Final Report JOINT STORMWATER AGENCY PROJECT to STUDY URBAN SOURCES of MERCURY, PCBS and ORGANOCHLORINE PESTICIDES

Final Report

JOINT STORMWATER AGENCY PROJECT TO STUDY URBAN SOURCES OF MERCURY, PCBS AND ORGANOCHLORINE PESTICIDES

A collaborative project of the following agencies: • Santa Clara Valley Urban Runoff Pollution Prevention Program • Contra Costa Clean Water Program • San Mateo Countywide Stormwater Pollution Prevention Program • Marin County Stormwater Pollution Prevention Program • Vallejo Flood Control and Sanitation District • Fairfield-Suisun Sewer District

April 2002

Page Intentionally Blank

Final Report

JOINT STORMWATER AGENCY PROJECT TO STUDY URBAN SOURCES OF MERCURY, PCBS AND ORGANOCHLORINE PESTICIDES

Prepared by: Kinnetic Laboratories, Inc. 307 Washington Street Santa Cruz, California 95060

In Association with: EOA, Inc. 1410 Jackson Street Oakland, California 94612

April 2002

Page Intentionally Blank

TABLE OF CONTENTS

List of Tables...... ii

List of Figures...... iii

EXECUTIVE SUMMARY...... ES-1

1.0 INTRODUCTION ...... 1 1.1 Background and Overview...... 1 1.2 Project Objectives ...... 2

2.0 METHODS...... 3 2.1 Selection of Sampling Sites...... 3 2.2 Sampling Procedures...... 20 2.2.1 Initial Sampling Equipment Decontamination ...... 20 2.2.2 Field Sampling Procedures...... 20 2.3 Analytical Methods and Reporting Limits ...... 24 2.4 Field and Laboratory Quality Control...... 24 2.4.1 Field Sampling Quality Control...... 24 2.4.2 Laboratory Quality Control ...... 24 2.5 Data Analysis ...... 26 2.5.1 Descriptive Statistics...... 26 2.5.2 Hypothesis Testing...... 26 2.5.3 Stormwater Load Estimations...... 26 2.5.3.1 Watersheds and Land Use ...... 28 2.5.3.2 Runoff Coefficients...... 30 2.5.3.3 Rainfall ...... 31 2.5.3.4 Contaminant Concentrations...... 34

3.0 RESULTS AND DISCUSSION ...... 35 3.1 Major Land Use...... 48 3.2 Urban and Nonurban ...... 51 3.2.1 Comparison iment with Recent San Francisco Bay Sediments...... 52 3.3 Stormwater Load Estimates...... 57 3.3.1 Assessment of potential impacts of cleanup activities on PCB loading ...... 57 3.4 Top 15th Percentile Sites...... 62 3.5 PCB Homologs...... 68

4.0 CONCLUSIONS AND RECOMMENDATIONS...... 73 4.1 Conclusions...... 73 4.2 Recommendations...... 74

5.0 REFERENCES...... 75 Appendix A: Summary of Analytical Results for PCBs and Organochlorine Pesticides. Appendix B: Site Descriptions, Field Notes, Site Photos (Photos provided on CD only) Appendix C: Sediment Particle Size Data Results Appendix D: QA/QC Assessment Appendix E: Laboratory Data Sheets (Provided on CD only)

LIST OF TABLES

Table ES.1 Summary of Estimated Stormwater Loads to San Francisco Bay from 17 Watersheds and Portions Contributed by Urban Sources...... ES-3 Table 1. Number of sampling locations by jurisdiction and drainage type...... 3 Table 2. Station location percentages by jurisdiction and drainage type...... 4 Table 3. Sampling locations and station identifications 2000-2001...... 15 Table 4. Parameters, methods, and holding times for analysis of sediments ...... 22 Table 5. Comparison of Regional Monitoring Program target method detection limits (MDLs) for PCB congeners with project specific MDLs and method reporting limits (MRLs) ...... 23 Table 6. Comparison of Regional Monitoring Program target method detection limits (MDLs) for organochlorine pesticides with project specific MDLs and method reporting limits (MRLs)...... 24 Table 7. Summary of laboratory quality control performance measures and data quality objectives for measurement of mercury, PCBs, organochlorine pesticides, TOC and grain size in sediments...... 25 Table 8. Summary of drainage areas and land use within each hydrologic area tributary to San Francisco Bay ...... 30 Table 9. Estimates of impervious fraction and calculated runoff coefficients for each major land use category...... 31 Table 10. Runoff coefficients used by SFEI (2000)...... 31 Table 11. Summary of data used to estimate average annual rainfall in each region...... 33 Table 12. Average annual rainfall values used for each hydrologic area...... 34 Table 13. Estimates of the range and average concentration of suspended solids in stormwater runoff from the five major land use classifications...... 34 Table 14. Summary of first year (2000) data results from each sampling location ...... 36 Table 15. Summary of data results from 2001 survey at each sampling location ...... 38 Table 16. Summary of 2001 data normalized to the fine fraction (<62.5 microns)...... 41 Table 17. Statistical summary of results by major land use classification ...... 49 Table 18. Statistical comparison of analyte concentration in sediments from storm drainages in four major land use categories...... 51 Table 19. Statistical summary of results by urban and nonurban land use categories...... 53 Table 20. Statistical comparison of analyte concentration in sediments from urban and nonurban (open) storm drainages...... 54 Table 21. Estimates of annual loading (lbs/year) of PCBs, total mercury, chlordane and DDT to San Francisco Bay from each major watershed...... 59 Table 22. Estimates of annual loading (lbs/year) of PCBs, total mercury, chlordane and DDT to San Francisco Bay from each major land use category...... 60 Table 23. Comparison of current annual PCB load estimates with two source reduction scenarios...... 60 Table 24. List of sites identified with the top 15th percentile concentrations of total PCBs and mercury from sampling conducted in 2001...... 64 Table 25. List of sites identified with the top 15th percentile concentrations of total chlordane and total DDT from sampling conducted in 2001...... 65

LIST OF FIGURES

Figure 1. Sampling sites showing locations of detail maps ...... 5 Figure 2. Sampling locations in northern Marin County ...... 6 Figure 3. Sampling locations in southern Marin County ...... 7 Figure 4. Sampling locations in northern San Mateo County ...... 8 Figure 5. Sampling locations in southern San Mateo County...... 9 Figure 6. Sampling locations in Santa Clara County ...... 10 Figure 7. Sampling locations in western Contra Costa County ...... 11 Figure 8. Sampling locations in eastern Contra Costa County...... 12 Figure 9. Sampling locations in the region of Vallejo...... 13 Figure 10. Sampling locations in the region of Fairfield...... 14 Figure 11. Hydrologic areas (HA) used to develop stormwater runoff loading estimates for San Francisco Bay ...... 29 Figure 12. Location of rain gauges used for estimates of annual precipitation...... 32 Figure 13. Concentrations of total PCBs measured in storm drains throughout the study area. 44 Figure 14. Concentrations of total mercury measured in storm drains throughout the study area...... 45 Figure 15. Concentrations of total chlordane measured in storm drains throughout the study area...... 46 Figure 16. Concentrations of total DDTs measured in storm drains throughout the study area...... 47 Figure 17. Comparison of concentrations and distributional properties of total PCBs, total mercury, DDT and chlordane in the fine fraction of storm drain sediments from four major land use categories...... 50 Figure 18. Comparison of concentrations and distributional properties of total PCBs, total mercury, chlordane and DDT in storm drains from urban and open land use areas tributary to San Francisco Bay...... 55 Figure 19. Comparison of concentrations of total PCBs, mercury, chlordane and DDT in storm drains from urban and nonurban (open) land use areas with levels measured in San Francisco Bay sediments...... 56 Figure 20. PCB annual loading estimates based upon existing conditions two cleanup scenarios for the top 15th percentile sites...... 61 Figure 21. Ranked distribution of total PCBs and mercury...... 66 Figure 22. Ranked distribution of total chlordane and mercury...... 67 Figure 23. Distribution of PCB homologs measured in standard formulations of Arochlors...... 70 Figure 24. Concentrations of PCB homologs in sediments from all 2001 sampling sites with total PCB concentrations greater than 300 ug/Kg fines...... 71 Figure 25. Concentrations of PCB Homologs in sediments from all 2001 sampling sites with total PCB concentrations between 100 and 300 ug/Kg fines...... 72

iii

Page Intentionally Blank

iv Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

EXECUTIVE SUMMARY

This report summarizes the results from two years of sampling conducted to document the presence and distribution of contaminants of concern in sediments collected from storm drains. During the first year of the program sediments were collected from 61 sites from Marin, San Mateo, Santa Clara Valley, Contra Costa and Solano Counties. Sampling locations were selected to characterize the average, range and variability of mercury and PCB concentrations in urban residential/commercial catchments, urban industrial catchments, open space, and downstream sites representative of entire watersheds. Sediments from the first year were tested to determine particle size distributions, total organic carbon (TOC), percent moisture, PCB congeners, total mercury, and methyl mercury. During the second year, field surveys were designed to emphasize industrial catchments. Organochlorine pesticides, including chlordane and DDTs, were added to the tests conducted on the sediment and methyl mercury was eliminated from the analytical suite.

During the second year’s survey, samples from 70 sites were submitted to the laboratory for analysis. Approximately two-thirds of the sampling locations were from areas dominated by industrial land use activities. Sediment sampling procedures were based on those in use by the Regional Monitoring Program for Trace Substances in the San Francisco Estuary and the USGS National Water Quality Assessment Program. To assure comparability of results, these procedures were coordinated with complimentary work being conducted by the Alameda Countywide Clean Water Program. Seven field replicates, submitted blind to the laboratories performing the analyses, were included to assess the extent of field variability since this could have a strong influence on data interpretation. In addition, a thorough laboratory control program included the use of method blanks, matrix spike/spike duplicates, surrogate spikes, laboratory control samples (LCS), certified reference materials (CRM), and laboratory replicates.

Summary statistics were developed for major land use classifications including the number of samples, the percentage of the samples with values above project reporting limits, minimum and maximum detected values, the median, 25th and 75th percentiles, mean, standard deviation and coefficient of variation. In the previous year, statistics were developed based upon natural logarithm transformations and evaluation was based upon characteristics of the lognormal distribution. Upon normalization to fines of the data for both years, portions of the data set were found to depart substantially from the lognormal distribution. In addition, handling of below detection limit data was becoming an issue for a few of the target analytes. In order to allow inclusion of the censored (below detection limit) data, a method known as regression on order statistics was used to obtain estimates of the distribution of data below project detection limits.

Two hypotheses were established to address objectives established for the second year of the sediment testing program. These included:

Ho: There was no difference in storm drain sediment contaminant levels among residential/commercial and industrial land use sites. H1: There was no difference in storm drain sediment contaminant levels among urban (industrial, residential/commercial and mixed) and nonurban (open) land use classifications.

The combined two-year data set of this sampling project together with data provided by the Alameda Countywide Clean Water Program and appropriate sites from separate PCB case studies resulted in 164 sampling units for PCBs consisting of 68 industrial sites, 31 residential/commercial sites, 44 mixed land

ES-1 Urban Sources of Mercury and PCBs-Final Report April 2001

use sites, and 21 open space sites. Slightly fewer sites (152) were tested for mercury. Organochlorine pesticide data were available from 73 locations.

All data analysis was based upon measured concentrations normalized to the sediment particles less than 62.5 microns (fine fraction).

A nonparametric analysis of variance, the Kruskal Wallis test, indicated that concentrations of total PCBs, total mercury and total chlordane were all significantly different among the four land use classifications. No significant differences (p=0.134) were found in the concentration of DDT in storm drain sediments from different land use classifications. A Bonferonni Inequality test indicated that concentrations of total PCB and total mercury in open land use sites were significantly lower than the remaining three categories. No significant differences were evident in the concentrations of total PCBs and mercury measured from industrial and residential/commercial land use sites. Chlordane was found to be significantly higher in sediments from residential/commercial land use sites.

Similar tests were conducted after reclassification of sampling sites to urban and nonurban (open) sites. Mann-Whitney tests indicated that samples taken from urban environments had significantly higher levels of total PCBs, total mercury, chlordane, and DDT. Median concentrations of constituents in samples collected from urban sites vs. nonurban sites were over 100 times greater for total PCBs, three times greater for total mercury, 40 times greater for chlordane and 100 times greater for DDT compounds. In the case of both chlordane and DDT, representative coverage of open land use activities was not accomplished. Only four samples were taken from open land use areas since the emphasis was directed towards improving coverage of industrial land use areas for better characterization of PCB sources.

The concentrations of contaminants in sediments from Bay area storm drains were used as the basis to develop planning level estimates of loading to San Francisco Bay. Estimation procedures were generally based on the SIMPLE model. Several key assumptions were necessary to develop input parameters. The first assumption was that concentrations of contaminants present in sediments from storm drains were 100 percent associated with particles less than 62.5 microns in size (fine fraction). The second major assumption was that the concentration of contaminants associated with the embedded fine fraction was representative of the concentration of the contaminants in suspended sediments in stormwater discharges.

Load estimates were developed based upon the 25th percentile, 75th percentile and median concentrations in order to provide a framework for assessing the extent of uncertainty associated with estimates (Table ES.1). The overall load for PCBs was estimated to range from 19 to 228 pounds per year with a median estimate of 87 pounds per year. Over 99 percent of the PCBs are from urban sources. Mercury loads were estimated at 131 to 578 pounds per year with a median of 271 pounds per year. In the case of mercury, a substantial portion (12 to 22 percent) was estimated to come from open and agricultural land use portions of the watersheds. Estimates of mercury loads associated with stormwater discharges specifically do not include large point sources associated with historic mercury mines located within San Francisco Bay watersheds. These sources were not characterized in this study.

Far less information is currently available for chlordane and DDT. Although initial estimates of loads from agricultural and open land use areas are very low, they are only based upon measurements from four sites. Preliminary estimates were made for these two constituents based upon data from this year. Based upon this limited set of data it is estimated that 48 to 261 pounds of chlordane and 2.3 to 55 pounds of DDT are discharged to the Bay from both urban and nonurban areas on an annual basis (Table ES.1).

An evaluation of the possible effects of remediation at locations with PCB concentrations in the top 15 percent of all urban sites was conducted to examine sensitivity to such measures. Clean up to an objective equivalent to the 75th percentile reduced estimated high-end loads by 24%. Use of the median urban concentration as a cleanup objective reduced high-end load estimates by approximately 50%. ES-2 Urban Sources of Mercury and PCBs-Final Report April 2002

Table ES.1 Summary of estimated stormwater loads to San Francisco Bay from 17 watersheds and portions contributed by urban sources

San Francisco Bay Urban Sources Watersheds Pounds Kilograms Pounds Kilograms PCB, Median 87 39.5 87 39.5 PCB, High 228 103.6 227 103.2 PCB, Low 19 8.6 19 8.6

Total Hg, Median 271 123.2 210 95.5 Total Hg, High 578 262.7 497 225.9 Total Hg, Low 131 59.5 115 52.3

Chlordane, Median 85 38.6 84 38.2 Chlordane, High 261 118.6 260 118.2 Chlordane, Low 48 21.8 48 21.8

DDT, Median 23 10.5 23 10.5 DDT, High 51 23.2 50 22.7 DDT, Low 2.3 1.0 2.3 1.0

Ranking the data and comparing distributional characteristics identified sites with particularly high concentrations of the target constituents. Several sites were found to have high concentrations of several of the target constituents. The highest concentration of PCBs (124,198 ug/Kg fines) was encountered in sediments collected from a manhole in an industrial catchment in San Carlos (SMC020). This site receives runoff from an area that includes Delta Star, a former generator of PCBs and manufacturer of transformers. High levels of mercury (11.3 mg/Kg fines), total chlordane (11,971 ug/Kg fines), and DDT (24,541 ug/Kg fines) were also encountered at this site. HCHs1 were also detected at this site in the highest concentrations encountered in the study (3,611 ug/Kg fines). PCBs and other pollutants have been found in soil samples from the Delta Star site. The Regional Board has adopted site cleanup requirements (Order No. 99-062).

The second highest concentration of PCBs (28,003 ug/Kg fines) was found in sediments collected from Colma Creek (SMC024). An oil and grease sheen has been observed at this site and at previous sites downstream of this sampling location. The sampling point is located inside a culvert under Collins Road. Several sources appear to converge at this point. All conveyances go underground at this point and mapping information was limited. The catchment includes a number of automobiles sales and maintenance activities. This site also had elevated levels of total chlordane (1,633 ug/Kg fines).

Sediments collected from the Pulgas Creek Pump Station (SMC023) in San Carlos contained the third highest concentration of PCBs (25,044 ug/Kg fines). This pump station drains an industrial catchment. Sediments at this site were anoxic. The composition of the PCBs at this site showed signs of weathering typical of anoxic conditions. Congeners with three to four chlorine atoms were more abundant than expected in typical Arochlor formulations. Sediments from this pump station also contained high levels of chlordane (902 ug/Kg fines) and DDT (1,410 ug/Kg fines).

1 HCH refers to the sum of the alpha, beta, gamma (lindane) and delta isomers of hexachlorocyclohexane.

ES-3 Urban Sources of Mercury and PCBs-Final Report April 2001

Sediments collected from a manhole near Nebraska Street in Vallejo (VFC006) contained the fourth highest concentrations of PCBs (21,231 ug/Kg fines) encountered this year. This is primarily a residential catchment that ultimately discharges through Austin Creek. Unlike some of the other sites with high levels of PCBs, other target constituents were not present at elevated levels at this site.

Sediments sampled for CCC032 in Richmond was comprised of a composite of four catch basins located on the north side of West Cutting Boulevard near South 1st Street. PCBs in this composite sample were measured at 12,538 ug/Kg fines. Composite samples came from catch basins in front of the Richmond Distribution Center, a PG&E substation, and a machine shop. Measurements of mercury, chlordane, and DDT were all within the top 15 percent at this site.

The highest concentrations of mercury were measured in sediments from three sites in Marin County (MCS007, MCS011, and MCS010), three in San Mateo County (SMC025, SMC030, and SMC020) and one Santa Clara Valley site (SCV028). The Marin County sites were all located in the lower portions of the Pacheco and Arroyo San Jose watersheds. Both ultimately discharge through Pacheco Pond to San Francisco Bay. The San Mateo sites include SMC020 where high PCBs were also encountered. Other San Mateo sites included (SMC025) which is an industrial area where sediments were collected from a manhole located at the eastern end of Beatty Avenue. SMC030 is located in the Jefferson Branch of Redwood Creek just northwest of Harrison Street where it runs underneath Duane Street. Drainage for this branch of the Creek includes Stulsaft Park on Farm Hill Boulevard. During redevelopment of the Park, it was discovered that the Park was the site of a mercury mine in the 1950’s. San Mateo County Health Service Agency and Redwood City discovered residual mercury contamination in the soils and have planned remediation efforts before continuing with the redevelopment process.

