, 1-2 I— SI SO O S l - 1 ' S I ' P —i ,OJ ON to' SO V l OJ OJ to to 00' SI OS :Oh' , p Oh os SJ OJ: 42- SO: >-i iO0. to OS 00 s i 42, SJ to 42. s i ,42.-.OS hO OJ s l ,:Vi' OJ 00 so to' .CO to : O J : 0 Oh ;cs-to: 42. p . 0 S O - Oh 00 so : ON P Co so to OJ :.O0' to Oh s i 00 ; p VO Z81 April 10. 05 March 31, 05 March.?!, 05 18:40 :oo. 42. U i SI t o o •.1—1 S4 ' © • © to CO o CO s CO o :SO !U1 : p 00. 42. p S I Ui , o OS' U l . © Os P U) to U l o •ui-p " o CO © SO Ui . © CO 42. • © -4 CO :os : U l • o i ps ' © U i .CO • p OS u i 0 0 U l : : *-- o to '42. •so: U l ;oo:' .'00 .so 0 0 t o Ul Ul , to 1—I :tO' U l : 1 i 42. U) oo: SI U l oo to oc Os •NO oc © U l , OS Os so U i .'00': :ui:- 42. to os oo © ' ON 'ON to 4=-. © ' 4> .0> t o :co: U i © OS i o U l : U l Ul :oo. U i to s i to U l OS to ••©. ' - 4 1 i -4 •vO • U i . © SO' © S | ' 1—» U l o Os' S I ' • 'S i ' to 42-U l to. U l so Ul. tO - O to 4>:: : t o 42. ON' uv 42, o 42. t o . t o • O : 42-; p os. o -OS u> p -CO p ,U) . © OC p -1 i Os S I h-' >—* oo OS.. 42. 1—' S I U i S i CO " © ; : 0 42-:© U l •oo; CO U l OS SI NO 0 0 :w 42. NO 42. ON U l : S P u » S i OO t o : U l Ul 42,; CO-Ul to u i co; to so to : U i , 1 i OS OS .42. 'Ul'-oo 42-1—1 U l CO • © . U l . S J t ° ;:co; s i so. i—' ; O0 CO i—* • U). .Ul. r -NO to so u» to: 42.. .o; U i 42-OS-U l : OS s i •ON-'42.. U i s i 42. OS :OS' :42>" 1. i to ' U i s i ON • oc-• OO t o OS' oo: u i S I U l 42-NO to to CO :ui. •42-: uv 0 s s i .U l OS U l ON' Ul:. U i *.:• t o ON: S): t o Ul to :'Co ;ui OS: t o ,42. . U i U l to . - 4 to s i to S I to -SI to "00 os Os p so t o • o : tO' U i ; • 1—-S I NO U i 42. U l NO o Ul U i U i ! fe: Ui u i 1—1 U l . U i to oo' 'OS ••©•• so 1—-Os-l o oo ' d ' ;Os ICS:. to to. © to 'OS to CO U l OS '.Ui.: U t 42-to OS' S I t o U l to :co U i OS NO to :Ui u i CO 'l-r-to : © . so to s i U l : r - ' CS 42. • to so 0 0 42.. ui -U l to oo 42, so U i Ul-o -co-to 42-. 42,; oc to OS os •CO —1 to tO U i 42-OO to so .42, OS os © to 42.. SI S I . 1—' U l . to © ,42-: 1 i Ul o SO to S I U) U i U) - o to to U i : 42.. 42. os-U i OS to © to Mi O S U l so ' - 4 : 42. ; u i " to. to to . U " to Ui: CO U l -4 U l : • to. • S I 1—' ' © • to U l 42-CO. © • • to •*=> .42., U l . os © • OS ON - O to to- t O 42-:3 42. o 42-o to 'SO'-oo 42,. S4 s/» US' to 1—' 42, oo -OS; :tO' NO' i to so 42-421 '42. .to U l SO to , u i - o t o C o so-to: i—1 42. .'42V :©-. ;os - y U l to ' © to S I to. : o 'ON. O ' CO © oo 'OO-:00 Os U l 4 i so PS ' u « -ON' ON -4 'OS h-> iOS' to i .Ul o £ ; to NO CO so 42. - 4 S I -~4 • CO 1—1 U i to to >4 '42. SO-'OS ON U l -4 to oc. U l OS O S ON "Ul : u i oo-.Ul os t o u i :42-. 42. S O p . co U l to ; s o •42, 1 i - o o •so 42-. u i o o © • © ; to 0 0 : S I ' 12--1—i Os U i Ul U i Ul. to ^4 t o © 4>: t o 42-42. U i 42. OS •~4 OS U i to-U i OS -O .42-;os :<->: NO 42-U l . S I OS U l U l ' NO' oo-i to - o . U) 42. U i CO.-U l 0 0 Ui OS 42. U l to-U) Cs >— 42, 'so: • O " so © 42. CO : -o . 236 S I C O -ob-. to oc: 42-to so o >-4 OS U) U) o OS oo U l to t o S i to U i : to' ~4 to S I U l •OS U i ' .O OS so' 42.' Ul . SO. :os so • o S I ;tO: OS U i . •oo: 1 to U i 42. S I to :00 :'OC O 0 ' 42. so' 42. U l to to to U i 20 to .so : 0C ©-h--oc 0 s OC to .'42,: - U i to-to, U) to CO to to to U l to U l OS <y U) 0 0 to 0 0 42. to so o <J OS .Ui 42. U i to. 1-2* to - o oo U l 42-U l 42. O . to to U l : © ' ' to s j • U l - o to S J to so U l U i U i -Ui. U l . Ul -4 - o U l : SO '—1 CO SO CO U l £81 May 2, 05 May 2, 05 April 10, 05 1—<" OJ to t-2|' to 1—1 4^ \—1 to 1—* to £:. to. to P S O OC 'OO. i-l s l Sj ' to p NO 8 s o - OJ p 1 : %: P P s O J p 8 422 P s Oh "OJ p : p . 42. Oh to P ' o Oh OJ P to 00 42. 1—-os P 1 -4 oo oo; p : SO' •oo •i> P 'OJ' •00 ' OS p Oh NO: i—* 6b OJ •—l" OJ s i to os 42-SQ OJ" NO OJ ON OJ to to OJ to to ,42-ps o> .'-si vp to p ' to 0 0 ,to 'OJ, -OJ,: s J to' OJ 00: :Oh, :o<i SO OJ to - P : to s j •<5> to 0 0 00' :tO. to - 4 ' so .to ;oo. bo t—' >—. ; O J s j to -CO-42. oo to —4 42., to •sl to s l Vt SO •S4 OJ to Oh to V i OJ 0 0 , 0 J : OC s i ov; to :oh. OS OJ to; to Oh: OJ Oh P OJ ON OJ, OJ: s j Oh to p p p 42. co; to, p t—1 Oh,; s j 42. .!—1 - o 42. OS •42. SO 0 0 OJ P Oh \—» 0 0 OS p 'OO : \ 0 ' - J ; OO: • © ' • OJ" ,°> ;42-: - 4 to o so r-: ON, ob :4> .ON oh *2J-42. Oh oo'-to N )— :ON:: Vt 42-p OJ O S OJ' OJ to 'OJ V i SO •O OJ Oft 42, OJ Oh I—*' :OJ to; S O Oh :4i —1 i>i V l Oh OS-S J •so: s | o to Oh Oh s i ON. 42. 42. OJ to-to s l OJ • p ; t ° . OJ s l t—' Vt 42. to s l 'OC so '27,60 3403 5174 6882 5395 6952 4070 7178 ?403 9061 7719 9559 to , 42. so s i ; 42.' Oh NO 42-NO , o ; to; Co o p 1376 © LZZ to so —1 OJ o s | -O. SD oc V l OJ SD Oh s i :cv s i OJ OS s j - 4 '42-oo "-o. ,42. to: 0 0 OS ,oo s i 42v. oo •so 42.: p p to' to' to OO • P: to SJ 42, ;s4 o to to s j to NO V l os OJ OS oo SO S O ,1—'.< Oh: SO. :Oh', OS oo SO O S NO o P to '22J ; o s4 to NO to ' O J ' p . to Oft 42.