One Marin County site, MCS008, was found to have high levels of both chlordane (15,396 ug/Kg fines) and DDT (2,054 ug/Kg fines). Sediment was collected from a corrugated metal outfall near 31 Pamaron Way that flows into Pacheco Creek. A followup investigation was conducted by Marin County to evaluate potential sources.

ES-4 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

1.0 INTRODUCTION

1.1 Background and Overview

The Project Plan for the first year of this project, “Joint Stormwater Agency Project to Study Urban Sources of Mercury and PCBs,” (EOA, Inc. 2000) was developed, approved, and submitted to the Regional Board on July 1, 2000 as a collaborative project of the following agencies:

• Santa Clara Valley Urban Runoff Pollution Prevention Program (SCVURPPP) • Contra Costa Clean Water Program (CCCWP) • San Mateo Countywide Stormwater Pollution Prevention Program (SMCSPPP) • Marin County Stormwater Pollution Prevention Program (MCSPPP) • Vallejo Flood Control and Sanitation District (VFCSD) • Fairfield-Suisun Sewer District (FSSD).

The first year Project Plan (EOA, Inc. 2000) provided a summary of developments that lead to implementation of this project. The overall goal of the project was to provide data to support development of TMDLs for PCBs and mercury in San Francisco Bay. During the first year of this study, the Regional Board requested that three primary questions be addressed. These included:

1. Are PCBs distributed somewhat evenly throughout the landscape resulting in relatively similar PCB discharges from storm water conveyance systems? 2. Are storm drains or other surface drainage pathways sources of PCBs in themselves? 3. Are there specific locations within watersheds where prior or current use of PCBs result in land sources contributing to continued PCBs discharges through stormwater conveyance systems?”

During the first year, deposited sediments were collected from 61 monitoring stations located within jurisdictions of the six stormwater agencies funding this investigation. Data from an additional 22 sites sampled and analyzed by the Alameda Countywide Clean Water Program were used to augment the data set. All sediments were analyzed to determine particle size distributions, total organic carbon, percent moisture, PCB congeners, total mercury, and methyl mercury.

Sampling locations were selected to characterize the average, range and variability of mercury and PCB concentrations of each of the following four categories:

• Urban residential/commercial catchments. • Urban industrial catchments. • Open space. • Downstream sites representative of entire watersheds.

Consideration was also given to providing adequate spatial coverage within the study area.

Based upon results from the first year’s program, a work plan was developed to acquire further information on potential sources of contaminants in storm drainages in San Francisco Bay watersheds (EOA Inc., 2001). This work plan was based upon input from the Regional Board and various stormwater agencies. The new work plan called for scaling back on mercury investigation by elimination of methyl mercury, adding chlorinated pesticides to the analytical list, and focusing upon identification of smaller drainages with high industrial use as well as areas where PCBs were historically used, stored or released. In addition, the work plan called for case studies to be performed in areas where high concentrations of PCBs were encountered during the previous year. The case studies were intended to

1 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002 isolate and identify significant sources within the target drainages. In most situations, the case studies are being conducted independently by each discharger.

1.2 Project Objectives

The second year work was designed to augment and extend the initial data from the fall 2000 survey. The general goal was to continue to assist Regional Board staff in developing TMDLs. This second year of work focused on PCBs, but included gathering additional data on mercury and beginning to collect data on organochlorine pesticides (especially the “legacy” pesticides chlordane, DDT, and dieldrin).

The specific objectives of the work were:

• Gathering further data on the distribution of total mercury and PCBs homologs in different land uses in urbanized watersheds draining to San Francisco Bay,

• Beginning to gather data on the distribution of organochlorine pesticides in different land uses in urbanized watersheds draining to San Francisco Bay,

• Attempting to identify additional conveyance systems with elevated concentrations of PCBs in embedded sediments, and

• Calculating order–of–magnitude estimates of loadings of PCBs, mercury, and chlorinated pesticides to the Bay from urban runoff conveyance systems.

2 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

2.0 METHODS

2.1 Selection of Sampling Sites

Stations were selected to characterize the average, range and variability of mercury and PCB concentrations within each of the following four categories:

• Urban residential/commercial catchments. • Urban industrial catchments. • Open space. • Downstream sites representative of entire watersheds.

In addition, stations were initially apportioned to be within the participating agencies’ jurisdiction as follows:

• Santa Clara Valley Urban Runoff Pollution Prevention Program (SCVURPPP) – 34% • Contra Costa Clean Water Program (CCCWP) – 24% • San Mateo Countywide Stormwater Pollution Prevention Program (SMCSPPP) – 24% • Marin County Stormwater Pollution Prevention Program (MCSPPP) – 6% • Vallejo Flood Control and Sanitation District (VFCSD) – 6% • Fairfield-Suisun Sewer District (FSSD). – 6%

The initial scope of work called for the selection of approximately 60 monitoring stations in all. The Vallejo Flood Control and Sanitation District opted to fund sampling at an additional five stations in lieu of conducting a case study investigation. This included resampling at one of the previous years sites where the highest concentrations of PCBs were encountered. The Marin County Stormwater Pollution Prevention Program also independently increased sampling sites by six to a total of nine in order to conduct an investigation of sources of contaminants in the two major watersheds flowing to Pacheco Pond. Fifty-eight monitoring stations were selected during the investigation and selection of stations (Table 1).

Table 1. Number of sampling locations by jurisdiction and drainage type.

DRAINAGE SCVURPPP SMCSPPP CCCWP MCSPPP VFCSD FSSD TOTAL TYPE Industrial 13 11 13 4 1 2 44 Res./Com. 1 4 2 4 11 Open 2 1 1 4 Mixed 4 1 3 2 1 11 Total 20 16 14 9 8 3 70

Final apportionment of stations within jurisdictions was close to target levels set for each jurisdiction (Table 2) when additional samples taken by the Marin and Vallejo programs are considered. As designed, the emphasis this year was on improving coverages in industrial land use areas.

3 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 2. Station location percentages by jurisdiction and drainage type.

DRAINAGE SCVURPPP SMCSPPP CCCWP MCSPPP VFCSD FSSD TOTAL TYPE % % % % % % % Industrial 18.1 15.3 18.1 5.6 1.4 2.8 62.9 Res./Com. 1.4 5.6 0 2.8 6.9 0 15.7 Open 2.8 0 1.4 0 2.8 0 5.7 Mixed 5.6 1.4 0 4.2 2.8 1.4 15.7 Total 28.6 22.9 20.0 12.9 11.4 4.3 100

4 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 1. Sampling sites showing locations of detail maps

5 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 2. Sampling locations in northern Marin County

6 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 3. Sampling locations in southern Marin County.

7 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 4. Sampling locations in northern San Mateo County.

Figure 5. Sampling locations in southern San Mateo County

Figure 6. Sampling locations in Santa Clara County

10 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 7. Sampling locations in western Contra Costa County.

11 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 8. Sampling locations in eastern Contra Costa County.

12 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 9. Sampling locations in the region of Vallejo.

13 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 10. Sampling locations in the region of Fairfield.

Table 3. Sampling Locations and Station Identifications 2000-2001.

STATION SAMPLING WATERSHED SAMPLING GENERAL SITE LOCATION LATITUDE LONGITUDE ID DATE PROPERTIES LOCATION

555 Burke Street (San Jose) 37º 18.620' 121º 51.167' Manhole SCV001 30 Oct. 2000 Industrial 285 Leo Street (San Jose) 37º 18.668' 121º 51.860' Manhole Sunol Street and Auzerais Avenue (San Jose) 37º 19.263' 121º 54.345' Catchbasin SCV002 30 Oct. 2000 Industrial West Home Street near. Sunol Street (San Jose) 37º 19.207' 121º 54.253' Catchbasin Fairchild Drive east of Ellis Street (Mountain View) 37º 24.195' 122º 03.032' Manhole SCV003 27 Oct. 2000 Industrial Intersection of National Ave. and Fairchild Drive (Mountain View) 37º 24.248' 122º 03.213' Manhole SCV004 Intersection of Transport St. and Industrial Avenue (Palo Alto) 24 Oct. 2000 37º 25.455' 122º 05.900' Industrial Catchbasin SCV005 Tahoe Ave. (near Lawrence Exp. And Kiefer, Santa Clara) 27 Oct. 2000 37º 22.545' 121º 59.875' Industrial Manhole SCV006 601 Walsh Avenue (Santa Clara) 27 Oct. 2000 37º 22.210' 121º 56.792' Industrial Manhole SCV007 Ringwood near Montegue Expressway (Milpitas) 27 Oct. 2000 37º 24.267' 121º 54.028' Industrial Outfall SCV008 Charleston Retention Basin (Mountain View) 24 Oct. 2000 37º 25.308' 122º 04.322' Industrial Detention Basin SCV009 Bellew Pump Station (Milpitas) 31 Oct. 2000 37º 25.087' 121º 55.785' Industrial Pump Station SCV010 Murphy Ranch Pump Station (Milpitas) 31 Oct. 2000 37º 24.089' 121º 55.532' Industrial Pump Station SCV011 Oak Creek Pump Station (Milpitas) 31 Oct. 2000 37º 24.395' 121º 55.248' Industrial Pump Station SCV012 West Virginia Street and the Guadalupe River (San Jose) 30 Oct. 2000 37º 19.207' 121º 53.343' Res./Com. Outfall SCV014 South Corner A Street and Knowles (Los Gatos) 30 Oct. 2000 37º 15.687' 121º 57.700' Res./Com Catchbasin SCV015 Calabassas Creek near Mount Eden Road (Saratoga) 26 Oct. 2000 37º 16.132' 122º 03.638' Open Open Channel SCV016 San Francisquito Creek near Alpine Road (Santa Clara County) 27 Oct. 2000 37º 24.758' 122º 11.655' Open Open Channel SCV017 Coyote Creek at Gilroy Hot Springs (Santa Clara County) 26 Oct. 2000 37º 05.963' 121º 28.375' Open Open Channel Lincoln Ave & Savaker Street (San Jose) 37º 19.068' 121º 54.482' Catchbasin SCV018 Northrup & Lincoln (San Jose) 30 Oct. 2000 37º 19.050' 121º 54.502' Mixed Catchbasin La Fenetre Apartments on Parkmoor Avenue (San Jose) 37º 19.007' 121º 54.495' Catchbasin Sunnyoaks Industrial Center (Campbell) 37º 16.307' 121º 57.148' Catchbasin SCV019 26 Oct. 2000 Mixed 1260 Dell Avenue (Campbell) 37º 16.063' 121º 57.210' Catchbasin SCV021 Cottle Road South (just south of Santa Teresa Blvd, San Jose.) 30 Oct. 2000 37º 14.088' 121º 48.232' Mixed Open Channel SCV022 Old Oakland near Montegue Expressway (Milpitas) 27 Oct. 2000 37º 24.267' 121º 54.028' Mixed Outfall SCV024 Calabassas Creek between Lakeside Dr. and Scott Blvd. (Sunnyvale) 31 Oct. 2000 37º 23.285' 121º 59.215' Mixed Open Channel SCV025 Gold Street Pump Station (Alviso) 21 Sept. 2001 37º 25.694' 121º 58.551' Mixed Pump Station

STATION SAMPLING WATERSHED SAMPLING GENERAL SITE LOCATION LATITUDE LONGITUDE ID DATE PROPERTIES LOCATION

SCV026 Oakmead Pump Station (San Jose) 21 Sept. 2001 37º 24.868' 121º 57.810' Mixed Pump Station SCV027 River Oaks Pump Station (San Jose) 21 Sept. 2001 37º 24.082' 121º 56.691' Industrial Pump Station SCV028 Newhall Street and Campbell Avenue (San Jose) 5 Oct. 2001 37º 20.743' 121º 55.535' Res./Com. Manhole SCV029 McKendrie Street and Stockton Avenue (San Jose) 5 Oct. 2001 37º 20.730' 121º 55.110' Industrial Catchbasin SCV030 Lorentz Barrel and Drum (San Jose) 10 Sept. 2001 37º 19.186' 121º 51.940' Industrial Catchbasin SCV031 Needles Drive and 10th Street (San Jose) 10 Sept. 2001 37º 18.832' 121º 51.608' Industrial Catchbasin SCV032 1850, 1854, and 1760 7th Street (San Jose) 10 Sept. 2001 37º 18.657' 121º 51.726' Industrial Catchbasin SCV033 Stauffer Chemical/GE Datasonics (1887 Monterey Rd. San Jose) 10 Sept. 2001 37º 18.541' 121º 52.008' Industrial Manhole SCV034 5th and 7th Streets (San Jose) 5 Oct. 2001 37º 21.722' 121º 54.060' Industrial Manhole SCV035 Hendy Avenue (Sunnyvale) 25 Sept. 2001 37º 22.562' 122º 01.349' Industrial Catchbasin SCV036 Robert Avenue (Santa Clara) 21 Sept. 2001 37º 21.881' 121º 56.567' Industrial Catchbasin SCV037 Commercial Street and Arques Avenue 25 Sept. 2001 37º 22.853' 122º 00.453' Industrial Catchbasin Commercial Avenue (Sunnyvale) 37º 25.501' 122º 06.005' Catchbasin SCV038 25 Sept. 2001 Industrial Independence Avenue (Sunnyvale) 37º 25.288' 122º 05.794' Catchbasin SCV039 Fabian Way (Palo Alto) 25 Sept. 2001 37º 25.563' 122º 06.249' Industrial Manhole SCV041 Lower Adobe Creek (Palo Alto) 25 Sept. 2001 37º 25.773' 122º 06.319' Mixed Open Channel SCV042 Lower Stevens Creek (Mountain View) 25 Sept. 2001 37º 24.389' 122º 04.135' Mixed Open Channel SCV043 Upper Stevens Creek (Santa Clara County) 25 Sept. 2001 37º 17.019' 122º 07.334' Open Open Channel SCV044 Wrigley Creek (North American Transformer, Milpitas) 25 Sept. 2001 37º 24.949' 121º 53.646' Industrial Open Channel SCV045 Upper Adobe Creek (Santa Clara County) 2 Oct. 2001 37º 20.824' 122º 09.742' Open Open Channel SMC001 Douglas Pump Station (Redwood City) 24 Oct. 2000 37º 29.230' 122º 12.392' Industrial Pump Station SMC002 Broadway Pump Station (Redwood City) 24 Oct. 2000 37º 29.233' 122º 12.822' Industrial Pump Station SMC003 Maple Ave. and South Canal (South San Francisco) 25 Oct. 2000 37º 39.027' 122º 24.833' Res./Com. Pump Station SMC004 Veterans Pump Station (Redwood City) 24 Oct. 2000 37º 29.832' 122º 14.223' Res./Com. Pump Station SMC005 Coyote Point Pump Station (San Mateo) 25 Oct. 2000 37º 35.213' 122º 19.918' Res./Com. Open Channel SMC006 Cordilleras Creek at Edmunds (San Mateo County) 24 Oct. 2000 37º 28.513' 122º 16.972' Open Open Channel SMC007 McGarvey Creek and West Union Confluence (San Mateo County) 25 Oct. 2000 37º 26.665' 122º 17.463' Open Open Channel SMC008 Bear Gulch Creek nr. Woodside Road (Woodside) 24 Oct. 2000 37º 24.975' 122º 16.147' Open Open Channel SMC009 South Maple Pump Station (South San Francisco) 25 Oct. 2000 37º 38.657' 122º 25.002' Mixed Pump Station SMC010 Colma Creek near West Orange Avenue (South San Francisco) 25 Oct. 2000 37º 39.190' 122º 25.533' Mixed Open Channel

STATION SAMPLING WATERSHED SAMPLING GENERAL SITE LOCATION LATITUDE LONGITUDE ID DATE PROPERTIES LOCATION

SMC011 Bradford Pump Station (Redwood City) 24 Oct. 2000 37º 29.332' 122º 13.622' Mixed Pump Station SMC012 San Mateo Creek near South Grant Street (San Mateo) 25 Oct. 2000 37º 34.203' 122º 19.115' Mixed Open Channel SMC013 Poplar Creek near N. Bayshore Blvd. and E. Poplar Ave. (San Mateo) 25 Oct. 2000 37º 34.857' 122º 19.407' Mixed Open Channel SMC015 Adrian Road @ David Road Pump Station (Burlingame) 6 Sept. 2001 37º 35.633' 122º 22.095' Industrial Pump Station SMC016 1740 Rollins Road Pump Station (Burlingame) 6 Sept. 2001 37º 35.875' 122º 22.627' Industrial Pump Station SMC017 1592 Marsten @ Rollins Road Pump Station (Burlingame) 6 Sept. 2001 37º 35.529' 122º 21.955' Industrial Pump Station SMC019 O’Connor Street Pump Station (East Palo Alto) 20 Sept. 2001 37º 27.659' 122º 07.455' Industrial Pump Station SMC020 Delta Star (San Carlos) 20 Sept. 2001 37º 31.054' 122º 15.829' Industrial Manhole SMC021 Bayport Avenue and Center Street (San Carlos) 20 Sept. 2001 37º 29.921' 122º 14.755' Industrial Manhole SMC022 Bransten Road (San Carlos) 20 Sept. 2001 37º 30.378' 122º 15.213' Industrial Catchbasin SMC023 Pulgas Creek Pump Station (San Carlos) 25 Sept. 2001 37º 30.290' 122º 14.949' Industrial Pump Station SMC024 Colma Creek @ Collins Avenue (Colma) 6 Sept. 2001 37º 40.442' 122º 27.421' Res./Com. Open Channel SMC025 Beatty Avenue (Brisbane) 20 Sept. 2001 37º 42.399' 122º 23.887' Industrial Manhole SMC026 430 East Grand Avenue (South San Francisco) 6 Sept. 2001 37º 39.051' 122º 23.022' Industrial Manhole SMC027 214 Shaw Road (South San Francisco) 6 Sept. 2001 37º 38.478' 122º 24.576' Industrial Manhole SMC028 Belmont Creek (San Carlos) 20 Sept. 2001 37º 31.228' 122º 15.964' Mixed Open Channel SMC029 Holly Street and Industrial Road (San Carlos) 20 Sept. 2001 37º 30.742' 122º 15.531' Res./Com. Open Channel SMC030 Jefferson Branch of Redwood Creek (Redwood City) 4 Oct. 2001 37º 28.851' 122º 14.074' Res./Com. Open Channel SMC031 Redwood Creek Main Branch (Redwood City) 4 Oct. 2001 37º 28.823' 122º 14.074' Res./Com. Open Channel CCC001 Goodrich Ave and Richmond Parkway (Richmond) 27 Oct. 2000 37º 58.490' 122º 21.900' Industrial Open Channel CCC002 Giant Street at Rheem Creek (Richmond) 27 Oct. 2000 37º 58.617' 122º 21.212' Industrial Catchbasin CCC003 209 Parr Boulevard (Richmond) 27 Oct. 2000 37º 58.062' 122º 22.178' Industrial Manhole CCC005 Parr Boulevard and Richmond Parkway (Richmond) 27 Oct. 2000 37º 58.068' 122º 22.512' Industrial Catchbasin CCC006 Hensley Street and Castro Street (Richmond) 27 Oct. 2000 37º 56.803' 122º 22.008' Industrial Catchbasin CCC007 Marina Vista Road (near Shell just west of I-680, Martinez) 27 Oct. 2000 38º 01.523' 122º 07.037' Industrial Catchbasin CCC009 Willow Creek near private road to Tefler (Pittsburg) 25 Oct. 2000 38º 01.808' 121º 54.883' Industrial Open Channel CCC010 Pine Creek near Castle Rock Road (Contra Costa County) 25 Oct. 2000 37º 53.605' 121º 59.630' Open Open Channel CCC011 Bollinger Creek North of Crow Canyon (Contra Costa County) 27 Oct. 2000 37º 47.102' 122º 00.083' Open Open Channel CCC012 San Ramon Creek near Sycamore Valley Road (Danville) 27 Oct. 2000 37º 48.673' 121º 59.037' Res./Com. Open Channel CCC013 Tassajarra Creek near Finley and Old School Road (Contra Costa County) 27 Oct. 2000 37º 48.185' 121º 51.652' Open Open Channel