-42-to p to oo 42-.42. to. o V l to to OS to V i OS to so os to 00-'oo to :to Oh. so O J 42. to OJ' :.ob: :-OJ-OJ to OJ O J to s i OJ .(—» •OJ OJ ps 42.-to V l ON, O oo p ON p t—< 42- 42, 1102 2068 3132. 3449. 3808 4777 2152 3654 4297 6113 4446 61.20 Cft 0 0 ob' Oh s l Oh to •OJ: O N ON. 42-s j 1054 Co Oh to to OJ OJ P . 42-r-1 to tO: o v OS OJ s i . OJ • H-'•' Oh S l OJ 42- 42-Oh 0 0 2f2-• oo :OJ" 42. V l Oh Oh s i OC. ON ft r- ps- to 42- OJ ON s l to ON O J to p p p ob ON s l .s l 4152 2628 2952 2664 1490: o v V i ,Oh Oh O ' Oh to 42. 42.. o . OO: OJ s j OJ to s i S» OJ • p s i V l to tO-V l Os 0 0 O : Oft NO -t—1 42. t—'' Vt Vt : s l 'OO, to OJ s l J—' to - 4 to so 16608:. 8964 10872 724.8. •<?V oo-'.OO' •oo to -la-CS 2083 to O J OS 1885 .1.253 SO OS' P . r-* O s j 42. OJ' to CO Oh S) Oh 1360 OJ SJ o , 0 J ' Oh OJ.: NO CO * -5543 o 42. 1.335 Vt Oh 138000 8.5080 129348 114696 899)3 57932 ,20350 to SJ ON P SO 26119 16181 13309 15932 OJ H—' p SJ 4825 22960 1,868 1708 7493 OS CO s J :.SO CO-p s j 1—1 p P Vt .36504 19476 21324 15256 12690 6132 386? 3127 to os: ,t—»-s i : 1.55.1 to sJ 0 0 1489 ' © P o ON to OJ 42. s l P ON s l Vt 2346 o s l P o Vt 1781 to o.-june 21, os: SO' ::VO vo. u i VO CY 'OS: os 4> vo, • s i S j OO : U i C O •,U1 • .o . o : o O S o ON to to. U) 42. 1—1 to to •42.' i--i ' O : 1-2 42. Vp: to U l 1—4 U i to J> 8 . © 0 0 to; Os S | <J o ; s i u i t-i'- i-l t O ' IT- U l 0 0 to to to U l . .4=2 oo os U i U l •s i to. U i o vo SJ Cs S | 1—f* U ) to tO: •Ul V l . oc. vo: ,u> t o OO: U l _ to to U l U i to :VO oo to OS to oo w to 1—1 ; u i 42, -o J i oo U ) " U i U l U l . to U i U i 42. U ) S i S J 42, o : ' 0 0 - ^ 1 VO , o OS OC to 1—I 1—2 U l 42. U l oc - 0 0 42, 42. U l Os o vo V 0 . ..SI oc 0 0 U l to o - :ui 42. 42. OS 'SO OS: U i o -42. Os May 31, 05 421 O U i ; •SJ : Os ,42-.SO •SO to O to u i So to to - o to OS so ,vo 4i. so tO: OS • o SO to h— 1—1 to _ to to. U ) 42. O O ' o Ui U l U l 4>- U l to U l U l vo z; ° ° os: S J -4 42. 42,, to o U i U) SI •oo. Cs . : •1—1' u i o" 1—I o o so 0 0 OO O ; U l 42. r O U l . O , U » ' U i 0 0 U i ; IT-U l : . o : so-to •2i to OS' 1—* to 'ui , : Os to to ,-o S J 1—2. oo U l . o .so OS to to ' o 4^i so to to U l to. U l to so •May 31,05 ;ui to t-^ § 8 S to i — : 0 - so Os. s i to •vo' s i to o :.SJ V 0 : to so to ' to :42v I— 42. 4* to U l U l 42. 0 0 v—• m: to U i ' OS 42, s i : so 42. O SI 42. OS SO 42. to: Ul U l . 42; 42. >4 Ui OS U l to U l 0 Os' S J 42. OS 0 : U i SI 1—» 0 0 . • o 0 1—1 —i . u i - .ui. to S4 vp s | 0 0 0 0 ' so OS so to 42; to .Ul Ul to U i Ui ,_, so 0 0 so 1—I S j OS . 0 0 42. 0 •OS Ui 42. Ui U l ,42, ' 0 ' 42. . 0 ; to to .* — ' 2 '^ 42, .1-—' U l U l 42,. U l —J vo 0 0 U l 0 0 0 0 0 0 42, .42.. to ,42, to _ Ul 42, U l 0 0 . 0 OS 0 OS Ul- 1—I U l 0 0 U) h—' so I—1 0 0 os to OS oc o OS to 42, to May 14, 05 to, o .42,; U) : u i . 42. 42. U l 0 - Ui . © • to P 42. p O ; 0 Ui • 0 SO 42. u i : OS, 104, ^ui: ' U i ' O . U l OS SO ;oc to J _ to OS ' U i : 0 SI • 0 0 ; to 0 0 •to U : • O U i to 42. 0 U i . s i to S i Ul. - 0 SO: Os 00 so; 42. O S •o U l - U l ' 42; OS o . Cs to to 42.' so i!l so O S 42. s j U l - to : U i 0 0 0 •OO SI: U l U i oo-- •>! 42. May 14, •t-; s i 1— i-i 42. os s i OS o o to, .so. to to. OS. 42. , © • to U l •so-so •so to h- ,—1 OS oc 12-'' U l OS •op r O 42, O 0 0 ,42. OO OS •op o 42. o to o ,so Ul to, w - J 42. 42. os to - » U i so to to 0 0 0 0 i—1 0 0 to u i 0 s i Os U l 18:20 30 .150 120 64 169 5327 1164 47 56 676 112 354 936 29593 6465 18:40. 54. 1.48 213 76 102 2408; 480 29 88 806 472 •236 3,19 7526 1500 19:00 41 114 53 60 52 1643 .305 29 34 304 261 747 656 20533 3816 g 19:20 2? 90 13 50 51 1203 166 45 47 82,1 251 2478 2562 60144 83.16 19:50 15 65 •27' 33 30 914 158 27 20 242 118 822 751 22838 3948 20:20 16 77 13 37 25, 1005 150 29 17 235 155 1871 1269 50246 7504 (1) *stands for < detection limit Perlite Particle Size and Adsorption Capacity Test Data Table E-2 Used perlite particle size distribution and trace metal concentration Sieve # Diameter ( i um) Perlite Weight Percentage (%) 60 250 5.51 40 420 3.06 30 595 1.4 20 841 1.31 10 2000 1.92 4 4750 69.4 3.5 5660 7.96 2 9420 9.09 Table E-3 New perlite particle size distribution Sieve # Diameter ( ^ m ) Perlite Weight Percentage (%) 2 9520 100 j 6680 98.4 4 4760 85.6 10 2000 23.94 20 841 11.92 30 595 2.59 35 500 0.09 186 Table E-4 Single metal adsorption capacity test data Vol. Solution (ml) Mn Effluent Cone, (mg/1) Vol. Solution (ml) Zn Effluent Cone, (mg/1) 10 0.00 10 0.18 20 20 1.56 30 4.74 30 3.90 40 6.31 40 5.73 50 6.63 50 5.85 60 7.66 60 6.79 70 7.82 70 6.84 80 8.22 80 7.28 90 8.51 90 8.50 100 8.54 100 8.98 110 9.86 110 9.58 120 9.9 120 10.00 187 Table E-5 M n and Zn co-existence adsorption capacity test data M n and Z n Vol . Solution (ml) M n Effluent Cone, (mg/1) V o l . Solution (ml) Z n Effluent Cone, (mg/1) 10 0.00 10 0.00 20 0.81 20 0.24 30 1.57 30 0.96 40 2.04 40 1.07 50 3.52 50 2.62 60 3.55 60 2.74 70 3.84 70 2.77 80 4.24 80 2.96 90 4.40 90 3.27 100 4.73 100 3.58 110 