STATION SAMPLING WATERSHED SAMPLING GENERAL SITE LOCATION LATITUDE LONGITUDE ID DATE PROPERTIES LOCATION

CCC014 Alhambra Creek near Brockwood Drive (Contra Costa County) 27 Oct. 2000 37º 57.622' 122º 07.490' Open Open Channel CCC015 Kellogg Creek and Vasco (Contra Costa County) 25 Oct. 2000 37º 52.317' 121º 41.942' Open Open Channel CCC016 Walnut Creek above Willow Pass Road (Concord) 27 Oct. 2000 37º 57.968' 122º 03.302' Mixed Open Channel CCC017 Rheem Creek near Giant Highway (Richmond) 27 Oct. 2000 37º 58.632' 122º 21.208' Mixed Open Channel CCC018 Wildcat Creek at 3rd Street (Richmond) 27 Oct. 2000 37º 57.623' 122º 22.005' Mixed Open Channel CCC019 Gregory Lane at Contra Costa Blvd. (Pleasant Hill) 25 Oct. 2000 37° 56.863' 122° 03.618' Res./Com. Open Channel CCC020 Stege Marsh (intersection of Regatta and open channel, Richmond) 1 Oct. 2001 37º 55.078' 122º 20.300' Industrial Open Channel CCC021 Pittsburgh Avenue (Richmond) 1 Oct. 2001 37º 57.790' 122º 22.287' Industrial Catchbasin CCC022 Landini Road and Erickson Road (Walnut Creek) 21 Sept. 2001 37º 57.484' 122º 02.241' Industrial Catchbasin CCC023 End of Canal Road (Richmond) 1 Oct. 2001 37º 54.376' 122º 21.852' Industrial Catchbasin CCC024 San Pablo Creek Outfall near Third Street (Richmond) 1 Oct. 2001 37º 58.048' 122º 22.010' Industrial Outfall CCC025 West Gertrude (Richmond) 1 Oct. 2001 37º 57.240' 122º 22.486' Industrial Catchbasin CCC026 Open Channel on Goodrick Avenue (Richmond) 1 Oct. 2001 37º 58.433' 122º 21.917' Industrial Open Channel CCC027 Sand Creek at Empire Mine Road (Contra Costa County) 21 Sept. 2001 37º 57.107' 121º 48.271' Open Open Channel CCC028 James Donlan Blvd. at Sommersville Road (Antioch) 3 Oct. 2001 37º 59.241' 121º 51.177' Industrial Outfall CCC029 Mt. Diablo Creek downstream of US Naval Weapon Station (Concord) 3 Oct. 2001 38º 00.824' 122º 00.906' Industrial Open Channel CCC030 Cerrito Creek at Pierce Road (Richmond) 2 Oct. 2001 37º 53.975' 122º 18.556' Industrial Open Channel CCC031 Wright Avenue at Harbor Way (Richmond) 1 Oct. 2001 37º 55.274' 122º 21.664' Industrial Catchbasin CCC032 West Cutting Blvd. Near South 1st Street (Richmond) 1 Oct. 2001 37º 55.523' 122º 22.146' Industrial Catchbasin CCC033 2501 Concord Blvd. near PG&E Substation (Concord) 21 Sept. 2001 37º 58.648' 122º 01.735' Industrial Catchbasin MCS001 Miller Creek at end of Marinwood (Marinwood) 30 Oct. 2000 38º 01.842' 122º 32.287' Open Open Channel MCS002 Mahone Creek near intersection of E St. and First Street (Larkspur) 30 Oct. 2000 37º 58.257' 122º 32.007' Mixed Open Channel MCS003 Novato Creek at Downtown Novato Center (Novato) 30 Oct. 2000 38º 06.437' 122º 34.732' Mixed Open Channel MCS004 Arroyo Corte Madera Del Presidio at La Goma (Mill Valley) 30 Oct. 2000 37º 53.872' 122º 32.192' Mixed Open Channel MCS005 Pacheco Pond (Novato) 25 June 2001 38º 04.670' 122º 31.879' Mixed Pond MCS006 Mouth of Arroyo San Jose into Pacheco Pond (Novato) 25 June 2001 38º 04.230' 122º 31.737' Mixed Open Channel MCS007 Outfall into Arroyo San Jose (Novato) 25 June 2001 38º 04.219' 122º 31.868' Res./Com. Outfall MCS008 Outfall into Pacheco Creek (Novato) 25 June 2001 38º 03.962' 122º 31.808' Industrial Outfall MCS009 Pacheco Creek downstream of Railroad Tracks (Novato) 25 June 2001 38º 03.713' 122º 31.641' Industrial Open Channel MCS010 Outfall at the end of Commercial Boulevard (Novato) 25 June 2001 38º 04.063' 122º 31.755' Industrial Catchbasin

STATION SAMPLING WATERSHED SAMPLING GENERAL SITE LOCATION LATITUDE LONGITUDE ID DATE PROPERTIES LOCATION

MCS011 Pacheco Creek at the end of Pacheco Creek Road (Novato) 25 June 2001 38º 03.033' 122º 33.245' Mixed Open Channel MCS012 Unlined Drainage Ditch near Lovell (San Rafael) 25 June 2001 37º 57.989' 122º 31.349' Industrial Open Channel MCS013 Open Ditch through the San Rafael Improvement Club (San Rafael) 25 June 2001 37º 58.495' 122º 32.334' Res./Com. Open Channel VFC001 Austin Creek at Hwy 37 (Vallejo) 31 Oct. 2000 38º 07.590' 122º 16.153' Res./Com. Open Channel VFC001 Austin Creek at Hwy 37 (Vallejo) 16 Oct. 2001 38º 07.585' 121º 16.155' Res./Com. Open Channel VFC002 Stream bed below flume exiting Chabot Lake (Vallejo) 31 Oct. 2000 38º 08.467' 122º 14.790' Res./Com. Open Channel VFC003 Hiddenbrooke (Vallejo) 31 Oct. 2000 38º 08.503' 122º 10.209' Open Detention Basin VFC004 Midway Street (Vallejo) 16 Oct. 2001 38º 05.446' 121º 14.867' Industrial Open Channel VFC005 Lake Dalwigk (Vallejo) 16 Oct. 2001 38º 05.676' 121º 14.512' Mixed Detention Basin VFC006 Nebraska Street (Vallejo) 16 Oct. 2001 38º 06.880' 121º 15.128' Res./Com. Manhole VFC007 Hiddenbrooke/Sky Valley (Vallejo) 16 Oct. 2001 38º 08.472' 122º 10.278' Open Detention Basin VFC008 Lemon Street (Vallejo) 16 Oct. 2001 38º 05.392' 121º 14.415' Res./Com. Open Channel VFC009 Fairgrounds Drive at Coach Lane (Vallejo) 16 Oct. 2001 38º 07.848' 121º 13.719' Mixed Open Channel VFC010 Sonoma Boulevard at Couch Street (Vallejo) 16 Oct. 2001 38º 06.953' 121º 15.257' Res./Com. Open Channel FSS001 Union Creek and Highway 12 (Solano County) 31 Oct. 2000 38º 13.777' 121º 58.547' Industrial Open Channel FSS002 Suisun Creek and Cordelia (Fairfield) 31 Oct. 2000 38º 13.487' 122º 06.427' Open Open Channel FSS003 Green Valley Creek and Cordelia (Cordelia) 3 Nov. 2000 38º 12.712' 122º 07.765' Res./Com. Open Channel FSS004 Huntington Drive (Fairfield) 21 Sept. 2001 38º 16.651' 121º 58.523' Industrial Catchbasin FSS005 Outfall into Ledgewood Creek at Cordelia Road (Solano County) 21 Sept. 2001 38º 14.012' 122º 03.478' Mixed Outfall FSS006 McCoy Creek at Walters Court (Fairfield) 21 Sept. 2001 38º 16.148' 121º 59.498' Industrial Open Channel SCV = Santa Clara Valley Urban Runoff Pollution Prevention Program SMC = San Mateo Countywide Stormwater Pollution Prevention Program CCC = Contra Costa Clean Water Program MCS = Marin County Stormwater Pollution Prevention Program VFC = Vallejo Flood Control and Sanitation District FSS = Fairfield-Suisun Sewer District

19 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

2.2 Sampling Procedures

General procedures used for sediment sampling were based on those in use by the Regional Monitoring Program for Trace Substances in the San Francisco Estuary (RMP) and the USGS National Water Quality Assessment Program (NAWQA). Development of sampling procedures was coordinated with complementary work being conducted by the Alameda Countywide Clean Water Program to assure comparability of results.

2.2.1 Initial Sampling Equipment Decontamination

Sediment sampling equipment was prepared in the laboratory a minimum of four days prior to sampling. Cleaning methods followed protocols adapted from the NOAA National Status and Trends Program for use by the Regional Monitoring Program (Bell et al., 1999). Sampling equipment included:

• 6” x 6” Petite Ponar Grab (all 316 Stainless Steel) • Tefzel-coated sampling scoops and spoons • Tefzel-coated compositing buckets • Wash bottles and storage containers for deionized water • Wash bottles for hydrochloric acid and methanol

Prior to sampling, all equipment was thoroughly cleaned. Equipment was soaked (fully immersed) for three days in 2% Micro® solution and deionized water. Equipment was rinsed three times in deionized water and let dry in a clean place. Equipment was then rinsed with a 1.0% solution of hydrochloric acid, followed by a rinse with deionized water to eliminate the acid. A rinse was then conducted with petroleum ether, followed by another set of three rinses with deionized water. Equipment was then allowed to dry in a clean place.

The cleaned Ponar Grab and compositing buckets were wrapped in aluminum foil until used in the field. All other equipment was stored in clean ZiplocTM bags until used in the field.

2.2.2 Field Sampling Procedures

Final coordinates for sediment sampling were determined in the field using a Garmin GPS. Field log sheets were compiled for each site that recorded the sampling date, crew members names, narrative description of the sampling site (general location) and all specific substations (places where samples are collected for compositing, including whether sediment was under water or exposed), and the method used to collect sediment. To reduce variability within any sampling site, a minimum of three equal volume replicates were taken whenever possible. These replicates were mixed to produce a single site composite.

Similar to procedures established for NAWQA (Shelton and Capel, 1994), one of two sampling methods was used to collect sediments, depending upon the site-specific conditions:

1. Tefzel-coated spoons/scoops – This was the primary sampling method used at most open channel or catchbasin sites. Spoon/scoops were used to collect sediments adjacent to obstructions and other areas for which other sampling gear was inappropriate. Samples were then placed directly into the compositing bucket.

2. Stainless Steel Petite Ponar Grab – The Ponar Grab was used for collection in shallow and deeper open channels, and at pump stations. The Ponar Grab is a self-closing sampler using a spring- loaded Pinch-PinTM that releases when the sampler impacts the bottom and the lowering cable or

20 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

line becomes slack. The top of each scoop has a removable stainless steel screen (583 micron) to allow water to flow through the sampler during descent. This lessens the frontal shock wave created by descent and reduces surface disturbance. Both screens are covered with neoprene rubber flaps that open during descent for water flow through, and close during retrieval to prevent sample wash out. After the Ponar Grab was retrieved, the surface water was allowed to drain off, and sediment was removed with a Tefzel-coated spoon/scoop and placed in the compositing bucket.

In order to be considered acceptable, the grab samples were required to satisfy a set of quality criteria. Samples were rejected if the Ponar grab did not close fully allowing sample to wash out or if removal of the overlying water resulted in significant wash out of sediment fines.

The compositing bucket was covered with aluminum foil when not in use. No sieving of sediments was performed in the field, however, larger debris and cobble were removed from the samples using a Tefzel- coated spoon. Sieving was later performed by the laboratory to remove excess cobble and debris prior to their analysis. At the conclusion of sample collection at each site, all sediment was composited in the buckets and then subsampled for distribution to the appropriate laboratories. Disposable powder free nitrile gloves were worn while collecting and compositing samples to mitigate potential contamination. Gloves were changed between each location to reduce the potential for cross-contamination.

All sampling equipment was rinsed with native water, when possible, or deionized water between uses at different depositional areas. All sampling equipment used at a particular sampling location was field cleaned prior to use at a different sampling location. The field-cleaning protocol involved 1) removal of sediments using a scrub brush and native water or deionized water; 2) scrubbing of sampling gear and compositing equipment with a 2% Micro® solution and deionized water; 3) rinse with deionized water; 4) rinse with a 1.0% solution of HCl; 5) rinse with methanol; and 6) rinse with deionized water.

At the conclusion of sample processing at each sampling location, all samples were wrapped in protective material and stored on ice in the field. At the conclusion of a day’s sampling, all samples were either stored overnight on dry ice or removed to a freezer for temporary storage prior to distribution to the analytical laboratories.

2.3 Analytical Methods and Reporting Limits

Analytical methods and detection limits for sediments are summarized in Tables 4 through 6. All detection limits were based upon dry weight with the assumption that sediments will contain no more than 20% water by weight. Laboratory method detection limits (MDL) and method reporting limits (MRL) are summarized in Table 5 for specific PCB congeners being analyzed. The list of PCB congeners analyzed included the 40 congeners listed in the Regional Monitoring Program QAPP (Lowe, Hoenicke, and Davis 1999), two extra congeners included in the Alameda County program (Applied Marine Science 2000), and 12 additional congeners. During the first year of the program, the laboratory inadvertently analyzed PCB199 rather than PCB201 due to minor differences in nomenclature between the IUPAC naming and number conventions utilized by this program and BZ (Ballschmiter and Zell 1980) conventions used by the laboratory. This was corrected during the second season of testing. Project detection limits for individual organochlorine pesticides analyzed during the second season of testing are summarized in Table 6.

Samples were collected in a single, 1-liter, widemouth glass container for transfer to the laboratory where the samples were homogenized, sieved through a 2 millimeter screen and then subsampled for each of the analyses in Tables 4 through 6. All samples were maintained at 2-4 degrees centigrade while being stored and during the shipping process.

21 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 4. Parameters, methods, and holding times for analysis of sediments

Method Holding Times Units Parameter Method Reporting (dry wt.) Limit 14 days extract Individual PCB Congeners ng/g EPA 8082 0.5 40 days analysis 14 days extract Organochlorine Pesticides ng/g EPA 8081 A 0.1 40 days analysis Total Mercury ng/g EPA 7471a 2-4 28 days Particle Size (Puget Sound Protocol ASTM % NS1 6 months2 with hydrogen peroxide digestion) D422M/PSEP EPA Percent Moisture % 0.1 6 months2 160.3/PSEP ASTM TOC % D4129- 0.1 14 days 82/PSEP

ASTM=American Society for Testing and Materials, PSEP=Puget Sound Estuary Project

1. NS indicates that the Target Detection Limit is not specified. 2. Maximum recommended limits if samples are sealed and refrigerated during storage.

22 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 5. Comparison of Regional Monitoring Program target method detection limits (MDLs) for PCB congeners with project specific MDLs and method reporting limits (MRLs)

PCB RMP Target Laboratory Laboratory PCB Congener Name Congener MDL1 MDL MRL (IUPAC) Number ng/g ng/g ng/g (IUPAC) 8 2,4'-Dichlorobiphenyl 1 0.5 0.5 18 2,2',5-Trichlorobiphenyl 1 0.2 0.5 28 2,4,4'-Trichlorobiphenyl 1 0.09 0.5 31 2,4',5-Trichlorobiphenyl 1 0.5 33 2,3',4'-Trichlorobiphenyl 1 0.1 0.5 44 2,2',3,5'-Tetrachlorobiphenyl 1 0.07 0.5 49 2,2',4,5'-Tetrachlorobiphenyl 1 0.1 0.5 52 2,2',5,5'-Tetrachlorobiphenyl 1 0.07 0.5 56 2,3,3',4'-Tetrachlorobiphenyl 1 0.5 60 2,3,4,4'-Tetrachlorobiphenyl 1 0.3 0.5 66 2,3',4,4'-Tetrachlorobiphenyl 1 0.09 0.5 70 2,3',4',5-Tetrachlorobiphenyl 1 0.5 0.5 74 2,4,4',5-Tetrachlorobiphenyl 1 0.5 0.5 772 3,3',4,4'-Tetrachlorobiphenyl NA 0.5 812 3,4,4',5-Tetrachlorobiphenyl NA 0.5 87 2,2',3,4,5'-Pentachlorobiphenyl 1 0.07 0.5 902 2,2',3,4',5-Pentachlorobiphenyl NA 0.5 95 2,2',3,5',6-Pentachlorobiphenyl 1 0.5 0.5 97 2,2',3,4',5'-Pentachlorobiphenyl 1 0.5 0.5 99 2,2',4,4',5-Pentachlorobiphenyl 1 0.5 0.5 101 2,2',4,5,5'-Pentachlorobiphenyl 1 0.2 0.5 105 2,3,3',4,4'-Pentachlorobiphenyl 1 0.4 0.5 110 2,3,3',4',6-Pentachlorobiphenyl 1 0.05 0.5 1142 2,3,4,4',5-Pentachlorobiphenyl NA 0.5 118 2,3',4,4',5-Pentachlorobiphenyl 1 0.07 0.5 1232 2,3',4,4',5'-Pentachlorobiphenyl NA 0.5 1262 3,3',4,4',5-Pentachlorobiphenyl NA 0.5 128 2,2',3,3',4,4'-Hexachlorobiphenyl 1 0.2 0.5 132 2,2',3,3',4,6'-Hexachlorobiphenyl 1 0.5 0.5 138 2,2',3,4,4',5'-Hexachlorobiphenyl 1 0.3 0.5 141 2,2',3,4,5,5'-Hexachlorobiphenyl 1 0.5 0.5 149 2,2',3,4',5',6-Hexachlorobiphenyl 1 0.5 0.5 151 2,2',3,5,5',6-Hexachlorobiphenyl 1 0.5 0.5 153 2,2',4,4',5,5'-Hexachlorobiphenyl 1 0.2 0.5 156 2,3,3',4,4',5-Hexachlorobiphenyl 1 0.09 0.5 1572 2,3,3',4,4',5'-Hexachlorobiphenyl NA 0.5 158 2,3,3',4,4',6-Hexachlorobiphenyl 1 0.3 0.5 1662 2,3,4,4',5,6-Hexachlorobiphenyl NA 0.5 1672 2,3',4,4',5,5'-Hexachlorobiphenyl NA 0.5 1692 3,3',4,4',5,5'-Hexachlorobiphenyl NA 0.5 170 2,2',3,3',4,4',5-Heptachlorobiphenyl 1 0.3 0.5 174 2,2',3,3',4,5,6'-Heptachlorobiphenyl 1 0.5 0.5 177 2,2',3,3',4,5',6'-Heptachlorobiphenyl 1 0.5 0.5 180 2,2',3,4,4',5,5'-Heptachlorobiphenyl 1 0.3 0.5 183 2,2',3,4,4',5',6-Heptachlorobiphenyl 1 0.08 0.5 1842 2,2',3,4,4',6,6'-Heptachlorobiphenyl NA 0.5 187 2,2',3,4',5,5',6-Heptachlorobiphenyl 1 0.2 0.5 1892 2,3,3',4,4',5,5'-Heptachlorobiphenyl NA 0.5 194 2,2',3,3',4,4',5,5'-Octachlorobiphenyl 1 0.07 0.5 195 2,2',3,3',4,4',5,6-Octachlorobiphenyl 1 0.5 0.5 1992 2,2',3,3',4,5,5',6'-Octachlorobiphenyl NA 0.5 0.5 201 2,2',3,3',4,5',6,6'-Octachlorobiphenyl 1 0.5 203 2,2',3,4,4',5,5',6-Octachlorobiphenyl 1 0.5 2062 2,2',3,3',4,4',5,5',6-Nonachlorobiphenyl NA 0.07 0.5 2092 2,2',3,3',4,4',5,5',6,6'-Decachlorobiphenyl NA 0.08 0.5

MDL=Method Detection Limit, MRL=Method Reporting Limit, NA= Not Applicable, not reported by the RMP. 1. Lowe, S., R. Hoenicke, and Jay Davis 1999. 1999 Quality Assurance Project Plan, Regional Monitoring Program for Trace Substances. RMP Contribution #33 2. Congeners not included in 1999 RMP QAPP.