4.91 110 3.86 120 5.00 120 4.15 130 5.00 130 4.90 140 5.00 140 5.00 188 Table E-6 Ca and Mg interference test data Mn (Ca/Mg) Zn (Ca/Mg) Vol. Solution Effluent Cone, (mj Vol. Solution Effluent Cone. ( mg/1) (ml) Mn Ca Mn (ml) Zn Ca Mn 10 0.92 1.44 0.59 10 1.61 2.67 1.29 20 2.90 3.11 1.41 20 2.51 2.73 1.56 30 3.39 1.53 30 2.99 3.34 1.40 40 3.35 3.51 1.49 40 3.27 3.35 1.36 50 3.47 3.60 1.61 50 3.78 3.47 1.47 60 3.80 3.73 1.63 60 3.94 3.90 1.53 70 3.91 3.74 1.67 70 4.37 4.49 1.61 80 4.07 3.87 1.68 80 4.45 4.07 1.63 90 4.07 3.91 1.70 90 4.97 4.33 1.70 100 4.17 4.16 1.70 100 4.99 4.61 1.74 110 4.84 4.91 2.00 110 5.00 5.02 2.04 189 A p p e n d i x F T r a n s p o r t a t i o n a n d L a n d Use M a p s CM m W S o > o < —1 *> '-'j'f* a a: oos, PI ^  1SB3 fflffi \ps~ni! SIR CjiiS^ C TrjiWJ^l im$°nW i f" —<|NFg Site Uk, .5 & sfclsfc HjWf /(fit i O^JUt''' r9? iliS*/* W» SS6 !!989 9(23;; .rii" m i • • - - M l ^ * J f ^ 7 C 4 P M ^ **** =& Figure F- l Morning peak hour traffic volume in the north half of Vancouver 190 SO H1 era' c 3 -o ft o c o < c 3 o c "iff 37b " f ls - : : * » LA r 610 \$?0 S70 i3S6 7S^^'' : : : :ip : : ; ; := »<>.« u A U T O V O L U M E S , 66* 153 7)1' "i.ea...ssj ITTT * 303 M30 )K SI 2SS las': E I I H E / Z P R O J E C T : B u x n a b j S u b - A r e a M o d e l S C E N A R I O 1 0 0 0 : 1 9 9 9 B u r n a b y B a s e T H R E S H O L D : U P P E R : 9 9 9 9 9 9 U I H D O U : 1 2 . 4 6 3 / 2 1 . -4527 2 4 . 9 7 3 / 3 0 . 6 3 5 4 0 5 - 0 9 - 2 3 1 6 : 2 3 M O D U L E : 6 . 1 2 C . V . R . D ; > N X I E 3 CQ o c H _3 O > 3 O - C <U Q. C o 3 E ON Figure F-4 Morning peak hour traffic volume in the south half of Burnaby 193 O 01 01 H S « J w w 7 h o-10 t-^ 7 — 0 0 , 1 a > 10 o W a o > o IH 9n V Jfl « o w o w Pi « & Figure F-5 Morning peak hour traffic volume in the Coquitlam 194 c • • D • • • Di it:;. 0 o i I™ V • Figure F-6 Land use map of the Brunette River Watershed in 2001 195 SubC at ch_ldus e Agricultural M Residential - Rural Residential ~ Single Farriily and Duplexes Residential - Townhouse and Low-rise Apartments HBResidential ~ High~ri se Apartments Commercial _ Residential/Mixed C omm er c i al CU Ins ti tut i onal . . J Indus tri al • industrial - Extractive S i Transportation, Communication and Utilities JRecreation and Frotected Natural Areas Protected Watershed HI Harvesting and Research EZIOpen and Undeveloped Lakes and Water Bodies 196 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0100954/manifest

Comment

Related Items