23 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 6. Comparison of Regional Monitoring Program target method detection limits (MDLs) for organochlorine pesticides with project specific MDLs and method reporting limits (MRLs)

RMP Approximate Minimum Organochlorine Target Laboratory Laboratory Pesticides MDL1 MDL RL ng/g ng/g ng/g alpha-BHC 1 0.1 1 beta-BHC 1 0.1 1 gamma-BHC (lindane) 1 0.2 1 delta-BHC 1 0.4 1 Heptachlor 1 0.1 1 Aldrin 1 0.2 1 Heptachlor Epoxide 1 0.1 1 gamma-Chlordane 1 0.1 1 Endosulfan I 1 0.1 1 alpha-Chlordane 1 0.1 1 Dieldrin 1 0.3 1 4,4’-DDE 1 0.2 1 Endrin 1 0.1 1 Endosulfan II 1 0.2 1 4,4’-DDD 1 0.2 1 Endrin Aldehyde 1 0.4 1 Endosulfan Sulfate 1 0.2 1 4,4’-DDT 1 0.2 1 Endrin Ketone 1 0.2 1 Methoxychlor 1 0.2 1 Hexachlorobenzene 1 0.05 1 Chlorpyrifos 1 0.06 1 Oxychlordane 1 0.1 1 2,4’-DDE 1 0.05 1 trans-Nonachlor 1 0.04 1 2,4’-DDD 1 0.04 1 cis-Nonachlor 1 0.04 1 2,4’-DDT 1 0.2 1 Mirex 1 0.06 1

2.4 Field and Laboratory Quality Control

2.4.1 Field Sampling Quality Control

Six field replicate samples were collected contemporaneously with an original sample and submitted blind to the laboratories performing the analyses. Identical recovery techniques were used for both the original and the replicate sample, and both samples were treated in an identical manner during storage, transportation, and analysis.

Data quality objectives are not established for field replicates. Replicates were included in this study to assess the extent of field variability since this could have a strong influence on data interpretation.

2.4.2 Laboratory Quality Control

A thorough laboratory quality control program was established in the 2000 Quality Assurance Project Plan (Kinnetic Laboratories, Inc. 2000) and the 2001 update (Kinnetic Laboratories, Inc. 2001). This

24 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002 program included use of method blanks, matrix spike/spike duplicates, surrogate spikes, laboratory control samples (LCS), certified reference materials (CRM) and laboratory replicates. Table 7 summarizes the frequency of each control check for each analyte and established data quality objectives.

Table 7. Summary of laboratory quality control performance measures and data quality objectives for measurement of mercury, PCBs, organochlorine pesticides, TOC and grain size in sediments.

Minimum QA Sample QA Measure Criteria Frequency Contamination by One per batch Method Blank

One per 20 field

samples Certified Reference Material (TOC every 15 RPD (if n=2) <35% samples) Precision RSD (if n>2) <35% RSD of last 7 CRMs <35% RPDs TOC <20% Hg <30% One per 20 field Laboratory Replicates Precision PCBs <40% samples OP Pest <40% RSD PSDs <20% Laboratory Control Samples (LCS) One per 20 field Within 20-25% (required for TOC and Accuracy and Precision samples consensus value Grain Size, optional for PCBs and Hg) Percent Recoveries TOC 85-115% Matrix Spikes One per 20 field Accuracy Hg 60-130% (Spike and Spike Duplicate) samples PCBs 30-150% OC Pest 30-150% Percent Recoveries Surrogate Spikes % Recovery One per sample PCBs 30-150% OC Pest 30-150%

1. Table based upon Tables 4a through 4c of the Regional Monitoring Program QAPP (Lowe, Heineke and Davis 1999). MDL=Method Detection Limit; RPD=Relative Percent Difference; RSD=Relative Standard Deviation

25 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

2.5 Data Analysis

2.5.1 Descriptive Statistics

Summary statistics included the number of samples, the percentage of the samples with values above project reporting limits, minimum and maximum detected values, the median, 25th and 75th percentiles, mean, standard deviation and coefficient of variation. In the previous year, statistics were developed based upon natural logarithm transformations and evaluation based upon characteristics of the lognormal distribution. Upon normalization to fines of the data for both years, portions of the data set were found to depart substantially from the lognormal distribution. In addition, handling of below detection limit data was becoming an issue for a few of the target analytes. In order to allow inclusion of the censored (below detection limit) data, a method known as regression on order statistics (ROS) was used. An Excel add-in tool developed by Caltrans (Shumway andArzari, 2000) was used to determine descriptive statistics for each target analyte. This tool was developed based upon procedures described by Helsel and Cohn (1988) and Helsel (1980).

The ROS method is based upon mapping the ordered (both censored and uncensored), log transformed data onto a probability scale. A least squares line is then fit by regressing the log transformed concentrations to the uncensored data on the probability scale. Censored data points are then assigned values based upon the ordered plotting positions and the regression equation. Summary statistics are then developed based upon the complete data set including estimated values for censored data.

2.5.2 Hypothesis Testing

Two hypotheses were established to address the primary objectives of this study. These included:

Data were first normalized to the organic-free, fine fraction (<62.5 micron) of the sediment samples. Data were first log transformed and examined graphically as an initial screening process to evaluate normality. In a few cases, data was noted to depart substantially from normality. Due to both normality issues and problems with addressing below detection limit data, nonparametric statistical tests were used to address the two hypotheses. The Kruskal-Wallis one-way analysis of variance test was used for comparison involving more than two land use categories. This was followed by Bonferonni’s Inequality Test, a nonparametric, a posterior test. The Mann-Whitney test was used for two-sample comparisons. Statistical significance for all tests was based upon the probability of Type 1 or α-error being less than or equal to 0.05.

2.5.3 Stormwater Load Estimations

Estimates of contaminant loading to San Francisco Bay associated with stormwater were developed with the general procedures outlined in the Simple Method (Schueler 1987) since it requires a relatively modest amount of information to obtain rough, planning level estimates of loading. Many of the input values for the estimates were based upon Davis et al. (2000) in order to provide consistent structure. The

26 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Simple Method estimates loads for chemical constituents as the product of annual runoff volume and pollutant concentrations, as:

L=0.226*R*C*A

Where: L = Annual Load (lbs) R = Annual Runoff (inches) C = Land Use Specific Pollutant Concentration (mg/L) A = Area (acres) 0.226 = Unit Conversion Factor

Pollutant concentrations (C) were developed for each major land use category (residential, commercial, industrial, agriculture and open) based upon the assumption that the concentration of each constituent in suspended solids could be estimated from storm drain sediment concentrations that were normalized to the fine fraction. This was defined as the proportion of the sample consisting of sediments less than 62.5 microns after digestion by hydrogen peroxide to remove organic material.

The median was used as the best estimate of the concentration of each pollutant in the fine fraction. A range of concentrations was developed based upon use of the 25th and 75th percentile concentrations associated with each land use. For two constituents, only four measurements were available from open land use areas. In these special cases, the minimum and maximum values were used in lieu of the 25th and 75th percentile values and the average value was used in place of the median.

Pollutant concentrations (C) were calculated for each land use by the following formulas:

Cm = (Median Sediment Concentration, normalized to fines) * (Average TSS Concentration) th C25 = (25 Percentile Sediment Concentration, normalized to fines) * (Average TSS Concentration) th C75 = (75 Percentile Sediment Concentration, normalized to fines) * (Average TSS Concentration)

Annual Runoff was calculated as:

R=P*Pj*Rv

Where: R = Annual Runoff (inches) P = Annual Rainfall (inches) Pj = Fraction of annual rainfall events that produce runoff Rv = Runoff Coefficient

Estimates of runoff coefficients (Rv) were calculated for each major land use category based upon a linear relationship with the impervious fraction associated with each category. The following relationship established by Schueler (1987) was used to obtain estimates of runoff coefficients.

Rv=0.05+0.91Ia

Where: Ia = Impervious Fraction

Estimates of annual runoff (P) were obtained from long term rainfall records maintained by the National Climatic Data Center. A default value of 0.9 was used for the fraction of annual rainfall that is expected to produce runoff (Pj).

27 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

2.5.3.1 Watersheds and Land Use

Watershed delineations were based upon CALWATER (Version 2.0) hydrologic areas (CDF&G 1999) used by the San Francisco Estuary Institute (SFEI; Davis et al. 2000) to estimate contaminant loads to coastal waters in the Bay area as part of a legislative mandate (AB 1429). The SFEI project elected to eliminate all areas of each watershed greater than 20 square miles that were located above a dam. The primary rationale for this approach was the assumption that significant proportions of particulates would be trapped in the reservoirs, substantially reducing transport of the particulates and associated contaminants to the receiving waters. For our purposes, only the seventeen watersheds entering San Francisco Bay were used to estimate contaminant loads (Figure 11).

Davis et al. (2000) also developed generalized land use within each watershed based upon the 1995 Association of Bay Area Governments (ABAG 1996) land use data set. The approximately 160 detailed classifications were reduced to five categories: residential, commercial, industrial, agriculture and open. Data from Davis et al were used to determine the total area of each watershed and the percentage of each major land use category (Table 8).

28 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Figure 11. Hydrologic areas (HA) used to develop stormwater runoff loading estimates for San Francisco Bay . Hydrologic subareas shown for each HA. Rain gauge sites marked on the map were used to develop annual rainfall estimates for each HA.

29 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April 2002

Table 8. Summary of drainage areas and land use within each hydrologic area tributary to San Francisco Bay1

Drainage Watersheds Area (Hydrologic Areas) (Sq. Meters) Residential Commercial Industrial Agricultural Open Alameda Creek 940,853,470 10 3 3 11 73 Berkeley 87,585,261 57 16 18 0 9 Concord 648,267,817 35 7 6 5 47 Concord (220731) 283,955,162 25 10 7 9 49 Concord (220732) 212,544,012 44 4 1 1 50 Concord (220733) 121,715,016 39 6 7 0 47 Concord (220734) 30,053,627 46 9 26 6 12 Coyote Creek 473,402,458 23 6 7 10 53 East Bay Cities 537,837,394 44 9 12 1 34 Fairfield 877,893,352 7 3 2 28 60 Fairfield (220721) 226,198,776 12 1 5 12 70 Fairfield (220722) 131,685,843 0 0 0 13 86 Fairfield (220723/26) 410,248,260 8 6 2 48 36 Fairfield (220724/25) 109,760,473 0 0 0 1 99 Fremont Bayside 191,146,170 26 6 11 8 49 Guadalupe River 215,171,511 47 8 5 5 35 Napa River 937,888,979 10 3 1 24 62 Novato 183,975,415 23 7 1 13 56 Palo Alto 593,745,251 43 10 8 1 39 Petaluma River 377,643,849 14 1 2 35 48 Pinole 152,427,916 33 5 12 0 49 San Francisco Bayside 28,764,911 58 39 2 0 1 San Mateo Bayside 426,680,239 41 10 12 0 37 San Rafael 157,659,876 50 8 1 0 41 Sonoma Creek 429,766,542 8 1 1 36 54 GRAND TOTALS 7,260,710,411 23.8 5.5 5.1 13.8 51.7 1. Source: Davis et al. 2000. Contaminant Loads from Stormwater to Coastal Waters in the San Francisco Bay Region. (Data excludes areas located above dams that were greater than 20 square miles.)

2.5.3.2 Runoff Coefficients

Default impervious fractions (Ia) were estimated for each major land use category based upon impervious fraction estimates developed for ABAG land use categories in the Santa Clara Basin Watershed Management Initiative, Volume 1, Watershed Characteristics Report (2000). A final weighted average impervious fraction was developed based upon the relative amount of each ABAG land use category within the five reclassified land use categories. The resulting estimates of impervious fraction and runoff coefficients for each land use are summarized in Table 9.

As a comparison, runoff coefficients used by Davis et al. (2000) are presented in Table 10. Davis used the average values for each land use as the best estimate. Resulting coefficients for commercial, industrial and agricultural land use were very similar, however, our estimates are substantially higher for residential land use and substantially lower for open land use.

30 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 9. Estimates of impervious fraction and calculated runoff coefficients for each major land use category.

Residential Commercial Industrial Agricultural Open

Ia (Impervious 0.60 0.94 0.91 0.025 0.01 Fraction) R v 0.60 0.91 0.89 0.07 0.06 (Runoff Coefficient)

Table 10. Runoff coefficients used by SFEI (2000)1

Rv (Runoff Coefficient) Residential Commercial Industrial Agricultural Open

Low 0.20 0.60 0.60 0.05 0.10 Average (Best) 0.35 0.90 0.90 0.10 0.25 High 0.50 0.95 0.95 0.20 0.50 1.Values based upon BASMAA (1996) and SCCWRP (2000)

2.5.3.3 Rainfall

Estimates of annual average rainfall were developed for each hydrologic area based upon long-term records from National Climatic Data Center (NCDC) Cooperative rain gauges located throughout the region. Data were available from 25 stations (Table 11, Figure 12). Data from one to five stations were averaged to obtain the best estimate of rainfall within each hydrologic area (Table 12). In most cases, gauges are located in urban areas at lower elevations. This tends to result in a bias towards lower total rainfall for each watershed. This, in turn, may result in lower overall estimates of TSS loading. However, since the particular contaminants of concern in this study are most strongly associated with the urban environment, this approach is not likely to introduce a significant bias in estimates of contaminant loading.

Figure 12. Location of rain gauges used for estimates of annual precipitation. Numbers refer to NCDC Coop designations.

Table 11. Summary of data used to estimate average annual rainfall in each region.

NCDC COOP Station Name Record Record Annual Average Latitude Longitude Station Start Date End Date Rainfall (inches) (deg:min) (deg:min) Number

047880 San Rafael Civic Center 7/1/48 12/31/00 35.85 38:00 122:31 044500 Kentfield 7/1/48 12/31/00 48.70 37:57 122:33 046826 Petaluma Fire Station 7/1/48 12/31/00 25.24 38:14 122:38 047769 San Francisco WSO AP 7/1/48 12/31/00 20.07 37:37 122:23 047864 San Mateo 7/1/48 12/31/78 18.77 37:32 122:18 047339 Redwood City 7/1/48 12/31/00 19.96 37:29 122:14 049792 Woodside Fire Station 3/1/73 11/30/00 30.27 37:26 122:15 047912 Santa Clara University 1/3/31 5/31/76 14.04 37:21 121:56 047821 San Jose 7/1/48 12/31/00 14.49 37:21 121:54 045123 Los Gatos 7/1/48 12/31/00 24.78 37:14 121:58 046144 Newark 7/1/48 12/31/00 14.62 37:31 122:02 040693 Berkeley 1/1/19 12/31/00 23.34 37:52 122:15 045378 Martinez Water Plant 2/1/70 12/31/00 19.59 38:01 122:07 047070 Port Chicago Naval Depot 7/8/48 9/30/75 15.72 38:01 122:01 045915 Mount Diablo Junction 4/1/52 12/31/00 24.10 37:52 121:56 047661 St. Mary’s College 7/1/48 7/31/81 28.20 37:50 122:06 046336 Oakland Museum 10/1/70 12/31/00 23.31 37:48 122:16 049185 Upper San Leandro Fltr 7/1/48 12/31/00 25.39 37:46 122:10 046335 Oakland WSO AP 7/1/48 9/30/00 18.03 37:44 122:12 044997 Livermore 4/1/30 12/31/00 14.48 37:40 121:46 045993 Mt. Hamilton 7/1/48 12/31/00 23.19 37:20 121:39 042934 Fairfield 12/4/50 12/31/00 22.44 38:17 122:04 046074 Napa State Hospital 2/1/17 12/31/00 24.64 38:17 122:16 047772 San Fran Mission Dolore 1/1/14 12/31/00 20.88 37:46 122:26 047767 San Francisco Richmond 7/1/48 12/31/00 19.90 37:46 122:30

33 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 12. Average annual rainfall values used for each hydrologic area.

Hydrologic Area Average Rainfall Alameda Creek 18.84 Berkeley 23.34 Concord 19.80 Coyote Creek 14.38 East Bay Cities 21.91 Fairfield 22.44 Fremont Bayside 14.62 Guadalupe River 24.78 Napa River 24.64 Novato 30.54 Palo Alto 20.74 Petaluma River 25.24 Pinole 21.47 San Francisco Bayside 20.39 San Mateo Bayside 22.27 San Rafael 48.70 Sonoma Creek 24.94

2.5.3.4 Contaminant Concentrations

Contaminant load concentrations for each constituent were based on the assumption that 100 percent of each contaminant was associated with the fine (less than 62.5 micron) fraction of the storm drain sediments. It was also assumed that concentration of contaminants in the fine fraction were representative of the concentration of the constituents in the suspended material in the stormwater discharges as measured by Total Suspended Solids (TSS). This requires the further assumption that TSS measures obtained by autosamplers in stormwater monitoring programs are limited to particles less than 62.5 microns. Data from regional programs conducted in the San Francisco Bay area were used as input for suspended sediment concentrations from each major land use category (Table 13). Since no information was available for sediment loads from agricultural activities in the Bay area, a generalized estimate for California developed by the Southern California Coastal Water Research Project (2000) was utilized. For our purposes, only the average TSS concentrations were used to develop load estimates. Variability was only evaluated by utilization of the 25th and 75th percentile concentrations of each specific contaminant. Additional variability introduced by estimates of TSS for each land use category (Table 13) and differences in annual rainfall were not included in the load estimates.

Table 13. Estimates of the range and average concentration of suspended solids in stormwater runoff from the five major land use classifications.

TSS (mg/L) Residential1 Commercial1 Industrial1 Agricultural2 Open1

Low 28 30 49 646 27 Average (Best) 90 98 157 2068 85 High 286 312 502 6618 272

1. Data as summarized from BASMAA (1996) by Davis et al. (2000) 2. Data from SCCWRP (2000)

34 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

3.0 RESULTS AND DISCUSSION

The field survey was conducted during two time periods. Initial sampling was completed at all Marin County sites on June 25, 2001. The remaining sites were sampled over a five-week period between September 6th and October 16th, 2001.

All surveys were completed before the occurrence of any substantial rainfall in the Bay area. A minor event yielding only 0.10 to 0.13 inches of rain occurred on September 25th, 2001. This event was not large enough to produce runoff of any significance.

As general reference, a summary of the results of 2000 survey is presented in Table 14. These summary data are limited to constituents that were also measured during the second year of the program.

Data from the 2001 survey are summarized in Tables 15 and 16. Bulk chemistry results are summarized in Table 15 for all major constituents analyzed during this year’s program. Organic compounds are grouped into the major categories of PCBs, DDTs, chlordane, HCH2, and endosulphan. Complete breakdown of chemical results for each of these categories is included in Appendix A. Table 16 provides of summary of all organic data normalized to the percent of organic-free, fines (silt/clay sized particles) present in the sediments. These data were used for all statistical comparisons and for estimation of contaminant loads.

Primary data (total PCBs, mercury, chlordane and DDT) are graphically summarized in Figures 13 through 16. All graphical summaries are based upon data that have been normalized to the fine fraction (particles less than 62.5 microns). In the case of total PCBs and mercury, the graphical summaries include results of both the 2000 and 2001 sediment testing programs. Alameda County data is limited to testing conducted during the 2000 program. The 2001 program in Alameda County consisted primarily of second round sampling at the same sites they investigated in the first year of the program.

2 HCH refers to the sum of the alpha, beta, gamma (lindane) and delta isomers of hexachlorocyclohexane

35 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 14. Summary of first year (2000) data results from each sampling location

Total PCBs Mercury, Total Mercury, Total PCBs normalized to normalized to Total Organic Solids, Total Site Name Total % Fines1 (ug/Kg) fines fines Carbon (%) (%) (mg/Kg) (ug/Kg fines) (mg/Kg fines) Contra Costa County CCC001 184.1 229 0.47 0.58 80.5 4.61 47.2 CCC002 194.6 1052 0.26 1.41 18.5 3.64 69.9 CCC003 174.9 723 0.36 1.49 24.2 2.08 68.1 CCC005 171.8 366 0.28 0.60 47.0 1.9 76.0 CCC006 165.0 2705 0.72 11.8 6.1 2.38 77.6 CCC007 15.7 121 0.68 5.23 13.0 3.25 73.5 CCC007FR2 23.3 197 0.31 2.63 11.8 3.26 71.3 CCC009 4.7 4.9 0.07 0.07 96.2 1.66 50.8 CCC010 4.6 7.0 0.04 0.06 65.4 4.38 24.0 CCC011 0.0 0.0 0.04 0.38 10.6 0.54 75.3 CCC012 4.3 23 0.03 0.16 18.5 0.93 85.1 CCC013 0.0 0.0 0.03 0.11 26.4 0.26 67.4 CCC014 0.0 0.0 0.03 0.07 43.6 1.17 76.9 CCC015 0.0 0.0 0.02 0.06 33.7 1.25 54.2 CCC016 19.7 55 0.15 0.42 35.9 4.69 34.0 CCC017 133.3 479 0.11 0.40 27.8 2.56 50.8 CCC018 5.6 9.1 0.1 0.16 61.5 1.03 60.4 CCC019 58.2 1293 0.19 4.22 4.5 4.24 65.1 Fairfield Suisun FSS001 114.1 161 0.06 0.08 70.9 2.71 58.7 FSS002 0.3 1.3 0.03 0.13 22.9 1.57 37.2 FSS003 0.2 1.0 0.02 0.10 19.7 0.69 59.7 Marin County MCS001 0.0 0.0 0.03 0.09 32.5 3.12 75.5 MCS002 72.3 1721 0.36 8.57 4.2 1.5 66.5 MCS003 0.0 0.0 0.05 1.39 3.6 0.38 73.8 MCS004 18.9 48 0.09 0.23 39.4 1.58 37.5 Santa Clara County SCV001 26746 61912 1.08 2.50 43.2 6.93 48.2 SCV002 654.4 1674 0.33 0.84 39.1 10.1 42.1 SCV002FR 132.6 285 0.29 0.62 46.5 7.85 37.3 SCV003 92.2 3179 0.1 3.45 2.9 1.33 71.9 SCV004 95.9 505 0.15 0.79 19.0 7.36 58.2 SCV005 66.3 102 0.68 1.05 64.8 15.9 36.9 SCV006 56.0 467 1.18 9.83 12.0 3.9 69.7 SCV007 12.2 321 0.1 2.63 3.8 2.36 71.2 SCV007FR 12.0 273 0.15 3.41 4.4 2.1 67.7 SCV008 34.2 44.3 0.13 0.17 77.2 4.08 34.9 SCV009 57.0 91.1 0.16 0.26 62.6 10.2 38.8 SCV009FR 17.6 28.1 0.16 0.26 62.7 5.55 41.8 SCV010 24.0 136 0.07 0.40 17.7 2.38 58.6 SCV011 14.1 371 0.13 3.42 3.8 11.2 69.2 SCV012 124.0 579 4.26 19.9 21.4 1.56 64.1 SCV014 30.5 258 0.18 1.53 11.8 3.78 71.3 SCV015 1.1 6.5 0.07 0.41 16.9 0.66 68.9

36 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Total PCBs Mercury, Total Mercury, Total PCBs normalized to normalized to Total Organic Solids, Total Site Name Total % Fines1 (ug/Kg) fines fines Carbon (%) (%) (mg/Kg) (ug/Kg fines) (mg/Kg fines) SCV016 29.8 113 0.04 0.15 26.4 1.63 54.2 SCV017 0.2 7.7 0.03 1.15 2.6 0.51 62.7 SCV018 38.4 267 0.20 1.39 14.4 3.51 56.7 SCV018FR 50.3 359 0.21 1.50 14.0 3.51 54.1 SCV019 101.2 1281 0.31 3.92 7.9 5.59 57.2 SCV021 10.0 89 0.12 1.07 11.2 2.38 69.6 SCV022 328.0 8200 0.09 2.25 4.0 2.44 66.0 SCV024 13.9 1069 0.05 3.85 1.3 0.05U 69.9 San Mateo County SMC001 69.8 71 0.17 0.17 98.5 1.52 34.5 SMC002 116.2 2371 0.07 1.43 4.9 1.86 73.9 SMC003 225.3 711 0.17 0.54 31.7 5.91 49.3 SMC004 83.7 1947 0.11 2.56 4.3 1.44 73.4 SMC005 187.4 424 0.2 0.45 44.2 3.77 36.4 SMC006 5.8 43 0.04 0.30 13.4 0.79 76.8 SMC007 0.0 0.0 0.03 0.57 5.3 0.29 69.8 SMC008 3.6 4.6 0.04 0.05 77.7 2.35 36.4 SMC009 476.9 2110 0.24 1.06 22.6 3.83 56.0 SMC009FR 226.8 1187 0.21 1.10 19.1 2.94 65.8 SMC010 194.8 566 0.06 0.17 34.4 3.63 38.6 SMC011 338.8 360 0.44 0.47 94.1 4.4 24.5 SMC012 12.4 258 0.05 1.04 4.8 0.82 70.9 SMC013 89.5 1904 0.11 2.34 4.7 1.35 69.7 Vallejo VFC001 144.7 228 0.18 0.28 63.4 3.01 52.7 VFC002 0.8 1.1 0.15 0.21 70.3 2.57 35.6 VFC003 0.3 0.4 0.29 0.39 75.3 2.08 33.6

1. % Fines indicates the organic-free fraction of the sediment less than 62.5 microns. 2. The suffix “FR” indicates that the sample was a field replicate.

Table 15. Summary of data results from 2001 survey at each sampling location.

Total Mercury, Total Total HCH3 Endosulfan4 Aldrin Dieldrin Endrin Chlorpyrifos Mirex Site Name % Fines TOC (%) PCBs Total DDT1 Chlordane2 (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (mg/Kg) (ug/Kg) (ug/Kg) Contra Costa County CCC020 22 5.0 145 0.14 18 47 ND ND ND ND ND 4.8 ND CCC021 20 2.1 18 0.14 22 2.3 ND ND ND ND ND ND ND CCC022 39 6.3 26 0.23 ND 53 ND ND ND ND ND 4.2 ND CCC023 15 3.0 242 0.34 43 4.3 ND ND ND ND ND ND ND CCC024 72 1.3 16 0.08 14 8.0 ND ND ND 6.5 ND ND ND CCC024FR5 78 3.2 116 0.20 89 16 ND ND ND 5.9 ND ND ND CCC025 25 3.5 144 0.27 66 41 ND ND ND 15 ND ND ND CCC026 61 6.7 156 0.47 17 11 ND ND ND 6 ND 23 ND CCC027 59 0.9 ND 0.03 0.16 0.42 0.19 ND ND ND ND ND ND CCC028 78 2.1 7 0.07 ND 1.3 ND ND ND ND ND ND ND CCC029 24 1.8 ND 0.07 6 ND ND ND ND ND ND ND ND CCC030 77 7.7 297 0.63 74 203 ND ND ND ND ND 24 ND CCC031 23 7.7 1153 0.41 215 ND ND ND ND ND ND ND 6.6 CCC032 16 2.7 1945 0.95 ND ND ND ND ND ND ND ND ND CCC033 44 13 13 0.32 ND ND ND ND ND ND ND ND ND Fairfield Suisun FSS004 15 1.0 ND 0.08 ND 0.82 ND ND ND ND ND ND ND FSS005 82 3.9 1 0.14 23 ND ND ND ND ND ND ND ND FSS006 90 3.6 ND 0.12 ND ND ND ND ND ND ND ND ND Marin County MCS005 96 2.6 0.24 0.15 19 4.3 ND ND ND ND ND ND ND MCS006 31 3.7 16 0.27 15 55 ND ND ND 3.4 ND ND ND MCS007 1 1.6 2 0.38 1.8 8.6 ND ND ND ND ND ND ND MCS008 4 1.8 4 0.24 83 622 ND ND ND ND ND ND ND MCS009 50 4.8 13 0.22 132 25 ND ND ND ND ND ND ND MCS009FR 53 3.0 6 0.14 102 20 ND ND ND ND ND ND ND MCS010 4 2.1 36 0.33 22 14 2.4 ND ND ND ND 46 ND MCS011 2 1.7 ND 0.26 0.74 0.11 0.64 ND ND ND ND ND ND

Total Mercury, Total Total HCH3 Endosulfan4 Aldrin Dieldrin Endrin Chlorpyrifos Mirex Site Name % Fines TOC (%) PCBs Total DDT1 Chlordane2 (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (ug/Kg) (mg/Kg) (ug/Kg) (ug/Kg) MCS012 75 1.5 52 0.38 41 52 ND ND 11 28 ND 3 ND MCS013 9 6.7 14 0.21 38 47 ND ND ND 6.6 ND ND ND Santa Clara County SCV025 49 1.5 42 0.32 23 12 ND ND ND ND ND 4.1 ND SCV026 13 1.7 3 0.14 56 ND ND ND ND ND ND ND ND SCV027 13 2.9 ND 0.27 108 0.53 ND ND ND ND ND ND ND SCV028 3 1.0 165 0.80 40 51 ND ND ND ND ND ND ND SCV029 12 2.4 768 0.08 177 63 ND ND ND ND ND ND ND SCV030 19 5.3 60 0.30 12 ND ND ND ND ND ND ND ND SCV031 31 6.5 47 1.9 6.8 9.1 ND ND ND ND ND ND ND SCV032 23 5.3 294 0.29 23 23 ND ND ND ND ND ND ND SCV033 17 2.1 1215 0.30 ND 86 ND ND ND ND ND ND ND SCV034 47 11 135 3.04 ND 11300 ND ND ND ND ND ND ND SCV035 26 7.1 338 0.14 55 14 ND ND ND ND ND ND ND SCV036 17 6.3 84 0.18 ND 98 ND ND ND ND ND ND ND SCV037 13 6.0 70 0.13 39 22 ND ND ND ND ND ND ND SCV037FR 14 6.5 51 0.12 27 17 ND ND ND ND ND ND ND SCV038 18 7.3 593 0.34 80 110 ND ND ND ND ND 6.1 ND SCV039 10 53 312 0.05 ND ND ND ND ND ND ND 11 ND SCV041 18 2.0 3 0.03 8.1 5.8 ND ND ND ND ND 1.3 ND SCV041FR 26 2.2 1 0.04 2.1 15 ND ND ND ND ND ND ND SCV042 17 2.2 33 0.06 3.8 6.9 ND ND ND ND ND ND ND SCV043 13 1.0 ND 0.09 0.11 0.12 ND ND ND ND ND ND ND SCV044 22 1.4 533 0.05 37 ND ND ND ND ND ND ND ND SCV045 6 0.5 0.44 0.02 ND 0.24 ND ND ND ND ND ND ND San Mateo County SMC0026 11 2.2 61 NS7 47 13 ND ND ND ND ND 4.7 ND SMC0096 4 0.8 70 NS 30 11 ND ND ND ND ND ND ND SMC0116 78 2.3 124 NS 11 28 ND ND ND ND ND 25 ND SMC015 53 7.6 63 0.12 1572 63 ND ND ND ND ND 18 ND SMC016 4 1.3 81 0.10 18 45 ND ND ND ND ND 4.2 ND SMC016FR 3 1.2 87 0.11 31 65 ND ND ND ND ND ND ND SMC017 87 6.4 139 0.35 ND 21 ND ND ND ND ND 29 ND SMC019 23 2.5 70 0.13 92 34 ND ND ND 6.8 ND 4.3 0.85 SMC020 16 7.5 20294 1.84 4010 1956 590 ND ND ND ND ND ND SMC021 40 11 1222 0.92 59 166 ND ND ND ND ND ND ND

1. DDT indicates the sum of 4,4’DDT, 2,4’DDT, 4,4’DDE, 2,4’DDE, 4,4’DDD, and 2,4’DDD. 2. Total chlordane indicates the sum of alpha-chlordane, beta-chlordane, cis-nonachlor, trans-nonachlor, heptachlor, heptachlor epoxide and oxychlordane. 3. HCH refers to the sum of the alpha, beta, gamma (lindane) and delta isomers of hexachlorocyclohexane. 4. Endosulfan refers to the sum of endosulfan I, endosulfan II, and endosulfan sulfate. 5. The suffix “FR” indicates that the sample was a field replicate. 6. Data from 2001 resampling at three San Mateo County pump stations in support of PCB case studies. 7. NS=Not Sampled These analyses not included with the case study.

Table 16. Summary of 2001 data normalized to the fine fraction (<62.5 microns).

Mercury, Total Total PCBs Total DDT1 HCH3 Endosulfan4 Aldrin Dieldrin Endrin Chlorpyrifos Mirex Site Name Total Chlordane2 (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (mg/Kg fines) (ug/Kg fines) Contra Costa County CCC020 648 0.63 79 211 ND ND ND ND ND 22 ND CCC021 91 0.70 109 12 ND ND ND ND ND ND ND CCC022 68 0.59 ND 135 ND ND ND ND ND 11 ND CCC023 1624 2.3 287 29 ND ND ND ND ND ND ND CCC024 22 0.11 19 11 ND ND ND 9.0 ND ND ND CCC024FR5 149 0.26 114 20 ND ND ND 7.5 ND ND ND CCC025 568 1.1 261 163 ND ND ND 59 ND ND ND CCC026 254 0.76 27 18 ND ND ND 10 ND 37 ND CCC027 ND 0.05 ND 0.7 ND ND ND ND ND ND ND CCC028 9 0.09 ND 1.7 ND ND ND ND ND ND ND CCC029 ND 0.29 25 ND ND ND ND ND ND ND ND CCC030 386 0.82 96 264 ND ND ND ND ND 31 ND CCC031 4915 1.7 916 ND ND ND ND ND ND ND 28 CCC032 12538 6.1 ND ND ND ND ND ND ND ND ND CCC033 29 0.72 ND ND ND ND ND ND ND ND ND Fairfield Suisun FSS004 ND 0.52 ND 5 ND ND ND ND ND ND ND FSS005 1.5 0.17 28 ND ND ND ND ND ND ND ND FSS006 ND 0.13 ND ND ND ND ND ND ND ND ND Marin County MCS005 0.25 0.16 20 4.5 ND ND ND ND ND ND ND MCS006 52 0.88 49 179 ND ND ND 11 ND ND ND MCS007 259 40 188 898 ND ND ND ND ND ND ND MCS008 89 5.9 2054 15396 ND ND ND ND ND ND ND MCS009 25 0.44 264 50 ND ND ND ND ND ND ND MCS009FR 12 0.26 192 38 ND ND ND ND ND ND ND MCS010 1006 9.3 604 399 67 ND ND ND ND 1292 ND MCS011 ND 14 40 5.9 34 ND ND ND ND ND ND MCS012 69 0.51 54 69 ND ND 15 37 ND 4.0 ND MCS013 166 2.5 441 550 ND ND ND 77 ND ND ND

Mercury, Total Total PCBs Total DDT1 HCH3 Endosulfan4 Aldrin Dieldrin Endrin Chlorpyrifos Mirex Site Name Total Chlordane2 (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (ug/Kg fines) (mg/Kg fines) (ug/Kg fines) Santa Clara County SCV025 87 0.66 46 25 ND ND ND ND ND 8.4 ND SCV026 27 1.1 450 ND ND ND ND ND ND ND ND SCV027 ND 2.0 813 4.0 ND ND ND ND ND ND ND SCV028 5398 26 1307 1680 ND ND ND ND ND ND ND SCV029 6306 0.66 1453 515 ND ND ND ND ND ND ND SCV030 318 1.6 60 ND ND ND ND ND ND ND ND SCV031 155 6.2 22 30 ND ND ND ND ND ND ND SCV032 1270 1.3 98 97 ND ND ND ND ND ND ND SCV033 7082 1.7 ND 501 ND ND ND ND ND ND ND SCV034 289 6.5 ND 24296 ND ND ND ND ND ND ND SCV035 1315 0.54 214 54 ND ND ND ND ND ND ND SCV036 480 1.0 ND 561 ND ND ND ND ND ND ND SCV037 548 1.0 309 172 ND ND ND ND ND ND ND SCV037FR 368 0.87 198 126 ND ND ND ND ND ND ND SCV038 3381 1.9 456 627 ND ND ND ND ND 35 ND SCV039 3125 0.50 ND ND ND ND ND ND ND 110 ND SCV041 18 0.17 45 33 ND ND ND ND ND 7.3 ND SCV041FR 2.7 0.15 8.1 58 ND ND ND ND ND ND ND SCV042 197 0.36 23 41 ND ND ND ND ND ND ND SCV043 ND 0.67 0.8 0.9 ND ND ND ND ND ND ND SCV044 2443 0.23 171 ND ND ND ND ND ND ND ND SCV045 7.4 0.34 ND 4.0 ND ND ND ND ND ND ND San Mateo County SMC0026 560 NS7 427 123 ND ND ND ND ND 43 ND SMC0096 1735 NS 734 268 ND ND ND ND ND ND ND SMC0116 159 NS 14 36 ND ND ND ND ND 32 ND SMC015 119 0.23 2973 119 ND ND ND ND ND 34 ND SMC016 1942 2.4 425 1074 ND ND ND ND ND 100 ND SMC016FR 2695 3.4 969 2003 ND ND ND ND ND ND ND SMC017 160 0.40 ND 24 ND ND ND ND ND 33 ND SMC019 304 0.56 397 148 ND ND ND 29 ND 19 3.7 SMC020 124198 11 24541 11971 3611 ND ND ND ND ND ND SMC021 3018 2.3 146 410 ND ND ND ND ND ND ND SMC022 2709 0.65 220 93 ND ND ND ND ND ND ND SMC023 25044 3.6 1410 902 ND ND ND ND ND ND ND

1. DDT indicates the sum of 4,4’DDT, 2,4’DDT, 4,4’DDE, 2,4’DDE, 4,4’DDD, and 2,4’DDD. 2. Total chlordane indicates the sum of alpha-chlordane, beta-chlordane, cis-nonachlor, trans-nonachlor, heptachlor, heptachlor epoxide and oxychlordane. 3. HCH refers to the sum of the alpha, beta, gamma (lindane) and delta isomers of hexachlorocyclohexane. 4. Endosulfan refers to the sum of endosulfan I, endosulfan II, and endosulfan sulfate. 5. The suffix “FR” indicates that the sample was a field replicate. 6. Data from 2001 resampling at three San Mateo County pump stations in support of PCB case studies 7. NS=Not Sampled These analyses not included with the case study..

43 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

Figure 13. Concentrations of total PCBs measured in storm drains throughout the study area. All data normalized to the fine fraction (<62.5 micron). Includes all data from 2000 and 2001 surveys. Alameda data limited to 2000 survey.

44 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

Figure 14. Concentrations of total mercury measured in storm drains throughout the study area. All data normalized to the fine fraction (<62.5 micron). Includes all data from 2000 and 2001 surveys. Alameda data limited to 2000 survey.

45 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

Figure 15. Concentrations of total chlordane measured in storm drains throughout the study area. All data normalized to the fine fraction (<62.5 micron).

46 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

Figure 16. Concentrations of total DDT measured in storm drains throughout the study area. All data normalized to the fine fraction (<62.5 micron).

47 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

3.1 Major Land Use

Several approaches were used to examine distributional patterns of PCBs, total mercury, chlordane and DDT in stormwater conveyances. The first of these was to aggregate data in the major land use categories listed under watershed characteristics in Table 1.

For PCBs and total mercury, data from both the 2000 and 2001 testing programs were used. Data provided by the Alameda Countywide Clean Water Program for the first year of monitoring was incorporated into this analysis. Watershed characteristics were assigned using information provided on the Program’s web site and same criteria were used to categorize sites sampled by this study. The combined data set resulted in a total of 164 samples that were analyzed for total PCBs and 153 samples analyzed for total mercury.

A statistical summary of data grouped by major land use characteristics is presented in Table 17. Summary statistics included the number of samples, the percentage of the samples with values above project reporting limits, minimum and maximum detected values, the median, 25th and 75th percentiles, mean, standard deviation and coefficient of variation. HCHs, aldrin, dieldrin, chlorpyrifos and mirex were uncommon in the storm drain sediments at the project detection limits. Endosulphan and endrin were never detected above detection limits. Box plots were used to provide a graphical comparison of the distributional characteristics of each of the four most common constituents (total PCB, mercury, chlordane and DDT) within each major land use category (Figure 17). All box plots are based upon log transformations of data normalized to the fine fraction of the sediment. A log(x+1) transformation was utilized to accommodate the nondetect values that were treated as zero.

A Kruskal-Wallis one-way analysis of variance and Dunn’s test were used to determine if significant differences existed among major land use characteristics and, if so, which watershed characteristics differed (Table 18). These tests indicated that concentrations of total PCBs, total mercury and chlordane were significantly different (p<0.05) among the four land use categories. No statistically significant differences were evident in concentrations of DDT among land use categories.

Application of Bonferonni’s inequality test indicated sediments from industrial and residential/commercial land use categories contained similar concentration of total PCBs. The median concentration of total PCBs from measurements at 68 industrial sites was 378 ug/Kg fines while the median concentration at 31 residential/commercial sites was 327 ug/Kg fines. Concentrations of total PCBs in open land use sites were significantly lower than all other categories with a median concentration of 2.2 ug/Kg fines. Concentrations of total mercury in storm drain sediments were found to be similar in residential/commercial (1.40 mg/Kg fines), industrial (1.22 ug/Kg fines) and mixed (0.75 ug/Kg fines) land use sites. Open land use sites had significantly lower levels of total mercury with a median concentration of 0.21 ug/Kg fines.

Similar analyses were also performed for both total chlordane and DDT. These analyses were conducted as an initial examination of these data but low replication in open land use areas must be considered when interpreting the results. Concentrations of total chlordane were found to be significantly higher in residential/commercial (89 ug/Kg fines) catchments. The median concentration of chlordane in mixed land use catchments was 7.4 ug/Kg fines while sediments from open land use drainages had a median concentration of just 1.0 ug/Kg fines.

Concentrations of DDT were statistically similar in sediments from all four land use categories.

48 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

Table 17. Statistical summary of results by major land use classification. All data normalized to the fine fraction (<62.5 microns). Bolded values represent exact calculations Units % Min Max Percentiles Parameter n Mean SD CV (Dry Wt) Detected Detected Detected 25th 50th 75th Industrial Total PCBs ug/Kg fines 68 94.1 4.84 124198 4455 19121 4.29 94 445 2264 Total Mercury mg/Kg fines 65 100 0.07 28.79 2.44 4.57 1.87 0.504 0.84 2.34 Total Chlordane ug/Kg fines 45 77.8 1.7 24296 1315 4849 3.7 2.8 69.0 404 Total DDT ug/Kg fines 45 68.9 19.1 24541 881 4915 5.6 16.4 97.9 426 Total HCH ug/Kg fines 45 4.4 2.4 590 13.2 Endosulphan ug/Kg fines 45 0 Aldrin ug/Kg fines 45 2.2 11 11 0.2 Dieldrin ug/Kg fines 45 13.3 4.4 2.8 1.5 Endrin ug/Kg fines 45 0 Chlorpyrifos ug/Kg fines 45 31 3.0 46 4.2 Mirex ug/Kg fines 45 4.4 0.85 6.6 0.2 Residential/Commercial Total PCBs ug/Kg fines 31 100 1.02 28003 2224 6629 2.98 96 431 713 Total Mercury mg/Kg fines 28 100 0.10 39.6 4.64 9.68 2.09 0.601 1.13 2.54 Total Chlordane ug/Kg fines 11 100 54.3 3744 964 1185 1.2 327 550 1633 Total DDT ug/Kg fines 11 72.7 7.2 1307 239 455 1.9 17.7 161 188 Total HCH ug/Kg fines 11 0 Endosulphan ug/Kg fines 11 0 Aldrin ug/Kg fines 11 0 Dieldrin ug/Kg fines 11 36.4 6.6 70 11 Endrin ug/Kg fines 11 0 Chlorpyrifos ug/Kg fines 11 18.2 45 1000 95 Mirex ug/Kg fines 11 Mixed Total PCBs ug/Kg fines 41 97.6 0.25 9113 724 2010 2.78 15.8 89.1 567 Total Mercury mg/Kg fines 36 100 0.15 8.57 1.31 1.80 1.37 0.240 0.55 2.03 Total Chlordane ug/Kg fines 13 84.6 4.5 268 65 87 1.3 5.2 34 109 Total DDT ug/Kg fines 13 84.6 14.1 734 123 240 1.9 17.1 39.8 98.0 Total HCH ug/Kg fines 13 7.7 0.64 0.64 0.1 Endosulphan ug/Kg fines 13 0 Aldrin ug/Kg fines 13 0 Dieldrin ug/Kg fines 13 7.7 3.4 3.4 0.52 Endrin ug/Kg fines 13 0 Chlorpyrifos ug/Kg fines 13 30.8 1.3 25 2.0 Mirex ug/Kg fines 13 0 Open Total PCBs ug/Kg fines 22 54.5 0.40 113 9.4 30 3.14 0.31 1.16 6.95 Total Mercury mg/Kg fines 22 100 0.05 14.0 0.92 4.10 4.44 0.086 0.32 0.43 Total Chlordane ug/Kg fines 4 75 0.72 4.0 1.6 2.2 1.3 3.3 Total DDT ug/Kg fines 4 25 0.82 0.82 Total HCH ug/Kg fines 4 25 0.19 0.19 Endosulphan ug/Kg fines 40 Aldrin ug/Kg fines 40 Dieldrin ug/Kg fines 40 Endrin ug/Kg fines 40 Chlorpyrifos ug/Kg fines 40 Mirex ug/Kg fines 4 0

49 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

15 4

)

E 1 2

10 I

G 1

B A 0

O ( 5

N ( -1

L -2

0 -3 l d n . al d n . ria e e m tri xe pe om st ix p o s i O C u M O ./C du M ./ nd s In es I Re R WATERSHED PROPERTIES WATERSHED PROPERTIES

12 12

10 ) 10

)

S 8 F 8

A / 6 T 6

(

N 4 4

(

2 L 2

0 0 al d n . al d n . tri xe pe om tri xe pe om s i O C s i O C du M ./ du M ./ n es n es I R I R WATERSHED PROPERTIES WATERSHED PROPERTIES

Horizontal Bar=Median, Upper Box Hinge=Third Quartile, Lower Box Hinge=First Quartile, Upper Whisker=Upper Hinge+(1.5*(Median-Upper Hinge)), Lower Whisker=Lower Hinge-(1.5*(Median-Lower Hinge)), *=Outside Value, o=Very Outside Value

Figure 17. Comparison of concentrations and distributional properties of total PCBs, total mercury, DDT and chlordane in the fine fraction of storm drain sediments from four major land use categories.

50 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

Table 18. Statistical comparison of analyte concentration in sediments from storm drainages in four major land use categories.

Analyte Probability2 Ranked Land Use Classifications3 (High to Low Concentrations)

Total PCBs1 <0.000 IND RES/COM MIXED OPEN

Total Mercury <0.000 RES/COM IND MIXED OPEN

Total Chlordane <0.000 RES/COM IND MIXED OPEN

Total DDT 0.134 RES/COM IND MIXED OPEN

1. Summation of all 54 congeners. 2. Significance of Kruskal-Wallis test based upon p<0.05. Significant probabilities are bolded. 3. Underlined land use categories indicate categories that are not significantly different (p<0.05) based upon Bonferonni’s inequality.

3.2 Urban and Nonurban

Land use types were further aggregated to compare concentrations of PCBs, mercury, chlordane and DDT in the urban portions of the study area with nonurban or open land use. Urban sites were considered to include industrial, residential/commercial, and mixed land use types. Nonurban were considered synonymous with open land use. Reclassification of sampling sites in this manner resulted in 143 urban and 21 nonurban sites for PCBs. A statistical summary of urban and nonurban data is presented in Table 19. As in the previous section, summary statistics are included only for contaminants that were commonly detected in the storm drain sediments. HCHs, aldrin, dieldrin, chlorpyrifos and mirex were uncommon in the storm drain sediments at the project detection limits. Endosulphan and endrin were never detected above detection limits. Box plots showing the distributional characteristics for each of the four commonly encountered sediment contaminants are presented in Figure 18. As in section 3.1, box plots are based upon natural log transformations of the concentrations of each contaminant after normalizing to the fine fraction. For plotting purposes, an ln(x+1) transformation was used to allow inclusion of nondetect values that were treated as zero.

Statistical comparisons (Table 20) using the Mann-Whitney test indicated that samples taken from urban environments had significantly higher levels of total PCBs, total mercury, chlordane and DDT. The median concentration of total PCBs and DDT in urban samples was on the order of two orders of magnitude or roughly 100 times the median value in open land use areas. Concentrations of chlordane in urban storm drains were approximately 40 times higher than observed in open land use areas. The median concentration of total mercury, however, was only 3 times greater in sediments from urban land use areas.

In the case of both chlordane and DDTs, the number of samples from open land use is limited to four sites. Although the differences between urban and nonurban sites was large and statistically significant, the numbers of samples from nonurban areas are far too low to be considered representative of open land use areas. Low numbers of open land use sites were sampled this year since the focus was on PCBs and improving coverage of areas characterized by industrial land use activities.

51 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

3.2.1 Comparison of Storm Drain Sediments with Recent San Francisco Bay Data

Data from the1999 and 2000 Regional Monitoring Program (RMP; SFEI 2000) and the Bay Protection and Toxic Cleanup Program (BPTCP; Hunt et al. 1998) were used as reference points for comparison with concentrations measured in Bay area storm drainages. For consistency, data obtained from these two programs were all normalized to the fine fraction. In 1999 and 2000, the RMP measured PCBs, mercury, chlordane and DDT in Bay sediments at 47 sites. Between one and three samples were taken at each site during this time period.

Data from these two programs suggest that concentrations of PCBs, chlordane and DDTs in Bay sediments are far less than those measured in storm drain sediments from urban areas (Figure 18). As an example, the median of PCB measurements by the RMP was 5.7 ug/Kg fines. The median concentration of PCBs reported by the BPTCP was 22.5 ug/Kg fines, approximately 1/10 of the median concentration in the urban storm drains. The median in urban storm drainages was 311 ug/Kg fines. In contrast, PCBs in sediments collected from open land use catches had a median concentration of only 1.16 ug/Kg fines.

In the case of total mercury, concentrations measured in sediments from urban stormwater drainages were found to be relatively comparable to values from the 1999 and 2000 RMP surveys as well as the earlier BPTCP studies. The median concentration of total mercury was 0.93 mg/Kg fines in urban storm drain sediments during the current study compared to 0.39 mg/Kg fines for all San Francisco Bay stations sampled by the RMP in 1999 and 2000 and 0.46 mg/Kg fines at BPTCP sites. The median concentration of mercury from open land use sites was 0.32 mg/Kg fines.

In all cases the contaminant concentrations in sediments sampled from open land use drainages contained levels near or lower than found in San Francisco Bay sediments or in sediments from urban storm drains (Figure 18).

52 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

Table 19. Statistical summary of results by urban and nonurban land use categories. All data normalized to the fine fraction (<62.5 microns). Bolded values represent exact calculations

Units % Min Max Percentiles Parameter n Mean SD CV (Dry Wt) Detected Detected Detected 25th 50th 75th Urban Total PCBs ug/Kg fines 140 96.4 0.25 124198 2868 13576 4.73 61.4 311 1254 Total Mercury mg/Kg fines 129 100 0.07 39.6 2.60 5.54 2.13 0.43 0.93 2.35 Total Chlordane ug/Kg fines 69 82.6 1.7 24296 1023 3957 3.9 8.5 69 433 Total DDT ug/Kg fines 69 72.5 7.2 24541 636 3966 6.2 13.5 78.9 353 Total HCH ug/Kg fines 69 4.3 0.64 590 Endosulphan ug/Kg fines 69 0 Aldrin ug/Kg fines 69 1.4 11 11 0.16 Dieldrin ug/Kg fines 69 15.9 3.4 70 2.8 Endrin ug/Kg fines 69 0 Chlorpyrifos ug/Kg fines 69 29 1.3 1000 18.4 Mirex ug/Kg fines 69 2.9 0.85 6.6 0.1 Nonurban (Open) Total PCBs ug/Kg fines 22 54.5 0.40 113 9.4 30 3.14 0.31 1.16 6.95 Total Mercury mg/Kg fines 22 100 0.05 14.0 0.92 4.10 4.44 0.086 0.32 0.43 Total Chlordane ug/Kg fines 4 75 0.72 4.0 1.6 2.2 1.3 1.6 3.3 Total DDT ug/Kg fines 4 25 0.82 0.82 Total HCH ug/Kg fines 4 25 0.19 0.19 Endosulphan ug/Kg fines 4 0 Aldrin ug/Kg fines 4 0 Dieldrin ug/Kg fines 4 0 Endrin ug/Kg fines 4 0 Chlorpyrifos ug/Kg fines 4 0 Mirex ug/Kg fines 4 0

53 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

Table 20. Statistical comparison of analyte concentration in sediments from urban and nonurban (open) storm drainages.

Urban Nonurban Analyte Probability1 Median Median

Total PCBs (ug/Kg fines) <0.000 311 1.16

Total Mercury (mg/Kg fines) <0.000 0.93 0.32

Total Chlordane (ug/Kg fines) <0.001 69 1.6

Total DDT (ug/Kg fines) <0.001 78.9 0.82

1. Mann-Whitney Test. Significance based upon p<0.05. Significant probabilities are bolded.

54 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

PCBs MERCURY

15 4

)

1 2

)

10 S

I 1

H 0

(

5 L N -1

-2

0 -3 Nonurban Urban Nonurban Urban URBAN URBAN

CHLORDANE DDT

12 12

) 10 10

)

S F 8 8

A / 6

6 T

(

4 N 4

(

L 2 2

0 0 Nonurban Urban Nonurban Urban

Figure 18. Comparison of concentrations and distributional properties of total PCBs, total mercury, chlordane and DDT in storm drains from urban and open land use areas tributary to San Francisco Bay.

55 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

PCBs MERCURY

10000.0

)

m p 10.00

(

p 1000.0

(

100.0 a

e 1.00

B 10.0

1.0 0.10

n n P P an an P CP a a M TC rb rb M T rb rb R P U u R P U u B n B on No N STUDY STUDY

CHLORDANE DDT

10000.0 10000.0

)

(

n 1000.0

(

i 1000.0

i a

c 100.0

i 100.0

n e

a 10.0

o 10.0

C 1.0 1.0 n n P CP a a P P an an M T rb rb M TC rb rb R P U u R P U u B on B n N No STUDY STUDY RMP=Regional Monitoring Program, BPTCP=Bay Protection and Toxics Cleanup Program, Urban and Nonurban (Open) indicate land use classifications used in this study.

Figure 18. Comparison of concentrations of total PCBs, mercury, chlordane and DDT in storm drains from urban and nonurban (open) land use areas with levels measured in San Francisco Bay sediments.

56 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002

3.3 Stormwater Load Estimates

Estimates of stormwater loads to San Francisco Bay are summarized by both hydrologic area (Table 21) and by major land use activity (Table 22). In both cases the low and high estimates of loading are based upon the 25th and 75th percentile estimates for contaminant concentrations and loading is estimated in terms of pounds per year.

Several major assumptions were necessary to develop these load estimates. It was first necessary to assume that contaminants in sediments from storm drains were 100 percent associated with particles less than 62.5 microns in size. Secondly, it had to be assumed that this fine fraction of the bedded sediment was representative of the suspended material measured as Total Suspended Solids (TSS) by stormwater monitoring programs. Considerable controversy exists regarding the latter assumption. There is substantial concern that TSS measurements are inherently inaccurate when measuring suspended material in storm water. The Suspended Sediment Concentration (SSC) method (ASTM D3977-97) is currently recommended for this purpose. USGS (Gray et al. 2000) analyzed results from 3,235 paired samples using both procedures and found the TSS method to be negatively biased by a factor of 25 to 34 percent and there was no clear relationship between the two procedures. Using good subsampling procedures, automated sampling procedures and TSS methods can include suspended material as large as 1 to 2 mm. Inconsistent procedures used by different sampling groups and laboratories introduce additional variability in TSS results. Using either the SSC or TSS methods, it is likely that the particle size composition of suspended material included with these measurements includes material larger than 62.5 microns. Since the particle size distribution of suspended matter in stormwater is rarely measured, it is unknown what proportion of measured TSS is represented by particles greater than 62.5 microns. This problem (violation of assumptions) would actually be exacerbated if the SSC method were in common use.

The overall load estimate for PCBs from urban sources was estimated to range from 19 and 228 pounds per year with a median estimate of 87 pounds per year (Table 22). Urban mercury loads were estimated at 115 to 497 pounds per year with a median of 210 pounds per year. Estimates of mercury loads associated with stormwater discharges specifically do not include large point sources associated with historic mercury mines located within San Francisco Bay watersheds. These sources were not characterized in this study.

Far less information is currently available for chlordane and DDT. Preliminary estimates were made based upon data from this year. Based upon this limited set of data it is estimated that 48 to 260 pounds of chlordane and 2.3 to 50 pounds of DDT are discharged to the Bay from urban sources on an annual basis (Table 22).

3.3.1 Assessment of potential impacts of cleanup activities on PCB loading

Alternative scenarios were analyzed to assess the possible effects of cleaning storm drains and/or contaminant sources contributing to those storm drains with high concentrations of PCBs. This evaluation was based upon the assumption that cleanup activities were conducted at all top 15 percentile sites. It was further assumed that the cleanup activities would be able to effectively reduce the sources to levels equivalent to either the 75th percentile (1,254 ug/Kg fines) or the 50th percentile (311 ug/Kg fines) concentrations in urban storm drains. The results of this assessment are summarized in Table 23 and Figure20.

This assessment indicated that cleaning up the top 15th percentile PCB sources to levels consistent with the 75th percentile urban site reduces the upper range of the loading estimate by 24% but has no impact on either the median or low estimates. PCB loads under this scenario were estimated to range from 19 to 173 pounds per year. By setting the cleanup standard to the median concentration of 311 ug/Kg fines, load

57 Draft - Urban Sources of Mercury, PCBs and Organochlorine Pesticides March, 2002 estimates were further reduced to 19 to 112 pounds per year, a reduction of just over 50% in the upper estimate from the baseline estimate. The estimate of the median load was reduced by 28%.

Table 21. Estimates of annual loading (lbs/year) of PCBs, total mercury, chlordane and DDT to San Francisco Bay from each major watershed.

Median High Low Median High Low Median High Low Median High Low WATERSHED Acres PCBs PCBs PCBs Total Hg Total Hg Total Hg Chlordane Chlordane Chlordane DDTs DDTs DDTs (lbs) (lbs) (lbs) (lbs) (lbs) (lbs) (lbs) (lbs) (lbs) (lbs) (lbs) (lbs) Alameda Creek 232597 4.8 13.7 1.0 17 35 7.7 4.3 13.3 2.4 1.2 2.9 0.13 Berkeley 21653 3.1 9.2 0.7 7 18 4.1 2.7 8.6 1.5 0.8 2.0 0.09 Concord 160264 9.4 23.9 2.1 25 57 13.0 9.3 28.7 5.3 2.6 5.3 0.25 Coyote Creek 115864 4.1 12.0 0.9 12 26 5.8 3.6 11.3 2.0 1.0 2.6 0.12 East Bay Cities 132964 12.5 35.9 2.7 30 72 16.5 11.1 34.8 6.2 3.2 7.7 0.35 Fairfield 217032 4.0 11.1 0.9 23 41 9.0 3.9 11.7 2.1 1.0 2.4 0.11 Fremont Bayside 47255 2.1 6.8 0.5 6 13 2.9 1.7 5.3 0.9 0.5 1.4 0.06 Guadalupe River 53194 4.6 10.5 1.0 12 28 6.4 4.9 15.0 2.9 1.3 2.4 0.12 Napa River 231864 4.8 10.7 1.1 26 47 10.2 5.5 16.2 3.1 1.4 2.5 0.12 Novato 45482 2.5 4.7 0.5 8 17 3.9 3.0 8.8 1.7 0.8 1.2 0.06 Palo Alto 148253 11.7 29.9 2.6 29 68 15.6 11.4 35.3 6.5 3.2 6.7 0.31 Petaluma River 93361 2.4 6.1 0.5 14 24 5.3 2.4 7.3 1.3 0.6 1.3 0.06 Pinole 37306 2.8 8.9 0.6 7 16 3.6 2.2 7.1 1.2 0.7 1.9 0.08 San Francisco Bayside 7111 0.9 1.6 0.2 2 5 1.2 1.1 3.2 0.6 0.3 0.4 0.02 San Mateo Bayside 105483 10.0 28.7 2.2 24 57 13.0 8.7 27.5 4.9 2.5 6.2 0.28 San Rafael 38976 5.9 10.6 1.3 16 35 8.3 7.3 21.7 4.3 1.9 2.7 0.14 Sonoma Creek 106246 1.6 3.9 0.4 13 22 4.6 1.7 5.0 0.9 0.4 0.9 0.04

TOTALS 1794905 87 228 19.2 271 578 131 85 261 48 23 51 2.3

59 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April 2002

Table 22. Estimates of annual loading of PCBs, total mercury, chlordane and DDT to San Francisco Bay from each major land use category.

Res Com Ind Ag Open All 17 Bay Watersheds Urban Areas Only Constituent (lbs/year) (lbs/year) (kg/year) (lbs/year) (kg/year)

PCBs Median 46 17 24 0.19 0.02 87 39.5 87 39.5 High 76 28 123 1.2 0.15 228 103.6 227 103.2 Low 35 13 21 0.37 0.05 19 8.6 19 8.6

Total Mercury Median 120 45 46 53.6 6.8 271 123.2 210 95.5 High 269 101 127 72 9.1 578 262.7 497 225.9 Low 64 24.0 27 14 1.8 131 59.5 115 52.3

Chlordane Median 58 22 3.7 0.67 0.09 85 38.6 84 38.2 High 173 65 22 0.67 0.09 261 118.6 260 118.2 Low 35 13 0.2 0.12 0.02 48 21.8 48 21.8

DDT Median 15 5.2 3.0 0.001 0.004 23 10.5 23 10.5 High 20 7.4 23 0.14 0.02 51 23.2 50 22.7 Low 1.1 0.4 0.9 0.00 0.00 2.3 1.0 2.3 1.0

Table 23. Comparison of current annual PCB load estimates with two source reduction scenarios. 1) cleanup of the top 15th percentile sites to the 75th percentile concentration (1254 ug/Kg fines) and to the 50th percentile concentration (311 ug/Kg fines).

PCBs Annual Load Percentage (lbs/year) Reduction Existing Conditions Median 87 NA High 228 NA Low 19 NA Cleanup to 75th Percentile Median 87 0 High 173 24 Low 19 0 Cleanup to 50th Percentile Median 62 28 High 112 51 Low 19 0

60 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

PCB Load Estimates Existing Estimate

250 Cleanup to 75th percentile 200 Cleanup to 50th percentile

150

100 Pounds per Year 50

Figure 20. PCB annual loading estimates based upon existing conditions two cleanup scenarios for the top 15th percentile sites. Error bars represent the 25th and 75th percentile load estimates. Bars indicate the median load estimate.

61 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

3.4 Top 15th Percentile Sites

Table 24 lists sites with concentrations of PCBs and mercury in the top 15th percentile of all sediments tested in the past two years. All concentrations are normalized to the fine fraction. Shaded areas indicate sites that were sampled in the previous year. Samples taken in 2000 are summarized in the table but are not discussed below. Many of the sites with higher concentrations from last year have been the subject of case studies. Figure 21 illustrates the general distribution of PCBs and total mercury measured at all sites.

Highest concentrations of PCBs were encountered in sediments collected from a manhole in an industrial catchment in San Carlos (SMC020). The sample was taken in front Spacesonic, on the west side of Industrial Road south of Quarry Road. This site receives runoff from an area that includes Delta Star (270 Industrial Road), a former generator of PCBs and manufacturer of transformers. PCBs and other pollutants have been found in soil samples from the site. The Regional Board has adopted site cleanup requirements (Order No. 99-062).

The second highest concentration was found in sediments collected from Colma Creek (SMC024). An oil and grease sheen has been observed at this site and at previous sites downstream of this sampling location. The sampling point is located inside a culvert under Collins Road. Several sources appear to converge at this point. All conveyances go underground at this point and mapping information was limited. The catchment includes a number of automobile sales and maintenance activities.

Sediments collected from the Pulgas Creek Pump Station (SMC023) in San Carlos contained the third highest concentration of PCBs. Sediments at this site were anoxic. The composition of the PCBs at this site showed signs of weathering typical of anoxic conditions. Congeners with three to four chlorine atoms were more abundant than expected in typical Arochlor formulations.

Sediments collected from a manhole near Nebraska Street in Vallejo (VFC006) contained the fourth highest concentrations of PCBs encountered this year. This is primarily a residential catchment that ultimately discharges through Austin Creek.

Samples collected at SMC040 contained the fifth highest levels of PCBs. This site was actually part of a case study investigation conducted to examine sources of PCBs to the South Maple Pump Station in Redwood City.

Sediment sampled for CCC032 was comprised of a composite of four catch basins located on the north side of West Cutting Boulevard near South 1st Street. PCBs in this composite sample were measured at 12,538 ug/Kg fines in the fine fraction of the sediment. Catch basin number one was located west of South 1st Street in front of the Richmond Distribution Center. Catch basin number two was located west of South 1st Street in front of the PG&E substation. Catch basin number three was located on the east side of South 1st Street at the intersection of West Cutting Boulevard next to Precision Machining. Catch basin number four was located on West Cutting Boulevard east of South 1st Street near the eastern edge of Precision Machining.

62 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002 this branch of the Creek includes Stulsaft Park on Farm Hill Boulevard. During redevelopment of the Park, it was discovered that the Park was the site of a mercury mine in the 1950’s. San Mateo County Health Service Agency and Redwood City discovered residual mercury contamination in the soils and have planned remediation efforts before continuing with the redevelopment process.

Table 25 lists sites that were found to have concentrations of total chlordane and DDT in sediments that ranked within the top 15th percentile of all sites investigated in 2001. These two groups of compounds were only measured during the past year. The general distribution of total chlordane and DDT among all sites is illustrated in rank order in Figure 22.

Many of the sites noted to have sediments with higher levels of PCBs and mercury were also found to have higher concentrations of both chlordane and DDT. Sediments from both the Delta Star site (SMC020) in San Carlos and Pulgas Creek Pump Station (SMC023) had high concentrations of both chlordane and DDTs. The highest concentration of chlordane (24,541 ug/Kg fines) was found in sediments from an industrial catchment in San Jose (SCV034). Sediment was collected from a manhole located at the convergence of 5th and 7th Streets. High levels of chlordane were also encountered at VFC006 in Vallejo where elevated levels of PCBs were also found.

63 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 24. List of sites identified with the top 15th percentile concentrations of total PCBs and total mercury from sampling conducted in both 2000 and 2001. (All data are normalized to the fine fraction.)

Concentration in Fine Site Date Collected Area Property Fraction Total PCBs (ug/Kg fines) SMC020 20-Sep-01 Industrial 124198 SCV0011 30-Oct-00 Industrial 61912 SMC024 06-Sep-01 Res./Com. 28003 SMC023 25-Sep-01 Industrial 25044 VFC006 16-Oct-01 Res./Com. 21231 SMC040 02-Oct-01 Industrial 13842 CCC032 01-Oct-01 Industrial 12538 ALA004 26-Sep-00 Mixed 9113 SCV022 27-Oct-00 Mixed 8200 SCV033 10-Sep-01 Industrial 7082 SCV029 05-Oct-01 Industrial 6306 SCV028 05-Oct-01 Res./Com. 5398 CCC031 01-Oct-01 Industrial 4915 SCV038 25-Sep-01 Industrial 3381 SCV003 27-Oct-00 Industrial 3179 SCV039 25-Sep-01 Industrial 3125 SMC021 20-Sep-01 Industrial 3018 SMC022 20-Sep-01 Industrial 2709 CCC006 27-Oct-00 Industrial 2705 SMC038 04-Oct-01 Res./Com. 2525 SCV044 25-Sep-01 Industrial 2443 SMC025 20-Sep-01 Industrial 2376 SMC002 24-Oct-00 Industrial 2372 SMC009 25-Oct-00 Mixed 2110 Total Mercury (mg/Kg fines) MCS007 25-Jun-01 Res./Com. 39.6 SMC025 20-Sep-01 Industrial 28.8 SCV028 05-Oct-01 Res./Com. 26.1 SCV012 30-Oct-00 Res./Com. 19.9 MCS011 25-Jun-01 Mixed 14.0 SMC030 04-Oct-01 Res./Com. 11.8 CCC006 27-Oct-00 Industrial 11.8 SMC020 20-Sep-01 Industrial 11.3 SCV006 27-Oct-00 Industrial 9.8 MCS010 25-Jun-01 Industrial 9.3 MCS002 30-Oct-00 Mixed 8.6 SCV034 05-Oct-01 Industrial 6.5 SCV031 10-Sep-01 Industrial 6.2 CCC032 01-Oct-01 Industrial 6.1 MCS008 25-Jun-01 Industrial 5.9 CCC007 27-Oct-00 Industrial 5.2 ALA010 28-Sep-00 Mixed 4.6 CCC019 25-Oct-00 Res./Com. 4.2 SCV019 26-Oct-00 Mixed 3.9 SCV024 31-Oct-00 Mixed 3.8 SMC023 25-Sep-01 Industrial 3.6

1. Shaded lines indicate samples from the 2000 survey

64 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Table 25. List of sites identified with the top 15th percentile concentrations of total chlordane and total DDT from sampling conducted in 2001. All data are normalized to the fine fraction.

Concentration in Fine Site Area Property Fraction (ug/Kg fines) Total Chlordane SCV034 Industrial 24296 MCS008 Industrial 15396 SMC020 Industrial 11971 VFC006 Res/Com 3744 SCV028 Res/Com 1680 SMC024 Res/Com 1633 SMC016 Industrial 1074 SMC023 Industrial 902 MCS007 Res/Com 898 SMC031 Res/Com 857 SCV038 Industrial 627 SMC025 Industrial 616 Total DDT SMC020 Industrial 24541 SMC015 Industrial 2973 MCS008 Industrial 2054 SCV029 Industrial 1453 SMC023 Industrial 1410 SCV028 Res/Com 1307 CCC031 Industrial 916 SCV027 Industrial 813 SMC009 Mixed 734 SMC025 Industrial 607 MCS010 Industrial 604 SCV038 Industrial 456 SCV026 Mixed 450

65 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Total PCBs - Normalized to Fine Fraction

80000

70000

60000

50000

40000

30000

20000 Total PCBs/Fine Fraction (ppb) Fraction PCBs/Fine Total

10000

0 1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 103 109 115 121 127 133 139 145 151 157 163

Total Mercury - Normalized to Fine Fraction

Total Hg/fine fraction (ppm) 10

0 1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 103 109 115 121 127 133 139 145 151 157 163

Figure 21. Ranked distribution of total PCBs and mercury. (All data normalized to the fine fraction. Concentration of 124,198 ug/Kg fines of PCBs at SMC020 is offscale)

66 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Total Chlordane - Normalized to Fines

10000 SCV034 = 24,296 ug/Kg MCS008 = 15,396 ug/Kg 9000 SMC020 = 11,971 ug/Kg 8000

7000

6000

5000

4000

3000 Chlordane (ug/Kg fines)

2000

1000

0 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73

Total DDT - Normalized to Fines

4000 SMC020 = 24,541 ug/Kg 3500

3000

2500

2000

1500 DDT (ug/Kg fines)

1000

500

0 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73

Figure 22. Ranked distribution of total chlordane and DDT. (All data normalized to the fine fraction. Concentrations of chlordane at SCV034 (24,296 ug/Kg fines), MCS008 (15,396 ug/Kg fines) and SMC020 (11,971 ug/Kg fines) and a DDT concentration of 24,541 ug/Kg fines at SMC020 are all offscale)

67 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

3.5 PCB Homologs

PCBs in the United States were primarily manufactured by the Monsanto Corporation under the trade name of Arochlors. These commercial preparations consisted of complex mixtures of PCB congeners. A number of different grades were manufactured with names like Arochlor 1016, 1242,1248, 1254 and 1260. In the case of Arochlors designated by 12xx, the last two digits indicated the percentage of chlorine by weight in the mixture. Arochlor 1016 was a unique formulation that consisted mostly of tri- and tetrabiphenyls with pentachlorobiphenyls and above removed. Arochlor 1016 was intended to replace Arochlor 1242 and contained 41% chlorine by weight (SWRCB 1983).

The degree of chlorination was determined roughly by the amount of time that biphenyls were exposed to anhydrous chlorine. Due to the manufacturing process, the specific composition of different batches was always slightly different.

Figure 23 illustrates differences in the composition of several Arochlor formulations as analyzed by the Agency for Toxic Substances and Disease Registry (ATSDR) (1997) and Frame et al. (1996). These data have been modified to reflect percentages of each homolog that would be expected based on the 54 congeners analyzed as part of this project. These illustrations demonstrate the substantial variation that can result from different batches of Arochlors. It also illustrates the general trend of increasing abundance of the penta-, hexa-, and heptachlorobiphenyls in the mixtures as percent chlorine by weight increases from 42 to 60 percent. Hepta- and octachlorobiphenyls are only present in substantial proportions in Arochlor 1260, the most highly chlorinated PCB mixture in this comparison.

In the natural environment, PCBs often do not appear to match specific Arochlor formulations. Typical reasons for this are either the effects of weathering processes or the possibility of multiple sources of different formulations. The lighter, less chlorinated PCB homologs typically decay faster in the weathering process (SWRCB 1983) due to being more volatile and soluble. Different ambient conditions, however, can have a major influence on the weathering process. Dechlorination occurs primarily in association with anaerobic conditions. Under aerobic conditions volatilization of lower weight congeners may be more significant. The dechlorination process of weathering also results in an accumulation of ortho-chlorinated congeners and losses of meta- and para-chlorinated congeners. Although there are some exceptions, degradation rates are typically inverse to the degree of chlorination. Homologs with at least four chlorine atoms (tetrachlorobiphenyls) degrade extremely slowly. Thus weathered samples of a specific Arochlor may tend to contain higher percentages of the more chlorinated congeners or homologs than the original sample if volatilization is the primary degradation process. The anaerobic dechlorination process generally results in lowering proportions of highly chlorinated congeners while increasing relative composition of less chlorinated congeners.

The relative percentage of PCB homologs in typical commercial formulations was compared against the composition of PCB homologs measured in storm drain sediments from all sites with total PCB concentrations greater than 100 ug/Kg fines (Figure 24 and 25). Comparison of the environmental samples of PCB concentrations provides some insight as to the age of the source as well as useful information for tracking the source when unique patterns are encountered. Nevertheless, interpretation can be confounded by lack of knowledge as to the original characteristics of the PCB source or possible multiple sources with different initial formulations. Figure 24 illustrates the relative composition of PCBs in samples from 16 sites with bulk sediment concentrations greater than 300 ug/Kg fines. Samples with concentrations in this range provide information that is more useful since relative proportions of each homolog are not highly biased by relatively small differences or analytical variability. Figure 25 illustrates the relative composition of homologs for all sites with total PCB concentrations between 100 and 300 ug/Kg fines. As noted last year, PCB mixtures at most sites were typically similar to Arochlor

68 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

1254 with enhanced abundance of homologs with 7 or more chlorine atoms. Some significant exceptions exist to this generalization.

Three sites, SMC023, CCC032 and SCV032, had unusually high proportions of tri- and tetrachlorobiphenyls. SMC023 is the Pulgas Creek Pump Station located in San Carlos in San Mateo County. CCC032 was a set of four catch basins sampled on West Cutting Boulevard near 1st Street in Richmond. SCV032 was a composite of sediment from three catchbasins located at 1850, 1854 and 1760 7th Street in San Jose. All areas were classified as industrial catchments. The high proportions of lower molecular weight, less chlorinated congeners combined with relatively high proportions of higher molecular weight, more chlorinated congeners is a relatively unique characteristic. PCBs in sediments from these sites either are from two or more Arochlor formulations or show evidence of a successive dechlorination due to weathering. Since two of these sites are from relatively small areas (an estimated 3- 4 acres), it is unlikely that the PCBs in these samples are from multiple sources.

Samples collected from a residential catchment in Vallejo (VFC006) were unique in having disproportionate quantities of heavier, more chlorinated (heptachlorobiphenyl and greater) congeners. Sediments collected from this site were covered by water.

Overall, examination of the composition of various PCB homologs has not provided highly useful information. Studies that have examined weathering of PCBs in sediment have primarily been conducted in sediment profiles from receiving water environments. In storm drains, it is impossible to know how long sediments have been exposed to different environmental conditions. In the best case, it can be assumed that the conditions existing at the time of sample collection were relatively stable for the past six to seven months. Since long-term environmental conditions impact the weathering process, the lack of knowledge of this history confounds interpretation of the data. Nevertheless, in cases where high concentrations of PCBs are measured and the homolog signature is unique, these data may still provide some assistance in tracing the source. This would be most effective in small catchments where distances from sources are relatively small.

Homolog Composition of Standard Aroclor Formulations

100%

90%

80% Deca 70% Nona Octa 60% Hepta 50% Hexa Penta 40% Tetra Tri Relative Composition 30% Di

20%

10%

0% 1016(a) 1016 (b) 1242 (a) 1242 (b) 1248 (a) 1248 (b) 1254 (a) 1254 (b) 1254 (b) 1260 (a) 1260 (b) Arochlors

Figure 23. Distribution of PCB homologs measured in standard formulations of Arochlors.

Data Source: Modified from U.S. EPA Region 5, Toxic Reduction Team, PCB Congener Database. Relative composition of homologs within each Arochlor was based only on congeners measured in conjunction with the current study. Original data was summarized by EPA Region 5 Team from (a) Agency for Toxic Substances and Disease Registry (ATSDR) 1997 and (b) Frame, G. M., Cochran, J.W., and Bosewadt, S.S. 1996.

Distribution of PCB Homologs

100%

90%

80% DECA 70% NONA OCTA 60% HEPTA 50% HEXA

40% PENTA TETRA 30% TRI 20% DI

10%

0% SMC SMC SMC CCC VFC SMC SCV CCC SCV SCV SCV SMC VFC VFC SCV SCV 020 024 023 032 006 021 033 031 029 038 044 029 001 010 035 039 DECA 033731910253000403 NONA 220 0 41 6 160 13 0 4 6 11 0 9 8 19 3 12 OCTA 1330 51 146 29 378 68 26 22 74 36 0 56 23 44 30 9 HEPTA 4740 466 416 73 402 186 60 105 328 109 40 107 109 91 146 23 HEXA 5260 4828 410 335 185 367 355 392 275 209 202 131 128 110 130 75 PENTA 4610 8419 286 458 95 391 630 401 72 195 219 94 91 75 28 165 TETRA 3530 2810 506 578 20 180 141 197 8 30 66 25 11 5 1 26 TRI 604 158 446 419 0 15 3 25 0 0 7 0 0 0 0 0 DI 045044000600000000 Concentrations (ug/Kg) of PCB Homologs

Figure 24. Concentrations of PCB homologs in sediments from all 2001 sampling sites with total PCB concentrations greater than 300 ug/Kg fines.

Distribution of PCB Homologs

100% 90% DECA 80% NONA OCTA 70% HEPTA 60% HEXA 50% PENTA TETRA 40% TRI 30% DI 20% 10% 0% SMC CCC SCV SMC CCC SCV CCC CCC CCC SMC SMC SMC SCV CCC CCC CCC SMC SMC CCC 025 030 032 022 023 028 026 020 025 025 017 031 034 039 002 042 011 026 041 FD DECA 0302000000103020004 NONA 0345203000002000314 OCTA 0015200014000203000020 HEPTA 24 12 104 59 56 14 43 26 42 30 16 11 49 13 16 5 34 14 23 HEXA 83 34 141 97 70 42 37 56 63 94 47 30 56 50 48 29 41 40 24 PENTA 154 63 26 47 33 50 38 47 29 93 56 86 0 60 48 80 31 36 37 TETRA 17 86 0 11 5 34 10 11 4 22 15 6 22 9 17 10 8 15 12 TRI 082120150553220008880 DI 191200000000000000000

Concentrations (ug/Kg) of PCB Homologs

Figure 25. Concentrations of PCB Homologs in sediments from all 2001 sampling sites with total PCB concentrations between 100 and 300 ug/Kg fines.

72 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

4.0 CONCLUSIONS AND RECOMMENDATIONS

4.1 Conclusions

The following conclusions are based upon analysis of data from the full two years of sediment testing conducted in San Francisco Bay area storm drains. In all cases, data analyses were based upon sediment contaminant concentrations normalized to the fraction of the sediment less than 62.5 microns (fine fraction). This general assumption was based upon evidence that suggests that all contaminants of concern tend to be associated with the fine fraction.

• Concentrations of total PCBs and total mercury, normalized to the fine fraction, are significantly different among the four major land use categories. In each case, sediments from sites representative of open land use contained significantly lower levels of both total PCBs and total mercury. No statistically significant differences were evident among samples collected in storm drains from areas characterized as residential/commercial and industrial land use. The highest concentrations of total PCBs were associated with residential/commercial and industrial land use areas.

• Preliminary statistical comparisons were conducted on the distributions of total chlordane and DDT compounds in Bay area storm drains. These tests indicate that concentrations of total chlordane in sediments from residential/commercial land use areas are significantly higher than all other land use classifications. No statistically significant differences were evident in the distribution of DDT compounds among the four land use classifications.

• Concentrations of both total PCBs and mercury are significantly higher in urban storm drains when compared to open (nonurban) land use.

• Preliminary comparisons of concentrations of total chlordane and DDT compounds in sediments from urban and nonurban storm drains suggest that concentrations of both are significantly higher in the urban environment. Currently, however, concentrations of these compounds in the nonurban environment are not considered to be adequately characterized.

• Planning level estimates of stormwater loads were developed for 17 watersheds that drain directly to San Francisco Bay. Urban sources within these watersheds were estimated to contribute 19 to 227 pounds of PCBs per year and 115 to 497 pounds of total mercury per year. Similar estimates were developed for total chlordane and DDT, however, results are considered preliminary due to limited spatial coverage. A total of 48 to 260 pounds of chlordane and 2.3 to 50 pounds of DDT were estimated based upon current data.

• An evaluation of the possible effects of remediation at locations with PCB concentrations in the top 15 percent of all urban sites was conducted to examine sensitivity to such measures. Clean up to an objective equivalent to the 75th percentile reduced estimated high-end loads by 24%. Use of the median urban concentration as a cleanup objective reduced high-end load estimates by approximately 50%.

• Median concentrations of total PCBs, total chlordanes, and DDT compounds measured in urban storm drain sediments are roughly two orders of magnitude greater than median concentrations measured in receiving water sediments by the San Francisco Bay Regional Monitoring Program (RMP).

73 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

• The median concentration of mercury in urban storm drains was generally comparable to levels measured in the Bay by the RMP. The median concentration of mercury in urban storm drains was 0.93 mg/Kg fines compared to 0.39 mg/Kg fines at all RMP sites during the 1999 and 2000 surveys.

4.2 Recommendations

The project work group met March 26, 20023 to discuss the results of the studies, including PCBs case studies performed by individual stormwater programs.4 Based on the first two years of work, the following recommendations were developed with the assistance of the work group:

• The participating agencies should redirect resources used the past two years for PCBs watershed characterization to case study work in areas with elevated PCBs. The objectives will include continuing to develop a better understanding of PCBs sources, developing and costing controls,5 and estimating the impact of such controls on reducing loads of urban runoff PCBs to San Francisco Bay. This information will help determine the feasibility and effectiveness of PCBs control options and inform future monitoring efforts. The overall goal is to assist Regional Board staff to prepare the urban runoff implementation portion of the PCBs TMDL for the Bay.

• Bay area stormwater agencies should continue working with Regional Board staff on the design and implementation of the PCBs case studies. The project work group should also continue to meet to coordinate PCBs case study work and share information.

• Bay area stormwater agencies should work with Regional Board staff and the BASMAA Monitoring Committee to determine future regional watershed monitoring needs, strategies and priorities for mercury, chlorinated pesticides and other pollutants of concern. The participating agencies may also wish to begin planning additional source investigations in areas with elevated levels of DDTs and chlordanes, depending on their monitoring priorities and available resources. Such future additional efforts should also be prioritized and coordinated through the BASMAA monitoring committee.

3 The meeting was attended by Jon Konnan (SMSTOPPP/SCVURPPP), Adam Olivieri (SCVURPPP), Liz Lewis (MCSTOPPP), Jack Betourne (VFCSD), Chris Sommers (CCCWP), Larry Bahr (FSSD), Kevin Cullen (FSSD), Arleen Feng (ACCWP), Fred Hetzel (RWQCB), James Downing (City of San Jose) and Marty Stevenson (KLI).

4 Individual stormwater programs will submit separate reports to the Regional Board documenting the results of PCBs case studies.

5 Preliminary information on PCBs controls is included in the Santa Clara Valley Urban Runoff Program’s March 1, 2002 Control Program for PCBs, Attachment 2 (Preliminary Comparison of Potential Options to Eliminate or Reduce Discharges of PCBs from Urban Runoff Conveyance Systems). This document was handed out at the March 6, 2002 BASMAA Monitoring Committee meeting.

74 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

5.0 REFERENCES

Association of Bay Area Governments (ABAG). 1996. Existing Land Use in 1995: Data for Bay Area Counties and Cities. ABAG, Oakland, CA, Publication Number. P96007EQK.

Agency for Toxic Substances and Disease Registry (ATSDR) 1997. Toxicological profiile for polychlorinated biphenyls (PCBs). Prepared by: Research Triangle Institute under Contract No. 205-93-0606. Prepared for: Agency for Toxic Substances and Disease Registry, Public Health Service, U.S. Department of Health and Human Services. September 1997. Original sources: Table 3-5 references Albro and Parker, 1979, and Albro et al., 1981.

Albro, P.W. and C. E. Parker, 1979. Comparison of the compositions of Aroclor 1242 and Aroclor 1016. J Chromatogr 169:161-166. 1979.

Albro P.W., J.T. Corbett, and J.L. Schroeder. 1981. Quantitative characterization of polychlorinated biphenyl mixtures (Aroclor 1248, 1254, and 1260) by gas chromatography using capillary columns. J Chromatogr 205:103-111.

Ballschmiter, K. and M. Zell. 1980. Analysis of polychlorinated biphenyls (PCB) by glass capillary gas chromatography. Fresenius Z. Anal. Chem. 302:20-31. 1980.

Bay Area Stormwater Management Agencies Association (BASMAA), 1996 San Francisco Bay Area Stormwater Runoff Monitoring Data Analysis 1988-1995. Prepared by URS Greiner Woodward- Clyde.

Bell, D., Gold, J. and Salop, P. 1999. Field Sampling Manual for the Regional Monitoring Program for Trace Substances. San Francisco Estuary Institute, Richmond, CA. Version 1 (January). URL: [http://www.sfei.org/rmp/documentation/fom/fom_html].

California Department of Fish and Game (CDF&G). 1999. The California Watershed Map (CALWATER 2.2)

Davis, J.A., L.J. McKee, J.E. Leatherbarrow, and T.H. Daum. 2000. Contaminant Loads from Stormwater to Coastal Waters in the San Francisco Bay Region, Comparison to other pathways and recommended approach for future evaluation. SFEI, September 2000.

EOA, Inc. 2001. Work Plan: Joint Stormwater Agency Project – Year Two, Investigation of Urban Sources of Mercury, PCBs and Organochlorine Pesticides. A collaborative project of the Santa Clara Valley Urban Runoff Pollution Prevention Program, Contra Costa Clean Water Program, San Mateo Countywide Stormwater Pollution Prevention Program, Marin County Stormwater Pollution Prevention Program, Vallejo Flood Control and Sanitation District, and Fairfield-Suisun Sewer District.

Frame G. M., J. W. Cochran, and S.S. Boewadt 1996. Complete PCB congener distributions for 17 Aroclor mixtures determined by 3 HRGC systems optimized for comprehensive, quantitative, congener-specific analysis. J. High Resol. Chromatogr., 19:657-668. December 1996. [Aroclor composition data were taken from a spreadsheet (aroclor_frame.xls) condensed by G.M. Frame from the research results reported in the above publication and posted on EPA Region 5 web site (http://www.epa.gov/toxteam/pcbid/down.htm).]

75 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

Gray, J.R., Glysson, G.D., Turcios, L.M., and Schwarz, G.E., 2000. Comparability of suspended-sediment concentration and total suspended solids data: U.S. Geological Survey Water-Resources Investigation Report 00-4191, 20 p

Gunther, A.J., P. Salop, D. Bell, A. Feng, J. Wiegel, and R. Wood. 2001. Initial Characterization of PCB, Mercury, and PAH Contamination in the Drainages of Western Alameda County. Report prepared for the Alameda Countywide Clean Water Program.

Helsel, Dennis R. and Cohn, Timothy A. 1988. Estimations of Descriptive Statistics for Multiply Censored Water Quality Data. Water Resources Research. Vol 24. No. 12, pp 1997-2004

Helsel, Dennis H. 1990. Less than obvious. Statistical treatment of data below the detection limit. Environmental Science and Technology. Vol. 24, No. 12, pp 1766-1774

Hunt, J.W., B. Anderson, B., B. Philips, J. Newman, R. Tjeerdema, K. Taberski, C. Wilson, M. Stephenson, H.M. Puckett, R. Fairey, and J. Oakden 1998. Sediment Quality and Biological Effects in San Francisco Bay. Bay Protection and Toxic Cleanup Program

Kinnetic Laboratories, Inc. 2000. Quality Assurance Project Plan: Joint Stormwater Agency Project to Study Urban Sources of Mercury and PCBs. A collaborative project of the Santa Clara Valley Urban Runoff Pollution Prevention Program, Contra Costa Clean Water Program, San Mateo Countywide Stormwater Pollution Prevention Program, Marin County Stormwater Pollution Prevention Program, Valley Flood Control and Sanitation District, and Fairfield-Suisun Sewer District.

Kinnetic Laboratories, Inc. 2001. Quality Assurance Project Plan: Joint Stormwater Agency Project to Study Urban Sources of Mercury, PCBs and Organochlorine Pesticides. A collaborative project of the Santa Clara Valley Urban Runoff Pollution Prevention Program, Contra Costa Clean Water Program, San Mateo Countywide Stormwater Pollution Prevention Program, Marin County Stormwater Pollution Prevention Program, Valley Flood Control and Sanitation District, and Fairfield-Suisun Sewer District.

Lowe, S., R. Hoenicke, and Jay Davis 1999. 1999 Quality Assurance Project Plan, Regional Monitoring Program for Trace Substances. RMP Contribution #33

San Francisco Estuary Institute. 2000. Regional Monitoring Program for Trace Substances, 1998 Annual Report.

Santa Clara Basin Watershed Management Initiative 2000. Volume 1, Watershed Characteristics Report.

Schueler, T. 1987. Controlling urban runoff: a practical manual for planning and designing urban BMPs. Metropolitan Washington Council of Governments. Washington, D.C.

Shelton, L. R. and Capel, P. D. 1994. Guidelines for Collecting and Processing Samples of Stream Bed Sediment for Analysis of Trace Elements and Organic Contaminants for the National Water- Quality Assessment Program. U.S. Geological Survey, Sacramento, CA. U.S.G.S. Open-File Report, 94-458. URL: [http://water.wr.usgs.gov/pnsp/pest.rep/bs-t.html].

Shumway, Robert H. and Azani, Rahman S. 2000. Statistical Approaches to Estimating Mean Water Quality Concentrations with Detection Limits. Report to Caltrans, October 20, 2000

76 Urban Sources of Mercury, PCBs and Organochlorine Pesticides April, 2002

SCCWRP 2000. Pollutant Mass Emissions to the Coastal Ocean of California: Initial Estimates and Recommendations to Improve Stormwater Emission Estimates, Appendix A Estimates of Mass Emissions to the Southern California Bight Region, November 10, 2000.

State Water Resources Control Board (SWRCB). 1983. Polychlorinated Biphenyls (PCBs). Special Projects Report No. 83-1sp.